On: 20 July 2015, At: 00:36 This article was ... - Fisheries · a Department of Wildlife, Fisheries and Aquaculture, Mississippi State University, Box 9690, Mississippi State, Mississippi

This article was downloaded by: [Department Of Fisheries]On: 20 July 2015, At: 00:36Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: 5 Howick Place,London, SW1P 1WG

Click for updates

North American Journal of AquaculturePublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/unaj20

Preharvest Feeding Strategy to Enhance Long-ChainPolyunsaturated and Polyunsaturated Fatty AcidComposition of the Tail Muscle of Freshwater PrawnsMacrobrachium rosenbergii Grown in Earthen PondsLouis R. D’Abramoa

a Department of Wildlife, Fisheries and Aquaculture, Mississippi State University, Box 9690,Mississippi State, Mississippi 39762, USAPublished online: 16 Oct 2014.

To cite this article: Louis R. D’Abramo (2015) Preharvest Feeding Strategy to Enhance Long-Chain Polyunsaturated andPolyunsaturated Fatty Acid Composition of the Tail Muscle of Freshwater Prawns Macrobrachium rosenbergii Grown in EarthenPonds, North American Journal of Aquaculture, 77:1, 1-7, DOI: 10.1080/15222055.2014.936539

To link to this article: http://dx.doi.org/10.1080/15222055.2014.936539

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

http://crossmark.crossref.org/dialog/?doi=10.1080/15222055.2014.936539&domain=pdf&date_stamp=2014-10-16

http://www.tandfonline.com/loi/unaj20

http://www.tandfonline.com/action/showCitFormats?doi=10.1080/15222055.2014.936539

http://dx.doi.org/10.1080/15222055.2014.936539

http://www.tandfonline.com/page/terms-and-conditions


North American Journal of Aquaculture 77:1–7, 2015C© American Fisheries Society 2015ISSN: 1522-2055 print / 1548-8454 onlineDOI: 10.1080/15222055.2014.936539

ARTICLE

Preharvest Feeding Strategy to Enhance Long-ChainPolyunsaturated and Polyunsaturated Fatty AcidComposition of the Tail Muscle of Freshwater PrawnsMacrobrachium rosenbergii Grown in Earthen Ponds

Louis R. D’Abramo*Department of Wildlife, Fisheries and Aquaculture, Mississippi State University, Box 9690, MississippiState, Mississippi 39762, USA

AbstractThree independent experiments were sequentially conducted during separate growing seasons (2009, 2010, and

2012) to determine whether a preharvest change in feed could achieve a desired enhancement of the long-chain(LC) polyunsaturated fatty acid (PUFA) content or the PUFA content of the tail muscle in freshwater prawnsMacrobrachium rosenbergii. Juvenile freshwater prawns were stocked into 0.05-ha earthen ponds, cultured for 111–122 d, and then harvested. For the first experiment, there were three treatments: exclusive feeding of a commercialprawn feed for 122 d after stocking; exclusive feeding of range cubes for 122 d after stocking; and administeringrange cubes for 84 d followed by prawn feed for the final 38 d prior to harvest. The second experiment consisted offour treatments: range cubes were fed exclusively for the entire growing season (116 d) or were fed until replacedby the prawn feed at 39, 24, or 17 d prior to harvest. The LC-PUFA profile for the tail muscle of harvested prawnswas equivalent between prawns given the prawn feed for the entire growing season and those given the range cubessubstituted by the prawn feed for the final 38 or 39 d before harvest. The proportional levels of LC-PUFA in the tailmuscle of prawns that received the prawn feed during the final 24 or 17 d prior to harvest were slightly lower. In thethird experiment (two treatments), range cubes were administered either throughout the growing season (112 d) orfor 84 d followed by the feeding of range cubes sprayed with flaxseed oil (2% weight/weight) for the final 28 d. Thepercentage (as total fatty acids) of linolenic acid (18:3[n-3]), the principal fatty acid in flaxseed oil, increased 7.8 timesin the feed and 3.3 times in the tail muscle of harvested freshwater prawns.

The highly fluctuating and ever-increasing costs of fish mealand fish oil ingredients included in formulations of aquaculturefeeds, combined with concerns about the limited availability andsustainability of these ingredients, have stimulated research ef-forts throughout the past two decades to find alternative feedstuffingredients that can successfully serve as complete or partialsubstitutes (NRC 2011). However, efforts directed at substitu-tion most often result in significantly lower and unacceptablegrowth, in part due to the presumed insufficient provision of es-sential polyunsaturated fatty acids (PUFAs) and long-chain (LC)PUFAs, particularly those of the n-3 (linolenic) family (e.g.,linolenic acid, 18:3[n-3]; docosahexaenoic acid, 22:6[n-3]; and

*E-mail: [email protected] March 20, 2014; accepted June 12, 2014

eicosapentaenoic acid, 20:5[n-3], where the number precedingthe colon is the number of carbon atoms, the number after thecolon is the number of double bonds, and the number after thehyphen indicates the position of the first double bond from themethyl end), that are contained in these dietary ingredients.

Low-input culture practices for the freshwater prawn Mac-robrachium rosenbergii over the past 10 years have evolvedin response to attempts to secure effective management of theunique biological characteristics of this crustacean species sothat optimal growth and mean harvest size, a reduction in therange of size at harvest, and noteworthy positive revenue can berealized (D’Abramo et al. 2009). Due to an inverse relationship

1

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:36

20

July

201

5

2 D’ABRAMO

between stocking density and growth, comparatively lowerstocking densities (24,700–29,640 prawns/ha) yield a higherindividual mean weight and reduced size variability at harvest(Daniels et al. 1995). Lower stocking densities therefore trans-late into a reduction in variable (operational) costs (primarilythe cost of juveniles for stocking and the cost of feed) and alarger harvested prawn that commands a higher selling price,which together contribute to optimization of net return. Lowerstocking densities correspond to a lower resident prawn biomassduring the growing season, and the contribution of natural foodin the pond that is proportionately available for consumptionper prawn therefore increases.

The greater availability of pond-derived nutrient sources tohelp satisfy the nutrient requirements of the freshwater prawnis a byproduct of the low-input production strategy. As a result,nutritionally incomplete and less-expensive “feeds” (primarilydue to lower levels of crude protein) can be used as nutrientsources during the entire growing season, as has been demon-strated by the successful use of range cubes (a commerciallyavailable livestock feed supplement) as an exclusive feed pro-vision (D’Abramo et al. 2010). Production achieved with rangecubes under low-input pond culture conditions was not sig-nificantly different from that achieved when a commerciallyavailable prawn feed containing fish meal and fish oil was fedat 100% or 60% of the range cube feeding rate (D’Abramoet al. 2010). The range cubes notably contain no fish meal orfish oil—sources of LC-PUFAs that are essential for optimalgrowth of freshwater prawns (D’Abramo and Sheen 1993). Inthe absence of LC-PUFAs in the range cubes, LC-PUFAs thatare derived from consumption of pond organisms (worms, in-sect larvae, etc.) are apparently sufficient to satisfy the essentialLC-PUFA requirements of this species.

The results of the freshwater prawn culture investigationsreported by D’Abramo et al. (2010) were the impetus for a se-ries of three follow-up experiments conducted during separatepond grow-out seasons (2009, 2010, and 2012). The presentexperiments were designed to evaluate different feeds and com-binations of feeds and feeding strategies to understand the mech-anisms for the provision of dietary LC-PUFAs and PUFAs, asreflected in the proportional levels of these fatty acids in theabdominal muscle of pond-cultured freshwater prawns.

METHODSProcedures common to all experiments.—All of the experi-

ments described were conducted in small (0.05–0.06-ha) earthenponds that contained vertically orientated substrate (orange plas-tic fencing) equivalent to 25% of the bottom surface area (cal-culation based on the surface of a single side, mesh included,length × height). Ponds were filled with well water approx-imately 1 month prior to the stocking of juvenile freshwaterprawns. Pre-stocking management practices included an initialorganic fertilization (cottonseed meal or soybean meal) at a rateof 224 kg/ha and an inorganic fertilization consisting of a combi-

TABLE 1. Feeding rates (percentage of total feed administered) used forfreshwater prawns during the growing season in each of the three grow-outexperiments.

Weeks 2009 2010 2012

1–5 13.8 13.8 13.86–10 31.0 34.5 34.511–15 34.2 37.7 37.716–18 21.0 14.0 14.0

nation of nitrogen and phosphorus applied as a liquid (13-38-0)at 4.68 L/ha (in 2009) or as a powder (12-48-8) at 3.4–5.6 kg/ha(in 2010 and 2012). Liquid agricultural lime was applied asneeded to increase water hardness. The hardness (35.5 mg/L)and alkalinity (77 mg/L) of the well water used to fill the exper-imental ponds were low, resulting in a weak buffering capacityand a proclivity for the occurrence of rare but rapid increases inpH to levels exceeding 9.5.

After the initial application, cottonseed meal or soybean mealwas then added at 16.8 kg/ha every other day within 1 weekof stocking. The total amount of the feed (range cubes and/orprawn feed) supplied for the duration of a growing season wasbased on a 3:1 ratio for the weight of range cubes fed : an-ticipated total weight of harvested freshwater prawns. Whenreduced feeding rates of prawn feed were part of the manage-ment strategy, the rates were calculated based on those used forrange cubes. The daily rates of feed addition (prawn feed and/orrange cubes) were based on proportions of the total amountto be fed during the growing season, as assigned to differenttime periods (weeks) during the growing season (Table 1), anddiffered slightly among experiments. Dissolved oxygen levelsmeasured by a YSI Model 550 or 550A were managed dailyto remain above 3 mg/L through multiple readings in the earlymorning, mid-morning, afternoon, and evening (as needed basedon the time of the growing season). A fixed in-pond aerator wasavailable to provide the equivalent of 9.88–12.35 hp/ha (1 hp =746 W) as needed. Any additional aeration that might be neededto maintain the minimum desired level of dissolved oxygen wasprovided by a tractor-driven power take-off paddlewheel. ThepH of the water in each pond was routinely monitored during themid-afternoon every 3–4 d, depending on weather conditions.

At the termination of each experiment, each pond was drainedand all freshwater prawns were harvested from the pond bottom;the prawns were then counted and weighed to determine sur-vival, mean harvest weight, and production. Due to the logisticsof labor (time) and the number of ponds in each experiment, theharvest of all ponds occurred over a range of days. Schedulingwas determined so that the mean harvest date was essentiallyequivalent among the replicate ponds for each treatment. Theperiod of time between initiation and completion of harvest ofall ponds that were part of an experiment did not exceed 7 dand generally was within 4 d. For each treatment, abdominalmuscle tissue was obtained from four freshwater prawns that

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:36

20

July

201

5

ENHANCEMENT OF PRAWN FATTY ACID COMPOSITION 3

were collected at harvest from each pond assigned to that treat-ment, and the samples were then combined for analysis. Thesetissue samples as well as samples of the different feeds used inthe experiments were then submitted to Eurofins Scientific, Inc.(Des Moines, Iowa), for determination of the composition ofconstituent fatty acids (expressed as a percentage of total fattyacids).

Experiment 1 (2009).—Juvenile freshwater prawns (meanweight = 0.22 g) were stocked into the experimental ponds ata rate of 27,170 prawns/ha from June 3 through June 8, 2009.There were three treatments, with five replicates per treatment.Two treatments consisted of the exclusive feeding of either rangecubes or a commercial prawn feed (Rangen, Buhl, Idaho) forthe entire 122-d growing season. The prawn feed contained LC-PUFAs that were derived specifically from menhaden fish oiland fish meal ingredients, and it was administered at 40% of thefeeding rate used for range cubes (Table 1). The reduced feed-ing rate was based on the assumption that the prawn feed wasformulated to be nutritionally complete and therefore would beeffective at the low stocking densities. In addition, the reductionwas based on a desire to reduce feeding costs, as the prawn feedwas twice the cost of the range cubes. The third treatment con-sisted of administering range cubes for the first 84 d (12 weeks)of the growing season, followed by a switch to the commercialprawn feed (at 50% of the feeding rate for range cubes) for thefinal 38 d prior to harvest. The increase from 40% to 50% of thefeeding rate for range cubes was based on the assumption thatas the end of the growing season approached, additional prawnfeed might be necessary to meet the needs of a comparativelyhigher resident biomass of freshwater prawns in the ponds.

Experiment 2 (2010).—As a follow-up to the results obtainedin experiment 1, experiment 2 was designed to evaluate whethershorter durations of time for administering the prawn feed priorto harvest could produce the same results as achieved by feedingthe prawn feed during the final 38 d prior to harvest. Juvenilefreshwater prawns (mean weight = 0.61 g) were stocked intothe experimental ponds at a rate of 24,700 prawns/ha on June11, 2010. There were four treatments, with three replicate pondsper treatment. Range cubes that were fed exclusively for 116 dserved as the control treatment. The three other experimentaltreatments consisted of a switch from range cubes to prawnfeed at 39 d (repeated for comparison with results obtained inexperiment 1), 24 d, or 17 d prior to harvest. As in experiment1, the feeding rate used with prawn feed was 50% of that usedwith range cubes.

Experiment 3 (2012).—Experiment 3 was based on the re-sults of experiment 2. Juvenile freshwater prawns were stockedinto the experimental ponds at 24,700 prawns/ha on June 11,2012. There were two treatments, with four replicate pondsper treatment. Again, feeding of range cubes during the entiregrowing season (112 d) served as the control. The experimentaltreatment consisted of administering range cubes for the first 84d (12 weeks) of the growing season, followed by the feedingof range cubes that were sprayed with organic, cold-pressed,

TABLE 2. Ranges and means ( ± SD) of survival, harvest weight, and yield offreshwater prawns that were fed range cubes for the entire growing season, prawnfeed for the entire growing season, or range cubes followed by a commercialprawn feed for 38 d prior to harvest (experiment 1; n = number of replicateponds for each treatment). Values in parentheses are for a replicate (one pond inthe range cubes + prawn feed treatment) that was not included in the calculationof means.

Survival Weight ProductionTreatment (%) (g) (kg/ha)

Range cubes(n = 5)

66.1–83.776.2 ± 6.6

36.7–48.043.2 ± 5.0

739–1,064910 ± 121

Prawn feed(n = 5)

57.1–82.269.4 ± 9.4

32.6–49.842.3 ± 6.4

714–1,031856 ± 162

Range cubes+ prawnfeed (at 50%feeding rate;n = 4)

71.1–79.575.9 ± 3.8(49.9)

37.9–54.143.2 ± 7.4(27.4)

768–1,153894 ± 180(376)

pure flaxseed oil (2% weight/weight; Dr. Adorable, Inc.) for thefinal 28 d (4 weeks) prior to harvest. The initial stocking weightof prawns was 0.39 g in the control treatment and 0.45 g inthe flaxseed oil treatment. The objective of the experiment wasto determine whether supplementation of linolenic acid, a pri-mary PUFA in flaxseed oil, to range cubes (essentially lackingin 18:3[n-3], 2.42%) would be reflected in enhanced levels ofthis PUFA in the feed and in the abdominal muscle tissue ofharvested prawns.

Statistical methods.—Using the general linear models pro-cedure (PROC GLM) of the Statistical Analysis System (SASInstitute 1999), ANOVA was used to determine whether therewere significant differences in mean survival, mean individualweight, and mean total weight harvested per hectare among thetreatments in experiment 1 and those in experiment 2. The re-sponses for the treatments in experiment 3 were not submittedto statistical analysis due to the different mean stocking weights.When survival in any harvested pond was less than 50%, dataderived from that pond were not included in calculation of themeans for the corresponding treatment. Feeding strategy was theindependent variable, and survival, growth, and production werethe dependent variables. Differences between means were evalu-ated by the least-significant-difference procedure in the Statisti-cal Analysis System. Percentage data were arcsine–square roottransformed prior to statistical analysis. All differences wereconsidered significant at P-values less than 0.05.

RESULTS

Experiment 1Ranges and means of freshwater prawn survival, harvest

weight, and total production for each treatment in experiment1 are presented in Table 2. One pond that was part of the treat-ment consisting of range cubes followed by the prawn finishing

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:36

20

July

201

5

4 D’ABRAMO

TABLE 3. Polyunsaturated fatty acid (PUFA) and long-chain PUFA composition (percentage of total fatty acids) of range cubes, prawn feed, and the abdominalmuscle tissue of freshwater prawns harvested from ponds in experiments 1 and 2. Prawns were fed range cubes for the entire growing season; prawn feed for theentire growing season; or range cubes followed by a commercial prawn feed for 38 d prior to harvest (experiment 1; 2009) or 39, 24, or 17 d prior to harvest(experiment 2; 2010). For days prior to harvest, the corresponding percentage of the growing season is presented in parentheses). Asterisks indicate values lessthan 0.10%.

Prawn muscle tissue

Entire growing season Prawn feed, days prior to harvest

Fatty Range Prawn Range cubes, Range cubes, Prawn feed,acid cubes feed 2009 2010 2009 38 (32.0%) 39 (33.6%) 24 (20.7%) 17 (14.7%)

18:2(n-6) 44.81 22.09 27.41 26.05 15.43 21.73 17.56 18.38 21.4218:3(n-3) 2.42 3.77 2.58 2.08 2.54 3.22 1.99 2.53 2.3320:4(n-6) * 1.31 5.14 6.06 3.73 4.07 4.97 6.46 5.3020:5(n-3) * 7.96 4.26 4.19 12.59 10.15 10.17 7.68 7.6822:5(n-3) * 1.55 0.59 0.82 0.68 0.91 1.11 0.86 0.5922:6(n-3) * 3.69 0.27 0.39 2.83 2.38 2.65 1.68 1.70

feed had poor survival. Data from this pond are reported butwere excluded from the calculations and from statistical analy-sis. The switch to the prawn finishing feed represented 32.0% ofthe total growing season. Mean survival, mean harvest weight,and mean production were not significantly different among thedifferent feeding strategy treatments. The results of fatty acidanalysis (expressed as a percentage of total fatty acids) revealedno detectable (<0.1%) n-3 LC-PUFA (20:5[n-3], 22:5[n-3], and22:6[n-3]) in the range cubes and a total of 13.20% n-3 LC-PUFA in the prawn feed (Table 3). The tail muscle tissue ofprawns harvested from ponds in each treatment contained LC-PUFA. Quantitatively, the combined level of n-3 LC-PUFA inprawns that received the commercial prawn diet either for theentire growing season or for the final 38 d was 16.10% and13.44%, respectively, and these levels were notably higher thanthe n-3 LC-PUFA level (5.12%) in prawns that were fed therange cubes for the entire growing season. The increase wasprimarily attributable to higher levels of 20:5(n-3).

Experiment 2Ranges and means of survival, harvest weight, and total pro-

duction for experiment 2 are presented in Table 4. Due to poorsurvival, data collected from two ponds (one pond that wasswitched from range cubes to prawn feed at 39 d prior to har-vest, and one pond that was switched from range cubes to prawnfeed at 24 d prior to harvest) were not included in the calcula-tions or in statistical analysis. Use of the prawn finishing feedfor the final 39 d represented 33.6% of the total growing sea-son; use of prawn feed for the final 24 d represented 20.7%of the growing season; and use of prawn feed for the final 17d represented 14.7% of the growing season. The means of sur-vival, harvest weight, and total production were not significantlydifferent among the treatments. The sum of the relative propor-tions of n-3 LC-PUFAs in the tail muscle of prawns that wereswitched from range cubes to the prawn diet at 39 d prior to

harvest (13.93%) was very similar to that determined for thesame treatment (38 d) in experiment 1 (13.44%). The summedrelative percentages of n-3 LC-PUFAs in the muscle tissue wereslightly lower when the change to the prawn diet occurred at 24d (10.2%) or 17 d (10.0%) before harvest.

Experiment 3In experiment 3, the use of range cubes with the sprayed-

on flaxseed oil represented 25.0% of the total growing season.Ranges and means of survival, harvest weight, and total pro-duction for experiment 3 are presented in Table 5. Due to poorsurvival, data collected from one replicate pond in the sprayed-on flaxseed oil treatment were not included in the calculation ofmeans. The initial stocking weight differed between treatments,thus precluding a statistical analysis to determine whether signif-icant differences existed between means for the two treatments.A higher stocking weight commonly yields a higher individualmean weight at harvest and consequently higher production.The successful provision of dietary linolenic acid via sprayed-on flaxseed oil is evident through an examination of the fattyacid profiles of the modified range cubes and the prawn abdom-inal muscle (Table 6). Linolenic acid (as a percentage of totaldietary fatty acids) increased from 2.47% in the standard rangecubes to 19.23% (7.8 times) in the range cubes supplementedwith flaxseed oil. This increase was reflected in the linolenicacid content of prawn abdominal muscle, which increased from1.64% in the control to 5.38% (3.3 times) in the flaxseed oiltreatment. Lower percentages of 18:2(n-6) (linoleic acid) in themuscle tissue principally corresponded to these increases.

DISCUSSIONAmong the three experiments, poor survival was encountered

in 4 of the 27 total ponds that were part of the experimental de-signs. Results from these four ponds were not included in thecalculation of means or in the statistical analysis based on the

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:36

20

July

201

5


TABLE 4. Ranges and means ( ± SD) of survival, harvest weight, and yield of freshwater prawns that were fed range cubes for the entire growing season orrange cubes followed by a commercial prawn feed for 39, 24, or 17 d prior to harvest (experiment 2; n = number of replicates for each treatment). Values inparentheses are for replicates (one pond in the 39-d treatment and one pond in the 24-d treatment) that were not included in the calculation of means.

Treatment Survival (%) Weight (g) Production (kg/ha)

Range cubes (control; n = 3) 84.8–90.2; 87.4 ± 2.7 46.0–54.1; 49.6 ± 4.1 1,026–1,136; 1,072 ± 57Range cubes + prawn feed (at 50%

feeding rate of range cubes)39 d (33.6% of the growing season; n = 2) 73.1–93.8; 83.5 ± 14.6

(22.2)49.5–55.0; 52.3 ± 3.9(82.4)

996–1,149; 1,072 ± 109(452)

24 d (20.7% of the growing season; n = 2) 91.6–97.1; 94.4 ± 3.9(45.3)

41.5–50.0; 45.8 ± 6.0(43.9)

1,029–1,133; 1,081 ± 74(493)

17 d (14.7% of the growing season; n = 3) 59.1–92.1; 77.9 ± 17.0 44.7–61.2; 51.6 ± 8.6 896–1,020; 970 ± 66

conclusion that poor survival was not the result of a specific feedor feed strategy treatment. Rather, despite robust and conscien-tious efforts to manage and maintain good water quality amongall ponds in each experiment, high pH was presumed to be thecause of the mortality. As previously stated, the well water thatwas used to fill the ponds had some unique chemical charac-teristics (low alkalinity in combination with low hardness) thatcontributed to a reduction in the pH-buffering capacity of thewater. This condition is the basis for the periodic occurrenceof sudden and lethal increases in pH (>9.5), especially duringconditions of high phytoplankton blooms and sunlight.

The semi-intensive culture of the freshwater prawn (i.e., ini-tial stocking densities of 24,000–27,000 prawns/ha) serves toeffectively manage the species’ biology so as to attain the de-sired harvest characteristics of large size and a narrow size range(Daniels et al. 1995). The lower resident prawn biomass asso-ciated with semi-intensive culture allows for more natural biotain the pond to be available and consumed per prawn during thegrowing season. The results of experiment 1 indicate that a feedformulation with no source of essential n-3 LC-PUFA yields amean harvest weight and mean production that are equivalent tothose achieved with a commercial prawn feed containing thesefatty acids as derived from fish meal and fish oil. Thus, through

TABLE 5. Ranges and means ( ± SD) of survival, harvest weight, and yieldof freshwater prawns that were fed range cubes during the entire growing season(112 d) or range cubes for 84 d followed by range cubes treated with sprayed-onflaxseed oil during the final 28 d (25.0%) of the growing season (experiment3; n = number of replicates for each treatment). Values in parentheses are fora replicate (one pond in the flaxseed oil treatment) that was not included in thecalculation of means.

Survival Weight ProductionTreatment (%) (g) (kg/ha)

Range cubes(control; n = 4)

75.2–95.788.6 ± 9.2

28.0–40.234.3 ± 5.6

522–927760 ± 47

Range cubes withsprayed-onflaxseed oil (n = 3)

83.9–92.389.4 ± 4.8(14.8)

39.4–47.142.0 ± 4.4(87.1)

897–980926 ± 47(318)

the consumption of natural biota, the freshwater prawns’ re-quirement for LC-PUFA and/or PUFA (much lower than thelevels required by marine shrimp) can be satisfied without fishmeal or fish oil ingredients, which are common sources of thesefatty acids but are not feedstuff components of the range cubes.During a study of freshwater prawns grown in earthen ponds,Tidwell et al. (1995) observed increases in the prawns’ con-sumption of benthic macroinvertebrates in response to the useof nutrient-deficient feeds, the absence of organic fertilization,or the lack of a pelleted feed.

Successful culture in the absence of fish meal and fish oil infeeds carries added significance, as it is a vital metric in deter-mining whether the farming of the specific fish or crustaceanspecies is sustainable. The pond biota are also a supplementalsource of dietary protein, thereby permitting lower protein lev-els in the feeds and a corresponding reduction in the total costof the feed.

TABLE 6. Polyunsaturated fatty acid (PUFA) and long-chain PUFA compo-sition (percentage of total fatty acids) of feed and abdominal muscle tissue offreshwater prawns that were fed range cubes during the entire growing season(control) or range cubes for 84 d followed by range cubes treated with sprayed-on flaxseed oil during the final 28 d (25.0%) of the growing season. Asterisksindicate values less than 0.10%.

Control Flaxseed oil treatment

Fatty Range Prawn Range cubes Prawnacid cubes tissue (with oil) tissue

16:0 20.40 20.84 15.41 20.7218:0 2.23 10.46 2.90 10.6916:1(n-9) 0.40 2.56 0.27 2.8018:1(n-9) 20.10 19.94 21.03 21.1018:2(n-6) 51.40 30.66 38.64 22.0218:3(n-3) 2.47 1.64 19.23 5.3820:4(n-6) * 4.25 * 5.3220:5(n-3) * 3.39 * 3.5022:6(n-3) * * * *

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:36

20

July

201

5

6 D’ABRAMO

FIGURE 1. Daily mean water temperatures in experimental ponds for the 40 d prior to harvest of freshwater prawns (experiment 2). Arrows indicate the switchfrom range cubes to commercial prawn feed in the treatment ponds at 39, 24, or 17 d prior to harvest.

The proportional increase in LC-PUFA realized in the ab-dominal muscle of freshwater prawns after they received thecommercial prawn feed as a finishing feed during the final third(∼6 weeks) of the growing season was equivalent to the in-crease achieved by administering the prawn feed exclusivelyfor the entire growing season. This equivalent level achievedduring the final 6 weeks of the growing season translates to afeed cost increase of $0.59 per kilogram, whereby harvestedprawns with an enhanced LC-PUFA content—a possible con-sumer preference—are produced.

Relative to the results obtained when the prawn feed was ad-ministered starting at 39 d prior to harvest, the total proportionalcontent of LC-PUFAs in prawns harvested from ponds some-what declined when the switch to prawn feed occurred at 24 or17 d prior to harvest. This observation could be the result of areduced rate of feed consumption corresponding to a decline inpond water temperatures toward the end of the growing season(Figure 1). For example, during the 39-, 24- and 17-d periods inwhich prawn feed was administered, the percentage of time forwhich midday pond water temperatures were recorded as beinggreater than 27◦C was 71.8, 54.2, and 35.3%, respectively.

Although mean harvest weight and production were notablyhigher for freshwater prawns that received range cubes withsprayed flaxseed oil as finishing feed, these results presumably

reflect the higher stocking weight of the prawns used in thistreatment. The sprayed-on oil may have improved the waterstability of the range cubes, which are normally confined toterrestrial use, thereby enhancing nutrient delivery. Supplementsof 18:3(n-3) in experimental semi-purified diets fed to juvenilefreshwater prawns produced proportional increases in this PUFAwithin the neutral and polar lipids of the tissue (D’Abramoand Sheen 1993). However, these increases were not associatedwith significant increases in weight relative to that of prawnsreceiving a non-supplemented control diet. Reigh and Stickney(1989) also observed a reduction in growth gained by juvenilefreshwater prawns when pure 18:3(n-3) was added at a level of1% to the diet.

The successful results achieved when flaxseed oil wassprayed onto the range cubes indicate that nutritional enhance-ment of an existing feed can accomplish the same desired effectas a complete change of feed. Incomplete nutrient sources, suchas range cubes, can themselves be nutritionally enhanced withLC-PUFA, PUFA, or both and then can be fed during a definedpre-harvest period to change the fatty acid profile of the prawns’abdominal muscle tissue. The tissue of wild freshwater prawnscontains particularly low proportions of 18:3(n-3) (Chanmugamet al. 1983). Modification of the PUFA composition, specifi-cally 18:3(n-3), of prawn abdominal muscle by applying a feed

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:36

20

July

201

5


management strategy in which range cubes are sprayed withflaxseed oil has important implications in being able to meet thedemands of health-conscious consumers. In the poultry industry,modification of the feed given to laying hens is already practicedto produce eggs that contain higher “omega-3” (linolenic acid)levels than conventional eggs (Scheideler 2003).

Use of a “finishing” feed containing a supplemented level ofa desired nutrient and that is fed during a specific time periodprior to harvest would seem to have application in the nutritionalfortification or enhancement of other farmed seafood products.Lipid-soluble nutrients as well as pigments could also be sup-plemented to nutritionally incomplete feeds via sprayed-on oil.Such efforts would presumably be cost effective if the productis marketed as “value-added,” whereby a higher selling price istypically commanded.

The present results indicate that for the farming of freshwa-ter prawns at a temperate latitude, a change to an LC-PUFA-or PUFA-containing feed for the final third (∼6 weeks) of thegrowing season fully achieves changes in the fatty acid compo-sition of prawn abdominal muscle that are equivalent to changesobtained by administering that feed for the entire 18 weeks. Atlower latitudes, where pond culture temperatures may be higherat harvest, the duration of time required for applying the finish-ing feed might be shorter because the rates (amounts) of foodconsumption would be correspondingly higher.

ACKNOWLEDGMENTSI thank the staff at the South Farm Aquaculture Unit for their

assistance in maintaining the protocols for experimental pondmanagement and for collecting prawn harvest data and prawnsfor fatty acid analysis. Special thanks to Mack Fondren for

preparing the flaxseed oil-sprayed diets and to Brian Davis (De-partment of Wildlife, Fisheries, and Aquaculture, MississippiState University) for assistance in statistical analysis.

REFERENCESChanmugam, P., J. Donovan, C. J. Wheeler, and D. H. Hwang. 1983. Differences

in the lipid composition of freshwater prawn (Macrobrachium rosenbergii)and marine shrimp. Journal of Food Science 42:1440–1441.

D’Abramo, L. R., T. R. Hanson, and C. L. Ohs. 2010. Pelleted sources ofnutrition and the effect of stocking size-graded juveniles in low-input farmingof the freshwater prawn Macrobrachium rosenbergii in earthen ponds. Journalof the World Aquaculture Society 41:841–857.

D’Abramo, L. R., and S.-S. Sheen. 1993. Polyunsaturated fatty acid nutrition injuvenile freshwater prawn Macrobrachium rosenbergii. Aquaculture 115:63–86.

D’Abramo, L. R., J. L. Silva, and M. O. Frinsko. 2009. Sustainable farming offreshwater prawns and the assurance of product quality. Mississippi Agricul-tural and Forestry Experiment Station Technical Bulletin 1188.

Daniels, W. H., L. R. D’Abramo, M .W. Fondren, and M. D. Durant. 1995.Effects of stocking density and feed on pond production characteristics andrevenue of harvested freshwater prawns Macrobrachium rosenbergii stockedas size-graded juveniles. Journal of the World Aquaculture Society 26:38–47.

NRC (National Research Council). 2011. Nutrient requirements of fish andshrimp. National Academies Press, Animal Nutrition Series, Washington,D.C.

Reigh, R. C., and R. R. Stickney. 1989. Effects of purified dietary fatty acids onthe fatty acid composition of freshwater shrimp, Macrobrachium rosenbergii.Aquaculture 77:157–174.

Scheideler, S. E. 2003. Flaxseed in poultry diets: meat and eggs. Pages 423–428in L. U. Thompson and S. C. Cunnane, editors. Flaxseed in human nutrition,2nd edition. American Oil Chemists’ Society Publishing, Urbana, Illinois.

Tidwell, J. H., C. D. Webster, J. D. Sedlaceck, P. A. Webster, W. L. Knight, S.J. Hill, L. R. D’Abramo, W. H. Daniels, M. J. Fuller, and J. L. Montanez.1995. Effects of complete and supplemental diets and organic pond fertil-ization on production of Macrobrachium rosenbergii and associated benthicmacroinvertebrate populations. Aquaculture 138:169–180.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:36

20

July

201

5


Click for updates


Evaluation of Hydrogenated Soybean Oil in Feeds forHybrid Striped Bass Fed in Conjunction with FinishingPeriods of Different DurationsKenson Kanczuzewskia & Jesse T. Trushenskiaa Center for Fisheries, Aquaculture, and Aquatic Sciences and Departments of Zoology andAnimal Science, Food and Nutrition, Southern Illinois University Carbondale, 1125 LincolnDrive, Room 173, Carbondale, Illinois 62901, USAPublished online: 23 Oct 2014.

To cite this article: Kenson Kanczuzewski & Jesse T. Trushenski (2015) Evaluation of Hydrogenated Soybean Oil in Feeds forHybrid Striped Bass Fed in Conjunction with Finishing Periods of Different Durations, North American Journal of Aquaculture,77:1, 8-17, DOI: 10.1080/15222055.2014.936540








http://dx.doi.org/10.1080/15222055.2014.936540




ARTICLE

Evaluation of Hydrogenated Soybean Oil in Feeds forHybrid Striped Bass Fed in Conjunction with FinishingPeriods of Different Durations

Kenson Kanczuzewski and Jesse T. Trushenski*Center for Fisheries, Aquaculture, and Aquatic Sciences and Departments of Zoology and AnimalScience, Food and Nutrition, Southern Illinois University Carbondale, 1125 Lincoln Drive, Room 173,Carbondale, Illinois 62901, USA

AbstractWe evaluated the production performance and tissue composition of hybrid striped bass, i.e., sunshine bass (female

White Bass Morone chrysops × male Striped Bass M. saxatilis (initial weight = 110.6 g) raised to a marketable size(final weight = 575 g) on grow-out feeds containing graded levels of fish oil and hydrogenated soybean oil (100% fishoil, 50% fish oil and 50% soybean oil, 25% fish oil and 75% soybean oil, or 100% soybean oil) in conjunction with fin-ishing periods of different durations (4, 8, or 12 weeks of feeding the 100% fish oil feed prior to harvest). Productionperformance varied significantly among the feeding regimens, but none of the experimental groups were signifi-cantly different from the 100% fish oil control group. However, performance tended to follow fish oil consumption,with regimens providing more fish oil during grow-out yielding marginally superior growth and growth efficiency.Fillet fatty acid profiles varied considerably among the regimens prior to finishing. Fillets of fish fed diets contain-ing increasing amounts of hydrogenated soybean oil contained more monounsaturated fatty acids (18:1[n-9]) and(n-6) and fatty acids (18:2[n-6]) and less long-chain polyunsaturated fatty acids and (n-3) fatty acids (20:5[n-3] and22:6[n-3]). Despite major differences in dietary levels of saturated fatty acids (SFAs), fillet levels of SFAs did notvary appreciably. Profile differences arising during grow out were reversed by finishing, to a greater or lesser extent,depending on the magnitude of prefinishing profile distortion and the duration of the finishing period. Utilization ofhydrogenated soybean oil merits further consideration and research, but our results suggest that this feedstuff hasvalue as a supplemental lipid source, if not a complete fish oil substitute, in feeds for hybrid striped bass.

The aquaculture industry is increasingly reliant upon indus-trially compounded aquafeeds (Tacon and Metian 2008), and theavailability of cost-effective feedstuffs has been identified as aserious constraint to continued growth within the aquaculturesector (FAO 2012). Fish oil is a high quality source of energy andlong-chain polyunsaturated fatty acids (LC-PUFAs; C20 or C22

fatty acids with three or more double bonds), but it is increas-ingly costly (FAO 2014). As a result, fish oil is commonly sparedwith less expensive plant-derived and animal-derived oils andfats in aquafeeds. Assuming essential fatty acid requirementsare met, fish fed reduced or fish-oil-free feeds generally per-form as well as those fed fish-oil-based feeds; however, sparing

*Corresponding author: [email protected] March 21, 2014; accepted June 13, 2014

LC-PUFA-rich fish oil with alternative lipids typically results ina marked decrease in levels of LC-PUFAs in the edible tissues(reviews by Rosenlund et al. 2011; Perez-Sanchez et al. 2013;Trushenski and Bowzer 2013).

Previous research has demonstrated that feeding diets con-taining primarily saturated fatty acids (SFAs; fatty acids with nodouble bonds) and monounsaturated fatty acids (MUFAs; fattyacids with one double bond) mitigates tissue fatty acid profilechange and the loss of LC-PUFAs associated with feeding re-duced fish oil feeds, particularly those rich in medium-chainpolyunsaturated fatty acids (MC-PUFAs; C18 fatty acids withtwo or more double bonds) (Laporte and Trushenski 2011;

8

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:40

20

July

201

5

SOYBEAN OIL IN HYBRID STRIPED BASS FEEDS 9

Trushenski et al. 2011a, 2011b, 2011d, 2013a, 2013b; Tur-chini et al. 2011a; Ramezani-Fard et al. 2012; Trushenski andKanczuzewski 2013; Woitel et al. 2014a, 2014b). Using SFA-rich and MUFA-rich lipids in lieu of MC-PUFA-rich lipids hasbeen suggested as one approach to minimizing LC-PUFA lossamong fish fed reduced or fish-oil-free feeds during grow out(Trushenski and Bowzer 2012). Additionally, finishing feedscan be used to compensate for any losses in fillet LC-PUFAs.These feeds, typically containing high levels of fish oil, areused to augment fillet levels of LC-PUFAs prior to harvest (Bellet al. 2003; Glencross et al. 2003; Izquierdo et al. 2005; Laneet al. 2006; Trushenski et al. 2008; Trushenski and Boesenberg2009; Trushenski et al. 2009; Thanuthong et al. 2011). Com-bining these two approaches has also proven promising. For ex-ample, complete or near-complete fillet profile restoration hasbeen achieved after finishing Cobia Rachycentron canadum andRainbow Trout Oncorhynchus mykiss raised on reduced fish oiland rendered animal-fat-based grow-out feeds containing highlevels of SFAs and MUFAs (Trushenski et al. 2011c; Gause andTrushenski 2013; Woitel et al. 2014b).

Soybean oil is routinely hydrogenated to increase its shelf-life and to alter its other physical properties. During the hydro-genation process, unsaturated carbons in fatty acids are satu-rated, i.e., double bonds between carbon atoms are convertedto single bonds between carbon and hydrogen atoms. Conse-quently, hydrogenated soybean oil contains little of its originalMC-PUFA content and high levels of SFAs or MUFAs, depend-ing on the hydrogenation process that was applied (Table 1).Hydrogenated soybean oils have been used to spare fish oilin feeds for Rainbow Trout (Trushenski et al. 2011a), Cobia(Trushenski et al. 2013b; Woitel et al. 2014b), White SeabassAtractoscion nobilis (Trushenski et al. 2013a), Largemouth BassMicropterus salmoides (Laporte and Trushenski 2011), and sun-shine bass (female White Bass Morone chrysops × male StripedBass M. saxatilis; Trushenski and Kanczuzewski 2013). Theseshort-term trials suggest that SFA-rich, hydrogenated soybeanoil is effective as an energy source and is advantageous in limit-ing LC-PUFA loss from the tissues. However, this lipid has yetto be tested in the context of a longer-term trial including theapplication of finishing feeds prior to harvest. Accordingly, weevaluated the production performance and tissue compositionof sunshine bass raised on grow-out feeds containing gradedlevels of fish oil and hydrogenated soybean oil in conjunctionwith finishing periods of different durations.

METHODSFeed formulation, manufacturing, and analysis.—Feed for-

mulations were based on those used in a previous experimentscreening various soybean oils as alternatives to fish oil in sun-shine bass feeds (Trushenski and Kanczuzewski 2013; Table 2).Feeds were identical except for the source of supplemental lipid,containing menhaden fish oil (100% FO; control and finish-ing feed), hydrogenated soybean oil (100% SO), or blends ofthese lipids (50% SO, 75% SO). Ingredients were mixed with a

TABLE 1. Typical fatty acid composition (g/100 g fatty acid methyl esters)of menhaden fish oil, soybean oil, and hydrogenated soybean oil we tested onsunshine bass. Values represent least-square means of triplicate samples.

Menhaden Soybean HydrogenatedFatty acid fish oil oil soybean oil

14:0 8.7 0.1 0.616:0 18.9 11.1 21.218:0 3.4 4.4 76.4SFAsa 33.1 16.1 98.516:1(n-7) 11.4 0.1 0.018:1(n-7) 3.2 1.5 0.018:1(n-9) 5.9 21.8 0.8MUFAsb 21.8 23.5 0.816:2(n-4) 1.4 0.0 0.016:3(n-4) 1.2 0.0 0.318:2(n-6) 1.7 52.9 0.220:4(n-6) 0.9 0.0 0.0(n-6)c 3.4 52.9 0.218:3(n-3) 2.0 7.4 0.118:4(n-3) 4.4 0.1 0.020:4(n-3) 1.7 0.0 0.020:5(n-3) 12.9 0.0 0.022:5(n-3) 2.3 0.0 0.022:6(n-3) 15.3 0.0 0.0(n-3)d 38.8 7.5 0.1PUFAse 45.1 60.4 0.7LC-PUFAsf 33.5 0.0 0.0MC-PUFAsg 8.9 60.4 0.4(n-3):(n-6) 11.3 0.1 0.4

aSum of all saturated fatty acids, i.e. fatty acids without double bonds; includes 12:0,15:0, 17:0, 20:0, and 22:0 in addition to other individually reported fatty acids.

bSum of all monounsaturated fatty acids, i.e., fatty acids with a single double bond;includes 14:1, 15:1, 17:1, 20:1(n-9), 22:1(n-9), and 22:1(n-11) in addition to other individ-ually reported fatty acids.

cSum of all (n-6) fatty acids; includes 18:3(n-6), 20:2(n-6), and 20:3(n-6) in additionto other individually reported fatty acids.

dSum of all (n-3) fatty acids; includes 20:3(n-3) in addition to other individuallyreported fatty acids.

e Sum of all polyunsaturated fatty acids, i.e. fatty acids with ≥2 double bonds.fSum of all long-chain polyunsaturated fatty acids, i.e. fatty acids with ≥20 carbon

atoms and ≥3 double bonds; includes 20:3(n-3) in addition to other individually reportedfatty acids.

gSum of all fatty acids with 18 carbon atoms and ≥2 double bonds.

cutter-mixer (model CM450; Hobart Corporation, Troy, Ohio),pelleted in a food grinder (1.5 hp electric grinder, Cabela’s, Syd-ney, Nebraska) and dried at 38◦C with a commercial-grade fooddehydrator (Harvest Saver R-5A; Commerical Dehydrator Sys-tems, Eugene, Oregon) to approximately 952 g/kg dry matter.Feeds were stored at −20◦C for the duration of the feeding trial.Triplicate samples of feed were analyzed to confirm proximate(Table 2) and fatty acid composition (Table 3). Samples werelyophilized (Freezone 6; Labconco Corporation; Kansas City,Missouri) to determine moisture content and then pulverized.Protein (LECO FP-528; LECO Corporation, St Joseph, Michi-gan) and ash (gravimetric determination following incinerationin muffle furnace, 600◦C for 4 h) content were determined for

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:40

20

July

201

5

10 KANCZUZEWSKI AND TRUSHENSKI

TABLE 2. Formulation and proximate composition of sunshine bass feeds containing graded levels of fish oil (FO) and hydrogenated soybean oil (SO).

Ingredient 100% FO 50% SO 75% SO 100% SO

Proximate composition, dry matter g/kgMenhaden fish meala 200.0 200.0 200.0 200.0Soybean meal 300.0 300.0 300.0 300.0Menhaden fish oila 98.8 49.0 24.0 0.0Hydrogenated soybean oilb 0.0 49.9 74.0 98.0Wheat bran 202.0 202.0 202.0 202.0Corn gluten meal 140.0 140.0 140.0 140.0Carboxymethyl cellulose 20.0 20.0 20.0 20.0Sodium phosphate 15.0 15.0 15.0 15.0Dicalcium phosphate 15.0 15.0 15.0 15.0Vitamin premixc 1.2 1.2 1.2 1.2Mineral premixd 1.0 1.0 1.0 1.0Stay-Ce 2.0 2.0 2.0 2.0Choline chloride 6.0 6.0 6.0 6.0

Proximate composition, least-square mean ± SE (n = 3)Dry matter 947 ± 0 954 ± 0 957 ± 1 951 ± 2Protein 426 ± 1 422 ± 7 427 ± 1 434 ± 1Lipid 136 ± 3 139 ± 6 137 ± 5 130 ± 10Ash 102 ± 3 103 ± 3 106 ± 1 104 ± 3

aOmega Protein, Inc., Houston, Texas.bArcher Daniels Midland, Decatur, Illinois.cFormulated to contain 25.000% L-ascorbyl-2-polyphosphate, 14.000% RRR-alpha tocopheryl acetate, 13.160% vitamin K, 12.500% inositol, 12.500% nicotinic acid, 7.500%

riboflavin, 6.250% calcium pantothenate, 2.500% pyridoxine hydrochloride, 1.250% thiamine mononitrate, 1.000% vitamin A palmitate, 0.500% cyanocobalamin, 0.450% folic acid,0.125% biotin, and 0.010% cholecalciferol in a cellulose base.

dFormulated to contain 24.897% zinc oxide, 14.933% ferrous sulfate, 3.470% manganese oxide, 0.967% cupric carbonate, 0.262% potassium iodide, 0.060% sodium selenate, and0.030% cobalt carbonate in a cellulose base.

e35% ascorbyl monophosphate, Roche Vitamins, Inc., Parsippany, New Jersey.

each sample. Lipid content was determined gravimetrically fol-lowing chloroform-methanol extraction (Folch et al. 1957). Re-served crude lipid samples were subjected to acid-catalyzedtransmethylation performed overnight at 50◦C (Christie 1982).The resultant fatty acid methyl esters (FAMEs) were separatedusing a gas chromatograph equipped with a flame-ionizationdetector fitted with a permanently bonded polyethylene gly-col, fused-silica capillary column (Omegawax 250, 30 m ×0.25 mm inner diameter, 0.25 µm film; Supelco, Bellefonte,Pennsylvania). The injection volume was 1.0 µL, helium wasthe carrier gas (30 cm/s, 205◦C), and the injector temperaturewas 250◦C. A split injection technique (100:1) was used, andthe temperature program was as follows: 50◦C held for 2 min,increased to 220◦C at 4◦C/min, and held at 220◦C for 15 min.Individual FAMEs were identified by reference to external stan-dards (Supelco 37 Component FAME Mix, PUFA-1, and PUFA-3; Supelco, Bellefonte, Pennsylvania).

Experimental design and feeding trial.—The feeds were usedin combination to create a control group and nine experimentalfeeding regimens:

• 100% FO only: fish were fed the 100% FO feed for theduration of the 21-week experiment; this regimen wasdesignated as the control group

• 50% SO + 4 weeks, 50% SO + 8 weeks, 50% SO +12 weeks: fish were fed the 50% SO feed in combina-tion with 4, 8, or 12 weeks of finishing with the 100%FO feed

• 75% SO + 4 weeks, 75% SO + 8 weeks, 75% SO +12 weeks: fish were fed the 75% SO feed in combina-tion with 4, 8, or 12 weeks of finishing with the 100%FO feed

• 100% SO + 4 weeks, 100% SO + 8 weeks, 100%SO + 12 weeks: fish were fed the 100% SO feed incombination with 4, 8, or 12 weeks of finishing withthe 100% FO feed

A water recirculation system consisting of 40, 170-L tanksequipped with mechanical and biological filtration units anda supplemental aeration system was used for the trial. Eachtank was stocked with randomly distributed juvenile sunshinebass (5 fish/tank; mean = 110.6 g, SE = 0.4), and the afore-mentioned feeding regimens were randomly assigned to tanksin quadruplicate (N = 4). Temperature and dissolved oxygenwere monitored daily throughout the study using a YSI OxygenMeter (model 550, Yellow Springs, Ohio,) and averaged 23.9◦C(range = 18.0–29.7◦C) and 7.4 mg/L (range = 6.1–9.6 mg/L).Fish husbandry and data collection practices (see below) were

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:40

20

July

201

5


TABLE 3. Least-square means ( ± SE) of fatty acid composition (g/100 g fatty acid methyl esters) of feeds containing graded levels of fish oil (FO) andhydrogenated soybean oil (SO) fed to sunshine bass. Values of SE < 0.1 are reported as 0.0.

Fatty acid(s) 100% FO 50% SO 75% SO 100% SO

14:0 7.9 ± 0.1 4.4 ± 0.0 2.8 ± 0.0 1.3 ± 0.016:0 19.1 ± 0.1 17.1 ± 0.0 15.6 ± 0.0 15.0 ± 0.018:0 3.6 ± 0.0 36.5 ± 0.1 51.7 ± 0.1 65.0 ± 0.0SFAsa 31.0 ± 0.2 58.4 ± 0.0 70.6 ± 0.1 81.8 ± 0.016:1(n-7) 10.0 ± 0.1 5.6 ± 0.0 3.4 ± 0.0 1.5 ± 0.018:1(n-7) 3.0 ± 0.0 1.8 ± 0.0 1.2 ± 0.0 0.6 ± 0.018:1(n-9) 8.7 ± 0.0 5.8 ± 0.0 4.5 ± 0.0 3.3 ± 0.0MUFAsb 22.7 ± 0.0 13.6 ± 0.0 9.4 ± 0.0 5.5 ± 0.016:2(n-4) 1.2 ± 0.0 0.7 ± 0.0 0.4 ± 0.0 0.2 ± 0.016:3(n-4) 1.4 ± 0.0 0.7 ± 0.0 0.4 ± 0.0 0.2 ± 0.018:2(n-6) 9.1 ± 0.1 7.3 ± 0.1 7.3 ± 0.1 6.8 ± 0.120:4(n-6) 1.0 ± 0.0 0.6 ± 0.0 0.4 ± 0.0 0.2 ± 0.0(n-6)c 10.3 ± 0.0 8.0 ± 0.1 7.6 ± 0.1 7.0 ± 0.118:3(n-3) 1.9 ± 0.0 1.2 ± 0.0 0.9 ± 0.0 0.7 ± 0.018:4(n-3) 2.9 ± 0.0 1.5 ± 0.0 0.9 ± 0.0 0.4 ± 0.020:4(n-3) 1.5 ± 0.0 0.8 ± 0.0 0.5 ± 0.0 0.2 ± 0.020:5(n-3) 13.7 ± 0.0 7.4 ± 0.0 4.4 ± 0.1 1.8 ± 0.022:5(n-3) 2.3 ± 0.0 1.2 ± 0.0 0.8 ± 0.0 0.3 ± 0.022:6(n-3) 11.2 ± 0.1 6.3 ± 0.0 4.0 ± 0.1 1.9 ± 0.0(n-3)d 33.3 ± 0.2 18.4 ± 0.0 11.6 ± 0.2 5.3 ± 0.1PUFAse 46.4 ± 0.2 28.0 ± 0.0 20.0 ± 0.1 12.7 ± 0.0LC-PUFAsf 29.8 ± 0.2 16.4 ± 0.0 10.1 ± 0.2 4.5 ± 0.1MC-PUFAsg 13.8 ± 0.1 10.1 ± 0.1 9.1 ± 0.1 7.9 ± 0.1(n-3):(n-6) 3.2 ± 0.0 2.3 ± 0.0 1.5 ± 0.0 0.8 ± 0.0

aSaturated fatty acids; sum of all fatty acids with no double bonds; includes 10:0, 12:0, and 20:0 in addition to individually reported SFAs.bMonounsaturated fatty acids; sum of all fatty acids with a single double bond; includes 20:1(n-9) in addition to individually reported MUFAs.cSum of all (n-6) fatty acids; includes 20:2(n-6) and 20:3(n-6) in addition to individually reported (n-6) fatty acids.dSum of all (n-3) fatty acids.ePolyunsaturated fatty acids; sum of all fatty acids with two or more double bonds; includes 20:2(n-6) and 20:3(n-6) in addition to individually reported PUFAs.fLong-chain polyunsaturated fatty acids; sum of all C20 and C22 fatty acids with three or more double bonds; includes 20:3(n-6) in addition to individually reported LC-PUFAs.gMedium-chain polyunsaturated fatty acids; sum of all C18 fatty acids with two or more double bonds.

conducted according to the standards of the Southern IllinoisUniversity Institutional Animal Care and Use Committee underAnimal Care and Use Protocol 08-027.

Data collection.—After approximately 9 weeks (60 d offeeding), one fish was randomly selected from each tank for de-termination of tissue composition prior to finishing (see below).Fish assigned to the +12-weeks regimens were then switchedto the 100% FO feed for the remaining 12 weeks of the trial;all other fish were fed previously assigned feeds until such timethe +8-weeks and +4-weeks regimens were also switched tothe 100% FO feed. After approximately 21 weeks of culture—60–116 d of grow out and 28–84 d of finishing, depending onfeeding regimen—all remaining fish were euthanized by tricainemethanesulfonate (MS-222) overdose (about 200 mg/L bath im-mersion until cessation of opercular movement) followed bysingle cranial pithing. Each fish was weighed individually anddissected to collect liver and intraperitoneal fat mass for deter-mination of hepatosomatic index (HSI) and liposomatic index(LSI). Samples of white muscle were collected, packaged in

sterile sample bags (Whirlpak, Nasco, Fort Atkinson, WI), andfrozen (−80◦C) prior to fatty acid analysis as described previ-ously for feed samples.

Standard production performance metrics were calculatedaccording to the following formulae, where BW is body weight:

Weight gain (%)

= 100 × average final individual BW − average initial individual BW

average individual initial weight,

Feed conversion ratio (FCR)

= average individual feed consumption (dry matter)

average individual weight gain,

Specific growth rate (SGR, % BW/d)

= 100 × loge (final BW) − loge(initial BW)

d of feeding,

Feed intake (% BW/d) = 100

× total dry matter intake/(initial individual BW × final individual BW)0.5

d of feeding,

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:40

20

July

201

5


TABLE 4. Production performance of sunshine bass fed diets containing different levels of fish oil (FO) and hydrogenated soybean oil (SO) in conjunctionwith finishing periods of different durations. Excluding the range of individual final weights reported, values represent least-square means; pooled standard error(PSE) and P-values are also reported. Means with common letter labels are not significantly different (P > 0.05). Abbreviations: SGR = specific growth rate,BW = body weight, FCR = food conversion ratio, HSI = hepatosomatic index, and LSI = Liposomatic index.

50% SO 75% SO 100% SO

100% + 4 + 8 + 12 + 4 + 8 + 12 + 4 + 8 + 12Parameter FO only weeks weeks weeks weeks weeks weeks weeks weeks weeks PSE P-value

Survival (%) 95 100 100 100 100 90 100 100 100 100 5 0.529Initial individual

weight (g)113 109 112 113 111 109 110 110 110 109 2 0.145

Final individualweight (g)

620 zy 602 zy 632 z 612 zy 607 zy 550 zy 540 zy 552 zy 535 zy 502 y 37 0.014

Range, individualweights (g)

494–860 435–754 433–777 373–803 429–774 418–688 387–721 411–735 429–811 389–642 NA NA

Weight gain (%) 450 452 466 440 445 403 391 404 385 362 23 0.047b

SGR (% BW/d) 1.2 1.2 1.2 1.2 1.1 1.1 1.1 1.1 1.1 1.1 0.0 0.057Feed consumption

(g/fish)a593 614 640 626 650 529 561 640 601 542 47 0.141

Feed intake(%BW/d)a

1.6 z 1.7 zy 1.7 zy 1.6 zy 1.7 zy 1.5 y 1.6 zy 1.8 zy 1.7 zy 1.6 zy 0.1 0.038

FCRa 1.17 x 1.25 yx 1.23 yx 1.26 yx 1.31 zyx 1.19 x 1.31 zyx 1.45 z 1.42 z 1.38 zy 0.04 < 0.001HSI 1.3 1.3 1.3 1.4 1.3 1.0 1.4 1.3 1.2 1.2 0.2 0.409LSI 2.4 2.5 2.7 2.7 2.6 2.1 2.5 2.4 2.5 2.2 0.3 0.596Fish oil

consumption(g/fish)

58 z 37 xwv 46 yx 49 zy 27 v 29 wv 40 yxw 15 u 27 v 34 wv 3 <0.001

aBased on feed dry matter.bAlthough the omnibus ANOVA test indicated a significant treatment effect, posthoc Tukey’s HSD pairwise comparison tests failed to identify differences among means.

Hepatosomatic index (HSI) = 100 × liver weight

BW, and

Liposomatic index (LSI) = 100 × total viscera weight

BW.

Coefficient of distance (Djh) values (Turchini et al. 2006), com-paring overall fatty acid profiles between the experimental reg-imens and the 100% FO only control group, were calculated asfollows:

Djh =[

n∑i=1

(Pij − Pih)2

]1/2

where Pij is the mean percent content of fatty acid i in thecontrol treatment (100% FO only) and Pih is the mean percentcontent of fatty acid i in an experimental regimen. Only majorindividual fatty acids (≥1% of total quantified FAME; no fattyacid groupings, e.g., SFAs, MUFAs) were used in the calculationof Djh.

Statistical analysis.—All production performance, filletlipid, and fatty acid data were analyzed using one-way ANOVAwith Tukey (honestly significant difference [HSD] pairwisecomparison tests for parameters exhibiting significant treatmenteffects). Additionally, production performance, fillet lipid, andpostfinishing–final fatty acid data from the experimental regi-mens (i.e., all regimens excluding the 100% FO only controlgroup) were subjected to two-way ANOVA to determine thesignificance of fish oil sparing (50% SO, 75% SO, or 100%

SO) and finishing duration ( +4 weeks, +8 weeks, or +12weeks) as main and interactive effects. When significant maineffects were observed, Tukey’s HSD tests were used to com-pare means within these main effects (i.e., to compare lev-els within factors). All procedures were completed using thePROC GLIMMIX function of SAS 9.2 (SAS Institute, Cary,North Carolina). In all cases, replicate tanks served as ex-perimental units (N = 4 for control group, N = 12 for ex-perimental regimens for prefinishing fatty acid data; N = 4for all regimens for all other data), and effects or differenceswere considered significant at α = 0.05. Coefficient of dis-tance values were not subjected to formal statistical analysisbecause of insufficient numbers of replicates (Djh is calculatedfrom fatty acid means; thus, a single value is calculated foreach regimen); values are presented for illustration purposesonly.

RESULTSOne-way ANOVA indicated that several production perfor-

mance parameters varied significantly among the feeding reg-imens (Table 4). Final individual weight was significantly dif-ferent between regimens with the largest (50% SO + 8 weeks;632 g) and smallest fish (100% SO + 12 weeks; 502 g), butnone of the experimental regimens were significantly differentfrom the 100% FO control group; moreover, when expressed interms of percent gain, growth did not vary significantly amongregimens. However, results of the two-way analysis of the

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:40

20

July

201

5


TABLE 5. P-values generated by two-way ANOVA tests of production performance of sunshine bass fed graded levels of fish oil (FO) and hydrogenated soybeanoil (SO). Descriptions of main effects are also porvided (see Table 4 for abbreviations).

Fish-oil-sparing Finishing duration Interaction Description of significantParameter P-value P-value P-value main effect(s)

Survival (%) 0.381 0.381 0.425 Not applicableInitial individual

weight (g)0.194 0.831 0.155 Not applicable

Final individualweight (g)

0.002 0.287 0.508 50% SO > 75% SO = 100% SO

Weight gain (%) 0.006 0.201 0.820 50% SOY > 100% SO, but 50% SO = 75%SO and 75% SO = 100% SO

SGR (% BW/d) 0.008 0.222 0.819 50% SOY > 100% SO, but 50% SO = 75%SO and 75% SO = 100% SO

Feed consumption(g/fish)

0.247 0.109 0.201 Not applicable

Feed intake (% BW/d) 0.129 0.050 0.193 75% SO < 100% SO, but 50% SO = 75%SO and 50% SO = 100% SO

FCR < 0.001 0.112 0.168 50% SO = 75% SO < 100% SOHSI 0.472 0.203 0.532 Not applicableLSI 0.277 0.824 0.431 Not applicableFish oil consumption

(g/fish)< 0.001 < 0.001 0.235 50% SO > 75% SO > 100% SO

+ 12 weeks > + 8 weeks > + 4 weeks

experimental regimens indicated that growth tended to followfish oil consumption, regimens receiving the most fish oil duringgrow-out (i.e., fish fed the 50% SO feed) yielding higher weightgain and SGR values (Table 5). Feed conversion ratios alsovaried significantly among feeding regimens. Specifically, FCRvalues (1.38–1.45) were significantly elevated among all threeregimens using the 100% SO feed compared with the 100%FO regimen (1.17; Table 4). Results of the two-way analysisalso supported this conclusion, indicating higher FCR valuesamong fish fed the 100% SO feed during grow-out than thosefed the 50% SO and 75% SO feeds (Table 5). Feed intake alsovaried significantly among regimens; however, none of the ex-perimental regimens were significantly different from the 100%FO control group (Table 4). The two-way analysis revealedsignificant effects of fish-oil-sparing on feed intake; however,the effect was unclear, intake being higher among fish fed the100% SO feed, but only for those fed the 75% SO feed dur-ing grow out (Table 5). No differences in HSI (1.0–1.4) or LSI(2.1–2.7) were revealed by one-way or two-way analysis, sug-gesting that although growth varied somewhat, the compositionof weight gained was equivalent among treatments (Tables 4,5). Expectedly, cumulative fish oil consumption varied amongthe regimens (Table 4), higher dietary inclusion rates and longerfinishing durations resulting in increased fish oil consumption(Table 5).

Fish-oil-sparing significantly affected prefinishing fillet fattyacid profile (Tables 6, 7). Compared to fillets of fish fed the100% FO feed, those of fish fed the hydrogenated soybean oil-

based grow-out diets contained higher levels of MUFAs (about31–36% versus 28% FAMEs) and (n-6) fatty acids (about 10–11% versus 9% FAMEs) and lower levels of LC-PUFAs (about20–25% versus 29% FAMEs) and (n-3) fatty acids (about 20–26% versus 30% FAMEs). The degree of profile distortion wasa function of dietary hydrogenated soybean oil content, dietscontaining more soybean oil yielding fillets with more distortedprofiles and Djh values (Table 6).

Finishing effectively reversed the distortions associated withfeeding the hydrogenated soybean oil-based grow-out diets,with longer finishing periods resulting in more comprehen-sive fillet modification (Tables 7, 8). At the end of the trial,LC-PUFAs were significantly reduced in only the 75% SO + 4weeks (about 28% FAME) and 100% SO + 4 weeks (about 24%FAME) regimens compared with the 100% FO regimen (about32% FAME). Similarly, (n-3) fatty acids were reduced onlyin those regimens providing the least fish oil, i.e., 75% SO + 4weeks, 100% SO + 8 weeks, and 100% SO + 4 weeks regimens.Conversely, regimens providing the least amount of fish oil ex-hibited increased levels of MUFAs, MC-PUFAs, and (n-6) fattyacids. Coefficient of distance values indicated less fillet fattyacid profile distortion among fish fed greater amounts of fish oil,whether in the context of grow out or finishing (Tables 6, 7). Bothone-way and two-way analysis of fillet lipid content (6.3–8.5%,dry matter basis) indicated equivalency among all regimens(Tables 7, 8).

A few mortalities were observed but were the result of fishjumping from tanks.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:40

20

July

201

5


TABLE 6. Prefinishing fatty acid composition (g/100 g fatty acid methyl esters) and coefficient of distance (Djh) values for sunshine bass fillets. Values representleast-square means of major fatty acids (≥1% of total fatty acid methyl esters). Where pooled SE values varied among means because of unequal replicationamong treatment groups, ranges are reported; pooled SE <0.1 are reported as 0.0. Means with common letter labels are not significantly different (P > 0.05).

Fatty acid(s)a 100% FO 50% SO 75% SO 100% SO Pooled SE P-value

14:0 5.5 z 4.9 y 4.5 x 4.1 w 0.1–0.2 <0.00116:0 21.0 y 21.6 zy 21.7 zy 21.9 z 0.2–0.3 0.03518:0 3.6 y 4.3 z 4.4 z 4.6 z 0.2 <0.001SFAs 30.2 31.0 30.6 30.7 0.3–0.4 0.30916:1(n-7) 9.3 z 8.7 zy 8.2 y 7.4 x 0.2–0.3 <0.00118:1(n-7) 3.3 z 3.2 zy 3.1 y 2.8 x 0.1 <0.00118:1(n-9) 14.7 x 17.6 yx 19.7 y 24.4 z 0.9–1.2 <0.00120:1(n-9) 1.2 y 1.2 y 1.3 y 1.4 z 0.0–0.1 <0.001MUFAs 28.5 y 30.7 y 32.2 y 36.0 z 0.7–1.2 <0.00118:2(n-6) 7.1 w 8.7 x 9.5 y 10.1 z 0.2–0.3 <0.00120:4(n-6) 1.4 1.4 1.4 1.2 0.1 0.037b

(n-6) 8.7 x 10.2 y 11.0 z 11.5 z 0.2–0.3 <0.00118:3(n-3) 1.3 z 1.3 z 1.3 z 1.2 y 0.0 <0.00118:4(n-3) 1.5 z 1.2 y 1.0 x 0.8 w 0.0–0.1 <0.00120:4(n-3) 1.1 z 0.9 y 0.8 x 0.6 w 0.0 <0.00120:5(n-3) 11.1 z 9.5 y 8.7 y 7.1 x 0.3–0.5 <0.00122:5(n-3) 2.5 z 2.2 zy 2.1 y 1.8 x 0.1 <0.00122:6(n-3) 12.9 z 11.1 zy 10.4 y 8.7 x 0.6–0.9 <0.001(n-3) 30.5 z 26.3 y 24.3 y 20.2 x 1.0–1.4 <0.001PUFA 41.3 z 38.4 z 37.2 z 33.3 y 1.1–1.6 <0.001LC-PUFAs 29.3 z 25.4 zy 23.7 y 19.7 x 1.0–1.5 <0.001MC-PUFAs 10.0 x 11.2 y 11.7 zy 12.0 z 0.2–0.3 <0.001(n-3):(n-6) 3.5 z 2.6 y 2.2 x 1.8 w 0.1 <0.001Djh 0.0 4.3 6.8 12.1 NA NA

aAll fatty acid abbreviations are as described in Table 3.bAlthough the omnibus ANOVA test indicated a significant treatment effect, posthoc Tukey’s HSD pairwise comparison tests failed to identify differences among means.

TABLE 7. P-values generated by two-way ANOVA tests of sunshine bass fillet fatty acid composition (in terms of primary fatty acid groupings) and lipidcontent, and description of significant main effects.

Fish-oil-sparing Finishing duration Interaction Description of significantFatty acid(s)a P-value P-value P-value main effect(s)

SFAs 0.544 0.834 0.841 Not applicableMUFAs 0.031b <0.001 0.066 + 4 weeks > + 8 weeks = + 12 weeksPUFAs 0.008 <0.001 0.028 50% SO = 75% SO > 100% SO

+ 4 weeks < + 8 weeks = + 12 weeksMC-PUFAs <0.001 <0.001 0.246 50% SO = 75% SO < 100% SO

+ 4 weeks > + 8 weeks > + 12 weeksLC-PUFAs 0.001 <0.001 0.034 50% SO = 75% SO > 100% SO

+ 4 weeks < + 8 weeks = + 12 weeks(n-3) <0.001 <0.001 0.011 50% SO = 75% SO > 100% SO

+ 4weeks < + 8 weeks = + 12 weeks(n-6) <0.001 <0.001 0.010 50% SO < 75% SO < 100% SO

+ 4 weeks > + 8 weeks > + 12 weeksLipid (% dry matter) 0.124 0.913 0.660 Not applicable

aAll fatty acid abbreviations are as described in Table 3.bAlthough the omnibus ANOVA test indicated a significant treatment effect, posthoc Tukey’s HSD pairwise comparison tests failed to identify differences among means.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:40

20

July

201

5


TABLE 8. Post-finishing fatty acid composition (g/100 g fatty acid methyl esters), coefficient of distance (Djh) values, and lipid content for hybrid striped bassfillets. Values represent least-square means of major fatty acids (≥ 1% of total fatty acid methyl esters). Where pooled SE values varied among means becauseof unequal replication among treatment groups, ranges are reported; pooled SE <0.1 are reported as 0.0. Means with common letter labels are not significantlydifferent (P > 0.05).

50% SO 75% SO 100% SO

Fatty 100% + 4 + 8 + 12 + 4 + 8 + 12 + 4 + 8 + 12 Pooledacid(s)a FO only weeks weeks weeks weeks weeks weeks weeks weeks weeks SE P-value

14:0 5.2 z 4.8 zy 5.1 z 5.0 zy 4.3 yx 4.5 zyx 4.8 zy 4.0 x 4.5 zyx 4.8 zy 0.2 <0.00116:0 21.0 21.0 20.8 20.8 21.0 21.2 21.2 21.6 21.3 21.1 0.3 0.30618:0 3.5 3.8 3.6 3.6 4.0 3.9 3.8 4.1 3.9 3.8 0.2 0.166SFAs 29.8 29.6 29.5 29.5 29.4 29.7 29.9 29.7 29.9 29.8 0.3–0.4 0.93516:1(n-7) 9.2 z 8.4 zyx 8.9 zy 8.8 zy 7.7 yx 7.8 yx 8.6 zy 7.2 x 7.9 yx 8.4 zyx 0.4 <0.00118:1(n-7) 3.4 z 3.3 zyx 3.4 z 3.3 zyx 3.0 wv 3.1 yxw 3.3 zy 2.8 v 3.1 xw 3.2 zyxw 0.1 <0.00118:1(n-9) 12.7 x 15.4 yx 14.6 yx 13.3 x 17.0 y 14.3 yx 14.4 yx 20.8 z 16.3 y 14.2 yx 0.8–0.9 <0.00120:1(n-9) 1.0 y 1.2 zy 1.2 zy 1.1 y 1.2 zy 1.0 y 1.1 y 1.3 z 1.2 zy 1.1 y 0.1 0.001MUFAs 26.4 y 28.2 zy 28.1 y 26.4 y 28.9 zy 26.4 y 27 .4 y 32.2 z 28.5 zy 26.9 y 1.2–1.3 0.00118:2(n-6) 6.7 u 8.5 xw 7.9 wv 7.2 vu 9.2 y 8.2 w 7.4 vu 10.2 z 8.9 yx 7.8 wv 0.2 <0.00120:4(n-6) 1.4 zy 1.3 zy 1.3 zy 1.4 zy 1.3 zy 1.4 z 1.4 zy 1.1 y 1.3 zy 1.4 zy 0.1 0.018(n-6)c 8.3 u 9.9 xw 9.3 wv 8.8 vu 10.6 y 9.8 xw 8.9 vu 11.6 z 10.4 yx 9.4 wv 0.2 <0.00118:3(n-3) 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.2 1.3 1.3 0.0 0.37918:4(n-3) 1.5 z 1.4 zy 1.5 z 1.5 z 1.2 yx 1.3 zy 1.4 zy 1.0 x 1.2 yx 1.3 zy 0.1 <0.00120:4(n-3) 1.1 z 1.0 zyx 1.1 zy 1.1 zy 0.9 xw 1.0 zyxw 1.0 zyxw 0.8 v 0.9 w 1.0 yxw 0.0 <0.00120:5(n-3) 11.2 z 10.4 zy 10.7 zy 11.3 z 9.9 y 10.9 zy 10.7 zy 8.3 x 10.0 y 10.7 zy 0.3 <0.00122:5(n-3) 2.5 z 2.4 zy 2.4 zy 2.6 z 2.3 y 2.4 zy 2.5 zy 2.0 x 2.3 y 2.5 zy 0.1 <0.00122:6(n-3) 15.9 z 13.7 zy 14.0 zy 15.5 z 13.5 zy 15.4 z 15.1 z 11.5 y 13.7 zy 15.2 z 0.9 0.001(n-3) 33.5 z 30.2 zy 31.0 zy 33.2 z 29.1 y 32.3 zy 31.8 zy 24.7 x 29.4 y 31.9 zy 1.1–1.2 <0.001PUFAs 43.8 z 42.2 z 42.3 z 44.1 z 41.6 zy 44.0 z 42.7 z 38.1 y 41.7 zy 43.3 z 1.1–1.2 <0.001LC-PUFAs 32.3 z 29.0 zy 29.6 zy 32.0 zy 28.0 yx 31.4 zy 30.8 zy 23.9 x 28.4 zy 31.0 zy 1.2–1.3 <0.001MC-PUFAs 9.5 u 11.2 yxw 10.7 xwv 10.0 vu 11.7 zy 10.7 xwv 10.0 vu 12.5 z 11.4 yx 10.4 wv 0.3 <0.001(n-3):(n-6) 4.0 z 3.0 xw 3.3 yx 3.8 zy 2.7 w 3.3 yx 3.6 y 2.1 v 2.8 w 3.4 yx 0.1 <0.001Djh 0.0 4.1 3.0 1.0 6.0 2.8 2.2 10.6 5.2 2.3 NA NALipid (% dry matter) 7.5 8.2 8.5 7.7 7.6 6.3 7.2 6.6 6.8 7.5 1.2 0.787

aAll fatty acid abbreviations/acronyms are as described in Table 3.

DISCUSSIONIn our study, sunshine bass grew relatively well and effi-

ciently regardless of whether they were fed diets containing fishoil or hydrogenated soybean oil; fish in all regimens reached amarketable size (>1 lb [454 g]) by the end of the feeding trialand maintained a FCR of 1.5 or less. However, differences infinal weights and FCRs were observed, with superior productionperformance generally associated with less fish oil replacement.Growth suppression has been reported among sunshine bass fedreduced or fish oil-free feeds (Lewis and Kohler 2008; Trushen-ski 2009). In these cases, essential fatty acid deficiencies wereimplicated as a contributing factor. Sunshine bass reportedly re-quire 0.5–1.0% (n-3) LC-PUFAs (as 20:5[n-3] and/or 22:6[n-3])in the diet (NRC 2011). Depending on fish oil inclusion rates, ourdiets contained 0.45–3.15% (n-3) LC-PUFAs (weight percent ofFAMEs converted to grams fatty acid/kg feed derived via con-version factor of 0.93 g fatty acid/g fish lipid [Weihrauch et al.1977]), suggesting the 100% SO feed may have been marginallydeficient with respect to these nutrients. Alternatively, it is pos-sible that hydrogenated soybean oil is not as well-digested asother lipids, and production performance was affected by minorreductions in digestible energy content of the soybean-oil-basedfeeds. Fish may not utilize SFAs as well as other fatty acids

(Hua and Bureau 2009); however, this has not translated to ma-jor differences in performance in our research with SFA-richlipids in various taxa (Trushenski et al. 2011a, 2011c, 2013b;Laporte and Trushenski 2011; Crouse et al. 2013; Trushenskiand Kanczuzewski 2013; Woitel et al. 2014a, 2014b). If the100% SO feed was inadequate with respect to essential fattyacid or digestible energy content, the magnitude of these inade-quacies was likely minor, given the associated effects. However,it is possible that feeding this diet exclusively, i.e., without fin-ishing with the 100% FO feed, would lead to a greater degreeof growth suppression.

The influence of fatty acid intake on tissue fatty acid com-position is one of the most well-researched topics in fish nu-trition (Turchini et al. 2011b). Hundreds of experiments with awide range of taxa, including sunshine bass (Fair et al. 1993;Nematipour and Gatlin 1993; Wonnacott et al. 2004; Lewisand Kohler 2008; Trushenski 2009; Trushenski et al. 2008,2009, 2011c; Trushenski and Kanczuzewski 2013), have demon-strated that fish tissues mirror dietary fatty acid profile and re-spond to changes in fatty acid intake. The differences in tissuecomposition we observed during grow out (i.e., depletion offish oil-associated LC-PUFAs, and accumulation of fatty acidsfound in the alternative lipid) and finishing (i.e., restoration of

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:40

20

July

201

5


LC-PUFAs and depletion of alternative lipid-associated fattyacids) are generally consistent with the precepts of tissue fattyacid profile change in fishes (Turchini et al. 2011b). However,as noted previously, sparing fish oil with SFA-rich or MUFA-rich alternatives typically does not induce tissue compositionalchange to the extent observed with MC-PUFA-rich alternatives(Trushenski et al. 2008, 2011a, 2011b, 2011d, 2013a, 2013b;Turchini et al. 2011a; Ramezani-Fard et al. 2012; Trushenskiand Kanczuzewski 2013; Woitel et al. 2014a, 2014b). Our re-sults support this paradigm. Though the hydrogenated soybeanoil-based feeds contained very high levels of SFAs (about 58–82% FAMEs) relative to the 100% FO feed (∼31% FAMEs),these differences were not proportionally reflected in the filletprofiles before (about 30–31% FAMEs in all regimens) or afterfinishing (about 30% FAMEs in all regimens). Our observa-tions of LC-PUFA retention during grow out and augmentationduring finishing also support the SFA-related “omega-3-sparingeffect” (Turchini et al. 2011a). Dietary levels of LC-PUFAs weremarkedly reduced in the hydrogenated soybean oil-based feeds(about 4.5–16.4% FAMEs) compared with the 100% FO feed(about 30% FAMEs). However, tissue levels were only mod-erately reduced prior to finishing (about 20–25% versus 30%FAMEs) and were comparable among most of the regimensafter finishing (about 28–32% FAMEs in all regimens except100% SO + 4 weeks [24% FAMEs]). Although fish-oil-sparingwith hydrogenated soybean oil reduced dietary levels of MUFAs(about 6–14% versus 23% FAMEs), tissue levels were elevatedamong fish fed these diets prior to finishing (about 31–36%versus 28% FAMEs), though the effect was diminished by fin-ishing (26–32% versus 26% FAMEs). This inverse pattern ofreduced dietary levels but elevated tissue levels of MUFAs hasbeen observed in previous research with SFA-rich aquafeeds(Trushenski 2009; Trushenski and Kanczuzewski 2013) andsuggests some ingested SFAs may be desaturated prior to tissuedeposition (Trushenski 2009). Regardless, the degree of profiledistortion we observed is comparatively minor with that ob-served in sunshine bass (Trushenski and Kanczuzewski 2013)and other taxa fed MC-PUFA-rich fish oil alternatives (Gauseand Trushenski 2013; Woitel et al. 2014a). In our previous in-vestigation of soybean oils in sunshine bass feeds, we fed ju-venile fish comparable diets containing 50:50 blends of fish oiland traditional (MC-PUFA-rich), low alpha-linolenic acid (MC-PUFA-rich), partial hydrogenated soybean oil (MUFA-rich), orfully hydrogenated soybean oil (SFA-rich). In this previous ex-periment, fish exhibited fillet Djh values ranging from 3.9 to17.6, the fully hydrogenated soybean oil-based feed yieldingthe lowest value (Trushenski and Kanczuzewski 2013). Here,we found the fish fed the 50% SO feed exhibited a very similarvalue prior to finishing (4.3). It is noteworthy that the highest Djhvalue observed (10.6) was in the 100% SO + 4 weeks regimen,indicating long-term feeding of a hydrogenated soybean-oil-based, fish-oil-free feed yields Djh values that are still consid-erably lower than the value associated with short-term feedingof diets containing 50:50 blends of fish oil and MC-PUFA-

rich soybean oils (17.3–17.6; Trushenski and Kanczuzewski2013).

In conclusion, partially replacing fish oil with hydrogenatedsoybean oil does not appear to substantially affect productionperformance of sunshine bass; however, further research re-garding the use of this ingredient in the context of complete fishoil replacement is warranted. Finishing is a successful strategyfor correcting the effects of fish-oil-sparing and increasing fil-let LC-PUFA content prior to harvest, and the use of SFA-rich,hydrogenated soybean-oil-based grow-out feeds appears advan-tageous in this context. We have demonstrated that sunshine basswith a nutritionally beneficial fatty acid profile can be success-fully raised using significantly less dietary fish oil; however,the price of hydrogenated soybean oil (about US$2,376/metricton in December 2013; Stephanie Block, Archer Daniels Mid-land, personal communication) relative to fish oil (∼$1,600–2,400/metric ton during 2013; FAO 2014) will determine itsrelevance as an ingredient in sunshine bass feeds.

ACKNOWLEDGMENTSWe thank Curtis Crouse, Carlin Fenn, John Bowzer, Chris

Bowzer, Artur Rombenso, Andy Coursey, Frank Woitel, JustinRosenquist, Bonnie Mulligan, Matt Young, Patrick Blaufuss,and Brian Gause for their assistance with data collection. Wealso thank Archer Daniels Midland for the donation of the hy-drogenated soybean oil we used in this work. Finally, we thankthe Illinois Soybean Association for financial support of theresearch described herein under grant 09-ISA-35-409-3.

REFERENCESBell, J. G., F. McGhee, P. J. Campbell, and J. R. Sargent. 2003. Rapeseed oil

as an alternative to marine fish oil in diets of post-smolt Atlantic Salmon(Salmo salar): changes in flesh fatty acid composition and effectiveness ofsubsequent fish oil “wash out.” Aquaculture 218:515–528.

Christie, W. W. 1982. Lipid analysis, 2nd edition. Pergamon, Oxford, UK.Crouse, C. C., R. A. Kelley, J. T. Trushenski, and M. J. Lydy. 2013. Use of

alternative lipids and finishing feeds to improve nutritional value and foodsafety of hybrid striped bass. Aquaculture 408-409:58–69.

Fair, P. H., W. P. Williams, and T. I. J. Smith. 1993. Effect of dietary menhadenoil on growth and muscle fatty acid composition of hybrid Striped Bass,Morone chrysops × M. saxatilis. Aquaculture 116:171–189.

FAO (Food and Agriculture Organization of the United Nations). 2012. Thestate of the world fisheries and aquaculture. FAO, Rome.

FAO (Food and Agriculture Organization of the United Nations). 2014.International commodity prices. Available: http://www.fao.org/economic/est/statistical-data/est-cpd/en./ (March 2014).

Folch, J., M. Lees, and G. H. Sloane-Stanely. 1957. A simple method forthe isolation and purification of total lipids from animal tissues. Journal ofBiological Chemistry 276:497–507.

Gause, B. R., and J. T. Trushenski. 2013. Sparing fish oil with beef tallow infeeds for Rainbow Trout: effects of inclusion rates and finishing on productionperformance and tissue fatty acid composition. North American Journal ofAquaculture 75:495–511.

Glencross, B. D., W. E. Hawkins, and J. G. Curnow. 2003. Restoration of thefatty acid composition of Red Seabream (Pagrus auratus) using a fish oilfinishing diet after grow-out on plant oil based feeds. Aquaculture Nutrition9:409–418.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:40

20

July

201

5


Hua, K., and D. P. Bureau. 2009. Development of a model to estimate digestiblelipid content of salmonid fish feeds. Aquaculture 286:271–276.

Izquierdo, M. S., D. Montero, L. Robaina, M. J. Caballero, G. Rosenlund, andR. Gines. 2005. Alterations in fillet fatty acid profile and flesh quality inGilthead Seabream (Sparus aurata) fed vegetable oils for a long term period.Recovery of fatty acid profiles by fish oil feeding. Aquaculture 250:431–444.

Lane, R. L., J. T. Trushenski, and C. C. Kohler. 2006. Modification of filletcomposition and evidence of differential fatty acid turnover in sunshine bassMorone chrysops × M. saxatilis following change in dietary lipid source.Lipids 41:1029–1038.

Laporte, J., and J. Trushenski 2011. Growth performance and tissue fatty acidcomposition of Largemouth Bass fed diets containing fish oil or blends or fishoil and soy-derived lipids. North American Journal of Aquaculture 73:435–444.

Lewis, H. A., and C. C. Kohler. 2008. Minimizing fish oil and fish meal withplant-based alternatives in sunshine bass diets without negatively impact-ing growth and muscle fatty acid profile. Journal of the World AquacultureSociety 39:573–585.

Nematipour, G. R., and D. M. Gatlin. 1993. Effects of different kinds of dietarylipid on growth and fatty acid composition of juvenile sunshine bass, Moronechrysops ♀ × M. saxatilis ♂. Aquaculture 141:154.

NRC (National Research Council). 2011. Nutrient requirements of fish andshrimp. National Academies Press, Washington, D.C.

Perez-Sanchez, J., L. Benedito-Palos, and G. F. Ballester-Lozano. 2013. Di-etary lipid sources as a means of changing fatty acid composition in fish:implications for food fortification. Pages 41–54 in V. R. Preedy, R. Sri-rajaskanthan, and V. B. Patel, editors. Handbook of food fortification andhealth, from concepts to public health applications, volume 2. Humana Press,New York.

Ramezani-Fard, E., M. S. Kamarudin, S. A. Harmin, and C. R. Saad. 2012.Dietary saturated and omega-3 fatty acids affect growth and fatty acid profilesof Malaysian Mahseer. European Journal of Lipid Science and Technology114:185–193.

Rosenlund, G., G. Corraze, M. Izquierdo, and B. E. Torstensen. 2011. The effectsof fish oil replacement on nutritional and organoleptic qualities of farmed fish.Pages 487–522 in G. M. Turchini, W.-K. Ng, and D. R. Tocher, editors. Fishoil replacement and alternative lipid sources in aquaculture feeds. CRC Press,Boca Raton, Florida.

Tacon, A. G. J., and M. Metian. 2008. Global overview on the use of fishmeal and fish oil in industrially compounded aquafeeds: trends and futureprospects. Aquaculture 285:146–158.

Thanuthong, T., D. S. Francis, S. D. Senadheera, P. L. Jones, and G. M. Turchini.2011. Fish oil replacement in Rainbow Trout diets and total dietary PUFAcontent: I. Effects on feed efficiency, fat deposition and the efficiency of afinishing strategy. Aquaculture 320:82–90.

Trushenski, J. T. 2009. Saturated lipid sources in feeds for sunshine bass: al-terations in production performance and tissue fatty acid composition. NorthAmerican Journal of Aquaculture 71:363–373.

Trushenski, J. T., P. Blaufuss, B. Mulligan, and J. Laporte. 2011a. Growthperformance and tissue fatty acid composition of Rainbow Trout reared onfeeds containing fish oil or equal blends of fish oil and traditional or novelalternative lipids. North American Journal of Aquaculture 73:194–203.

Trushenski, J. T., and J. Boesenberg. 2009. Influence of dietary fish oil con-centration and finishing duration on beneficial fatty acid profile restora-tion in sunshine bass Morone chrysops × M. saxatilis. Aquaculture 296:277–283.

Trushenski, J. T., J. Boesenberg, and C. C. Kohler. 2009. Influence of grow-out feed fatty acid composition on finishing success in Nile Tilapia. NorthAmerican Journal of Aquaculture 71:242–251.

Trushenski, J. T., and J. C. Bowzer. 2012. Having your omega 3 fatty acids andeating them, too: strategies to ensure and improve the long-chain polyunsatu-rated fatty acid content of farm-raised fish. Pages 319–340 in F. De Meester,R. R. Watson, and S. Zibadi, editors. Sustainable long chain omega-3 fattyacids in cardiovascular and mental health. Springer, New York.

Trushenski, J. T., and J. C. Bowzer. 2013. Having your omega 3 fatty acids andeating them too: strategies to ensure and improve the long-chain polyunsatu-rated fatty acid content of farm-raised fish. Pages 319–340 in F. De Meester,R. R. Watson, and S. Zibadi, editors. Omega-6/3 fatty acids, functions, sus-tainability strategies and perspectives. Humana Press, New York.

Trushenski, J. T., B. Gause, and H. A. Lewis. 2011c. Selective fatty acidmetabolism, not sequence of dietary fish oil intake, prevails in fillet fattyacid profile change in sunshine bass. North American Journal of Aquaculture73:204–211.

Trushenski, J. T., and K. L. Kanczuzewski. 2013. Traditional and modifiedsoy oils as substitutes for fish oil in feeds for hybrid Striped Bass (Moronechrysops × M. saxatilis). North American Journal of Aquaculture 75:295–304.

Trushenski, J. T., H. A. Lewis, and C. C. Kohler. 2008. Fatty acid profile ofsunshine bass: I. Profile change is affected by the initial composition anddiffers among tissues. Lipids 43:629–641.

Trushenski, J., B. Mulligan, D. Jirsa, and M. Drawbridge. 2013b. Sparing fishoil with soybean oil in feeds for White Seabass: effects of inclusion rate andsoybean oil composition. North American Journal of Aquaculture 75:305–315.

Trushenski, J. T., J. Rosenquist, and B. Gause. 2011b. Growth performance,tissue fatty acid composition, and consumer appeal of Rainbow Trout rearedon feeds containing terrestrial derived rendered fats. North American Journalof Aquaculture 73:468–478.

Trushenski, J. T., M. Schwarz, H. Lewis, J. Laporte, B. Delbos, R. Takeuchi,and L. A. Sampaio. 2011d. Effect of replacing dietary fish oil with soybeanoil on production performance and fillet lipid and fatty acid composition ofjuvenile Cobia Rachycentron canadum. Aquaculture Nutrition 17:437–447.

Trushenski, J. T., M. Schwarz, F. Woitel, and F. Yamamoto. 2013a. Saturatedfatty acids limit the effects of replacing fish oil with soybean oil with orwithout phospholipid supplementation in feeds for juvenile Cobia. NorthAmerican Journal of Aquaculture 75:316–328.

Turchini, G. M., D. S. Francis, and S. S. De Silva. 2006. Modification of tissuefatty acid composition in Murray Cod (Maccullochella peelii peelii, Mitchell)resulting from a shift from vegetable oil diets to a fish oil diet. AquacultureResearch 37:570–585.

Turchini, G. M., D. S. Francis, S. P. S. D. Senadheera, T. Thanuthong, and S. S.De Silva. 2011a. Fish oil replacement with different vegetable oils in MurrayCod: evidence of an “omega-3 sparing effect” by other dietary fatty acids.Aquaculture 315:250–259.

Turchini, G. M., W.-K. Ng, and D. R. Tocher. 2011b. Fish oil replacementand alternative lipid sources in aquaculture feeds. CRC Press, Boca Raton,Florida.

Weihrauch, J. L., L. P. Posati, B. A. Anderson, and J. Exler. 1977. Lipid con-version factors for calculating fatty acid contents of foods. Journal of theAmerican Oil Chemists’ Society 54:36–40.

Woitel, F. R., J. T. Trushenski, M. H. Schwarz, and M. L. Jahncke. 2014a. Morejudicious use of fish oil in Cobia feeds: I. Assessing the relative merits ofalternative lipids. North American Journal of Aquaculture 76:222–231.

Woitel, F. R., J. T. Trushenski, M. H. Schwarz, and M. L. Jahncke. 2014b. Morejudicious use of fish oil in Cobia feeds: II. Effects of graded fish oil sparingand finishing. North American Journal of Aquaculture 76:232–241.

Wonnacott, E. J., R. L. Lane, and C. C. Kohler. 2004. Influence of dietaryreplacement of menhaden oil with canola oil on fatty acid composition ofsunshine bass. North American Journal of Aquaculture 66:243–250.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:40

20

July

201

5


Click for updates


Landlocked Fall Chinook Salmon Ovarian Fluid Turbidityand Egg SurvivalKristen H. Becketab, Michael E. Barnesa, Dan J. Durbenb & Timothy M. Parkera

a McNenny State Fish Hatchery, South Dakota Department of Game, Fish and Parks, 19619Trout Loop, Spearfish, South Dakota 57783, USAb Black Hills State University, 1200 University Street, Spearfish, South Dakota 57799, USAPublished online: 06 Nov 2014.

To cite this article: Kristen H. Becket, Michael E. Barnes, Dan J. Durben & Timothy M. Parker (2015) Landlocked FallChinook Salmon Ovarian Fluid Turbidity and Egg Survival, North American Journal of Aquaculture, 77:1, 18-21, DOI:10.1080/15222055.2014.951811








http://dx.doi.org/10.1080/15222055.2014.951811




TECHNICAL NOTE

Landlocked Fall Chinook Salmon Ovarian Fluid Turbidityand Egg Survival

Kristen H. BecketMcNenny State Fish Hatchery, South Dakota Department of Game, Fish and Parks, 19619 Trout Loop,Spearfish, South Dakota 57783, USA; and Black Hills State University, 1200 University Street, Spearfish,South Dakota 57799, USA

Michael E. Barnes*McNenny State Fish Hatchery, South Dakota Department of Game, Fish and Parks, 19619 Trout Loop,Spearfish, South Dakota 57783, USA

Dan J. DurbenBlack Hills State University, 1200 University Street, Spearfish, South Dakota 57799, USA

Timothy M. ParkerMcNenny State Fish Hatchery, South Dakota Department of Game, Fish and Parks, 19619 Trout Loop,Spearfish, South Dakota 57783, USA

AbstractWe assessed the turbidity measurements of ovarian fluid and

ovarian fluid mixed with lake water, distilled water, and well waterfor their potential to predict egg survival of landlocked fall ChinookSalmon Oncorhynchus tshawytscha from Lake Oahe, South Dakota.Turbidity of the ovarian fluid alone ranged from 23.1 to 101.7 NTU,and mean turbidity increased with the addition of any of the watersources. Egg survival to the eyed stage ranged from 0% to 61%,and was not significantly correlated with ovarian fluid turbidity orthe turbidities of any of the ovarian fluid and water combinations.

Eggs of landlocked fall Chinook Salmon Oncorhynchustshawytscha from Lake Oahe, South Dakota, typically exhibitpoor survival during hatchery incubation, and survival variesdramatically from female to female (Barnes et al. 2000a, 2003).Because of the highly variable egg survival and the relativelyhigh cost associated with incubating potentially nonviable eggsduring hatchery rearing, the ability to eliminate poor-qualityspawns immediately after removal from the female would leadto improvements in spawning efficiencies and cost reductions.

Subjective qualitative assessments of egg quality have beenused to decide which spawns should be included in hatchery

*Corresponding author: [email protected] March 5, 2014; accepted June 29, 2014

production of salmonids (ADFG 1983), but these assessmentscannot consistently identify nonviable spawns (Springate andBromage 1984; Lahnsteiner et al. 1999; Barnes et al. 2000b).Additionally, objective quantitative egg assessments have alsobeen used with mixed results (Satia et al. 1974; Fauvel et al.1993; Lahnsteiner et al. 1999, 2001; Barnes et al. 2000b,2003).

Ovarian fluid turbidity, particularly when ovarian fluid ismixed with water, has shown some promise in predicting futureegg survival (Wojtczak et al. 2004; Tabrizi et al. 2011). Using aqualitative assessment, Wojtczak et al. (2004) reported that theovarian fluid from poor quality eggs of Rainbow Trout O. mykissbecame turbid when mixed with hatchery water. Tabrizi et al.(2011) used total suspended solids (TSS) as an indirect measureof the turbidity of ovarian fluid mixed with water and observedsignificant correlations of TSS with egg survival in CaspianBrown Trout Salmo trutta caspius. Nero et al. (2013) quantita-tively measured ovarian fluid turbidity in Lake Oahe ChinookSalmon and determined that it could only provide some ba-sis for the retention or disposal of selected spawns if brokeneggs were present. However, there is no published informa-tion examining the possible usefulness of quantitative turbidity

18

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:41

20

July

201

5

TECHNICAL NOTE 19

TABLE 1. Sample number (N) and mean ( ± SE), minimum, and maximum values for length, postspawn weight, egg data, and turbidity measurements forspawning female, landlocked, fall Chinook Salmon.

Variable N Mean ± SE Median Minimum Maximum

Female length (mm TL) 26 671 ± 11 660 587 768Female weight (g) 26 2,644 ± 150 2,444 1,732 4,489Egg size (eggs/mL) 21 5.21 ± 0.21 4.81 3.79 7.35Total eggs 21 2,606 ± 233 2,708 431 4,760Percent survival 21 21.3 ± 4.0 16.7 0 61.1Turbidity (NTU)

Ovarian fluid 26 47.4 ± 3.8 44.1 23.1 101.7Ovarian fluid with distilled water 26 68.3 ± 50.0 5.6 1.9 1,294Ovarian fluid with lake water 26 161.5 ± 92.4 7.0 3.7 2,066Ovarian fluid with well water 26 73.3 ± 56.5 4.7 1.9 1,470

measurements of ovarian fluid mixed with water as a predictorof survival of eggs from landlocked fall Chinook Salmon, orwhether correlations are dependent on the water source used.Thus, the objective of this study was to determine whether theturbidity of ovarian fluid mixed with different water sources canbe used as a predictor of survival of landlocked fall ChinookSalmon eggs.

METHODSLandlocked fall Chinook Salmon from Lake Oahe were

spawned at Whitlocks Spawning Station, near Gettysburg, SouthDakota, on October 22, 2013. All fish were anesthetized in watersaturated with carbon dioxide. Milt was collected from males,pooled in a container, and kept on ice until used. If the femaleswere ripe, as indicated by the extrusion of a small number ofeggs with a small amount of manual pressure on the abdomen,the fish were stunned by percussion. Eggs were then expressedinto a mesh net using compressed oxygen, allowing the ovarianfluid to drain into a plastic container below the net. The ovarianfluid was poured from the plastic container into a 15-mL cen-trifuge tube. Four different 10-mL samples were prepared foreach female: 10 mL of pure ovarian fluid, and 1 mL of ovarianfluid mixed with 9 mL of either distilled water, surface waterfrom Lake Oahe (total hardness as CaCO3, 220 mg/L; alkalinityas CaCO3, 160 mg/L; pH, 8.2; total dissolved solids, 440 mg/L),or well water (total hardness as CaCO3, 360 mg/L; alkalinity asCaCO3, 210 mg/L; pH, 7.6; total dissolved solids, 390 mg/L)from McNenny State Fish Hatchery, Spearfish, South Dakota.The turbidity of each sample was measured using a LaMotte2020e turbidity meter (LaMotte Company, Chestertown, Mary-land).

After removal of the ovarian fluid from the eggs, spawningand egg transport continued as per Nero et al. (2013). Twenty-sixfemales were spawned and only two females showed any brokeneggs (fewer than five per spawn). No overripe eggs (Barnes et al.2003) were observed from any of the females. Each spawn wasplaced into an individual plastic bag and spawns were stored on

ice in a cooler for transportation. Lengths and weights of eachfemale were taken postspawn.

Upon arrival at McNenny Hatchery, the eggs were disinfectedin a 100-mg/L buffered free iodine solution for 10 min. Afterdisinfection, the eggs from each female were inventoried usingwater displacement (Piper et al. 1982) and then placed intoseparate vertical flow incubation trays (Marisource, Payallup,Washington). Because only 21 incubation trays were available,five randomly chosen spawns were not incubated separatelyfor the collection of egg survival data (including one of thespawns with some broken eggs). The eggs were incubated inwell water and treated daily with formalin as described by Neroet al. (2013). Removal of dead eggs occurred on incubationday 28 (eyed egg stage) and the remaining viable eyed eggswere re-inventoried by water displacement. Percent survival wascalculated by dividing the number of eyed eggs by the initialnumber of eggs, and multiplying this amount by 100. Datawere analyzed with regression and correlation analysis usingthe SPSS (9.0) statistical analysis program (SPSS, Chicago,Illinois) with significance predetermined at P < 0.05.

RESULTSTwenty-six female Chinook Salmon having a TL ranging

from 587 to 768 mm were spawned; the ovarian fluid turbidityranged from 23.1 to 101.7 NTU. Turbidity increased after theaddition of the various types of water to the ovarian fluid (Ta-ble 1). Eyed egg survival ranged from 0% to 61% and was notsignificantly correlated with ovarian fluid turbidity (Figure 1)nor with the turbidities of any of the ovarian fluid and watercombinations (Figure 2).

Significant correlations were observed among turbidity mea-surements of the ovarian fluid alone and the ovarian fluid mixedwith different water types (Table 2). Correlations between ovar-ian fluid turbidity and the ovarian fluid and water combinationsranged from r = 0.426–0.505, while high correlations (r =0.898 and higher) occurred among the turbidities of the ovarianfluid combined with either lake water, distilled water, or well

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:41

20

July

201

5

20 BECKET ET AL.

FIGURE 1. Survival (%) to the eyed stage of eggs from landlocked fall Chi-nook Salmon in relation to ovarian fluid turbidity.

water. The turbidity of ovarian fluid mixed with either lake, dis-tilled, or well water was significantly and positively correlatedwith postspawn female weight and negatively correlated withthe total number of eggs per spawning female. Female lengthand weight were also correlated.

DISCUSSIONThe results indicate that turbidity of ovarian fluid or ovarian

fluid mixed with distilled water, lake water, or well water is nota good predictor of future egg survival, at least for landlockedChinook Salmon spawns within the ranges of turbidity observedin this study. These results are similar to that reported by Neroet al. (2013), who also observed no significant correlation be-tween pure ovarian fluid turbidity and egg survival in landlockedChinook Salmon spawns that did not have broken eggs.

FIGURE 2. Survival (%) to the eyed stage of eggs from landlocked fall Chi-nook Salmon in relation to the turbidity of ovarian fluid mixed with either lakewater, distilled water, or hatchery well water.

TABLE 2. Significant correlations between ovarian fluid (OF) turbidity mea-surements (NTU) from female, landlocked, fall Chinook Salmon.

Variable 1 Variable 2 N r P

Ovarianfluid

OF + lakewater

26 0.426 0.030

Ovarianfluid

OF + distilledwater

26 0.502 0.009

Ovarianfluid

OF + wellwater

26 0.505 0.009

OF + lakewater

OF + distilledwater

26 0.917 0.000

OF + lakewater

OF + wellwater

26 0.898 0.000

OF +distilledwater

OF + wellwater

26 0.999 0.000

The quantitative results from this study are difficult to com-pare with the qualitative turbidity assessment of Wojtczak et al.(2004). Wojtczak et al. (2004) visually observed that upon con-tact with water, two types of Rainbow Trout ova could be dis-tinguished: ova causing turbid water and ova that did not causesuch an effect. They concluded that turbidity of the ovarian fluidmixed with water was caused by the coagulation of egg lipidsand lipoproteins in the water and reported that fertilization rateswere lower in the spawns where the ovarian fluid mixed withwater became turbid. However, Wojtczak et al. (2004) did notdirectly measure turbidity. Tabrizi et al. (2011), using TSS asan indirect turbidity measurement, noted significant correlationsbetween the TSS of both ovarian fluid and ovarian fluid mixedwith water and egg survival. However, the relationship betweenTSS and actual turbidity measurements is inconsistent, partic-ularly at higher turbidity levels (Duchrow and Everhart 1971;Sorenson et al. 1977), making comparisons between the resultsof this study and Tabrizi et al. (2011) capricious. In the onlyother study that quantitatively measured ovarian fluid turbidityin NTUs, Nero et al. (2013) found some correlation between tur-bidity and egg survival in spawns containing broken eggs. Therewere only two spawns with broken eggs in the current study, andin these only a very small number of eggs were actually broken.Just as with Nero et al. (2013), there was no correlation betweenovarian fluid turbidity and subsequent egg survival.

Unlike previous ovarian fluid turbidity studies (Wojtczaket al. 2004; Tabrizi et al. 2011; Nero et al. 2013) no spawnsin this study contained overripe eggs, only two females pro-duced spawns with broken eggs, and only one such spawnwas included in the egg survival portion of the study. Bro-ken or overripe eggs likely diffuse lipoproteins, glycoproteins,phosphoproteins, and other egg contents (Heming and Bud-dington 1988) into the ovarian fluid leading to an increase inturbidity, thereby making turbidity a probable indicator of eggsurvival in spawns with broken eggs (Tabrizi et al. 2011). The

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:41

20

July

201

5

TECHNICAL NOTE 21

lack of correlation between turbidity and egg survival found inthis study suggests that turbidity may not be a useful tool forscreening spawns when no broken or overripe eggs are present.The turbidity measured when ovarian fluid samples were mixedwith water may have indicated the presence of egg contentsin the ovarian fluid. However, the wide range of egg survivalobserved in this study suggests there is another mechanism af-fecting egg survival in the absence of overripe or broken eggsthat may not be detectable with turbidity measurements.

The strong correlations between the turbidity measurementsof the ovarian fluid mixed with three different water sourcessuggest that the water source used, while potentially impactingthe absolute measured turbidity, does not affect the trend of tur-bidity measurements for an ovarian fluid sample. Thus, a higherturbidity measurement using one water source will also gen-erally register a higher turbidity measurement with a differentwater source.

In conclusion, the results of this study indicate that the useof ovarian fluid turbidity, or turbidity of ovarian fluid combinedwith water, cannot be used to accurately predict subsequent eggsurvival in landlocked fall Chinook Salmon, at least in the rangeof turbidities experienced in this study and in the absence of anybroken eggs. The significant and high correlations among theturbidities of ovarian fluid mixed with the three different watersources used in this study suggest that comparative evaluationsbetween salmon spawning studies using the turbidity of ovarianfluid mixed with water can be made regardless of the watersource used.

ACKNOWLEDGMENTSThe authors thank Sarah Zimmerman, Patrick Nero, Eric

Krebs, Bob Hanten, and the crew at Whitlocks Spawning Stationfor their assistance with this study.

REFERENCESADFG (Alaska Department of Fish and Game). 1983. Fish culture manual.

ADFG, Juneau.Barnes, M. E., R. P. Hanten, R. J. Cordes, W. A. Sayler, and J. Carreiro. 2000a.

Reproductive performance of inland fall Chinook Salmon. North AmericanJournal of Aquaculture 62:203–211.

Barnes, M. E., R. P. Hanten, W. A. Sayler, and R. J. Cordes. 2000b. Viabilityof inland fall Chinook Salmon spawn containing overripe eggs and reliabilityof egg viability estimates. North American Journal of Aquaculture 62:237–239.

Barnes, M. E., W. A. Sayler, R. J. Cordes, and R. P. Hanten. 2003. Potentialindicators of egg viability in landlocked fall Chinook Salmon spawn with orwithout the presence of overripe eggs. North American Journal of Aquaculture65:49–55.

Duchrow, R. M., and W. H. Everhart. 1971. Turbidity measurement. Transac-tions of the American Fisheries Society 4:682–690.

Fauvel, I. C., M. H. Omnes, M. Suquet, and Y. Normant. 1993. Reliable as-sessment of overripening in turbot (Scopthalmus maximus) by a simple pHmeasurement. Aquaculture 117:107–113.

Heming, T. A., and R. K. Buddington. 1988. Yolk absorption in embryonicand larval fishes. Pages 408–446 in W. S. Hoar and D. J. Randall, editors.The physiology of developing fish, part A. Eggs and larvae, fish physiology,volume XI. Academic Press, San Diego, California.

Lahnsteiner, F., B. Urbanyi, A. Horvath, and T. Weismann. 2001. Bio-markers for egg quality determination in cyprinid fish. Aquaculture 195:333–352.

Lahnsteiner, F., T. Weismann, and R. A. Patzner. 1999. Physiological and bio-chemical parameters for egg quality determination in Lake Trout, Salmo truttalacustris. Fish Physiology and Biochemistry 20:375–388.

Nero, P. A., M. E. Barnes, and M. M. Wipf. 2013. Turbidity of landlocked fallChinook Salmon ovarian fluid in relation to egg survival. Open Fish ScienceJournal 6:75–77.

Piper, R. G., I. B. McElwain, L. E. Orme, J. P. McCraren, L. G. Fowler, and J. R.Leonard. 1982. Fish hatchery management. U.S. Fish and Wildlife Service,Washington, D.C.

Satia, B. P., L. R. Donaldson, L. S. Smith, and J. N. Nightingale. 1974. Com-position of ovarian fluid and eggs of the University of Washington strain ofRainbow Trout (Salmo gairdneri). Journal of the Fisheries Research Boardof Canada 31:1769–1799.

Sorenson, D. L., M. M. McCarthy, E. J. Middlebrooks, and D. B. Porcella.1977. Suspended and dissolved solids effects on freshwater biota: a review.U.S. Environmental Protection Agency, Environmental Research Laboratory,Report 600/3-77-042, Corvallis, Oregon.

Springate, J., and N. Bromage. 1984. Rainbow Trout egg and fry losses, checkon quality. Fish Farmer 7(3):24–25.

Tabrizi H., S. A. Nezami, R. Lorestani, and S. Shamspour. 2011. Brokeneggs influence on fertilization capacity and viability of eggs, turbidityand pH of ovarian fluid and fertilization water in the endangered CaspianBrown Trout, Salmo trutta caspius. International Journal of Biology 3:161–6.

Wojtczak, M., R. Kowalski, S. Dobosz, K. Goryczko, H. Kuzminski, J. Glo-gowski, and A. Ciereszko. 2004. Assessment of water turbidity for evaluationof Rainbow Trout (Oncorhynchus mykiss) egg quality. Aquaculture 242:617–624.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:41

20

July

201

5


Click for updates


Beta-Glucans and Mannan Oligosaccharides EnhanceGrowth and Immunity in Nile TilapiaKhaled M. Selima & Rasha M. Redaa

a Department of Fish Diseases and Management, Faculty of Veterinary Medicine, ZagazigUniversity, Zagazig, Sharkia 44511, EgyptPublished online: 24 Nov 2014.

To cite this article: Khaled M. Selim & Rasha M. Reda (2015) Beta-Glucans and Mannan Oligosaccharides Enhance Growth andImmunity in Nile Tilapia, North American Journal of Aquaculture, 77:1, 22-30, DOI: 10.1080/15222055.2014.951812








http://dx.doi.org/10.1080/15222055.2014.951812




ARTICLE

Beta-Glucans and Mannan Oligosaccharides EnhanceGrowth and Immunity in Nile Tilapia

Khaled M. Selim* and Rasha M. RedaDepartment of Fish Diseases and Management, Faculty of Veterinary Medicine, Zagazig University,Zagazig, Sharkia 44511, Egypt

AbstractWe studied the effects of a combination of dietary beta-glucans (β-G) and mannan oligosaccharides (MOS) on

Nile Tilapia Oreochromis niloticus. Three-hundred-sixty fingerlings (mean mass ± SD = 8.7 ± 0.4 g) were separatedinto three groups (G1, G2, and G3) of 120 fish; G1 (control group) was fed a basal diet, whereas G2 and G3 were fedprebiotic-supplemented diets at final levels of 1.5 and 3.0 g/kg feed, respectively. Each group was subdivided into twosubgroups: subgroup A was fed for 60 d to evaluate growth performance, nutrient utilization, intestinal morphometry,and body composition; and subgroup B was fed for 30 d to evaluate immune status and disease resistance. The bestgrowth and feed utilization were observed in G3. There was no significant difference in final body weight or weightgain between G2 and G1 after 30 d, whereas both variables were significantly higher in G2 than in G1 after 60 d. Atthe end of the feeding period, G2 had a better feed conversion ratio than G1. Villus height, number of goblet cells, andnumber of intraepithelial lymphocytes were greatest in G3, followed by G2 and then G1. Whole-body protein contentand fat content were higher in G3 than in G2 and G1. Only G3 had significantly higher serum total protein, albumin,and globulin than G1. Serum killing percentage and phagocytic activity were significantly higher in G3 than in G1

and G2, whereas serum lysozyme activity was significantly higher in G3 and G2 than in G1. The nitric oxide assayindicated a significant effect in G3 compared with G1 after 30 d. Fish that were fed the prebiotic mixture had betterrelative percent survival than G1 fish after challenge with Yersinia ruckeri. Dietary supplementation with β-G andMOS in combination improves the performance of Nile Tilapia.

Several food supplements have been used to improve growthperformance, modulate immune status, and increase disease re-sistance in various fish species. Bacterial diseases are majorproblems encountered in fish aquaculture and have been com-bated mainly with antibiotics (Austin and Austin 2007). How-ever, the use of antibiotics poses threats, such as the creationof antibiotic-resistant genes, immunosuppression, destabiliza-tion of beneficial gastrointestinal bacteria, and accumulationof antibiotics in the musculature (Alderman and Hastings 1998;Teuber 2001; Romero et al. 2012). Thus, antibiotic use is strictlyregulated worldwide, and the use of alternatives in aquacultureis receiving greater attention.

A prebiotic is a nondigestible food ingredient that benefi-cially affects the host by improving the balance of its intestinalmicroflora (Gibson et al. 2004; Qiang et al. 2009; Rurangwa

*Corresponding author: [email protected] April 12, 2014; accepted June 29, 2014

et al. 2009). Beta-glucans (β-G) and mannan oligosaccharides(MOS) are prebiotics commonly used in fish and are natu-rally occurring polysaccharides found in the cell walls of theyeast Saccharomyces cerevisiae. Other sources, such as brew-ers’ yeast, torula yeast Candida utilis, fungi, and algae, arealso currently used (Raa 1996). Prebiotics stimulate growthperformance, improve nutrient availability, modulate microbialcolonization, improve gut development, modulate innate andacquired immunity responses, enhance resistance to the de-velopment of potential pathogens, minimize the side effectsassociated with vaccines and drug therapy, and absorb myco-toxins found in nutrients (Sakai 1999; Szilagyi 2002; Li andGatlin 2004; Torrecillas et al. 2007; Dalmo and Bogwald 2008;Andrews et al. 2009; Yousefian and Amiri 2009; Geraylou et al.2013; Selim et al. 2014).

22

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:41

20

July

201

5

ENHANCEMENT OF NILE TILAPIA GROWTH 23

Enteric redmouth disease (ERM), or yersiniosis, is an in-fectious acute or subacute disease of fish caused by the Gram-negative, rod-shaped bacterium Yersinia ruckeri. This diseaseleads to severe mortality (60–70%) and high economic losses(Roberts 1983; Frerichs and Roberts 1989). Enteric redmouthdisease was first reported in the 1950s in the USA and thenspread throughout the world, especially during the last twodecades (Lesel et al. 1983; Lucangeli et al. 2000; Tobback et al.2007). It is characterized by septicemia with hemorrhagic zoneson external surfaces (particularly around the mouth) and in inter-nal organs (particularly the intestines; Austin and Austin 2007).Yersinia ruckeri is an opportunistic pathogen that commonly in-habits the small intestine, spleen, and liver of asymptomatic fish(Busch 1978). Stress factors, such as heat discomfort, oxygendepletion, overcrowding, pollution, bad handling, and infectiousdisease, can lead to immunosuppression and cause outbreaks ofERM (15–18◦C is suitable for symptomatic disease; Guguianuand Miron 2002). Antibiotic control is ineffective for the acuteform of ERM; consequently, dealing with ERM depends onthe use of vaccines and immunostimulators (Tebbit et al. 1981;Ghittino 1985).

The majority of prior studies have used one prebiotic ina fish’s diet. However, few studies have used a combinationof yeast polysaccharides, although several commercial prepa-rations have been formulated for aquaculture use. The yeastpolysaccharides β-G and MOS have two different modes ofaction. Few data are available about their application for pro-moting growth or immunostimulation. In the present study, thecombined effects of β-G and MOS on Nile Tilapia Oreochromisniloticus were evaluated. The polysaccharides were provided asa dietary supplement at two levels, and the fish were monitoredfor possible effects on growth performance, food utilization,gut morphology, body composition, immune response, and re-sistance to Y. ruckeri infection.

METHODSFish.—Three-hundred-sixty Nile Tilapia fingerlings (mean

mass ± SD = 8.7 ± 0.4 g) were obtained from AbbassaFish Hatchery, Sharkia Province, Egypt. Fish were apparentlyhealthy and free from any serious diseases, parasites, or skinlesions. For 15 d, the fish were acclimated to experimental con-ditions in glass aquaria (80 × 40 × 30 cm) and were fed a basaldiet. The water temperature, pH, dissolved oxygen, and salinitywere measured daily, while carbon dioxide, carbonate hard-ness, ammonia-nitrogen, nitrite-nitrogen, and nitrate-nitrogenwere measured once per week according to standard methods(APHA et al. 1998). About 25% of the water was changed withthe daily water sampling.

Experimental design and diet preparation.—Fingerlingswere separated randomly into three groups (G1, G2, and G3)to measure growth and immunity parameters. Each group wassubdivided into two subgroups (A and B; each subgroup had trip-licates of 20 fish): subgroup A was used for growth performance

TABLE 1. Ingredient and chemical composition of the experimental diets fedto three groups of Nile Tilapia (G1 = control group, no dietary supplementation;G2 = diet supplemented with prebiotic combination at 1.5 g/kg; G3 = dietsupplemented with prebiotic combination at 3.0 g/kg).

Ingredient G1 G2 G3

Ingredient (% dry weight)Fish meal 25 25 25Meat meal 20 20 20Soybean meal 22 22 22Yellow corn 26.5 26.5 26.5Fish oil 5 5 5Vitamin and mineral premixa 1.5 1.5 1.5Prebiotic combination (g/kg

feed)0.0 1.5 3.0

aPharma mix, batch number 02100033.

evaluation for 60 d, after which samples were collected for bodycomposition analysis and intestinal histology; and subgroup Bwas used for twice-monthly immunological examination fol-lowed by intraperitoneal challenge with Y. ruckeri to examinefish resistance.

The control group (G1) was fed a basal diet without sup-plementation; the diet contained approximately 39.9% crudeprotein and 10.89% crude lipid (Table 1). The treatment groups(G2 and G3) were fed an animal health care product contain-ing a combination of β-G (at 300 g/kg) and MOS (at 180 g/kg;Immunowall, Pro Vet Care, Brazil). The product was added tobasal diets at 1.5 g/kg for G2 and 3.0 g/kg for G3. Fish were fedthree times daily at a rate of 5% of fish live body weight.

Growth variables.—The fish were weighed every 2 weeks toassess growth performance. The final body weight, weight gain,specific growth rate, and feed conversion ratio were determinedas described by De Silva and Anderson (1995) and Yanbo andZirong (2006). The protein efficiency ratio was calculated asdescribed by Stuart and Hung (1989).

Intestinal histology.—After 8 weeks of feeding, three por-tions of the intestine were fixed in Bouin’s solution: the proximalpart (between the end of the gastric pylorus and the beginningof the spiral intestine), middle part (spiral intestine), and distalpart (between the end of the spiral intestine and 2 cm before theanus). The samples were dehydrated in an ascending alcohol se-ries (70–100%) and embedded in paraffin; tissue sections werethen stained with hematoxylin and eosin. The heights of intesti-nal villi in all parts were measured by using Image J version1.36 (National Institutes of Health). Mean villus height for eachsection was based on an average of 10 villus heights/section.Other intestinal sections were stained with Alcian blue to countgoblet cells and intraepithelial lymphocytes (IELs) within 10intestinal folds (Samanya and Yamauchi 2002).

Chemical analysis of fish body composition.—At the end ofthe feeding period for subgroup A, five fish from each replicatewere randomly selected, autoclaved, ground into a homoge-neous slurry, oven-dried, and reground. Carcass samples were

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:41

20

July

201

5

24 SELIM AND REDA

analyzed for dry matter and ash (as described by AOAC 2000);crude protein (N × 6.25) was measured by the Kjeldahl methodusing a Kjeltec auto analyzer (Model 1030; Tecator, Hgans,Sweden), and total lipid was measured using the method ofBligh and Dyer (1959).

Nonspecific immune analysis.—In subgroup B, blood sam-ples (5 samples/replicate) were collected from the caudal bloodvessels every 2 weeks and were divided into two portions. Oneportion was collected with heparin and was used to measurephagocytic activity; the white blood cells were separated fromthe peripheral blood of the tested fish in the different experimen-tal groups by layering over an equal volume of Ficoll–Hypaquesolution followed by centrifugation at 1,000 rotations/min for30 min (Ourth and Chung 2004). Heat-inactivated Candidaalbicans was used to determine phagocytic activity (Kumariand Sahoo 2006). The second portion was collected withoutanticoagulant and was centrifuged at 3,000 rotations/min for15 min. Serum was collected and stored at −20◦C until use.Levels of serum total protein and albumin were measured spec-trophotometrically by using specific kits (Spectrum Diagnostics,Egyptian Company for Biotechnology, Obour City, Cairo,Egypt). Globulin levels were determined by direct subtractionof the albumin values from the total protein values (Coles 1974).Serum lysozyme, nitric oxide, and serum bactericidal activitywere also determined by using published methods (Ellis 1990;Kajita et al. 1990; Rajaraman et al. 1998).

Challenge test.—After 30 d, the fish in subgroup B wereinjected intraperitoneally with 0.1 mL of Y. ruckeri (1.5 ×108 cells/mL) that had been previously isolated from moribundfish and confirmed to be pathogenic. Challenged fish were keptunder observation for 14 d to record ERM. Samples were takenfrom the kidney and intestine of dead and clinically infected fishto re-isolate Y. ruckeri (Eissa et al. 2008). The average mortalityamong all replicates was used for calculating relative percentsurvival (RPS; Amend 1981).

Statistical analysis.—Means and SDs were calculated foreach variable. The data were analyzed by ANOVA and posthoc Duncan’s multiple comparisons to identify the significantlydifferent groups. Fisher’s exact probability test was used to ana-lyze the results of RPS calculations. All statistical analyses wereperformed with GraphPad Prism 4 software (GraphPad Soft-ware, Inc., San Diego, California) with a significance level α of0.05.

RESULTS

Growth PerformanceThe growth performance of and feed utilization by Nile

Tilapia that were fed two different concentrations of β-G andMOS in combination for 60 d are presented in Figure 1. After 30d, G3 had a significantly higher final body weight, weight gain,and specific growth rate than the control group G1; G2 also hada significantly higher weight gain than G1 (Figure 1A, B). Thefeed conversion ratio after 30 d was significantly lower in G3

FIGURE 1. Growth performance (mean + SD) of Nile Tilapia that received adietary combination of beta glucans (β-G) and mannan oligosaccharides (MOS)at two levels (G1 = control group, no dietary supplementation; G2 = dietsupplemented with prebiotic combination at 1.5 g/kg; G3 = diet supplementedwith prebiotic combination at 3.0 g/kg) for 2 months: (A) initial body weight(IBW), final body weight after 30 d (FBW1), final body weight after 60 d(FBW2), weight gain after 30 d (WG1), and weight gain after 60 d (WG2);and (B) specific growth rate after 30 d (SGR1), specific growth rate after 60 d(SGR2), feed conversion ratio after 30 d (FCR1), feed conversion ratio after 60d (FCR2), and protein efficiency ratio (PER). Means with different letters (a, b,and c) are significantly different (one-way ANOVA: P < 0.05).

than in G1 (Figure 1B). After 60 d, G3 fish had a significantlyhigher final body weight than G2 and G1; both G2 and G3 hadsignificantly greater weight gain and specific growth rate thanG1. The best (lowest) feed conversion ratio was observed for G3,followed by G2 and then G1 (Figure 1B). The protein efficiencyratio was significantly higher in G3 than in G2 and G1 (Figure1B).

Intestinal HistologyMicroscopic examination of the intestine showed that intesti-

nal villus height (Figure 2A) and microvillus density in all partsof the intestine exhibited significant increases in G3 comparedwith both G2 and G1 and were significantly greater in G2 thanin G1. The prebiotic-fed fish had significantly more goblet cellsthan the controls. The number of goblet cells in the proximaland middle parts of the intestine was significantly greater inG3 than in G2 and was significantly greater in G2 than in G1

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:41

20

July

201

5


FIGURE 2. Intestinal histology (mean + SD) of Nile Tilapia that received adietary combination of beta glucans (β-G) and mannan oligosaccharides (MOS)at two levels for 2 months: (A) villus height; (B) number of goblet cells; and(C) number of intraepithelial lymphocytes in the proximal, middle, and distalparts of the intestine. Means with different letters (a, b, and c) are significantlydifferent (one-way ANOVA: P < 0.05). See Figure 1 for a definition of treatmentgroups (G1–G3).

(Figure 2B). The number of goblet cells in the distal portion ofthe intestine showed no significant differences between G3 andG2. Fish belonging to G3 showed a significant increase in IELnumbers in all parts of the intestine relative to fish in G2 andG1 (Figure 2C). The number of IELs in the proximal and distalparts of the intestine was significantly greater in G2 than in G1,whereas the number of IELs in the middle part of the intestinedid not differ between these two groups.

FIGURE 3. Body composition (mean + SD) of Nile Tilapia that received adietary combination of beta glucans (β-G) and mannan oligosaccharides (MOS)at two levels for 2 months: (A) percent moisture and percent crude protein; and(B) percent crude lipid and percent ash content. Means with different letters (aand b) are significantly different (one-way ANOVA: P < 0.05). See Figure 1for a definition of treatment groups (G1–G3).

Body Composition AnalysisPercentages of crude protein and lipid were significantly

higher in G3 than in G2 and G1 (Figure 3A, B). However, thewhole-body moisture content was highest in G1. The ash contentwas not significantly affected by prebiotic dietary supplementa-tion.

Nonspecific Immune ResponseTotal serum protein, albumin, and globulin were significantly

higher in G3 than in G1 (Figure 4). There were no significantdifferences in these variables between G1 and G2. After both 15and 30 d, the serum killing percentages and phagocytic activitypercentages were significantly higher in G3 than in G2 and G1

(Figure 5). Lysozyme activity was significantly greater in G3 andG2 than in G1. The nitric oxide assay exhibited no significantdifference among the three groups after 15 d, whereas the assayindicated a significant increase for G3 relative to G1 after 30 d(Figure 5).

Challenge TestThe challenge test showed that administering dietary

prebiotic to Nile Tilapia enhanced their resistance to

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:41

20

July

201

5

26 SELIM AND REDA

FIGURE 4. Total protein content, albumin content, and globulin content(mean + SD) in Nile Tilapia that received a dietary combination of beta glu-cans (β-G) and mannan oligosaccharides (MOS) at two levels for 1 month.Means with different letters (a and b) are significantly different (one-wayANOVA: P < 0.05). See Figure 1 for a definition of treatment groups(G1–G3).

Y. ruckeri infection (Figure 6). Moribund fish were character-ized by hemorrhaging in different parts of the body and aroundthe mouth as well as congestion in internal organs. The percentmortality was significantly different among G3 (26 ± 4%), G2

(55 ± 4%), and G1 (65 ± 5%). Accordingly, the highest RPSwas observed for G3, which had significantly greater RPS thanG2 and G1.

FIGURE 6. Percent mortality and relative percent survival (RPS; mean +SD) of Nile Tilapia that received a dietary combination of beta glucans (β-G) and mannan oligosaccharides (MOS) at two levels for 30 d and then werechallenged with Yersinia ruckeri. Means with different letters (a, b, and c) aresignificantly different (Fisher’s exact probability test: P < 0.05). See Figure 1for a definition of treatment groups (G1–G3).

DISCUSSIONPrebiotic supplementation is of great interest in aquacul-

ture. Prebiotics have beneficial effects on fish health (Gangulyet al. 2013). Our results indicate that supplementation with β-Gand MOS (0.3% in feed by weight) improved growth perfor-mance in and feed utilization by Nile Tilapia. Use of prebioticscontaining β-G and MOS was previously found to improve the

FIGURE 5. Nonspecific immune response (mean + SD) in Nile Tilapia that received a dietary combination of beta glucans (β-G) and mannan oligosaccharides(MOS) at two levels for 1 month: (A) serum killing percentage; (B) serum lysozyme activity (µg/mL); (C) phagocytic activity; and (D) serum nitric oxide (mmol−1).Means with different letters (a, b, and c) are significantly different (one-way ANOVA: P < 0.05). See Figure 1 for a definition of treatment groups (G1–G3).

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:41

20

July

201

5


growth performance of Common Carp Cyprinus carpio, Bel-uga Huso huso, and sea cucumbers Apostichopus japonicus (Guet al. 2011; Ta’ati et al. 2011; Ebrahimi et al. 2012). Ye et al.(2011) reported that the use of two types of prebiotic (fructo-oligosaccharides and MOS) in the diets of Olive Flounder Par-alichthys olivaceus yielded better growth performance than theuse of a single product. Welker et al. (2012) reported that NileTilapia given diets supplemented with 0.1% β-G showed im-provements in weight gain and feed utilization efficiency. Pre-vious studies found that the addition of β-G or MOS had nopositive effect on Nile Tilapia growth (Whittington et al. 2005;Sado et al. 2008; Shelby et al. 2009), although the administrationof whole S. cerevisiae did improve Nile Tilapia growth (Lara-Flores et al. 2003). Many studies have reported that MOS stimu-late growth in other fishes, such as Yellow Catfish Pelteobagrusfulvidraco, Rainbow Trout Oncorhynchus mykiss, Olive Floun-der, Atlantic Salmon Salmo salar, European Bass Dicentrarchuslabrax, and Gilthead Bream Sparus auratus (Staykov et al. 2007;Torrecillas et al. 2007; Helland et al. 2008; Gultepe et al. 2010;Ye et al. 2011; Wu et al. 2014). In parallel with these studies,a few reports have shown a positive correlation between β-Gadministration and growth in species such as the mirror carp (avariant of Common Carp), Pacific Red Snapper Lutjanus peru,and marron (crayfish) Cherax spp. (Sang and Fotedar 2010;Guzman-Villanueva et al. 2014; Kuhlwein et al. 2014). In con-trast, other studies have found no effect of β-G or MOS additionon the growth of Gulf Sturgeon Acipenser oxyrinchus desotoi(Pryor et al. 2003), European Bass (Bagni et al. 2005), RainbowTrout (Dobsikova et al. 2012), or Atlantic Salmon (Refstie et al.2010).

As these studies suggest, the use of β-G and MOS in aqua-culture can have a positive effect, a negative effect, or no effecton growth. The variation in results could be due to differences inthe chemical structure of the additives, preparation of the com-pounds, the origin and inclusion level of the prebiotic, fish ageand species, and experimental conditions (Akrami et al. 2009).The action of polysaccharides produced from the cell walls ofyeasts varies according to structural complexity and differencesin fermentation and processing procedures, which can causedifferences in subsequent effects (Sado et al. 2008). Mannanoligosaccharides might improve growth through modulation ofintestinal microbiota, thereby increasing villus integrity and re-sistance to pathogenic bacteria with a consequent increase in theefficiency of digestion and absorption (Fernandez et al. 2002;Andrews et al. 2009; Dimitroglou et al. 2009). On the otherhand, β-G can be degenerated by glucanases, thus producingenergy and building protein for growth (Misra et al. 2006). NileTilapia treated with β-G showed a significant increase in foodintake, which should be reflected in their growth rate (Shelbyet al. 2009).

Our results showed that a combined application of β-G andMOS significantly increased the intestinal villus heights, thenumber of goblet cells, and the number of IELs in all partsof the intestine. Zhu et al. (2012) reported that supplementa-

tion with yeast polysaccharides (0.1, 0.2, and 0.3% of feed byweight)—primarily β-G and MOS—increased the height of in-testinal folds and increased the number of goblet cells in Chan-nel Catfish Ictalurus punctatus. Administering MOS increasedmicrovillus length and density in Rainbow Trout, Red Drum Sci-aenops ocellatus, and White Bass Morone chrysops × StripedBass Morone saxatilis hybrids (Dimitroglou et al. 2009; Zhouet al. 2010; Anguiano et al. 2012). Several authors have reportedno significant differences in intestinal histology after MOS sup-plementation in hybrid tilapia (Nile Tilapia × Blue TilapiaOreochromis aureus), Gulf Sturgeon, and European Bass (Pryoret al. 2003; Genc et al. 2007; Torrecillas et al. 2007). A recentinvestigation of mirror carp demonstrated that β-G preparationhad no significant effects on the intestinal absorptive surface orgoblet cells, although it did result in higher leukocyte infiltrationof the anterior portion of the intestine (Kuhlwein et al. 2014).Zhu et al. (2012) showed that diet supplementation with 0.1%or 0.2% yeast polysaccharides increased the intestinal gobletcell numbers in Channel Catfish. Increased microvillus heightand a healthy mucosal epithelium that is rich in goblet cells andIELs are positively correlated with growth enhancement andfeed utilization by improving nutrient digestion and absorption,intestinal secretions, and resistance to opportunistic indigenousbacteria (Staykov et al. 2007; Dimitroglou et al. 2009; Gentenet al. 2009).

Supplementation with β-G and MOS at 0.3% increased thecrude protein and fat content of Nile Tilapia. Ebrahimi et al.(2012) showed that a combination of β-G and MOS added to thediet at 2.5 g/kg improved the crude protein content in CommonCarp fingerlings. In addition, supplementation with whole S.cerevisiae improved the protein and lipid content of Nile Tilapia(Lara-Flores et al. 2003; Abdel-Tawwab et al. 2008). Similarcrude protein improvements have been observed in hybrid tilapia(Nile Tilapia × Blue Tilapia) and Rainbow Trout fed a dietsupplemented with MOS at 4.5 g/kg (Genc et al. 2007; Yilmazet al. 2007). This can be attributed to increased feed intakeand improved digestibility and utilization of nutrients (Misraet al. 2006; Andrews et al. 2009; Shelby et al. 2009). Razeghiet al. (2012) reported improvement in the crude lipid contentof Beluga that were fed MOS at 2 g/kg. In contrast, AtlanticSalmon that received a diet supplemented with MOS at 10 g/kgshowed a significant decrease in body protein content (Hellandet al. 2008).

Our results show that dietary β-G and MOS improve non-specific immunity levels in Nile Tilapia and consequently in-crease their resistance to Y. ruckeri challenge. The effects ofdietary prebiotics on nonspecific immune responses vary con-siderably among studies, even for the same measured variablesor fish species. We found increased nonspecific immunity lev-els after β-G and MOS supplementation (0.3%), as evidencedby increased serum globulin, serum killing activity, phago-cytic activity, and lysozyme activity after 15 d of feeding andby improved nitric oxide activity after 30 d. Torrecillas et al.(2007) found that feeding a mixture of MOS and β-G for 60 d

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:41

20

July

201

5

28 SELIM AND REDA

improved immune status and stress resistance in European Bass.Gu et al. (2011) reported a synergistic effect between dietary β-G and MOS in A. japonicus, as indicated by prolonged elevationof immune indices (relative to a single supplementation with β-G or MOS) and increased total nitric oxide synthesis activity.El-Boshy et al. (2010) reported a significant increase in cel-lular and humoral immunity in Nile Tilapia that received β-Gat 0.1% of the diet for 3 weeks. In contrast, Nile Tilapia thatwere given the same concentration of β-G for 4 weeks showedno improvements in immune variables except for the respira-tory burst (Welker et al. 2012). Feeding of different types ofglucans or whole yeast was reported to increase nonspecificimmunity in tilapias (Wang et al. 2007; El-Boshy et al. 2010).Administration of β-G or MOS increased the activity of non-specific defense mechanisms in Common Carp, Walking CatfishClarias batrachus, Rohu Labeo rohita, Rainbow Trout, PacificRed Snapper, and A. japonicus (Sahoo and Murkherjee 2002;Kumari and Sahoo 2006; Staykov et al. 2007; Gu et al. 2010;Guzman-Villanueva et al. 2014).

In the absence of specific opsonization, alternative immuneresponses could depend on the presence of mannose receptorsand toll-like receptors (TLRs) in microbes, which bind to man-nose and glucans, leading to enhanced phagocytic and bacteri-cidal abilities in phagocytes and neutrophils (Ofek et al. 1995;Engering et al. 1997; Stafford et al. 2003; Bricknell and Dalmo2005; Rebl et al. 2009). Mannan oligosaccharides bind with andblock receptors on pathogens, preventing their colonization orinvasion of the host; MOS also enhance the liver’s secretion ofa material rich in mannose-binding lectin, which binds the bac-terial capsule and triggers the complement cascade (Janeway1993; Newman 2001; Fernandez et al. 2002). Nonspecific pro-tection against pathogenic agents is more important than specificprotection because it is an earlier immune response (Ellis 2001;Bricknell and Dalmo 2005). The use of injected or ingested glu-cans increased the effectiveness of a vaccine against yersiniosisin Rainbow Trout and Atlantic Salmon (Robertsen et al. 1990;Siwicki et al. 2004). In addition, the use of glucan in combi-nation with vitamin C increased the levels of specific antibodyagainst Y. ruckeri in Rainbow Trout (Verlhac et al. 1996). Previ-ous reports have shown that when incorporated into feed, MOSor β-G increase resistance to pathogenic bacteria in different fishspecies (Sahoo and Murkherjee 2002; Kumari and Sahoo 2006;Torrecillas et al. 2007; Welker et al. 2007). Some studies havereported that MOS inhibit colonization of the chicken smallintestine by Enterobacteriaceae (summarized by Oyofo et al.1989). In contrast, RPS was not affected in Nile Tilapia thatwere given a diet supplemented with β-G and then challengedwith Streptococcus iniae or Edwardsiella tarda (Whittingtonet al. 2005; Shelby et al. 2009).

Our results show that a combination of β-G and MOS ad-ministered at 0.3% by weight of feed influences growth per-formance, intestinal histology, and body composition in NileTilapia. This prebiotic improves the nonspecific immune sta-tus of fingerlings, as indicated by bactericidal activity, phago-

cytic activity, lysozyme activity, and nitric oxide level, and alsoimproves protection against infection with Y. ruckeri, therebyacting as a robust immunostimulant.

ACKNOWLEDGMENTSThis work was supported by a grant from the postgraduate

studies sector of Zagazig University to R.M.R. We thank HatemM. Selim, Selim Pharm Co., for supplying us with prebioticcombination. Authors K.M.S. and R.M.R. contributed equallyto this work.

REFERENCESAbdel-Tawwab, M., A. M. Abdel-Rahman, and E. M. Ismael. 2008. Eval-

uation of commercial live bakers’ yeast, Saccharomyces cerevisiae as agrowth and immunity promoter for fry Nile Tilapia, Oreochromis niloti-cus (L.) challenged in situ with Aeromonas hydrophila. Aquaculture 280:185–189.

Akrami, R., A. M. Hajimoradloo, A. Matinfar, and A. M. Abedian Kinari.2009. Effect of dietary prebiotic inulin on growth performance, intestinal mi-croflora, body composition and hematological parameters of juvenile Beluga,Huso huso (Linnaeus 1758). Journal of the World Aquaculture Society 40:771–779.

Alderman, D. J., and T. S. Hastings. 1998. Antibiotic use in aquaculture: devel-opment of antibiotic resistance–potential for consumer health risks. Interna-tional Journal of Food Sciences and Technology 33:139–155.

Amend, D. F. 1981. Potency testing of fish vaccines. Development in BiologicalStandardization 49:447–454.

Andrews, S. R., N. P. Sahu, A. K. Pal, and S. Kumar. 2009. Haematologicalmodulation and growth of Labeo rohita fingerlings: effect of dietary mannanoligosaccharide, yeast extract, protein hydrolysate and chlorella. AquacultureResearch 41:61–69.

Anguiano, M., C. Pohlenz, A. Buentello, and D. M. Gatlin. 2012. The effectsof prebiotics on the digestive enzymes and gut histomorphology of RedDrum (Sciaenops ocellatus) and hybrid Striped Bass (Morone chrysops ×M. saxatilis). British Journal of Nutrition 4:1–7.

AOAC (Association of Official Analytical Chemists). 2000. Official methods ofanalysis. AOAC International, Gaithersburg, Maryland.

APHA (American Public Health Association), American Water Works Associ-ation, and Water Pollution Control Federation. 1998. Standard methods forexamination of water and wastewater, 16th edition. APHA, Washington, D.C.

Austin, B., and D. A. Austin. 2007. Bacterial fish pathogens: diseases of farmedand wild fish, 4th edition. Praxis Publishing, Chichester, UK.

Bagni, M., N. Romano, M. G. Finoia, L. Abelli, G. Scapigliati, P. G. Tiscar, M.Sartie, and G. Marino. 2005. Short- and long-term effects of a dietary yeastbeta-glucan (Macrogard) and alginic acid (Ergosan) preparation on immuneresponse in Sea Bass (Dicentrarchus labrax). Fish and Shellfish Immunology18:311–325.

Bligh, E. G., and W. J. Dyer. 1959. A rapid method of total lipid extrac-tion and purification. Canadian Journal of Biochemistry and Physiology 37:911–917.

Bricknell, I., and R. A. Dalmo. 2005. The use of immunostimulants in fish larvalaquaculture. Fish and Shellfish Immunology 19:457–472.

Busch, R. A. 1978. Enteric red mouth disease (Hagerman strain). Marine Fish-eries Review 40:42–51.

Coles, E. 1974. Veterinary clinical pathology, 2nd edition. Saunders, Philadel-phia.

Dalmo, R. A., and J. Bogwald. 2008. Beta-glucans as conductors of immunesymphonies. Fish and Shellfish Immunology 25:384–396.

De Silva, S. S., and T. A. Anderson. 1995. Fish nutrition in aquaculture. Chap-man and Hall, London.

Dimitroglou, A., D. L. Merrifield, R. Moate, S. J. Davies, P. Spring,J. Sweetman, and G. Bradley. 2009. Dietary mannan oligosaccharide

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:41

20

July

201

5


supplementation modulates intestinal microbial ecology and improves gutmorphology of Rainbow Trout, Oncorhynchus mykiss (Walbaum). Journal ofAnimal Science 87:3226–3234.

Dobsikova, R., J. Blahova, A. Franc, J. Jakubik, I. Mikulikova, H. Modra,K. Novotna, and Z. Svobodova. 2012. Effect of β-1.3/1.6-D-glucan derivedfrom oyster mushroom Pleurotus ostreatus on biometrical, haematological,biochemical, and immunological indices in Rainbow Trout (Oncorhynchusmykiss). Neuroendocrinology Letters 33:96–106.

Ebrahimi, G. H., H. Ouraji, M. K. Khalesi, M. Sudagar, A. Barari, M. ZareiDangesaraki, and K. H. Khalili. 2012. Effects of a prebiotic, immunogen, onfeed utilization, body composition, immunity and resistance to Aeromonashydrophila infection in the Common Carp Cyprinus carpio (Linnaeus) fin-gerlings. Journal of Animal Physiology and Animal Nutrition 96:591–599.

Eissa, A. E., M. Moustafa, M. Abdelaziz, and N. A. Ezzeldeen. 2008. Yersiniaruckeri infection in cultured Nile Tilapia, Oreochromis niloticus, at a semi-intensive fish farm in lower Egypt. African Journal of Aquatic Science33:283–286.

El-Boshy, M. E., A. M. El-Ashram, F. M. Abdelhamid, and H. A. Gadalla. 2010.Immunomodulatory effect of dietary Saccharomyces cerevisiae, β-glucan andlaminaran in mercuric chloride treated Nile Tilapia (Oreochromis niloticus)and experimentally infected with Aeromonas hydrophila. Fish and ShellfishImmunology 28:802–808.

Ellis, A. E. 1990. Lysozyme assays. Pages 101–103 in J. S. Stolen, T. C. Fletcher,D. P. Anderson, B. S. Roberson, and W. B. Van Muiswinkel, editors. Tech-niques in fish immunology. SOS Publications, Fair Haven, New Jersey.

Ellis, A. E. 2001. Innate host mechanism of fish against viruses and bacteria.Developmental and Comparative Immunology 25:827–839.

Engering, A. J., M. Cella, D. Fluitsma, M. Brockhaus, E. C. Hoefsmit, A.Lanzavecchia, and J. Pieters. 1997. The mannose receptor functions as ahigh capacity and broad specificity antigen receptor in human dendritic cells.European Journal of Immunology 27:2417–2425.

Fernandez, F., M. Hinton, and B. V. Gils. 2002. Dietary mannan-oligosaccharides and their effect on chicken caecal microflora in relationto Salmonella enteritidis colonization. Avian Pathology 31:49–58.

Frerichs, G. N., and R. J. Roberts. 1989. The bacteriology of teleosts. Pages289–320 in R. J. Roberts, editor. Fish pathology, 2nd edition. Bailliere Tindall,London.

Ganguly, S., K. C. Dora, S. Sarkar, and S. Chowdhury. 2013. Supplementationof prebiotics in fish feed: a review. Reviews in Fish Biology and Fisheries23:195–199.

Genc, M. A., E. Yilmaz, E. Genc, and M. Aktas. 2007. Effects of dietary mannanoligosaccharides (MOS) on growth, body composition and intestine and liverhistology of the hybrid tilapia (Oreochromis niloticus × O. aureus). IsraeliJournal of Aquaculture 59:10–16.

Genten, F., E. Terwinghe, and A. Danguy. 2009. Atlas of fish histology. SciencePublishers, Enfield, New Hampshire.

Geraylou, Z., C. Souffreau, E. Rurangwa, L. De Meester, C. M. Courtin,J. A. Delcour, J. Buyse, and F. Ollevier. 2013. Effects of dietaryarabinoxylan-oligosaccharides (AXOS) and endogenous probiotics on thegrowth performance, non-specific immunity and gut microbiota of juvenileSiberian Sturgeon (Acipenser baerii). Fish and Shellfish Immunology 35:766–775.

Ghittino, P. 1985. Tecnologia e patologia in aquacoltura, volume 2, patologia.[Technology and pathology in aquaculture, volume 2, pathology.] TipografyEmilio, Bono, Torino, Italy.

Gibson, G. R., H. M. Probert, J. Van Loo, R. A. Rastall, and M. B. Roberfroid.2004. Dietary modulation of the human colonic microbiota: updating theconcept of prebiotics. Nutrition Research Reviews 17:259–275.

Gu, M., H. Ma, K. Mai, W. Zhang, N. Bai, and X. Wang. 2011. Effects of dietaryβ-glucan, mannan oligosaccharide and their combinations on growth perfor-mance, immunity and resistance against Vibrio splendidus of sea cucumber,Apostichopus japonicus. Fish and Shellfish Immunology 31:303–309.

Gu, M., K. Mai, W. Zhang, Q. Ai, X. Wang, and N. Bai. 2010. Immune responseof sea cucumber Apostichopus japonicus coelomocytes to several immunos-timulants in vitro. Aquaculture 306(1–4):49–56.

Guguianu, E., and L. Miron. 2002. Elemente de ihtiopatologie. [Elements offish pathology.] PIM, Iasi, Romania. (In Romanian.)

Gultepe, N., S. Salnur, B. Hossu, and O. Hisar. 2010. Dietary supplementa-tion with mannan oligosaccharides (MOS) from Bio-Mos enhances growthparameters and digestive capacity of Gilthead Sea Bream (Sparus aurata).Journal of Aquaculture Nutrition 17:482–487.

Guzman-Villanueva, L. T., F. Ascencio-Valle, M. E. Macıas-Rodrıguez, and D.Tovar-Ramırez. 2014. Effects of dietary β-1,3/1,6-glucan on the antioxidantand digestive enzyme activities of Pacific Red Snapper (Lutjanus peru) af-ter exposure to lipopolysaccharides. Fish Physiology and Biochemistry 40:827–837

Helland, B., S. J. Helland, and D. M. Gatlin. 2008. The effect of dietary sup-plementation with mannanoligosaccharide, fructooligosaccharide or galac-tooligosaccharide on the growth Atlantic Salmon (Salmo salar). Aquaculture283:163–167.

Janeway, C. A. Jr. 1993. How the immune system recognizes invaders. ScientificAmerican 269:72–79.

Kajita, Y., M. Sakai, S. Atsuta, and M. Kobayash. 1990. The immunomodu-latory effects of levamisole on Rainbow Trout, Oncorhynchus mykiss. FishPathology 25:93–98.

Kuhlwein, H., D. L. Merrifield, M. D. Rawling, A. D. Foey, and S. J. Davies.2014. Effects of dietary β-(1,3)(1,6)-D-glucan supplementation on growthperformance, intestinal morphology and haemato-immunological profile ofMirror Carp (Cyprinus carpio L.). Journal of Animal Physiology and AnimalNutrition 98:279–289.

Kumari, J., and P. K. Sahoo. 2006: Non-specific immune response of healthyand immunocompromised Asian Catfish (Clarias batrachus) to several im-munostimulants. Aquaculture 255:133–141.

Lara-Flores, M., M. A. Olvera-Novoa, B. E. Guzman-Mendez, and W. Lopez-Madrid. 2003. Use of the bacteria Streptococcus faecium and Lactobacillusacidophilus, and the yeast Saccharomyces cerevisiae as growth promoters inNile Tilapia (Oreochromis niloticus). Aquaculture 216:193–201.

Lesel, R., M. Lesel, F. Gavini, and A. Vuillaume. 1983. Outbreak of entericred mouth disease in Rainbow Trout, Salmo gairdneri richardson, in France.Bulletin of the European Association of Fish Pathologists 6:385–387.

Li, P., and D. M. Gatlin III. 2004. Dietary brewers yeast and the prebiotic Grobi-otic TMAE influence growth performance, immune responses and resistanceof hybrid Striped Bass (Morone chrysops × M. saxatilis) to Streptococcusiniae infection. Aquaculture 231:445–456.

Lucangeli, C., A. Morabito, A. Caprioli, L. Achene, L. Busani, E. Mazzolini,A. Fabris, and A. Macri. 2000. Molecular fingerprinting of strains of Yersiniaruckeri serovar O1 and Photobacterium damsela subsp. piscicida isolated inItaly. Veterinary Microbiology 76:273–281.

Misra, C. K., B. K. Das, S. C. Mukherjee, and P. Pattnaik. 2006. Effect of longterm administration of dietary β-glucan on immunity, growth and survival ofLabeo rohita fingerlings. Aquaculture 255:82–94.

Newman, K. E. 2001. Effect of mannan oligosaccharide on the microflora andimmunoglobulin status of sows and piglet performance. Journal of AnimalScience 79:189.

Ofek, I., J. Goldhar, Y. Keisari, and N. Sharon. 1995. Nonopsonic phagocytosisof microorganisms. Annual Review of Microbiology 49:239–276.

Ourth, D., and T. Chung. 2004. Purification of antimicrobial factor from granulesof Channel Catfish peripheral blood leucocytes. Biochemical and BiophysicalResearch Communications 313:28–36.

Oyofo, B. A., J. R. DeLoach, D. E. Corrier, J. O. Norman, R. L. Ziprin, and H.H. Mollenhauer. 1989. Prevention of Salmonella typhimurium colonizationof broilers with D-mannose. Poultry Science 68:1357–1360.

Pryor, G. S., J. B. Royes, F. A. Chapman, and R. D. Miles. 2003. Man-nan oligosaccharides in fish nutrition: effects of dietary supplementation ongrowth and GI villi structure in Gulf of Mexico Sturgeon. North AmericanJournal of Aquaculture 65:106–111.

Qiang, X., C. YongLie, and W. QianBing. 2009. Health benefit application offunctional oligosaccharides. Carbohydrate Polymers 77:435–441.

Raa, J. 1996. The use of immunostimulatory substances in fish and shellfishfarming. Reviews in Fisheries Sciences 4:229–288.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:41

20

July

201

5

30 SELIM AND REDA

Rajaraman, V., B. J. Nonnecke, S. T. Franklin, D. C. Hammell, and R. L.Horst. 1998. Effect of vitamins A and E on nitric oxide production by bloodmononuclear leukocytes from neonatal calves fed milk replacer. Journal ofDairy Science 81:3278–3285.

Razeghi, M. M., R. Akrami, S. H. Ghobadi, K. A. Denji, N. Ezatrahimi, and A.Gharaei. 2012. Effect of dietary mannan oligosaccharide (MOS) on growthperformance, survival, body composition, and some hematological parame-ters in Giant Sturgeon juvenile (Huso huso Linnaeus 1754). Fish Physiologyand Biochemistry 38:829–835.

Rebl, A., T. Goldammer, and H. Seyfert. 2009. Toll-like receptor signaling inbony fish. Veterinary Immunology and Immunopathology 134(3–4):139–150.

Refstie, S., G. Baeverfjord, R. R. Seim, and O. Elvebø. 2010. Effects of dietaryyeast cell wall β-glucans and MOS on performance, gut health, and salmonlice resistance in Atlantic Salmon (Salmo salar) fed sunflower and soybeanmeal. Aquaculture 305:109–116.

Roberts, M. S. 1983. A report of an epizootic in hatchery reared Rainbow Trout,Salmo gairdneri richardson at an English trout farm, caused by Yersiniaruckeri. Journal of Fish Disease 6:551–552.

Robertsen, B., G. Rorstad, R. Engstad, and J. Raa. 1990. Enhancement of non-specific resistance in Atlantic Salmon, Salmo salar L., by a glucan fromSaccharomyces cerevisiae cell walls. Journal of Fish Disease 13:391–400.

Romero, J., G. Feijoo, and P. Navarrete. 2012. Antibiotics in aquaculture–use,abuse and alternatives. Pages 159–198 in E. D. Carvalho, G. S. David, andR. J. Silva, editors. Health and environment in aquaculture. InTech, Rijeka,Croatia.

Rurangwa, E., J. L. Laranja, R. Van Houdt, Y. Delaedt, Z. Geraylou, T. Van deWiele, J. Van Loo, V. Van Craeyveld, C. M. Courtin, J. A. Delcour, and F.Ollevier. 2009. Selected nondigestible carbohydrates and prebiotics supportthe growth of probiotic fish bacteria mono-cultures in vitro. Journal of AppliedMicrobiology 106:932–940.

Sado, R. J., A. J. D. A. Bicudo, and J. E. P. Cyrno. 2008. Feeding dietary man-nan oligosaccharide to juvenile Nile Tilapia (Oreochromis niloticus) has noeffect on hematological parameters and showed decreased feed consumption.Journal of the World Aquaculture Society 39:821–826.

Sahoo, P. K., and S. C. Murkherjee. 2002. The effect of dietary immunomodu-lation upon Edwardsiella tarda vaccination in healthy and immunocompro-mised Indian Major Carp (Labeo rohita). Fish and Shellfish Immunology12:1–16.

Sakai, M. 1999. Current research status of fish immunostimulants. Aquaculture172:63–92.

Samanya, M., and K. Yamauchi. 2002. Histological alterations of intestinal villiin chickens fed dried Bacillus subtilis var. natto. Comparative Biochemistryand Physiology Part A Molecular and Integrative Physiology 133:95–104.

Sang, H. M., and R. Fotedar. 2010. Effects of dietary beta-1,3-glucan on thegrowth, survival, physiological and immune response of marron, Cheraxtenuimanus (Smith 1912). Fish and Shellfish Immunology 28:957–960.

Selim, K. M., H. El-Hofy, and R. Khalil. 2014. The efficacy of three mycotoxinadsorbents to alleviate aflatoxin B1-induced toxicity in Oreochromis niloticus.Aquaculture International 22:523–540.

Shelby, R. A., C. Lim, M. Yildirim-Aksoy, T. L. Welker, and P. H. Klesius.2009. Effects of yeast oligosaccharide diet supplements on growth and diseaseresistance in juvenile Nile Tilapia, Oreochromis niloticus. Journal of AppliedAquaculture 21:1–11.

Siwicki, A. K., K. Kazun, E. Gtbski, E. Terech-Majewska, P. Baranowski,and S. Trapkowska. 2004. The effect of beta-1.3/1.6-glucan in diets on theeffectiveness of anti-Yersinia ruckeri vaccine – an experimental study inRainbow Trout (Oncorhynchus mykiss). Polish Journal of Food and NutritionSciences 13/54(2):59–61.

Stafford, J. L., K. K. Ellestad, K. E. Magor, M. Belosevic, and B. G. Magor.2003. A toll-like receptor (TLR) gene that is up-regulated in activated Goldfishmacrophages. Developmental and Comparative Immunology 27:685–698.

Staykov, Y., P. Spring, S. Denev, and J. Sweetman. 2007. Effect of a mannanoligosaccharide on the growth performance and immune status of RainbowTrout (Oncorhynchus mykiss). Aquaculture International 15:153–161.

Stuart, J. S., and S. O. Hung. 1989. Growth of juvenile White Sturgeon(Acipenser transmontanus) fed different proteins. Aquaculture 76:303–316.

Szilagyi, A. 2002. Lactose—a potential prebiotic. Alimentary Pharmacologyand Therapeutics 16:1591–1602.

Ta’ati, R., M. Soltani, M. Bahmani, and A. A. Zamini. 2011. Effects of theprebiotics immunoster and immunowall on growth performance of juvenileBeluga (Huso huso). Journal of Applied Ichthyology 27:796–798.

Tebbit, G., J. D. Erickson, and R. B. Water. 1981. Development and use ofYersinia ruckeri bacterins to control enteric redmouth disease. Developmentsin Biological Standardization 49:395–401.

Teuber, M. 2001. Veterinary use and antibiotic resistance. Current Opinion inMicrobiology 4:493–499.

Tobback, E., A. Decostere, K. Hermans, F. Haesebrouck, and K. Chiers. 2007.Yersinia ruckeri infections in salmonid fish. Journal of Fish Disease 30:257–268.

Torrecillas, S., A. Makol, M. J. Caballero, D. Montero, L. Robaina, F. Real, andJ. Sweetman. 2007. Immune stimulation and improved infection resistancein European Sea Bass (Dicentrarchus labrax) fed mannan oligosaccharides.Fish and Shellfish Immunology 23:969–981.

Verlhac, V., J. Gabaudan, A. Obach, W. Schuep, and R. Hole. 1996. Influence ofdietary glucan and vitamin C on non-specific and specific immune responsesof Rainbow Trout (Oncorhynchus mykiss). Aquaculture 43:123–133.

Wang, W. S., S. W. Hung, Y. H. Lin, C. Y. Tu, M. L. Wong, S. H. Chiou, and M. T.Shieh. 2007. The effects of five different glycans on innate immune responsesby phagocytes of hybrid tilapia and Japanese Eels Anguilla japonica. Journalof Aquatic Animal Health 19:49–59.

Welker, T. L., C. Lim, M. Aksoy, and P. H. Klesius. 2012. Use of diet crossover todetermine the effects of β-glucan supplementation on immunity and growthof Nile Tilapia, Oreochromis niloticus. Journal of the World AquacultureSociety 43:335–348.

Welker, T. L., C. Lim, M. Yildirim-Aksoy, R. Shelby, and P. H. Klesius. 2007.Immune response and resistance to stress and Edwardsiella ictaluri challengein Channel Catfish, Ictalurus punctatus, fed diets containing commercialwhole-cell yeast or yeast subcomponents. Journal of the World AquacultureSociety 38:24–35.

Whittington, R., C. Lim, and P. H. Klesius. 2005. Effect of dietary β-glucanlevels on the growth response and efficacy of Streptococcus iniae vaccine inNile Tilapia, Oreochromis niloticus. Aquaculture 248:217–225.

Wu, Z., Y. Yu, X. Chen, H. Liu, J. Yuan, Y. Shi, and X. Chen. 2014. Effect ofprebiotic konjac mannanoligosaccharide on growth performances, intestinalmicroflora, and digestive enzyme activities in Yellow Catfish, Pelteobagrusfulvidraco. Fish Physiology and Biochemistry 40:763–771.

Yanbo, W., and X. Zirong. 2006. Effect of probiotic for Common Carp (Cypri-nus carpio) based on growth performance and digestive enzymes activities.Animal Feed Science and Technology 127:283–292.

Ye, J. D., K. Wang, F. D. Li, and Y. Z. Sun. 2011. Single or combined effects offructo- and mannan oligosaccharide supplements and Bacillus clausii on thegrowth, feed utilization, body composition, digestive enzyme activity, innateimmune response and lipid metabolism of the Japanese Flounder Paralichthysolivaceus. Journal of Aquaculture Nutrition 17:902–911.

Yilmaz, E., M. A. Genc, and E. Genc. 2007. Effects of dietary mannan oligosac-charides on growth, body composition, and intestine and liver histologyof Rainbow Trout, Oncorhynchus mykiss. Israeli Journal of AquacultureBamidgeh 59:182–188.

Yousefian, M., and M. S. Amiri. 2009. A review of the use of prebioticin aquaculture for fish and shrimp. African Journal of Biotechnology 8:7313–7318.

Zhou, Q. C., J. A. Buentello, and D. M. Gatlin III. 2010. Effects of dietary pre-biotics on growth performance, immune response and intestinal morphologyof Red Drum (Sciaenops ocellatus). Aquaculture 309:253–257.

Zhu, H., H. Liu, J. Yan, R. Wang, and L. Liu. 2012. Effect of yeast polysac-charide on some hematologic parameter and gut morphology in ChannelCatfish (Ictalurus punctatus). Fish Physiology and Biochemistry 38:1441–1447.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:41

20

July

201

5


Click for updates


Effects of Sodium Chloride and Long-Term, Low-Concentration Exposures to Hydrogen Peroxide on NewZealand Mud SnailsRandall W. Oplingera & Eric J. Wagnera

a Utah Division of Wildlife Resources, Fisheries Experiment Station, 1465 West 200 North,Logan, Utah 84332, USAPublished online: 24 Nov 2014.

To cite this article: Randall W. Oplinger & Eric J. Wagner (2015) Effects of Sodium Chloride and Long-Term, Low-Concentration Exposures to Hydrogen Peroxide on New Zealand Mud Snails, North American Journal of Aquaculture, 77:1,31-36, DOI: 10.1080/15222055.2014.951810








http://dx.doi.org/10.1080/15222055.2014.951810




TECHNICAL NOTE

Effects of Sodium Chloride and Long-Term,Low-Concentration Exposures to Hydrogen Peroxideon New Zealand Mud Snails

Randall W. Oplinger* and Eric J. WagnerUtah Division of Wildlife Resources, Fisheries Experiment Station, 1465 West 200 North,Logan, Utah 84332, USA

AbstractNew Zealand mud snails Potamopyrgus antipodarum (NZMS)

are an invasive species and keeping NZMS out of hatcheries isimportant because the snail can be spread through the stocking offish. We conducted a series of experiments to evaluate the effective-ness of using hydrogen peroxide to kill NZMS and sodium chlorideto force NZMS to withdraw into their shells. The lowest concen-trations of hydrogen peroxide that produced 100% mortality ofNZMS were 750 mg/L for a 24-h exposure and and 75 mg/L for a96-h exposure. We found that NZMS treated with sodium chloridewithdrew into their shell and released from the substrate. Treat-ments with sodium chloride (717 mg/L) caused 70% of NZMS towithdraw into their shells and release from the substrate within thefirst minute of exposure, which could be beneficial in flushing themfrom a hatchery system provided there is sufficient discharge.

Aquatic invasive species (AIS) can have significant, adverseeffects on species conservation (Rahel et al. 2008), ecosystemecology (Olden et al. 2004), and human economic interests(Pimentel et al. 2005; Horsch and Lewis 2009). Traditionally,the unintentional spread of AIS has been linked to ballast wa-ter, anglers, and boaters (Johnson et al. 2001; Ricciardi 2006;Rothlisberger et al. 2010). Aquaculture activities have also beenassociated with the spread of AIS species (e.g., bighead carpHypophthalmichthys nobilis and silver carp H. molitrix intothe Mississippi River drainage; Freeze and Henderson 1982).It is generally accepted that AIS species can propagate in fishhatcheries and be spread through stocking or escapement ifproper hazard analysis and critical point (HAACP) proceduresare not in place.

The New Zealand mud snail Potamopyrgus antipodarum(NZMS), native to New Zealand, was first found in the UnitedStates in 1987 (Bowler 1991) and has since spread through-

*Corresponding author: [email protected] October 15, 2013; accepted July 6, 2014

out western North America (Kerans et al. 2005). Ecologically,NZMS are problematic because they can outcompete with nativeinvertebrates (Hall et al. 2003). Several fish hatcheries in west-ern North America have NZMS. When hatchery rainbow troutOncorhynchus mykiss consume NZMS, the consumed snails cansurvive digestion (Oplinger et al. 2009). Thus fish can spreadNZMS by consuming snails prior to being loaded on a haul-ing truck and then by excreting live NZMS after being stocked.Analyses of stomach samples from a hatchery with NZMS sup-port estimates that one live NZMS could be excreted into thewild for every 25,000 fish stocked (R. W. Oplinger, unpublisheddata).

It is imperative that hatcheries identify whether NZMS arepresent and take steps to prevent the inadvertent movement ofthe species. Snails are seldom eradicated from hatcheries exceptwithin defined areas, and such efforts are often hampered bythe costs of control and eradication. The elimination of NZMSoften requires facility reconstruction. Research has demon-strated that several chemicals, including household ammonia(full strength); benzethonium chloride (1,940 ppm); CloroxCommercial Solutions 409 Cleaner, Degreaser, and Disinfec-tant (full strength; active ingredient 0.3% alkyl dimethyl benzylammonium chloride); copper sulfate (504 ppm free copper con-centration); Pinesol (50% dilution), formalin (60,000 mg/L);potassium permanganate (10,000 mg/L); and hydrogen perox-ide (7,500 mg/L), are effective against NZMS with 5–15 minof exposure (Watton and Hawkes 1984; Hosea and Finlayson2005; Oplinger and Wagner 2009). Research has also demon-strated that NZMS withdraw into their shell when exposed tosudden increases in sodium chloride concentration (Oplingeret al. 2009). While not lethal, sodium chloride treatments couldprovide a simple, low-cost method that facilitates flushing snailsout of raceways and pipes. Despite this research, limitations

31

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:41

20

July

201

5

32 OPLINGER AND WAGNER

exist. First, it may be difficult to sustain high-concentrationchemical treatments in a flow-through aquaculture system, andchemical discharges may exceed what is permitted by the U.S.Environmental Protection Agency (EPA).

Because of these limitations, longer term, exposures to lowerconcentrations of chemicals may provide a more practical meansfor controlling NZMS at a hatchery. Also, the sodium chlo-ride exposure requirements (application duration) required foreliciting a shell withdrawal response are not known. In thismanuscript, we determined the toxicity of hydrogen peroxideto NZMS in both 24- and 96-h exposures to NZMS. Also, weconducted two experiments that evaluated the effectiveness ofdifferent sodium chloride application regimens on withdrawalbehavior. The benefit of forcing withdrawal is that snails thathave withdrawn are not attached to the substrate and can beswept downstream. The intention of these experiments was todevise an inexpensive and simple method for reducing the num-ber of NZMS within a hatchery.

METHODSThe NZMS used for all experiments were from a population

housed at the Fisheries Experiment Station (FES; Utah Divi-sion of Wildlife Resources) in Logan, Utah. All experimentswere performed using FES well water (pH = 7.6, hardness =175 mg/L, conductivity = 375 µS/cm) at temperatures rangingbetween 12◦C and 16◦C.

Hydrogen peroxide experiment.—We determined the acutetoxicity of hydrogen peroxide to NZMS using 24- and 96-hexposures. For these trials, groups of 20 NZMS were placedin 400-mL beakers containing 300 mL of water. The beakerswere darkened by covering with black plastic sheeting. Hydro-gen peroxide (Western Chemical, Ferndale, Washington) wasadded to the beakers in sufficient quantity to produce the de-sired static treatment concentrations. For the 24-h trial, ac-tive ingredient concentrations of 0, 10, 25, 50, 100, 200, 500,750, and 1,000 mg/L were tested; 0, 10, 25, 50, 75, 100, and200 mg/L concentrations were tested for the 96-h trial. Fivereplicate beakers, each containing 20 NZMS were exposed toeach combination of concentration and exposure duration (10replicate beakers were tested at 0 mg/L for both the 24- and96-h trials). At the end of each trial, the contents of the beakerwere poured through a 100-µm sieve and rinsed three times withfresh hatchery water. The contents of the sieve were added backto the beaker and the beaker was refilled. After 96 h of recovery,the number of live snails (considered alive if they responded totactile stimulation) in each beaker was counted. Neonate sur-vival and production was not measured. The survival data wereanalyzed by estimating LC50 values with a probit analysis. Theanalysis was performed using the PROC PROBIT procedure inSAS (SAS Institute 1998).

Sodium chloride experiment.—The purpose of this trial wasto determine whether sodium chloride application leads to in-creased numbers of NZMS entering the drift. Two regimens of

sodium chloride (Western Sun Solar Salt, Salt Lake City, Utah,label states >99.6% sodium chloride) treatment were tested:2 min of continuous salt application (717 mg/L after dilutioninto artificial stream), and 2 min of pulsed salt application (saltapplied for first 30 s, then no salt applied in seconds 31–60, thensalt applied in seconds 61–90, and no salt applied for last 30 s).The effects of these treatments were compared with that of acontrol (2 min of no sodium chloride application). The experi-ment was conducted in an artificial stream that was constructedusing a plastic housing gutter (80.0 cm long × 11.5 cm wide).A 3.0-cm-tall baffle was positioned 20.0 cm downstream of thehead of the stream. The stream was positioned over a large tankfilled with water and a pump provided water from that tank tothe stream. Water flow was set at 0.15 m/s (controlled using avalve), and effluent from the stream flowed back into the largetank, creating a recycle system.

One hundred NZMS were added to the artificial stream andwere given 5 min to acclimate. The flow in the stream wasslowly increased over a period of 10 min until a discharge of2.3 L/s was reached. Salt was applied and the number of snailsthat entered the drift was counted for 2 min. In the controltreatment, the 2-min counting period began immediately afterthe target water velocity was reached. For tracking purposes, wecounted continuously for 2 min the snails that were dislodged,dividing this counting period into twelve 10-s intervals. A totalof five replicate groups of 100 NZMS were exposed to eachtreatment. Data were reported as the percentage of snails and thecumulative percentage of snails in the stream that released fromthe substrate during each 10-s counting increment. Data wereanalyzed as a two-way analysis of variance (ANOVA) usingR Statistical Software (Hornik 2013). The dependent variablewas the percentage or cumulative percentage of NZMS thatdislodged, and the two independent variables were salt treatmentand counting interval. A Box–Cox transformation (Box and Cox1964) was used to force the data to meet the assumptions ofthe analysis. Results were considered significant at P < 0.05.Significant differences were evaluated using Tukey’s HonestlySignificant Difference Test (Kuehl 2000).

Sodium chloride hatchery discharge trial.—This trial tookplace in the effluent of the FES Hatchery. NZMS were unex-pectedly found in the effluent and the Utah Division of WildlifeResources responded by drying the effluent 4 d after discovery.This trial was quickly planned and executed to take advantageof this narrow window of opportunity and no replicates couldbe performed. The density of NZMS in the discharge substratewas <100/m2. The effluent is approximately 180 m long ×3.0 m wide × 0.33 m deep (average depth) with a substrateof sand, gravel, and small cobbles. Periphyton and macrophytesoccupy approximately 25% of the substrate, the remainder be-ing bare. The average water velocity was approximately 0.4 m/s.Drift nets were constructed from a vinyl material with 1.0 mmmesh and installed by fitting into slots in three existing con-crete structures, thus subdividing the stream into three sections.The upstream section was approximately 40 m long, the middle

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:41

20

July

201

5

TECHNICAL NOTE 33

section was approximately 55 m long, and the downstream sec-tion was approximately 85 m long.

Sodium chloride was applied by adding 227 kg of salt directlyto the stream. The sodium chloride dissolved over a period ofapproximately 5 min. The concentration of sodium chloridein the stream was not measured but should have exceeded the717 mg/L used in the laboratory experiments (average concen-tration would be ∼1,900 mg/L if all the salt dissolved at samerate over the 5-min dissolution period). The nets were installed1 h prior to salt treatment then removed to determine the baselinerate of NZMS drift. The nets were reinstalled, salt was applied,and 1 h later, the number of NZMS in the nets was again deter-mined. The effectiveness of the salt treatment was assessed bycomparing the number of NZMS caught in each net both beforeand after salt application.

Other possible irritants.—Following the same treatment de-sign as in the sodium chloride experiment, we determinedwhether NZMS release from substrate when subjected to 2 minof continuous exposure to hydrogen peroxide (1,500 mg/L afterdilution into stream), hot sauce (Tabasco brand, 500 mg/L afterdilution), sand (10 g/L after dilution), and silt (10 g/L after di-lution). A single replicate group of 100 NZMS was exposed toeach irritant.

RESULTS

Hydrogen Peroxide ExperimentThe estimated 24-h LC50 value was 37.5 mg/L (95% confi-

dence limits: 21.0–61.4 mg/L), and no NZMS survived treat-ment at concentrations of 750 mg/L or more (Table 1). Inthe 96-h trial, no NZMS survived treatment at concentrationsof 75 mg/L or more (Table 1), and the estimated LC50 was11.0 mg/L (95% confidence limits: 5.4–14.9 mg/L).

Sodium Chloride ExperimentThe application of sodium chloride led to large numbers of

NZMS releasing from the substrate (Figure 1). Depending onthe treatment, between 60% and 67% of the snails releasedwithin the first 30 s after salt application. Approximately 70%of NZMS released from the substrate after 1 min of applica-tion. The proportion of snails that released from the substrate isasymptotic and peaks at around 75% (Figure 1).

Significant differences in the proportion and the cumulativeproportion of NZMS that released were observed both amongtreatments and time intervals, and a significant treatment ×time interaction was observed (all P < 0.01). For both responsevariables, a significantly greater proportion of NZMS releasedwhen salt was applied than in the control (both P < 0.01). Nosignificant difference in NZMS release was observed, however,between the two salt application methods (Figure 1, top panel;both P ≥ 0.56). No differences in release were observed throughthe first 50 s (all P ≥ 0.58). After the first 50 s, significant reduc-tions in the proportion of NZMS that became dislodged wereobserved (compared with that in the first 50 s, all P ≤ 0.04).

TABLE 1. Percent survival and standard deviation (SD) of New Zealand mudsnails exposed to various concentrations of hydrogen peroxide for either 24 or96 h. Five replicate groups of 20 snails were subjected to each combinations ofexposure time × hydrogen peroxide concentration (10 replicates were testedin 0 mg/L trials).

Exposure Concentration Survival (%)time (h) (mg/L) mean SD

24 0 90.5 25.010 94 10.725 92 9.250 14 10.7

100 4.2 4.2200 1 2.2500 1 2.2750 0 0

1,000 0 096 0 94 15.2

10 95 8.725 25 34.050 13 10.375 0 0

100 0 0200 0 0

The largest cumulative increase in dislodgement (Figure 1, bot-tom panel) occurred between the first two 10-s time intervals(4.1% ± 12.1% increase, Q2,11 = 4.73, P < 0.01). The cumu-lative proportion of snails that released became asymptotic after40 s and no significant increases in dislodgement were observedafterwards (all P ≥ 0.21).

Sodium Chloride Hatchery Discharge TrialDuring the observation period prior to the salt application,

one NZMS was found in both the middle and downstream nets.No NZMS were found in the upstream-most net. One hour aftersalt treatment, one snail was found in the downstream net andno snails were found in the other two nets.

Other Possible IrritantsWe found that NZMS did not release from the substrate when

exposed to hydrogen peroxide, hot sauce, sand, or silt. We didnot test higher concentrations of these irritants because froma practical sense, it would be costly or not feasible to provide2 min of continuous exposure to higher concentrations of theseirritants.

DISCUSSIONThe results from these studies show that 24-h exposures to

hydrogen peroxide concentrations of at least 750 mg/L and 96-hexposures to concentrations of at least 75 mg/L killed all NZMS.The 96-h LC50 value that we determined for hydrogen perox-ide (11.0 mg/L) is similar to that of many invertebrate species,

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:41

20

July

201

5


Time Since Experiment Initiation (s)

10 20 30 40 50 60 70 80 90 100 110 120

Tot

al P

erce

nt R

elea

sed

-20

0

20

40

60

80

100

Per

cent

age

Rel

easi

ng fr

om S

ubst

rate

0

10

20

30

40

502 min Continuous SaltTwo, 30 s Pulses of SaltControl

FIGURE 1. Percent of New Zealand mud snails that release from the substrateand enter the drift during the first 2 min after the addition of a 3% sodiumchloride solution to an artificial stream (total concentration 717 mg/L afterdilution). Three treatment conditions were tested: (1) 2 min of continuous saltapplication, (2) two 30-s pulses of salt, and (3) control. Data wererecorded in10-s intervals and both the percentage of snails that released in each 10-s interval(top panel) and the cumulative percentage of snails that released (bottom panel)are shown. The discharge in the stream was 2.3 L/s and the sodium chloridesolution was applied at a rate of 0.055 L/s. Five replicate groups of 100 snailswere subjected to each treatment. Error bars represent 1 SD.

including Gammarus sp. (96-h LC50 = 4.42 mg/L; Kay et al.1982), Daphnia carinata (48-h LC50 = 5.6 mg/L; Reichwaldtet al. 2012), and Physa sp. (96-h LC50 = 17.7 mg/L; Kay et al.1982). Our results provide an alternative treatment regimen tothe 15-min treatment at 7,500 mg/L (Oplinger and Wagner 2009)and also kills 100% of NZMS. Long-term, low-concentrationdrips of hydrogen peroxide may be appealing because it couldbe logistically difficult to continuously apply hydrogen peroxideat a rate of 7,500 mg/L (e.g., effluent concerns or safety issuescaused by rapid dispersal of concentrated chemical). Despitethe fact that long-term exposures could be simpler to apply, theycould also be more costly. In a situation where water is flowingcontinuously, completely killing NZMS in a 15-min treatmentwill require 10.4% of the hydrogen peroxide of a 24-h expo-sure and 26.0% of the hydrogen peroxide of a 96-h exposure.Interestingly, a 96-h exposure requires 40.0% of the hydrogen

peroxide of a 24-h exposure and is thus more cost effective.Possibly the biological effects of hydrogen peroxide exposureaccumulate with time, which would allow for the use of dis-proportionally less hydrogen peroxide during the 96-h exposurethan the 24-h exposure. However, beyond hatchery logistics andsafety concerns, the concentration of hydrogen peroxide neededto kill NZMS in a 15-min exposure could exceed what is permit-ted by the EPA (acute toxicity benchmark = 0.7 mg/L; Schmidtet al. 2007). The Food and Drug Administration limits the usehydrogen peroxide (Perox-Aid; Eka Chemicals, Marietta, Geor-gia) on freshwater salmonids to a maximum of 100 mg/L for60 min. Thus, when removal of NZMS is desired, hydrogen per-oxide is best applied in raceways that do not contain fish. Statesand other countries may have even more stringent regulationson the use of hydrogen peroxide.

We also found that high concentrations of sodium chlo-ride can be used to force NZMS to withdraw into theirshell. In contrast to hydrogen peroxide, sodium chloride isnot highly lethal to most snails, including NZMS (Keffordand Nugegoda 2005; Oplinger et al. 2009). Concentrations ofsodium chloride required to elicit a behavioral response inNZMS are also nontoxic to trout (Suomalainen et al. 2005).Research on the snail Melanoides tuberculata found that ap-proximately 75% of individuals survived 24-h of exposure tosodium chloride concentrations of 50,000 mg/L (Mitchell et al.2007). While not tested on NZMS, it is possible that sodiumchloride treatment has nonlethal negative effects on the snail.For example, elevated salinity can decrease the growth and fe-cundity of Physa acuta (Kefford and Nugegoda 2005). Forcingsnails into their shell could be beneficial for hatcheries that haveNZMS because snails that have withdrawn into their shell arenot attached to the substrate and are easily swept away by flow.Thus sodium chloride treatments could be used to reduce thenumber of NZMS in raceways or pipes. Reducing the numberof NZMS could help reduce the costs and increase the effective-ness of other eradication methods such as chemical treatmentsand could decrease the likelihood that NZMS are moved to otherhabitats by other vectors such as by wildlife. Sodium chlorideis inexpensive and readily available and is thus ideal for use ina hatchery setting.

The response of NZMS to sodium chloride is almost imme-diate and the majority of NZMS that withdraw into their shellwhen sodium chloride is applied do so within the first 30 s af-ter application is initiated. Regardless, approximately 25% ofNZMS are resistant to sodium chloride treatment and our re-search shows there is little benefit to treating for periods longerthan 60 s. Unpublished laboratory tests show that the numberof snails that withdraw into their shell does not increase withtreatments up to 10 min, at concentrations up to 2,000 mg/L,and under various scenarios where “rest periods” are providedbetween intervals of sodium chloride application (Oplinger, un-published data). Thus, on the basis of these findings, we feelthat 717 mg/L sodium chloride for 60 s is sufficient for forcingapproximately 75% of NZMS into their shell and the use of

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:41

20

July

201

5

TECHNICAL NOTE 35

higher concentrations and longer exposure durations add littlebenefit. Lower concentrations may also be effective but have notbeen tested. We performed similar experiments with hydrogenperoxide, hot sauce, sand, and silt but did not observe a simi-lar withdrawal response. Thus the response we observed withsodium chloride may not be typical of the response of NZMSto most perturbations. Blandford and Little (1983) hypothe-sized that there are sensory organs in the tentacles of NZMSthat allow them to sense rapid changes in osmotic pressure.This may explain why we did not observe a similar responsewhen NZMS were exposed to chemicals other than sodiumchloride.

Some observations were made during our trials that maybe beneficial when designing sodium chloride treatments toreduce NZMS at fish hatcheries. Mainly, sodium chloride ap-parently works best when NZMS are exposed to an immediate,high concentration of NaCl versus exposing NZMS graduallyto increasing concentrations. In our experiments, we controlledthe application of a liquid sodium chloride solution using apump, creating an almost immediate increase in concentration.Some observational trials were performed where an appropriateamount of dry rock salt was added directly to the artificial streamto produce a sustained concentration of 717 mg/L during 60 sof dissolution. This slow release forced less than 10% of NZMSto withdraw into their shell (Oplinger, unpublished data). Theseresults were corroborated by our field trial, where we addeddry rock salt directly to the effluent stream and observed noincrease in the number of NZMS drifting downstream. The lowrelease rate of NZMS in the field trial may be been causedby the slow dissolution rate of the rock salt or by entrapmentof drifting NZMS by the sediment and vegetation. We think itprobable that sodium chloride treatments work best in hatcheryraceways and pipes where there is consistent flow and few ob-stacles slowing the downstream drift of NZMS. Oplinger et al.(2009) exposed NZMS to an 11,000 mg/L sodium chloride so-lution and the majority of the snails responded by withdrawinginto their shell; however, over 90% of the snails reemerged andwere active 5 min after being added to the salt solution. This isnot surprising, given that NZMS are known to colonize habi-tats with salinities ranging from fresh to saltwater (Alonso andCastro-Dıez 2008). As a result, the effect of sodium chloride onNZMS appears to be temporary, and repeated daily or weeklytreatments may be required to “push” NZMS downstream.

These results provide insight into two new methods forcontrolling NZMS. Our data show that long-term, low-concentration drips of hydrogen peroxide could potentially beused to eliminate NZMS from hatcheries. Sodium chloride treat-ments could be used as a tool to help reduce the numbers ofNZMS at a hatchery by coercing NZMS out of raceways andpipes. Sodium chloride could also potentially be used in the wildto slow the upstream movement of NZMS towards critical habi-tats until more permanent solutions are implemented. WhetherNZMS become acclimated to repeated sodium chloride expo-sures is not known and should be evaluated in future research.

Future work should also address how changes in water quality,temperature, and velocity influence the response of NZMS tohydrogen peroxide and sodium chloride.

ACKNOWLEDGMENTSFunding for this research was provided by the Federal Aid in

Sport Fish Restoration Program, project F-96-R, and the UtahDivision of Wildlife Resources. This manuscript was greatlyimproved by the comments from several anonymous reviewers.

REFERENCESAlonso, A., and P. Castro-Dıez. 2008. What explains the invading success of

the aquatic mud snail Potamopyrgus antipodarum (Hydrobiidae, Mollusca)?Hydrobiologia 614:107–116.

Blandford, P., and C. Little. 1983. Salinity detection by Hydrobia ulvae (Pen-nant) and Potamopyrgus jenkinsi Smith (Gastropoda:Prosobranchia). Journalof Experimental Marine Biology and Ecology 68:25–38.

Bowler, P. A. 1991. The rapid spread of the freshwater hydrobiid snail Pota-mopyrgus antipodarum (Gray) in the middle Snake River, southern Idaho.Proceedings of the Desert Fishes Council 21:173–182.

Box, G. E. P., and D. R. Cox. 1964. An analysis of transformations. Journal ofthe Royal Statistical Society Series B 26:211–243.

Freeze, M., and S. Henderson. 1982. Distribution and status of Bighead Carp andSilver Carp in Arkansas. North American Journal of Fisheries Management2:197–200.

Hall, R. O., J. L. Tank, and M. F. Dybdahl. 2003. Exotic snails dominate nitrogenand carbon cycling in a highly productive stream. Frontiers in Ecology andthe Environment 1:408–411.

Hornik, K. 2013. The R FAQ. Available: http://CRAN.R-project.org/doc/FAQ/R-FAQ.html. (October 2013).

Horsch, E. J., and D. J. Lewis. 2009. The effects of aquatic invasive species onproperty values: evidence from a quasi-experiment. Land Economics 85:391–409.

Hosea, R. C., and B. Finlayson. 2005. Controlling the spread of New Zealandmud snails on wading gear. California Fish and Game, Office of Spill Pre-vention and Response, Administrative Report 2005-02, Sacramento.

Johnson, L. E., A. Ricciardi, and J. T. Carlton. 2001. Overland dispersal ofaquatic invasive species: a risk assessment of transient recreational boating.Ecological Applications 11:1789–1799.

Kay, S. H., P. C. Quimby Jr., and J. D. Ouzts. 1982. Hydrogen peroxide: apotential algicide for aquaculture. Proceedings of the Southern Weed ScienceSociety 35:275–289.

Kefford, B. J., and D. Nugegoda. 2005. No evidence for a critical salinitythreshold for growth and reproduction in the freshwater snail Physa acuta.Environmental Pollution 134: 377–383.

Kerans, B. L., M. F. Dybdahl, M. M. Gangloff, and J. E. Jan-not. 2005. Potamopyrgus antipodarum: distribution, density, and ef-fects on native macroinvertebrate assemblages in the greater Yellowstoneecosystem. Journal of the North American Benthological Society 24:123–138.

Kuehl, R. O. 2000. Design of experiments: statistical principles of researchdesign and analysis, 2nd edition. Brooks/Cole Publishing, Pacific Grove,California.

Mitchell, A. J., M. S. Hobbs, and T. M. Brandt. 2007. The effect ofchemical treatments on red-rim melania Melanoides tuberculata, an ex-otic aquatic snail that serves as a vector of trematodes to fish and otherspecies in the USA. North American Journal of Fisheries Management 27:1287–1293.

Olden, J. D., N. L. Poff, M. R. Douglas, M. E. Douglas, and K. D. Fausch. 2004.Ecological and evolutionary consequences of biotic homogenization. Trendsin Ecology and Evolution 19:18–24.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:41

20

July

201

5


Oplinger, R., P. Brown, and E. Wagner. 2009. Effect of sodium chloride,tricaine methanesulfonate, and light on New Zealand mud snail behavior,survival of snails defecated from Rainbow Trout, and effects of epsomsalt on snail elimination rate. North American Journal of Aquaculture 71:157–164.

Oplinger, R., and E. Wagner. 2009. Toxicity of common aquaculture disinfec-tants to New Zealand mud snails and mud snail toxicants to Rainbow Trouteggs. North American Journal of Aquaculture 71:229–237.

Pimentel, D., R. Zuniga, and D. Morrison. 2005. Update on the economicand environmental costs associated with alien-invasive species in the UnitedStates. Ecological Economics 52:273–288.

Rahel, F. J., B. Bierwagen, and Y. Taniguchi. 2008. Managing aquatic speciesof conservation concern in the face of climate change and invasive species.Conservation Biology 22:551–561.

Reichwaldt, E. S., L. Zheng, D. J. Barrington, and A. Ghadouani. 2012. Acutetoxicological response of Daphnia and Moina to hydrogen peroxide. Journalof Environmental Engineering 138:607–611.

Ricciardi, A. 2006. Patterns of invasion in the Laurentian Great Lakes in relationto changes in vector activity. Diversity and Distributions 12:425–433.

Rothlisberger, J. D., W. L. Chadderton, J. McNulty, and D. M. Lodge. 2010.Aquatic invasive species transport via trailered boats: what is being moved,who is moving it, and what can be done. Fisheries 35:121–132.

SAS Institute. 1998. SAS/STAT user’s guide version 8.0, volume 1. SAS Insti-tute, Cary, North Carolina.

Schmidt, L. J., M. P. Gaikowski, and W. H. Gingerich. 2007. Environmental as-sessment for the use of hydrogen peroxide in aquaculture for treating externalfungal and bacterial diseases of cultured fish and fish eggs. U.S. GeologicalSurvey, Report 2007-5639, La Crosse, Wisconsin.

Suomalainen, L. R., M. Tiirola, and E. T. Valtonen. 2005. Treatment of colum-naris disease of Rainbow Trout: low pH and salt as possible tools? Diseasesof Aquatic Organisms 65:115–120.

Watton, A. J., and H. A. Hawkes. 1984. The acute toxicity of ammonia and cop-per to the gastropod Potamopyrgus jenkinsi (Smith). Environmental PollutionSeries A Ecological and Biological 36:17–29.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:41

20

July

201

5


Click for updates


Initial Characterization of Embryonic Development inNorth American BurbotJoshua P. Eganab, Ryan D. Johnsonc, Paul J. Andersde & Kenneth D. Cainfg

a Conservation Biology Graduate Program, University of Minnesota, 1987 Upper BufordCircle, St. Paul, Minnesota 55108, USAb Bell Museum of Natural History, University of Minnesota, St. Paul, Minnesota 55108, USAc Idaho Department of Fish and Game, 3101 South Powerline Road, Nampa 83686, Idaho, USAd Cramer Fish Sciences, 317 West 6th Street, Suite 204, Moscow, Idaho 83843, USAe Department of Fish and Wildlife Sciences, University of Idaho, Post Office Box 441136,Moscow, Idaho 83844-1136, USAf Department of Fish and Wildlife Sciences and Aquaculture Research Institute, University ofIdaho, 875 Perimeter Drive MS1136, Moscow, Idaho 83844-1136, USAg National Centre for Marine Conservation and Resource Sustainability, University ofTasmania, Locked Bag 1370, Launceston, Tasmania 7250, AustraliaPublished online: 15 Dec 2014.

To cite this article: Joshua P. Egan, Ryan D. Johnson, Paul J. Anders & Kenneth D. Cain (2015) Initial Characterizationof Embryonic Development in North American Burbot, North American Journal of Aquaculture, 77:1, 37-42, DOI:10.1080/15222055.2014.955156








http://dx.doi.org/10.1080/15222055.2014.955156




TECHNICAL NOTE

Initial Characterization of Embryonic Development in NorthAmerican Burbot

Joshua P. EganConservation Biology Graduate Program, University of Minnesota, 1987 Upper Buford Circle, St. Paul,Minnesota 55108, USA; and Bell Museum of Natural History, University of Minnesota, St. Paul,Minnesota 55108, USA

Ryan D. JohnsonIdaho Department of Fish and Game, 3101 South Powerline Road, Nampa, Idaho 83686, USA

Paul J. AndersCramer Fish Sciences, 317 West 6th Street, Suite 204, Moscow, Idaho 83843, USA;and Department of Fish and Wildlife Sciences, University of Idaho, Post Office Box 441136, Moscow,Idaho 83844-1136, USA

Kenneth D. Cain*Department of Fish and Wildlife Sciences and Aquaculture Research Institute, University of Idaho,875 Perimeter Drive MS1136, Moscow, Idaho 83844-1136, USA;and National Centre for Marine Conservation and Resource Sustainability, University of Tasmania,Locked Bag 1370, Launceston, Tasmania 7250, Australia

AbstractDeclining wild populations and increasing commercial inter-

est are encouraging development of aquaculture methods for theNorth American Burbot Lota lota maculosa. A current focus ofBurbot aquaculture research is reducing high mortality that istypically associated with culture of Burbot embryos and larvae.However, no Burbot embryonic or larval staging systems are cur-rently available to provide a comparative baseline for studies ofthese life stages. To help address this gap, we examined Burbot em-bryonic development from egg fertilization until the onset of thelarval period at first feeding. Ontogeny was characterized usingdiagnostic morphological features visible with stereo microscopy.Six developmental periods were characterized (cleavage, blastula,gastrula, segmentation, swim bladder inflation, and first feeding),along with 15 developmental stages. Water temperature rangedfrom 3◦C to 5◦C for the duration of this study. Cleavage cycleswere approximately 8 h/division. Segmentation began at 12 d post-fertilization (DPF) and continued until first hatch at 33 DPF. Firstfeeding was observed at 45 DPF, 12 d after first hatch. Results pre-sented here are expected to help refine Burbot culture methodologyand research efforts aimed at conservation, management, and com-mercial production. Future research should use additional cohorts

*Corresponding author: [email protected] May 21, 2014; accepted August 11, 2014

and expose embryos to a range of culture conditions to better un-derstand the factors governing embryonic survival and variabilityin the timing of events during development.

Burbot Lota lota are the only freshwater species in the codfamily Gadidae and exhibit a circumpolar distribution above40◦N (Stapanian et al. 2010). Phylogenetic investigation iden-tified two genetically distinct Burbot lineages that have beenclassified as subspecies: Eurasian Burbot L. l. lota, which rangesfrom Europe to the Great Slave Lake, Canada; and North Amer-ican Burbot L. l. maculosa, which are found south of the GreatSlave Lake in North America (Van Houdt et al. 2003). Despitethis broad distribution, many populations of both subspecies arecurrently experiencing depletion, imperilment, and extirpationacross major portions of their ranges (reviewed in Stapanianet al. 2010; Barron et al. 2012). Burbot population declineshave been attributed to a wide range of anthropogenic factors,including habitat loss, river impoundment and regulation, pollu-tion and water quality changes, invasive species, climate change,and overexploitation (Stapanian et al. 2010).

37

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:42

20

July

201

5

38 EGAN ET AL.

Following these population declines and increasing commer-cial aquaculture interests, aquaculture development has recentlybegun for both Burbot subspecies (Barron et al. 2013b). How-ever, these efforts are hindered by high mortality during theembryonic and larval life stages (Wolnicki and Kaminski 2001;Jensen et al. 2008a, 2008b, 2008c; Barron et al. 2012, 2013a,2013b). Reducing this mortality requires establishment of rear-ing protocols that account for differences in optimal environ-mental conditions through ontogeny, as has been done for morewidely cultured fish species. For example, embryonic and larvalmortality has been reduced during aquaculture of Atlantic CodGadus morhua by identifying water exchange rates, light inten-sities, and temperatures best suited to different periods duringearly development (reviewed in Brown et al. 2003). However,before life stage-specific rearing protocols can be developed fora species, its embryonic and larval development must be charac-terized to allow scientists and fish culturists to experiment uponand communicate about events during development (Hall et al.2004). To date, no studies have characterized Burbot embryonicand larval development. To help fill this gap, we examined em-bryonic development of North American Burbot from fertiliza-tion until the onset of the larval life history period at first feeding.

Ontogeny in fishes has been systematically characterized us-ing a standard developmental staging series for species of in-terest (e.g., Hall et al. 2004; Fujimura and Okada 2007). Adevelopmental staging series is a chronological set of develop-mental periods, generally demarcated by unique morphologicaltraits, used to characterize changes during the development oforganisms. This approach provides a standard framework thatallows comparison of ontogenetic development within and be-tween species (Fujimura and Okada 2007). In this paper, weprovide results of an initial developmental staging series forNorth American Burbot from fertilization to the onset of exoge-nous feeding.

METHODSBurbot embryological stages were determined using a stag-

ing system initially applied to a closely related species, At-lantic Cod, as reported by Hall et al. (2004). General life historyperiods were assigned using criteria reported by Balon et al.(1990). Ontogeny was characterized using diagnostic morpho-logical features visible with standard stereo microscopy. Em-bryos were incubated in filtered, UV-sterilized, and dechlori-nated municipal water. Water temperature was stable, with agradual increase from 3◦C to 5◦C over the 45-d study. Devel-opmental status of early life stages was reported using dayspostfertilization (DPF), days posthatch (DPH), and degree-dayspostfertilization (DDPF). One degree-day equals 1◦C above 0◦Cfor 24 h.

Gametes from two broodstock sources were used in thisstudy and were from captive-reared and wild-captured fish. Bothsources originated from Moyie Lake in southeastern BritishColumbia, Canada. Gametes were manually collected, fertil-

ized, and enumerated from captive broodstock according toJensen et al. (2008c). Although eggs from captive broodstockhad estimated mean fertilization rates of 85%, high embryo mor-tality (nearly 100%) by 8 DPF necessitated the use of additionalembryos. Additional gametes were subsequently collected andfertilized on the ice at Moyie Lake during the week of February12, 2012, following Neufeld et al. (2011). Fertilized eggs weretransported to the University of Idaho Aquaculture Research In-stitute (UI-ARI, Moscow, Idaho). These gametes were found tobe at the same developmental stage at 8 DPF as the captive ga-metes and were therefore used to characterize Burbot ontogenyfrom 9 DPF until the study was terminated, when Burbot beganfeeding exogenously. After fertilization, embryos were accli-mated and stocked into 1.2-L Imhoff cones (Aquatic Ecosys-tems, Apopka, Florida) modified to make upwelling incubatorsfollowing Jensen et al. (2008c). Flow rates in incubators keptembryos slowly moving at all times. A 24-h photoperiod wasmaintained for the duration of the study.

The first group of wild-collected fertilized eggs were ini-tially sampled at 0800, 1200, and 1600 hours to determinethe embryonic cleavage period. A cleavage period of approx-imately 8 h was observed, so embryos were then sampled ev-ery 8 h until the end of the cleavage period. After cleavagewas complete, embryos were sampled daily at 1400 hours.For each sampling event, 10 embryos were randomly pipettedfrom the incubators, euthanized with buffered MS-222 (tricainemethanesulfonate) at approximately 250 mg/L (Western Chem-ical), and photographed with an AmScope 10 MP stereo mi-croscope (Amscope, Irvine, California). Resulting photos wereedited for clarity and focus using Adobe PhotoShop Elements10 (www.Adobe.com). Any unintentionally sampled dead em-bryos were discarded and randomly replaced with viable em-bryos. Variation in the developmental stages observed at eachsampling event is reported. As in other embryological studies(e.g., Kimmel et al. 1995), later stage embryos were manuallyremoved from the chorion (between 26 and 32 DPF) to bettervisualize development.

RESULTSOur preliminary developmental staging system for North

American Burbot divides the embryogenesis into six distinctperiods (cleavage, blastula formation, gastrula, segmentation,swim bladder inflation, and first feeding) and 15 embryonicstages (Table 1). Each stage is identified by a unique set of diag-nostic morphological features. Resulting developmental periodsand embryonic stages are described below, summarized in Ta-ble 1, and illustrated in Figures 1 and 2.

Cleavage PeriodAt 8 h postfertilization, eggs entered the one-cell stage (10/10

sampled embryos at one-cell stage; Figure 1A). At this stage,the second cell division was beginning to bisect the first cellalong the animal pole, forming two blastomeres of equal size.Second cleavage was characterized by distinct cell formation

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:42

20

July

201

5

TECHNICAL NOTE 39

TABLE 1. Timing, duration, and characteristics of embryonic development of North American Burbot reared at 3–5◦C, including DPF and DDPF. Observationsuntil 8 DPF and from 9 DPF until the termination of this study originated from captive and wild broodstock, respectively. YSL = yolk syncytial layer; GI =gastrointestinal.

Period Stage DPF DDPF Characteristics

Cleavage 1-cell 0 1 Initial cell formation16-cell 1 4 Cells become increasingly irregular in sizeCap 2 8 Cells exhibit grainy, caplike appearance on yolk

Blastula YSL formation 3 12 Observed as a clear layer between cap and yolkSphere 4 16 Cell mass depresses yolkBlastodisc elongation 5 17 Yolk is no longer depressed, cell mass elongates and

resides on top of yolkGastrula Germ ring 5 18 Marks beginning of epiboly

25% epiboly 5 20 Coverage of yolk via cell mass begins50% epiboly 6 24 Embryo begins to develop axes

Segmentation Shield 13 52 Medial axis swells, creating shieldlike appearanceEyespot 15 60 Observed as solid masses of bulging cells on either side

of neural plate (no pigment yet)Hatch 33 136 Embryos begin to display motion and are free from

chorionGolden eye 33 136 Eyes exhibit golden color under light

Swim bladder inflation Free embryo 40 183 Observed as a clear sac directly above GI tractFirst feeding Larva 45 229 Distinguished by food in GI tract

perpendicular to the first, doubling the number of cells to four.By the third cleavage (1 DPF), embryos entered the 16-cellstage (8/10 sampled embryos were at 16-cell stage and 2/10were at 8-cell stage) and individual blastomeres exhibited con-siderable size variation (Figure 1B). Cell size decreased withadditional cleavage and became increasingly variable amongembryos. By 2 DPF, embryos entered the 128-cell stage (10/10sampled embryos at 128-cell stage). Cells exhibited a grainycaplike appearance, with several blastomeres becoming clearlyvisible (Figures 1C).

Blastula PeriodAt 3 DPF, a solid ball of cells (blastula) formed, consisting

of ∼ 500 cells (Figure 1D). At this point, no synchrony in celldivision was observed and the embryos entered a midblastulatransition (MBT) phase. The peripheral blastomeres and thoseadjacent to the yolk began to fuse, creating the YSL at ∼ 3DPF (10/10 sampled embryos at YSL formation stage). Thislayer was difficult to visualize from the side but was clearlyrecognized when viewed from below (Figure 1D). Continuedcell division appeared to exert pressure on the yolk as the em-bryos entered the sphere stage (7/10 sampled embryos at spherestage and 3/10 at YSL formation stage) at 4 DPF (Figure 1E).Pressure eventually subsided as the blastodisc began to elongateand reposition itself atop the yolk by about 5 DPF.

Gastrula Formation PeriodThe gastrula period was initially marked by the appear-

ance of a germ ring (8/10 embryos at germ ring stage, 1/10

at sphere stage, and 1/10 at nearly 25% epiboly) at approxi-mately 5 DPF (Figure 1F). The germ ring was clearly visiblewhen viewed from beneath but could only be partially observedin side view. These developments were associated with earlystages of epiboly (coverage of the yolk when viewed fromthe side). At approximately 25% epiboly (∼5 DPF; 8/10 em-bryos at 25% epiboly and 2/10 at germ ring stage), a cav-ity was noticeable beneath the blastodisc (Figure 1G). Hallet al. (2004) identified a similar cavity in Atlantic Cod em-bryos and proposed that it may be yolk filled. By about 50%epiboly (∼6 DPF; 10/10 embryos at 50% epiboly), a triangu-lar mass of cells was apparent (Figure 1H) and at this stageembryos began to exhibit dorsoventral, anterior, and posterioraxes, and the point of the triangular mass of cells formed theanterior axis of the embryo. Following the completion of epi-boly, cells in the epiblast gave rise to the ectoderm and cellsof the hypoblast gave rise to the mesoderm and the endoderm(∼7–12 DPF).

Segmentation PeriodOrganogenesis began midway through epiboly, with devel-

opment of a notochord being the first noticeable change at thisstage of development (∼13 DPF). Neuralation followed organo-genesis, appearing as a swelling of the medial axis of the em-bryonic shield (10/10 sampled embryos at shield stage; Figure1I). Optic primordial cells (eye spots) and brain developmentwere first noticeable at 15 DPH, along with the initial braindevelopment (Figure 1J, K).

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:42

20

July

201

5

40 EGAN ET AL.

FIGURE 1. Embryonic developmental stages of the North American Burbot. (A) 1-cell stage (8 h postfertilization); (B) 16-cell stage (1 DPF); (C) “cap” created∼128-cell stage (2 DPF); (D) YSL development (3 DPF); (E) sphere stage (4 DPF); (F) onset of germ ring (5 DPF); (G) asymmetric yolk sac formation at ∼ 25%epiboly shown between arrows; (H) 50% epiboly showing formation of triangular cell mass; (I) neuralation and notochord development (14 DPF) shown betweenarrows; (J) eye spot development (15 DPF) at arrow; (K) beginnings of brain formation seen between optic primordial cells (16 DPF). Observations until 8 DPFand from 9 DPF until the termination of this study originated from captive and wild broodstock, respectively. Scale bars = 200 µm.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:42

20

July

201

5

TECHNICAL NOTE 41

FIGURE 2. Stages of late embryonic and early larval development. (A) Em-bryos still in chorion (25 DPF); (B) embryos removed from chorion (26 DPF)with GI tract clearly visible; (C) hatch (33 DPF) and golden-eye stage; (D) lowerjaw has become extended and all sections of GI tract—foregut (FG), midgut(MG), and hindgut (HG)—have become identifiable; (E) swim bladder inflation(40 DPF) shown by arrow; (F) first feeding made obvious by food in GI tract atarrows. Observations until 8 DPF and from 9 DPF until the termination of thisstudy originated from captive and wild broodstock, respectively. Scale bars =500 µm.

Hatch and Free Embryo StagesEmbryos remained tightly curled in the chorion and organs

were clearly visible at 25 DPF (Figure 2A). The GI tract wasclearly observable when larvae were removed from the chorionat 26 DPF (Figure 2B). Eye spots initially appeared as solidmasses of bulging cells on either side of the neural plate andcontinued developing until the embryos entered the “golden-eye” stage (10/10 sampled embryos at golden-eye stage), wheneyes took on a golden appearance at 33 DPF (Figure 2C). Hatchwas first observed at 33 DPF/60 DDPF and lasted for severaldays. Free embryos had a large yolk sac, which sustained en-dogenous feeding for approximately 14 d (Figure 2C). Duringthe first few days after hatch, free embryos exhibited sporadicswimming, as swim bladder inflation had not yet occurred. Be-tween 37 and 38 DPF, the three sections of the intestine (foregut,midgut, and hindgut) became identifiable, and the lower jaw be-came extended (Figure 2D). Swim bladder inflation began by40 DPF (Figure 2E), when larvae began orienting themselves inthe water column and exhibiting early shoaling behavior. Larvaethen began feeding on saltwater rotifers Brachionus spp., whichwere clearly visible in the midgut and hindgut by 45 DPF (7/10sampled embryos feeding at 45 DPF; Figure 2F). The studyended at the beginning of the larval life history period, markedby the onset of exogenous feeding, at 45 DPF/229 DDPF.

DISCUSSIONThis study provides results from the first quantitative devel-

opmental staging series for early life stages of North AmericanBurbot, a species in its relative infancy in terms of aquacul-ture techniques and production. Timing of initial embryoniccell cleavage within egg lots appeared to be fairly synchronous.Burbot in this study (reared at 3–5◦C) exhibited an averagecleavage time of ∼ 8 h. For comparison, Hall et al. (2004) noteda cleavage time of 2.25 h at 7◦C for a related species, AtlanticCod, which was about half the time of cleavage observed forBurbot in this study. Differences may have been due to species-specific growth rates or different temperatures in the respectivestudies. Involution (the process by which an expanding epithe-lium turns over on itself and continues to grow in the oppositedirection along its basal margin) occurred much earlier in Bur-bot (∼20–25% epiboly) than in Zebrafish Danio spp., whichoccurs at approximately 50% epiboly (Kimmel et al. 1995).Somitogenesis (development of somites, the mesodermal layeron both sides of the developing neural tube) in Burbot beganbefore completion of epiboly, unlike in Zebrafish, which do notundergo somitogenesis until completion of epiboly (Kimmelet al. 1995).

This initial staging series for Burbot provides a solid em-pirical foundation for additional early life ontogeny studies forthis previously uncultured and historically unmanaged species(Paragamian and Willis 2000; Stapanian et al. 2010). WhileBurbot early life behaviors and biological requirements remainrelatively unknown, they currently constitute major bottlenecksin aquaculture production. Information generated by this study,particularly regarding the emerging relationships between time,thermal units, and embryonic and larval development in Burbotcan also help identify recruitment requirements and characterizerecruitment failure in managed wild and hatchery-supplementedpopulations. Finally, results of this study are expected to fur-ther develop and refine Burbot culture methods for conserva-tion and commercial applications, and guide additional researchneeded to inform Burbot conservation and management. Fur-ther research with larger samples sizes is needed to gain amore comprehensive understanding of variability in the tim-ing of events during embryogenesis within and between cohortsand the impacts of environmental factors such as temperature,dissolved oxygen, and photoperiod on the early development ofBurbot.

ACKNOWLEDGMENTSThis project was supported by the Kootenai Tribe of Idaho

and the Bonneville Power Administration (Project 198806400;contract 37267). We appreciate the continued support of theKootenai Tribe of Idaho; the British Columbia Ministry ofForests, Lands, and Natural Resource Operations; the IdahoDepartment of Fish and Game; Cramer Fish Sciences; the U.S.Fish and Wildlife Service; and UI-ARI.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:42

20

July

201

5

42 EGAN ET AL.

REFERENCESBalon, E. K. 1990. Epigenesis of an epigeneticist: the development of some

alternative concepts on the early ontogeny and evolution of fishes. GuelphIchthyology Reviews 1:1–46.

Barron, J. M., N. R. Jensen, P. J. Anders, J. P. Egan, and K. D. Cain. 2013a. Sup-pression of cannibalism during larviculture of Burbot through size grading.North American Journal of Aquaculture 75:556–561.

Barron, J. M., N. R. Jensen, P. J. Anders, J. P. Egan, S. C. Ireland, and K. D.Cain. 2012. Effects of temperature on the intensive culture performance oflarval and juvenile North American Burbot (Lota lota maculosa). Aquaculture364-365:67–73.

Barron, J. M., N. R. Jensen, P. J. Anders, J. P. Egan, S. C. Ireland, and K. D. Cain.2013b. Effects of stocking density on survival and yield of North AmericanBurbot reared under semi-intensive conditions. Transactions of the AmericanFisheries Society 142:1680–1687.

Brown, J. A., G. Minkoff, and V. Puvanendran. 2003. Larviculture of At-lantic Cod (Gadus morhua): progress, protocols and problems. Aquaculture227:357–372.

Fujimura, K., and N. Okada. 2007. Development of the embryo, larva andearly juvenile of Nile Tilapia Oreochromis niloticus (Pisces: Cichlidae). De-velopmental staging system. Development Growth and Differentiation 49:301–324.

Hall, T. E., P. Smith, and I. A. Johnston. 2004. Stages of embryonic devel-opment in the Atlantic Cod Gadus morhua. Journal of Morphology 259:255–270.

Jensen, N. R., S. C. Ireland, J. T. Siple, S. R. Williams, and K. D. Cain.2008a. Evaluation of egg incubation methods and larval feeding regimesfor North American Burbot. North American Journal of Aquaculture 70:162–170.

Jensen, N. R., S. R. Williams, S. C. Ireland, J. T. Siple, M. D. Neufeld, andK. D. Cain. 2008b. Preliminary captive Burbot spawning observations. Pages155–165 in V. L. Paragamian and D. H. Bennett editors. Burbot: ecology, man-agement, and culture. American Fisheries Society, Symposium 59, Bethesda,Maryland.

Jensen, N. R., M. D. Zuccarelli, S. J. Patton, S. R. Williams, S. C. Ireland,and K. D. Cain. 2008c. Cryopreservation and methanol effects on Burbotsperm motility and egg fertilization. North American Journal of Aquaculture70:38–42.

Kimmel, C. B., W. W. Ballard, S. R. Kimmel, B. Ullman, and T. F. Schilling.1995. Stages of embryonic development of the zebrafish. DevelopmentalDynamics 203:253–310.

Neufeld, M. D., C. A. Davis, K. D. Cain, N. R. Jensen, S. C. Ireland, and C.Lewandowski. 2011. Evaluation of methods for the collection and fertilizationof Burbot eggs from a wild stock for conservation aquaculture operations.Journal of Applied Ichthyology 27:9–15.

Paragamian, V. L., and D. W. Willis, editors. 2000. Burbot: biology, ecology,and management. American Fisheries Society, Fisheries Management SectionPublication 1, Spokane, Washington.

Stapanian, M. A., V. L. Paragamian, C. P. Madenjian, J. R. Jackson, J. Lap-palainen, M. J. Evanson, and M. D. Neufeld. 2010. World-wide status ofBurbot and conservation measures. Fish and Fisheries 11:34–56.

Van Houdt, J. K., B. Hellemanns, and F. A. M. Volckaert. 2003. Phylogenetic re-lationships among Paelearctic and Nearctic Burbot Lota lota: Pleistocene ex-tinctions and recolonization. Molecular Phylogenetics and Evolution 29:599–612.

Wolnicki, J., and L. R. Kaminski. 2001. Influence of water temperature ongrowth, survival, condition, and biological quality of juvenile Burbot Lotalota (L.). Archives of Polish Fisheries 9:79–86.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:42

20

July

201

5


Click for updates


Effect of Air Exposure and Resubmersion on theBehavior and Oxidative Stress of Pacific White ShrimpLitopenaeus vannameiHui-Ling Liua, Shi-Ping Yanga, Cheng-Gui Wanga, Siu-Ming Chana, Wang-Xiong Wanga, Zhen-Hua Fenga & Cheng-Bo Suna

a Fishery College, Guangdong Ocean University, 40 East Jiefang Road, Xiashan District,Zhangjiang City, Guangdong Province 524025, ChinaPublished online: 15 Dec 2014.

To cite this article: Hui-Ling Liu, Shi-Ping Yang, Cheng-Gui Wang, Siu-Ming Chan, Wang-Xiong Wang, Zhen-Hua Feng &Cheng-Bo Sun (2015) Effect of Air Exposure and Resubmersion on the Behavior and Oxidative Stress of Pacific White ShrimpLitopenaeus vannamei, North American Journal of Aquaculture, 77:1, 43-49, DOI: 10.1080/15222055.2014.955157








http://dx.doi.org/10.1080/15222055.2014.955157




ARTICLE

Effect of Air Exposure and Resubmersion on the Behaviorand Oxidative Stress of Pacific White Shrimp Litopenaeusvannamei

Hui-Ling Liu, Shi-Ping Yang,* Cheng-Gui Wang, Siu-Ming Chan, Wang-XiongWang, Zhen-Hua Feng, and Cheng-Bo SunFishery College, Guangdong Ocean University, 40 East Jiefang Road, Xiashan District, Zhangjiang City,Guangdong Province 524025, China

AbstractThe effect of air exposure on antioxidant activities in Pacific white shrimp Litopenaeus vannamei was studied. The

behavioral changes in the shrimp and the levels of superoxide dismutase (SOD), malondialdehyde (MDA), and totalantioxidant competence (T-AOC) in the muscles and hepatopancreas were determined after the shrimp were exposedto air and then resubmersed in water. Results showed that the duration of air exposure significantly influenced shrimpsurvival. The maximum air exposure period during which the shrimp could remain alive was 30 min. After 10 minof air exposure, the shrimp could survive when they were resubmersed in water. The T-AOC in the hepatopancreasand muscles was significantly decreased in shrimp that were exposed to air for 20 min. The MDA content in thehepatopancreas was significantly higher for the 20-, 30-, and 40-min air exposure groups than for the control group.During the resubmersion period, the MDA content in the shrimp hepatopancreas and muscles increased. For shrimpthat were exposed to air for 10 min, SOD activities in the hepatopancreas and muscles were restored after 3 h ofresubmersion in water. Our results indicate that air exposure can cause oxidative damage to Pacific white shrimp,but the damage can be reversed after the shrimp are resubmersed.

During the consumption of oxygen (O2) by respiring cells,approximately 0.1–0.2% of O2 is converted to reactive oxygenspecies (ROS), including superoxides (O−

2 ) and hydrogen per-oxide (H2O2; Fridovich 2004). Although ROS act as messengermolecules in normal cell functions (Wang et al. 2009), exces-sive ROS can be produced, thereby contributing to increasedoxidative stress (Kim et al. 2009). Oxidative stress is definedas the imbalance between the production and elimination offree radicals. In a balanced process, ROS are generated andthen eliminated by protective mechanisms (Durackova 2010).Excessive ROS can damage cellular macromolecules, includingnucleic acids, membrane lipids, and proteins (Wang et al. 2009).Various environmental stressors, including pathogen infection(Munoz et al. 2000; Mathew et al. 2007), acute salinity or pHchanges (Wang et al. 2009), temperature stress (Zhou et al.2010), hypoxia (Zenteno-Savın et al. 2006), and air exposure

*Corresponding author: [email protected] June 13, 2014; accepted August 6, 2014

(Romero et al. 2007, 2011), have also been shown to induceROS generation in aquatic animals.

The Pacific white shrimp Litopenaeus vannamei is one of themost important cultured species in the shrimp farming indus-try. Global production of Pacific white shrimp is approximately2,328,000,000 kg (Ma et al. 2013). In a shrimp-grade culturesystem, live shrimp are removed from the water for transportbetween ponds. Previous studies were conducted to developtechniques for transporting live shrimp and for increasing theirsurvival rate during a short-term transport period. For example,shrimp can be chilled and anesthetized during transport (Salinand Jayasree-Vadhyar 2001; Lorsingkum et al. 2011). However,shrimp usually suffer from various stressors during transportbecause of handling, air exposure, disturbance, and changes inother environmental factors. Air exposure is also consideredone of the more severe stressors to crustaceans, causing serious

43

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:42

20

July

201

5

44 LIU ET AL.

metabolic and respiratory disturbances (Sang and Fotedar 2005;Barrento et al. 2011).

Studies have yet to provide information on the behavioralresponses and changes in oxidative-related variables of Pacificwhite shrimp after they are subjected to air exposure and re-submersion. In the present study, we investigated the effects ofair exposure and resubmersion on the behavior and oxidativestress of Pacific white shrimp. Our objective was to develop astrategy for handling shrimp with minimal stress and mortalityincidences after air exposure during transport between ponds.

METHODSExperimental animals.—Experiments were conducted at the

Marine Biology Research Base of Guangdong Ocean Univer-sity (Zhanjiang, Guangdong, China). A batch of healthy Pacificwhite shrimp (mean body weight ± SD = 13.5 ± 0.5 g) wasbrought to the laboratory from a shrimp pond. Only shrimp inthe intermolt stage were selected and used for the present study.The shrimp were maintained in an air-conditioned room (nor-mal water temperature) at 28◦C and received no food for at least12 h before the experiment was conducted.

Air exposure and resubmersion.—Air exposure was per-formed by placing the shrimp in individual opaque foam tanks(120-L capacity) without water; wet gauze was used in the tanksto maintain air humidity. A single layer of shrimp was placed ineach foam tank. The shrimp were then subjected to air exposurestress for 0 (control), 10, 20, 30, or 40 min at 28◦C. After 10 minof air exposure, 120 shrimp were divided into five groups (15tanks at a density of 8 shrimp/tank) and were resubmersed inaerated seawater (28◦C) for 1.0, 1.5, 2.0, 2.5, or 3.0 h. Fromeach of the 10-, 20-, 30-, and 40-min air exposure treatments,four groups of eight shrimp were resubmersed in aerated seawa-ter (28◦C) for 30 min. The swimming mode, jumping behavior,branchial motion breathing, heartbeat, and appendage motionof shrimp were observed after the shrimp were subjected toair exposure and resubmersion. The absence of a heartbeat wasconsidered an indication of death.

Sample collection and analysis.—Hepatopancreas and mus-cle samples were collected for use in determining antioxidantvariables. Shrimp were anesthetized before their tissues wereexcised using a previously described method (Luedemana andLightnera 1992). The excised hepatopancreas and muscle tis-sues were homogenized in tris-HCl buffer (pH 7.4) at 4◦C. Thesamples were then centrifuged at 4,000 × gravity for 10 min at4◦C, and the clear supernatant was used immediately to analyzeantioxidant variables according to the methods of Yang et al.(2010). Superoxide dismutase (SOD) activity, malondialdehyde(MDA), and total antioxidant competence (T-AOC) were evalu-ated using commercial kits (Nanjing Jiancheng BioengineeringInstitute, China) according to the manufacturer’s instructions.

Statistical analysis.—Data describing the antioxidant vari-ables for the hepatopancreas and muscles were analyzed by one-way ANOVA and Duncan’s multiple comparisons of means to

determine statistical differences. Statistical analyses were per-formed using SPSS version 11.5 for Windows (SPSS, Chicago,Illinois).

RESULTS

BehaviorSudden air exposure and hypoxia caused behavioral changes

in Pacific white shrimp, including manic behavior, frequentjumping, abdominal appendage motion, branchial motion, andincreased heartbeat. After 10 min of air exposure, the shrimpshowed manic behavior characterized by occasional jumping.Mortality was not observed in shrimp that had been exposedto air for 30 min. However, after 40 min of air exposure, noheartbeat was observed, indicating that the shrimp had died.Therefore, 30 min was the maximum period of air exposure forwhich shrimp could remain alive.

After 10 min of air exposure, a portion of the shrimp thatwere resubmersed could swim in equilibrium; some of theseshrimp swam in a leaning fashion, whereas other shrimp couldnot swim. All of the shrimp could still swim after 10 min ofair exposure and 30 min of resubmersion. The maximum airexposure time that allowed 100% of the shrimp to survive afterair exposure and resubmersion was 10 min; 30 min was themaximum air exposure time that enabled any shrimp to remainalive. However, all of the shrimp in the 30-min air exposuregroup had died by the end of the 30-min resubmersion period(Table 1).

Effect of Air Exposure on Hepatopancreas AntioxidantVariables

The SOD activity in the hepatopancreas gradually increasedafter the shrimp were exposed to air, although no significant dif-ference was observed in SOD activity (P > 0.05). The shrimphepatopancreas was the only tissue in which the MDA contentwas affected significantly after air exposure (P < 0.05). TheMDA content in the hepatopancreas reached the maximum value(18.0 units [U]/g of tissue) after 30 min of air exposure. Air ex-posure also significantly affected T-AOC in the hepatopancreas(P < 0.05). The T-AOC decreased after 10 and 20 min of airexposure and then gradually increased after 30–40 min of airexposure (Figure 1).

Effect of Air Exposure on Muscle Antioxidant VariablesThe SOD activity in the muscles significantly decreased af-

ter 30 min of air exposure (P < 0.05). Although no signifi-cant difference in MDA content was observed in the muscles(P > 0.05), a gradual increase in MDA content within the rangeof 2.326–2.974 U/g of tissue was detected after air exposure.The T-AOC remained unchanged after 10 min of air exposure.However, a significant decrease in T-AOC was observed in themuscles after 20 min of air exposure. The minimum T-AOC

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:42

20

July

201

5

AIR EXPOSURE EFFECT ON PACIFIC WHITE SHRIMP 45

TABLE 1. Effect of air exposure and resubmersion on the behavior of Pacific white shrimp.

Behavior

Air exposuretime (min) During air exposure Immediately after resubmersion After 30 min of resubmersion

0 Frequent jumping; abdominalappendage motion; branchialmotion; heartbeat

10 Occasional jumping; abdominalappendage motion; branchialmotion; heartbeat

Three shrimp swam inequilibrium, three shrimp swamin a leaning pattern, and twoshrimp could not swim

Seven shrimp swam inequilibrium; one shrimp swam ina leaning pattern

20 Gill cavity motion; heartbeat;abdominal appendage motion intwo shrimp

One shrimp swam in leaningpattern, and the remainingshrimp could not swim; gillcavity motion; heartbeat

One shrimp swam in equilibrium,two shrimp swam in a leaningpattern, and the remainingshrimp died

30 Heartbeat Heartbeat All shrimp died40 All shrimp died

value (1.05 U/g of tissue) was detected after 40 min of air ex-posure (Figure 2).

Effect of Resubmersion on Hepatopancreas AntioxidantVariables

Air exposure followed by resubmersion did not elicit anysignificant effect on hepatopancreas SOD activity (P > 0.05),which remained in the range of 11.57 to 12.30 U/mg of protein.The MDA content in the hepatopancreas varied differentiallywith air exposure followed by resubmersion, exhibiting agradual increase. Significantly high MDA content was observedafter 1.5 and 2.5 h of resubmersion (P < 0.05). In contrast toSOD activity and MDA content, the T-AOC in the hepatopan-creas showed a different pattern of fluctuation characterizedby an initial increase and then a decrease at one or more timepoints. Regardless of the fluctuations, T-AOC consistentlyexhibited a significantly higher value after resubmersion thanbefore resubmersion (P < 0.05). The maximum T-AOC value(35.97 U/g of tissue) in the hepatopancreas was observed after1.5 h of resubmersion (Figure 3).

Effect of Resubmersion on Muscle Antioxidant VariablesThe SOD activity, MDA content, and T-AOC in the muscles

were significantly affected by air exposure and subsequent re-submersion (P < 0.05). The SOD activity in the muscles showedfluctuating patterns, with minimum values being observed after2 h of resubmersion. The MDA content in the muscles signif-icantly increased within the first 2 h of resubmersion, and themaximum MDA content was detected after 2 h of resubmer-sion. A significant decrease in muscle T-AOC was observed

after 2 h of resubmersion (P < 0.05). After 3 h of resubmersion,muscle T-AOC was restored to values similar to those observedimmediately after resubmersion (P > 0.05; Figure 4).

DISCUSSIONIn Asia, particularly in China, live shrimp are a more popular

cuisine than head-on dead shrimp. During size-grading in a cul-ture facility, live shrimp are transported from one pond to anotherin the absence of water. Upon arrival at the target site, shrimpare subjected to air exposure procedures, including catching,weighing, and processing. However, these procedures can causeconsiderable stress. The behavioral response of shrimp is a goodvisual indicator of physiological stress. Appendage motion andshock response were impaired by air exposure. Thirty minuteswas the maximum air exposure period for which shrimp couldremain alive. Our results are consistent with the results of pre-vious studies (Shinji et al. 2012). On the basis of behavioralresults, we found that 10 min was the maximum air exposuretime that allowed for the survival of shrimp after the exposureand resubmersion.

During air exposure, the O2 exchange level in the gills is prob-ably at its minimum or negligible. For crustaceans, exposure toair disrupts oxygen consumption and leads to great variationsin physiological variables (Fotedar and Evans 2011). Duringair exposure, the sinus gland peptide-G of Pacific white shrimpwas secreted from the X-organ–sinus gland into the hemolymph,thereby increasing glucose levels in the hemolymph (Shinji et al.2012). Crustaceans that are exposed to air also exhibit a reduc-tion in osmoregulatory capacity (Sang and Fotedar 2005), anincrease in lactose concentration (Taylor and Waldron 1997),

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:42

20

July

201

5

46 LIU ET AL.

FIGURE 1. Effect of air exposure on (a) superoxide dismutase (SOD) activity,(b) malondialdehyde (MDA) content, and (c) total antioxidant competence (T-AOC) in the hepatopancreas of Pacific white shrimp. Different lowercase lettersindicate significant differences between time points (mean ± SD). Note thedifference in the units along the y-axis.

FIGURE 2. Effect of air exposure on (a) superoxide dismutase (SOD) activity,(b) malondialdehyde (MDA) content, and (c) total antioxidant competence (T-AOC) in the muscles of Pacific white shrimp. Different lowercase letters indicatesignificant differences between time points (mean ± SD). Note the differencein the units along the y-axis.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:42

20

July

201

5


FIGURE 3. Effect of resubmersion on (a) superoxide dismutase (SOD) activ-ity, (b) malondialdehyde (MDA) content, and (c) total antioxidant competence(T-AOC) in the hepatopancreas of air-exposed Pacific white shrimp. Differentlowercase letters indicate significant differences between time points (mean ±SD). Note the difference in the units along the y-axis.

FIGURE 4. Effect of resubmersion on (a) superoxide dismutase (SOD) activ-ity, (b) malondialdehyde (MDA) content, and (c) total antioxidant competence(T-AOC) in the muscles of air-exposed Pacific white shrimp. Different lower-case letters indicate significant differences between time points (mean ± SD).Note the difference in the units along the y-axis.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:42

20

July

201

5

48 LIU ET AL.

reductions in granular cell proportions, greater bacterial con-centrations in the system, and a longer clotting time (Fotedaret al. 2001).

Aerobic organisms possess a baseline status of antioxidantsystems to maintain a balance between production and removalof endogenous ROS and other pro-oxidants (Correia et al. 2003).The antioxidant system of aerobic organisms is a complex, adap-tive system for maintaining cellular functions and preventing ox-idative stress. Organisms can adjust their antioxidant defensesto maintain the concentration of ROS as a strategy for pro-tecting tissues against oxidative damage during re-oxygenation(Romero et al. 2007). The process by which organisms adjusttheir antioxidant defenses is called “preparation for oxidativestress” (Hermes-Lima et al. 1998). The present results demon-strate that air exposure affects the oxidative status of Pacificwhite shrimp. Although no significant difference was observedin hepatopancreas SOD activities, a gradual increase in SODactivity was evident. Several aquatic animals have exhibitedsimilar patterns of increase in antioxidant enzyme activities un-der limited O2 conditions. For example, significant increases inSOD activity of the muscles, hepatopancreas, and hemolymphwere observed in false king crabs Paralomis granulosa afterthey were subjected to air exposure (Romero et al. 2007). Asignificant increase in SOD activity was also observed in thehemocytes of Farrer’s scallops Chlamys farreri that were ex-posed to air for 2 h at 25◦C (Chen et al. 2007). The muscles ofPacific white shrimp showed a higher SOD activity in responseto hypoxia than under controlled conditions (Parrilla-Taylor andZenteno-Savın 2011), indicating that the changes in SOD ac-tivity are physiological responses for avoiding harmful effectscaused by air exposure.

The muscle SOD activity exhibited its lowest level after 2 h ofresubmersion and consistently increased during the succeedingthree time points. In Pacific white shrimp that were exposed tohypoxia, a similar increasing pattern of muscle SOD activity wasfound after 5 h of re-oxygenation (Parrilla-Taylor and Zenteno-Savın 2011). Almost all of the analyzed tissues in false kingcrabs showed an increase in SOD activity during the first 2 h ofresubmersion (Romero et al. 2011).

The T-AOC is the sum of enzymatic (SOD, catalase, andglutathione peroxidase) and nonenzymatic (ascorbate, urate, vi-tamin E, pyruvate, glutathione, taurine, and hypotaurine) an-tioxidants (Mahfouz et al. 2009). After 10 min of air exposure,T-AOC in the shrimp hepatopancreas decreased significantly,but no significant difference was observed in SOD activity; thisindicates that nonenzymatic antioxidants can be consumed inthe hepatopancreas during air exposure. High concentrations ofglutathione, thioredoxin, and carotenoids are found in the bod-ies of crustaceans (Lee and Shiau 2004; Chien and Shiau 2005;Aispuro-Hernandez et al. 2008). The T-AOC levels in the hep-atopancreas of Pacific white shrimp after 1.0, 1.5, 2.0, 2.5, and3.0 h of resubmersion were significantly higher than the level ob-served immediately after resubmersion (P < 0.05). The T-AOCin the muscles decreased significantly after 2 h of resubmersion

(P > 0.05) and increased steadily during the next hour of re-submersion. This result can be attributed to the production ofnonenzymatic antioxidants in the hepatopancreas. Romero et al.(2011) indicated that nonenzymatic antioxidants are major ac-tive compounds in the hepatopancreas during air exposure. Highconcentrations of carotenoids and their derivatives are found inthe organs of crayfish (Sagi et al. 1995). Pannunzio and Storey(1998) found that glutathione, a low-molecular-weight antiox-idant, increased in common periwinkles Littorina littorea afterexposure to anoxia and re-oxygenation.

The MDA content in the hepatopancreas and muscles of Pa-cific white shrimp increased as exposure time was prolonged.However, only the MDA content of the hepatopancreas showed asignificant change during air exposure. These results indicatedthat the hepatopancreas was more prone to oxidative damagethan muscles. In the hepatopancreas and muscle of false kingcrabs, lipid peroxidation levels increased after the crabs weresubjected to air exposure (Romero et al. 2007). Moreover, theMDA content in the hepatopancreas and muscles increased dur-ing resubmersion, and muscle MDA content peaked after 2 h ofresubmersion. Although an increase in T-AOC during the resub-mersion period was considered, the antioxidant defense systemwas insufficient to prevent lipid peroxidation in tissues.

In summary, air exposure altered the oxidant–antioxidantstatus and triggered the antioxidant responses of severalenzymes in Pacific white shrimp. Oxidative damage can bereversed when the shrimp are returned to their normal habitatconditions, but in many cases the stress is irreversible andeventually causes the death of the animal. The results of thisstudy are useful in defining the maximum period for whichthe shrimp can be exposed to air and returned to water withoutsuffering any irreversible damages.

ACKNOWLEDGMENTSThe current research was supported by the Guangdong Nat-

ural Science Foundation (Grant S2011040000169) and the Pro-gram of Ocean and Fishery Bureau of Guangdong (GrantsA201201B01 and GD2012-A03-012). We also thank all ofthe individuals who dedicated their time to assisting in ourexperiments.

REFERENCESAispuro-Hernandez, E., K. D. Garcia-Orozco, A. Muhlia-Almazan, L. del Toro-

Sanchez, R. M. Robles-Sanchez, J. Hernandez, G. Gonzalez-Aguilar, G.Yepiz-Plascencia, and R. R. Sotelo-Mundo. 2008. Shrimp thioredoxin is apotent antioxidant protein. Comparative Biochemistry and Physiology Part C148:94–99.

Barrento, S., A. Marques, P. Vaz-Pires, and M. Leonor Nunes. 2011. Cancerpagurus (Linnaeus 1758) physiological responses to simulated live trans-port: influence of temperature, air exposure and AQUI-S. Journal of ThermalBiology 36:128–137.

Chen, M. Y., H. S. Yang, M. Delaporte, S. J. Zhao, and K. Xing. 2007. Immuneresponses of the scallop Chlamys farreri after air exposure to different tem-peratures. Journal of Experimental Marine Biology and Ecology 345:52–60.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:42

20

July

201

5


Chien, Y. H., and W. C. Shiau. 2005. The effects of dietary supplementation ofalgae and synthetic astaxanthin on body astaxanthin, survival, growth, and lowdissolved oxygen stress resistance of kuruma prawn, Marsupenaeus japonicusBate. Journal of Experimental Marine Biology and Ecology 318:201–211.

Correia, A. D., M. H. Costa, O. J. Luis, and D. R. Livingstone. 2003. Age-relatedchanges in antioxidant enzyme activities, fatty acid composition and lipidperoxidation in whole body Gammarus locusta (Crustacea: Amphipoda).Journal of Experimental Marine Biology and Ecology 289:83–101.

Durackova, Z. 2010. Some current insights into oxidative stress. PhysiologicalResearch 59:459–469.

Fotedar, S., and L. Evans. 2011. Health management during handling and livetransport of crustaceans: a review. Journal of Invertebrate Pathology 106:143–152.

Fotedar, S., E. Tsvetnenko, and L. Evans. 2001. Effect of air exposure on theimmune system of the rock lobster Panulirus cygnus. Marine and FreshwaterResearch 52:1351–1355.

Fridovich, I. 2004. Mitochondria: are they the seat of senescence? Aging Cell3:13–16.

Hermes-Lima, M., J. M. Storey, and K. B. Storey. 1998. Antioxidant defensesand metabolic depression: the hypothesis of preparation for oxidative stress inland snails. Comparative Biochemistry and Physiology Part B 120:437–448.

Kim, Y. S., F. Ke, and Q. Y. Zhang. 2009. Effect of β-glucan on activity ofantioxidant enzymes and Mx gene expression in virus infected Grass Carp.Fish and Shellfish Immunology 27:336–340.

Lee, M. H., and S. Y. Shiau. 2004. Vitamin E requirements of juvenile grassshrimp, Penaeus monodon, and effects on non-specific immune responses.Fish and Shellfish Immunology 16:475–485.

Lorsingkum, T., C. Limsuwan, W. Purivirojkul, and N. Chuchird. 2011. Eval-uation of isoeugenol as an anesthetic for Pacific white shrimp (Litopenaeusvannamei). Proceedings of the 49th Kasetsart University Annual Conference3:57–64.

Luedemana, R. A., and D. V. Lightnera. 1992. Development of an in vitroprimary cell culture system from the penaeid shrimp, Penaeus stylirostris andPenaeus vannamei. Aquaculture 101(3–4):205–211.

Ma, Z., R. Wan, X. Song, and L. Gao. 2013. The effect of three culture methodson intensive culture system of Pacific white shrimp (Litopenaeus vannamei).Journal of Ocean University of China 12:434–440.

Mahfouz, R., R. Sharma, D. Sharma, E. Sabanegh, and A. Agarwal. 2009.Diagnostic value of the total antioxidant capacity (TAC) in human seminalplasma. Fertility and Sterility 91:805–811.

Mathew, S., K. Ashok Kumar, R. Anandan, P. G. Viswanathan Nair, and K.Devadasan. 2007. Changes in tissue defence system in white spot syndromevirus (WSSV) infected Penaeus monodon. Comparative Biochemistry andPhysiology Part C 145:315–320.

Munoz, M., R. Cedeno, J. Rodrıguez, W. P. van der Knaap, E. Mialhe, andE. Bachere. 2000. Measurement of reactive oxygen intermediate productionin haemocytes of the penaeid shrimp, Penaeus vannamei. Aquaculture 191:89–107.

Pannunzio, T. M., and K. B. Storey. 1998. Antioxidant defenses and lipidperoxidation during anoxia stress and aerobic recovery in the marine gastro-pod Littorina littorea. Journal of Experimental Marine Biology and Ecology221:277–292.

Parrilla-Taylor, D. P., and T. Zenteno-Savın. 2011. Antioxidant enzyme activitiesin Pacific white shrimp (Litopenaeus vannamei) in response to environmentalhypoxia and reoxygenation. Aquaculture 318:379–383.

Romero, M., M. Ansaldo, and G. A. Lovrich. 2007. Effect of aerial exposureon the antioxidant status in the sub-Antarctic stone crab Paralomis granulose(Decapoda: Anomura). Comparative Biochemistry and Physiology Part C146:54–59.

Romero, M., F. Tapella, M. P. Sotelano, M. Ansaldo, and G. A. Lovrich. 2011.Oxidative stress in the sub-Antarctic false king crab Paralomis granulosaduring air exposure and subsequent re-submersion. Aquaculture 319:205–210.

Sagi, A., M. Rise, K. Isam, and S. M. Arad. 1995. Carotenoids and their deriva-tives in organs of the maturing female crayfish Cherax quadricarinatus. Com-parative Biochemistry and Physiology Part B 112:309–313.

Salin, K. R., and K. Jayasree-Vadhyar. 2001. Effect of different chilling ratesfor cold anaesthetization of Penaeus monodon (Fabricius) on the survival, du-ration and sensory quality under live storage in chilled sawdust. AquacultureResearch 32:145–155.

Sang, H. M., and R. Fotedar. 2005. The effects of air exposure on hemolymphosmoregulatory capacity of brown tiger shrimp, Penaeus esculentus, andwestern king shrimp, Penaeus latisulcatus, reared at different salinities. Jour-nal of Applied Aquaculture 16(3–4):79–94.

Shinji, J., B. J. Kang, T. Okutsu, K. Banzai, T. Ohira, N. Tsutsui, and M.N. Wilder. 2012. Changes in crustacean hyperglycemic hormones in Pacificwhiteleg shrimp Litopenaeus vannamei subjected to air-exposure and low-salinity stresses. Fisheries Science 78:833–840.

Taylor, H. H., and F. M. Waldron. 1997. Respiratory responses to air-exposure inthe southern rock lobster, Jasus edwardsii (Hutton) (Decapoda: Palinuridae).Marine and Freshwater Research 48:889–898.

Wang W. N., J. Zhou, P. Wang, T. T. Tian, Y. Zheng, Y. Liu, W. J. Mai, andA. L. Wang. 2009. Oxidative stress, DNA damage and antioxidant enzymegene expression in the Pacific white shrimp, Litopenaeus vannamei whenexposed to acute pH stress. Comparative Biochemistry and Physiology PartC 150:428–435.

Yang, S. P., Z. H. Wu, J. C. Jian, and X. Z. Zhang. 2010. Effect of marine redyeast Rhodosporidium paludigenum on growth and antioxidant competenceof Litopenaeus vannamei. Aquaculture 309(1–4):62–65.

Zenteno-Savın, T., R. Saldierna, and M. Ahuejote-Sandoval. 2006. Superoxideradical production in response to environmental hypoxia in cultured shrimp.Comparative Biochemistry and Physiology Part C 142:301–308.

Zhou, J., L. Wang, Y. Xin, W. N. Wang, W. Y. He, A. L. Wang, and Y. Liu.2010. Effect of temperature on antioxidant enzyme gene expression and stressprotein response in white shrimp, Litopenaeus vannamei. Journal of ThermalBiology 35:284–289.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:42

20

July

201

5


Click for updates


Development and Evaluation of an Acoustic Deviceto Estimate Size Distribution of Channel Catfish inCommercial PondsBradley T. Goodwillera, Rachel V. Beechamb, J. D. Heffingtona & James P. Chambersc

a National Center for Physical Acoustics, University of Mississippi, 1 Coliseum Drive,University, Mississippi 38677, USAb Department of Natural Science and Environmental Health, Mississippi Valley StateUniversity, 14000 Highway 82 West, Itta Bena, Mississippi 38941, USAc Department of Mechanical Engineering, University of Mississippi, 1 Chucky Mullins Drive,University, Mississippi 38677, USAPublished online: 15 Dec 2014.

To cite this article: Bradley T. Goodwiller, Rachel V. Beecham, J. D. Heffington & James P. Chambers (2015) Developmentand Evaluation of an Acoustic Device to Estimate Size Distribution of Channel Catfish in Commercial Ponds, North AmericanJournal of Aquaculture, 77:1, 50-54, DOI: 10.1080/15222055.2014.956197








http://dx.doi.org/10.1080/15222055.2014.956197




TECHNICAL NOTE

Development and Evaluation of an Acoustic Deviceto Estimate Size Distribution of Channel Catfishin Commercial Ponds

Bradley T. Goodwiller*National Center for Physical Acoustics, University of Mississippi, 1 Coliseum Drive, University,Mississippi 38677, USA

Rachel V. BeechamDepartment of Natural Science and Environmental Health, Mississippi Valley State University,14000 Highway 82 West, Itta Bena, Mississippi 38941, USA

J. D. HeffingtonNational Center for Physical Acoustics, University of Mississippi, 1 Coliseum Drive, University,Mississippi 38677, USA

James P. ChambersDepartment of Mechanical Engineering, University of Mississippi, 1 Chucky Mullins Drive, University,Mississippi 38677, USA

AbstractAs one step in the continued effort to utilize acoustic methods

and techniques to the betterment of catfish aquaculture, an acous-tic “catfish sizer” was designed to determine the size distribution ofChannel Catfish Ictalurus punctatus in commercial ponds. The cat-fish sizer employed a custom-built 460 kHz piezoelectric transducerwith a 20◦ beam angle. The sizer was built at the National Centerfor Physical Acoustics and tested in commercial catfish ponds inMississippi and Arkansas in June and October. To test the system,fish were collected in seven different ponds using a cut seine. Theywere weighed and then sent through a pipe into the open pond. Thetransducer was mounted underwater at the end of the pipe. As theindividual fish exited the pipe, they passed through the acousticfield created by the transducer and the echo detected by the trans-ducer was recorded. In three of the seven ponds, calibration testswere run. For these calibration tests, the same process was usedwith the added step that a jet of bubbles was introduced in front ofthe transducer after every fifth fish. Analysis of the measured fishweights and acoustic data were used to produce a Gaussian predic-tion model. This model was applied to the data from the blind teststo predict the average weight and the SD of fish in each pond. Thedevice and methodology worked well, with significant differencesbetween the measured and calculated weights found in only one ofthe seven ponds tested.

*Corresponding author: [email protected] January 21, 2014; accepted August 6, 2014

The catfish farming industry is a major revenue generator inthe Southeast and nationwide (NASS 2009; MSU 2011). Thereis a need for new techniques and technologies that can improveand optimize the various facets of the industry, including pondinventory. Without accurate inventory information, farmers canmake errors in decisions regarding proper feeding, restockingamounts, and treatments for diseases.

There are several inventory estimation techniques that re-quire historical data on stocking densities and mortality rates(Wyk et al. 1997). In addition to the burden of requiring ac-curate records, these methods rely on guesswork to accountfor diseases and other unobserved phenomenon that affect theinventory of the pond. Developing new methods for determin-ing inventory that reduce labor requirements as well as elimi-nate the dependence on accurate record keeping would improveprofitability.

This manuscript describes an attempt to empirically correlatethe acoustic echo from individual fish with their size in a waythat provides an accurate estimate of the size distribution of anentire pond. As a supplement to the Aquascanner pond inven-tory system (Chambers et al. 2010), a system utilizing acousticbackscatter was designed and tested to gain information about

50

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:42

20

July

201

5

TECHNICAL NOTE 51

FIGURE 1. Schematic of the catfish sizer from two angles.

the weight of individual Channel Catfish Ictalurus punctatusand thus to estimate the size distribution of catfish in ponds.If coupled with the information provided by the Aquascanner(Chambers et al. 2005), this would give an improved descriptionof the total biomass of catfish ponds.

METHODSDesign.—The catfish sizer was designed to be a portable,

field-deployable unit (Figure 1). All of the data collection equip-ment was attached to a PVC raft which could be either stakeddown or tied off to reduce movement. The raft was equippedwith a 15-cm-diameter PVC pipe angled down into the watersuch that half of the pipe was above water and half was be-low. At the underwater exit of the pipe, there was a circularmetal band approximately 35 cm in diameter which served asthe transducer mount. This ring was mounted to the pipe on oneside, so that the pipe and the ring were offset from each other.Thus, the exiting fish were forced to be a minimum of 20 cmaway from the transducer. Since the transducer used had a nearfield of 7 mm, the fish were constrained to be in the far field ofthe transducer. This means that the beam from the transducerwas well formed and free from any distortion due to the edgesof the transducer. However, since the wavelength was smallerthan the size of the fish, multiple echoes and interference ef-fects did occur at the surface of the fish. For these and severalother reasons, an absolute measurement of target strength wasimpractical. Thus, the focus of this study was purely empiricalin nature. The acoustic signal generated was a 20-cycle sinewave with frequency 460 kHz transmitted every 20 ms. Thetransducer was used as both the sending and receiving element.

When the acoustic pulse encountered a change in impedance(density or sound speed), a portion of the wave, proportional tothe change in density, was reflected back. This reflected wavewas then incident on the transducer, which generated a voltageproportional to the amplitude of the incident wave. An analogcircuit was used to integrate the voltage across the transduceruntil sampled by the system. The integrated voltage was sampled25,000 times/s. While this sampling rate is below the Nyquistrate, the integration via the analog circuit makes the data col-lected analogous to the root mean square value, or the acousticenergy recorded by the transducer. The electronics necessary tooperate the transducer and collect data were mounted on top ofthe raft. The system was interfaced (via wireless access point)with a computer on the bank of the pond, allowing the collecteddata to be transferred wirelessly and stored for later analysis.

Experimental process.—A subsample of fish from the testpond was collected by seining a fraction of the pond using a cutseine. For calibration and testing purposes, the individual fishwere weighed to the 100th pound before being placed into thePVC pipe. The fish then swam down the pipe which was outsideof the seine net, allowing them to swim back into the open pond.As the fish exited the pipe they swam through the acoustic field,and the resulting echo was detected by the transducer.

The experiment was conducted in seven different test ponds,and in three of these ponds a calibration test was conducted. Intotal, 745 fish were tested ranging in size from 0.3 to 7.36 lb;82 of these fish were used for calibration. For the calibrationtests, the acoustic signal from individual fish was matched totheir measured weight. Unfortunately, the acoustic detection andcomputer processing were not perfect. For various reasons, theanalyzed acoustic data did not always report the same number

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:42

20

July

201

5

52 GOODWILLER ET AL.

FIGURE 2. Processed acoustic data showing the echoes from fish at roughly25–30 cm as well as the air bubbles used for calibration at 50 s and range5–10 cm.

of signatures as fish sampled. In some cases, the automatedroutine was unable to identify fish echoes from the backgroundnoise, resulting in undercounts. In other cases, fish twisted andmoved as they went down the pipe which resulted in identifyingtwo distinct echoes for the same fish, thus overestimating thenumber of test fish. To deal with this issue, a stream of airbubbles was introduced in front of the transducer after everyfifth fish. The air bubble provided a large signal directly in frontof the transducer. Since the fish were constrained to be 20 cmaway from the transducer, this signal was easily identifiable.If a fish was missed, the air bubble echo allowed the user todiscard the appropriate section of data while still being able toproperly match the acoustic data with the measured weights onthe remaining fish.

Blind tests were conducted in each pond after the calibrationtests. The only operational difference between the two was theintroduction of the air bubbles. For the blind tests, missed orovercounted fish were a reality and were considered to be partof the system.

Data analysis.—The data analysis process was broken intoseveral parts. First, the raw acoustic data were analyzed by acustom Matlab (MathWorks 2013) routine to find the fish andthen to determine their acoustic signature. For the calibrationtests, the program generated a plot of acoustic amplitude ver-sus time and position as well as an array of acoustic signa-tures. Figure 2 shows one of the calibration test plots. Theseplots were generated for the calibration tests so that missedfish could be located. To determine the acoustic signatures, theprogram utilized a threshold detection routine. The minimumthreshold and minimum time between detections was manuallypicked to minimize the number of under- and overcounted fishon the calibration data. Once these two parameters were deter-mined, the program was run on all of the calibration data aswell as the blind data. The acoustic signatures from the cali-bration runs were matched with the measured weights of theindividual fish. Any set between bubble echoes that did not re-turn five signatures was thrown out since the matching couldnot be done correctly. Once properly matched, the calibrationdata were plotted in data analysis software called GraphicalAnalysis (Vernier Software and Technology 2014), which wasused to find the model that best predicted the mean weight ofthe individual ponds as well as the SD from the mean. Thismodel was applied to the blind data and used to calculate theweights of the tested fish. We used t-tests for significance testingto compare the measured and calculated weights of fish in eachblind trial. Data were analyzed with SAS 8.02 (SAS Institute2001), and significance testing was done at the 0.05 probabilitylevel.

FIGURE 3. Calibration data with the manually determined Gaussian fit plotted.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:42

20

July

201

5

TECHNICAL NOTE 53

TABLE 1. Mean weights and SDs, both measured and calculated, for each of the seven test ponds and the combined calibration data.

Measured mean Calculated mean Error of Measured Calculated ErrorPond number weight (lb) weight (lb) mean estimate (%) SD (lb) SD (lb) of SD (%)

1 1.51 1.38 8.61 0.51 0.76 49.02 2.45 1.74 28.98 1.12 1.24 10.73 1.53 1.43 6.54 0.76 0.86 13.24 1.81 1.58 12.71 0.84 1.25 48.85 2.34 2.15 8.12 1.19 1.51 26.86 0.84 0.89 5.95 0.39 0.53 35.97 1.43 1.49 4.20 0.33 0.84 154Calibration data 1.92 1.83 4.83 1.86 2.25 21.18

RESULTSMeasured fish weights in the study ranged from 0.3 to 7.36

lb, and SDs ranged from 0.33 to 1.19 lb. The Graphical Analysissoftware used a least-squares routine to determine a Gaussianmodel to fit the calibration data. However, the root mean squareerror was not the desired metric of success, as it is a measureof how close each individual point is to its predicted value.The goal of the catfish sizer was not necessarily to predict thesize of individual fish but rather to accurately estimate the sizedistribution of the entire pond. The metrics of interest, therefore,were the average weight and the SD from that average. To thisend, the constants of the Gaussian curve were manually adjustedto minimize the errors in both the mean values of the predictedversus measured weights as well as the SDs for the calibrationdata (Figure 3).

Utilizing this model applied to the blind data, the catfishsizer provided estimates of the average weights of the testedfish. Comparisons of measured and calculated weights usingT-test showed no significant differences for ponds 1 (P =

FIGURE 4. Plot of the mean weights with SEs, both measured and calculated,for each of the seven test ponds.

0.22, N = 71), 3 (P = 0.41, N = 91), 4 (P = 0.11, N =109), 5 (P = 0.37, N = 102), 6 (P = 0.44, N = 114), and7 (P = 0.67, N = 111). A significant difference between the mea-sured and calculated weights was found for pond 2 (P = 0.004,N = 65). The mean values and SEs for measured and calculatedweights for each pond can be seen in Figure 4 and are tabulatedin Table 1.

DISCUSSIONIt has been shown for various species of fish that a fish’s target

strength is correlated to its size (Kubecka and Duncan 1998). Anobject’s target strength is essentially a measure of that object’sacoustic reflectivity. Obtaining an absolute measure of targetstrength can be complicated and difficult, particularly in a fieldsetting. From a technical standpoint, numerous properties of thehardware used and the acoustic media must be precisely knownor independently measured. In addition to the technical issues,there are biological considerations that complicate matters fur-ther. Extensive work has been done investigating several of thepotential variables (Frouzova et al. 2005), with most studiesshowing that fish orientation relative to the transducer is crit-ically important (Love 1977). The purpose of this study wasto investigate the possibility of bypassing the theoretical deter-mination of actual target strength and simply use an empiricalmethod to estimate fish size. Therefore, it is not unexpected thatthe catfish sizer was not very accurate on a fish-by-fish basis.However, the results show that many of these biological con-siderations tend to average themselves out when investigatinglarger sample sizes, allowing reasonable predictions for averageweight and SD.

It should be noted that this fitting method imposes constraints.The fit cannot predict a size less than 0.76 lb or greater than 7.24lb. While this is a limitation, it covers the span of most food-sized fish.

The work detailed here presents a method of determiningthe size distribution of an arbitrary pond with no prior informa-tion of size distribution. While the results are not as accurateas was hoped, they are not unusable. There are several facets ofthis project that would benefit from further development, par-ticularly in reducing the number of missed fish. It is estimated

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:42

20

July

201

5

54 GOODWILLER ET AL.

that approximately 10% of the blind test fish were missed. Theuse of two transducers parallel to the path of the fish wouldpotentially allow for more accurate detection. This would havethe added benefit of potentially providing the length of the fish,which could then be used to determine size. The frequency ofthe transducer used was chosen due to availability of materials.Different frequencies may provide better results. Another areafor improvement is the postprocessing routine. The routine wasdeveloped to best model the entire size range of fish tested. Ifbasic information about the size of the fish in a pond of interestis known (e.g., brood sized or fingerling), then different fittingmethods could be applied to better suit the specific size range ofinterest. This work is intended to show the potential of an acous-tic sizing device to be a boon to the catfish aquaculture industry.

REFERENCESChambers, J., C. D. Minchew, and R. Beecham. 2005. Biomass assessment

in commercial catfish ponds. Journal of the Acoustical Society of America117:2554.

Chambers, J. P., H. E. Bass, K. E. Gilbert, and D. E. Kleinert. 2010. Underwaterbiomass assessment device and method. U.S. patent 7,688,675. March 30,2010.

Frouzova, J., J. Kubecka, H. Balk, and J. Frouz. 2005. Target strength of someEuropean fish species and its dependence on fish body parameters. FisheriesResearch 75:86–96.

Kubecka, J., and A. Duncan. 1998. Acoustic size vs. real size relationships forcommon species of riverine fish. Fisheries Research 35:115–125.

Love, R. H. 1977. Target strength of an individual fish at any aspect. Journal ofthe Acoustical Society of America 62:1397–1403.

MathWorks. 2013. Matlab R2013a. MathWorks, Natick, Massachusetts.MSU (Mississippi State University). 2011. Mississippi agriculture, forestry and

natural resources factbook 2011. MSU Extension Service, Mississippi Stateand Mississippi Agricultural Extension Statistics Service, Jackson.

NASS (National Agricultural Statistics Service). 2009. 2007 Census of agricul-ture. NASS, Geographic Area Series, Part 51, Washington, D.C.

SAS. 2001. SAS, version 8.02. SAS Institute, Cary, North Carolina.Vernier Software and Technology. 2014. Graphical analysis. Vernier Software

and Technology, Beaverton, Oregon.Wyk, P. V., M. Masser, D. Heikes, and H. S. Killian. 1997. Inventory assessment

methods for catfish ponds. Southern Regional Aquaculture Center, Stoneville,Mississippi.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:42

20

July

201

5


Click for updates


Effect of Feed Pellet Characteristics on Growth andFeed Conversion Efficiency of Largemouth Bass Raisedin PondsJames Tidwella, Shawn Coylea & Leigh Anne Brighta

a Aquaculture Research Center, Kentucky State University, 103 Athletic Road,Frankfort,Kentucky 40601, USAPublished online: 16 Dec 2014.

To cite this article: James Tidwell, Shawn Coyle & Leigh Anne Bright (2015) Effect of Feed Pellet Characteristics on Growthand Feed Conversion Efficiency of Largemouth Bass Raised in Ponds, North American Journal of Aquaculture, 77:1, 55-58, DOI:10.1080/15222055.2014.960116








http://dx.doi.org/10.1080/15222055.2014.960116




COMMUNICATION

Effect of Feed Pellet Characteristics on Growth and FeedConversion Efficiency of Largemouth Bass Raised in Ponds

James Tidwell,* Shawn Coyle, and Leigh Anne BrightAquaculture Research Center, Kentucky State University, 103 Athletic Road, Frankfort,Kentucky 40601, USA

AbstractWe investigated the growth, survival, and feed conversion of

Largemouth Bass Micropterus salmoides fed one of two sizes offloating pellets or a sinking pellet for 151 d in 0.04-ha ponds. Thefloating and sinking pellets were made from almost identical for-mulas. The fish were fed once daily to apparent satiation basedon observed feeding activity. There was no significant difference insurvival, growth, or feed conversion of fish fed the larger floatingpellet versus those fed the standard-size floating pellet or sinkingpellet. Fish fed the sinking pellet were significantly larger at har-vest than fish fed either size of floating pellet. The efficiency offeed utilization was not decreased with the sinking pellet. Theseresults indicate the Largemouth Bass accept and utilize sinkingpellets well. However, their use on commercial scale farms wouldprobably require some management modifications.

The Largemouth Bass Micropterus salmoides is the largestmember of the North American sunfish family, Centrarchi-dae, and shows promise as a commercial aquaculture species(Cochran et al. 2009). There have been a number of studiesevaluating their nutritional requirements (Tidwell et al. 1996;Coyle et al. 2000; Bright et al. 2005) and the ability to utilizealternative feed ingredients (Tidwell et al. 2005, 2007; Subhadraet al. 2006). Fish diets must be formulated to meet the animals’nutritional requirements and must be manufactured into pelletsthat the fish will readily accept and utilize efficiently. The abil-ity of fish to detect, ingest and utilize a feed can be affected byphysical characteristics such as pellet density (floating, sinking,sinking rate), size (shape, diameter, and length), color (contrast),and texture (hardness; Jobling et al. 2001).

Early commercial fish diets were manufactured by steam pel-leting into densely compressed particles that sink. Later, extru-sion processed feeds that float on the water surface became morecommon. Floating feed is a valuable management tool because

*Corresponding author: [email protected] May 5, 2014; accepted August 25, 2014

it allows the farmer to see how much, and how actively, the fisheat (Mgbenka and Lovell 1984). Extrusion can also increase thewater stability of the pellet (Lovell 1989), as well as increasethe digestibility of carbohydrates (Hardy and Barrows 2002)and potentially deactivate certain antinutrition factors (Sorensen2012).

Cost is a disadvantage of extruded floating feeds comparedwith pelleted sinking. Extrusion requires more expensive equip-ment than pelleting (Mgbenka and Lovell 1984). Also, energycosts are higher because more steam is used (the ingredientmix is heated to a higher temperature, and extruded feeds re-quire heat in drying, while pelleted feeds do not (Mgbenkaand Lovell 1984). However, in some species the lower nutrientdensity and greater pellet stability of extruded diets prolongedgastric emptying time compared with steam pellets and conse-quently reduced feed intake (Hilton et al. 1981; Venou et al.2009) and even growth (Booth et al. 2000, 2002; Honorato et al.2010).

The size and shape of the feed may also affect the amounteaten (Jobling et al. 2001). Particle size of fish feeds shouldbe as large as possible to minimize nutrient leaching (Lovell1989). Optimally, pellet size should be 25–50% of the speciesmouth width (Jobling et al. 2001). Kubitza and Lovshin (1997)fed age-1 Largemouth Bass (initial weight, 62 g) a 7-mm troutpellet and reported that pellet size may have been too small.They indicated that Largemouth Bass strike at individual pelletsand many strikes would be required to satiate a fish. Kubitzaand Lovshin (1997) proposed that by feeding a large pellet toLargemouth Bass, less feeding energy would be expended permeal and faster growth might be attained.

The objective of our study was to evaluate the relative per-formance of Largemouth Bass fed feeds with different pelletcharacteristics (sinking or floating and pellet size) under practi-cal pond conditions.

55

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5

56 TIDWELL ET AL.

TABLE 1. Formulation of diets manufactured as two sizes of floating pelletsor as sinking pellets and fed to juvenile Largemouth Bass in ponds.

Ingredient or composition (%)

Item Floating Sinking

IngredientSoybean meal (48%) 20.85 18.58Poultry by-product meal 16.00 15.00Fish meal 10.00 10.00Distillers dried grains 0.00 10.00Rice bran–high fat 6.49 8.00Corn 10.57 7.78Poultry blood meal 7.50 7.35Wheat flour 10.00 3.00Fish–poultry oil blend 9.88 8.92Mono/di-calcium

phosphate0.70 0.70

D-methionine 0.30 0.29Vitamin premix 0.25 0.25Trace mineral premix 0.05 0.05

Analyzed composition

Moisture 9.0 9.1Protein 43.4 43.6Lipid 16.8 18.4Fiber 2.7 2.6Ash 7.8 7.4Nitrogen-free extract 20.2 18.9

METHODSAge-1 (subadult) feed-trained Largemouth Bass averaging

184.7 g (SD, 35.3) and 22.8 mm (SD, 1.3) were indiscriminatelystocked into nine, 0.04-ha earthen ponds (volume ≈ 550 m3) ata density of 10,000 bass/ha. New water was only added to re-place evaporation. There were three replicate ponds per dietarytreatment. Fish were fed once daily (0800 hours) to apparentsatiation based on feeding response. Total amount of diet con-sumed was recorded daily. The three experimental diets wereformulated to contain 43% protein and 17% lipid (Table 1). Allformulations were similar, with only minor differences effectedby pellet density, size, and ability to float. Pellets were manufac-tured by Burris Mill and Feed (Franklinton, Louisiana). Floatingpellets were produced with diameters of 5.5 mm (control) and13.0 mm (large); sinking pellets were 5.5 mm, as manipulatedvia pellet expansion during manufacture.

After 151 d of feeding each pond was seined three times andthen drained to ensure that all fish were removed. Harvestedfish were bulk weighed and counted into holding tanks. Onceall fish were harvested from a pond, a indiscriminate sample of40 fish from each pond were individually weighed and measuredfor total length.

Water quality.—In each pond, water temperature and dis-solved oxygen were measured twice daily using an YSI 85 DOmeter (YSI Company, Yellow Springs, Colorado). Total am-monia, nitrite, pH, and alkalinity were measured three timesper week using a HACH DR/2500 spectrophotometer (HACH,Loveland, Colorado).

Statistics.—Various growth performance characteristicswere calculated: condition factor (K = [average weight/averageharvest length3]100); specific growth rate (SGR) or percent bodyweight gain/day was calculated as SGR = [(logeWf − logeWi) /t]100, where Wf = final weight (g), Wi = initial weight (g), and t =time (d); feed conversion ratio or FCR = total diet fed (g)/totalwet weight gain (g); and feed cost per unit of weight gain =manufactured diet cost/weight gain (kg), where cost (US$) wasbased on actual invoice amounts. Treatments were statisticallycompared using analysis of variance at α= 0.05 via Statistix ver-sion 8.0 (Statistix Analytical Software, Tallahassee, Florida). Ifsignificant differences were found, treatment means were sepa-rated using Fisher’s Least Significant Difference method (Steeleand Torrie 1980). All percentage and ratio data were arcsine-transformed prior to analysis (Zar 1984). However, data arepresented untransformed to facilitate comparisons.

RESULTS AND DISCUSSIONThere were no significant treatment differences (P > 0.05)

in any of the measured water quality variables over the studyperiod (Table 2). Values for un-ionized ammonia nitrite were allat concentrations considered acceptable (Roseboom and Richey1977; Palachek and Tomasso 1984).

Pellet SizesBoth pellet sizes of the floating diet were well utilized by

the fish throughout the experiment. Harvest data indicated thatstandard size (control, 5.5 mm) versus large (13.0) floating pel-lets produced no significant (P > 0.05) impact on productioncharacteristics during second year growth of Largemouth Bassin ponds (Table 3). However, fish fed the 5.5-mm sinking pelletshad significantly greater (P ≤ 0.05) growth (average weight gainand SGR) than fish fed either size of floating pellet.

Although Kubitza and Lovshin (1997) proposed that Large-mouth Bass might benefit from increased pellet sizes, they didnot conduct a controlled comparison. While our data, however,showed larger pellets of no benefit, Nortvedt and Tuene (1995)reported that feeding larger pellets improved feed conversionefficiencies in Atlantic Halibut Hippoglossus hippoglossus. Lin-ner and Brannas (1994) reported that in Arctic Char Salvelinusalpinus responded more rapidly to larger pellets, but pellet con-sumption was better at intermediate sizes. That is agrees withour study; i.e., bass actively accepted the larger pellets, but thatdid not subsequently result in improved production.

Floating Versus Sinking PelletsThe Largemouth Bass fed the 5.5-mm sinking pellets were

significantly larger (P ≤ 0.05) at harvest (629 g) than those

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5

COMMUNICATION 57

TABLE 2. Mean ± SE water quality results from ponds containing Largemouth Bass fed two sizes of floating pellets or sinking pellets tested in three replicateponds per diet. There were no significant differences (P > 0.05) among treatments.

Feed type

Variable Standard floating Large floating Sinking

Morning temperature (◦C) 21.1 ± 0.6 20.3 ± 0.1 20.9 ± 0.1Afternoon temperature (◦C) 22.4 ± 0.0 22.1 ± 0.2 22.6 ± 0.1Morning dissolved oxygen (mg/L) 7.3 ± 0.1 7.3 ± 0.0 7.0 ± 0.2Afternoon dissolved oxygen (mg/L) 12.4 ± 0.9 1 0.9 ± 0.2 10.9 ± 0.4Total ammonia (mg/L) 0.67 ± 0.13 0.69 ± 0.04 0.89 ± 0.04Un-ionized ammonia (mg/L) 0.132 ± 0.01 0.115 ± 0.01 0.124 ± 0.01Nitrite (mg/L) 0.017 ± 0.01 0.041 ± 0.01 0.029 ± 0.02pH 8.7 ± 0.1 8.5 ± 0.0 8.5 ± 0.1Alkalinity (mg/L) 97.8 ± 2.6 101.7 ± 4.4 101.4 ± 2.8

fed a similar size floating pellet (566 g). Specific growth ratewas also significantly higher (P ≤ 0.05) in fish fed the sinkingpellet (2.6 g/d) than in those fed the floating pellet (2.2 g/d).There was no significant difference (P > 0.05) in survival(> 90%), total yield, condition factor, FCR, or average dailyfeed consumption among fish fed the similar size floating orsinking diets. Total yields in our study were higher than for theLargemouth Bass reported by Cochran et al. (2009), who com-pared diets containing different fish meal levels, but stocked ata lower density (8,650/ha versus 10,000/ha). The fish we fedthe 5.5-mm sinking pellets, compared with those fed the large(13 mm) floating pellets, had production metric differences simi-lar to the differences between the 5.5-mm sinking versus floatingpellets.

The larger average size at harvest and higher SGR, of Large-mouth Bass fed sinking pellets are in agreement with findingsof Booth et al. (2000, 2002) for Silver Perch Bidyanus bidyanusand with Honorato et al. (2010) for Pacu Piaractus mesopotam-icus. According to Sorensen (2012), the lower bulk density (lessweight per unit volume) of floating pellets compared with sink-ing pellets would result in fish fed floating pellets becomingphysically satiated at a lower energy intake. This is supportedby Mgbenka and Lovell (1984), who reported that sinking feeds

had less bulk per unit of weight (i.e., were more dense) and lessdigestible energy. Dense pellets with low energy concentrationswould allow fish to consume more of the nutrients essential togrowth before becoming satiated by energy intake or stomachfullness.

There are also behavioral aspects associated with feedingsinking pellets versus floating. Kubitza and Lovshin (1997) pro-posed that use of sinking pellets might result in more wastedfeed (i.e., higher FCR) because Largemouth Bass were not likelyto pick up pellets off the bottom. However, we found FCR wasnot different among fish fed floating or sinking pellets, indi-cating the efficient use of sinking pellets. Both studies utilizedsatiation feeding. Other behavioral traits might also be affectthe suitability of different pellet characteristics. Kubitza andLovshin (1997) also observed that, when using floating pellets,Largemouth Bass were reluctantly forced into surface waterswith high light levels and temperatures (≥30◦C). Cochran et al.(2009) also observed that Largemouth Bass were hesitant tofeed on floating pellets at the surface on bright sunny days, butwould readily consume pellets that would “slow-sink” throughthe water column.

Another potential positive aspect of the use of sinking feedsfor Largemouth Bass is that the manufacture of sinking pellets

TABLE 3. Mean ± SD of production metrics for age-1 pond-cultured (151 d) Largemouth Bass fed two different floating pellet sizes (5.5 and 13 mm) or asinking pellet (5.5 mm) of similar diet formulations. Different lowercase letters indicate statistical significance (P < 0.05).

Production variable Standard floating Large floating Sinking

Average weight (g) 565.6 ± 11.0 y 580.6 ± 6.5 y 629.0 ± 9.8 zSurvival (%) 94.2 ± 2.2 z 94.6 ± 0.4 z 90.1 ± 2.3 zTotal yield (kg/ha) 5,321.1 ± 28.0 z 5,491.8 ± 59.3 z 5,669.5 ± 220.5 zSpecific growth rate (g/d) 2.2 ± 0.1 y 2.3 ± 0.0 y 2.6 ± 0.1 zCondition factor 1.8 ± 0.0 z 1.9 ± 0.0 z 1.9 ± 0.0 zFeed conversion ratio 1.7 ± 0.1 z 1.7 ± 0.1 z 1.5 ± 0.0 zFeed consumption (kg/ha/d) 34.1 ± 0.9 z 36.2 ± 0.7 z 33.7 ± 1.3 zFeed cost per unit of weight gain (US$/kg) 1.4 ± 0.0 z 1.4 ± 0.1 z 1.3 ± 0.0 z

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5

58 TIDWELL ET AL.

allows lower dietary carbohydrate levels. To produce float-ing pellets a formulation must contain ≥20% carbohydrate forproper expansion (Lovell 1989). However, as a strict carnivore,Largemouth Bass have been found to be negatively impacted bycarbohydrate levels of 20% (Goodwin et al. 2002; Amoah et al.2008). Use of sinking pellets would allow more flexibility infeed formulation and reduction of carbohydrate content. Also, itwould increase the potential of more localized feed productionbecause steam pelleting facilities are more common and lesscomplex than extruders.

A potential approach to capitalize on the positive results ofsinking pellets, without sacrificing the positive feed manage-ment characteristics of floating pellets, could be to feed primar-ily sinking feed but to mix in some floating pellets as “indicatorpellets” to assist in monitoring feeding activity and consump-tion, as evaluated by Mgbenka and Lovell (1984) in ChannelCatfish Ictalurus punctatus. They proposed the feeding of a15:85 ratio of extruded to pelleted diets to reduce costs whileproviding the same management benefits of feeding observation.

In summary, we found no benefit from increased pellet sizesof floating diets for feeding Largemouth Bass. We also foundthat Largemouth Bass fed sinking pellets had higher growthrates than those fed similar-seized floating pellets.

ACKNOWLEDGMENTSWe would like to thank Russell Neal for assistance in the

daily care of the culture system. This research was partiallyfunded by a U.S. Department of Agriculture/Cooperative StateResearch, Education, and Extension Service grant to KentuckyState University (KSU) under agreement KYX-80-91-04A. Ad-ditional support was provided by Kentucky’s Regional Univer-sity Trust Fund to the Aquaculture Program as KSU’s Programof Distinction.

REFERENCESAmoah, A., S. D. Coyle, C. D. Webster, R. M. Durborow, L. A. Bright, and

J. H. Tidwell. 2008. Effects of graded levels of carbohydrate on growth andsurvival of Largemouth Bass, Micropterus salmoides. Journal of the WorldAquaculture Society 39:397–405.

Booth, M., G. Allan, A. Evans, and V. Gleeson. 2002. Effects of steam pellet-ing or extrusion on digestibility and performance of Silver Perch Bidyanusbidyanus. Aquaculture Research 33:1163–1173.

Booth, M. A., G. L. Allen, and R. Warner-Smith. 2000. Effects of grinding, steamconditioning and extrusion of a practical diet on digestibility and weight gainof Silver Perch, Bidyanus bidyanus. Aquaculture 182:287–299.

Bright, L. A., S. D. Coyle, and J. H. Tidwell. 2005. Effect of dietary lipid leveland protein energy ratio on growth and body composition of Largemouth BassMicropterus salmoides. Journal of the World Aquaculture Society 36:129–134.

Cochran, N. J., S. D. Coyle, and J. H. Tidwell. 2009. Evaluation of reducedfish meal diets for second year growth of the Largemouth Bass, Micropterussalmoides. Journal of the World Aquaculture Society 40:735–743.

Coyle, S. D., J. H. Tidwell, and C. D. Webster. 2000. Response of LargemouthBass Micropterus salmoides to dietary supplementation of lysine, methionine,and highly unsaturated fatty acids. Journal of the World Aquaculture Society31:89–95.

Goodwin, A. E., R. L. Lochmann, D. M. Tieman, and A. J. Mitchell. 2002.Massive hepatic necrosis and nodular regeneration in Largemouth Bass feddiets high in available carbohydrate. Journal of the World Aquaculture Society33:466–477.

Hardy, R. W., and F. T. Barrows. 2002. Diet formulation and manufacture. Pages506–600 in J. E. Halver and R. W. Hardy, editors. Fish nutrition. AcademicPress, London.

Hilton, J. W., C. Y. Cho, and S. J. Slinger. 1981. Effect of extrusion processingand steam pelleting diets on pellet durability, pellet water absorption andthe physiological response of Rainbow Trout Salmo gairdneri. Aquaculture25:185–194.

Honorato, C. A., L. C. Almeida, C. Da Silva Nunes, D. J. Carneiro, and G.Moraes. 2010. Effects of processing on physical characteristics of diets withdistinct levels of carbohydrates and lipids: the outcomes on the growth ofPacu (Piaractus mesopotamicus). Aquaculture Nutrition 16:91–99.

Jobling, M., E. Gomes, and J. Dias. 2001. Feed types, manufacture and ingredi-ents. Pages 25–48 in D. Houlihan, T. Boujard, and M. Jobling, editors. Foodintake in fish. Blackwell Scientific Publications, Oxford, UK.

Kubitza, F., and L. L. Lovshin. 1997. Pond production of pellet-fed advancedjuvenile and food-size Largemouth Bass. Aquaculture 149(3-4):253–262.

Linner, J., and E. Brannas. 1994. Behavioral response to commercial food of dif-ferent sizes and self-initiated food size selection by Arctic Char. Transactionsof the American Fisheries Society 123:416–422.

Lovell, T. 1989. Feed formulation and processing. Pages 107–128 in T. Lovell,editor. Nutrition and feeding of fish. Van Nostrand Reinhold, New York.

Mgbenka, B. O., and R. T. Lovell. 1984. Feeding combinations of extrudedand pelleted diets to Channel Catfish in ponds. Progressive Fish-Culturist46:245–248.

Nortvedt, R, and S. Tuene. 1995. Multivariate evaluation of feed for AtlanticHalibut. Chemometrics and Intelligent Laboratory Systems 29:271–282.

Palachek, R. M., and J. R. Tomasso. 1984. Toxicity of nitrite to Channel Catfish(Ictalurus punctatus), tilapia (Tilapia aurea), and Largemouth Bass (Mi-cropterus salmoides): evidence for a nitrite exclusion mechanism. CanadianJournal of Fisheries and Aquatic Sciences 41:1739–1744.

Roseboom, D. P., and D. L. Richey. 1977. Acute toxicity of residual chlorineand ammonia to some native Illinois fishes. Illinois State Water Survey Reportof Investigation 85.

Sorensen, M. 2012. A review of the effects of ingredient composition andprocessing conditions on the physical qualities of extruded high-energyfish feed as measured by prevailing methods. Aquaculture Nutrition 18:233–248.

Steele, R. G. D., and J. H. Torrie. 1980. Principles and procedures of statistics:a biometrical approach, 2nd edition. McGraw-Hill, New York.

Subhadra, B., R. Lochmann, S. Rawles, and R. Chen. 2006. Effect of dietarylipid source on the growth, tissue composition, and hematological parametersof Largemouth Bass Micropterus salmoides. Aquaculture 255:210–222.

Tidwell, J. H., S. D. Coyle, and L. A. Bright. 2007. Effects of different typesof dietary lipids on growth and fatty acid composition of Largemouth Bass.North American Journal of Aquaculture 69:257–264.

Tidwell, J. H., S. D. Coyle, L. A. Bright, and D. K. Yasharian. 2005. Evaluationof plant and animal source proteins for replacement of fish meal in practicaldiets for the Largemouth Bass Micropterus salmoides. Journal of the WorldAquaculture Society 36:454–463.

Tidwell, J. H., C. D. Webster, and S. D. Coyle. 1996. Effects of di-etary protein level on second year growth and water quality for Large-mouth Bass (Micropterus salmoides) raised in ponds. Aquaculture 145(1-4):213–223.

Venou, B., M. N. Alexis, E. Fountoulaki, and J. Haralabous. 2009. Performancefactors, body composition and digestion characteristics of Gilthead Sea BreamSparus aurata fed pelleted or extruded diets. Aquaculture Nutrition 15:390–401.

Zar, J. H. 1984. Biostastistical analysis, 2nd edition. Prentice-Hall, EnglewoodCliffs, New Jersey.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5


Click for updates


Growth Performance of Hybrid Striped Bass, RainbowTrout, and Cobia Utilizing Asian Carp Meal-BasedAquafeedsJohn Bowzera & Jesse Trushenskiaa Center for Fisheries Aquaculture and Aquatic Sciences, Southern Illinois UniversityCarbondale, 1125 Lincoln Drive, Carbondale, Illinois 62901, USAPublished online: 16 Dec 2014.

To cite this article: John Bowzer & Jesse Trushenski (2015) Growth Performance of Hybrid Striped Bass, RainbowTrout, and Cobia Utilizing Asian Carp Meal-Based Aquafeeds, North American Journal of Aquaculture, 77:1, 59-67, DOI:10.1080/15222055.2014.960117








http://dx.doi.org/10.1080/15222055.2014.960117




ARTICLE

Growth Performance of Hybrid Striped Bass, Rainbow Trout,and Cobia Utilizing Asian Carp Meal-Based Aquafeeds

John Bowzer and Jesse Trushenski*Center for Fisheries Aquaculture and Aquatic Sciences, Southern Illinois University Carbondale,1125 Lincoln Drive, Carbondale, Illinois 62901, USA

AbstractFish meal sparing is more difficult for nutritionally demanding carnivorous fishes, but economic considerations

and the limited supply of fish meal continue to incentivize investigations of alternative protein sources for aquafeeds.A promising alternative to traditional, marine-origin fish meal is fish meal derived from undesirable freshwaterspecies, such as the invasive Asian carp Hypophthalmichthys spp. To assess the relative value of such ingredients,we evaluated growth performance of juvenile hybrid Striped Bass (White Bass Morone chrysops × Striped Bass M.saxatilis; initial weight, 21.9 ± 0.2 g [mean ± SE]), Rainbow Trout Oncorhynchus mykiss (15.1 ± 0.2 g), and CobiaRachycentron canadum (57.2 ± 0.5 g) reared for 8 weeks on practical diets containing different levels of menhadenfish meal (MFM), Asian carp meal (CFM), or a 50:50 blend of these ingredients such that 0, 20, 40, or 60% of theestimated digestible protein content was derived from fish meal. Growth performance was generally consistent acrosstaxa, and weight gain tended to increase with fish meal inclusion, regardless of its origin. However, Cobia did performbetter on CFM-based diets, suggesting that MFM or CFM can yield improved performance for some taxa or lifestages, but these differences are likely to be marginal in most circumstances. We conclude CFM is a suitable andperhaps lower-cost alternative to MFM in feeds for carnivorous fishes.

Feed accounts for a substantial portion of overall operatingcosts (40–50%) for intensive aquaculture facilities (Cheng et al.2004). Protein is typically the most expensive component ofanimal feed, and this is particularly problematic in aquafeedsbecause fish have a much higher protein demand compared withother livestock (Keembiyehetty and Gatlin 1992). Historically,fish meal was used as a primary ingredient in aquafeeds be-cause of its high protein density, favorable amino acid profile,and digestibility and palatability to aquatic livestock. However,fish meal inclusion rates have been declining in recent decades(Tacon and Metian 2008; Tacon et al. 2011) because it is consid-erably more expensive than other alternative plant- and animal-derived protein sources (Gatlin et al. 2007; Welch et al. 2010;FAO 2014). However, sparing fish meal with these alternativescan be particularly difficult for nutritionally demanding carniv-orous fishes, which may not readily accept or perform as wellon diets with reduced levels or free of fish meal. Fish meal has


been the primary protein source for carnivorous species becauseit generally meets all essential amino acid requirements, is pro-tein dense, and is highly palatable; but, due to the high cost ofthis ingredient and its overall high inclusion levels in the diets,there are considerable economic incentives to reduce fish mealuse in aquafeeds (Watanabe 2002).

A potential alternative to traditional marine-origin fish mealis a freshwater fish meal rendered from invasive species such asSilver Carp Hypophthalmichthys molitrix and Bighead Carp H.nobilis (hereafter referred to collectively as Asian carp). Asiancarp have become particularly abundant in the Mississippi Riverbasin (McClelland et al. 2012; Tsehaye et al. 2013), and sinceinvasive species typically have adverse effects on native pop-ulations and economically important activities (Vitousek et al.1997; Lodge et al. 2006), a variety of control and eradicationstrategies have been investigated and deployed. Of these, controlthrough increased harvest pressure is a primary focus because it

59

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5

60 BOWZER AND TRUSHENSKI

is currently the only viable short-term control strategy (Conoveret al. 2007). Considerable fishing pressure is needed to success-fully control Asian carp in this system (Tsehaye et al. 2013);thus, market research and development have been undertaken toincrease the demand for Asian carp (Nelson 2013; Varble andSecchi 2013). However, since Asian carp are not a preferredfood fish in the United States, industrial applications, such asrendering into protein meals for livestock feeding, are consid-ered the most promising to incentivize harvest (Conover et al.2007; Bowzer et al. 2013).

Asian carp meal has several advantages over other alterna-tive protein sources. Utilization of plant-origin (i.e., derivativesof oilseeds or grains) and animal-origin (i.e., by-products oflivestock processing) feedstuffs in aquafeeds can be restrictedby inadequate nutrient levels, digestibility, or palatability; thepresence of antinutritional factors; or practical limitations to in-clusion in the diet (e.g., high cost, complications related to feedmanufacturing; Glencross et al. 2007). For instance, soybeanmeal is one of the most promising alternative protein sources(Gatlin et al. 2007), and despite its routine use in aquafeeds(Hendricks 2003) high inclusion levels can result in undesirableeffects such as poor palatability (Adelizi et al. 1998), low feedconversion efficiency (Davies and Morris 1997), and gut enteri-tis (Heikkinen et al. 2006; Iwashita et al. 2008). Fish meals,including Asian carp meal, do not appear to present any of thesechallenges (Bowzer et al., in press). Furthermore, while mostmarine fisheries are unlikely to withstand greater fishing pres-sure (FAO 2012), greater harvest of Asian carp would likelyimprove the ecological integrity of the Mississippi River basin.

Rendering Asian carp into fish meal may increase demandand associated harvest pressure on these invasive fish, but de-mand for Asian carp meal is dependent on the product’s qualityand pricing. Current pricing estimates for Asian carp meal rangefrom US$600–650/metric ton (P. Hitchens, Southern IllinoisUniversity Carbondale, personal communication), considerablylower than current pricing for marine-origin fish meal ($1,500–2,000/metric ton: FAO 2014). Additionally, an industry must beestablished to produce the product to supply the market. Giventhat, development of large-scale Asian carp meal productionfacilities has been relatively slow due to initial high risk of in-vestment in a new product and little information regarding itsnutritional value, contaminant load, or potential volume until re-cently. Researchers have begun to fill in many of these data gapssuch as establishing that the digestibility of Asian carp meal iscomparable to that of other fish meals in aquafeeds (Bowzeret al., in press), and its practical feeding value in LargemouthBass Micropterus salmoides feeds is similar to menhaden fishmeal (Bowzer et al. 2014). As this information has becomeavailable, interest and investment in Asian carp rendering hasincreased; a facility specializing in Asian carp meal productionrecently opened in Grafton, Illinois (Moon 2014). Therefore,Asian carp meal appears to be a promising protein source foraquafeeds, but its utilization by a range of nutritionally demand-ing carnivorous fish species has yet to be fully determined.

Accordingly, we assessed the growth performance of hybridStriped Bass (White Bass Morone chrysops × Striped BassM. saxatilis), Rainbow Trout Oncorhynchus mykiss, and CobiaRachycentron canadum fed diets containing different levels ofmenhaden fish meal (MFM) or Asian carp meal (CFM).

METHODSFish were held in recirculating aquaculture systems with con-

tinuous aeration and mechanical and biological filtration units.Trials with hybrid Striped Bass (initial weight, 21.9 ± 0.2 g[mean ± SE]; Keo Fish Farms, Keo, Arkansas) and RainbowTrout (15.1 ± 0.2 g; Crystal Lake Fisheries, Ava, Missouri)were conducted at the Center for Fisheries, Aquaculture, andAquatic Sciences (CFAAS) at Southern Illinois University Car-bondale, Carbondale, Illinois, while the Cobia (57.2 ± 0.5 g;Troutlodge Marine Farms, Vero Beach, Florida) trial was con-ducted at the Virginia Seafood Agricultural Research and Ex-tension Center at Virginia Tech, Hampton, Virginia. Stockingdensity, tank size, system filtration, and water quality for eachsystem are described in Table 1. Water temperature and dis-solved oxygen were measured daily (YSI 550 temperature–oxygen meter; Yellow Springs Instruments, Yellow Springs,Ohio), whereas other water quality conditions (total ammonianitrogen, nitrite-nitrogen, nitrate-nitrogen, alkalinity, and salin-ity [Cobia trial only]) were measured weekly for each system(Hach DR 2800 portable spectrophotometer, digital titrator, andreagents; Hach Company, Loveland, Colorado). Water qualityconditions were maintained within suitable ranges throughoutthe trials (Table 1).

Practical diets were formulated to meet the known nutritionalrequirements of hybrid Striped Bass, Rainbow Trout, and Co-bia (NRC 2011) and to contain approximately 40:12%, 42:13%,and 45:12% digestible protein : lipid, respectively (Table 2). Thediets contained different levels of menhaden fish meal (MFM;Special Select, Omega Protein., Houston, Texas), Asian carp fishmeal (CFM; Protein Products, Gainsville, Florida), or a 50:50blend of these ingredients such that 0% (0 FM), 20% (20 MFM,20 CFM), 40% (40 MFM, 40 CFM), or 60% (60 MFM, 60CFM, 60 Blend) of the estimated digestible protein content wasderived from a fish meal source. Estimates of digestible pro-tein content were based on apparent digestibility coefficientsreported for hybrid Striped Bass and Rainbow Trout (Barrowset al. 2012; Bowzer et al., in press); digestibility coefficientswere not available for Cobia; therefore, these feeds were for-mulated using Rainbow Trout values. Errors were made in theformulation of two diets (10% lipid in 60 MFM diet versus 12%in other diets for Cobia; 45% digestible protein in 60 CFM dietversus 40% in other diets for hybrid Striped Bass). Diets weremanufactured and analyzed in triplicate to determine proximatecomposition (Table 3) using standard methods described in de-tail by Rombenso et al. (2013). Minor differences resulting fromthe formulation errors mentioned above and/or inaccuracies iningredient weighing or mixing were observed; however, these

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5

ASIAN CARP MEAL-BASED AQUAFEEDS FOR CARNIVOROUS FISH 61

TABLE 1. Recirculation system characteristics and water quality observed during feeding trials for hybrid Striped Bass, Rainbow Trout, and Cobia. Waterquality values represent least-squares means ± SE.

Parameter Hybrid Striped Bass Rainbow Trout Cobia

Number of tanksa 24 24 24Tank volume (L) 150 190 300Stocking density (fish/tank) 10 10 10Biofiltration Trickle-down biofilter Trickle-down biofilter Fluidized bed biofilterMechanical filtration Bead filter Bead filter Bead filterAeration Blower/airstones Blower/airstones Blower/airstonesTemperature (◦C) 23.6 ± 1.1 15.3 ± 1.2 26.7 ± 1.1Dissolved oxygen (mg/L) 7.7 ± 0.3 9.4 ± 0.5 6.1 ± 0.5pH 7.7 ± 0.1 7.9 ± 0.3 7.4 ± 0.3Salinity (‰) <5 <5 24.4 ± 0.8Alkalinity (mg/L) 181 ± 54 189 ± 40 162 ± 22NO3

−-N (mg/L) 15.6 ± 6.5 3.8 ± 2.8 17.6 ± 7.9NO2

−-N (mg/L) 0.04 ± 0.05 0.03 ± 0.02 0.16 ± 0.08Total ammonia nitrogen (mg/L) 0.26 ± 0.15 0.12 ± 0.10 0.64 ± 0.36

aRainbow Trout study used only 24 of 36 tanks in the system.

unintended differences did not appear substantial enough to giverise to differences in growth performance (see Results). Feedswere randomly assigned to triplicate tanks (N = 3), and fishwere fed assigned diets once daily to apparent satiation for 8weeks. At the end of each trial, fish were harvested, counted,and group-weighed by tank to assess growth performance interms of the following metrics:

Weight gain (%)

= 100 × average individual final weight − average individual initial weight

average individual initial weight

Feed conversion ratio (FCR)

= average individual feed consumption (dry matter)

average individual weight gain

Specific growth rate (SGR, % body weight/d)

= 100 × loge (final individual weight) −loge(initial individual weight)

d of feeding

Feed intake (FI, % body weight/d) =

100 × total dry matter intake/(initial individual weight × final individual weight)0.5

d of feeding

Three fish per tank were randomly selected and euthanized byan overdose of tricaine methanesulfonate (MS-222; ∼200 mg/Lin culture water, fish immersed until opercular ventilation hadceased for 10 min) for individual weighing and dissection to cal-culate hepatosomatic index (HSI, all taxa) and viscerosomaticindex (VSI, hybrid Striped Bass and Rainbow Trout only) asfollows:

HSI = 100 × (liver weight/whole body weight)

VSI = 100 × (total viscera weight/whole body weight)

Data from each trial were analyzed separately by one-wayANOVA (PROC GLIMMIX) to determine the significance of

differences between dietary treatments (SAS version 9.3, SASInstitute, Cary, North Carolina). Tukey’s Honestly SignificantDifferences (HSD) pairwise comparison post hoc tests wereused to compare means when omnibus tests indicated signif-icant differences among treatment groups. Additionally, datafrom the MFM and CFM series (i.e., excluding the 60 Blendand 0 FM) for each study were analyzed by two-way ANOVA(PROC GLIMMIX) to determine the significance of fish mealtype and inclusion level as main and interactive effects. Wheninteractive effects were not significant, they were removed fromthe analysis and only the main effects were tested. All effectswere considered significant at P < 0.05.

RESULTSAll diets were well accepted and hybrid Striped Bass, Rain-

bow Trout, and Cobia performed in a manner generally con-sistent with our previous experience with these taxa; however,growth performance was observed to vary among dietary treat-ment groups (Table 4) and was influenced by fish meal typeand inclusion level (Table 5), depending on the performancemetric and taxon. Survival was 100% in each trial, except fortwo mortalities in the Cobia trial that were not related to dietarytreatment.

Growth of hybrid Striped Bass varied significantly amongdietary treatments and was influenced by fish meal inclusionlevel, but not fish meal type. Results of omnibus tests indicatedfinal weight, weight gain, and SGR were significantly affectedby dietary treatment, but pairwise comparisons did not revealsignificant differences among dietary treatment groups. Growthwas significantly reduced among fish fed diets containing lessthan 60% fish meal protein. Rainbow Trout growth did not varysignificantly among dietary treatments and was not influencedby fish meal type or inclusion level, but numeric trends linking

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5


TABLE 2. Feed formulations (g/kg, on as-fed basis) of diets where 60, 40, 20, or 0% of digestible protein consists of fish meal (Asian carp meal [CFM],menhaden fish meal [MFM], or blend) fed to hybrid Striped Bass, Rainbow Trout, and Cobia.

Ingredient 60 MFM 40 MFM 20 MFM 60 CFM 40 CFM 20 CFM 60 Blend 0 FM

Hybrid Striped BassMenhaden fish meal 358.7 239.2 119.6 0.0 0.0 0.0 180.0 0.0Asian carp meal 0.0 0.0 0.0 429.3 286.2 143.1 215.0 0.0Soybean meal 207.3 355.0 487.0 314.6 338.3 453.7 203.9 487.9Poultry byproduct meal 67.0 120.0 190.0 135.1 152.6 232.8 80.0 399.6Wheat bran 275.3 186.5 98.3 47.6 137.9 75.3 239.1 16.9Menhaden fish oil 62.5 70.1 75.9 44.1 55.8 65.8 52.9 66.3Carboxymethyl cellulose 20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0Choline chloride 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0Mineral premix 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0Vitamin premix 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2Stay C 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0Protein (crude) 460.0 474.3 491.4 551.2 497.9 507.7 475.8 531.8Protein (digestible) 400.0 400.0 400.0 450.0 400.0 400.0 400.0 400.0Lipid (crude) 120.0 120.0 120.0 120.0 120.0 120.0 120.0 120.0

Rainbow TroutMenhaden fish meal 376.7 251.1 125.6 0.0 0.0 0.0 188.3 0.0Asian carp meal 0.0 0.0 0.0 450.8 300.5 150.3 225.4 0.0Soybean meal 200.0 200.0 200.0 150.0 200.0 200.2 200.0 243.7Poultry byproduct meal 183.1 318.5 453.8 160.6 281.2 435.0 155.2 560.0Wheat bran 91.7 85.5 79.3 108.9 83.2 78.0 90.0 55.1Menhaden fish oil 69.3 65.7 62.1 50.6 56.0 57.2 62.0 62.1Blood meal 50.0 50.0 50.0 50.0 50.0 50.0 50.0 50.0Carboxymethyl cellulose 20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0Choline chloride 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0Mineral premix 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0Vitamin premix 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2Stay C 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0Protein (crude) 517.5 519.2 520.9 523.0 522.4 522.5 519.9 522.0Protein (digestible) 420.0 420.0 420.0 420.0 420.0 420.0 420.0 420.0Lipid (crude) 130.0 130.0 130.0 130.0 130.0 130.0 130.0 130.0

CobiaMenhaden fish meal 403.6 269.0 134.5 0.0 0.0 0.0 201.8 0.0Asian carp meal 0.0 0.0 0.0 483.0 322.0 161.0 242.0 0.0Soybean meal 164.5 263.6 408.2 158.2 287.8 370.7 137.9 358.7Poultry byproduct meal 70.0 120.0 177.1 110.0 162.0 225.2 80.0 320.0Wheat bran 248.6 169.7 84.1 136.8 86.3 58.2 199.6 71.1Menhaden fish oil 39.2 68.5 76.9 37.8 52.7 65.6 49.5 73.9Soy protein isolate 45.0 80.0 90.0 45.0 60.0 90.0 60.0 147.1Carboxymethyl cellulose 20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0Choline chloride 6.0 6.0 6.0 6.0 6.0 6.0 6.0 6.0Mineral premix 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0Vitamin premix 1.2 1.2 1.2 1.2 1.2 1.2 1.2 1.2Stay C 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0Protein (crude) 507.6 517.5 532.2 543.9 547.0 550.4 522.1 556.4Protein (digestible) 450.0 450.0 450.0 450.0 450.0 450.0 450.0 450.0Lipid (crude) 100.0 120.0 120.0 120.0 120.0 120.0 120.0 120.0

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5


TABLE 3. Analyzed composition (g/kg, on as-fed basis) of diets where 60, 40, 20 or 0% of digestible protein consists of fish meal (Asian carp meal [CFM],menhaden fish meal [MFM], or blend) fed to hybrid Striped Bass, Rainbow Trout, and Cobia.

Parameter 60 MFM 40 MFM 20 MFM 60 CFM 40 CFM 20 CFM 60 Blend 0 FM

Hybrid Striped BassDry matter 943.2 943.7 941.0 937.8 941.6 940.1 936.7 931.8Protein (crude) 447.1 456.0 474.5 574.0 491.1 491.6 472.6 525.6Lipid (crude) 125.7 128.4 131.7 130.5 123.5 126.0 125.3 127.2Ash 107.3 103.6 94.6 131.2 112.4 90.4 116.8 91.6

Rainbow TroutDry matter 940.1 943.2 941.0 936.7 937.8 941.6 943.7 931.8Protein (crude) 545.9 568.5 582.2 552.5 569.8 571.6 557.1 574.8Lipid (crude) 138.4 140.5 143.9 133.4 137.4 139.1 142.4 140.7Ash 134.1 123.6 103.8 144.8 123.2 111.4 229.7 105.1

CobiaDry matter 940.1 943.2 941.0 936.7 937.8 941.6 943.7 931.8Protein (crude) 545.9 568.5 582.2 552.5 569.8 571.6 557.1 574.8Lipid (crude) 138.4 140.5 143.9 133.4 137.4 139.1 142.4 140.7Ash 134.1 123.6 103.8 144.8 123.2 111.4 229.7 105.1

TABLE 4. Growth performance (final weight, weight gain, SGR, and FCR), FI, HSI, and VSI of hybrid Striped Bass, Rainbow Trout, and Cobia fed diets where60, 40, 20, or 0% of digestible protein consists of fish meal (Asian carp meal [CFM], menhaden fish meal [MFM], or Blend). Values represent least-squared means;pooled SEs and P-values resulting from one-way ANOVA tests are also provided. Letters indicate significant treatment differences (P < 0.05).

Parameter 60 MFM 40 MFM 20 MFM 60 CFM 40 CFM 20 CFM 60 Blend 0 FM Pooled SE P-value

Hybrid Striped BassInitial weight (g) 21.9 22.0 21.9 21.9 22.0 21.7 22.0 21.8 0.2 0.77Final weight (g)a 72.3 61.3 58.1 73.1 61.5 59.9 68.3 63.7 4.6 0.03Weight gain (%)a 230 179 165 233 180 176 211 192 21 0.03SGR (% body weight/d)a 2.2 1.9 1.8 2.2 1.8 1.8 2.1 1.9 0.1 0.04FCR 1.1 1.0 1.1 1.1 1.1 1.1 1.1 1.1 0.1 0.89FI (% body weight/d) 2.6 2.1 2.1 2.6 2.3 2.2 2.3 2.3 0.2 0.03HSI 1.5 1.4 1.2 1.5 1.4 1.4 1.3 1.3 0.1 0.08VSI 6.7 7.3 6.8 7.2 7.0 7.5 6.8 7.4 0.5 0.65

Rainbow TroutInitial weight (g) 14.9 15.1 15.1 15.1 15.1 15.1 15.1 15.0 0.2 0.97Final weight (g) 45.7 45.5 38.7 48.3 45.5 43.6 43.4 41.6 5.3 0.74Weight gain (%) 206 200 158 221 201 188 188 178 35 0.75SGR (% body weight/d) 1.8 1.7 1.4 1.6 1.9 1.8 1.7 1.6 0.3 0.71FCR 1.0 0.9 1.1 0.9 1.0 1.0 1.0 1.0 0.1 0.65FI (% body weight/d) 2.1 1.9 1.9 2.0 2.2 2.1 2.0 1.9 0.2 0.83HSI 1.8 1.9 2.3 1.8 2.0 2.0 2.2 2.0 0.2 0.05VSI 12.4 12.1 12.1 11.6 11.5 11.5 11.4 12.6 0.5 0.17

Cobiab

Initial weight (g) 57.2 57.8 56.9 57.1 56.8 57.3 57.3 57.0 3.8 0.31Final weight (g) 217.9 zy 221.3 zy 203.1 zy 249.4 z 238.2 zy 206.9 zy 242.4 zy 186.2 y 16.9 0.02Weight gain (%) 281 zy 283 zy 257 zy 337 z 333 z 273 zy 323 z 227y 26 0.01SGR (% body weight/d) 2.4 zy 2.4 zy 2.3 zy 2.7 z 2.7 z 2.4 zy 2.6 z 2.2 y 0.1 <0.01FCR 1.7 zy 1.6 zy 1.8 zy 1.5 y 1.5 y 1.7 zy 1.6 y 1.9 z 0.1 <0.01FI (% body weight/d) 4.5 4.4 4.4 4.3 4.3 4.5 4.5 4.3 0.2 0.86HSI 1.9 1.7 1.8 1.6 1.8 1.7 1.7 3.4 0.9 0.45

aAlthough the omnibus test indicated a significant treatment effect, the more conservative Tukey’s HSD pairwise comparison test failed to identify differences among means.bVSI was not assessed in Cobia.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5


TABLE 5. Growth performance (final weight, weight gain, SGR, and FCR), FI, HSI, and VSI of hybrid Striped Bass, Rainbow Trout, and Cobia fed diets inwhich 20, 40, or 60% of digestible protein consisted of fish meal provided by Asian carp meal (CFM) or menhaden fish meal (MFM), excluding the 60 Blend.Values represent least-squared means; pooled SEs and P-values resulting from two-way ANOVA tests are also provided. Letters indicate significance differences(P < 0.05). The interaction term was not significant for all parameters; it was removed from the analysis and only the main effects were tested and reported.

Fish meal type Inclusion level

Parameter MFM CFM Pooled SE P-value 20% 40% 60% Pooled SE P-value

Hybrid Striped BassInitial weight (g) 22.0 21.9 0.1 0.49 21.8 22.0 21.9 0.1 0.37Final weight (g) 63.9 64.8 2.5 0.70 59.0 y 61.4 y 72.7 z 3.0 <0.01Weight gain (%) 191 196 11 0.66 170 y 179 y 232 z 14 <0.01SGR (% body weight/d) 1.9 2.0 0.1 0.62 1.8 y 1.9 y 2.2 z 0.1 <0.01FCR 1.1 1.1 0.0 0.66 1.1 1.1 1.1 0.1 0.95FI (% body weight/d) 2.2 2.3 0.1 0.33 2.1 2.2 2.6 0.1 <0.01HSI 1.3 1.4 0.1 0.13 1.3 1.4 1.5 0.1 0.08VSI 6.9 7.3 0.3 0.32 7.2 7.2 6.9 0.4 0.79

Rainbow TroutInitial weight (g) 15.0 15.1 0.1 0.72 15.1 15.1 15.0 0.1 0.74Final weight (g) 43.3 45.8 2.7 0.37 41.2 45.5 47.0 3.3 0.23Weight gain (%) 188 203 18 0.41 173 201 213 22 0.21SGR (% body weight/d) 1.7 1.8 0.1 0.51 1.6 1.8 1.7 0.2 0.55FCR 1.0 1.0 0.1 0.68 1.1 1.0 1.0 0.1 0.32FI (% body weight/d) 2.0 2.1 0.1 0.47 2.0 2.0 2.1 0.1 0.80HSI 2.0 1.9 0.1 0.42 2.2z 1.9 zy 1.8 y 0.1 0.03VSI 12.2 z 11.5 y 0.3 0.02 11.8 11.8 12.0 0.3 0.80

Cobiaa

Initial weight (g) 57.3 57.1 2.4 0.30 57.1 57.3 57.1 2.9 0.82Final weight (g) 214.1 231.5 10.2 0.11 205.0 229.7 233.6 12.5 0.08Weight gain (%) 273 y 314 z 15 0.02 265 308 309 19 0.05SGR (% body weight/d) 2.4 y 2.6 z 0.1 0.02 2.4 2.5 2.5 0.1 0.04FCR 1.7 z 1.6 y 0.0 <0.01 1.7 z 1.6 y 1.6 y 0.1 <0.01FI (% body weight/d) 4.4 4.4 0.1 0.69 4.5 4.4 4.4 0.1 0.67HSI 1.8 1.7 0.1 0.29 1.7 1.7 1.8 0.1 0.98

aVSI was not analyzed in Cobia.

improved growth with higher fish meal inclusion were evident.Growth of Cobia varied significantly among dietary treatmentgroups and was influenced by fish meal type and inclusion level.Although not statistically significant in each study, growth wasgenerally greater among fish fed the CFM-based feeds, particu-larly at the higher inclusion levels. Overall, similar trends wereobserved among studies (i.e., better growth with higher fish mealinclusion and CFM).

For hybrid Striped Bass, FCR, HSI, and VSI did not varyamong dietary treatments and were unaffected by fish meal typeor inclusion level. However, FI increased with inclusion level,regardless of fish meal type. In Rainbow Trout, FCR, FI, HSI,and VSI did not vary among dietary treatments. Additionally,FCR and FI were not influenced by fish meal type or inclusionlevel, but Rainbow Trout HSI was significantly increased amongfish fed diets with lower fish meal inclusion levels. Furthermore,although VSI was unaffected by fish meal inclusion level, it was

affected by fish meal type. Rainbow Trout fed diets containingMFM had higher VSI. Cobia FI and HSI did not vary amongdietary treatments and were unaffected by fish meal type orinclusion level; however, FCR was significantly reduced amongCobia fed CFM diets and diets with higher fish meal inclusionlevels.

DISCUSSIONOur results indicate that growth performance is broadly con-

sistent among hybrid Striped Bass, Rainbow Trout, and Co-bia fed MFM or CFM. This suggests that CFM (∼$650/metricton; P. Hitchens, Southern Illinois University Carbondale, per-sonal communication) is a cost-effective alternative to tradi-tional marine-origin fish meals ($1,500–2,000/metric ton: FAO2014). Given that, substituting traditional fish meals with CFMin the diets of other carnivorous fishes will not likely affectgrowth performance negatively. Any differences in growth per-

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5


formance associated with utilizing CFM will likely be minimaland the result of nuances in dietary needs and formulations.Therefore, CFM appears to be a suitable, cost-effective substi-tute to traditional marine-origin fish meals in the diets of carniv-orous fish. Published information comparing fish meal types islimited, but appears generally supportive. For example, in a sim-ilar study with juvenile Largemouth Bass fed diets containingCFM, there were no significant differences in growth perfor-mance when compared with menhaden fish meal (Bowzer et al.2014). Kop and Korkut (2010) performed a comparable studywith juvenile Rainbow Trout assessing three different marine-based fish meals: Peruvian fish meal (traditional commercialfish meal), locally produced (Izmir, Turkey) anchovy fish meal,and locally produced fish meal from the by-products of severalspecies (Gilthead Sea Bream Sparus aurata, Sea Bass Dicen-trarchus labrax, and Rainbow Trout). Their results indicatedthat growth performance was not influenced by fish meal type.

As long as the product used is fresh, properly stored, and ofhigh quality, the effect of fish meal type on growth performanceof fish seems minor. The level of fish meal inclusion, however,appears to have a strong influence on feed intake and growthperformance. Typically, the more fish meal that is includedin the diet, the better the growth. Yet, due to its variable butgenerally high cost, fish meal sparing is common practice indiet formulations for many species including hybrid StripedBass (Rawles et al. 2011), Rainbow Trout (Baboli et al. 2013),and Cobia (Trushenski et al. 2013). These studies have reportedsuccess (i.e., similar growth performance among fish fed fishmeal-based and alternative protein-based feeds) in partiallyreplacing fish meal with a variety of alternative plant- andanimal-based protein meals provided that some amount offish meal was included in the diet. Correspondingly, our datasuggest that some inclusion of fish meal, regardless of origin,is important to ensure rapid, efficient growth. This also impliessome fish meal sparing is possible without affecting growthperformance. Therefore, it is not unexpected that there werelittle to no significant differences in growth provided that asufficient amount of fish meal was included in the diet.

Feed conversion and intake values were within expectationsfor each study. However, Cobia demonstrated better feed con-version when fed diets containing CFM and higher inclusionlevels of fish meal. This better feed conversion may explain thegreater growth performance observed in Cobia fed diets con-taining CFM and higher inclusion levels. Minor differences infeed intake can certainly influence growth performance, butgiven that no clear trends emerged, suggesting reduced in-take occurred due to fish meal origin or lower inclusion levels,better feed conversion seems to be a reasonable explanation forthe improved growth observed with CFM and higher fish mealinclusion diets in Cobia. Although the nutritional compositionof CFM and MFM are relatively similar, as demonstrated byBowzer et al. (in press), small differences in the crude proteincontent and amino acid profile of these meals may explain theslightly better FCR observed in this study. Additionally, better

feed conversion with higher inclusion levels of fish meal in thediet is typical of many carnivorous fishes, and this was clearlydemonstrated by Salze et al. (2010) who replaced fish meal withgraded levels of soy protein concentrate in the diets of juvenileCobia. Therefore, the better FCR values for diets having higherfish meal inclusion was not surprising, and the small differencebetween fish meal types is likely due to minor differences in thecomposition of the meals.

There were no apparent trends for hybrid Striped Bassor Cobia in relation to organosomatic indices, but RainbowTrout were leaner (i.e., lower VSI) when fed CFM-based diets.Additionally, Rainbow Trout fed diets with lower inclusion lev-els of fish meal had slightly larger livers than fish fed dietswith higher inclusion levels of fish meal. The lower fat deposi-tion in Rainbow Trout fed CFM diets can at least partially beattributed to those diets having less crude lipid coupled withhigher ash content compared with the MFM diets. These dis-crepancies in the diets are relatively small, but likely contributedto the minor differences in lipid deposition observed betweenfish meal types. It is also well documented in Rainbow Troutthat liver size varies with the digestible carbohydrate content ofthe diet (Kaushik et al. 1989; Escaffre et al. 2007) due to anapparent inefficiency to utilize this macronutrient (Skiba-Cassyet al. 2013); however, it is not likely this is the key factor thatcontributed to differences in liver size in this study since thecarbohydrate content remained similar among the diets. Poultryby-product was the main ingredient to replace fish meal in thelower inclusion level diets, and therefore, it is likely the keyfactor contributing to the variation in liver size. Steffens (1994)demonstrated that methionine and lysine supplementation wasnecessary to maintain proper growth in Rainbow Trout whensparing or substituting fish meal with poultry by-product. Al-though our diets were formulated to meet all known nutrientrequirements of Rainbow Trout and were therefore not supple-mented with additional crystalline amino acids, differences inessential and nonessential amino acid availability due to higherinclusion levels of poultry by-product meal may have causedthe variation in liver size observed in this study (Steffens 1994).

In conclusion, CFM appears to be a cost-effective alterna-tive protein source to traditional marine-origin fish meal in thediets of carnivorous fishes such as hybrid Striped Bass, Rain-bow Trout, and Cobia. Some subtle differences in growth andorganosomatic indices were observed among the taxa we inves-tigated, but these differences were marginal. Caution should betaken when considering full-scale, commercial production giventhat these fish were reared under optimal conditions for only an8-week feeding trial. Further long-term studies should be con-ducted under conditions typical of intensive commercial cultureof these fish (i.e., temperature and oxygen fluctuations, stress,and disease), which may adversely affect growth or survival, butdue to the similarity of CFM to traditional marine-origin fishmeals, we believe these concerns are relatively minor. There-fore, further development of a CFM industry could not onlyproduce a cost-effective alternative to traditional marine-origin

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5


fish meals, but also encourage harvest of invasive Asian carp inU.S. waterways.

ACKNOWLEDGMENTSWe extend our thanks to the Illinois Soybean Association for

supporting this research project under grant 12-10-59-240-550-10. We also thank Omega Protein, Tyson, and Darling Interna-tional for the donation of feedstuffs used to prepare the feedsevaluated in this work. We thank Crystal Lakes Fisheries for thedonation of Rainbow Trout fingerlings used in this work. PaulHitchens was instrumental in arranging the manufacturing anddelivery of the Asian carp meal to our facility. We also thankMichael Schwarz, Steve Urick, and their staff at the VirginiaSeafood Agricultural Research and Extension Center for con-ducting the Cobia feeding trial. Finally, we thank Jonah May,Michael Page, Chris Jackson, and Kelli Barry of the CFAAS forhelp with data collection and analysis.

REFERENCESAdelizi, P. D., R. R. Rosati, K. Warner, Y. V. Wu, T. R. Muench, M. R. White,

and P. B. Brown. 1998. Evaluation of fish-meal free diets for Rainbow Trout,Oncorhynchus mykiss. Aquaculture Nutrition 4:255–262.

Baboli, M. J., M. Dawodi, and A. Gorjipor. 2013. Effect of replacement fishmeal by poultry meal on growth, survival, and body composition of RainbowTrout (Oncorhynchus mykiss). International Research Journal of Applied andBasic Sciences 5:296–300.

Barrows, F. T., T. F. Gaylord, W. Sealey, and S. D. Rawles. 2012.Database of nutrient digestibilities of 355 traditional and novel feed in-gredients for trout and hybrid Striped Bass. Available: http://www.ars.usda.gov/Main/docs.htm?docid=21905. (October 2014).

Bowzer, J. C., A. Bergman, and J. T. Trushenski. 2014. Growth performance ofLargemouth Bass Micropterus salmoides fed fish meal derived from Asiancarp. North American Journal of Aquaculture 76:85–89.

Bowzer, J. C., J. T. Trushenski, and D. C. Glover. 2013. Potential of Asian carpfrom the Illinois River as a source of raw materials for fish meal production.North American Journal of Aquaculture 75:404–415.

Bowzer, J. C., J. T. Trushenski, S. Rawles, T. G. Gaylord, and F. T. Barrows.In press. Apparent digestibility of Asian carp- and common carp-derived fishmeals in feeds for hybrid Striped Bass Morone chrysops ♀ × M. saxatilis♂ and Rainbow Trout Oncorhynchus mykiss. Aquaculture Nutrition. DOI:10.1111/anu.12136.

Cheng, Z. J., R. W. Hardy, and H. J. Huige. 2004. Apparent digestibility coeffi-cients of nutrients in brewer’s and rendered animal by-products for RainbowTrout (Oncorhynchus mykiss (Walbaum)). Aquaculture Research 35:1–9.

Conover, G., R. Simmonds, and M. Whalen. 2007. Management and controlplan for Asian carps in the United States. Aquatic Nuisance Species TaskForce, Asian Carp Working Group, Washington, D.C.

Davies, S., and P. Morris. 1997. Influence of multiple amino acid supplementa-tion on the performance of Rainbow Trout, Oncorhynchus mykiss (Walbaum),fed soya based diets. Aquatic Research 28:65–74.

Escaffre, A., S. Kaushik, and M. Mambrini. 2007. Morphometric evaluation ofchanges in the digestive tract of Rainbow Trout (Oncorhynchus mykiss) dueto fish meal replacement with soy protein concentrate. Aquaculture 273:127–138.

FAO (Food and Agriculture Organization of the United Nations). 2012. Thestate of world fisheries and aquaculture. FAO, Rome.

FAO (Food and Agriculture Organization of the United Nations). 2014. Com-modity price index. FAO, Economic and Social Development Department,Rome. Available: http://www.fao.org/economic/est/prices. (March 2014).

Gatlin, D. III., F. Barrows, D. Bellis, P. Brown, J. Capen, K. Dabrowski, T. G.Gaylord, R. W. Hardy, E. M. Herman, G. Hu, A. Krogdahl, R. Nelson, K.E. Overturf, M. Rust, W. Sealey, K. Skonberg, E. J. Souza, D. Stone, and R.F. Wilson. 2007. Expanding the utilization of sustainable plant products inaquafeeds: a review. Aquaculture Research 38:551–579.

Glencross, B. D., M. Booth, and G. L. Allan. 2007. A feed is only as good as itsingredients-a review of ingredient evaluation strategies for aquaculture feeds.Aquaculture Nutrition 13:17–34.

Heikkinen, J., J. Vielma, O. Kemilainen, M. Tiirola, P. Eskelinen, T. Kiuru, D.Navia-paldanius, and A. von Wright, 2006. Effects of soybean meal baseddiet on growth performance, gut histopathology and intestinal microbiota ofjuvenile Rainbow Trout (Oncorhynchus mykiss). Aquaculture 261:259–268.

Hendricks, J. D. 2003. Adventidious toxins. Pages 602–641 in J. E. Halver andR. W. Hardy, editors. Fish nutrition, 3rd edition. Academic Press, New York.

Iwashita, Y., T. Yamamoto, H. Furuita, T. Sugita, and N. Suzuki. 2008. Influenceof certain soybean antinutritional factors supplemented to a casein-basedsemipurified diet on intestinal and liver morphology in fingerling RainbowTrout Oncorhynchus mykiss. Fisheries Science 74:1075–1082.

Kaushik, S. J., F. Medale, B. Fauconneau, and D. Blanc. 1989. Effectof digestible carbohydrates on protein/energy utilization and on glucosemetabolism in Rainbow Trout (Salmo gairdneri R.). Aquaculture 79:63–74.

Keembiyehetty, C. N., and D. M. Gatlin III. 1992. Dietary lysine requirement ofjuvenile hybrid Striped Bass (Morone chrysops × M. saxatilis). Aquaculture104:271–277.

Kop, A., and A. Y. Korkut. 2010. Effects of diets with difference fish meal originson the performance of Rainbow Trout Oncorhynchus mykiss (Walbaum).Journal of Animal and Veterinary Advances 9:581–583.

Lodge, D. M., S. Williams, H. J. MacIsaac, K. R. Hayes, B. Leung, S. Reichard,R. N. Mack, P. B. Moyle, M. Smith, D. A. Andow, J. T. Carlton, and A.McMichael. 2006. Biological invasions: recommendations for U.S. policyand management. Ecological Applications 16:2035–2054.

McClelland, M. A., G. G. Sass, T. R. Cook, K. S. Irons, N. N. Michaels, T. M.O’Hara, and C. S. Smith. 2012. The long-term Illinois River fish populationmonitoring program. Fisheries 37:340–350.

Moon, J. 2014. Fish plant begins equipment testing. The Telegraph(April 30). Available: www.thetelegraph.com/news/news/1293393/fish-plant-begins-equipment-testing#.u90iy2jnbqa. (August 2014).

Nelson, L. R. 2013. Analysis of factors that affect the meat quality of invasiveAsian carps harvested from the Illinois River for export to China. Master’sthesis. Southern Illinois University, Carbondale.

NRC (National Research Council). 2011. Nutrient requirements of fish andshrimp. National Academy Press, Washington, D.C.

Rawles, S. D., K. R. Thompson, Y. J. Brady, L. S. Metts, M. Y. Aksoy, A.L. Gannam, R. G. Twibell, S. Ostrand, and C. D. Webster. 2011. Effectsof replacing fish meal with poultry by-product meal and soybean meal andreduced protein level on the performance and immune status of pond-grownsunshine bass (Morone chrysops × M. saxatilis). Aquaculture Nutrition17:708–721.

Rombenso, A., C. Crouse, and J. T. Trushenski. 2013. Comparison of traditionaland fermented soybean meals as alternatives to fish meal in hybrid StripedBass feeds. North American Journal of Aquaculture 75:197–204.

Salze, G., E. McLean, P. Rush Battle, M. H. Schwarz, and S. R. Craig. 2010.Use of soy protein concentrate and novel ingredients in the total eliminationof fish meal and fish oil in diets for juvenile Cobia, Rachycentron canadum.Aquaculture 298:294–299.

Skiba-Cassy, S., S. Panserat, M. Larquier, K. Dias, A. Surget, E. Plagnes-Juan,S. Kaushik, and I. Seiliez. 2013. Apparent low ability of liver and muscleto adapt to variation of dietary carbohydrate:protein ratio in Rainbow Trout(Oncorhynchus mykiss). British Journal of Nutrition 109:1359–1372.

Steffens, W. 1994. Replacing fish meal with poultry by-product meal in dietsfor Rainbow Trout, Oncorhynchus mykiss. Aquaculture 124:27–34.

Tacon, A. G. J., M. R. Hasan, and M. Metian. 2011. Demand and supply of feedingredients for farmed fish and crustaceans. FAO Fisheries and AquacultureTechnical Paper 564.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5


Tacon, A. G. J., and M. Metian. 2008. Global overview on the use of fishmeal and fish oil in industrially compounded aquafeeds: trends and futureprospects. Aquaculture 285:146–158.

Trushenski, J. T., M. Schwarz, W. V. N. Pessoa, B. Mulligan, C. Crouse,B. Gause, F. Yamamoto, and B. Delbos. 2013. Amending reduced fish-meal feeds with marine lecithin, but not soy lecithin, improves thegrowth of juvenile Cobia and may attenuate heightened responses to stresschallenge. Journal of Animal Physiology and Animal Nutrition 97:170–180.

Tsehaye, I., M. Ctalano, G. Sass, D. Glover, and B. Roth. 2013. Prospects forfishery-induced collapse of invasive Asian carp in the Illinois River. Fisheries38:445–454.

Varble, S., and S. Secchi. 2013. Human consumption as an invasive speciesmanagement strategy, a preliminary assessment of the marketing potential ofinvasive Asian carp in the US. Appetite 65:58–67.

Vitousek, P. M., C. M. D’Antonio, L. L. Loope, M. Rejmanek, and R.Westbrooks. 1997. Introduced species: a significant component of human-caused global environmental change. New Zealand Journal of Ecology 21:1–16.

Watanabe, T. 2002. Strategies for further development of aquatic feeds. FisheriesScience 68:242–252.

Welch, A., R. Hoenig, J. Stieglitz, D. Benetti, A. Tacon, N. Sims, and B.O’Hanlon. 2010. From fishing to the sustainable farming of carnivorousmarine finfish. Reviews in Fisheries Science 18:235–247.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5


Click for updates


Positive Correlation Between Inhibition of Branchialand Renal Carbonic Anhydrase and Ammonia Producedby Cultured Silver Catfish Rhamdia quelenLuciana R. Souza-Bastosa, Leonardo P. Bastosb & Carolina A. Freirea

a Departamento de Fisiologia, Setor de Ciências Biológicas, Universidade Federal do Paraná,Curitiba, Paraná, Brazilb Instituto de Tecnologia para o Desenvolvimento, Curitiba, Paraná, BrazilPublished online: 16 Dec 2014.

To cite this article: Luciana R. Souza-Bastos, Leonardo P. Bastos & Carolina A. Freire (2015) Positive Correlation BetweenInhibition of Branchial and Renal Carbonic Anhydrase and Ammonia Produced by Cultured Silver Catfish Rhamdia quelen,North American Journal of Aquaculture, 77:1, 68-75, DOI: 10.1080/15222055.2014.960118








http://dx.doi.org/10.1080/15222055.2014.960118




ARTICLE

Positive Correlation Between Inhibition of Branchialand Renal Carbonic Anhydrase and Ammonia Producedby Cultured Silver Catfish Rhamdia quelen

Luciana R. Souza-BastosDepartamento de Fisiologia, Setor de Ciencias Biologicas, Universidade Federal do Parana, Curitiba,Parana, Brazil

Leonardo P. BastosInstituto de Tecnologia para o Desenvolvimento, Curitiba, Parana, Brazil

Carolina A. Freire*Departamento de Fisiologia, Setor de Ciencias Biologicas, Universidade Federal do Parana, Curitiba,Parana, Brazil

AbstractSilver Catfish Rhamdia quelen, a native South American catfish relevant in intensive culture systems in Brazil, were

exposed to three ammonia concentrations added to the water (0.5, 1.0, and 2.0 mg/L as ammonium chloride). After 5and 24 h, measured water ammonia levels increased to maximum levels of 15.8–41.1 mg/L, reflecting the addition ofammonia excreted by the fish. Aquaria were aerated but kept closed, and pH and temperature were kept constant.Hematocrit, plasma ammonia, osmolality, Na+ , Cl−, K+ , glucose, and cortisol were assayed. Branchial and musclehydration levels and branchial and renal specific activities of the carbonic anhydrase (CA) were determined. SilverCatfish did not show stress responses or signs of osmoregulatory disturbance upon the increased water ammonialevels. However, there was inhibition of the CA in gills and kidneys, especially after 24 h, in a dose-dependent mannerto the total ammonia accumulated in the water. Although Silver Catfish are tolerant of increase in ammonia, careshould be taken to limit increases in ammonia levels and time of exposure because the severity of the deleteriouseffects will certainly increase. In addition, results have shown that the fish carbonic anhydrase (branchial and renal)is a sensitive biomarker of effect of ammonia.

Most of the ammonia produced in teleost fish is excretedthrough the gill epithelium via passive diffusion of the non-ionized form (NH3; Wilkie 1997, 2002; Evans et al. 2005; Ipand Chew 2010). Increases in ammonia concentrations are arather common outcome of intensive farming systems, poten-tially compromising water quality and fish health (Hargreavesand Kucuk 2001; Pereira and Mercante 2005; Carneiro et al.2009a; Ip and Chew 2010). Ammonia levels in fish ponds arehighly variable and hard to control because they depend onthe density and metabolic rate of the animals, on the type and


amount of food administered, and on the quality and frequencyof tank-cleaning protocols (Hargreaves and Kucuk 2001; Pereiraand Mercante 2005; Liu et al. 2008; Miron et al. 2008).

Excess ammonia is toxic to most teleost fishes, although withremarkable variability in the tolerances and different experimen-tal conditions (e.g., Wilkie 1997, 2002; Randall and Tsui 2002;Ip and Chew 2010). Ammonia toxicity may be potentiated whenassociated to other stress factors, such as alterations in pH,temperature, salinity, time of transport, density, and handling(Wilkie 1997, 2002; Randall and Tsui 2002; Miron et al. 2008;

68

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5

AMMONIA INHIBITS FISH CARBONIC ANHYDRASE 69

Carneiro et al. 2009a). The effects of internal accumulation ofammonia in freshwater fish have been extensively studied andinvolve disturbances in all physiological systems (e.g., Randalland Tsui 2002; Ip and Chew 2010). Ammonia inhibits the activ-ity of the enzyme carbonic anhydrase (CA), thus damaging vitalprocesses such as osmoregulation, gas exchange, excretion, andacid–base balance (ArasHisar et al. 2004). High internal lev-els of ammonia may even hamper survival and growth of thefish (Wilkie 1997, 2002; Hargreaves and Kucuk 2001; Ran-dall and Tsui 2002; Pereira and Mercante 2005; Miron et al.2008). Despite the growing interest in the CA as a pollutionbiomarker, its inhibition has been essentially related to the pres-ence of heavy metals and pesticides/herbicides (Lionetto et al.2012).

A South American native, the Silver Catfish Rhamdia quelen(Quoy and Gaimard 1824; Heptapteridae; also popularly knownas jundia) is widely distributed across the continent and is con-sidered a promising species in intensive fish culture systems,mainly in southern Brazil (Barcellos et al. 2001; Barcellos et al.2004a, 2004b; Miron et al. 2008; Carneiro et al. 2009a, 2009b;Souza-Bastos and Freire 2009). This species is quite tolerant ofculture conditions, has good fertilization rate and development,excellent commercial acceptance, and has been considered forreplacement in the culture of exotic species (Barcellos et al.2001, 2004b; Golombieski et al. 2003; Carneiro et al. 2009a,2009b; Souza-Bastos and Freire 2009). Studies on the effect ofammonia on this species have been restricted to lethal concen-tration determinations that have indicated variation in ammoniatoxicity at different levels of pH, temperature, or stocking den-sity (Golombieski et al. 2003; Miron et al. 2008; Carneiro et al.2009a, 2009b).

Our study objective was to determine the effects of increasedwater ammonia on branchial and renal carbonic anhydrase ac-tivity and on other osmoregulatory and stress markers in SilverCatfish. The focus was on the putative effects of ammonia inclosed systems, without additional effects or interaction be-tween ammonia and other stress factors, such as increased den-sity, transport, changes in pH, and water temperature. Increaseddoses of water ammonia were obtained through the initial addi-tion of low ammonia concentrations, accompanied by ammoniaproduced by the experimental fish.

METHODSAnimals and laboratory maintenance.—Silver catfish juve-

niles (mean = 57.7 g, SD = 1.6; 18.9 cm TL, SD = 0.2; n =64) were obtained from culture facilities of the Pontifıcia Uni-versidade Catolica do Parana (PUC-PR) (25◦35′S, 49◦13′W) inSao Jose dos Pinhais, Parana, Brazil. Fishes were transported(about 40 min) in local freshwater and under constant aera-tion to the Laboratory of the Federal University of Parana inCuritiba.

In the laboratory, fish were placed in two stock tanks of∼250 L containing dechlorinated tap water with constant aer-

ation, biological filtration, a natural photoperiod of about 12 hlight : 12 h dark, and temperature of 20◦C (SD, 2). In each 250-Ltank stocking density was 40 fish (or 0.16 fish/L) as approxi-mately that recommended for optimal growth for Silver Catfishof this same size-class (0.1 fish/L in Barcellos et al. 2004b). Eachstock tank received a total of 15 mL (0.0675 g/mL) of acriflavinechloridrate (Aqualife Labcon) and 5 g of common salt (sodiumchloride) for parasite removal, in a single event. This salt concen-tration presents no osmotic stress to this species (Souza-Bastosand Freire 2009). After a week of daily replacements of 50% ofthe tank water (always adding plain dechlorinated freshwater)with no medication or salt, we believed that fish were free of themedication, the salt, and potential infections. They remained inthe stock tanks containing plain freshwater, where they accli-mated for additional 3 weeks. During these 21 d, 50% of thewater was replaced every other day. Fish were fed daily, alwaysin the morning, a total amount of about 20 g of commercialfood used for this species (extruded, Supra, 32% protein). Afterallowing 40 min for fish to feed, the unconsumed food was re-moved, and the walls and filters of the tanks were cleaned. Nomortality was recorded during the acclimation period.

Experiments and blood and tissue sampling.—After the ac-climation period, fish were divided into eight groups: four ex-perimental conditions, each at a 5-h period of exposure and eachat 24-h period. The four experimental exposure conditions wereno initial addition of ammonia (control or D0) and initial expo-sure to doses of ammonia (NH4Cl/L) at 0.5 mg/L of water (D1),1.0 mg/L (D2), and 2 mg/L (D3), each dose being tested at 5 and24 h. As each condition (ammonia concentration and time) wastested in four replicates, requiring 32 aquaria (each, 20 L), eachstocked with two similarly sized fish, so eight fish were testedfor each conditions group. This fish density is not stressful forthe about 60-g Silver Catfish, which were grown from fry (Bar-cellos et al. 2004b). These added levels of ammonia result inmuch lower levels of NH3 (nonionized ammonia) than the levelreported as lethal for Silver Catfish juveniles by Carneiro et al.(2009b): NH3 at1.9 mg/L of water = LC50 at 96 h.

The experimental tanks were prepared about 20 h beforeadding the fish, in order to assure that the water was entirelydechlorinated. Constant aeration was provided before and af-ter placing the fish. Fish were not fed during the 24-h periodbefore being placed in the experimental aquaria. During the ex-periments, aquaria remained covered but under aeration. Watertemperature was measured, and one water sample was collectedat the beginning (T0) at 5 h, and at 24 h from each aquarium.The water samples were measured for pH and total ammonia(Boyd and Tucker 1992).

In all groups, ammonia levels built during the periods ofexposure because significant amounts of ammonia were pro-duced and excreted by the fish to the water, increasing levelsof sublethal ammonia. The fraction of un-ionized ammonia inthe water samples was calculated using the formula suggestedby Liu et al. (2008), which uses the equivalence coefficient(Keq) determined by Emerson et al. (1975). Levels of total

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5

70 SOUZA-BASTOS ET AL.

FIGURE 1. Relationship between added ammonia as NH4Cl and (A) total and(B) un-ionized ammonia levels measured in the aquaria water holding SilverCatfish exposed to experimental conditions: the addition of 0 (D0), 0.5 (D1), 1(D2), or 2 (D3) mg/L of NH4Cl, where ammonia was measured initially (0 h),and after 5 and 24 h after placing fish in the aquaria. The dotted lines (trends ofdirect relationship between added ammonia and measured ammonia levels, forthe same exposure time) and solid lines (trends of steeply increasing ammonialevels over time for each experimental condition) were fitted by eye. Increasedammonia levels over time are noted even in the D0 group which received noinitial addition of ammonia, as a result of ammonia produced by the fish. Theobvious difference between the metabolically produced ammonia and the addedammonia is represented by the solid (metabolic, across time) and dotted (added,across groups) arrows. Pearson correlation coefficients and probability levelsare written to the right of dotted lines (N = 4 for each correlation); asterisksdenote significant correlations.

ammonia and unionized ammonia increased drastically alongtime for all experimental conditions, and the addition of am-monium chloride also led to proportionally higher levels ofammonia in the water (Figure 1). Significant correlations be-tween amount of initially added ammonia and measured ammo-nia levels (both total and un-ionized ammonia) occurred onlyfor time = 0 (initial conditions), confirming the crescent lev-els of added ammonia along the experimental conditions. Totalammonia levels after 24 h almost significantly correlated (Pear-

TABLE 1. The pH of aquaria water in which Silver Catfish were exposedto experimental levels of ammonia (NH4Cl in mg/L of water): 0 (control orD0), 0.5 (D1), 1 (D2), or 2.0 (D3) at the start of the experiment (T0) and after5 and 24 h of holding. No significant differences were found among the fourexperimental conditions. Water temperature was kept constant at 18◦C (SE, 0.1)for all aquaria throughout the experiment.

Experimental condition Time Mean pH ± SE

D0 T0 7.3 ± 0.165 h 7.3 ± 0.0724 h 7.4 ± 0.04

D1 T0 7.3 ± 0.105 h 7.2 ± 0.1024 h 7.4 ± 0.03

D2 T0 7.3 ± 0.085 h 7.2 ± 0.0924 h 7.4 ± 0.08

D3 T0 7.3 ± 0.055 h 7.2 ± 0.0424 h 7.3 ± 0.01

son coefficient, 0.947) with initial added ammonia (P = 0.053;Figure 1A). For all groups, individual variation in amount ofammonia produced and excreted during the 5 or 24 h of theexperiments caused variability in the levels measured and, thus,reduced the power and significance of the correlations. Still,the increasing levels of ammonia along the duration of the ex-posure and across treatments were very clear (Figure 1). Lev-els of pH, on the contrary, did not vary during the course ofthe experiments, and for all concentrations of ammonia tested(Table 1).

At the end of each experimental period (5 or 24 h), fishwere anesthetized individually with benzocaine (80 mg/L ofwater) until total absence of response to touch stimuli (at about2 min). Fish were measured (total length) and weighed (totalweight) and a blood sample (about 1 mL) was taken throughpuncture of the caudal vein using heparinized (Parinex, sodiumheparin) insulin syringes. The blood sample was immediatelycentrifuged for 5 min (about 3,000 × g) at room temperature;plasma obtained was kept at −20◦C until assayed for ammonia,osmolality, ions, glucose, and cortisol. Another small bloodsample was collected and deposited directly into heparinizedcapillary tubes for hematocrit determination. The fish were theneuthanized by spinal cord severance and tissues (gills, kidneysand muscle) were dissected. Water content in the in the largestgill arch (right side) and an axial muscle fragment (right side)was measured immediately after their removal from the fish.The remaining gill arches (both sides) and the entire kidneywere stored at −80◦C until assayed for the specific activity ofthe enzyme carbonic anhydrase. This experimental protocol wasapproved by the Ethics Committee on Animal Experimentationof the Federal University of Parana (certificate n◦241 issued onAugust 9, 2007).

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5


Blood and plasma determinations.—The blood sampled inheparinized capillaries tubes (Perfecta LTDA) was centrifugedat 3,000 × g for 10 min (Presvac CMH 28), and read across ahematocrit scale. The concentration of total ammonia was deter-mined in undiluted samples according to the protocol establishedby Verdouw et al. (1978), with absorbance read at 650 nm (UL-TROSPEC 2100 pro–Amersham Pharmacia Biotech). Plasmaosmolality was determined in undiluted samples (vapor pres-sure micro-osmometer Wescor VAPRO 5520). Chloride con-centrations were read in undiluted samples using a colorimet-ric method (Labtest, Brazil) with absorbance read at 470 nm,and Na+ and K+ were assayed in diluted samples (1:100,Ultra-pure water) through flame photometry (CELM FC–180,Brazil). Glucose concentration was assayed using a colorimetricmethod (Labtest, Brazil) in undiluted samples, with absorbanceread at 505 nm. Cortisol was determined in undiluted plasmasamples using a colorimetric microplate method (Human Diag-nostics Worldwide), with absorbance readings at 450 nm andreference wavelength of 630 nm (ELISA Tecan Infinite M200Chisto R.).

Branchial and muscular hydration.—Water content percent-ages were determined in 64 gills (mean of mass = 1.19 g, SE =0.01) and muscle slices (1.45 g, SE = 0.01). Tissue fragmentswere thawed and weighed (wet weight) in eppendorf tubes (bal-ance Bioprecisa FA2104 N, Brazil, precision = 0.1 mg). Sam-ples were then dehydrated for 24 h at 60◦C and were weighedagain (dry weight). Water weight loss was expressed as thepercentage of the initial wet weight of the sample.

Carbonic anhydrase specific activity.—Gill (mean =0.0980 g, SE = 0.003) and kidney (0.1431 g, SE = 0.005)samples (n = 64) were assayed for the carbonic anhydrase (CA)specific activity (Henry 1991; Vitale et al. 1999). The CA ac-tivity was detected (catalyzed rate; CR) through pH reductionover 20 s of measurement time (pH meter inoLAB pH Level1WTW, Germany), in 4◦C deionized water saturated with CO2,in the presence of tissue homogenized in phosphate buffer (pH7.5). Non-catalyzed rate (NCR) was quantified by the slope ofthe regression line of pH reduction in the absence of tissuehomogenate. Carbonic anhydrase activity (CAA) was calcu-lated based on Burnett et al. (1981) and Vitale et al. (1999):CAA = [(CR/NCR) − 1]/mg of total protein. Total proteincontent of tissue homogenates was determined according toBradford (1976), with absorbance readings at 595 nm (ELISATecan Infinite M200).

Statistical analysis.—Abiotic data from the aquaria wa-ter and biotic data from fish were analyzed through two-wayANOVA (total ammonia concentration × exposure time) fol-lowed by the Tukey post hoc analysis for locating differences.Pearson’s product moment correlation coefficient was used toassess the relationship between (1) initially added ammonia andmeasured ammonia levels at 0, 5, and 24 h, and (2) CA ac-tivities and levels of total ammonia measured in the aquariaafter 5 and 24 h. For all tests, the level of significance was α =0.05.

RESULTS

Blood and Plasma DeterminationsHematocrit (24–29%) and total plasma ammonia (3.75–

5.86 mg/L) of Silver Catfish remained unchanged upon am-monia addition throughout the experiments (Table 2). Plasmaosmolality ranged between 247 and 271 mOsm/kg of water.The only change was noted in the group that initially received2.0 mg/L of NH4Cl; it increased from 253 mOsm/kg after 5 hto 271 mOsm/kg after 24 h (Table 2).

As with hematocrit and total ammonia, plasma chloride(106–117 mM), potassium (2.5–3.0 mM), and cortisol (61–139 ng/mL) of Silver Catfish also remained stable. Plasmasodium (about 130 mM) showed the minimum and maximumvalues in the D0 group: 125 at 24 h and 137 mM at 5 h (Ta-ble 2). Plasma glucose (33–82 mg/dL) differed only within theD1 group, from 70 mg/dL after 5 h to 33 mg/dL after 24 h(Table 2).

Branchial and Muscular HydrationWater in the Silver Catfish gill tissue samples was 82–83%

and in muscle samples was 76–78%, neither changing under allexperimental conditions (Table 2).

Carbonic Anhydrase Specific ActivitySilver Catfish gill sample CA specific activity was always

lower at 24 h than was the respective values at 5 h, even for theD0 group. When different treatments were compared for the 5-hexposure, the lowest activity was found in the D1 group. For the24-h exposure, D1, D2, and D3 groups had lower CA activitythan the control D0 group (Figure 2A).

Kidney CA specific activity for the D0 group was also lowerat 24 h than at 5 h. However, in the kidney, no change in activitywas noted at 5 h, and an inhibition was detected only after 24 h,in the D3 group (Figure 2B).

DISCUSSIONSilver Catfish demonstrated to tolerate up to 24 h of exposure

to increasing levels of ammonia in the water. It could be demon-strated that the addition of ammonia as NH4Cl (0.5–2.0 mg/L)was nonsignificant compared with the ammonia produced andexcreted by the fish: total water ammonia levels reached 28.9–41.1 mg/L (0.247–0.333 mg/L) in our experiments. However,this initial ammonia introduction to the aquaria prompted in-creased levels of total ammonia from ammonia produced andexcreted by the fish, as note at 5 or 24 h. This protocol in-creased doses of ammonia for the experimental fish. As a con-sequence, a dose-dependent response of CA inhibition couldbe detected. Osmoregulatory homeostasis was maintained, nosign of stress was detected, and the most evident effect was thedose-dependent inhibition of the branchial and renal CA. Bycomparison, the maximum level of total ammonia tolerated bySilver Catfish fingerlings at 24 h was of 25 mg/L. This ammonialevel at 24 h caused mortality, but was associated to other stress

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5


TABLE 2. Blood and plasma metrics for Silver Catfish exposed to different initial doses of ammonia (i.e., conditions as explained in Table 1). For all values,there were no differences among groups exposed to different ammonia levels within the same exposure period (5 or 24 h); asterisks denote values that were at 24 hsignificantly different from the 5-h value.

Mean ± SE by experimental condition

D0 D1 D2 D3

Metric 5 h 24 h 5 h 24 h 5 h 24 h 5 h 24 h

Hematocrit (%) 26 ± 3.4 27 ± 4.1 25 ± 3.0 28 ± 1.9 29 ± 2.1 24 ± 2.4 24 ± 2.1 25 ± 2.8Total ammonia N

(mg/L)5.57 ± 0.55 5.54 ± 0.90 5.86 ± 1.67 4.77 ± 0.62 4.37 ± 0.36 5.28 ± 0.87 5.33 ± 0.71 3.75 ± 0.48

Osmolality(mOsm/kg ofwater)

247 ± 6.3 253 ± 6.4 257 ± 5.2 264 ± 4.4 261 ± 7.6 265 ± 3.7 253 ± 6.4 271 ± 4.8*

Cl− (mM) 108 ± 3.2 106 ± 4.4 109 ± 2.8 116 ± 1.9 117 ± 2.4 116 ± 4.9 109 ± 2.2 109 ± 3.6Na+ (mM) 137 ± 2.2 125 ± 4.0* 129 ± 2.5 133 ± 2.0 134 ± 2.2 131 ± 2.6 131 ± 2.7 131 ± 1.0K+ (mM) 2.5 ± 0.2 2.7 ± 0.1 2.6 ± 0.2 2.7 ± 0.1 2.7 ± 0.2 3.0 ± 0.2 3.0 ± 0.2 2.8 ± 0.2Glucose (mg/dL) 82 ± 16 68 ± 27 70 ± 11 33 ± 5* 43 ± 3 50 ± 11 58 ± 13 63 ± 22Cortisol (ng/mL) 98 ± 29 134 ± 42 139 ± 42 61 ± 14 110 ± 41 77 ± 16 119 ± 27 115 ± 21Gill hydration (%) 83 ± 0.5 83 ± 0.4 83 ± 0.4 82 ± 0.3 83 ± 0.2 82 ± 0.2 83 ± 0.3 82 ± 0.2Muscle hydration

(%)77 ± 0.3 76 ± 0.8 76 ± 0.8 76 ± 0.7 77 ± 0.8 78 ± 0.4 77 ± 0.8 77 ± 0.6

sources: transport, density, and temperature (Golombieski et al.2003). This is compatible with the fact that our fish did not showany mortality after 24 h in water with even higher total ammonialevels.

Hematocrit values were a little bit lower than the normalranges previously reported for Silver Catfish (range 37–51%;Borges et al. 2004) and are compatible with previous exposure ofthis catfish to ammonia (Carneiro et al. 2009a). In this last study,the Silver Catfish also showed no change in hematocrit after4 h of transport in closed systems with different fish densities(Carneiro et al. 2009a). Specimens of another Siluriform, theChannel Catfish Ictalurus punctatus also showed no change inhematocrit after 12 h of exposure to total ammonia at 25 mg/Lof water (Tomasso et al. 1980). Hematocrit is, thus, a very stableparameter.

When exposed to increasing concentrations of ammonia, itshould be more difficult for fish to get rid of ammonia becauseof the reduced outward diffusive gradients (revised in Wilkie1997, 2002; Randall and Tsui 2002; Ip and Chew 2010). So,increased plasma ammonia levels are expected. The increasein ammonia concentrations in the water did not cause changesin the concentration of total ammonia in plasma specimens ofSilver Catfish: 3.75–5.86 mg/L. Lower levels of plasma ammo-nia were detected in a previous study of Silver Catfish: 0.45 mg/Lin controls, increasing to 0.99 mg/L after fish were transportedfor 4 h (Carneiro et al. 2009a). Common Carp Cyprinus carpioalso showed an increase in total plasma ammonia from 5.11to 7.66 mg/L (conversion from µmol/L based on the molecularweight of NH3 = 17.03) after 12 h of transport in closed systems(Dobsikova et al. 2006). And also, there is high interspecific

variation in the toxicity of raised water ammonia levels, leadingto increased plasma ammonia. Toxic levels of total water ammo-nia have been set by the U.S. Environmental Protection Agencyfor more than 30 species of freshwater fishes, ranging from38 mg/L in Rainbow Trout Oncorhynchus mykiss to 124 mg/Lfor European Eels Anguilla anguilla (the most tolerant species)in pH 7.5 (USEPA 2009). Considering that un-ionized ammonia(NH3) is the toxic form of ammonia, values that are toxic (lead-ing to mortality) reported for freshwater fish (pH of about7) areabout 2.4 mg/L (LC50 24 h) for Channel Catfish Ictalurus punc-tatus and the Blue Tilapia Oreochromis aureus (Hargreaves andKucuk 2001; Biswas et al. 2006). As above, the Silver Catfishwas shown to be less sensitive than the Rainbow Trout. More-over, coherently, the maximum un-ionized ammonia levels wemeasured were one order of magnitude lower than the toxic lev-els reported for the Channel Catfish and the Blue Tilapia, thusalso not resulting in mortality or morbidity of our Silver Catfish.

Increase in water ammonia for up to 24 h caused few changesin the osmoregulatory homeostasis of the Silver Catfish. Infreshwater fish, exposure to increased ammonia hampers saltabsorption (e.g., Wilkie 1997, 2002; Randall and Tsui 2002;Evans et al. 2005), thus expectedly causing dilution of the ex-tracellular fluid such as found for chloride in Common Carp(Dobsikova et al. 2006), and for sodium here in Silver Catfish.Few studies have focused on the effects of ammonia increase onosmoregulatory homeostasis in freshwater fishes, but the liter-ature available shows compatible data; little changes in plasmaosmo-ionic concentrations upon ammonia increase were notedin Silver Catfish (Carneiro et al. 2009a, 2009b) and ChannelCatfish (Tomasso et al. 1980).

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5


FIGURE 2. (A) Branquial and (B) renal carbonic anhydrase specific activityof Silver Catfish subjected to ammonia in aquaria. Experimental conditionswere no added ammonia (control; N = 8) or initial addition of NH4Cl at 0.5(D1; N = 8), 1.0 (D2; N = 8), or 2.0 mg/L of water (D3; N = 7–8) for 5 or 24 h.Final total ammonia levels measured are in brackets. Different letters above barsindicate statistical differences between experimental conditions at 5 h (lowercase letters, white bars) and at 24 h (upper case letters, gray bars). Asterisksdenote significant differences between 5 and 24 h.

Compatibly, because extracellular osmotic ionic stability wasbasically maintained, we observed no change in muscle watercontent of the Silver Catfish. Silver Catfish have been shown todisplay a great capacity for regulation of tissue hydration, evenwhen exposed to increased water salinity (Souza-Bastos andFreire 2009). A longer period of exposure with further increasingwater ammonia levels could possibly disturb osmoregulatoryhomeostasis.

Besides the lack of significant osmoregulatory disturbance,no sign of stress induction was detected in the Silver Catfish;there was no increase in plasma cortisol values, which rangedbetween 61 and 139 ng/mL. Plasma cortisol values determinedin our study were higher than basal values already determinedfor Silver Catfish (about 30 ng/mL; Barcellos et al. 2001; 2004a,2004b). However, other studies of this species also showed largevariations in plasma cortisol levels, such as in situations of han-dling and confinement (30–300 ng/mL; Barcellos et al. 2001;2004a), transport (60–175 ng/mL; Carneiro et al. 2009a), andsalinity (146–178 ng/mL; Souza-Bastos and Freire 2009). Thus,

FIGURE 3. Carbonic anhydrase specific activity (CAA /mg protein) in Sil-ver Catfish tissues versus final total ammonia measured (TAN, mg/L) in theiraquaria: (A) gills at 5 h of exposure and at (B) 24 h, and kidney at (C) 5 h andat (D) 24 h. Values of Pearson’s product moment correlation coefficients (andP-value) indicate that the negative correlations are significant only after 24 h,for both tissues. Linear regression lines were, thus, drawn only for the data after24 h for these significant correlations.

our results may reflect the stress of handling the fish and main-taining them in closed systems, as already identified in otherstudies (Barcellos et al. 2001; 2004a, 2004b; Carneiro et al.2009a, 2009b; Souza-Bastos and Freire 2009). However, thereare few studies on the cortisol response associated with in-creased water ammonia; mostly, ammonia has been associatedwith other sources of stress. For example, when ammonia wasassociated with transport and increased density of Silver Catfish,cortisol increased from 50 to about170 ng/mL (Carneiro et al.2009a). However, as in our study, plasma cortisol of CommonCarp (202–213 ng/mL) was also not affected by increasing con-centrations of ammonia in water and plasma during transport(Dobsikova et al. 2006). Differently, ammonia alone (0.4 mg/L)decreased cortisol in Silver Catfis, from about 200 to about70 ng/mL after 24 h (Carneiro et al. 2009b). Thus, more studiesare needed to confirm whether ammonia alone is indeed a stronginducer of the cortisol stress response. Or, a good idea could beto associate other hormonal markers of stress to plasma cortisol,such as catecholamines (see Wendelaar Bonga 1997).

Following the same pattern shown for cortisol, plasma glu-cose ranged from 33 and 82 mg/dL. This variability is consistentwith previous data for Silver Catfish: 43–78 mg/dL (Borgeset al. 2004). Exposure to high concentrations of ammonia,

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5


associated (Dobsikova et al. 2006; Carneiro et al. 2009a) ornot (Carneiro et al. 2009b) to other sources of stress leads toincreased metabolic rate and consequent hyperglycemia. Ourdata do not fit into this hyperglycemia pattern; although therewas an isolated case of hypoglycemia (24 h, 0.5 mg/L addedammonium chloride), the main issue is the great variability inplasma glucose levels, which was due to other sources of stress(Wendelaar Bonga 1997; Barcellos et al. 2001).

The activity of the enzyme carbonic anhydrase (CA) deter-mined in the gills and kidneys of Silver Catfish was the mostsensitive variable evaluated in this study; it was clearly inhib-ited, in a dose-dependent manner to the levels of total ammoniameasured after 24 h of exposure (Figure 3). This differentialeffect on CA activity between 5 and 24 h highlights that theammonia in the water is indeed produced by the animal, so thelonger the time, the more it builds up and affects the enzyme.A direct effect of the ammonia on the carbonic anhydrase hasnot been frequently reported in the literature. Indeed, a recentreview highlighting the biomarker attributes of the CA did notmention ammonia, only metals and pesticides (Lionetto et al.2012). The only study found reported on branchial CA activityinhibition by ammonia (5.4 mg/L for 3 h) in the Rainbow Trout(ArasHisar et al. 2004).

A broad effect of CA inhibition is expected to directly in-terfere with salt absorption in freshwater, the ability to excreteCO2, and acid–base regulation (e.g., Henry 1996; Henry andSwenson 2000; ArasHisar et al. 2004; Georgalis et al. 2006a,2006b; Randall and Tsui 2006). In Silver Catfish, as discussedabove, the reduced branchial CA activity did not result in signif-icant loss of osmo-ionic homeostasis, which might be expectedgiven the branchial exchange of Na+ /H+ and Cl−/HCO3

− (e.g.,Henry 1996; Randall and Brauner 1998; Henry and Swenson2000; Evans et al. 2005). Also, with less CO2 hydration (CAinhibition), there is less CO2 removal from the fish (e.g., Ran-dall and Brauner 1998; Henry and Swenson 2000; Perry andGilmour 2006), and CO2 retention contributes to metabolic aci-dosis, which may have ensued with our Silver Catfish. Metabolicacidosis has been found in fish exposed to ammonia (Harg-reaves and Kucuk 2001; ArasHisar et al. 2004) and in fish in-jected with acetazolamide and benzolamide, inhibitors of theCA (e.g., Henry et al. 1988; Henry et al. 1995; Georgalis et al.2006a, 2006b). In addition, inhibition of kidney CA putativelycontributes to metabolic acidosis from reduced production ofH+ and HCO3

−, less retention of HCO3−, and reduced rates of

urine acidification (Georgalis et al. 2006a; Perry and Gilmour2006). It is important to mention that the toxicity of ammoniais dependent on pH and water temperature (e.g., Tomasso et al.1980; Hargreaves and Kucuk 2001). In agreement, increasedwater pH, leading to increased levels of un-ionized ammonia,reduced the tolerance to increased ammonia in Silver Catfish(Miron et al. 2008). So, the lack of a severe effect of ammo-nia on the Silver Catfish is compatible with the fact that watertemperature of the aquarium was fully controlled (18◦C, SE= 0.1), and water pH did not need to be adjusted with acid

or alkaline solutions throughout the experiment, remaining be-tween 7.2 and 7.4 and without significant differences betweenthe experimental conditions (P = 0.857).

In summary, metabolic acidosis may have happened in Sil-ver Catfish, although inversion in plasma Na/Cl ratios was notobserved (see Souza-Bastos and Freire 2009). So, even withsignificant CA inhibition from the increasing water ammonialevels after 24 h, neither stress nor osmoregulatory homeosta-sis breakdown were detected. The Silver Catfish proved to berelatively tolerant of increased concentrations of ammonia inthe water generated by their own metabolism. In conclusion, agreat amount of ammonia is actually produced by fish, whichare the real significant source of ammonia in tanks and transportcontainers or packages. Catfishes are tolerant of increased waterammonia, especially when the increase is not associated withother stresses, such as possibly changes in pH or temperature,handling, or increase in fish density. One must also keep in mindthat the duration of ammonia exposure (meaning more ammo-nia being excreted in the water by the fish) is extremely relevantconcerning the physiological disturbance of the animals, whichapplies specifically to intensive systems of fish culture, and fishtransportation. Moreover, we demonstrated here the value of theassay of the branchial and renal carbonic anhydrase activity asbiomarker of the effect of increased water ammonia levels.

ACKNOWLEDGMENTSAuthors wish to gratefully acknowledge the supply of Sil-

ver Catfish by the culture facilities of the Catholic Universityof Parana (LAPEP-PUC), the Doctorate fellowship awarded toL.R.S.B., and grants to C.A.F (Edital Universal 2007 and PQ-1D), all from the Brazilian National Council on Science andTechnology.

REFERENCESArasHisar, S., O. Hisar, T. Yanik, and S. M. Aras. 2004. Inhibitory effects of

ammonia and urea on gill carbonic anhydrase enzyme activity of RainbowTrout (Oncorhynchus mykiss). Environmental Toxicology and Pharmacology17:125–128.

Barcellos, L. J. G., L. C. Kreutz, R. M. Quevedo, I. Fioreze, L. Cericato,A. B. Soso, M. Fagundes, J. Conrad, R. K. Baldissera, A. Bruschi, and F.Ritter. 2004b. Nursery rearing of jundia, Rhamdia quelen (Quoy & Gaimard)in cages: cage type, stocking density and stress response to confinement.Aquaculture 232:383–394.

Barcellos, L. J. G., L. C. Kreutz, C. Souza, L. B. Rodrigues, I. Fioreze, R.M. Quevedo, L. Cericato, A. B. Soso, M. Fagundes, J. Conrad, L. A. Lac-erda, and S. Terra. 2004a. Hematological changes in jundia (Rhamdia quelenQuoy & Gaimard: Pimelodidae) after acute and chronic stress caused byusual aquacultural management, with emphasis on immunosuppressive ef-fects. Aquaculture 237:229–236.

Barcellos, L. J. G., V. M. Woehl, G. F. Wassermann, R. M. Quevedo, I. Ittzes,and M. H. Krieger. 2001. Plasma levels of cortisol and glucose in response tocapture and tank transference in Rhamdia quelen (Quoy & Gaimard), a SouthAmerican catfish. Aquaculture Research 32:121–123.

Biswas, J. K., D. Sarkar, P. Chakraborty, J. N. Bhakta, and B. B. Jana. 2006.Density dependent ambient ammonium as the key factor for optimizationof stocking density of Common Carp in small holding tanks. Aquaculture261:952–959.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5


Borges, A. L., V. Scotti, D. R. Siqueira, D. F. Jurinitz, and G. F. Wassermann.2004. Hematologic and serum biochemical values for jundia (Rhamdia que-len). Fish Physiology and Biochemistry 30:21–25.

Boyd, C. E., and C. S. Tucker. 1992. Water quality and pond soil analyses foraquaculture. Auburn University, Alabama Agricultural Experiment Station,Auburn, Alabama.

Bradford, M. M. 1976. A rapid and sensitive method for the quantification ofmicrogram quantities of protein utilizing the principle of protein-dye binding.Analytical Biochemistry 72:248–254.

Burnett, L. E., P. B. J. Woodson, M. G. Rietow, and V. C. Vilicich. 1981. Crab gillintra-epithelial carbonic anhydrase plays a major role in haemolymph CO2

and chloride ion regulation. Journal of Experimental Biology 92:243–254.Carneiro, P. C. F., P. H. S. Kaiseler, E. A. C. Swarofsky, and B. Baldisserotto.

2009a. Transport of jundia Rhamdia quelen juveniles at different loadingdensities: water quality and blood parameters. Neotropical Ichthyology7:283–288.

Carneiro, P. C. F., E. A. C. Swarofsky, D. P. E. Souza, T. M. Ramos, B. Bagli-oli, and B. Baldisserotto. 2009b. Ammonia-, sodium chloride-, and calciumsulfate-induced changes in the stress responses of jundia, Rhamdia quelen,juveniles. Journal of the World Aquaculture Society 40:810–817.

Dobsikova, R., Z. Svobodova, J. Blahova, H. Modra, and J. Velisek. 2006. Stressresponse to long distance transportation of Common Carp (Cyprinus carpioL.). Acta Veterinaria Brno 75:437–448.

Emerson, K., R. C. Russo, R. E. Lund, and R. V. Thurston. 1975. Aqueousammonia equilibrium calculations: effect of pH and temperature. Journal ofthe Fisheries Research Board of Canada 32:2379–2383.

Evans, D. H., P. M. Piermarini, and K. P. Choe. 2005. The multifunctional fishgill: dominant site of gas exchange, osmoregulation, acid-base regulation,and excretion of nitrogenous waste. Physiological Reviews 85:97–177.

Georgalis, T., K. M. Gilmour, J. Yorston, and S. F. Perry. 2006a. Roles ofcytosolic and membrane-bound carbonic anhydrase in renal control of acid-base balance in Rainbow Trout, Oncorhynchus mykiss. American Journal ofPhysiology 291:F407–F421.

Georgalis, T., S. F. Perry, and K. M. Gilmour. 2006b. The role of branchialcarbonic anhydrase in acid-base regulation in Rainbow Trout (Oncorhynchusmykiss). Journal of Experimental Biology 209:518–530.

Golombieski, J. I., L. V. F. Silva, B. Baldisserotto, and J. H. S. Silva. 2003.Transport of Silver Catfish (Rhamdia quelen) fingerlings at different times,load densities, and temperature. Aquaculture 216:95–102.

Hargreaves, J. A., and S. Kucuk. 2001. Effects of diel un-ionized ammoniafluctuation on juvenile hydrib Striped Bass, Channel Catfish, and Blue Tilapia.Aquaculture 195:163–181.

Henry, R. P. 1991. Techniques for measuring carbonic anhydrase in vitro: theelectrometric delta pH and pH stat assays. Pages 119–126 in S. J. Dodgson,R. E. Tashian, G. Gros, and N. D. Carter, editors. The carbonic anhydrases:cellular physiology and molecular genetics. Plenum, New York.

Henry, R. P. 1996. Multiple roles of carbonic anhydrase in cellular transport andmetabolism. Annual Review of Physiology 58:523–538.

Henry, R. P., R. G. Boutilier, and B. L. Tufts. 1995. Effects of carbonic anhydraseinhibition on the acid base status in lamprey and trout. Respiratory Physiology99:241–248.

Henry, R. P., N. J. Smatresk, and J. N. Cameron. 1988. The distribution ofbranchial carbonic anhydrase and the effects of gill and erythrocyte carbonicanhydrase inhibition in the Channel Catfish Ictalurus punctatus. Journal ofExperimental Biology 134:201–218.

Henry, R. P., and E. R. Swenson. 2000. The distribution and physiological signif-icance of carbonic anhydrase in vertebrate gas exchange organs. RespiratoryPhysiology 121:1–12.

Ip, Y. K., and S. F. Chew. 2010. Ammonia production, excretion, toxicity, anddefense in fish: a review. Frontiers in Physiology 1:1–20.

Lionetto, M. G., R. Caricato, M. E. Giordano, E. Erroi, and T. Schettino.2012. Carbonic anhydrase as pollution biomarker: an ancient enzyme with anew use. International Journal of Environmental Research and Public Health9:3965–3977.

Liu, H., S. Xie, X. Zhu, W. Lei, D. Han, and Y. Yang. 2008. Effects of dietaryascorbic acid supplementation on the growth performance, immune and stressresponse in juvenile Leiocassis longirostris Gunther exposed to ammonia.Aquaculture Research 39:1628–1638.

Miron, D. S., B. Moraes, A. G. Becker, M. Crestani, R. Spanevello, V. L. Loro,and B. Baldisserotto. 2008. Ammonia and pH effects on some metabolic pa-rameters and gill histology of Silver Catfish, Rhamdia quelen (Heptapteridae).Aquaculture 277:192–196.

Pereira, L. P. F., and C. T. J. Mercante. 2005. A amonia nos sistemas de criacaode peixes e seus efeitos sobre a qualidade da agua, uma revisao. [Ammoniain fish culture systems and its effects on water quality, a review.] Boletim doInstituto de Pesca 31:81–88.

Perry, S. F., and K. M. Gilmour. 2006. Acid-base balance and CO2 excretionin fish: unanswered questions and emerging models. Respiratory Physiologyand Neurobiology 154:199–215.

Randall, D. J., and C. Brauner. 1998. Interactions between ion and gas transfer infreshwater teleost fish. Comparative Biochemistry and Physiology A 119:3–8.

Randall, D. J., and T. K. N. Tsui. 2002. Ammonia toxicity in fish. MarinePollution Bulletin 45:17–23.

Randall, D. J., and T. K. N. Tsui. 2006. Tribute to R. G. Boutilier: acid-base transfer across fish gills. Journal of Experimental Biology 209:1179–1184.

Souza-Bastos, L. R., and C. A. Freire. 2009. The handling of salt by theneotropical cultured freshwater catfish Rhamdia quelen. Aquaculture 289:167–174.

Tomasso, J. R., C. A. Goudie, B. A. Simco, and K. B. Davis. 1980. Effectsof environmental pH and calcium on ammonia toxicity in Channel Catfish.Transactions of the American Fisheries Society 109:229–234.

USEPA (U.S. Environmental Protection Agency). 2009. Update aquatic lifeambient water quality criteria for ammonia–freshwater. USEPA, EPA-822-D-09–001, Washington, D.C.

Verdouw, H., C. J. A. van Echteld, and E. M. J. Dekkers. 1978. Ammoniadetermination based on indophenols formation with sodium salicylate. WaterResearch 12:399–402.

Vitale, A. M., J. M. Monserrat, P. Castilho, and E. M. Rodriguez.1999. Inhibitory effects of cadmium on carbonic anhydrase activity andionic regulation of the estuarine crab Chasmagnathus granulata (De-capoda, Grapsidae). Comparative Biochemistry and Physiology C 122:121–129.

Wendelaar Bonga, S. E. 1997. The stress response in fish. Physiological Reviews77:591–625.

Wilkie, M. P. 1997. Mechanism of ammonia excretion across fish gills. Com-parative Biochemistry and Physiology A 118:39–50.

Wilkie, M. P. 2002. Ammonia excretion and urea handling by fish gills: presentunderstanding and future research challenges. Journal of Experimental Zool-ogy 293:284–301.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:43

20

July

201

5


Click for updates


Variability in Size Traits of Sunshine Bass Larvae fromDifferent Male Striped BassS. E. Lochmanna & K. J. Goodwina

a Department of Aquaculture and Fisheries, University of Arkansas at Pine Bluff, 1200 NorthUniversity Drive, Mail Slot 4912, Pine Bluff, Arkansas 71601, USAPublished online: 16 Dec 2014.

To cite this article: S. E. Lochmann & K. J. Goodwin (2015) Variability in Size Traits of Sunshine Bass Larvae from DifferentMale Striped Bass, North American Journal of Aquaculture, 77:1, 76-81, DOI: 10.1080/15222055.2014.960538








http://dx.doi.org/10.1080/15222055.2014.960538




ARTICLE

Variability in Size Traits of Sunshine Bass Larvaefrom Different Male Striped Bass

S. E. Lochmann* and K. J. GoodwinDepartment of Aquaculture and Fisheries, University of Arkansas at Pine Bluff, 1200 North UniversityDrive, Mail Slot 4912, Pine Bluff, Arkansas 71601, USA

AbstractWe examined the amount of variability in size at hatch and size at yolk absorption among sunshine bass (female

White Bass Morone chrysops × male Striped Bass M. saxatilis) due to sire effects. Eggs from a female White Bass wereseparated into four aliquots. Each aliquot was fertilized by sperm from a different male Striped Bass. Single fertilizedeggs were place in 6-mL vials and incubated at 18◦C. Eggs were examined every 6 h until the first eggs hatched andevery 3 h thereafter. Approximately 50–100 yolk sac larvae were removed from the vials at hatching, photographed,and returned live to the vials when possible. At 5 d posthatch (dph), the remaining larvae were photographed andenumerated. Standard lengths of larvae at hatch and at 5 dph were estimated from photographs. Growth (i.e., increasein length) during the endogenous larval stage was calculated for a subset of individuals photographed at both times.This process was repeated in two subsequent trials. There were significant sire effects for size at hatch (P = 0.003),size at 5 dph (P < 0.001), and growth (P < 0.001). The sire effect explained 3.0–39.0% of the variability in these sizetraits. Genetic selection for favorable size traits at early life stages among both parent species could lead to largersunshine bass larvae at yolk absorption and eliminate the need for rotifers at first feeding. Rapid growth at early lifestages may be an indicator of rapid growth at later life stages. Efforts to select for rapid growth at several life stagescould enhance multiple aspects of the hybrid Striped Bass industry.

Production of Striped Bass Morone saxatilis and its hybridswas a fast-growing sector of the U.S. aquaculture industry inthe early 2000s (NASS 2006). The Striped Bass industry rankedfourth in sales among finfish aquaculture in the United States(NASS 2006). Production has declined since 2005 due to ris-ing production costs (M. Turano, Pentair Aquatic Eco-Systems,personal communication). Harrell (1997) pointed out that do-mestication of broodstock would likely increase efficiencies andfurther develop the industry. Harrell (1997) also suggested thatdomestication is, in fact, a form of selection for individuals thatadapt well to hatchery conditions. Several other investigatorshave outlined the benefits of genetic selection, once domesti-cated broodstock are widely available (Hallerman 1994; Kohler1997; Garber and Sullivan 2006).

A first step in selective breeding programs is to determinethe amount of variability in traits to be considered for geneticimprovement (Hallerman 1994; Garber and Sullivan 2006).

*Corresponding author: [email protected] February 11, 2014; accepted July 7, 2014

Variability in growth rates among Chesapeake Bay strains ofStriped Bass were examined by Woods et al. (1999). Variabilityin growth rates of Striped Bass were further examined amongfive strains, geographically distinct from the aforementionedChesapeake Bay strains, by Woods (2001). Both studies indi-cated significant differences in Striped Bass growth rates amonggeographic strains. Kohler et al. (2001) compared growth ratesamong three geographic strains of White Bass M. chrysops.There were significant differences in weight among strains ofWhite Bass at the end of phase II production. Fuller and McEn-tire (2011) found significant differences in length and weight atthe end of phase I production among 15 families of White Bass.Wang et al. (2006, 2007) found significant dam and sire effectson growth rates and carcass quality traits of juvenile sunshinebass (female White Bass × male Striped Bass). These studiescollectively indicate that growth rates are variable among strainsand families in both parent species and the hybrid. Those of

76

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5

PATERNAL EFFECTS ON SUNSHINE BASS LARVAE 77

Wang et al. (2006, 2007), in particular, illustrated the likelihoodthat genetic selection of either parent species could improvegrowth performance of the hybrid. However, growth rate is notthe only trait important for aquaculture production.

Genetic improvement of other traits, such as fillet dress-outor disease resistance, could also reduce production costs in thehybrid Striped Bass industry. Kohler et al. (2001) found differ-ences in fillet dress-out among geographic strains of White Bassat the end of phase III production. Couch (2006) found differ-ences in antimicrobial peptide activity among strains of StripedBass, indicating possible geographic differences in disease re-sistance. There have been a few examinations of variability ofother traits, such as egg and larval size among strains or fam-ilies of Striped Bass and White Bass. Larger eggs generallyproduce larger larvae at hatch (Chambers 1997). Larvae that arelarger at hatch should have a wider prey size spectrum (Labayet al. 2004) and would likely be better at avoiding predation inearthen ponds (Paradis et al. 1996). Brown et al. (1998) foundsignificant differences in egg weights and larval growth ratesamong four geographic strains of Striped Bass. Bergey et al.(2003) found significant differences in diameters of Striped Basseggs among nine different geographic strains. Lochmann et al.(2009) reported significant differences in egg size among fe-male White Bass and significant differences in length at hatchamong groups of sunshine bass produced using those same fe-male White Bass. However, the relationship between egg sizeand size at hatch was weak (Lochmann et al. 2009), so selec-tion for larger egg size does not appear likely to produce largerlarvae at hatch. Lochmann et al. (2012) reported a significantdam effect on length at hatch among groups of sunshine bassproduced using different White Bass females. The study designused by Lochmann et al. (2012) was not sufficient to determinethe sire effect on size at hatch. In the present investigation, wequantified the amount of variability in size traits at early lifestages among groups of sunshine bass attributable to differentStriped Bass males.

METHODSStriped Bass sperm and White Bass eggs were used to pro-

duce sunshine bass at Keo Fish Farm, Keo, Arkansas. Eggsfrom a single White Bass were combined with sperm from fourStriped Bass for each trial of the study. A unique White Bass fe-male and four unique Striped Bass males were used during eachof three trials. Females were collected from the lower WhiteRiver, Arkansas, earlier in the year. Males were either producedat Keo Fish Farm (trial 1, males A–D) from a Ouachita River,Arkansas, stock of Striped Bass or collected directly from theOuachita River earlier in the year (trials 2–3, males E–M). Cap-tive fish were fed Fathead Minnow Pimephales promelas andGolden Shiners Notemigonus chrysoleucas ad libitum duringcaptivity.

Spawning followed standard procedures for sunshine bassproduction. A single individual was responsible for all strip-

spawning. Briefly, human chorionic gonadotropin was used toinduce final maturation and ovulation. Females were examined6–8 h before expected ovulation (24 h at 18◦C) and every 2 hthereafter until they were determined to be ripe according toRees and Harrell (1990). A ripe female was strip-spawned bygently squeezing the abdomen. Eggs were divided equally intofour different dry plastic containers. The abdomen of a malewas first dried with a shop towel to avoid sperm activation, andthen milt was expressed with gentle palpitation. Milt from oneof the four males was used to fertilize the eggs in one of the fourcontainers. The process was repeated with three different males.No sperm extenders were used during the study, and no specificmilt-to-egg ratio was employed. When sperm had been addedto all four containers, a small amount of water was added toeach container and the ingredients were mixed thoroughly. Theeggs from each container were placed in separate McDonaldhatching jars. Eggs were treated with tannic acid (150 mg/Lfor ∼10 min) to minimize clumping. Aeration kept the eggsin suspension during the tannic acid treatment. Following thetreatment, tannic acid was flushed from the hatching jars for∼30 min with well water. Gentle aeration was used to keep theeggs in suspension during transport from Keo to the Aquacultureand Fisheries Center at the University of Arkansas at Pine Bluff.Well water was also transported from Keo to Pine Bluff for useduring egg incubation.

Fertilization and hatch rates were determined similarly ineach trial. After 3 h of water hardening, eggs were removedfrom a McDonald hatching jar, examined under a dissecting mi-croscope to determine whether fertilization had occurred, andthen placed individually into 6-mL uncapped glass vials filledwith well water at 18◦C. Blastoderm formation indicated fertil-ization. Only fertilized eggs were placed into vials. Unfertilizedeggs were counted to determine fertilization rates. One hundredand fifty eggs fertilized by each male were placed into 6-mLvials, which were then placed into three 50-slot racks. Rackswere then placed randomly on one of four shelves in an uprightincubator, three racks per shelf. Eggs were incubated at ∼18◦C.Temperatures in several of the vials were measured every 4–6 h. Well water from Keo was held at 18◦C and used to replacehalf the water in each vial daily. Hardness of Keo water in thecarboys was tested at the beginning of each trial. We did nothave access to a fiber optic dissolved oxygen probe, nor wouldthe pH probe fit through the vial mouth. Also, the amount ofwater necessary for the total ammonia nitrogen test exceededthe amount replaced daily from each vial. Hence, we tested acomposite sample of the water removed daily from vials for pH,total ammonia nitrogen (TAN), and dissolved oxygen (DO). Weacknowledge that a composite sample is suboptimal, but reportthe composite sample results rather than no water quality infor-mation. We also report the un-ionized ammonia concentrationbased on a formula for the un-ionized fraction of TAN (Emersonet al. 1975).

Yolk sac larvae and post-yolk-sac larvae were photographedduring each trial of the study. Eggs were examined every 6 h

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5

78 LOCHMANN AND GOODWIN

TABLE 1. Average (SD) water quality conditions measured during each week of this study.

Week Temperature (◦C) Dissolved oxygen (mg/L) pH Un-ionized ammonia (mg/L) Hardness (mg/L)

1 18.4 (0.4) 7.3 (1.7) 8.1 (0.1) 0.014 (0.004) 248 (15)2 18.2 (0.3) 7.5 (1.2) 7.6 (0.3) 0.008 (0.005) 221 (29)3 18.2 (0.4) 8.0 (1.0) 7.7 (0.3) 0.009 (0.005) 231 (39)

until hatching was first observed and every 3 h thereafter. Eggswere monitored until larvae had hatched or eggs became milkywhite and covered in fungus, indicating no possibility of hatch-ing. Upon observation with a lighted magnifying lamp, emergentyolk sac larvae were removed from the vial and photographed(2.5–4.0 × magnification) with a Spot Insight Color Model3.2.0 digital camera (Diagnostic Instruments, Sterling Heights,Michigan) mounted on a Leica MZ9.5 microscope (Leica Mi-crosystems, Buffalo Grove, Illinois) and interfaced with a com-puter. The duration of the egg stage was noted for each yolk saclarva. No more than 100 yolk sac larvae from each male werephotographed. When possible, yolk sac larvae were returned liveto their vials. We calculated two hatch rate estimates for eachmale based on the proportion of hatched larvae (n = 50) fromtwo of the three racks. The third rack, with its hatched larvae, aswell as the two racks containing yolk sac larvae previously mea-sured, remained in the incubator until 5 d posthatch (dph). Wecontinued to exchange half the water daily until most larvae hadabsorbed their yolk. Yolk absorption occurs at ∼5 dph at 18◦C.At 5 dph, the remaining larvae were photographed as before.For a small group of fish from most males, we were able to col-lect data for both length at hatch and length at yolk absorptionfrom the same individuals. Standard lengths were measured, bya single individual, for each larva with Image-Pro Plus softwareversion 4.5.1.22 (Media Cybernetics, Silver Spring, Maryland).The major and minor axes of the yolk sac at hatch and at 5dph were also measured. To determine precision of the lengthmeasurements a random subsample of 20 larvae were measuredtwice. We calculated the percent difference between the twoSL measurements, and then calculated the mean and SD of thepercent difference. To determine the accuracy of the length mea-surements, a ruler was photographed and a 1-mm increment onthe ruler was measured 20 times. We then calculated the meanand SD of the 1-mm increment measurements.

Two production and three larval size traits were examined toquantify variability in those characteristics among Striped Bassmales in this study. Estimates of fertilization and hatch rateswere determined for each Striped Bass male and the correlationbetween fertilization rate and hatch rate was calculated. Weexamined the sire effect on length at hatch and length at 5dph with mixed-effects nested ANOVA (MIXED procedure inStatistical Analysis System; SAS Institute 1990). We used themodel,

Yi jk = µ + TRIALi + SIRE(TRIAL)i j + εi jk,

where Yijk is an individual observation of length (at hatch or at5 dph), µ is the overall mean, TRIALi is the fixed effect of triali, SIRE(TRIAL)ij is the random effect of sire j nested in trial i,and εijk is the random residual (Bosworth et al. 1997; Wang et al.2006). The effect of trial was tested against the mean square ofthe sire nested in trial term (Zar 1999). We also calculated thecorrelation between mean length at hatch and mean length at 5dph. For a subset of fish (n = 112), we were able to determinegrowth (i.e., increase in length) based on photographs of thesame individual at hatch and at 5 dph. Three males (G, H,and J) had fewer than five individuals with estimates of size athatch and at 5 dph. Those males were excluded from growthanalyses because of small sample sizes, though not from thepreviously mentioned analyses of length at hatch and at 5 dph.We subtracted length at hatch from length at 5 dph to acquirea measure of linear growth over the 5-d period. We employedthe method outlined in Heyer et al. (2001) to calculate yolkvolume at hatch and at 5 dph. Yolk is generally exhausted in 5d at 18◦C. If yolk was completely exhausted and the yolk sacwas unidentifiable at 5 dph, the yolk volume was presumed tobe zero. We calculated the volume of yolk utilized during the5-d period by subtracting the yolk volume at 5 dph from theyolk volume at hatch. We used the same mixed-effects nestedANOVA approach to determine the effects of trial, yolk use, andsire nested in trial on growth. An alpha level of 0.05 was usedfor all statistical tests.

RESULTSWater quality was generally similar among trials during

the study. Average temperature ranged from 18.2◦C to 18.4◦C(Table 1). Average DO was above 7.0 mg/L each trial. Theaverage pH ranged from 7.6 to 8.1 during the study. Averageun-ionized ammonia was at or below 0.014 mg/L and hardnesswas consistently above 200 mg/L (Table 1). These water qualitycharacteristics are within the optimal ranges for sunshine bassproduction (Hodson and Hayes 1989; Rees and Harrell 1990).

Fertilization rate ranged from 77% to 97% during the study(Figure 1). Hatch rate estimates were fairly consistent betweenthe two racks but varied among eggs fertilized by Striped Bassmales. Average hatch rate ranged from 53% to 81%. All hatchrates during the second trial were lower than the lowest hatchrate observed during the first trail. Hatch rates in the third trailoverlapped hatch rate ranges from both previous trials. Fertil-ization and hatch rates were not correlated (r = −0.457, P =0.135). Hatch rates of eggs from a single White Bass female

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


FIGURE 1. Fertilization rate (filled bars) and mean hatch rate (open bars) foreggs fertilized by different Striped Bass males during this study. Error bars onmean hatch rate are 1 SD. Striped Bass males A–D represent trial 1, StripedBass males E–H represent trial 2, and Striped Bass males J–M represent trail 3.

and fertilized by Striped Bass males E–H were among the low-est hatch rates (Figure 1) even though they had some of thehighest fertilization rates.

The image analysis software allowed both precise and ac-curate measurements of SL. The mean (SD) percent differencebetween pairs of SL measurements was 0.405% (SD, 0.313).This represents an average difference between pairs of SL mea-surements of less than 0.02 mm. The mean length of the 1-mmincrement measurements was 1.016 mm (SD, 0.014).

Average length at hatch ranged from 2.83 to 3.06 mm SL.Length at hatch varied among trials (F = 8.50, df = 2, P =0.008). The random effect of SIRE(TRIAL) was significant aswell (F = 2.83, df = 9, P = 0.003). The entire model accountedfor approximately 8.7% of the variability in length at hatch(i.e., R-square = 1 − sum of squares [SS]error/SStotal), but theSIRE(TRIAL) effect accounted for only 3.0% of the variabilityin length at hatch (SSsire(trial)/SSmodel = 0.35).

Average length at 5 dph ranged from 3.87 to 4.16 mm SL.Length at 5 dph varied among trials (F = 5.07, df = 2, P =0.034). The random effect of SIRE(TRIAL) was also signifi-cant (F = 4.31, df = 9, P < 0.001). The model accounted for

2.5

3.0

3.5

4.0

4.5

5.0

A B C D E F G H J K L M

Ave

rage

SL

(m

m)

Striped Bass Male

Hatch

5 dph

FIGURE 2. Average length at hatch (open bars) and average length at 5 dph(filled bars) for larvae sired from different Striped Bass males during this study.Error bars are 1 SD. Striped Bass males A–D represent trial 1, Striped Bassmales E–H represent trial 2, and Striped Bass males J–M represent trail 3.

y = 0.7176x + 1.9281R² = 0.3114

3.85

3.90

3.95

4.00

4.05

4.10

4.15

4.20

2.80 2.85 2.90 2.95 3.00 3.05 3.10

Len

gth

at 5

dph

(m

m)

Length at hatch (mm)

FIGURE 3. Relationship between mean length (SL) at hatch and mean lengthat 5 dph for larvae sired from the 12 Striped Bass males used during this study.

15.0% of the variability in length at 5 dph, but the SIRE(TRIAL)effect accounted for almost half of that explained variability(SSsire(trial)/SSmodel = 0.47). Individual length at hatch was morevariable than individual length at 5 dph (Figure 2). Length athatch and length at 5 dph appeared related (Figure 3). However,the correlation was not significant (r = 0.552, P = 0.063).

Average growth during the 5-d period ranged from 0.59 to1.18 mm. Yolk use was somewhat less variable and rangedfrom 0.11 to 0.26 mm3 (Figure 4). The effects of trial andyolk use on growth were not significant. However, the effectof SIRE(TRIAL) was significant (F = 7.46, df = 6, P < 0.001).The model explained 42.6% of the variability in growth and, asexpected, the SIRE(TRIAL) effect accounted for most (92%) ofthe variability explained by the model.

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

A B C D E F G H J K L MGro

wth

(m

m)

or Y

olk

Use

(m

m3 )

Striped Bass Male

Growthyolk use

FIGURE 4. Average growth (open bars) and yolk use (filled bars) betweenhatch and 5 dph for groups of larvae sired from 9 of the 12 Striped Bass malesduring this study. Error bars are 1 SD. Striped Bass males A–D represent trial 1,Striped Bass males E–F represent trial 2, and Striped Bass males K–M representtrail 3. No larvae from males G, H, and J photographed at hatch survived to 5dph.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5

80 LOCHMANN AND GOODWIN

DISCUSSIONThis study, like several others, demonstrated a sire effect on

size traits of moronids and their hybrids. Wang et al. (2006,2007) conducted a 10 × 10 factorial cross with Striped Bassmales and White Bass females and found a significant sire effecton weight and length of older individuals (∼152 d postfertiliza-tion [dpf]) and on carcass quality traits at 389 dpf. Couch (2006)observed numerous sire effects on size traits among Striped Bassfamilies at the end of phase II and phase III production. Couchfound strong evidence for strain effects, but not for sire effects,on size traits of smaller phase I Striped Bass. Several studiesexamining family, stock, or strain effects on size traits were notspecifically designed to examine sire effects. For example, somestudies pooled sperm from multiple males (Bosworth et al. 1997;Woods et al. 1999; Kohler et al. 2001; Lochmann et al. 2009).Other studies created full-sib families (Woods 2001; Fuller andMcEntire 2011). This study demonstrated a sire effect on sizetraits at the earliest stages of development, and thus comple-ments studies demonstrating sire effects on size traits at laterlife stages.

The range of mean length at hatch in the present study (2.83–3.06 mm SL) was similar to ranges reported in other studies.Sunshine bass mean length at hatch among 12 female WhiteBass groups ranged from 2.89 to 3.09 mm SL (Lochmann et al.2009). Bosworth et al. (1997) reported sunshine bass had amean length at 9 h post-hatch of 3.01 mm SL. Lochmann et al.(2012) indicated a larger range of mean length at hatch (2.80–3.18 mm SL), but their study included the effect of differentincubation temperatures on sunshine bass length at hatch. Whenvariability in a size trait is high among families or strains there isa greater likelihood that genetic improvements through selectionwill favorably influence the trait.

The sire effect on growth was already apparent at yolk ab-sorption (∼5 dph). This study did not examine whether sireprogeny with higher growth performance up to yolk absorptionexhibited higher growth performance at later life history stages.However, Couch (2006) found that performance of fish duringphase II production was a good predictor of performance duringphase III production. Woods et al. (1999) calculated best lin-ear unbiased predictions of family merit and demonstrated thatfamilies with favorable performance characteristics for lengthand weight at young ages were often the families with favorableperformance characteristics at older ages. Size differences ob-served in the present study at yolk absorption could be a resultof inequalities of yolk, which is the source of energy duringthe endogenous larval stage. However, we found no relationshipbetween yolk utilized and growth during the endogenous larvalstage. We suggest that the size differences observed at yolk ab-sorption were an early manifestation of phenotypic or genotypicdifferences due to different sires.

Sire effect in this study accounted for less overall variabilityin length at hatch (3%) than dam effect from similar studies.Dam effect explained 11% of the variability in length at hatch

among 12 female White Bass groups (Lochmann et al. 2009).The effect of dam nested in time explained approximately 16%of the variability in length at hatch for 14 female White Bassgroups (Lochmann et al. 2012). These studies are comparableto the present study because dams from all three studies camefrom the Arkansas and White Rivers and sires came from theOuachita River. Wang et al. (2006) noted that the dam effect ex-plained twice the variability of sunshine bass length and weightat 152 dpf than did sire effect. Broad-sense heritability corre-sponds to the amount of explained variability, and one mightinfer that increases in length at hatch are more likely to comethrough selection among White Bass dams than selection amongStriped Bass sires. Nevertheless, several investigators have con-cluded that there is enough variation in growth among StripedBass individuals and strains to consider selective breeding toimprove size traits in the Striped Bass component of the con-tribution to hybrid Striped Bass (Woods 2001; Couch 2006).We argue that the sire effect on growth of sunshine bass inthe present study has already become evident at 5 dph and thatselection for improvement in growth at later life stages might ac-tually result in improvement of size traits at earlier life stages aswell.

Why is size at early life stages so important? Sunshine bassare considerably smaller at hatch than palmetto bass (femaleStriped Bass × male White Bass) because the White Bass fe-male used for producing sunshine bass produces a smaller eggand smaller larva (Bosworth et al. 1997). This creates a need touse rotifers at first feeding in sunshine bass that is not requiredfor palmetto bass (Ludwig 1993). However, most of the hybridStriped Bass industry relies on sunshine bass fingerlings be-cause of the relative ease of culturing and spawning White Bassfemales. Eklund et al. (2012) compared the cost of producingsunshine bass in earthen ponds and tanks at a variety of produc-tion scales. Production costs in tanks were considerably higherthan costs in ponds, though tank production allowed year-roundproduction of fingerlings. In one hypothetical scenario, where1 million phase I fingerlings are produced in sixteen 2,457-Ltanks, equipment for rotifer culture represented approximately33% of the annual investment costs and rotifer culture expensesrepresented approximately 96% of the annual variable costs ofproduction. Ludwig and Lochmann (2009) demonstrated thatsome sunshine bass larvae were able to consume microcystArtemia nauplii at first feeding, but survival was lower than sun-shine bass larvae fed rotifers Brachionus plicatilis at first feeding(38% and 94% survival, respectively). Genetic selection to in-crease size at yolk absorption could allow a greater proportionof sunshine bass to consume microcyst Artemia nauplii at firstfeeding, remove the need for rotifers during tank production,reduce the cost of tank production, allow year-round productionof fingerlings, and significantly improve efficiencies within theindustry. This study does not prove that the differences in sizetraits at early life history stages are genetic. The differences weobserved could be related to differences among the three trials.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


The sire effect might be phenotypic and related to male size orage. The sire effect also might be related to differences in spermquality associated with the distinct trials. Alternatively, the sireeffect could be genetic, or a combination of genetic and pheno-typic factors. Studies examining the heritability of size at hatchand size at yolk absorption, as well as studies that determinewhether rapid growth up to yolk absorption is a good predictorof growth performance at later life stages, would confirm theefficacy of this approach to improving efficiencies in the hybridStriped Bass industry.

ACKNOWLEDGMENTSWe are grateful to Mike Freeze and Mike Clark (Keo Fish

Farm) for providing eggs for the study. This project was sup-ported by funding from the Evans–Allen Program of the U.S.Department of Agriculture.

REFERENCESBergey, L. L., R. A. Rulifson, M. L. Gallagher, and A. S. Overton. 2003.

Variability of Atlantic coast Striped Bass egg characteristics. North AmericanJournal of Fisheries Management 23:558–572.

Bosworth, B. G., G. S. Libey, and D. R. Notter 1997. Egg, larval, and fingerlingtraits of crosses among Striped Bass (Morone saxatilis), White Bass (M.chrysops), and their F1 hybrids. Aquaculture 154:201–217.

Brown, J. J., A. Ehtisham, and D. O. Conover. 1998. Variation in larval growthrate among Striped Bass stocks from different latitudes. Transactions of theAmerican Fisheries Society 127:598–610.

Chambers, R. C. 1997. Environmental influences on egg and propagule sizes inmarine fishes. Pages 63–102 in R. C. Chambers and E. A. Trippel, editors.Early life history and recruitment in fish populations. Chapman and Hall,New York.

Couch, C. R. 2006. Microsatellite DNA marker-assisted selective breeding ofStriped Bass. Doctoral dissertation. North Carolina State University, Raleigh.

Eklund, P., C. Engle, and G. Ludwig. 2012. Comparative cost analysis of hy-brid Striped Bass fingerling production in ponds and tanks. North AmericanJournal of Aquaculture 74:39–53.

Emerson, K., R. C. Russo, R. E. Lund, and R. V. Thurston.1975. Aqueousammonia equilibrium calculations: effect of pH and temperature. Journal ofthe Fisheries Research Board of Canada 32:2379–2383.

Fuller, S. A., and M. M. McEntire. 2011. Variation in body weight and lengthamong families of fingerling White Bass after communal rearing. Journal ofApplied Aquaculture 23:250–255.

Garber, A. F., and C. V. Sullivan. 2006. Selective breeding for the hybrid StripedBass (Morone chrysops, Rafinesque × M. saxatilis, Walbaum) industry:status and perspectives. Aquaculture Research 37:319–338.

Hallerman, E. M. 1994. Toward coordination and funding of long-term geneticimprovement programs for Striped and hybrid bass Morone sp. Journal of theWorld Aquaculture Society 25:360–365.

Harrell, R. M. 1997. Hybridization and genetics. Pages 217–234 in R. M. Harrell,editor. Striped Bass and other Morone culture. Elsevier Press, Amsterdam.

Heyer, C. J., T. J. Miller, F. P. Binkowski, E. M. Caldarone, and J. M. Rice.2001. Maternal effects as a recruitment mechanism in Lake Michigan YellowPerch (Perca flavescens). Canadian Journal of Fisheries and Aquatic Sciences58:1477–1487.

Hodson, R. G., and M. Hayes. 1989. Hybrid Striped Bass: hatchery phase.Southern Regional Aquaculture Center, Publication 301, Stoneville, Missis-sippi.

Kohler, C. C. 1997. White Bass production and broodstock development. Pages169–184 in R. M. Harrell, editor. Striped Bass and other Morone culture.Elsevier Press, Amsterdam.

Kohler, C. C., R. J. Sheehan, J. J. Myers, J. B. Rudacille, M. L. Allyn, and A. V.Suresh. 2001. Performance comparison of geographic strains of White Bass(Morone chrysops) to produce sunshine bass. Aquaculture 202:351–357.

Labay, A. A., K. Collins, R. W. Standage, and T. M. Brandt. 2004. Gut content offirst-feeding wild darters and captive-reared Dusky Darters. North AmericanJournal of Aquaculture 66:153–157.

Lochmann, S. E., K. J. Goodwin, and C. L. Racey. 2012. Temperature andmaternal effects on hatch rate and length at hatch of hybrid bass (WhiteBass × Striped Bass) larvae. North American Journal of Aquaculture 74:283–288.

Lochmann, S. E., K. J. Goodwin, and C. L. Racey, and C. C. Green. 2009.Variability of egg characteristics among female White Bass Morone chrysopsand the relationship between egg volume and length at hatch of sunshine bass.North American Journal of Aquaculture 71:147–156.

Ludwig, G. M. 1993. Effects of trichlorfon, fenthion, and diflubensuron on thezooplankton community and on production of reciprocal-cross hybrid StripeBass fry in culture ponds. Aquaculture 110:301–319.

Ludwig, G. M., and S. E. Lochmann. 2009. Tank culture of sunshine bassfingerlings without using rotifers. North American Journal of Aquaculture71:224–228.

NASS (National Agriculture Statistics Service). 2006. Census of aquaculture(2005). U.S. Department of Agriculture, Washington, D.C.

Paradis, A. R., P. Pepin, and J. A. Brown. 1996. Vulnerability of fish eggs andlarvae to predation: review of the influence of the relative size of prey andpredator. Canadian Journal of Fisheries and Aquatic Sciences 53:1226–1235.

Rees, R. A., and R. M. Harrell. 1990. Artificial spawning and fry production ofStriped Bass and hybrids. Pages 43–72 in R. M. Harrell, H. H. Kerby, and R.M. Minton, editors. Culture and propagation of striped bass and its hybrids.American Fisheries Society, Southern Division, Striped Bass Committee,Bethesda, Maryland.

SAS Institute. 1990. SAS/STAT user’s guide, version 6, 4th edition. SAS Insti-tute, Cary, North Carolina.

Wang, X., K. E. Ross, E. Saillant, D. M. Gatlin, and J. R. Gold. 2006. Quanti-tative genetics and heritability of growth-related traits in hybrid Striped Bass(Morone chrysops ♀ × Morone saxatilis ♂). Aquaculture 261:535–545.

Wang, X., K. E. Ross, E. Saillant, D. M. Gatlin, and J. R. Gold. 2007. Geneticeffects on carcass-quality traits in hybrid Striped Bass (Morone chrysops ♀× Morone saxatilis ♂). Aquaculture Research 38:973–980.

Woods, L. C. III. 2001. Domestication and strain evaluation of Striped Bass(Morone saxatilis). Aquaculture 202:343–350.

Woods, L. C. III, E. M. Hallerman, L. Douglass, and R. M. Harrell. 1999.Variation in growth rates within and among stocks and families of StripedBass. North American Journal of Aquaculture 61:8–12.

Zar, J. H. 1999. Biostatistical analysis, 4th edition. Pearson Education, UpperSaddle River, New Jersey.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


Click for updates


Efficacy of Iodine for Disinfection of Lake SturgeonEggs from the St. Lawrence River, New YorkMarc Chalupnickia, Dawn Dittmana, Clifford E. Starliperb & Deborah D. Iwanowiczb

a U.S. Geological Survey, Great Lakes Science Center, Tunison Laboratory of Aquatic Science,3075 Gracie Road, Cortland, New York 13045, USAb U.S. Geological Survey, Leetown Science Center, National Fish Health Research Laboratory,11649 Leetown Road, Kearneysville, West Virginia 25430, USAPublished online: 16 Dec 2014.

To cite this article: Marc Chalupnicki, Dawn Dittman, Clifford E. Starliper & Deborah D. Iwanowicz (2015) Efficacy of Iodinefor Disinfection of Lake Sturgeon Eggs from the St. Lawrence River, New York, North American Journal of Aquaculture, 77:1,82-89, DOI: 10.1080/15222055.2014.963768








http://dx.doi.org/10.1080/15222055.2014.963768



North American Journal of Aquaculture 77:82–89, 2015American Fisheries Society 2015ISSN: 1522-2055 print / 1548-8454 onlineDOI: 10.1080/15222055.2014.963768

ARTICLE

Efficacy of Iodine for Disinfection of Lake Sturgeon Eggsfrom the St. Lawrence River, New York

Marc Chalupnicki* and Dawn DittmanU.S. Geological Survey, Great Lakes Science Center, Tunison Laboratory of Aquatic Science,3075 Gracie Road, Cortland, New York 13045, USA

Clifford E. Starliper and Deborah D. IwanowiczU.S. Geological Survey, Leetown Science Center, National Fish Health Research Laboratory,11649 Leetown Road, Kearneysville, West Virginia 25430, USA

AbstractOptimal fish husbandry to reduce the risk of disease is particularly important when using wild fish as the source

for gametes. The propagation and reestablishment of Lake Sturgeon Acipenser fulvescens in New York waters tobecome a viable self-sustaining population is considered a high priority by managers. While standard hatchery eggdisinfection practices have been used to prevent the transmission of diseases, data on the bacterial loads present onegg surfaces following iodine disinfection is lacking. Our study investigated the bacteria present on the outer surfaceof Lake Sturgeon eggs and the effectiveness of an iodine disinfection treatment in eliminating bacteria that could posea threat to egg survival and cause hatchery disease outbreaks. During the springs of 2011–2013, 12 to 41 differentspecies of bacteria were recovered from the outer egg surfaces prior to an iodine treatment; Aeromonas, Pseudomonas,Shewanella, and Chryseobacterium were the most common genera identified. Cohort eggs treated using the standardprotocol of a single treatment of 50 mg/L iodine for 30 min resulted in an average of 57.8% reduction in bacterialCFU/g. While this is a significant reduction, bacteria were not completely eliminated and hatchery managers shouldbe aware that pathogens could remain on Lake Sturgeon eggs following the standard iodine disinfection treatment.

The reestablishment of native fish species that have been lostor have had populations depressed to critically low levels hasbecome a high priority of the New York State Department ofEnvironmental Conservation (NYSDEC) in recent years. Propa-gation of hatchery-reared offspring has been the primary sourceto meet these demands. The containment and control of water-borne pathogens is critical to the functioning of hatchery sys-tems. Organisms that pose a threat to the successful productionof hatchery offspring need to be identified and dealt with tominimize their negative effects. The most effective way to ac-complish this is a proactive and preventative approach throughoptimal husbandry practices to maximize fish health and toprevent introductions of pathogens into the hatcheries. In NewYork, Lake Sturgeon Acipenser fulvescens has been identified


as a critical species for repopulation throughout the state due tocritically depressed numbers and its importance in a functionalecosystem to facilitate control of exotic invaders (i.e., RoundGoby Neogobius melanostomus, zebra mussel Dreissena poly-morpha, quagga mussel D. bugensis). To meet this demand,state and federal hatcheries that typically rear a suite of sportfish species have developed Lake Sturgeon culture protocols us-ing traditional standard disease control procedures. Therefore,incoming fertilized Lake Sturgeon eggs have been disinfectedwith the standard 30-min bath of 50 mg/L of the active ingredientpolymeric iodine (NYSDEC Oneida Lake Fish Culture Station,personal communication). This disinfection process has beenaccepted as a safe iodine concentration for use with no apparentdetrimental effect on egg survival (Bouchard and Aloisi 2002;

82

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5

IODINE DISINFECTION OF LAKE STURGEON EGGS 83

Yesaki et al. 2002; Aloisi et al. 2006; Van Eenennaam et al.2008).

Currently, little is known about the bacterial flora present onthe outer surfaces of Lake Sturgeon eggs prior to or followingiodine disinfection. Fujimoto et al. (2013) recently observeda higher diversity of bacterial communities on egg surfaces athigher stream velocities through early development but did notassess any disinfection protocols. Bouchard and Aloisi (2002)examined the effectiveness of different egg disinfectants withiodine being the most effective; however, those authors onlyrelated the egg disinfection success to hatch survival and not toegg surface bacterial colony enumeration. Common hatchery-cultured species such as Rainbow Trout Oncorhynchus mykiss,Walleye Sander vitreus, Atlantic Salmon Salmo salar, WhiteSturgeon A. transmontanus, Green Sturgeon A. medirostris, andAtlantic Sturgeon A. oxyrinchus oxyrinchus have received moreattention regarding disinfection effectiveness (Cipriano et al.2001; Yesaki et al. 2002; Mohler 2003; Drennan et al. 2006; VanEenennaam et al. 2008; Wagner et al. 2008; Dabrowski et al.2009; Chalupnicki et al. 2011). Those authors corroborated theuse of iodine as a successful disinfectant but related the suc-cesses to egg survival and not to bacterial CFU or directly toreducing bacterial loads. However, Wagner et al. (2008) statedthat to completely control surface bacteria on Rainbow Trouteggs, treatment would require an iodine concentration detrimen-tal to egg survival; in determining the optimal concentration ofiodine, the positive effects of iodine in reducing bacterial loadmust be weighed against the adverse effects on egg survival athigher concentrations. This was also confirmed by Chalupnickiet al. (2011) who noted a significant reduction of egg surfacebacteria of Atlantic Salmon eggs at the standard iodine disin-fecting concentration, but levels necessary for complete bacteriaelimination were not practical for embryo survival.

The objectives of this study were to (1) enumerate and iden-tify the bacteria on the outer surfaces of Lake Sturgeon eggs and(2) evaluate the effectiveness of the standard (NYSDEC) iodinedisinfection protocol to reduce the bacterial load present on theegg surfaces.

METHODSThe St. Lawrence River is the final outlet of the Great Lakes,

connecting the northeast corner of Lake Ontario to the AtlanticOcean. It is 1,197 km long and has a mean annual dischargeof 7,100 m3/s (USACE 2007). The river and its tributaries haveundergone extensive hydroelectric development, but the systemstill holds a resident population of adult Lake Sturgeon belowthe Robert Moses Saunders Dam. The NYSDEC, in collabo-ration with the U.S. Geological Survey and the U.S. Fish andWildlife Service, collected adult Lake Sturgeon from the SouthChannel of the St. Lawrence River near Massena, New York(45.00389◦N, 74.78616◦W) as an annual survey of the popu-lation and to collect gametes for hatchery propagation (Klindt2005).

In spring 2011, we collected 15 mL (50–100 eggs) of “green”(i.e., nonfertilized) eggs from each of eight gravid Lake Sturgeonto characterize and enumerate the bacteria present on the outeregg surfaces. Eggs were collected during the stripping processinto individual, sterile, 25-mL scintillation vials and were kepton ice until they were assayed for bacteria. In the springs of 2012and 2013, eggs were collected to evaluate the effectiveness ofthe standard NYSDEC disinfection process by comparing thebacteria on the surfaces of green eggs with those of fertilized,iodine-treated eggs. We again collected 15 mL of green eggsfrom six (three each in 2012 and 2013) gravid Lake Sturgeonto act as controls. After control samples were collected, theremainder of the eggs stripped from each of the six femaleswere each fertilized with the milt from one male (milt wasmixed with 1 L of well water in 2012 and 2013) for 1 min. Thefluid was decanted and the eggs were gently swirled in a mixtureof fuller’s earth and water for 40 min to coat the adhesive outerjelly layer. The fuller’s earth solution was poured off and theeggs were immersed for 30 min in water containing 50 mg/L ofpolymeric iodine (Argentyne, Argent Chemical Laboratories,Redmond, Washington). Following the immersion treatment,approximately 15 mL (50–100) of eggs were removed, placedin sterile, 25-mL scintillation vials, and kept on ice until bacterialassays and colony enumerations were performed at the Cortland,New York, laboratory approximately 12 h later. Morphometricdata were recorded from the egg-source females prior to theirrelease back into the river.

Aseptic techniques were employed to recover bacteria fromthe surfaces of 20 eggs from each female (10 treated with io-dine and 10 untreated controls). Eggs were weighed and in-dividual eggs were placed in an appropriate amount of sterile0.1% peptone–0.05% yeast extract (pH 7.0; Becton, Dickin-son and Company, Sparks, Maryland) to yield a 1:10 or 1:20(w/v) dilution. The eggs were mixed in the diluent for 30 s at75% maximum setting using a vortex mixer (Velp ScientificaWizard, Neu–Tec Group, Farmingdale, New York). The eggsretained their integrity during mixing. Serial 10-fold dilutionswere prepared in 0.1% peptone–0.05% yeast extract (through1 × 10−4) and 20-µL volumes from each dilution were placedon the surfaces of R2A agar (Becton, Dickinson and Com-pany) plates. The plates were incubated aerobically at 20◦C un-til the resulting bacterial colonies were distinguishable, whichwas within 48–72 h. Bacterial colonies were enumerated fromplates that had the lowest dilutions of single, isolated coloniesand are reported as CFU/g of egg after multiplication by allsample dilution factors. Colonies representing the different mor-phologies were transferred to R2A to ensure purity and developcultures for characterizations. In 2011, bacteria were character-ized using standard phenotypic identification procedures (Kone-man et al. 1992; MacFaddin 2000; Minana-Galbis et al. 2002;Beaz-Hidalgo et al. 2010).

Bacterial cultures developed from egg surfaces from 2012and 2013 were characterized based on the 16S rRNA PCR.The DNA was extracted using the DNA blood and tissue kit

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5

84 CHALUPNICKI ET AL.

(Qiagen, Valencia, California) according to the methods forgram-positive bacteria. The DNA was stored at 4◦C beforeamplification. For PCR amplification, a PCR cocktail contain-ing 1 µM of each universal primer, F63 (5′ - CAG GCC TAACAC ATG CAA GTC - 3′) and R1389 (5′- AGC GGC GGTGTG TAC AAG - 3′) (Marchesi et al. 1998), was added to Go-Taq Green Master Mix (Promega, Madison, Wisconsin). Allprimers were purchased from Integrated DNA Technologies(Coralville, Iowa). The primers were used with a PCR cyclingprofile consisting of a 2-min denaturation step at 94◦C; 35 cy-cles of 45 s at 94◦C, 30 s at 58◦C, and 2 min at 72◦C; and a7-min extension at 74◦C. After amplification, 5 µL of the PCRproduct was identified by electrophoresis at 90 V for 2 h on agel containing 1.2% I.D.NA agarose (FMC Bioproducts, Rock-land, Maine). Amplified PCR fragments averaged 910 bp inlength.

The PCR products were cleaned using QIAquick PCR Purifi-cation Kit (Qiagen). Sequencing reactions were performed usingthe Big Dye Cycle Sequencing Kit (Applied Biosystems, FosterCity, California) according to the manufacturer’s instructionsfor both the forward and reverse primer. The samples were thensubjected to the following PCR cycling profile: 25 cycles of 30 sat 96◦C, 15 s at 58◦C, and 4 min at 60◦C, and a 10-min exten-sion at 72◦C. The PCR sequencing reactions were cleaned withAgencourt CleanSEQ (Beckman Coulter Genomics, BeckmanCoulter, Brea, California) and loaded onto an the 3100 GeneticAnalyzer (Applied Biosystems). The DNA samples were se-quenced in both directions and analyzed with BioEdit softwareversion 7.2.0. Each sequence was compared with sequencesin the National Center for Biotechnology Information (NCBI)GenBank catalog to find similar sequences for use in taxonomicidentification and to confirm similarity among replicates.

Normality tests were performed on each year’s data set us-ing the Shapiro–Wilk normality test (Statistix 8.0, AnalyticalSoftware, Tallahassee, Florida). Statistical analyses (P < 0.05)for the nonnormal data were done using paired t-tests (Statistix8.0, Analytical Software) to test for significant differences inbacterial CFU/g between untreated and iodine-treated groups in2012 and 2013.

RESULTSThe mean lengths of Lake Sturgeon used for egg bacteria

analysis were 1,435, 1,557, and 1,497 mm and weights were 20,25, and 26 kg in 2011, 2012, and 2013, respectively (Table 1).The mean weights of Lake Sturgeon eggs were 0.025, 0.024,and 0.020 g per egg for 2011, 2012, and 2013, respectively. Themean bacterial CFU/g recovered from the outer surfaces of 10eggs per female ranged from 1.75 × 103 to 6.53 × 105 CFU/g(Table 2). A total of 191 bacterial isolates were identified fromthe Lake Sturgeon egg surfaces in 2011 including 12 speciesof Aeromonas (Table 3), with the predominant species beingA. encheleia, which comprised 20.15% of the bacteria iden-tified. Other Aeromonas species frequently recovered were A.

TABLE 1. Total number (N) and mean ± SE length and weight of gravidLake Sturgeon collected from the St. Lawrence River, New York.

Year N Length (mm) Weight (kg)

2011 8 1,435 ± 13.3 20.03 ± 1.372012 3 1,557 ± 43.3 25.7 ± 3.472013 3 1,497 ± 81.9 26.2 ± 4.48

eucrenophila (19.40%) and A. popoffii (17.91%). Two other bac-terial species were recovered in 2011: Pseudomonas fluorescens,which accounted for 20.15% of the bacterial cultures, and She-wanella putrefaciens, which accounted for 5.97%. Aeromonasencheleia, P. fluorescens, A. eucrenophila, and A. popoffii werethe most common bacterial species, comprising 77.61% of thetotal bacteria recovered in 2011.

In a comparison of the viable cell counts before and after thesingle treatment with 50 mg/L iodine, the overall mean countfor untreated eggs in 2012 was 5.76 × 105 CFU/g, whereas, themean for treated eggs was significantly less at 1.08 × 103 CFU/g(P < 0.01; Table 2). Similarly, the mean count for iodine-treatedeggs in 2013 was 8.67 × 102 CFU/g, which was significantlyless (P = 0.02) than the bacterial recovery of 3.09 × 104

CFU/g from untreated cohort eggs. Combining the bacterialcell counts from 2012 and 2013, there was a 57.80% reductionin the total bacteria recovered from iodine-treated eggs. Thegreatest mean bacterial recovery from 10 untreated eggs fromone female from both years was 1.72 × 106 CFU/g, whilethe lowest mean recovery from eggs following treatment withiodine was 2.80 × 102 CFU/g from a female in 2013 (Table 2).

In 2012 and 2013, 80 bacterial isolates were identified fromuntreated eggs and 55 isolates were identified from iodine-treated eggs. We identified 42 species of bacteria from the outersurfaces of untreated eggs and 32 species from the surfaces ofiodine-treated eggs (Table 4). In 2012, there was a similar num-ber of different bacterial species recovered from untreated andtreated eggs, 12 and 15 species, respectively, including sevenspecies that were recovered from both groups of eggs. The bac-teria recovered in greatest frequency from untreated eggs wereChryseobacterium sp. (15.79% of the total bacteria) and Kocu-ria spp. (15.79%). On the other hand, after the treatment withiodine Acinetobacter lwoffii accounted for 17.86% of the bac-teria and Bacillus spp. accounted for 24.99%; combined theycomprised 42.85% of the total bacteria recovered. There was agreater diversity in bacterial species identified from eggs in 2013compared with 2011 and 2012, particularly from untreated eggsfrom which 34 bacteria were identified (Table 4). Nineteen bac-terial species were recovered following treatment with iodine,eight of which were also identified from eggs prior to treatment.Chryseobacterium spp. accounted for 25.87% of bacteria recov-ered from untreated eggs and were the predominate bacteria re-covered. Brevibacterium sp. (11.11%) was the most commonlyrecovered bacterium from iodine-treated eggs in 2013. Whereas

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


TABLE 2. Mean ± SE bacterial CFU/g on the outer surfaces of Lake Sturgeon eggs (10 eggs from each individual) from gravid females collected in 2011(n = 8) and paired (Untreated, Treated = 50 mg/L iodine) mean bacterial CFU from gravid females collected in 2012 and 2013 (n = 6).

2011 2012 2013

Replicate Untreated Untreated Treated Untreated Treated

1 2.47 × 104 ± 1.35 × 104 1.72 × 106 ± 1.68 × 106 1.44 × 103 ± 2.00 × 102 7.76 × 103 ± 3.70 × 103 8.80 × 102 ± 2.72 × 102

2 6.71 × 104 ± 2.56 × 104 1.12 × 104 ± 1.68 × 103 1.28 × 103 ± 3.58 × 102 2.85 × 104 ± 9.39 × 103 2.80 × 102 ± 1.20 × 102

3 6.53 × 105 ± 9.72 × 104 8.60 × 102 ± 1.69 × 102 5.20 × 102 ± 2.65 × 102 5.66 × 104 ± 2.03 × 104 1.44 × 103 ± 5.44 × 102

4 2.37 × 103 ± 3.98 × 102

5 2.94 × 105 ± 3.85 × 104

6 1.75 × 103 ± 3.74 × 102

7 1.03 × 104 ± 1.36 × 103

8 5.39 × 104 ± 1.29 × 104

Mean 1.38 × 105 ± 1.02 × 105 5.76 × 105 ± 5.60 × 105 1.08 × 103 ± 1.74 × 102 3.09 × 104 ± 8.19 × 103 8.67 × 102 ± 2.18 × 102

P-value <0.01 0.02t-value −5.55 −2.70

Bacillus spp. was the most prevalent bacteria from treated eggsin 2012, none were identified from treated eggs in 2013. Fur-thermore, Aeromonas sp. was not identified in 2012 and 2013(Table 4), but was most prevalent from eggs in 2011 (Table 3).

DISCUSSIONThe purpose of disinfecting fish eggs is to eliminate, or at least

greatly reduce, pathogens to contain their spread in the hatchery,particularly when it involves gametes from free-ranging fishes.With a disease prevention strategy that includes disinfectingthe eggs prior to their entering a hatchery, resource managersshould be confident in knowing that precautions to reduce theintroduction of pathogens have been satisfied. However, thisstrategy may be compromised if the accepted egg disinfectionprotocol does not adequately eliminate pathogens. Total disin-fection may not be obtained without using elevated disinfectiontreatment levels that would be detrimental to egg survival.

TABLE 3. Percent composition (%) of bacteria recovered from the outersurfaces of Lake Sturgeon eggs collected in 2011.

Bacteria species % composition

Aeromonas encheleia 20.15Pseudomonas fluorescens 20.15Aeromonas eucrenophila 19.40Aeromonas popoffii 17.91Aeromonas tecta 5.97Shewanella putrefaciens 5.97Aeromonas molluscorum 2.99Aeromonas sp. 2.24Aeromonas hydrophila 1.49Aeromonas eucrenophila 0.75Aeromonas allosaccharophila 0.75Aeromonas caviae 0.75Aeromonas piscicola 0.75Aeromonas veronii bv. sobria 0.75

The results of our study in 2011 show the successful re-covery of bacteria from Lake Sturgeon eggs collected from theSt. Lawrence River. The CFU/g of the present study are sim-ilar to counts by Fujimoto et al. (2013) who reported 5,826CFU/g associated with Lake Sturgeon eggs collected from theBlack River, Michigan, where experimental water flow condi-tions were controlled in a laboratory setting and were similarto the water flow of the St. Lawrence River at the time theeggs were taken. Wagner et al. (2008) and Chalupnicki et al.(2011) also noted similar bacterial counts from untreated Rain-bow Trout and Atlantic Salmon eggs respectively, prior to aniodine disinfection treatment.

Although there was a significant decrease in bacterial countsfrom disinfected eggs in 2012 and 2013, there was still a rela-tively substantial bacterial flora present following the treatment.One explanation for this is the use of green eggs instead of usinguntreated fertilized eggs. Due to current biosecurity protocolsall fertilized eggs need to be treated with iodine after collectionto control pathogen transmission into hatcheries. Thus, the useof green eggs was the only plausible comparison to use, butthese conflictions need to be noted. The finding of flora presentfollowing the disinfection treatment was consistent with bacte-ria counts after treatments reported in the current literature andposes an important consideration of whether it is essential tocompletely eliminate all bacteria from egg surfaces or whetherthere is an acceptable bacterial load for hatcheries. Under thestandard salmonid disinfection protocol of 50 mg/L iodine for30 min, Chalupnicki et al. (2011) reported a 97% reduction inbacterial CFU/g on the outer surface of Atlantic Salmon eggs.Cipriano et al. (2001) showed that certain bacteria are not com-pletely eliminated using the standard federal egg disinfectionprotocol and, therefore, they suggested a second disinfection(i.e., double treatment) to ensure complete elimination of bac-teria. Wagner et al. (2008) noted a significant reduction in thenumber of bacteria on the outer surfaces of Rainbow Trout eggsthat were exposed to a single treatment with 100 mg/L iodine for15 min; however, this only resulted in a small number of eggs

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


TABLE 4. Percent composition (%) of bacteria recovered from the outer surfaces of Lake Sturgeon eggs collected in 2012 and 2013 with and without an iodinedisinfection treatment and the accession and percent identity.

2012 2013

Bacteria taxon Untreated Treated Untreated Treated Accession number % identity

Acetobacteraceae 3.70 JX067726.1 99Acinetobacter lwoffii 17.86 3.45 JF935069.1, HF952698.1,

EF204273.1, HQ696437.1,KC469965.1

98–100

Acinetobacter sp. 5.26 3.57 3.70 HQ634940.1, FJ228151.1,JF799960.1

98–99

Acinetobacterradioresistens

1.72 JN669142.1 99

Agrococcus sp. 7.14 FJ911533.1 99Arthrobacter sp. 7.41 HE613776.1 99Bacillus megaterium 3.57 AB703264.1 99Bacillus pumilus 5.26 10.71 KC844765.1, FR821657.1,

KC632219.1, JX293283.194–98

Bacillus safensis 5.26 3.57 KC843397.1 99Bacillus sp. 5.26 7.14 1.72 KC128896.1, JX083966.1,

JN215493.198–99

Brevundimonas diminuta 1.72 HG000005.1 99Brevibacterium sp. 5.26 3.57 5.17 11.11 AM981204.1 99Brevundimonas

naejangsanensis5.26 3.57 JQ684238.1 99

Chryseobacteriumhaifense

6.90 JQ684226.1, JX100819.1,JX122183.1

97–99

Chryseobacteriumpiscium

3.45 DQ862541.1 99

Chryseobacterium sp. 15.79 15.52 7.41 JX005906.1, JQ995770.1,JX657162.1, JX949547.1,JX287785.1, GQ246712.1

97–99

Comamonadaceae 1.72 AB461020.1 99Curtobacterium sp. 3.70 GU595334.1 97Devosia sp. 1.72 7.41 EF540486.1, FM955869.1 97–99Dyadobacter hamtensis 3.70 JX869991.1 96Flavobacteriaceae 1.72 EU523664.2 99Janibacter sp. 1.72 JX949809.1 98Kaistella sp. 6.90 3.70 HM196764.1 99Kocuria marina 3.57 NR025723.1 98Kocuria palustris 1.72 KC893993.1 99Kocuria rhizophila 10.53 NR074786.1 99Kocuria sp. 5.26 KC470544.1 99Leucobacter sp. 1.72 7.41 GQ344412.1, JN863503.1 98–99Microbacteriaceae 3.70 JQ977360.1 99Microbacterium sp. 3.70 KF358264.1 99Microbacterium

esteraromaticum1.72 KC430861.1 99

Microbacteriumfoliorum

1.72 KC853297.1 99

Microbacteriumoleivorans

1.72 KC764962.1 99

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


TABLE 4. Continued.

2012 2013

Bacteria taxon Untreated Treated Untreated Treated Accession number % identity

Microbacterium oxydans 3.45 KF055008.1, KC179045.1 99Microbacterium

phyllosphaerae3.70 EU714359.1 99

Micrococcus flavus 3.70 KC134360.1 99Micrococcus luteus 10.53 5.17 KF054946.1, FN667800.1,

KC814649.1, JX429841.199

Micrococcusyunnanensis

7.14 1.72 HF570087.1, HE800816.1 99

Moraxella sp. 3.57 GU595358.1 99Mycoplana bullata 1.72 EU977636.1 95Nocardioides sp. 1.72 FR682687.1 99Novosphingobium

panipatense1.72 3.70 JQ647885.1 99

Pantoea sp. 7.14 JN853250.1 99Planobacterium

taklimakanense5.26 JX100825.1 99

Planococcus sp. 3.70 FJ357647.1 99Pseudomonas sp. 7.41 KF202623.1, KC816552.1 99Pseudomonas putida 1.72 AP013070.1 97Rhizobium sp. 1.72 3.70 JX233522.1, KF003414.1 98–99Rhodococcus sp. 3.45 JQ928373.1, HQ256815.1 99Shewanella putrefaciens 5.26 7.14 (Phenotypic identification)Sphingobacterium sp. 3.45 7.41 JF327460.1, EF204462.1 98–100Sphingomonas

mucosissima1.72 HF930759.1 99

Sphingomonas sp. 1.72 KC464782.1, HQ256808.1,KC862017.1, JQ977427.1

96–99

Staphylococcus equorum 3.45 JN089357.1, HE647625.1 99Staphylococcus pasteuri 3.57 FJ217194.1 96Staphylococcus

saprophyticus1.72 GQ480491.1 100

Staphylococcus sp. 1.72 JQ977588.1 99Xanthomonas sp. 1.72 AB680639.1 95Unknown identification 15.79 7.14

in which no bacteria were recovered after an apparent thoroughdisinfection. Although Wagner et al. (2008) suggested a higherdose of iodine, users need to be aware of the possible lethaleffects to eggs of achieving apparent total bacterial elimina-tion. Leary and Pederson (1988) and Fowler and Banks (1991)both noted that 100 mg/L iodine during water hardening waslethal to egg survival for Rainbow Trout and Chinook SalmonO. tshawytscha, respectively. While Bouchard and Aloisi (2002)concluded that 50 mg/L iodine for 30-min during water harden-ing was not lethal to egg survival, they suggested further researchis needed to evaluate the bacteria present on eggs following thistreatment. Other chemicals including hydrogen peroxide and

formalin have been evaluated and are successful at reducing thetotal numbers of bacteria present on egg surfaces, but the con-centrations of these chemicals that are needed to be effective asdecontaminants are typically high and are lethal to egg survival(Rach et al. 1997, 1998; Wagner et al. 2008).

The presence of certain species of bacteria on the outer sur-faces of Lake Sturgeon eggs during this study was highly vari-able, although the effectiveness of an iodine treatment can bevery specific. In the 3 years of this study, we observed distinctbacterial communities on the eggs from each year and withfew strain similarities across years. The dynamics of freshwa-ter bacterial communities can vary greatly and be influenced

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


temporally and spatially on their ability to adhere and surviveon the outer egg surface (Anderson-Glenna et al. 2008; Lamyet al. 2009; Besemer et al. 2012; Fujimoto et al. 2013). As theSt. Lawrence River is the emptying point of all of the LaurentianGreat Lakes, it is no surprise to observe a vast and constantlychanging bacterial community. The shift of using river water towell water as a source during egg fertilization can greatly affectthe bacteria community as aeromonads are typically open-waterstrains, and this shift may be the reason for their disappear-ance in 2012–2013. Aeromonas, Pseudomonas, Shewanella, andChrysobacterium observed in abundant numbers in 2011–2013were eliminated or significantly reduced in our 2012–2013 tri-als. However, the inability to completely eliminate these bac-terial strains with a standard single iodine treatment poses anapparent risk to diseases and introductions of pathogens in theabsence of good husbandry. Cipriano et al. (2001) concludedthat a secondary 100-mg/L iodine treatment following waterhardening was needed to stop the growth of Staphylococcus.Wright and Snow (1975) reported 200 mg/L iodine was neededto completely eliminate Aeromonas liquefaciens on centrarchideggs, while Gee and Sarles (1942) found 56 mg/L iodine waseffective in killing A. salmonicida on trout eggs. Fujimoto et al.(2013) also noted that Proteobacteria, which includes most ofthe identified bacteria found in our study from 2012 to 2013,were present at the bottom of streams and rivers and not in thewater column. That observation is consistent with those in ourstudy of Lake Sturgeon, which primarily spend the majority oftheir time on the bottom.

In conclusion, the present study quantified and identified theaerobic bacterial communities on the outer surfaces of LakeSturgeon eggs in 2011–2013. A significant decrease (57.8%)in the total bacteria was demonstrated following a single io-dine treatment with 50 mg/L for 30 min. However, fish hatcherymanagers should be aware that the standard disinfection pro-tocol currently in place for common hatchery-reared speciesdoes not remove all bacteria from the surfaces of wild-originLake Sturgeon eggs. Therefore, treated eggs could potentiallybe a source for pathogens to be introduced to hatcheries. How-ever, given that the tested treatment levels did not compromiseegg survival, the control of bacterial numbers is perhaps suffi-cient to result in successful hatchery rearing. Additional studiesare needed to better understand the bacterial flora present onouter egg surfaces and to define effective but safe disinfec-tion protocols for gametes taken from free-ranging fishes anddestined for hatchery rearing for population augmentation orrestoration.

ACKNOWLEDGMENTSWe thank Roger Klindt, New York Department of En-

vironmental Conservation, Watertown, New York, and ScottSchlueter, U.S. Fish and Wildlife Service, Cortland, New York,for providing field support. We are also grateful to many otherswho helped in the Lake Sturgeon egg-taking process. This arti-cle is Contribution 1882 of the U.S. Geological Survey, Great

Lakes Science Center. Any use of trade, product, or firm namesis for descriptive purposes only and does not imply endorsementby the U.S. Government.

REFERENCESAloisi, D. R., R. Gordon Jr., N. J. Strazl, J. L. Walker, and T. R. Brady. 2006.

Genoa National Fish Hatchery Lake Sturgeon culture standard operatingprocedures. U.S. Fish and Wildlife Service, Great Lakes-Big Rivers Region,FDS-2006-003, Genoa, Wisconsin.

Anderson-Glenna, M. J., V. Bakkestuen, and N. J. W. Clipson. 2008. Spatialand temporal variability in epilithic biofilm bacterial communities along anupland river gradient. FEMS Microbiology Ecology 64:407–418.

Beaz-Hidalgo, R, A. Alperi, N. Bujan, J. L. Romalde, and M. J. Figueras. 2010.Comparison of phenotypical and genetic identification of Aeromonas strainsisolated from diseased fish. Systematic and Applied Microbiology 33:149–153.

Besemer, K., H. Peter, J. B. Logue, S. Langenheder, E. S. Lindstrom, L. J.Tranvik, and T. J. Battin. 2012. Unraveling assembly of stream biofim com-munities. Multidisciplinary Journal of Microbial Ecology 6:1459–1468.

Bouchard, H. J. III, and D. B. Aloisi. 2002. Investigations in concurrent disin-fection and de-adhesion of Lake Sturgeon eggs. North American Journal ofAquaculture 64:212–216.

Chalupnicki, M. A., H. G. Ketola, C. E. Starliper, and D. Gallagher. 2011.Efficacy and toxicity of iodine disinfection of Atlantic Salmon eggs. NorthAmerican Journal of Aquaculture 73:124–128.

Cipriano, R. C., B. M. Novak, D. E. Flint, and A. C. Cutting. 2001. Reappraisalof the federal fish health recommendation for disinfecting eggs of AtlanticSalmon in iodophor. Journal of Aquatic Animal Health 13:320–327.

Dabrowski, K., K. Ware, M. Jaroszewska, and K Kwasek. 2009. Evaluation ofWalleye embryo survival and larval viability after iodine treatment. NorthAmerican Journal of Aquaculture 71:122–129.

Drennan, J. D., S. E. LaPatra, J. T. Siple, S. Ireland, and K. D. Cain. 2006.Transmisson of White Sturgeon iridovirus in Kootenai River White SturgeonAcipenser transmontanus. Diseases of Aquatic Organisms 70:37–45.

Fowler, L. G., and J. L. Banks. 1991. A safe level of iodophor for treating eggsof fall Chinook Salmon during water hardening. Progressive Fish-Culturist53:250–251.

Fujimoto, M., J. A. Crossman, K. T. Scribner, and T. L. Marsh. 2013. Microbialcommunity assembly and succession on Lake Sturgeon egg surfaces as afunction of simulated spawning stream flow rate. Microbial Ecology 66:500–511.

Gee, L. L., and W. B. Sarles. 1942. The disinfection of trout eggs contaminatedwith Bacterium salmonicida. Journal of Bacteriology 44:111–126.

Klindt, R. M. 2005. New York State Department of Environmental ConservationLake Sturgeon egg take manual. New York State Department of Environmen-tal Conservation, Region 6 Publication, Watertown.

Koneman, E. W., S. D. Allen, W. M. Janda, P. C. Schreckenberger, and W.C. Winn Jr. 1992. Color atlas and textbook of diagnostic microbiology, 4thedition. Lippincott, Philadelphia.

Lamy, D., I. Obernosterer, M. Laghdass, L. F. Artigas, E. Breton, J. D. Grattep-anche, E. Lecuyer, N. Degros, P. Labaron, and U. Christaki. 2009. Temporalchanges of major bacterial groups and bacterial heterotrophic activity dur-ing a Phaeocystis globosa bloom in the eastern English Channel. AquaticMicrobial Ecology 58:95–107.

Leary, R. F., and J. Pederson. 1988. Effects of water hardening eggs in a Betadyneor erythromycin solution on hatching success, development, and the geneticcharacteristics of Rainbow Trout. University of Montana, Population GeneticsLaboratory Report 88/4, Missoula.

MacFaddin, J. F. 2000. Biochemical tests for identification of medical bacteria,3rd edition. Lippincott Williams and Wilkins, Philadelphia.

Marchesi, J., T. Sato, A. Weightman, T. Martin, J. Fry, S.Hiom, D. Dymock,and W. Wade. 1998, Design and evaluation of useful bacterium-specific PCRprimers that amplify genes coding for bacterial 16S rRNA. Applied andEnvironmental Microbiology 64:795–799.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


Minana-Galbis, D., M. Farfan, J. G. Loren, and M. C. Fuste. 2002. Biochemicalidentification and numerical taxonomy of Aeromonas spp. isolated from en-vironmental and clinical samples in Spain. Journal of Applied Microbiology93:420–430.

Mohler, J. W. 2003. A culture manual for the Atlantic Sturgeon. Fish andWildlife Service, Region 5 Publication, Hadley, Massachusetts.

Rach, J. J., M. P. Gaikowski, G. E. Howe, and T. M. Schreier. 1998. Evaluationof the toxicity and efficacy of hydrogen peroxide treatments on eggs of warm-and coolwater fishes. Aquaculture 165:11–25.

Rach, J. J., G. E. Howe, and T. M. Schreier. 1997. Safety of formalin treatmentson warm- and coolwater fish eggs. Aquaculture 149:183–191.

USACE (U.S. Army Corps of Engineers). 2007. John Glenn Great Lakes basinprogram biohydrological information base; appendix A: physical overviewof the Great Lakes-St. Lawrence River System. USACE, Buffalo, New York.

Van Eenennaam, J. P., J. Linares-Casenave, J.-B. Muguet, and S. Doroshov.2008. Induced spawning, artificial fertilization, and egg incubation tech-niques for Green Sturgeon. North American Journal of Aquaculture 70:434–445.

Wagner, E. J., R. E. Arndt, E. J. Billman, A. Forest, and W. Cavender. 2008.Comparison of the efficacy of iodine, formalin, salt, and hydrogen peroxidefor control of external bacteria on Rainbow Trout eggs. North AmericanJournal of Aquaculture 70:118–127.

Wright, L. D., and J. R. Snow. 1975. The effects of six chemicals for disinfectionof Largemouth Bass eggs. Progressive Fish-Culturist 37:213–216.

Yesaki, T. Y., R. Ek, J. Siple, J. P. Van Eenennaam, and S. I. Doroshov. 2002.The effects of iodophor disinfection and transportation on the survival tohatch of fertilized White Sturgeon (Acipenser transmontanus) eggs. Journalof Applied Ichthyology 18:639–641.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


Click for updates


The Effectiveness of Flow-Through or Static CopperSulfate Treatments on the Survival of Golden Shinersand Fathead Minnows Infected with FlavobacteriumcolumnareBradley D. Farmera, David L. Strausa, Benjamin H. Becka & Anita M. Kellyb

a U.S. Department of Agriculture, Agricultural Research Service, Harry K. Dupree StuttgartNational Aquaculture Research Center, Post Office Box 1050, 2955 Highway 130 East,Stuttgart, Arkansas 72160, USAb University of Arkansas at Pine Bluff, Lonoke Agricultural Center, Post Office Box 357, 2001Highway 70, East Lonoke, Arkansas 72086, USAPublished online: 16 Dec 2014.

To cite this article: Bradley D. Farmer, David L. Straus, Benjamin H. Beck & Anita M. Kelly (2015) The Effectiveness ofFlow-Through or Static Copper Sulfate Treatments on the Survival of Golden Shiners and Fathead Minnows Infected withFlavobacterium columnare , North American Journal of Aquaculture, 77:1, 90-95, DOI: 10.1080/15222055.2014.953280








http://dx.doi.org/10.1080/15222055.2014.953280




ARTICLE

The Effectiveness of Flow-Through or Static Copper SulfateTreatments on the Survival of Golden Shiners and FatheadMinnows Infected with Flavobacterium columnare

Bradley D. Farmer,* David L. Straus, and Benjamin H. BeckU.S. Department of Agriculture, Agricultural Research Service,Harry K. Dupree Stuttgart National Aquaculture Research Center, Post Office Box 1050,2955 Highway 130 East, Stuttgart, Arkansas 72160, USA

Anita M. KellyUniversity of Arkansas at Pine Bluff, Lonoke Agricultural Center, Post Office Box 357, 2001 Highway 70,East Lonoke, Arkansas 72086, USA

AbstractFour studies were conducted to compare the effects of copper sulfate (CuSO4), when delivered in either a flow-

through or a static system, on the survival of Golden Shiners Notemigonus crysoleucas and Fathead MinnowsPimephales promelas infected with Flavobacterium columnare. The treatment regimens were administered to fishin well water and were based on the recommended treatment rate (1% of alkalinity). Golden Shiners (experiments 1and 3) and Fathead Minnows (experiments 2 and 4) were treated separately. In experiments 1 and 2, the treatmentrate was 2.0 mg/L CuSO4 applied to a flow-through system daily for five consecutive days; control fish were untreated.Study durations were 7 d with the majority of the mortalities occurring in the first 5 d. Survival rates in the flow-through system were 77% and 69% for CuSO4-treated Golden Shiners and Fathead Minnows, respectively. Survivalin both species was significantly different from that in their untreated control fish, which was 33% for Golden Shinersand 41% for Fathead Minnows. Static applications of CuSO4 were investigated in experiments 3 and 4. Treatmentdoses were 0, 0.5, 1, 2, and 4 mg/L (approximately 0.0, 0.25, 0.5, 1, and 2 times the recommended rate). Survival ofFathead Minnows in this experiment was 10, 28, 52, 47, and 35% in the 0, 0.5, 1, 2, and 4 mg/L CuSO4 treatments,respectively. The corresponding survival of Golden Shiners was 49, 50, 65, 75, and 60% in the 0, 0.5, 1, 2, and 4 mg/LCuSO4 treatments, respectively.

Columnaris disease has long been recognized as capable ofinfecting any fish that swims in freshwater (Beck et al. 2012).The majority of research conducted on this disease has been ori-ented around food fish aquaculture, and information is lackingon how columnaris disease affects other areas of aquaculture,such as baitfish. The baitfish industry accounts for a large por-tion of U.S. aquaculture ($38,000,000 yearly according to NASS2006), and profit margins can be negatively affected by disease.One of the most prevalent diseases of baitfish is columnarisdisease (Stone et al. 1997).

*Corresponding author: [email protected] March 28, 2014; accepted August 2, 2014

Columnaris disease is a bacterial infection caused by theorganism Flavobacterium columnare, a long, slender, nonflag-ellated rod, 0.3–0.7 µm wide and 3–10 µm long; cells exhibitgliding motility (Farmer 2004). Colonies of F. columnare oncytophaga agar are flat, yellow, rhizoid, strongly adherent, andspread across the surface, forming irregular margins (Farmeret al. 2011). These bacteria form columnar aggregates on in-fected tissue that are often referred to as “haystacks.” Based onthe site of infection and appearance of infected tissues, the dis-ease is commonly known as “saddleback,” “fin rot,” or “cotton

90

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5

TREATING COLUMNARIS ON SHINERS AND MINNOWS WITH COPPER SULFATE 91

wool disease.” One common clinical sign of the disease is thepronounced erosion and necrosis of external tissues, the gillsoften being a site of major damage (Davis 1922). Presumptivediagnosis of columnaris disease is based on the presence of theclinical signs mentioned above and by the cell morphology ofF. columnare in wet mounts of tissue scrapings from infectedtissues (Noga 2010).

Adhesion is widely accepted as an essential prerequisite to in-fection; however, adhesion may not necessarily lead to success-ful colonization of the host. Firm bacterial adhesion is thoughtto allow F. columnare to withstand cleansing mechanisms op-erating on the surfaces of the gill, such as water flow, mucussecretion, and the frequent cellular turnover of the respiratoryepithelium (Decostere et al. 1999a, 1999b). The early attach-ment prior to the adhesion is the stage of pathogenesis at whicha chemical theraputant can be effective.

Numerous chemical and antibiotic treatments suggested ascurrently approved or available under authorization of an Inves-tigational New Animal Drug (INAD) exemption are effectiveagainst columnaris disease. Current feeding practices in the bait-fish industry rely heavily on natural foods, making medicatedfeed impractical (Lochmann et al. 2002). That means chemicaltreatments are the most highly utilized, copper sulfate (CuSO4),chloramine-T, potassium permanganate, formalin, and hydro-gen peroxide being commonly used (Boyd 2000; Noga 2010;Bowker et al. 2013). Copper sulfate has recently been shownto be effective against columnaris disease in Channel CatfishIctalurus punctatus (Darwish et al. 2012; Farmer et al. 2012,2013). However, the effectiveness of neither CuSO4 nor anyother compounds against columnaris has been investigated inbaitfish.

The objective of this study was to investigate the effects ofCuSO4 on farm-raised baitfish (Golden Shiners Notemigonuscrysoleucas and Fathead Minnows Pimephales promelas) in-fected with F. columnare. The CuSO4 treatments were adminis-tered in either a flow-through or a static system and a range ofdoses were investigated in static systems in both species.

METHODSOutbreak.—Hatchery-raised Golden Shiners and Fathead

Minnows were obtained from a local baitfish producer. Fishwere held indoors in two 180-L flow-through fiberglasstanks. The tanks were stocked with about 1,000 fish each andwere supplied with 24◦C well water. After one week, the fishwere suspected of developing a columnaris infection, as theywere observed grossly to have clinical signs of the disease, in-cluding “saddleback” and fin erosion with a mortality rate of20–40 (2–4% of the population) fish per day over a 3-d pe-riod. Five fish were necropsied, and wet mounts (400 × ) fromskin scrapes of lesion and gill clips revealed long flexing bacte-ria congregating together (i.e., “haystacks”). Samples from theskin and gill from these same fish were cultured for bacterialpathogens on selective Ordal’s and tryptic soy media with 5%sheep blood (Hawke and Thune 1992). Culture plates were incu-

bated at 28◦C for 24 h and observed for growth. A presumptivediagnosis of columnaris disease was made based on clinicalsigns and morphology of the bacteria.

Fish and experimental design.—After diagnosis, four exper-iments were conducted to assess the effectiveness of CuSO4 ona natural F. columnare epizootic in Golden Shiners and FatheadMinnows using fish from populations diagnosed with colum-naris. In experiments 1 and 2, 18-L tanks were randomly as-signed between the CuSO4 treatment and an untreated control.Stand pipes in test tanks allowed for 10 L of aerated well waterwith a flow rate of 30 mL/min (∼4 turnovers/d) in an “ultralow flow system” (Mitchell and Farmer 2010). Fish were ran-domly selected from the holding tank and groups of 10 fishwere stocked into each tank in four rounds, resulting in 40 fishper tank. Following stocking, tank treatments were assignedaccording to a randomization table. Fish were allowed to accli-mate for 1 h; dead and moribund fish were replaced. Since thefish already had columnaris, the usual 24 h acclimation periodwas reduced to 1 h; otherwise, all fish would have died priorto treatment. Moribund fish removed from the tanks were fixedin formalin for subsequent histological examination. During theexperiment, water temperature ranged from 23.3 to 24.0◦C anddissolved oxygen concentration from 7.2 to 8.6 mg/L. Ammo-nia was measured daily with a Hach DR4000 spectrophotometerwith a Low Range Ammonia test kit (Hach, Loveland, Colorado)from three randomly selected tanks, and measurements rangedfrom 0.8 to 1.7 mg/L total ammonia nitrogen. Total alkalinity(219 mg/L) and total hardness (118 mg/L) were determinedprior to treatment by titration (APHA et al. 2005). The durationof the experiment was 10 d.

In experiments 3 and 4, the treatments were applied statically;the only difference was fish species. The 18-L tanks were ran-domly assigned among five treatments, that is, CuSO4 at ratesof 0.5, 1, 2, and 4 mg/L and an untreated control. Test tankscontained 10 L of aerated well water. Fish were again randomlyselected from the holding tank and stocked in groups of 10 intoeach tank in two rounds, resulting in 20 fish per tank. Fish wereallowed to acclimate for 1 h, after which dead and moribundfish were replaced. During the experiment, water temperatureranged from 22.9 to 23.5◦C and dissolved oxygen from 7.4 to7.9 mg/L. Ammonia was measured daily from three randomlyselected tanks, and measurements ranged from 0.83 to 2.6 mg/Ltotal ammonia nitrogen. On the third and successive days, a 5-Lwater exchange was completed to maintain water quality. To-tal alkalinity (219 mg/L) and total hardness (118 mg/L) weredetermined prior to treatment, as above. The duration of theexperiment was 10 d. In experiments 1 and 2, fish were not fedduring the first 5 days of treatment but were offered food on day6 and throughout the remainder of the studies. Fish were notfed in experiments 3 or 4 i to maintain water quality. Animalcare and experimental protocols were approved by the HarryK. Dupree Stuttgart National Aquaculture Research Center In-stitutional Animal Care and Use Committee and conformed toAgricultural Research Service Policies and Procedures 130.4and 635.1.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5

92 FARMER ET AL.

Copper sulfate treatments.—Copper sulfate (Triangle Brandcopper sulfate; Freeport-McMoRan Sierrita, Green Valley,Arizona) treatments were applied by adding predeterminedamounts of stock solution directly into the tanks. Stock solu-tions were prepared in volumetric flasks at a concentration of10 g/L in deionized water. The target dose for CuSO4 was 2.2 mgL−1 d−1 for experiments 1 and 2 for five consecutive days. TheCuSO4 dose was based on the traditional recommendation of 1%of the alkalinity (MacMillan 1985). Water flow was held con-stant, so the duration of exposure to treatments exponentiallydeclined as a function of the turnover rate (∼5.5 h).

In experiments 3 and 4, the treatments were applied by addingpredetermined amounts of stock solution directly into the tanks.Target doses were 0.5, 1, 2, and 4 mg/L, which corresponded toapproximately 0.25, 0.5, 1, and 2 times the recommended ratebased on the alkalinity.

Bacteriology.—Fish were observed daily for clinical signsassociated with the disease. Fish exhibiting clinical signs andunable to maintain neutral buoyancy were considered moribundand removed for necropsy. Bacteriological samples were cul-tured on selective cytophaga agar (Hawke and Thune 1992) con-taining 5 mg/mL neomycin sulfate and 200 IU/mL polymyxinU. If present, a maximum of three moribund or dead fish weresampled from each tank daily. Cultures were incubated at 28◦Cfor 48 h and then scored as being positive or negative for growth,based on colony morphology (i.e., flat, yellow, rhizoid colonies).Samples were stored at 80◦C until DNA extraction. All DNA ex-tractions were performed according to the manufacturer’s direc-tions using a DNeasy Blood and Tissue Kit (Qiagen, Valencia,California). The extracted template DNA (consisting of catfishgenomic and bacterial DNA) was used for pathogen identityconfirmation by polymerase chain reaction (PCR) by utilizingthe species specific primers of Panangala et al. (2007) (Becket al. 2012; Farmer et al. 2012).

Histology.—Tissue samples were collected from fish in thestock tanks during the columnaris outbreak and immediatelyfixed in Davidson’s solution. Tissue samples were transferred to70% isopropanol 24–48 h after fixation and stored until routineparaffin embedding. Tissues were sectioned to 5–6-µm-thick,mounted on glass slides, and stained with hematoxylin and eosinor Giemsa for a microscopic description of the pathologies as-sociated with the columnaris disease outbreak.

Statistical analysis.—Survival data were analyzed withSigmaPlot 11 (SigmaPlot, San Jose, California) using Kaplan–Meier log-rank survival analysis, and all pair-wise multiple com-parisons used the Holm–Sidak method with adjusted P-values.Treatment was the fixed effect and replication was added as therandom effect. Treatment effects were considered significant atP ≤ 0.05.

RESULTS

Clinical Signs and HistologyMoribund fish displayed signs consistent with an F.

columnare infection; fish were lethargic, and minimal feeding

activity was noted. Gross pathologies were also typical; theskin of moribund fish initially showed depigmentation, whichbecame more diffuse as the infection progressed. Dermal ulcer-ation and cutaneous sloughing exposed the underlying muscletissue, and severely frayed fins, particularly caudal fins, wereobserved. Occasionally, the rostrum was observed having depig-mentation with yellow appearance. Long rod-like bacteria wereobserved in Giemsa-stained sections, and were associated withnecrotic lesions. A representative section is shown in Figure 1.

BacteriologyFlavobacterium columnare was cultured from infected tis-

sues of all fish sampled. Culture was attempted from 40 totalnecropsies, 10 from each experiment. Bacterial growth was con-sidered presumptively positive if at least one colony matchedthe colony morphology of F. columnare. An isolate’s identitywas confirmed by PCR. No other bacterial fish pathogens wereisolated.

HistologyCutaneous lesions featured distinct clusters of homogeneous

rod-shaped bacilli morphologically consistent with F. columnare(Figure 1). Lesions showed evidence of necrosis, typically in-volved the epidermis and dermis, and in some cases infiltratedinto the underlying skeletal muscle. A large number of inflam-matory cells were also associated with lesions.

Experiment 1 and 2 SurvivalThe mean ( ± SE) percent survival for the untreated control

fish in experiment 1 (Golden Shiners) and fish in experiment 2(Fathead Minnows) on day 10 was 33.3 (6.1)% and 40.6 (3.9)%,respectively. The percent survival in experiment 1 and 2 of therepeated application flow-through CuSO4-treated Golden Shin-ers and Fathead Minnows was 76.7 (5.0)% and 68.8 (3.7)%,

FIGURE 1. A representative hematoxylin and eosin section; Giemsa-stainedfrom the dermal lesion of a Golden Shiner colonized by F. columnare. The arrowindicates the presence of bacteria with long rod cell morphology, consistent withF. columnare infection.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


respectively. The increase in survival relative to that of the un-treated control fish was 43.4% for CuSO4-treated Golden Shin-ers and 28.2% for Fathead Minnows. Results of the survivalcurve analysis indicate that the rate of survival in CuSO4-treatedfish was significantly different from that of the untreated controlfish (Figure 2).

Experiment 3 and 4 SurvivalThe mean ( ± SE) percent survival for the untreated con-

trol fish in experiment 3 (Fathead Minnows) and experiment 4(Golden Shiners) on day 7 was 10.0 (3.87)% and 48.3 (11.7),respectively. The percent survival in experiment 3 of a singlestatic application of 0.5, 1, 2, and 4 mg/L CuSO4 was 28.3(15.9), 51.7 (7.26), 46.7 (11.7), and 35 (5.0)%, respectively.The percent survival in experiment 4 of a single static applica-tion of 0.5, 1, 2, and 4 mg/L CuSO4 was 50.0 (20.0), 65.0 (2.9),75.0 (2.9), and 60.0 (5.0)%, respectively. Results of the survivalcurve analysis indicate the rate of survival in all CuSO4-treatedFathead Minnows was significantly different from that of theuntreated control fish (Figure 3), whereas experiment 4 showedno differences between any of the treatment groups, includingthe untreated controls (Figure 4).

DISCUSSIONThe sum of these results is not surprising, CuSO4 has been

found to be effective for treating columnaris in a number ofstudies (Darwish et al. 2012: Farmer et al. 2012, 2013). Re-peated flow-through applications of CuSO4 were shown to

FIGURE 2. Kaplan–Meier survival curve of Fathead Minnows infected withF. columnare and treated with CuSO4 as a static bath. Treatment rates of 0.5,1.0, 2.0, and 4.0 mg/L CuSO4 resulted in survival rates significantly differentfrom that of the untreated control group (P ≤ 0.05). Statistical differences fromuntreated groups are indicated by asterisks.

FIGURE 3. Kaplan-Meier survival curve of Golden Shiners infected with F.columnare and treated with CuSO4 as a static bath. No statistical differencesbetween any of the treatments and the untreated controls were detected by theanalysis (P ≤ 0.05).

improve survival in catfish during an F. columnare epizootic(Farmer et al. 2012). What is novel is the response to staticdoses of CuSO4 over a variety of rates ranging from 0.25 to2 times the recommended rate. The results from experiment

FIGURE 4. Kaplan–Meier survival curve of Golden Shiners (GS) and FatheadMinnows (FHM) infected with F. columnare and treated with CuSO4 five timeson five consecutive days in flow-through well water. Treatment with 2 mg/LCuSO4 resulted in survival rates significantly different from those of the un-treated control groups (P ≤ 0.05). Statistical differences from untreated groupsare indicated by asterisks.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5

94 FARMER ET AL.

4 did not decisively confirm the effectiveness of the statictreatment, but this outcome was likely because of the high vari-ability in the nontreated control. In one replicate, only had 30%of the fish succumbed to the disease, while the other two repli-cates averaged 62.5%. If there had been a fourth replicate, it ispossible an outlier analysis could have deemed one tank outsidethe 95% confidence limit and possibly resulted in significantdifferences in CuSO4-treated fish versus nontreated fish.

It is difficult to amend an ongoing, natural epizootic into alaboratory study (Farmer et al. 2012). The disease must be ob-served, diagnosed, and confirmed and then must be of a singlepathogen etiology. The animals must be moved to the experi-mental units, which is an obvious stressor. The disease must alsocontinue on its natural course in this new environment. The cul-mination of these factors makes conducting studies of this kindrisky and challenging and can make the results difficult to inter-pret accurately, as exemplified in experiment 4. Future researchshould attempt to validate these results in an industry setting.

The described histopathologies were consistent with previ-ously described F. columnare–induced pathologies from con-trolled challenge studies (Darwish et al. 2012). In contrast toresults shown here, a previous study found that CuSO4 as a1 mg/L static bath was not effective in controlling mortalityof Channel Catfish experimentally infected with F. columnare(Thomas-Jinu and Goodwin 2004). Darwish et al. (2012) alsosuggest that such a desired effect might be masked when theinfection treated is severe or at an advanced pathogenesis stage.One possible explanation for these differences is that a naturalexposure is better suited for therapeutic intervention than areexperimental infection models.

The mechanism by which CuSO4 exerts a beneficial effectagainst an F. columnare infection in Channel Catfish is prob-ably multifaceted. Copper sulfate can have a direct inhibitoryeffect on bacteria (in the water or on the fish) and parasitesthrough displacing essential metals from their native bindingsites or through ligand interactions (Farmer et al. 2012). Thisdisplacement results in alteration of the conformational struc-ture of nucleic acids, proteins, the oxidative phosphorylationcascade, and osmotic balance (Borkow and Gabbay 2005).

In summary, CuSO4 is effective in increasing the survivalof both Golden Shiners and Fathead Minnows infected with F.columnare. The results suggest that, if it is an option, repeatedflow-through treatments are more effective. If static treatmentsare used, dosing CuSO4 at 1% or possibly 0.5% of the alkalin-ity should produce the best results; going any higher may becounterproductive. As always, water chemistry must be takeninto consideration, as CuSO4 can be toxic in water with lowalkalinity. Note also that chemotherapeutant effects are greatlyaffected by organic matter and will be rendered inert over timein pond applications. In addition, CuSO4 is toxic to planktonand pond treatment can be problematic, especially during thegrowing season. Future studies should investigate the applica-tion of CuSO4 (possibly a single application) for effectivenesson natural outbreaks of columnaris in ponds.

ACKNOWLEDGMENTSWe appreciate Cindy Ledbetter for her dedicated technical

assistance throughout the course of the study and the Harry SaulMinnow Farm in DeValls Bluff, Arkansas, for donating fish.Mention of trade names or commercial products in this articleis solely for the purpose of providing specific information anddoes not imply recommendation or endorsement by the U.S.Department of Agriculture.

REFERENCESAPHA (American Public Health Association), American Water Works Associ-

ation, and Water Pollution Control Federation. 2005. Standard methods forthe examination of water and wastewater, 21st edition. APHA, Washington,D.C.

Beck, B. H., B. D. Farmer, D. L. Straus, C. Li, and E. Peatman. 2012. Putativeroles for a rhamnose binding lectin in Flavobacterium columnare patho-genesis in Channel Catfish Ictalurus punctatus. Fish Shellfish Immunology33:1008–1015.

Borkow, G., and J. Gabbay. 2005. Copper as a biocidal tool. Current MedicinalChemistry 12:2163–2175.

Bowker, J. D., D. Carty, J. T. Trushenski, M. P. Bowman, N. Wandelear, andM. D. Matthews. 2013. Controlling mortality caused by external columnarisin Largemouth Bass and Bluegill with chloramine-T or hydrogen peroxide.North American Journal of Aquaculture 75:342–351.

Boyd, C. E. 2000. Water quality in warmwater fish ponds. Auburn UniversityAgricultural Experiment Station, Auburn, Alabama.

Darwish, A. M., A. J. Mitchell, and D. L. Straus. 2012. Evaluation of a four hourstatic copper sulfate treatment against experimental infection of Flavobac-terium columnare in Channel Catfish (Ictalurus punctatus). Journal of FishDiseases 43:688–695.

Davis, H. S. 1922. A new bacterial disease in freshwater fishes. U.S. Bureau ofFisheries Bulletin 38:63–67.

Decostere, A., F. Haesebrouck, E. Van Driessche, G. Charlier, and R.Ducatelle. 1999a. Characterization of the adhesion of Flavobacteriumcolumnare (Flexibacter columnaris) to gill tissue. Journal of Fish Diseases 22:465–74.

Decostere, A. H., J. F. Turnbull, and G. Charlier 1999b. Influence of water qual-ity and temperature on adhesion of high and low virulence Flavobacteriumcolumnare strains to isolated gill arches. Journal of Fish Diseases 22:1–11.

Farmer, B. D. 2004. Improved methods for the isolation and characterizationof Flavobacterium columnare. Master’s thesis. Louisiana State University,Baton Rouge.

Farmer, B. D., B. H. Beck, A. J. Mitchell, and D. L. Straus. 2013.Pretreating Channel Catfish with copper sulfate affects susceptibilityto columnaris disease. North American Journal of Aquaculture 75:205–211.

Farmer, B. D., B. H. Beck, and D. L. Straus. 2012. Effectiveness of cop-per sulfate and potassium permanganate on Channel Catfish infected withFlavobacterium columnare. North American Journal of Aquaculture 74:320–329.

Farmer, B. D., A. J. Mitchell, and D. L. Straus. 2011. The effect of high totalammonia concentration on the survival of Channel Catfish experimentallyinfected with Flavobacterium columnare. Journal of Aquatic Animal Health23:162–168.

Hawke, J. P., and R. L. Thune. 1992. Systemic isolation and antimicrobialsusceptibility of Cytophaga columnaris from commercially reared ChannelCatfish. Journal of Aquatic Animal Health 4:109–113.

Lochmann, R., N. Stone, and E. Park. 2002. Baitfish feeds and feeding prac-tices. Southern Regional Aquaculture Center, Publication 121, Stoneville,Mississippi.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


MacMillan, J. R. 1985. Infectious diseases. Pages 405–496 in C. S. Tucker,editor. Channel Catfish culture. Elsevier, Amsterdam.

Mitchell, A. J., and B. Farmer. 2010. Evaluation of an ultra low-flow waterdelivery system for small experimental tanks. North American Journal ofAquaculture 72:195–200.

NASS (National Agriculture Statistic Service). 2006. Census of aquaculture(2005), volume 3, special studies, part 2, AC02-SP-2. NASS, Washington,D.C.

Noga, E. J. 2010. Fish disease: diagnosis and treatment. Blackwell ScientificPublications, Ames, Iowa.

Panangala, V. S., C. A. Shoemaker, and P. H. Klesius. 2007. TaqMan real-time polymerase chain reaction assay for rapid detection of Flavobacteriumcolumnare. Aquaculture Research 38:508–517.

Stone, N., E. Park, L. Dorman, and H. Thomforde. 1997. Baitfish culture inArkansas: Golden Shiners, Goldfish, and Fathead Minnows. University ofArkansas at Pine Bluff, Pine Bluff Cooperative Extension Program, Publica-tion MP386, Pine Bluff.

Thomas-Jinu, S., and A. E. Goodwin. 2004. Acute columnaris infection in Chan-nel Catfish Ictalurus punctatus (Rafinesque): efficacy of practical treatmentsfor warmwater aquaculture ponds. Journal of Fish Diseases 27:23–28.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


Click for updates


Gonad Development in Triploid Ornamental Koi Carpand Results of Crossing Triploid Females with DiploidMalesBoris Gomelskya, Kyle J. Schneidera, Ammu Anil1ab & Thomas A. Delomasa

a Aquaculture Research Center, Kentucky State University, 103 Athletic Road, Frankfort,Kentucky 40601, USAb Present address: Department of Fisheries and Allied Aquacultures, Auburn University, 203Swingle Hall, Auburn, Alabama 36849, USA.Published online: 31 Dec 2014.

To cite this article: Boris Gomelsky, Kyle J. Schneider, Ammu Anil1 & Thomas A. Delomas (2015) Gonad Developmentin Triploid Ornamental Koi Carp and Results of Crossing Triploid Females with Diploid Males, North American Journal ofAquaculture, 77:1, 96-101







http://dx.doi.org/10.1080/15222055.2014.963766




COMMUNICATION

Gonad Development in Triploid Ornamental Koi Carpand Results of Crossing Triploid Females with Diploid Males

Boris Gomelsky,* Kyle J. Schneider, Ammu Anil,1 and Thomas A. DelomasAquaculture Research Center, Kentucky State University, 103 Athletic Road, Frankfort,Kentucky 40601, USA

AbstractGonad development in 4-year-old triploid and diploid orna-

mental koi, a variant of Common Carp Cyprinus carpio, from cor-responding heat-shocked and control progenies was investigated.Diploid males were normally mature. Triploid males from heat-shocked progeny demonstrated development of testes typical fortriploid fish; triploid males did not release sperm and their testeshad a pinkish color and were significantly reduced in size. Diploidfemales were normally mature and their gonadosomatic indices(GSIs) varied from 7.5% to 30.7% and the mean value was 21.3%.Triploid females had unexpectedly well-developed ovaries, whichwere filled with fully grown oocytes; their GSIs varied from 4.2%to 30.1% and the the mean value was 17.0%. Four triploid koifemales released large quantities (from 260,000 to 394,500 eggsper female) of ovulated eggs after hormonal injection. Eggs fromtriploid females were fertilized with sperm from normal diploid koimales. Mass mortality of hatched larvae occurred at the swim-upstage, but about 32,000 swim-up larvae were obtained and stockedfor further rearing. A total of 248 juveniles (or less than 1% fromthe number of stocked larvae) were collected from outdoor tanks.Ploidy analysis of juveniles (n = 110) showed that most of themwere aneuploid with ploidy ranging from 2.3n to 2.9n with a meanvalue of 2.6n; two juveniles were diploid (2n). This shows thattriploid koi females produced aneuploid eggs with a ploidy rangefrom haploid to diploid level with the modal ploidy level around1.5n, similar to the production of aneuploid spermatozoa observedearlier for triploid males in fish.

The primary purpose of induced polyploidy in aquacultureand fisheries is to obtain triploid fish, i.e., fish which have threehaploid chromosome sets in karyotypes. Artificially obtainedtriploid fish are considered to be genetically sterile, i.e., they arenot capable of producing viable progeny. Triploid fish are alsocharacterized by complete or partial reduction of the gonads(Benfey 1999; Piferrer et al. 2009).

The present article describes gonad development in 4-year-old triploid and diploid ornamental koi, a variant of Common

*Corresponding author: [email protected] address: Department of Fisheries and Allied Aquacultures, Auburn University, 203 Swingle Hall, Auburn, Alabama 36849, USA.Received April 21, 2014; accepted August 29, 2014

Carp Cyprinus carpio, from corresponding heat-shocked andcontrol progenies and presents the results of crossing triploidfemales with normal diploid koi males. Initially, analyzed pro-genies were obtained in a study on the effect of ploidy on scale-cover pattern in linear koi (Gomelsky et al. 2012). In that studytriploid linear fish in heat-shocked progeny exhibited a non-typical scale cover pattern characterized by the appearance ofadditional scales on the body; this type of scale cover was termed“multi-scaled linear.” All analyzed multi-scaled linear fish (n =43) were triploid. The control (not shocked) progeny obtainedfrom the same koi parents consisted of typical linear and scaledfish (Gomelsky et al. 2012). In 2012, when the fish reached theage of 4 years, it was proposed to complete the experiment andcollect data on gonad development. All fish from the controlprogeny were sacrificed and dissected; however, because dis-secting multi-scaled linear females from heat-shocked progenydemonstrated unexpectedly well-developed ovaries, the remain-ing fish were saved in order to investigate the reproductive abilityof triploid females in crosses.

METHODSGonad development in 110 scaled and typical linear fish

from the control progeny and 57 multi-scaled linear fish fromthe heat-shocked progeny was investigated in 2012. Fish weight,TL, and gonad weight were recorded; the gonadosomatic index(GSI) was determined as the percentage of gonad weight to fishweight. Digital photographs of the two lateral sides of all multi-scaled linear fish from the heat-shocked progeny and of somefish from the control progeny were taken using a Nikon D7000digital camera. Each multi-scaled linear fish has a unique scaledistribution pattern that can be used to identify individuals muchlike a fingerprint, as was recently described for mirror carp,another variant of Common Carp, by Huntingford et al. (2013).

96

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5

COMMUNICATION 97

In the previous study (Gomelsky et al. 2012), digital photographsof all multi-scaled linear fish, whose triploidy was determinedby two methods (comparison of erythrocyte nuclear size usinga Coulter counter and flow cytometric determination of DNAcontent), were taken for identification. Based on comparisonof images, individual fish with proven triploidy were identifiedamong the multi-scaled linear fish dissected in the present study.

In 2013, four, 5-year-old, multi-scaled linear females fromheat-shocked progeny were artificially spawned and crossedwith normal diploid koi males. Eggs from each female werefertilized with sperm taken from different koi males (one maleper cross). The technique of artificial spawning was the sameas described by Gomelsky et al. (2012). To induce ovarian fol-licle maturation in females and spermiation in males, fish par-ents were injected with carp pituitary extract (Argent ChemicalLaboratories, Redmond, Washington) at 3 mg/kg. Total weightof the eggs obtained from each female was recorded; asample of eggs taken from each female was weighed, and thenumber of eggs enumerated to determine the number of eggsper gram. Eggs were artificially inseminated in plastic bowls andwere treated with a water : cow milk mixture (volumetric ratio8:1) to remove egg adhesiveness. Embryos were incubated inMcDonald jars. The percentage of live embryos was determinedapproximately 20 h after fertilization at the early organogenesisstage; at this stage live embryos are easily distinguished fromunfertilized white eggs. Hatched larvae were collected in meshhapas (small, fine-mesh cages) placed in flow-through racewaytanks. Most swim-up larvae obtained in crosses were stockedin separate 20-m3 outdoor tanks for rearing; some larvae werestocked in aquaria in a recirculation system. After 4 monthsof rearing, tanks were drained and all juveniles were collected,weighed, and measured.

Ploidy of fish breeders as well as samples of the obtainedprogenies was determined by flow cytometric analysis of ery-throcyte nuclear DNA content using an Accuri C6 flow cytome-ter (Becton, Dickinson and Company). Blood samples were col-lected from the caudal vein in 3.0-mL Vacutainer tubes (Becton,Dickinson and Company) containing lithium heparin and placedon ice. For each blood sample, 1.0 µL of heparinized blood wasplaced into 12 × 75-mm polystyrene test tubes (Becton, Dick-inson and Company); 0.5 µL of heparinized blood taken fromLargemouth Bass Micropterus salmoides was also added to eachtest tube to act as an internal staining control. Then, 500 µL ofpropidium iodide staining solution (Biosure) was added to thetubes. After the 10-min incubation period in darkness, a sam-ple was placed on the sample injection port and 40,000 eventswere recorded for each sample. The relative DNA content wasdetermined as the ratio of sample fluorescence peak intensityto the internal standard (Largemouth Bass) fluorescence peakintensity.

RESULTSData on mean weight and GSI of 4-year-old fish are presented

in Table 1. Diploid males from control progeny were normally

mature and released sperm from the genital pore; GSI was 6.4± 1.8% (mean ± SD). Multi-scaled linear males from heat-shocked progeny demonstrated the development of testes typicalfor triploid fish; these males did not release sperm from thegenital pore and their testes had a pinkish color and were reducedin size. Mean GSI of multi-scaled males from heat-shockedprogeny (1.1%) was about six times smaller than that of malesfrom control progeny (6.4%) (see Table 1). Among the 29 multi-scaled linear males analyzed from heat-shocked progeny, six fishwhose triploidy was confirmed in the previous study (Gomelskyet al. 2012) were identified based on scale cover patterns inphotographs. The GSIs of these males varied from 0.6% to1.7% and had a mean value of 1.1%, which is exactly the sameas the mean value of GSI for all dissected multi-scaled linearmales from the heat-shock progeny (1.1%, see Table 1).

Diploid females from control progeny were mature and hadovaries filled with fully grown oocytes; GSI varied from 7.5%to 30.7% and had a mean value of 21.3% (SD, ± 4.4%) (seeTable 1). The appearance of one female from control progenyand its ovaries are shown in Figure 1. Investigation of multi-scaled linear females from heat-shocked progeny showed thatmany of them had well-developed ovaries, which were filledwith fully grown oocytes; GSI varied from 4.2% to 30.1% andthe mean value was 17.0 (SD, ± 7.5%) (see Table 1). Among28 multi-scaled linear females dissected in the present study,six fish were identified as those whose triploidy was determinedearlier. The GSIs of these fish varied from 6.0% to 26.1% andhad a mean value of 16.2%, which is close to the mean valueof the GSI for all dissected multi-scaled linear females fromheat-shocked progeny (17.0%, see Table 1). The appearance ofone identified triploid multi-scaled linear female and its ovariesare shown in Figure 2.

In the spring of 2013, 5-year-old multi-scaled linear fe-males from heat-shocked progeny were artificially spawnedand crossed with normal diploid koi males. Six females withprofoundly swollen abdomens were selected and hormonallyinjected. Four females ovulated after the injections and thestripped eggs looked typical for koi (Common Carp). Data onweights and reproductive characteristics of multi-scaled linearfemales as well as results of crosses are presented in Table 2. Theweight of stripped eggs obtained from females varied from 450to 778 g or from 12.0% to 23.6% of female weight. Numbersof eggs per gram, total numbers of eggs obtained from females,and percentages of live embryos are presented in Table 2. Ingeneral, the production of eggs from females, embryo incuba-tion, and hatching of larvae proceeded normally. However, massmortality of larvae occurred in the hapas for several days afterhatching; large quantities of hatched larvae did not swim upand died. Nevertheless, from 3,000 to 14,950 swim-up larvae(or 2.9–5.5% from number of live embryos the next day afterfertilization) were obtained in progenies (see Table 2); a total ofabout 32,000 swim-up larvae were obtained.

Most swim-up larvae obtained in crosses were stocked inseparate 20-m3 outdoor tanks for rearing. Survival of fish during

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5

98 GOMELSKY ET AL.

TABLE 1. Mean weight and GSI of 4-year-old koi from control diploid and heat-shocked triploid progenies.

GSI (%)Type of Scale cover Number of Mean ± SDprogeny types Fish ploidy Fish sex analyzed fish (n) weight (kg) Mean ± SD Range

Control Scaled andtypical linear

2n Males 53 1.57 ± 0.31 6.4 ± 1.8 1.4–10.0

Females 57 2.03 ± 0.45 21.3 ± 4.4 7.5–30.7

Heat-shocked

Multi-scaledlinear

3n Males 29 2.14 ± 0.46 1.1 ± 0.6 0.3–2.9

Females 28 2.61 ± 0.76 17.0 ± 7.5 4.2–30.1

4 months of rearing in outdoor tanks was low and varied from0.2% to 1.2% in different progenies; a total of 248 juveniles werecollected from the tanks. The numbers of collected juveniles perprogeny and their mean TLs are presented in Table 2. About 40%of juveniles had different morphological abnormalities, whichmostly were manifested as head deformities and a reduction ofdorsal fins.

Some larvae (600 larvae from females 1–3 and 900 larvaefrom female 4) were stocked for rearing in aquaria of recir-culation system. No survivors were observed in aquaria of therecirculating system at 5 months after larvae were stocked.

Triploidy of multi-scaled linear females, which were used incrosses, was confirmed by flow cytometric analysis. The relativeDNA content of the four females varied from 2.54 to 2.58 and

FIGURE 1. (A) Diploid scaled koi female from control progeny (TL =51.1 cm, weight = 2.36 kg). (B) Appearance of ovary in female’s body cavity(gonad weight = 644 g, GSI = 27.3%).

had a mean value 2.57 (SD, ± 0.02). The relative DNA contentdetermined for the five diploid koi males (four males used incrosses plus one male, which was hormonally injected but notused in crosses) varied from 1.70 to 1.75 and had a mean value1.72 (SD, ± 0.02). The ratio of mean DNA content in females tomean DNA content in diploid fish (2.57:1.72) was 1.49, whichis very close to a theoretical value of 1.50, thus confirmingtriploidy of females.

Ploidy of 110 juveniles (from 2 to 61 fish per progeny, seeTable 2) was determined by flow cytometry. Distribution ofjuveniles with regard to ploidy is shown in Figure 3. Most of thejuveniles were aneuploid and ploidy ranged from 2.3n to 2.9nand had a mean value of 2.6n; two juveniles were diploid (2n)and one fish had ploidy of 3.9n.

FIGURE 2. (A) Triploid multi-scaled linear koi female from heat-shockedprogeny (TL = 52.7 cm, weight = 2.74 kg). (B) Appearance of ovary in female’sbody cavity (gonad weight = 714 g, GSI = 26.1%).

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5

TAB

LE

2.R

esul

tsof

indu

ced

spaw

ning

oftr

iplo

idko

ifem

ales

and

thei

rcr

osse

sw

ithdi

ploi

dko

imal

es.

Wei

ghto

fst

ripp

edeg

gsN

umbe

rof

swim

-up

larv

aeN

umbe

rof

juve

nile

sfr

omou

tdoo

rta

nks

Fem

ale

num

ber

Fem

ale

wei

ght

(kg)

g%

a

Num

ber

ofeg

gspe

rgr

am

Tota

lnum

ber

ofst

ripp

edeg

gs%

ofliv

eem

bryo

sn

%b

n%

c

Mea

n±

SDju

veni

les

tota

lle

ngth

(cm

)

Num

ber

ofju

veni

les

anal

yzed

for

ploi

dy

13.

3077

823

.650

739

4,44

669

.014

,950

d5.

517

61.

210

.0±

3.2

612

3.60

492

13.7

640

314,

880

64.0

5,85

0d2.

936

0.6

13.5

±4.

023

33.

7271

419

.250

936

3,42

650

.08,

580d

4.7

320.

415

.9±

2.9

244

3.76

450

12.0

578

260,

100

26.0

3,00

0d4.

44

0.2

18.2

2

a Det

erm

ined

from

fem

ale

wei

ght.

b Det

erm

ined

from

num

ber

ofliv

eem

bryo

sth

ene

xtda

yaf

ter

fert

iliza

tion.

c Det

erm

ined

from

the

num

ber

ofst

ocke

dla

rvae

.d Fr

omth

ese

num

bers

600

(fro

mfe

mal

es1–

3)or

900

larv

ae(f

rom

fem

ale

4)w

ere

stoc

ked

inaq

uari

a;th

ere

stof

the

larv

aew

ere

stoc

ked

inou

tdoo

rta

nks.

99

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5

100 GOMELSKY ET AL.

FIGURE 3. Percent distribution of juveniles (n = 110) obtained from triploidkoi females with regard to ploidy.

DISCUSSIONAs mentioned above, in the previous study (Gomelsky et al.

2012) all 43 analyzed multi-scaled linear fish from heat-shockedprogeny were triploid. On this basis, it can be expected with ahigh probability that all multi-scaled linear fish dissected inthe present study were also triploid. The 12 dissected multi-scaled linear fish (six males and six females), whose triploidywas confirmed in the previous study, had gonad developmentsimilar to other fish from this group. The present study usedflow cytomery to confirm triploidy in four additional multi-scaled linear females, which were crossed with normal diploidkoi males.

Multi-scaled linear males from heat-shocked progeny de-monstrated a development of testes typical for triploids (Cherfaset al. 1994a; Benfey 1999; Piferrer et al. 2009). Testes of thesemales had pinkish (not white) color because of the absenceof sperm and were reduced in size compared with the testes ofdiploid males. Mean values of GSI for triploid and diploid malesrecorded in the present study (1.1% and 6.4%, respectively) wereclose to the corresponding mean GSI values reported earlier byCherfas et al. (1994a) for Common Carp (0.9% for triploid malesversus 4.7% for diploid males). Cherfas et al. (1994a) observedthat a few Common Carp triploid males released small amountsof sperm (with abnormal consistency) when stripped. In thepresent study, none of the 29 analyzed triploid males releasedsperm from the genital pore.

It is generally accepted that induced triploidy in fish affectsthe reproductive system of females more strongly than it doesin males. Usually triploid females develop very small ovariesthat do not contain advanced vitellogenic oocytes (Benfey 1999;Piferrer et al. 2009). In contrast, triploid koi females analyzedin this study developed large ovaries; in some triploid femalesGSI reached 25–30%. Earlier, Cherfas et al. (1994a) had also re-ported that some triploid Common Carp females had unexpect-edly well-developed mature ovaries; in some triploid femalesGSI reached 10%. Wu (1990) also noticed that a few CommonCarp triploid females had developed gonads containing oocytes

at maturation stage. No information on any crosses of triploidCommon Carp females has been previously reported.

In the present study, triploid koi females reacted normallyto hormonal injection and released large quantities of ovulatedeggs. Variability of ovulated eggs produced by triploid femaleswas not studied especially, but visually no asynchrony in eggsize was noticed as was described before for triploid females(Benfey 1996, 1999). Cases of artificially obtained triploid fe-males in fish being able to produce some progenies in crossesare very rare. Goudie (1988) reported the production of triploidoffspring in crosses of triploid females with triploid males inGrass Carp Ctenopharyngodon idella presuming that triploidfemales, the same as triploid males, produced aneuploid 1.5ngametes. Benfey (1996) described that no offspring survivedthrough the prehatch period after eggs produced by one triploidBrook Trout Salvelinus fontinalis female were fertilized by nor-mal haploid sperm. In contrast to females, reproductive abil-ity of triploid males has been studied in many fish species.As a rule, triploid males produced aneuploid spermatozoa withploidy range from haploid (n) to diploid (2n) level and the modalploidy level around 1.5n. This was shown by flow cytometricanalysis of sperm produced by triploid males in Rainbow TroutOncorhynchus mykiss (Benfey et al. 1986), Barfin Flounder Ve-rasper moseri (Mori et al. 2006), Tench Tinca tinca (Linhartet al. 2006), Atlantic Cod Gadus morhua (Peruzzi et al. 2009),and some other fish species. Aneuploidy of sperm produced bytriploid males can result in the appearance of aneuploid proge-nies in their crosses with normal diploid females. For example,Peruzzi et al. (2009) showed that in Atlantic Cod the ploidy oflarvae, generated by crossing triploid males with normal diploidfemales, ranged from nearly diploid (2.1n) to nearly triploid(2.75n) values and the mean ploidy level was around 2.4n. Mostjuveniles obtained from triploid females in the present studyhad ploidy ranging from diploid to nearly triploid (2.9n) levelsand the mean ploidy value was 2.6n. This showed that triploidkoi females produced aneuploid eggs with a ploidy range fromhaploid to diploid level and the modal ploidy level around 1.5n,similar to the production of aneuploid spermatozoa observedearlier for triploid males in fish.

Aneuploid progeny obtained from triploid koi females hadlow viability. Mass mortality of larvae was observed after hatch-ing; the yield of swim-up larvae was about 3–5% of the numberof live embryos the next day after fertilization (usually for nor-mal koi progenies this index is 60–80%). Survival of fish inoutdoor tanks was also very low, overall less than 1% (juvenilesurvival of normal koi progenies in tanks is usually 50–75%).Nevertheless, observed survival of some aneuploid koi juvenilescan be considered as an unexpected result. Usually, crossingtriploid male fish with diploid females results in the appearanceof completely nonviable aneuploid progeny (Lincoln 1981; Lin-coln and Scott 1984; Cherfas et al. 1994a; Peruzzi et al. 2009).Cherfas et al. (1994a) reported complete mortality of all em-bryos obtained after the fertilization of Common Carp eggswith sperm released by triploid males. In contrast, in the present

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5

COMMUNICATION 101

study some viable aneuploid juveniles from triploid femaleswere obtained in the same species. It should be noted that pro-genies obtained by using sperm from triploid males are usuallynot numerous because of the small amount of sperm producedby triploid males and the low concentration of spermatozoa. Inthe present study, triploid koi females had large fecundity, whichmade it possible to obtain more than 30,000 swim-up larvae inspite of high mortality at the swim-up stage. When mass quan-tities of progeny are raised, the probability that some aneuploidfish could survive increases.

Not many cases have been recorded of aneuploid fish reach-ing the juvenile stage. Ueda et al. (1991) demonstrated the via-bility of aneuploid fingerlings with 3.5n ploidy which were ob-tained by crossing allotriploid Rainbow Trout × Brook Trouthybrid males with normal diploid Rainbow Trout females. Cher-fas et al. (1994b) reported that fingerlings obtained by crossingallotriploid hybrid females of Prussian Crucian Carp Carassiusgibelio × Common Carp with diploid Common Carp maleswere aneuploid and had chromosome numbers from 110 to 190,which corresponded to a ploidy range from 2.2n to 3.8n. Zhangand Arai (1999) obtained aneuploid progeny with fish ploidyranging between 3n to 4n in the loach Misgurnus anguillicauda-tus (also known as Oriental Weatherfish) by crossing artificiallyproduced triploid males with natural tetraploid females.

Survivors in progenies arising from crossing triploid maleswith diploid females are typically believed to result from fertil-ization of eggs by euploid spermatozoa, which can arise withlow frequency from atypical triploid meiosis. For example, VanEenennaam et al. (1990) showed that rare surviving juvenilesobtained by crossing triploid Grass Carp males with diploid fe-males were diploid. In the present study, two diploid juvenileswere found in progeny obtained from triploid females. One ju-venile obtained from one triploid koi female had a ploidy level3.9n. Apparently, this fish resulted from the spontaneous sup-pression of the second meiotic division in an aneuploid (1.45n)egg. Earlier, Ueda et al. (1991) suggested a similar mechanismfor the explanation of the appearance of aneuploid fish with3.5n ploidy in progeny obtained by crossing allotriploid Rain-bow Trout × Brook Trout hybrid males with diploid RainbowTrout females.

Further studies will be aimed at a more detailed investigationof reproductive features of triploid koi females. The influenceof aneuploidy on the development of the reproductive system infish obtained from triploid females will also be investigated.

ACKNOWLEDGMENTSSupport for this study was provided by U.S. Department of

Agriculture National Institute of Food and Agriculture grantKYX-80-12-24A to Kentucky State University (KSU) and

Kentucky’s Regional University Trust Fund to the AquacultureProgram as KSU’s Program of Distinction.

REFERENCESBenfey, T. J. 1999. The physiology and behavior of triploid fishes. Reviews in

Fisheries Science 7:39–67.Benfey, T. J. 1996. Ovarian development in triploid Brook Trout (Salvelinus

fontinalis). Page 357 in F. W. Goetz and P. Thomas, editors. Proceedings ofthe fifth international symposium on the reproductive physiology of fish. FishSymposium 95, Austin, Texas.

Benfey, T. J., I. I. Solar, G. De Jong, and E. M. Donaldson. 1986. Flow-cytometric confirmation of aneuploidy in sperm from triploid Rainbow Trout.Transactions of the American Fisheries Society 115:838–840.

Cherfas, N. B., B. Gomelsky, N. Ben-Dom, Y. Peretz, and G. Hulata. 1994a.Assessment of triploid Common Carp (Cyprinus carpio L.) for culture. Aqua-culture 127:11–18.

Cherfas, N. B., B. I. Gomelsky, O. V. Emelyanova, and A. V. Recoubratsky.1994b. Induced diploid gynogenesis and polyploidy in Crucian Carp, Caras-sius auratus gibelio (Bloch), × Common Carp, Cyprinus carpio L., hybrids.Aquaculture Research 25:943–954.

Gomelsky, B., K. J. Schneider, R. P. Glennon, and D. A. Plouffe. 2012. Effectof ploidy on scale-cover pattern in linear ornamental (koi) Common CarpCyprinus carpio. Journal of Fish Biology 81:1201–1209.

Goudie, C. 1988. Some triploid Grass Carp can be induced to spawn. U.S. Fishand Wildlife Service, Research Information Bulletin 88-24, Washington, D.C.

Huntingford, F. A., F. L. Borcato, and F. O. Mesquita. 2013. Identifying indi-vidual Common Carp Cyprinus carpio using scale pattern. Journal of FishBiology 83:1453–1458.

Lincoln, R. F. 1981. Sexual maturation in triploid plaice (Pleuronectes platessa)and plaice × flounder (Platichthys flesus) hybrids. Journal of Fish Biology19:415–426.

Lincoln, R. F., and A. P. Scott. 1984. Sexual maturation in triploid Rain-bow Trout (Salmo gairdneri, Richardson). Journal of Fish Biology 25:385–392.

Linhart, O., M. Rodina, M. Flajshans, N. Mavrodiev, J. Nebesarova, D. Gela,and M. Kocour. 2006. Studies on sperm of diploid and triploid tench, Tincatinca (L.). Aquaculture International 14:9–25.

Mori, T., S. Saito, C. Kishioka, and K. Arai. 2006. Aquaculture performance oftriploid flounder Verasper moseri. Fisheries Science 72:270–277.

Peruzzi, S., G. Rudolfsen, R. Primicerio, M. Frantzen, and G. Kauric. 2009.Milt characteristics of diploid and triploid Atlantic Cod (Gadus morhua L.).Aquaculture Research 40:1160–1169.

Piferrer, F., A. Beaumont, J.-C. Falguiere, M. Flajshans, P. Haffray, and L.Colombo. 2009. Polyploid fish and shellfish: production, biology and applica-tions to aquaculture for performance improvement and genetic containment.Aquaculture 293:125–156.

Ueda, T., R. Sato, M. Iwata, A. Komaru, and J. Kobayashi. 1991. The viable3.5n trouts produced between diploid females and allotriploid males. JapaneseJournal of Genetics 66:71–75.

Van Eenennaam, J. P., R. K. Stocker, R. G. Thiery, N. T. Hagstrom, and S. I.Doroshov. 1990. Egg fertility, early development and survival from crossesof diploid female × triploid male Grass Carp (Ctenopharyngodon idella).Aquaculture 86:111–125.

Wu, C. 1990. Retrospects and prospects of fish genetics and breeding researchin China. Aquaculture 85:61–68.

Zhang, Q., and K. Arai. 1999. Aberrant meioses and viable aneuploid progeny ofinduced triploid loach (Misgurnus anguillicaudatus) when crossed to naturaltetraploids. Aquaculture 175:63–86.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


Click for updates


Proximate Composition of Bioflocs in Culture SystemsContaining Hybrid Red Tilapia Fed Diets with VaryingLevels of Vegetable Meal InclusionJosé Antonio López-Elíasa, Angélica Moreno-Ariasa, Anselmo Miranda-Baezab, Luis RafaelMartínez-Córdovaa, Martha Elisa Rivas-Vegab & Enrique Márquez-Ríosc

a Departamento de Investigaciones Científicas y Tecnológicas, Universidad de Sonora,Boulevard Colosio s/n, Edificio 7J, Hermosillo, Sonora 83000, Mexicob Laboratorio de Tecnologías de Cultivo de Organismos Acuáticos, Universidad Estatal deSonora, Carretera Huatabampo Km 5, Navojoa, Sonora 85800, Mexicoc Departamento de Investigación y Posgrado en Alimentos, Universidad de Sonora, BoulevardColosio s/n, Edificio 7F, Hermosillo, Sonora 83000, MexicoPublished online: 31 Dec 2014.

To cite this article: José Antonio López-Elías, Angélica Moreno-Arias, Anselmo Miranda-Baeza, Luis Rafael Martínez-Córdova,Martha Elisa Rivas-Vega & Enrique Márquez-Ríos (2015) Proximate Composition of Bioflocs in Culture Systems ContainingHybrid Red Tilapia Fed Diets with Varying Levels of Vegetable Meal Inclusion, North American Journal of Aquaculture, 77:1,102-109, DOI: 10.1080/15222055.2014.963767








http://dx.doi.org/10.1080/15222055.2014.963767




ARTICLE

Proximate Composition of Bioflocs in Culture SystemsContaining Hybrid Red Tilapia Fed Diets with Varying Levelsof Vegetable Meal Inclusion

Jose Antonio Lopez-Elıas and Angelica Moreno-AriasDepartamento de Investigaciones Cientıficas y Tecnologicas, Universidad de Sonora, Boulevard Colosios/n, Edificio 7J, Hermosillo, Sonora 83000, Mexico

Anselmo Miranda-Baeza*Laboratorio de Tecnologıas de Cultivo de Organismos Acuaticos, Universidad Estatal de Sonora,Carretera Huatabampo Km 5, Navojoa, Sonora 85800, Mexico

Luis Rafael Martınez-CordovaDepartamento de Investigaciones Cientıficas y Tecnologicas, Universidad de Sonora, Boulevard Colosios/n, Edificio 7J, Hermosillo, Sonora 83000, Mexico

Martha Elisa Rivas-VegaLaboratorio de Tecnologıas de Cultivo de Organismos Acuaticos, Universidad Estatal de Sonora,Carretera Huatabampo Km 5, Navojoa, Sonora 85800, Mexico

Enrique Marquez-RıosDepartamento de Investigacion y Posgrado en Alimentos, Universidad de Sonora,Boulevard Colosio s/n, Edificio 7F, Hermosillo, Sonora 83000, Mexico

AbstractBiofloc culture systems, which are based on the development of microorganisms that recycle inorganic nutrients and

organic matter, may contribute to the nutrition of some farmed species. Juvenile red tilapia (Nile Tilapia Oreochromisniloticus × Mozambique Tilapia O. mossambicus) cultured in saltwater were fed pelleted diets in which 0, 33, 67, or100% of the fish meal was substituted with a vegetable meal mix (corn, wheat, and sorghum meals). The proximatecomposition of the biofloc produced in the culture systems was evaluated. Four experimental diets and one controldiet (isocaloric and isoproteic) were randomly assigned to 15 experimental tanks. Samples of biofloc were periodicallycollected to measure the total suspended solids, organic matter, and ash content and to determine the protein, lipid,and carbohydrate contents. At the end of the study, variables describing red tilapia production were determined. Thebiofloc volume, total suspended solids, ash, and organic matter showed significant differences among treatments, butcarbohydrate (33.0–39.0%), lipid (2.6–3.5%), and protein (23.7–25.4%) levels were similar. No significant differenceswere observed in red tilapia survival, final biomass, or feed conversion ratio. We conclude that the substitution of fishmeal with vegetable meal in the pelleted feed had no adverse effect on the production response of saltwater-culturedred tilapia.

*Corresponding author: [email protected] June 25, 2014; accepted September 5, 2014

102

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5

DIET EFFECT ON BIOFLOC PROXIMATE COMPOSITION 103

“Biofloc” is a term used to designate the aggregation ofparticles in a colloidal dispersion. These aggregates containheterotrophic and nitrifying bacteria, microalgae, fungi, anddetritus as well as flagellates, ciliates, rotifers, and nematodes(Buford et al. 2004). Tilapias can be successfully farmed athigh concentrations of suspended solids; thus, tilapia speciesare good candidates for culture in biofloc systems (Emerencianoet al. 2013) since the suspended organic matter, detritus, and un-consumed food can provide an important nutritional source. Theretention of residues in a culture system and their conversion tobiofloc can be achieved by constant aeration and water columnagitation to keep nutrients available (Crab et al. 2012); it is alsonecessary to add carbon sources as a substrate for bacteria thatconsume organic matter (Azim et al. 2008).

The microbiota associated with bioflocs is able to removetoxic compounds and recycle nutrients (Audelo-Naranjo et al.2012). Bioflocs are also a source of live feed for farmed organ-isms, allowing for the partial substitution of formulated feeds(Tacon et al. 2002; De Schryver et al. 2008) while also reducingproduction costs and pollution. The proximate composition ofthe biofloc may vary according to several factors, such as culturedensity, aeration intensity, dissolved oxygen, carbon source andavailability, temperature, salinity, and pH as well as the type ofmicrobial community that has developed (Avnimelech 2007; DeSchryver et al. 2008; Martınez-Cordova et al. 2014).

Most of the previous biofloc experiments have focused on (1)calculating the effects of bioflocs on water quality characteristics(e.g., pH, dissolved oxygen, and total ammonia nitrogen [TAN])and (2) the production response of cultured organisms (e.g.,weight gain, feed conversion ratio [FCR], total biomass, andsurvival). However, there is little available information aboutthe biochemical composition of biofloc and how it varies inrelation to the type of diet provided or the age of the culture.

In recent years, the increased demand for and scarcity offish meal have caused increases in the price of formulated feed,thereby affecting the economic feasibility of aquaculture andnecessitating studies for the development and evaluation of al-ternative diets. Some studies have suggested biofloc technol-ogy (BFT) as an alternative to complement the nutritional re-quirements of shrimps (pink shrimp Farfantepenaeus paulensis,pinkspot shrimp F. brasiliensis, and Pacific white shrimp Litope-naeus vannamei) and Nile Tilapia Oreochromis niloticus (Azimand Little 2008; Kuhn and Boardman 2008; Ballester et al. 2010;Emerenciano et al. 2011). The benefits of biofloc-based culturesare related to the better use of nutrients, diminished productionof wastewater, reduction in farm surface area, decreases in en-vironmental impacts, and reduction in dependence on fish meal(Moreira de Souza et al. 2013).

Azim and Little (2008) studied the quality of biofloc that de-veloped in a culture system wherein two formulated feeds withdifferent protein content (24% and 35%) were administered toNile Tilapia; their results indicated that lipid and protein com-position and energy content did not show significant differencesamong treatments. Similar results were reported by Megahed

(2010) in comparing the quality of biofloc developed during theculture of green tiger prawns Penaeus semisulcatus that were feddiets with different levels of crude protein (31.15, 21.6, 18.45,and 16.25%); the protein (19.9%) and lipid (11.8%) content ofthe biofloc did not differ among treatments.

The feasibility of culturing tilapia in saltwater representsan opportunity for farming these fishes in enclosure facilitiesusing seawater (Miranda-Baeza et al. 2010). Advancement inthe knowledge of tilapia nutrition, including the use of BFT,may contribute to the sustainability of tilapia culture.

Based on the above information, we sought to evaluate twohypotheses: (1) in a BFT system, the unconsumed food, re-mains of dead organisms, and metabolites can be used bymicroorganisms to produce biomass of high nutritional qual-ity; and (2) the origin of the protein (animal or vegetable) inthe pelleted feed does not significantly affect biofloc compo-sition. We focused on evaluating the variation in biofloc prox-imate composition as a function of the fish meal substitutionlevel in the diet and the age of the tilapia culture in a BFTsystem.

METHODSExperimental culture.—Juvenile red tilapia (Nile Tilapia ×

Mozambique Tilapia O. mossambicus hybrids; mean individualweight ± SD = 16.2 ± 1.0 g) were stocked in fifteen 200-L, in-door tanks at a density of 2.0 kg/m3 (12–15 fish/tank dependingon the individual weight). The experimental fish were main-tained under constant aeration supplied by an electric blower(0.333 hp).

One control diet and four experimental diets (isocaloric, isoli-pidic, and isoproteic) were evaluated in triplicate over a 42-dperiod. In the formulation and preparation of the experimentaldiets, fish meal was included at four different levels: 0% (T0),10% (T10), 20% (T20), and 30% (T30). To obtain the desiredprotein level (35%), a mixture of vegetable meals (corn, wheat,and sorghum) was included at different proportions for eachdiet (100, 67, 33, or 0% of the fish meal was substituted, respec-tively; Table 1). The control diet consisted of a commercial feedformulated for tilapia, and the crude protein content and crudefat content were similar to those of the experimental diets (35%and 8%, respectively). In total, 1,200 g of feed were adminis-tered to each tank, with a feeding rate equal to 3% of the tilapiabiomass.

Fifteen days prior to the stocking of red tilapia, the indoorculture tanks (0.6 m wide, 1.00 m deep) were filled with 200 L(0.71-m depth) of filtered seawater (35‰) and then were inocu-lated with mature, microbe-dominated biofloc at 1%. The proto-col for biofloc formation and maturation consisted of adding or-ganic matter (commercial pulverized pellet at 0.025 g·L−1·d−1;∼500-µm diameter, 35% crude protein) and a carbon source(unrefined granulated sugar) to maintain a C:N ratio of 20:1.

Temperature was not controlled; rather, it was completely de-pendent on the local temperature variation (24–28◦C during the

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5

104 LOPEZ-ELIAS ET AL.

TABLE 1. Formulation and composition of experimental diets fed to redtilapia that were cultured in saltwater systems with biofloc (T0 = 0% fish meal;T10 = 10% fish meal; T20 = 20% fish meal; T30 = 30% fish meal).

Diet

Ingredient orcomponent T30 T20 T10 T0

Ingredients(g/kg of diet)

Soybeanconcentrate

200.0 373.9 549.8 725.0

Fish meal 300.0 200.0 100.0 0.0Cornmeal 376.1 150.0 114.5 150.0Wheat meal 50.0 150.0 150.0 50.0Sorghum meal 10.3 51.5 10.0 5.0Fish oil 32.5 43.5 44.6 38.9Soybean lecithin 10.0 10.0 10.0 10.0Vitamin premix 8.0 8.0 8.0 8.0Mineral premix 5.0 5.0 5.0 5.0Sodium phosphate 5.0 5.0 5.0 5.0Choline chloride 2.0 2.0 2.0 2.0Vitamin C 1.0 1.0 1.0 1.0Butylated

hydroxytoluene0.1 0.1 0.1 0.1

Proximatecomposition (%dry weight basis)

Crude protein 35 35 35 35Crude fat 7 8 8 7.196Crude fiber 3.128 3.526 4.157 4.996Ash 2.31 2.497 2.768 3.16

experiment). There was no water exchange during the experi-ment. Freshwater was added only to compensate for losses fromevaporation. Oxygen was maintained at high levels (4–6 mg/L).Dissolved oxygen and temperature were measured twice perday by using a multiparameter YSI 550A meter. The pH wasmeasured daily with a Denver Instruments UP-10 pH meter, andsalinity was determined every 3 d with a refractometer (AquaticEco-Systems).

During the first 10 d of culture, the carbon : nitrogen (C:N)ratio was maintained at approximately 20:1 by estimating thecarbon and nitrogen added in the feed and sugar. For the remain-der of the trial, the C:N ratio was reduced to 12:1.

Total ammonia nitrogen and nitrite-nitrogen measurement.—Samples (250 mL) of culture water were collected every 2 d andfiltered with glass-microfiber filters (Whatman GF/C, 47-mmdiameter). Total ammonia nitrogen was then determined by thesalicylate method (Hach 8155) and nitrite-nitrogen (NO2-N)was determined by the diazotization method (Hach 8507) inaccordance with the manufacturer’s protocols (Hach Co. 2005).

Biofloc volume estimation.—Samples of culture water wereplaced in 1-L Imhoff cones and were left for 45 min to allowsedimentation. Suspended solids on the walls were removedcarefully with a glass stirrer, and the samples were allowed tosettle for an additional 15 min. Biofloc volume (mL/L) was thenrecorded.

Biofloc biomass evaluation.—Samples (500 mL) of culturewater were taken from each tank on days 0, 7, 21, 35, and 42 ofthe experiment; the samples were stored in plastic bottles andwere kept frozen at −80◦C until analysis. Water samples werefiltered through glass microfiber filters (Whatman GF/C), andthe filters were dried in a convection oven (70◦C for 72 h) tocalculate the total suspended solids (TSS) content. The bioflocbiomass was calculated based on the difference in filter weightbefore and after filtration and drying. The ash content was eval-uated by incinerating the sample in a muffle furnace at 450◦Cfor 4 h. The organic matter in the biofloc was estimated by cal-culating the difference in weight between the TSS and the ashcontent.

Biofloc proximate composition.—To evaluate the total pro-tein content of the biofloc, the filtered samples were treatedwith a sodium hydroxide solution (NaOH, 0.1 N) for extrac-tion of protein according to Lowry’s method (Lowry et al.1951, as modified by Lopez-Elıas et al. 1995). Total lipid con-tent of the biofloc was determined by extraction with a mix-ture of chloroform–methanol–water (5:2.5:1) following Pande’smethod (Pande et al. 1963, as modified by Lopez-Elıas et al.1995). Total carbohydrate was evaluated by calculating the dif-ference between the weight of organic matter and the sum oftotal protein and lipid.

Production response of red tilapia.—The production re-sponse of red tilapia in the treatments was evaluated based onbiometric indices (individual weight and biomass), survival, andFCR.

Statistical analysis.—Biofloc development was depictedgraphically with mean volume, total particulate matter, andproximate composition. Repeated-measures ANOVA was usedto compare biofloc volume, total particulate matter, organic mat-ter, and ash as well as water quality variables among treatmentsfor each sampling day. A chi-square test of independence wasused to verify whether biofloc composition (carbohydrate, lipid,protein, and ash percentages) significantly differed among treat-ments (Sokal and Rohlf 1981). Survival (arcsine transformed),mean final weight, final biomass, and FCR of red tilapia wereanalyzed using one-way ANOVA. Tukey’s honestly significantdifference test (Zar 1996) was used to compare and rank treat-ment means. Statistical analyses were performed in Statisticaversion 7.0 for Windows (Statsoft, Inc.).

RESULTS

Water Quality VariablesThe pH averaged between 7.68 and 7.72 at the beginning of

the experiment and averaged 7.65–7.75 at the end of the study,

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


TABLE 2. Global mean ( ± SD) volume, total suspended solids (TSS), ash content, and organic matter content of biofloc in treatments with different levels offish meal inclusion in pelleted diets for red tilapia (C = control with 30% fish meal; see Table 1 for definition of experimental diets T0–T30). Within a row, meanswith different letters are significantly different (P < 0.05).

Variable T0 T10 T20 T30 C

Volume (mL/L) 9.22 ± 2.81 x 8.67 ± 3.74 yx 6.78 ± 5.26 zy 4.22 ± 2.63 z 5.56 ± 2.18 zyTSS (mg/L) 537.3 ± 108.1 y 351.3 ± 57.7 z 402.3 ± 151.2 z 404.0 ± 120.4 z 354.6 ± 88.9 zAsh (mg/L) 213.0 ± 50.1 y 116.9 ± 31.1 z 156.1 ± 102.17 zy 167.0 ± 63.9 zy 132.3 ± 60.6 zOrganic matter (mg/L) 324.2 ± 63.0 y 234.4 ± 38.3 z 246.1 ± 59.6 z 237.0 ± 59.6 z 222.3 ± 36.4 z

without significant differences among treatments (P > 0.05).Total ammonia nitrogen concentration was low in all treatments,and there were no significant differences (P > 0.05). A slightlyhigher mean TAN of 0.63 mg/L was recorded for the control,while T30 exhibited a lower value of 0.40 mg/L. The NO2-N concentration began to increase on day 6, and maximumlevels of 3–4 mg/L were observed on days 9–12. During theentire experiment, the mean NO2-N values ranged from 0.82to 0.85 mg/L and did not significantly differ among treatments(P > 0.05).

Biofloc VolumeMean biofloc volume showed significant differences among

treatments (P < 0.05); mean volume was 9.22 mL/L for T0,8.67 mL/L for T10, 6.78 mL/L for T20, and 4.22 mL/L forT30, whereas the mean for the control was 5.56 mL/L (Table 2).From the beginning to the end of the study, the highest bioflocvolumes (mL/L) were recorded for treatments with a lower fishmeal content. In general, the biofloc volume increased with

the age of the culture, but the rate of increase differed amongtreatments (Figure 1). On day 42 (the end of the culture period),the highest biofloc volume (19 mL/L) was observed for T0,while the lowest volume was observed for T30 (6.5 mL/L).

Biofloc BiomassMean TSS in the biofloc ranged from 351.3 mg/L in T10 to

537.3 mg/L in T0 (Table 2). The TSS concentration for T0 wassignificantly higher than values observed for the other treatments(P < 0.05). For T10, T20, T30, and the control, TSS values werestable through time (Figure 2A).

On day 0 (i.e., 1 d before the fish were stocked), lowerTSS concentrations within the range of 200–300 mg/L wererecorded, and there were no significant differences among treat-ments. On day 7 (the second sampling day), an increase ofapproximately 100 mg/L was observed in all treatments. Fromday 21 to day 42, T0 consistently presented the highest TSSconcentrations (Figure 2A).

FIGURE 1. Mean ( + SD) biofloc volume (mL/L) obtained in the treatments with different levels of fish meal inclusion in pelleted diets for red tilapia (C =control with 30% fish meal; experimental diets: 0 = 0% fish meal; 10 = 10% fish meal; 20 = 20% fish meal; 30 = 30% fish meal).

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


FIGURE 2. Changes in mean ( ± SD) values of (A) total suspended solids,(B) ash content, and (C) organic matter content of the biofloc that developedin red tilapia culture with different levels of fish meal inclusion in the pelleteddiets (see Figure 1 for definition of treatments).

Ash content in the biofloc of T10, T20, T30, and the con-trol was maintained at stable levels throughout the experi-ment; no significant changes related to culture age were ob-served (Figure 2B). The global mean ash content of the bioflocranged from 116.9 to 213.0 mg/L, with significant differences(Table 2).

Mean organic matter content in the biofloc varied from 222.3to 324.2 mg/L, and significant differences among treatmentswere identified (P < 0.05; Table 2). From day 7 to the end ofthe experiment, organic matter showed low variation in T10,T20, T30, and the control, whereas it increased over the cultureperiod in T0 (Figure 2C).

FIGURE 3. Changes in mean ( ± SD) values of (A) total protein, (B) totalcarbohydrate, and (C) total lipid of the biofloc developed in red tilapia culturewith different levels of fish meal inclusion in the pelleted diets (see Figure 1 fordefinition of treatments).

Biofloc Proximate CompositionThe protein concentration in the biofloc ranged from 46 to

70 mg/L on day 0 and increased until day 21, when the highestmeans were recorded; a general decrease was observed there-after (Figure 3A). At the end of the culture period (day 42),T0 had the highest protein concentration (132 mg/L). Consid-ering the whole experiment, the mean protein content in thebiofloc varied from 23.7% to 25.4% and showed no significantdifferences among treatments (P > 0.05; Table 3).

On day 0, the carbohydrate concentration in the bioflocwas similar among treatments, with means ranging from 70to 80 mg/L. Carbohydrate values reached 120–150 mg/L on day

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


TABLE 3. Proximate composition (global mean ± SD) of the biofloc (dry basis) developed with different levels of fish meal inclusion in pelleted diets for redtilapia (C = control with 30% fish meal; see Table 1 for definition of experimental diets T0–T30).


Ash (%) 39.3 ± 3.6 33.0 ± 5.4 35.8 ± 10.1 40.4 ± 4.9 35.9 ± 8.3Carbohydrate (%) 33.0 ± 3.9 39.0 ± 3.7 35.5 ± 4.7 32.2 ± 8.9 36.5 ± 5.2Protein (%) 24.5 ± 2.3 24.1 ± 5.3 23.7 ± 5.5 25.4 ± 9.3 23.9 ± 5.3Lipid (%) 2.6 ± 0.3 3.5 ± 0.6 2.8 ± 0.7 3.2 ± 0.8 2.9 ± 0.5

7 and increased continuously over the remainder of the cultureperiod (Figure 3B). Considering the entire experimental pe-riod, the carbohydrate content of the biofloc varied from 32.2%to 39.0%, without significant differences among treatments(P > 0.05; Table 3).

The lipid content did not show any clear trend during theculture period (Figure 3C). Mean lipid content ranged from 2.6%to 3.5% and did not significantly differ among the treatments(P > 0.05; Table 3).

Production Response of Red TilapiaThe survival of red tilapia did not present significant among-

treatment differences (P > 0.05); nevertheless, survival tendedto be greater when fish meal inclusion in the pelleted diet waslower (Table 4). Final individual weight was significantly differ-ent among treatments (P < 0.05): the lowest weight (45.0 g) wasobserved for T0, while the highest weight (64.1 g) was recordedfor T20. Final fish biomass varied from 5.8 to 7.1 kg/m3, andFCR ranged from 1.2 to 1.7; in both cases, no significant differ-ences were observed among the treatments (P > 0.05; Table 4).

DISCUSSIONAmmonia is the main metabolic waste of fish and is found in

two different forms in the water column: un-ionized ammonia(NH3) is the form most toxic to fish, whereas ionized ammonia(NH4

+ ) is less toxic (El-Sayed 2006). The form of ammoniapresent is closely correlated with pH; at high pH, un-ionizedammonia is dominant. Timmons et al. (2002) indicated that TANlevels less than 3.0 mg/L are safe for warmwater fish farmingif the pH is between 6.5 and 8.5. In our study, pH ranged from7.53 to 7.92, and under these conditions the nontoxic form of

ammonia was dominant; additionally, the TAN concentrationwas maintained under 0.7 mg/L due to the prior maturation of thebiofloc. The C:N ratios used in our experiment (20:1 and 12:1)contributed to maintaining a high concentration of heterotrophicbacteria, which have been demonstrated to remove ammonia andorganic matter from the water column (Avnimelech 2007).

Nitrite-nitrogen is an intermediate compound in the nitrifi-cation process; the increase in NO2-N during the first days wasa normal process attributable to the low rate at which nitrite-oxidizing bacteria (Timmons et al. 2002) increased in the bioflocaggregates. The peak in NO2-N and the subsequent decrease tolow concentrations indicated the occurrence of an efficient nitri-fication process in the biofloc culture, which has been reportedpreviously by several authors (Hari et al. 2006; Avnimelech2009; Emerenciano et al. 2012).

As the amount of vegetable meal in the diet increased, ahigher biofloc volume was observed; this effect could have beeninduced by the lower digestibility of vegetable meal, which cre-ates a higher amount of flocculated material (mainly fiber) thatcan be used as substrate by the bacteria and other microorgan-isms. Similarly, Soltan et al. (2008) found that when fish mealsubstitution by a mix of vegetable meals in a tilapia diet ex-ceeded 45%, the apparent digestibility decreased from 81% to73%. In the present study, the increase in biofloc volume fromthe beginning to the end of the trial was associated with the fishmeal content in the diet, since T0 showed the highest increase(from an initial volume of 3 mL/L to a final volume of 19 mL/L),while T30 recorded the lowest increase (from an initial volumeof 1.3 mL/L to a final volume of 5.6 mL/L).

The biofloc biomass recorded in this study accords withvalues reported by Scopel et al. (2011), who found that theTSS level in a Pacific white shrimp culture was greater when

TABLE 4. Mean ( ± SD) survival, final individual weight, initial and final biomass, and feed conversion ratio (FCR) of cultured red tilapia that received dietswith different levels of fish meal inclusion (C = control with 30% fish meal; see Table 1 for definition of experimental diets T0–T30). Within a row, means withdifferent letters are significantly different (P < 0.05).


Survival (%) 100 ± 0 100 ± 0 91 ± 16 83 ± 10 90 ± 16Final individual weight (g) 45.0 ± 1.5 z 53.5 ± 5.6 zy 64.1 ± 1.9 y 57.4 ± 18.7 zy 53.1 ± 5.1 zyInitial biomass (kg/m3) 2.0 2.0 2.0 2.0 2.0Final biomass (kg/m3) 6.0 ± 0.7 7.1 ± 0.1 6.8 ± 1.4 5.8 ± 1.0 6.3 ± 0.5FCR 1.5 ± 0.3 1.2 ± 0.0 1.4 ± 0.5 1.7 ± 0.5 1.5 ± 0.2

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


vegetable meal diets were administered than when fish meal di-ets were used. Scopel et al. (2011) suggested that the result wasdue to the lower digestibility of the vegetable ingredients.

In a study of Nile Tilapia cultures, Azim and Little (2008)compared experimental diets containing 35% and 24% proteinand observed TSS values of 597 and 560 mg/L, respectively,similar to TSS levels found in the present study. Avnimelech(2007) reported mean TSS values of 460–643 mg/L, close tovalues we found for T0. In contrast, Ray et al. (2010) observedno difference in TSS production when Pacific white shrimp weregiven fish meal diets versus vegetable meal diets.

Increases in TSS commonly affect the respiration of cul-tured organisms (gill obstruction) and dissolved oxygen diffu-sion into the water column (Crab et al. 2012; Moreira de Souzaet al. 2013). Nevertheless, the TSS values reported here (400–500 mg/L) are within the recommended range (200–500 mg/L)for a good productive response in biofloc cultures (Avnimelech2009).

The mean ash values of the biofloc in all treatments (33–40%) were similar to previous findings (39.2–41.1%: Ballesteret al. 2010; 39.8%: Becerra-Dorame et al. 2012). Maica et al.(2011) reported that the ash content in biofloc tended to increasewith increasing salinity: mean ash content was 22.12, 26.73,and 42.19% at salinities of 2, 4, and 25‰, respectively. In thepresent study, salinity was maintained at a constant level of35‰, similar to the experiments performed by Ballester et al.(2010) and Becerra-Dorame et al. (2012).

The organic matter content of biofloc ranged between 59%and 66% and therefore was very close to the values of 57–79%reported by Maica et al. (2011).

With regard to the proximate composition of the biofloc, thetotal protein content we found was similar to values reported byAzim and Little (2008; 28–31%), Ekasari et al. (2010; 30.4%),Becerra-Dorame et al. (2012; 11.5–17.5%), and Emerencianoet al. (2012; 37.9–38.4%). The results we recorded for carbo-hydrate content were close to previously published values of35.4% (Becerra-Dorame et al. 2012) and 29.1–34.9% (Emeren-ciano et al. 2012). At the end of the culture period (i.e., day 42),we observed that the protein content was decreasing, whereasthe carbohydrate content was still increasing; this result couldbe due to fiber accumulation in the culture tanks. Azim andLittle (2008), Ekasari et al. (2010), and Becerra-Dorame et al.(2012) reported that the mean lipid content in the biofloc was3.16–3.2%, 6–9%, and 6.5%, respectively, similar to the lipidcontent values observed during the present trial.

The production response of red tilapia was independent ofdietary treatment. Survival rate, final biomass, and FCR werenot affected by substituting fish meal with vegetable meal inthe diet. These results concur with those reported by Scopelet al. (2011) after substituting fish meal with vegetable meal inexperimental diets for Pacific white shrimp cultured in bioflocsystems; those authors found no significant difference in meanfinal biomass, FCR, or survival among the various diet treatmentgroups. Similar results were reported by Azim and Little (2008),

who found no significant difference in mean survival, weightgain, or FCR for Nile Tilapia that were fed diets with twoprotein contents during culture in biofloc systems.

In our study, the TSS, ash content, and volume of the bioflocwere greater in diet treatments with a lower fish meal inclusion,but we observed no negative effect on the final biomass or FCRof red tilapia. In BFT systems, the TSS concentration is affectedby fish consumption; tilapia are able to filter particulate matter,and in the absence of formulated feed they could use the bioflocas a unique nutritional source. Avnimelech (2007) reported adecrease in biofloc volume (from 40 mL/L to 20 mL/L) duringthe first 6 d of Nile Tilapia culture when the fish were deprivedof exogenous feed.

In biofloc cultures, the microbial communities improve thewater quality and recycle the organic matter (Buford et al. 2004);bacteria are consumed by other microorganisms and generatehigh concentrations of live biomass. More specific studies haverevealed that biofloccules used as feed in aquaculture have con-siderable amounts of essential amino acids and essential fattyacids, and the vitamin profiles are usually adequate for fish andcrustaceans (Martınez-Cordova et al. 2014). In BFT systems, thecultured organisms have a high-quality food source that is per-manently available (Browdy et al. 2001; Samocha et al. 2007)and complements their nutrition. Recent evidence indicates thatmicrobial flocs and biofilms act as probiotics (Becerra-Dorameet al. 2012; Martınez-Cordova et al. 2014); therefore, the nu-tritional condition of farmed organisms is a consequence notonly of the feed consumed but also of the presence of bacteriaand their exogenous enzymes, which contribute to the diges-tive physiology of the cultured species (Ziaei-Nejad et al. 2006;Zhou et al. 2009). In the present study, the bacterial communityof the biofloc may have improved the assimilation of vegetablemeal that was included in the experimental diets.

Based on the proximate composition of biofloc developedusing a pelleted diet with a low fish meal content, we concludethat the biofloc is a good, nutritious source of protein, lipid,and carbohydrate for cultured tilapias. These findings suggestthat tilapia producers who use biofloc systems can employ dietswith a low fish meal content to reduce production costs as wellas promote an environmentally friendly method of aquaculture.Future research on BFT should include evaluations of lowerfish meal inclusion in the diets, the amino acid and fatty acidcomposition of the biofloc, and the effects of biofloc-associatedmicroorganisms on the digestive activity of tilapias.

ACKNOWLEDGMENTSThis study was partially financed by the Sonora State

University project entitled “Evaluation of vegetable ingredi-ents to grow tilapia in seawater with biofloc” (Project C-PII/11/26ESU0057P/01). Angelica Moreno-Arias was supportedby a scholarship from the Mexican National Council of Sci-ence and Technology as part of a master’s degree. We thank

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


the University of Sonora for the support provided during thisresearch.

REFERENCESAudelo-Naranjo, J. M., L. R. Martınez-Cordova, S. Gomez-Jimenez, and D.

Voltolina. 2012. Intensive culture of Litopenaeus vannamei without waterexchange and with an artificial substrate. Hidrobiologica 22:1–7.

Avnimelech, Y. 2007. Feeding with microbial flocs by tilapia in minimal dis-charge biofloc technology ponds. Aquaculture 264:140–147.

Avnimelech, Y. 2009. Biofloc technology. World Aquaculture Society, BatonRouge, Louisiana.

Azim, M. E., and D. C. Little. 2008. The biofloc technology (BFT) in indoortanks: water quality, biofloc composition, and growth and welfare of NileTilapia (Oreochromis niloticus). Aquaculture 283:29–35.

Azim, M. E., D. C. Little, and J. E. Bron. 2008. Microbial protein production inactivated suspension tanks manipulating C:N ratio in feed and the implicationsfor fish culture. Bioresource Technology 2008:3590–3599.

Ballester, E. L. C., P. C. Abreu, R. O. Cavalli, M. Emerenciano, L. Abreu, andW. Wasielesky. 2010. Effect of practical diets with different protein levelson the performance of Farfantepenaeus paulensis juveniles nursed in a zeroexchange suspended microbial floc intensive system. Aquaculture Nutrition16:163–172.

Becerra-Dorame, M. J., M. Martınez-Porchas, L. R. Martınez-Cordova, M. E.Rivas-Vega, J. A. Lopez-Elıas, and M. A. Porchas-Cornejo. 2012. Produc-tion response and digestive enzymatic activity of the Pacific white shrimpLitopenaeus vannamei (Boone, 1931) intensively pregrown in microbial het-erotrophic and autotrophic-based systems. Scientific World Journal 2012:1–6.

Browdy, C. L., D. Bratvold, A. D. Stokes, and R. P. McIntosh. 2001. Perspectiveson the application of closed shrimp culture systems. Pages 20–34 in C. L.Browdy and D. E. Jory, editors. The new wave: proceedings of the specialsession on sustainable shrimp culture, aquaculture 2001. World AquacultureSociety, Baton Rouge, Louisiana.

Buford, M. A., P. J. Thompson, R. P. McIntosh, R. H. Bauman, and D. C. Pear-son. 2004. The contribution of flocculated material to shrimp (Litopenaeusvannamei) nutrition in a high-intensity, zero-exchange system. Aquaculture232:525–538.

Crab, R., T. Defoirdt, P. Bossier, and W. Verstraete. 2012. Biofloc technologyin aquaculture: beneficial effects and future challenges. Aquaculture 356–357:351–356.

De Schryver, P., R. Crab, T. Defoirdt, N. Boon, and W. Verstraete. 2008. Thebasics of bio-floc: the added value of aquaculture. Aquaculture 277:125–137.

Ekasari, J., R. Crab, and W. Verstraete. 2010. Primary nutritional content of bio-flocs cultured with different organic carbon sources and salinity. HAYATIJournal of Biosciences 17:125–130.

El-Sayed, A. M. 2006. Tilapia culture. CAB International, Wallingford, UK.Emerenciano, M., E. L. C. Ballester, R. O. Cavalli, and W. Wasielesky. 2011.

Effect of biofloc technology (BFT) on the early postlarval stage of pinkshrimp Farfantepenaeus paulensis: growth performance, floc compositionand salinity stress tolerance. Aquaculture International 19:891–901.

Emerenciano, M., E. L. C. Ballester, R. O. Cavalli, and W. Wasielesky. 2012.Biofloc technology application as a food source in a limited water ex-change nursery system for pink shrimp Farfantepenaeus brasiliensis (La-treille, 1817). Aquaculture Research 43:447–457.

Emerenciano, M., G. Gaxiola, and G. Cuzon. 2013. Biofloc technology (BFT):a review for aquaculture application and animal food industry. Pages 301–328 in D. Matovik, editor. Biomass now: cultivation and utilization. InTech,Rijeka, Croatia.

Hach. 2005. Procedures manual, DR5000 spectrophotometer. Hach, Loveland,Colorado.

Hari, B., B. Madhusoodana Kurup, J. T. Varghese, J. W. Schrama, M. C. J.Verdegem. 2006. The effect of carbohydrate addition on water quality and thenitrogen budget in extensive shrimp culture systems. Aquaculture 252:248–263.

Kuhn, D. D., and G. D. Boardman. 2008. Use of microbial flocs generated fromtilapia effluent as nutritional supplement for shrimp, Litopenaeus vannamei, inrecirculating aquaculture systems. Journal of the World Aquaculture Society39:72–82.

Lopez-Elıas, J. A., M. del, C. Baez-Duenas, and N. Huerta-Aldaz. 1995. Manualde tecnicas analıticas aplicadas al cultivo de microalgas. [Manual of analyticaltechniques applied to microalgae culture.] Departamento de InvestigacionesCientıficas y Tecnologicas, Universidad de Sonora, Mexico.

Lowry, O. H., N. J. Rosenbrough, A. L. Farr, and R. J. Randall. 1951. Proteinmeasurement with the Folin phenol reagent. Journal of Biological Chemistry193:265–275.

Maica, P. F., M. R. Borba, and W. Wasielesky. 2011. Effect of low salinityon microbial floc composition and performance of Litopenaeus vannamei(Boone) juveniles reared in a zero-water-exchange super-intensive system.Aquaculture Research 43:361–370.

Martınez-Cordova, L. R., M. Emerenciano, A. Miranda-Baeza, and M.Martınez-Porchas. 2014. Microbial-based systems for aquaculture of fishand shrimp: an updated review. Reviews in Aquaculture 6:1–18.

Megahed, M. E. 2010. The effect of microbial biofloc on water quality, survivaland growth of the green tiger shrimp (Penaeus semisulcatus) fed with differentcrude protein levels. Journal of the Arabian Aquaculture Society 5:119–142.

Miranda-Baeza, A., J. Lizarraga-Armenta, M. E. Rivas-Vega, J. A. Lopez-Elıas, and M. Nieves-Soto. 2010. Integrated culture of tilapia (Oreochromismossambicus × Oreochromis niloticus) and Pacific oyster (Crassostrea gi-gas) in a recirculation system. Journal of the World Aquaculture Society41:763–771.

Moreira de Souza, D., L. Godoy, W. Wasielesky, and E. L. C. Ballester. 2013.Contribucion de los microorganismos en ambientes acuaticos y en sistemasde cultivo. [Contribution of the microorganisms in aquatic environmentsand in culture systems.] Pages 1–24 in L. R. Martınez-Cordova and M.Martınez-Porchas, editors. Alimento natural en acuacultura. [Natural food inaquaculture.] AGT Editor SA, Mexico, City.

Pande, S. V., R. P. Khan, and T. A. Venkitasubramanian. 1963. Microdetermina-tion of lipids and serum total fatty acid. Analytical Biochemistry 6:415–423.

Ray, A. J., B. L. Lewis, C. L. Browdy, and J. W. Leffler. 2010. Suspendedsolids removal to improve shrimp (Litopenaeus vannamei) production and anevaluation of a plant-based feed in minimal-exchange, super-intensive culturesystems. Aquaculture 299:89–98.

Samocha, T. M., S. Patnaik, M. Speed, A. M. Ali, J. M. Burger, R. V. Almeida,Z. Ayub, M. Harisanto, A. Horowitz, and D. L. Brock. 2007. Use of molassesas carbon source in limited discharge nursery and grow-out systems forLitopenaeus vannamei. Aquaculture Engineering 36:184–191.

Scopel, B. R., R. Schveitzer, W. Q. Seiffert, V. Pierri, R. F. Arantes, F. N. Vieira,and L. A. Vinatea. 2011. Substitution of fish meal in diets for marine shrimpgrown in a biofloc system. Pesquisa Agropecuaria Brasileira 46:928–934.

Sokal, R. R., and F. J. Rohlf. 1981. Biometry, 2nd edition. Freeman, New York.Soltan, M. A., M. A. Hanafy, and M. I. A. Wafa. 2008. Effect of replacing

fish meal by a mixture of different plant protein sources in Nile Tilapia(Oreochromis niloticus L.) diets. Global Veterinaria 2:157–164.

Tacon, A. G. J., J. Cody, L. Conquest, S. Divakaran, I. P. Forster, and O. Decamp.2002. Effect of culture system on the nutrition and growth performanceof Pacific white shrimp Litopenaeus vannamei (Boone) fed different diets.Aquaculture Nutrition 8:121–137.

Timmons, M. B., J. M. Ebeling, F. W. Wheaton, S. T. Summerfelt, and B.J. Vinci. 2002. Recirculating aquaculture systems. Cayuga Aqua Ventures,Ithaca, New York.

Zar, J. 1996. Biostatistical analysis. Prentice Hall, Englewood Cliffs, New Jer-sey.

Zhou, X. X., Y. B. Wang, and W. F. Li. 2009. Effect of probiotic on larvaeshrimp (Penaeus vannamei) based on water quality, survival rate and digestiveenzyme activities. Aquaculture 287:349–353.

Ziaei-Nejad, S., M. H. Rezaei, G. A. Takami, D. L. Lovett, A. R. Mirvaghefi,and M. Shakouri. 2006. The effect of Bacillus spp. bacteria used as probioticson digestive enzyme activity, survival and growth in the Indian white shrimpFenneropenaeus indicus. Aquaculture 252:516–524.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


Click for updates


Guide for AuthorsPublished online: 22 Jan 2015.

To cite this article: (2015) Guide for Authors, North American Journal of Aquaculture, 77:1, 110-115







http://dx.doi.org/10.1080/15222055.2015.1002342



North American Journal of Aquaculture 77:110–115, 2015c© American Fisheries Society 2015

ISSN: 1522-2055 print / 1548-8454 onlineDOI: 10.1080/15222055.2015.1002342

NORTH AMERICAN JOURNAL OF AQUACULTURE

Guide for Authors

Editorial Policy

We encourage the submission of original papers on all aspectsof aquaculture, including broodstock selection and spawning,nutrition and feeding, health and water quality, facilities andproduction technology, and the management of ponds, pens,and raceways. We will consider papers dealing with ways to im-prove the husbandry of any aquatic species—marine or freshwa-ter, vertebrate or invertebrate—raised for commercial, scientific,recreational, enhancement, or restoration purposes that may beof interest to practitioners in North America.

Papers concerning fisheries science per se may be submittedto the American Fisheries Society’s (AFS) sister publicationTransactions of the American Fisheries Society; those dealingwith management may be submitted to the North AmericanJournal of Management; those dealing with the health of fishand other aquatic organisms may be submitted to the Journalof Aquatic Animal Health; and those with a focus on marineand estuarine species and habitats may be submitted to Marineand Coastal Fisheries: Dynamics, Management, and EcosystemScience.

Authors are cautioned not to republish original data withoutfull attribution and explicit permission; see “Dual Publication ofScientific Information” in Transactions of the American Fish-eries Society 110:573–574, 1981.

Manuscript Submission and Review

Manuscript Categories

Manuscripts may be submitted in any of the following cat-egories: (1) Articles are full reports of substantial, controlledresearch and critical reviews of such research; they will bejudged on their scientific merit and relevance to practical aqua-culture. (2) Communications are shorter papers reporting non-replicated experiments, the exploratory culture of new speciesor life stages, the testing of new techniques, and so forth; theywill be judged primarily on the biological insights that they pro-vide as well as their practical contribution to aquaculture. (3)Technical Notes are short papers that deal with operational im-provements, describe new techniques or equipment, or presentunusual observations; they will judged on their practical con-tribution and general interest. (4) Comments are critiques ofpapers published by this journal (responses to which will be in-

vited from the original authors), brief presentations of additionalobservations or data related to previously published papers, orshort discussions of technical issues of interest to the aquacul-tural community.

Submission Procedures

Manuscripts and associated correspondence should be sub-mitted at the journal’s online submission and tracking site,http://mc.manuscriptcentral.com/naja (this site may also beaccessed through the Publications section at the AmericanFisheries Society’s Web site, www.fisheries.org). Detailed in-structions, including acceptable file formats, are available at thesite.

Although the submission site permits authors to include acover letter, such letters are generally not necessary; they shouldbe included only when they contain information that cannoteasily be incorporated into the standard submission form.

The site also permits authors to recommend certain reviewersand/or to request that certain reviewers not be used. Recommen-dations are encouraged but are not required.

Review Process

Submitted papers will be critically reviewed by at least twoexperts in the relevant discipline(s) and evaluated by one of thejournal’s editors. A manuscript may be returned to its authorwithout review if it is judged to be of poor quality or inappro-priate for this journal.

All submissions are electronically screened for the inappro-priate use of material from previously published sources. Insubmitting a paper, you are stipulating that, except where ex-plicitly indicated otherwise, all of the statements, data, and otherelements reflect your own work and not that of others. All allu-sions to the work of others should be properly cited; exact quota-tions from other sources should be in quotation marks. Authorsare also cautioned not to repeat long passages from their ownpublications. Failure to follow these requirements may resultin rejection of the paper and, in extreme cases, restrictions onpublishing in this journal.

Authors have the option of anonymity; if they wish to exerciseit, they should prepare their manuscripts accordingly.

Review of manuscripts relies on volunteers and can be a fairlylengthy process. However, we strive to get decisions to authorsin 9–12 weeks. If revisions are requested, authors should makethem promptly, normally within 30 days of receiving the editor’sdecision (short extensions will be allowed if there are justifiabledelays). If a revision is not received within the allowed time,

110

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5

GUIDE FOR AUTHORS 111

the paper will be considered withdrawn; late revisions will betreated as new submissions and may have to go through thereview process again.

Publication Charges

Publication charges are US$100 per printed page and willbe billed when the paper is in proof. Full and partial subsidiesare available to voting members of the American Fisheries So-ciety who certify that grant or agency funds are not available.Manuscript reviews are not affected by requests for subsidies;however, at least one author must be (or become) an AFS mem-ber by the time that a paper is published. Every paper publishedin the journal is subject to a $30 fee to offset handling costs.Authors will receive a free PDF of the published article andmay purchase reprints of their papers from the printer whenthey receive their proofs.

Manuscript Preparation

Components

A typical manuscript will have the following components:Title page.—The title page should give the title of the pa-

per and the name(s) and complete mailing address(es) of theauthor(s). In addition to accurately reflecting the content of thepaper, the title should be short (preferably no more than 12words) and to the point. See a recent issue of the journal for theformat to use for authors’ names and addresses. A suggestedrunning head (shortened version of the title) may also be in-cluded on the title page. Keywords are not used in this journal,however, and so should not be included.

Abstract.—Articles, communications, and technical notes re-quire abstracts; comments do not. The abstract should consistof one paragraph (up to 300 words for an article and up to 200words for a communication or note) that concisely states whyand (generally) how the study was done as well as what theresults were and what they mean. It should not simply outlinethe contents (e.g., avoid statements to the effect that such-and-such is presented) or present the methods in detail. Citations andfootnotes are not allowed in abstracts, and abbreviations shouldbe used sparingly. Detailed statistical results (e.g., P-values)should be reserved for the main text. Because abstracts tend tobe more widely read than complete papers, authors should takecare to make them comprehensive, clear, and interesting.

Introduction.—The introduction should provide a context forthe work to be reported, particularly its purpose and importance.In doing so, it should present at least a summary review ofprevious literature on the subject.

Methods.—Descriptions of the methods employed in thestudy should be detailed enough to enable readers to repeatit. Previously published descriptions may be cited in lieu ofpresenting complete new ones provided that the sources arereadily available (in general, avoid citations to theses, disserta-

tions, agency reports, and similar sources in this instance). Ifmore than one method was used or a particular method entails aseries of major steps, present each method or step in a separatesubsection. Appropriate tables and figures can reduce the needfor detailed verbal descriptions of methods. Papers focusing en-tirely on techniques or models do not require a separate sectionon methods.

Results.—As a rule, it is preferable to present detailed re-sults in tables and/or figures and to devote the text to summarystatements and analyses. Display data in tables if numerical pre-cision is important, in figures if trends are paramount. Althoughthe presentation of a large amount of raw data is generally notmeaningful, data should not be refined to the point that thereader cannot verify the analyses or use the information forother purposes. In presenting the results of statistical tests, re-port the type of test, the test statistic, the degrees of freedom,and the significance level (P-value). Although the value 0.05 iscommonly used as the threshold in hypothesis testing, we haveno specific requirements in this area; in the interest of providinguseful information, authors should report all P-values. It is veryimportant that statistical designs and models be appropriate forthe studies in which they are used; we encourage authors tohave a statistician review their work before submitting a paperfor publication. Lastly, statistical results should be presented inbiologically meaningful terms rather than in purely statisticaljargon.

Discussion.—The merits of a paper can be greatly enhancedby a good discussion. In it authors should indicate the signif-icance of their research, how it relates to current knowledge,and any avenues that it suggests for further research. Informedspeculation is acceptable as long as it is clearly identified assuch. Authors should avoid merely restating their results and/or(re)summarizing the literature.

Acknowledgments.—In this section authors may acknowl-edge the sources of their funding and thank those who con-tributed directly to the project or the preparation of themanuscript. Dedications and acknowledgment of emotional sup-port from family and friends are not appropriate. If all authorsare employees of the U.S. Government, this section should statethat the mention of specific products does not constitute en-dorsement by their agency.

References.—References should be selected with a viewto relevance and availability, with preference given to peer-reviewed publications that are widely available. Internal reports,papers presented at conferences, articles in preparation, and soforth should be treated as unpublished and cited like personalcommunications (i.e., parenthetically in the text alone). Authorsshould obtain written permission to cite such material. Commonreference formats are given below; a more complete list is givenin chapter 8 of the AFS style guide, which is available at theAFS Web site as well as the manuscript submission site.

Footnotes.—Footnotes should be kept to a minimum. Typ-ically, they are used to report changes of address for au-thors, identify additional sources of data, or explain technical

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5

112 GUIDE FOR AUTHORS

nomenclature (e.g., ages of anadromous fish and structures offatty acids).

Tables.—In general, tables should be designed to presentrelated information as simply and directly as possible. A goodrule of thumb is to establish the point(s) that the table is intendedto make, then to select the information required to do that anddetermine the most logical order in which to present it. Detailedguidelines for the preparation of tables may be found in chapter12 of the AFS style guide, but a few of the more important onesmay be mentioned here:

1. We prefer to print tables in “portrait” orientation but willallow ones in “landscape” orientation as long as they takeup no more than one page.

2. Tables that are too long or too wide to fit on one page can becarried over to a facing page, but authors should try to avoidcreating tables that span more than two pages. In general,very large tables should appear as supplements in the onlineversion of the article only.

3. Tables should contain only three horizontal rules (lines)—one before the column headings, one after those headings,and one at the bottom of the table—and no vertical rules.

4. As a rule, captions should be detailed enough that thetable can be understood apart from the text (if there ismore than one table with the same general structure, onlythe first needs to have a detailed caption). Captions shouldbe written so as to stress the purpose of the table and notmerely list its contents in a mechanical way.

5. There should be only one set of column headings. If theinformation to be presented seems to require more thanthat, the table should be redesigned (e.g., by switching therows and columns) or split into two or more tables.

6. Bold, centered headings may be used within the body of thetable to distinguish different types of data as long as theydo not conflict with the column headings.

7. Only the first letter of a row or column heading should becapitalized (along with words or symbols that would becapitalized in ordinary text).

8. The data within the body of the table should not be crowded;if need be, blank rows can be inserted to separate data intological groups or provide guides for the eye.

9. Significant differences should be indicated by lowercaseletters, beginning with the letter “z” (“z” may mark eitherthe highest or the lowest value[s], but subsequent lettershave to follow suit); in most cases, there should be noomissions in the sequence of the letters. The letters shouldbe set on the same lines as the values to which they pertain(not as superscripts) and be separated from those values bysingle spaces.

10. Values less than 1.00 should be preceded by zeroes (e.g.,0.78).

11. Values need not be reported to all significant digits if a lessernumber of digits conveys the information in a meaningfulway.

12. Footnotes should be indicated by superscripted lowercaseletters, beginning with the letter “a”; the letters may appearin the row and column headings as well as the body of thetable but not in the caption. The footnotes per se should belisted on separate lines at the bottom of the table.

Figure captions.—Figure captions should appear togetherin a list rather than separately with each figure (however, thenumber of the figure and the name of the corresponding au-thor should be given outside the image area of each figure forpurposes of identification). Like table captions, figure captionsshould generally be detailed enough that the figure can be under-stood apart from the text. To the extent possible, however, paneldescriptions, (full) variable names, units of measure, legends,and so forth should be included in the figure itself rather thanin the caption; in no case should they be given in both places.Different panels may be designated “A,” “B,” and so forth, but itis preferable to give them substantive labels (e.g., “Treatment”and “Control”).

Figures.—Figures include visual materials such as graphs,maps, diagrams, and photographs. Figures have proved to beone of the most troublesome aspects of the publishing process.As the Journals Department has only limited ability to modifyfigures, they frequently have to be sent back to the authors forcorrection.

At the most fundamental level, figure design should followcertain commonsense principles: figures should be as simpleand straightforward as possible; have a high enough resolutionto be easily readable (300 dpi or more); and be consistent inthe use of lettering, line widths, and other graphic elements. Inaddition, they need to conform to AFS style. It is particularlyimportant to remember that most figures will be reduced by upto 50% when printed and thus need to be designed with this inmind. We recommend that authors use a copier to reduce eachfigure to the width of one or two printed columns (3.50 and7.25 inches, respectively), depending on the dimensions of theparticular figure, and verify that all elements are still legible.The following are particularly problematical: bold type (whichtends to blur), italic type (which tends to become less visi-ble), dashed lines (which tend to appear continuous) and dottedlines (which tend to disappear entirely). Additional guidelinesfor the preparation of figures may be found in the AFS styleguide.

In the print version of the journal, all figures will be repro-duced in black and white unless specific arrangements have beenmade with the Journals Department to cover the extra costs ofcolor printing. (In the online version, color figures will be repro-duced in color at no additional charge.) Because color printingis expensive, authors are advised not to use color to distinguishphenomena when other means (different shading, symbols, andso forth) are adequate. If you use color in a figure, avoid us-ing similar colors or shades that may be difficult for readers todistinguish. Also, in deference to readers with color blindness,avoid using red and green in the same figure.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


Digital files in EPS, TIFF, and PSD formats are preferred;figures should be submitted as separate files rather than beingimbedded in text files.

Mathematical and statistical expressions.—Chapter 4 of theAFS style guide covers the treatment of these expressions indetail, but a few general points may be mentioned here:

1. Symbols representing variables and parameters should beitalicized if they consist of single letters in the Latin alphabet(e.g., K and F). All other symbols except Greek letters maybe italicized or not, provided that the treatment is consis-tent (e.g., CPUE or CPUE); Greek letters should never beitalicized.

2. Natural logarithms may be expressed as loge or ln; logarithmswith other bases should identify the base (e.g., log10).

3. Long equations should be “broken” at logical points, nor-mally after an operator such as a plus or minus sign.

4. Definitions of variables and parameters may be run into thetext if only a few such terms are involved. If there are anumber of them or they are used in more than one equation,a list is preferable (see section 4.8 of the style guide).

5. Avoid the expressions “the mean length was 45.2 ± 3.84mm” and “the mean (±SD) length was 45.2 ± 3.84 mm”because they are at best awkward and at worst inaccurate.Use the expressions “the mean ± SD length was 45.2 ± 3.84mm” or “the mean length was 45.2 mm (SD, 3.84)” instead.

Appendices and supplements.—In addition to the standardelements of a paper, authors may submit certain supplemen-tary material, such as additional data or results, the derivationsof equations, computer code, and so forth. For publication pur-poses, such material will be treated either as an appendix (whichwill appear with the article in both the print and online versions)or as a supplement (which will appear only in the online ver-sion). Of course, all material that is essential to understandingan article should be included in the article itself. Closely relatedmaterial that will be of interest to a large number of readersmay be placed in an appendix. Other material may be madeavailable through a supplement if the editors deem it importantenough for readers to have ready access to. In terms of format,appendices should be regarded as extensions of articles and thusfollow AFS style strictly. Supplements, by contrast, may be inany format that is suitable for their contents; however, (1) thereshould be consistency between the symbols, abbreviations, andso forth used in the supplement and those used in the articleand (2) either the title of the supplement or the first paragraphshould make clear how it relates to the article.

Style and Format

Published articles represent the culmination of research ef-forts, often lengthy and highly sophisticated ones. To do thoseefforts justice, however, the articles must be well written; poorlywritten articles not only place an unnecessary burden on read-ers, they also cast doubt on the quality of the research itself.

Although some people naturally write better than others, mostcan develop the ability to write well through practice and atten-tion to detail. The introduction to the AFS style guide should bea particularly valuable resource in this regard; in a few pages, itidentifies the errors in composition mostly commonly encoun-tered in the papers submitted to AFS journals and shows how tocorrect them. We also encourage authors to have other fisheriesprofessionals critique their initial drafts with respect to presen-tation as well as substance. Authors whose native language isnot English should make a point of having English speakersreview their manuscripts before submission.

In writing for AFS journals, authors are also expectedto follow certain style conventions pertaining to capitaliza-tion, spelling, punctuation, mathematical expressions, technicalterms, and so forth. For instance, we require that the letter P(indicating the degree of statistical significance) be capitalizedas well as italicized, whereas some journals require that it belowercased. Although some of the more important style con-ventions are noted below, all of them are discussed in detail inthe AFS style guide. Authors would be well advised to becomefamiliar with the main elements of AFS style and to consult theguide frequently in preparing their manuscripts.

Resources for authors.—As suggested above, the principalresource on matters of style is the AFS style guide. Authorsmay also find it helpful to consult the Chicago Manual of Style(University of Chicago Press, Chicago) and Scientific Style andFormat (Council of Science Editors, Chicago), though the AFSstyle guide always takes precedence.

The standard resource for word usage and spelling is Web-ster’s Third New International Dictionary, as updated by thelatest edition of Merriam-Webster’s Collegiate Dictionary. Ap-pendix A of the AFS style guide shows the proper way to spellmany of the terms used in fisheries writing (some of which arenot in the dictionary), including terms for which our preferredspelling differs from that in the dictionary.

The standard resource for the common and scientific namesof North American fish species is the current edition of Commonand Scientific Names of Fishes from the United States, Canada,and Mexico (American Fisheries Society, Bethesda, Maryland).For other aquatic species, authors should follow the companionpublications World Fishes Important to North Americans andCommon and Scientific Names of Aquatic Invertebrates fromthe United States and Canada (the volumes Mollusks, Deca-pod Crustaceans, and Cnidaria and Ctenophora are currentlyavailable in the latter series).

In most cases, scientific names should be included only atfirst mention in the abstract and text; full common names (e.g.,“Coho Salmon” rather than simply “Coho”) should be usedelsewhere. The format for the first mention is

Coho Salmon Oncorhynchus kisutch,

in which all parts of the common name are capitalized and thescientific name follows the common name but is not given in

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5

114 GUIDE FOR AUTHORS

parentheses. See chapter 9 of the AFS style guide for additionalinformation on the treatment of species’ names; the acceptedplurals of fish names are given in Appendix C of the guide.

In papers about population dynamics, we prefer the notationused by W. E. Ricker in Computation and Interpretation of Bio-logical Statistics of Fish Populations (Fisheries Research Boardof Canada Bulletin 191, 1975). However, all symbols should bedefined anew in every paper. Our standard sources for chemicaland enzyme names are the current editions of the Merck Index(Merck & Co., Rahway, New Jersey) and Enzyme Nomencla-ture (Academic Press, San Diego, California), respectively. Thepreferred treatment of allozymes is noted in the article “GeneNomenclature for Protein-Coding Loci in Fish” by J. B. Shakleeet al. (Transactions of the American Fisheries Society 119:2–15, 1990). Additional information on the treatment of these andother technical matters may be found in chapter 11 of the AFSstyle guide.

Manuscript format.—As an aid to reviewers and editors, au-thors should

1. use a line spacing of at least space and a half for all com-ponents of the paper, including the title page, footnotes, andtables;

2. number all pages sequentially and provide continuous linenumbering beginning with the title page;

3. use a 12-point font throughout;4. use three levels of headings, as follows: for the major sec-

tions of the paper (Methods, Results, Discussion, Acknowl-edgments, and References), type them flush left with initialletters capitalized (except for prepositions and conjunctions)in ordinary type, preceded by “<A>” (e.g., <A>Methods);for subsections in Results and Discussion, type them flushleft with initial letters capitalized in ordinary type precededby “<B>” (e.g., <B>Treatment 1); and for subsections inMethods and sub-subsections in Results and Discussion, runthem into the text with only the initial letter of the first wordcapitalized, all words italicized, and followed by a periodand a long dash (e.g., Sampling design.—); and

5. turn off automatic hyphenation and justification.

General style conventions.—A detailed presentation of AFSstyle is beyond the scope of these guidelines. The followingconventions, however, are so general as to apply to virtuallyevery paper:

1. Only symbols and abbreviations included in Webster’s dic-tionaries or listed at the end of these guidelines (as well asat the back of each printed issue of the journal) may be usedwithout definition. All others should be defined at first use(e.g., index of biotic integrity [IBI]). Abbreviations shouldnot be introduced unless they are used at least two moretimes.

2. As a rule, either metric or English units may be used, butnot both. The only exceptions are a few quantities that are

typically expressed only one way (e.g, g [of medication]/lb[of feed]).

3. Single-digit numbers should be spelled out unless they areused with units of measure or in conjunction with largervalues (e.g., 8 Walleyes and 16 Saugers). Numbers with fouror more digits should contain commas; those less than 1.00should be preceded by zeroes.

4. Ratios involving two values or units of measure should beindicated by forward slashes (e.g., 0.30 g/d); ratios involvingthree such terms should be indicated by negative exponents(e.g., 0.01 g·g–1·d–1).

5. Ages of fish should be expressed by Arabic numerals and notcontain plus signs (e.g., a fish is age 1 [not age 1+] from theJanuary 1 after it hatches to the following December 31).

6. Dates should be expressed as month–day–year (e.g., January11, 2011. Note that the term “Julian day” does not mean dayof the year and should not be used in that context.

7. Time should be expressed in terms of the 24-hour clockfollowed by the word “hours” (e.g., 1435 hours rather than2:35 p.m.).

Reference formats.—Text citations should conform to theauthor–year system. Examples of common types are as follows:

(Johnson 1995)(Johnson and Smith 1996)(Johnson et al. 1997, 1998) [three or more authors](Johnson et al. 1999, 2001; Smith 2000)(Johnson 2000a, 2000b)(Johnson, in press)(E. M. Johnson, National Marine Fisheries Service, personal communication)

Note that with one exception citations should be listed inchronological order; the exception is that all citations to the sameauthor(s) should be grouped together (see the fourth exampleabove).

In reference lists, references should be in strict alphabeticalorder by authors’ last names; if there are two or more referenceswith the same authors, those references should then be listedchronologically. All authors must be named in references.

Detailed information on reference formats may be found inchapter 8 of the AFS style guide. The more common types areas follows:

Articles in journals

Pace, M. L., and J. D. Orcutt. 1981. The relative importance of protozoans, ro-tifers, and crustaceans in a freshwater zooplankton community. Limnologyand Oceanography 26:822–830.

Note that (1) except for the first author, authors’ initials comebefore their last names; (2) only the first word of the title of thearticle is capitalized (along with any other words that would becapitalized in ordinary text); and (3) the name of the journal isgiven in full.

Books

Krebs, C. J. 1989. Ecological methodology. Harper and Row, New York.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5


Chapters in books

Omernik, J. M. 1995. Ecoregions: a spatial framework for environmental man-agement. Pages 49–62 in W. S. Davis and T. P. Simon, editors. Biologicalassessment and criteria: tools for water resource planning and decisionmaking. Lewis Publishers, Boca Raton, Florida.

Government reports

Reports that are issued on a regular basis are treated much likearticles in journals (the principal difference being that page num-bers should not be given); other reports are treated like books:

Everest, F. H., C. E. McLemore, and J. F. Ward. 1980. An improved tri-tubecryogenic gravel sampler. U.S. Forest Service Research Note PNW-350.[journal format]

USEPA (U.S. Environmental Protection Agency). 1998. Water quality crite-ria and standards plan: priorities for the future. USEPA, 822-R-98-003,Washington, D.C. [book format]

Electronic publications

Formats for references to electronic publications are stillevolving. The important thing is to give the reader enough in-formation to be able to locate the reference easily.

If a book or report is only available online or is availablein print form but was accessed online, the reference should beformatted as follows:

Baldwin, N. A., R. W. Saalfield, M. R. Dochoda, H. J. Buettner, and R. L.Eshenroder. 2000. Commercial fish production in the Great Lakes, 1867–1996. Great Lakes Fishery Commission, Ann Arbor, Michigan. Available:www.glfc.org/databases/. (September 2000).

The month and year in parentheses indicate when the sitewas last accessed.

If a journal is available in print form, authors should usethe standard reference format even if they accessed the articleonline. If a journal is only available electronically, the formatdepends on the way(s) in which articles are designated. Twopossible formats are as follows:

Gallagher, M. B., and S. S. Heppell. 2010. Essential habitat information for age-0 rockfish along the central Oregon coast. Marine and Coastal Fisheries:Dynamics, Management, and Ecosystem Science [online serial] 2:60–72.DOI: 10.1577/C09-032.1

Kimmerer, W. J. 2004. Open-water processes of the San Francisco Estuary:from physical forcing to biological responses. San Francisco Estuary andWatershed Science [online serial] 2(1):article 1.

Dow

nloa

ded

by [

Dep

artm

ent O

f Fi

sher

ies]

at 0

0:44

20

July

201

5

Documents

On: 20 July 2015, At: 00:36 This article was ... - Fisheries · a Department of Wildlife, Fisheries and Aquaculture, Mississippi State University, Box 9690, Mississippi State, Mississippi