Proceedings Tri-State Dairy Nutrition Conference

April 8 & 9, 2003 Tri-State Dairy Nutrition Conference

1

Minimizing Subclinical Metabolic Diseases

Todd Duffield1

Department of Population MedicineUniversity of Guelph

Abstract

Subclinical ketosis is a common diseasein lactating dairy cows with a lactationalincidence rate above 40% in many herds. Thiscondition is associated with increased clinicaldisease risk, reduced milk production, andimpaired reproductive performance. On a herdbasis, subclinical ketosis is much more costlythan clinical ketosis. Prevention is achievedlargely through effective dry cow programs thatencompass both good nutrition and excellentcow management. However, certain additivesare helpful in reducing subclinical ketosis. Theseinclude propylene glycol, rumen protectedcholine, and where approved for dairy cattle –ionophores. Effective monitoring programs area critical component to managing subclinicalketosis. Monitoring is useful for assessment ofthe transition cow program and for identificationof individual animals to treat.

Introduction

Most periparturient abnormalities havesome metabolic element as a component of thesufficient cause of clinical disease. Themetabolic disturbance of milk fever can bemeasured through low serum calciumconcentrations. Negative energy balance, fatmobilization, and subsequent elevations inketone body concentrations play a contributingrole in the expression of fatty liver syndrome,clinical ketosis, and abomasal displacement. A

negative energy balance may also increase therisk of retained placenta, metritis, and mastitisthrough impaired immune function. A thirdcategory of metabolic disease in early lactationmight include rumen acidosis, which is markedby low rumen pH. Thus, calcium homeostasis,energy balance, and rumen pH are importantconsiderations for disease prevention intransition dairy cows (Goff and Horst, 1997).

In general, subclinical disease incidenceis far more common than clinical disease,frequently going unnoticed and may beassociated with significant clinical disease risks,impaired production, and reduced reproductiveperformance. Of the three major subclinicalmetabolic diseases, the most information existsfor subclinical ketosis. It is associated with bothlosses in milk production and increased risk ofperiparturient disease. Prevention depends onseveral factors, including proper transition cownutrition, management of body condition, andmay be helped through the use of certain feedadditives, such as niacin, propylene glycol,rumen protected choline, and ionophores. It iscommonly accepted that subclinicalhypocalcemia is an important disease, but verylittle information is published on the impact ofthis problem on subsequent risk of disease orproduction loss. Subclinical rumen acidosis isalso thought to be a major problem on manydairy farms, but it is difficult to measure andvery little controlled research exists for thissyndrome. This article will focus on the

1Contact at: Ontario Veterinary College, University of Guelph, Guelph, Ontario, Canada N1G 2W1, (519) 824-4120, FAX(519) 763-8621, Email: [email protected]


2

importance, prevention aspects, and monitoringstrategies for subclinical ketosis.

Is subclinical ketosis a disease?

Elevated levels of circulating ketonebodies occur in early lactation in response to thehomeorhetic drive to sustain high levels of milkproduction, at a time when dry matter intake isreduced (Baird, 1982). The three major ketonebodies are acetone, acetoacetate, and beta-hydroxybutyrate (BHBA). Subclinical ketosisis simply a condition marked by increased levelsof circulating ketone bodies without the presenceof the clinical signs of ketosis. Subclinicalketosis has been associated with increased riskof specific periparturient diseases (ketosis,displaced abomasum, metritis, and mastitis),decreased milk production, and impairedreproductive performance. If these impacts aretrue, then prevention or reduction in incidenceshould ameliorate some or all of the negativeeffects of this condition. Administration of amonensin controlled release capsule three weeksprecalving has decreased the incidence ofsubclinical ketosis, clinical ketosis, anddisplaced abomasum, and improved milkproduction (particularly in cows and herds atincreased risk of subclinical ketosis).

Association with periparturient disease

Cows in early lactation with subclinicalketosis had an increased risk of metritis four dayslater (Dohoo and Martin, 1984). However, moststudies have identified ketosis to be a resultrather than a cause of metritis. Cows havingsubclinical ketosis are at increased risk ofsubsequently developing clinical ketosis (Dohooand Martin, 1984). The relationship betweendisplaced abomasum and ketosis has beenidentified as bi-directional (Curtis et al., 1985;Grohn et al., 1989). That is, ketosis may be acause of displacement, and abomasaldisplacement may lead to ketosis. Correa et al.(1993) found that ketosis increased the risk ofabomasal displacement, but not the reverse.

However, ketosis, as an inciting or predisposingcause of abomasal displacement, can be furthersupported by some recent Guelph research.Elevated BHBA concentrations above 1000mmol/L increased the likelihood of abomasaldisplacement (Geishauser et al., 1997). Cowswith concentrations of BHBA at or above 1400µmol/L in the first two weeks post calving werethree times more likely to subsequently developeither clinical ketosis or abomasal displacement(Duffield, 1997).

Two studies have found a relationshipbetween the diagnosis of ketosis prior toidentifying mastitis (Dohoo and Martin, 1984;Syvajarvi et al., 1986). Mastitis increased therisk of ketosis in Finnish Ayrshires (Grohn etal., 1989). Hyperketonemic cows with BHBAblood levels above 1400 µmol/L were found tosuffer a more severe experimental mastitis thannormal cows (Kremer et al., 1993). There maybe some important immune functionimplications associated with decreased energybalance and subclinical ketosis. Recently, twoseparate studies have identified subclinicalketosis as a risk factor for the subsequentoccurrence of clinical mastitis (Leslie et al.,2000).

Impact on milk production

In general, there is consensus that anegative association between hyperketonemiaand milk production exists. In one study, theloss of production associated with a positive milkketone test was 1.0 to 1.4 kg/day of milk for alactation (Dohoo and Martin, 1984). Test daymilk production was negatively correlated withmilk acetone levels in several Scandinavianprojects (Andersson and Emanuelson, 1985;Gustafsson et al., 1993; Steen et al., 1996).Kauppinen (1984) reported that subclinicallyketotic cows had significantly higher annual milkyields than nonketotic cows. Herdt et al. (1981)found higher levels of BHBA in higherproducing cows, but individual milk tests werenot collected on the same day but preceded blood


3

measurement for BHBA. It is possible thathigher milk yields put cows at increased risk ofdeveloping subclinical ketosis. Increased levelsof milk production may be associated withincreased fat mobilization and a greater risk ofhyperketonemia.

Effect on milk components

Milk fat and milk protein aresignificantly altered in hyperketonemia. Milkfat percentage was increased in subclinicallyketotic cows (Miettenen, 1994; Miettenen andSetala, 1993). The association between milk fatand hyperketonemia is, presumably, because ofincreased availability of BHBA and fatty acidsfor milk fat synthesis. It is unclear whetherincreased levels of circulating ketones causeincreased milk fat, or if cows that are prone tohigher milk fat yields are more susceptible tosubclinical ketosis. Milk protein percentage hasbeen reported to be lower in cows withsubclinical ketosis (Miettenen, 1994; Miettenenand Setala, 1993). This may be the result of areduced energy supply, since milk proteinpercentage is positively associated with netenergy balance. Recently, we have identifiedan impact of subclinical ketosis on milk fatpercentage at first DHI test postcalving only, buta reduction in milk protein percentage for thefirst three DHI tests.

Impact on reproductive performance

Increasing the degree of negative energybalance in early lactation has been shown toincrease the interval from calving to firstovulation (Butler and Smith, 1989). Butler andSmith (1989) suggested that cows with a longerinterval from calving to first ovulationexperience a decrease in pregnancy at firstservice because conception risk is related to thenumber of ovulatory cycles that occur prior toinsemination (Stevenson and Call, 1983;Whitmore et al., 1974). Since hyperketonemiais a symptom of a disturbed energy metabolism,many authors have investigated the relationship

between subclinical ketosis and reproductiveperformance. No effect of either subclinical orclinical ketosis on individual cow fertility wasfound in two studies (Andersson andEmanuelson, 1985; Kaupinnen, 1984).However, significant correlations between theherd prevalence of hyperketonemia and herdmean intervals from both calving to first serviceand calving to last service have been noted(Andersson and Emanuelson, 1985). A linkbetween subclinical ketosis and the increasedincidence of cystic ovaries has also been reported(Andersson and Emanuelson, 1985; Dohoo andMartin, 1984). Miettenen and Setala (1993)found an increased interval from calving toconception in cows with high milk and fat yields.The associations between fertility and increasedfat and milk yields do not necessarily imply arelationship between impaired fertility andhyperketonemia. The duration of either clinicalor subclinical ketosis may be too short to exerta negative effect on calving interval. However,Whitaker et al. (1993) found cows with a betterenergy status at 14 days postpartum had areduced interval from calving to the onset ofcyclicity and fewer services per conception. Noeffect was observed when energy status wasevaluated at 21 days postpartum or at firstservice. This study was only conducted on 24cows within one herd. In a much larger dataset,Cook et al. (2001) reported significantly longercalving to conception intervals and higherculling rates in cows that had high milk acetoneconcentrations in early lactation. Initialscreening of our 1010 cow dataset from 1995indicates a significant reduction in 1st serviceconception rate for cows identified to besubclinically ketotic in the second weekpostpartum.

Cost of subclinical ketosis

When negative impacts of milkproduction losses, increased risk of disease, andreduced reproductive performance areconsidered, the cost of one cow with subclinicalketosis is estimated to be $78 U.S. (Geishauser


4

et al., 2001). This number will vary dependingon several variables, including the valueassigned for milk, impaired reproduction, andmetabolic disease. Regardless, the individualdisease value for subclinical ketosis is less thanclinical disease. However, because thesubclinical form is more prevalent, the cost atthe herd level is much higher. For example, ifan average dairy herd has an incidence rate forclinical ketosis of 5% and the disease costs$145.00, a 100-cow dairy herd would have a costof clinical ketosis of $725 in a year. Whereas,an average 100-cow dairy would have asubclinical ketosis incidence of 41%, with anannual cost of $3198.

How common is subclinical ketosis?

Before a program is instituted, theveterinarian and farm manager need to knowwhat the average incidence of subclinical ketosisis for the herd so that a reasonable and achievabletarget can be set. In a recent trial conducted atGuelph, the median incidence of subclinicalketosis (BHBA > 1400 µmol/L) in untreatedcows was 41% for the first nine weeks oflactation (Duffield et al., 1998). This wasroughly equivalent to two cows identified assubclinically ketotic per 10 cows examined ineach of the first and the second weeks post-calving. The range across 25 herds for the totalnine weeks was 8 to 80%. The four highest herdshad incidence rates above 65% and also had thelargest milk production response to prophylactictreatment.

Prevention

General Guidelines

Since ketosis occurs in early lactation,recommendations for prevention have focusedon the nutritional management of the dry andtransition cow. Detailed recommendations fornutrition during the dry period can be foundelsewhere (Oetzel, 1998). It is a commonrecommendation to divide the dry period into

two feeding groups: far-off and close-up(Radostits et al, 1994). Typically, far-off dietsfollow NRC (2001) guidelines for dry cows. Theclose-up diet is usually balanced according torecommendations that are halfway betweenthose for the dry cow and those for the earlylactation cow and should be fed starting at leastthree weeks before expected calving (Oetzel,1998). The goals of the transition diet(specifically designed to prevent subclinicalketosis) are to maximize dry matter intake(DMI) to provide adequate energy density(Oetzel, 1998). Avoidance of ketogenicfeedstuffs (Tveit et al., 1992) and increasedfrequency of feeding concentrates (Andersson,1988; Gustafsson et al., 1993) have beenadvocated as preventive measures againstsubclinical ketosis. The reduction ofoverconditioning cows in late lactation and theearly dry period, as well as lead feeding withconcentrates about three weeks prior to calving,have also been suggested as aids in prophylaxis(Andersson, 1988; Lean et al, 1991).Maximizing DMI and maintenance of aconsistent intake through the last three weeksprior to calving is likely the hallmark of asuccessful transition cow program. Recent workat Guelph (Tera Osborne, MS candidate,personal communication) indicates that a DMIof less than 12 kg/day per cow in the last threeweeks prior to calving substantially increases thesubsequent risk of subclinical ketosis (OddsRatio 5.7, P < 0.05). Achieving group DMItargets above an average of 12 kg/day per coware possible, and based on the above finding,should be a goal for the close-up group. Moreimportant than ration formulation and rationingredients, close attention should be paid to cowcomfort and environmental issues. These factorsinclude, but are not limited to, adequate penspace or stall space per cow, adequate feed bunkspace, sufficient and comfortable bedding,adequate water supply, and minimization of heatstress.


5

Feed Additives

In addition to good nutrition, certain feedadditives have been found beneficial in reducingsubclinical ketosis, when administeredprophylactically. Niacin fed prior to calving atthe rate of 3 to 6 g/day may be helpful in reducingblood levels of BHBA (Dufva et al., 1983; Fronkand Schultz, 1979). Propylene glycol has beenused successfully for the prevention ofsubclinical ketosis (Emery et al., 1964; Sauer etal., 1973). Treatment of cows for eight weeksstarting at calving with either 3 or 6% propyleneglycol in a concentrate mixture significantlyreduced the incidence of positive milk ketonetests (Fisher et al., 1973). Precalving oraltreatment with 300 g/day of propylene glycol for10 days lowered serum non-esterified fatty acids(NEFA) concentrations and improved somemeasures of reproductive performance in onestudy (Formigoni et al., 1996). A dose ofpropylene glycol of 1 L/day as an oral drenchfor nine days prior to calving decreased BHBA,and NEFA and increased glucose concentrations(Struder et al., 1993). It appears that a bolus ofpropylene glycol is necessary for maximumeffect, since mixing in a total mixed ration isnot as efficacious as either an oral drench orwhen mixed with a small quantity of grain(Christenson et al., 1995). Schultz (1958)reported that sodium propionate could be givento prevent clinical ketosis in dairy cattle.Propylene glycol requires repeated daily oraladministration and sodium propionate mayreduce feed intake (Sauer et al., 1989).Ionophores have been proposed as potentialprophylactic agents for reducing hyperketonemia(Lean et al., 1991; Tyler et al., 1992). In contrastto propylene glycol and sodium propionate,ionophores are relatively inexpensive and mucheasier to administer. Rumen protected cholinereduced liver triglycerides and increased liverglycogen (Piepenbrink and Overton, 2000).More recently, preliminary Guelph researchshows reductions in NEFA and BHBAconcentrations postcalving in cows receiving

rumen protected choline (Reashure™; BalchemEncapsulates, Slate Hill, NY) during transitioncompared to control cows. The cost benefit ofthis prevention tool needs to be investigated.

Ionophores

The gluconeogenic potential ofmonensin has attracted researchers to investigateits possible role as an antiketogenic agent in dairycattle. Rogers and Hope-Cawdery (1980) firstdescribed the beneficial effects of monensin forreducing the incidence of ketosis in a herd witha clinical ketosis problem. The antiketogenicproperties of monensin were later investigatedin a Canadian trial involving two levels ofmonensin and three groups of 12 Holstein cows(Sauer et al, 1989). Monensin included at 30g/ton of total ration (high group), decreased theincidence of subclinical ketosis and significantlyreduced blood BHBA concentrations in the firstthree weeks postpartum (Sauer et al., 1989). Theincidence of subclinical ketosis, defined as totalblood ketones > 9 mg/100 ml (900 µmol/L), wasreduced by 50% and blood BHBAconcentrations were reduced by 40% for the highmonensin group. Based on the average feedintakes observed in this trial, the low monensingroup received approximately 208 mg/day ofmonensin and the high group 399 mg/day. Themonensin treatment commencing at two to fourweeks prior to calving reduced serum BHBA andNEFA in lactating dairy cows during the first 28days postpartum when monensin was fed at 300or 450 mg/day, but not by a daily dose of 150mg/day of monensin (Thomas et al., 1993).Serum glucose was not influenced by monensinfeeding. Australian cows treated with amonensin controlled release capsule (CRC)during the first week postcalving hadsignificantly lower plasma BHBAconcentrations and tended to have higher glucoseconcentrations than controls (Abe et al., 1994).A CRC that delivers 335 mg/day of monensinsodium for 95 days reduced the incidence ofsubclinical ketosis by 50% and also decreased


6

the duration of the condition when it wasadministered 3 weeks prior to expected calving(Duffield et al., 1998).

The most closely linked diseasesoccurring subsequent to subclinical ketosis aredisplaced abomasum and clinical ketosis.Estimates of milk production loss range from300 to 450 kg (660 to 990 lb) for a lactation(Dohoo and Martin, 1984; Gustafsson et al.,1993). These losses must be weighed againstthe cost of any prophylactic measure.Administration of a monensin CRC precalvingreduced the incidence of clinical ketosis by 50%,abomasal displacement by 40%, and multipleillness by 40% (more than one disease) (Duffieldet al., 1999b). The milk production responsedepended on body condition and was 0.85 kg/day (1.9 lb/day) at peak lactation in cows with aprecalving body condition score (BCS) of 3.25to 3.75 and was 1.2 kg/day (2.6 lb/day) for thefirst 90 days of lactation in fat cows (BCS³4.0)(Duffield et al., 1999a). No milk productionresponse was noted in thin cows, presumablybecause they had the lowest BHBAconcentrations and were at decreased risk ofsubclinical ketosis. A subsequent Canadianstudy conducted in 45 dairy herds confirmed thatmonensin CRC reduces the incidence ofdisplaced abomasum (Duffield et al., 2002). Apooled summary of the two Canadian projectsshowed that monensin CRC reduced theincidence of displaced abomasum and clinicalketosis by 40% each. In addition to the impactof monensin on displaced abomasum andclinical ketosis, pooled anlaysis of the twoCanadian CRC studies showed that the incidenceof retained placenta tended to be lowered by 25%in monensin treated cows (P = 0.09). Monensinis currently not approved in the United Statesfor use in lactating dairy cows.

Monitoring subclinical ketosis

When do I test cows?

By most definitions, the theoreticaltesting period for transition cows would extendfrom three weeks prior to calving until threeweeks after calving. Practically however, themost important time periods are: during the lastweek prior to calving and within the first twoweeks after calving.

Precalving

It is unusual for cows to developsubclinical ketosis precalving because theetiology of the condition depends on thehomeorhetic drive for milk production.However, cows in an energy deficit precalvingwill start mobilizing energy reserves in the finalweek before parturition. This can be measuredvia serum or plasma NEFA. The challenge forthis precalving sample is predicting when theanimal is going to calve. In most cases, a serumbank needs to be established and then samplesare submitted retrospectively once the actualcalving date is known.

Postcalving

A ketone testing program shouldcommence after calving. The primary riskperiod for subclinical ketosis is the first monthof calving. Our work at Guelph has indicatedthat the first two weeks postcalving is the timeof peak incidence. In addition, the median daysto diagnosis of clinical ketosis and displacedabomasum were 11 days. Thus, in order to tryto prevent subclinical disease from becomingclinical disease (if that is possible), cows mustbe identified early. For these reasons, asubclinical ketosis monitoring program shouldfocus on the first two to three weeks of lactation.


7

What test do I use?

NEFA

This test should only be used precalvingon samples obtained within one week ofparturition. Unfortunately, these restrictionsmake the utility of this test limited. However, itmay serve useful in certain situations, such as aherd investigation or an intervention follow-up.The data for this variable is frequently rightskewed and thus averages can be verymisleading. One suggested threshold is 0.5units/L. In recent work, cows within one weekof calving with serum NEFA above this thresholdwere at a 3.5 times greater risk of subsequentlydeveloping a displaced abomasum. Whole herdinterpretation is best made by calculating aproportion of cows above a threshold value;however, at this point, there are not a lot of gooddata on an appropriate goal for this parameter.In a multi-herd 1060 cow study near Guelph,30% of cows were above 0.5 U/L during the lastweek prior to calving.

Serum BHBA

In contrast to NEFA, serum BHBAshould only be used postcalving. The first twoweeks are the primary risk period for subclinicalketosis, defined by a serum concentration of1400 umol/L BHBA or greater. Although BHBAis the most stable of the ketones, it is the mostsubject to variation associated with feed intake,thus all samples on a given farm should alwaysbe taken at the same time of day. In addition,hemolysis is known to artificially elevate values;therefore, hemolyzed samples should beavoided. Other disadvantages of serum BHBAis the cost (approximately $5.00 per sample) andthe laboratory turn around time (minimum 24hours). However, all things considered, serumBHBA analysis is the gold standard from whichto compare cowside tests. A reasonable goal isto have less than two cows per 10 with BHBAabove 1400 umol/L in the first two weeks post-calving.

Milk Ketone Tests

Most milk ketone tests measure acetoneand acetoacetate through a chemical reactionwith nitroprusside which causes a color changefrom white to either pink or purple. These testsin general are poorly sensitive in milk (< 40%)but highly specific (> 90%). One exception isthe milk ketone test that measures BHBA. It ismarketed in Europe as “Ketolac BHBA”, inJapan as “Sanketopaper”, and in Canada as“Keto-Test”. This test has a much highersensitivity in milk (> 60%) and reasonably goodspecificity (> 70%, up to 90%). This is a semi-quantitative test that allows choosing a lowerthreshold for screening to increase sensitivityand a higher threshold for diagnosis to increasespecificity.

Urine Ketone Tests

The urine ketone tests are based on thesame nitroprusside reaction as the milk powderketone tests. These tests are highly sensitive(approaching 100%) but are poorly specific.Thus, they are great tests for ruling outsubclinical ketosis with a negative test result.However, their use overestimates a subclinicalketosis problem because of a high probabilityof false positive reactions. If the urine test wasused to evaluate the goal of less than two cowsper 10 with BHBA above 1400 umol/L in thefirst two weeks post-calving, an adjustment ofthe goal to less than 5 cows per 10 with positiveurine ketone tests would be required (Table 1).More work needs to be done to fully assess theutility of urine ketone tests.

Selection and Interpretation of Cowside Tests

It is most likely that in screening a groupof fresh cows, there would be two possibleactions resulting from a test. One action mightbe to treat positive animals with the goal toprevent subsequent development of clinicaldisease. In this case, a high predictive value ofa positive test is desired so that normal animals


8

are not unnecessarily treated. The second actionmight be to compare the percentage of positivereactors to a goal for determining theeffectiveness of either the transition ration orsome prophylactic measure in reducing theincidence of subclinical ketosis. In this situation,the apparent prevalence is the parameter thatactually would be used. Note from Table 1 thatthe urine ketone test would substantiallyoverestimate the prevalence of subclinicalketosis, while the Ketocheck™ test wouldgrossly underestimate the prevalence. This doesnot preclude these tests from being used.However, the impact of the inherent sensitivityand specificity of the test must be rememberedwhen establishing goals and interventionthresholds.

Herd Disease Records

Herd records (computerized or paper) areimportant tools for monitoring the incidence ofperiparturient disease. Producers should setgoals for minimizing the incidence of metabolicdisease. Herd consultants should periodicallyreview herd performance relative to the goals.In addition, intervention levels should also beconsidered. Several diseases are associated withincreasing age, and this must be taken intoaccount when assessing herd performance. Forexample, in monitoring and comparing herdincidence of milk fever and clinical ketosis, it isimportant to stratify this by parity. A highproportion of first lactation animals will give aherd a much lower incidence of milk fever andclinical ketosis, since risk increases with age.

Can herd incidence of certain diseasesbe used to decide whether a herd has a problemwith subclinical ketosis? Herd level analysis ofour 1995/1996 dataset involving 25 dairy herdsindicates that the herd incidence of displacedabomasum is positively associated with theprobability of a herd having a high incidence(> 20% in the first two weeks of lactation) ofsubclinical ketosis. The only other predictivevariable was precalving body condition score

(higher average increased risk). If greater than10% of the herd had a BCS > 4.0 at three weeksprecalving, that herd was extremely likely tohave a problem with subclinical ketosis.

DHI Test Day Data

Since milk fat and milk proteinpercentages are altered in sublinical ketosis,these parameters have been investigated for theirutility in defining subclinical ketosis. Amongall protein and fat parameters, a protein to fatratio of < 0.75 was the best test for diagnosingsubclinical ketosis, at the cow level, in aCanadian study (Duffield et al., 1997). However,the protein to fat ratio was not a good test overall,having a sensitivity of 58% and a specificity of69%.

Using data from a 25 herd studyconducted in Guelph in 1995, the mediancumulative herd incidence of subclinical ketosiswas 41% in the first two months postcalving.Summary data for each herd from each cow’sfirst DHI test postcalving were used to assessthe protein to fat ratio as a test at the herd levelfor classifying a herd as a high or low incidenceherd for subclinical ketosis. A herd mean proteinto fat ratio of <0.78 yielded a sensitivity of 69%and a specificity of 75% for identifying herdswith subclinical ketosis problems. Further, ifmore than 40% of cows in the herd at 1st DHItest had a protein to fat ratio of less than or equalto 0.75, those herds were likely to be problemherds. This test had a sensitivity of 69% and aspecificity of 83%. Although more work needsto be done on herd level indicators of subclinicalketosis, herd level protein to fat ratios appear tobe better indicators of herd level issues thanindividual cow protein to fat ratios are ofidentifying cows with subclinical ketosisproblems.

A recent study was conducted throughthe Ontario DHI testing facility to evaluate theutility of milk acetone measurements fordiagnosing subclinical ketosis. At the cow level,


9

test day milk acetone values were not found tobe useful in identifying cows at risk fordeveloping clinical metabolic disease (displacedabomasum or ketosis) or clinical lameness. Thisis not an issue with the acetone test methodologybut more likely a problem with the timing of themilk sample collection relative to the occurrenceof the disease event. The highest prevalence ofsubclinical ketosis occurs within two weeks ofcalving. Since most of our DHI testing programshave an interval of 30 to 45 days and cows lessthan five days in milk on test day are notsampled, the probability of testing all cowswithin two weeks of calving is low. Therefore,from an implementation standpoint, the testingof routine DHI samples for the purpose ofidentifying cows at risk of subclinical or clinicaldisease is inefficient.

Conclusions

Subclinical ketosis is an important andcommon disease in lactating dairy cows.Prevention depends largely on effective dry cownutrition and management. However, certainfeed additives, such as ionophores and rumenprotected choline, may be beneficial. Given thecost of subclinical ketosis, the fact it is a commonproblem in early lactation, and the strongassociation with clinical disease, monitoringprograms for subclinical ketosis during the firstfew weeks of lactation may be warranted. Thereare several cowside tests for subclinical ketosisavailable; however, all of the current tests havetheir strengths and weaknesses. The design andfrequency of a subclinical ketosis monitoringprogram will depend on the purpose of theprogram and the frequency of disease within theherd.

References

Abe, N., I.J. Lean, A. Rabiee, J. Porter, and C.Graham. 1994. Effects of odium monensin onreproductive performance of dairy cattle. II.Effects on metabolites in plasma, resumption ofovarian cyclicity and oestrus in lactating cows.Aust. Vet. J. 71(9): 277-282.

Andersson, L. 1988. Subclinical ketosis in dairycows. Metabolic Diseases of Ruminants, Vet.Clin. N. Amer. Food Animal Practice. 4(2):233-248.

Andersson, L. and U. Emanuelson. 1985. Anepidemiological study of hyperketonaemia inSwedish dairy cows; determinants and therelation to fertility. Prev. Vet. Med. 3:449-462.

Baird, D.G. 1982. Primary ketosis in the highproducing dairy cow: clinical and subclinicaldisorders, treatment, prevention and outlook. J.Dairy Sci. 65:1-10.

Butler, W.R. and R.D. Smith. 1989.Interrelationships between energy balance andpostpartum reproductive function in dairy cattle.J. Dairy Sci. 72:767-783.

Curtis, C.R., H.N. Erb, C.J. Sniffen, R.D. Smith,and D.S. Kronfeld. 1985. Path analysis of dryperiod nutrition, postpartum metabolic andreproductive disorders, and mastitis in Holsteincows. J. Dairy Sci. 68:2347-2360.

Christensen, J.O., F.E. Rasmussen, and R.R.Grummer. 1995. Influence of propylene glycoldelivery method on plasma metabolites of feedrestricted cattle. J. Dairy Sci. 78 (Suppl. 1):240.(Abstr.)

Cook, N.B., W.R. Ward, H. Dobson. 2001.Concentrations of ketones in milk in earlylactation, and reproductive performance of dairycows. Vet. Rec. 148: 769-772.


10

Dohoo, I.R. and S.W. Martin. 1984. Subclinicalketosis: Prevalence and associations withproduction and disease. Can. J. Comp. Med.48:1-5.

Duffield, T.F. 1997. Effects of a monensincontrolled release capsule on energy metabolism,health, and production in lactating dairy cattle.DVSc Thesis dissertation, Univ. of Guelph.

Duffield T, R. Bagg, L. DesCoteaux , E.Bouchard, M. Brodeur, D. DuTremblay, G.Keefe, S. LeBlanc, and P. Dick. 2002.Prepartum monensin for the reduction of energyassociated disease in postpartum dairy cows. J.Dairy Sci. 85:397-405.

Duffield, T.F., D.F. Kelton, K.E. Leslie, K.Lissemore, and J.H. Lumsden. 1997. Use of testday milk fat and milk protein to predictsubclinical ketosis in Ontario dairy cattle. Can.Vet. Journal 38:713-718.

Duffield, T.F., K.E. Leslie, D. Sandals, K.Lissemore, B.W. McBride, J.H. Lumsden, P.Dick, and R. Bagg. 1999a. Effect of prepartumadministration of a monensin controlled releasecapsule on milk production and milkcomponents in early lactation. J. Dairy Sci. 82:272-279.

Duffield, T.F., K.E. Leslie, D. Sandals, K.Lissemore, B.W. McBride, J.H. Lumsden, P.Dick, and R. Bagg. 1999b. Effect of prepartumadministration of a monensin controlled releasecapsule on cow health and reproduction. J. DairySci. 82:2377-2384.

Duffield, T.F., D. Sandals, K.E. Leslie, K.Lissemore, B.W. McBride, J.H. Lumsden, P.Dick, and R. Bagg. 1998. Efficacy of monensinfor the prevention of subclinical ketosis inlactating dairy cows. J. Dairy Sci. 81: 2866-2873.

Dufva, G.S., E.E. Bartley, A.D. Dayton, and D.O.Riddell. 1983. Effect of niacin supplementationon milk production and ketosis of dairy cattle.J. Dairy Sci. 66:2329-2336.

Emery, R.S., N. Burg, L.D. Brown, and G.N.Blank. 1964. Detection, occurrence, andprophylactic treatment of borderline ketosis withpropylene glycol feeding. J. Dairy Sci. 47:1074-1079.

Fisher, L.J., J.D. Erfle, G.A. Lodge, and F.D.Sauer. 1973. Effects of propylene glycol orglycerol supplementation of the diet of dairycows on feed intake, milk yield and composition,and incidence of ketosis. Can. J. Anim. Sci. 53:289-296.

Formigoni, A., M.C. Cornil, A. Prandi, A.Mordenti, A. Rossi, D. Portetelles, and R.Renaville. 1996. Effect of propylene glycolsupplementation around parturition on milkyield, reproductive performance and somehormonal and metabolic characteristics in dairycows. J. Dairy Res. 63:11-24.

Fronk, T.J., and L.H. Schultz. 1979. Oralnicotinic acid as a treatment for ketosis. J. DairySci. 62:1804-1807.

Geishauser, T., K. Leslie, and K. Kelton. 2001.Monitoring subclinical ketosis in dairy herds.Comp. Cont. Ed. 23: s65-s71.

Goff, J.P., and R.L. Horst. 1997. Physiologicalchanges at parturition and their relationship tometabolic disorders. J. Dairy Sci. 80:1260-1268.

Grohn, Y.T., H.N. Erb, C.E. McCulloch, andH.S. Saloniemi. 1989. Epidemiology ofmetabolic disorders in dairy cattle: Associationamong host characteristics, disease, andproduction. J. Dairy Sci. 72:1876-1885.


11

Gustafsson , A.H., L. Andersson, and U.Emanuelson. 1993. Effect of hyperketonemia,feeding frequency and intake of concentrate andenergy on milk yield in dairy cows. Anim. Prod.56:51-60.

Herdt, T.H., J.B. Stevens, W.G. Olson, and V.Larson. 1981. Blood concentrations of β-hydroxybutyrate in clinically normal Holstein-Friesian herds and in those with a highprevalence of clinical ketosis. Am. J. Vet.Res.12(3):503-506.

Kauppinen, K. 1984. Annual milk yield andreproductive performance of ketotic and nonketotic dairy cows. Zbl. Vet. Med. A.31:694 -704.

Kremer, W.D.J., E.N. Noordhuizen Stassen, F.J.Grommers, Y.H. Schukken, R. Heeringa, A.Brand, and C. Burvenich. 1993. Severity ofexperimental Escherichia coli mastitis inketonemic and nonketonemic dairy cows. J.Dairy Sci. 76:3428-3436.

Lean, I.J., M.L. Bruss, R.L. Baldwin, and H.F.Trout. 1991. Bovine ketosis: a review. I.Epidemiology and pathogenesis. Vet. Bulletin.61(12):1209-1218.

Leslie, K.E., T.F. Duffield, Y.H. Schukken, andS.J. LeBlanc. 2000. The influence of negativeenergy balance on udder health. NationalMastitis Council, Regional MeetingProceedings. Pg. 25-33.

Miettenen, P.V.A. 1994. Relationship betweenmilk acetone and milk yield in individual cows.J. Vet. Med. A. 41:102-109.

Miettinen, P.V.A., and J.J. Setala. 1993.Relationships between subclinical ketosis, milkproduction and fertility in Finnish dairy cattle.Prev. Vet. Med. 17:1 8.

National Research Council. 2001. Nutrientrequirements of dairy cattle. 7th rev. ed. Natl.Acad. Sci., Washington, DC.

Oetzel, G.R. 1998. Dairy: NutritionManagement. Nutritional management of drydairy cows. Comp. Cont. Ed., March, FoodAnimal 391-396.

Piepenbrink, M.S., and T.R. Overton. 2000.Liver metabolism and production ofperiparturient dairy cattle fed rumen-protectedcholine. J. Dairy Sci. 83(Suppl. 1):257 (Abstr.).

Radostits, O.M., K.E. Leslie, and J. Fetrow.1994. J. Dairy cattle nutrition. In: Herd Health:Food Animal Production Medicine, 2nd ed. WBSaunders Co, Philadelphia, PA.

Rogers, P.A.M. and M.J. Hope-Cawdery. 1980.Monensin, ketosis and nitrate toxicity in cows.Vet. Rec. 106:311-312.

Sauer, F.D., J.D. Erfle, and L.J. Fisher. 1973.Propylene glycol and glycerol as a feed additivefor lactating dairy cows: an evaluation of bloodmetabolite parameters. Can. J. Anim. Sci.53:265-271.

Sauer, F.D., J.K.G. Kramer, and W.J. Cantwell.1989. Antiketogenic effects of monensin in earlylactation. J. Dairy Sci. 72:436-442.

Schultz, L.H. 1958. Use of sodium propionatein the prevention of ketosis in dairy cattle. J.Dairy Sci. 41:160-168.

Steen , A., O. Osteras, and H. Gronstol. 1996.Evaluation of additional acetone and ureaanalyses, and of the fat lactose quotient in cowmilk samples in the herd recording system inNorway. J. Vet. Med. A. 43:181-191.

Stevenson, J.S., and E.P. Call. 1983. Influenceof early estrus, ovulation, and insemination onfertility in postpartum Holstein cows.Theriogenology 19:367-375.


12

Struder, V.A., R.R. Grummer, and S.J. Bertics.1993. Effect of prepartum propylene glycoladministration on periparturient fatty liver indairy cows. J. Dairy Sci. 76:2931-2939.

Syvajarvi, J., H. Saloniemi, and Y. Grohn. 1986.An epidemiological and genetic study onregistered diseases in Finnish Ayrshire cattle.Acta. Vet. Scand. 27:223-234.

Thomas, E.E., S.E. Poe, R.K. McGuffey, D.H.Mowrey, and R.D. Allrich. 1993. Effect offeeding monensin to diary cows on milkproduction and serum metabolites during earlylactation. J. Dairy Sci. 76(Suppl. 1):280. (Abstr.)

Tveit, B., F. Lingaas, M. Svendsen, and O.V.Sjaastad. 1992. Etiology of acetonemia inNorwegian cattle. 1. Effect of ketogenic silage,season, energy level, and genetic factors. J. DairySci. 75:2421-2432.

Tyler, J.W., D.F. Wolfe, and R. Maddox. 1992.Clinical indications for dietary ionophores inruminants. Comp. Cont. Ed. 14(7):989-993.

Whitaker, D.A., E.J. Smith, G.O. da Rosa, andJ.M. Kelly. 1993. Some effects of nutrition andmanagement on the fertility of dairy cattle. Vet.Rec. 133:61-64.

Whitmore, H.L., W.J. Tyler, and L.E. Casida.1974. Effects of early postpartum breeding indairy cattle. J. Anim. Sci. 38:339-346.


13

Table 1. Use of cowside ketone tests in screening programs for identifying subclinical ketosis.

20% Prevalence 40% Prevalence 60% Prevalence PV1 PV2 AP3 PV PV AP PV PV APTest +ve -ve +ve -ve +ve -ve

Keto-Test using 62% 93% 23% 81% 83% 35% 91% 68% 48%100 µmmol/L

Ketocheck 90% 86% 8% 96% 70% 16% 98% 51% 23%(milk)

Urine 38% 100% 53% 62% 100% 65% 78% 100% 76%

1PV +ve: Predictive value of a positive test result.2PV –ve: Predictive value of a negative test result.3Apparent prevalence.


14


15

Nutritional Strategies to Promote Calf HealthDuring the Liquid Feeding Period

James K. Drackley1

1Contact at: 260 Animal Sciences Laboratory, 1207 West Gregory Drive, Urbana, IL 61801, (217) 244-3157, FAX (217)333-7088, Email: [email protected]

Department of Animal SciencesUniversity of Illinois

Abstract

Morbidity and mortality of pre-weaneddairy calves remains a significant problem in theUnited States. Appropriate nutrition is acritically important component of optimal healthfor young dairy calves. However, good nutritioncannot overcome the negative effects of poorenvironmental conditions and large pathogenloads. While no nutritional “magic bullets”exist, improving nutrition of calves in the overallcontext of good management can be expectedto benefit health. Perhaps, the biggest windowfor improvement lies in improving energy andprotein supply to calves during the first two tothree weeks of life. Considerable evidence incalves and in other species points to thedetrimental effects of inadequate milk intake onsusceptibility to disease or impaired function ofthe immune system. Many feed ingredients andadditives are available that have been purportedto improve health in calves. The evidence forefficacy of these materials is briefly reviewed inthis paper.

Introduction

Other than the period of transition frompregnancy to lactation, the milk-feeding periodfor calves represents the time of life when dairyanimals are subject to the greatest number ofhealth problems. According to results of therecently released Dairy 2002 survey from theUSDA-APHIS National Animal Health

Monitoring System (NAHMS), 8.7% of heifercalves born alive died before weaning. Scouringand other digestive problems were the reportedcauses of death in 62.1% of reported calf deaths,with respiratory problems representing 21.3%.It is frustrating that these figures have changedlittle since the 1992 NAHMS survey (NAHMS,1993), despite increased knowledge andeducational efforts on the importance ofappropriate colostrum management, nutrition,and health programs. While many dairy farmersachieve much lower calf mortality, these surveyresults indicate that health of the pre-weaneddairy calf is still a limiting factor on far too manyU.S. dairy farms.

Commercial and producer interests in theyoung calf seem to have increased recently.Several factors may be involved in this renewedinterest. Expansion of contract or custom rearingof heifers increases the importance of healthy,well-grown calves. The trend to larger and largerdairy operations has created new issues ofpracticality in management and places newemphasis on overall profitability. Implicationsand applications of “accelerated growth”schemes have drawn considerable interest.Perhaps the most important driver in the renewedinterest in calf nutrition and management hasbeen the high demand, short supply, and resultantsky-rocketing prices for pregnant heifers duringthe preceding two calendar years.


16

The authors of the NAHMS Dairy 2002report concluded their summary of results onmortality of U.S. dairy cattle as follows:“Management practices that reduce both calf andcow loss during or immediately subsequent tocalving should be reviewed.” (NAHMS, 2002).Given the sobering statistics on the lack ofprogress in lowering calf mortality, and therenewed interest in practices that may contributeto improved health of dairy calves, the objectiveof this paper is to address some aspects ofnutritional management during the liquid feedingperiod that may lead to improved calf health.

Calf health resides at the center ofcomplex interactions among the environment,disease-causing agents, and nutrition of the calf(Davis and Drackley, 1998). Breakdowns ordeficiencies of management that providesuboptimal environment (lack of sanitation,temperature or humidity stressors, inadequatehealth management protocols, etc.), poorcolostrum management (Davis and Drackley,1998; Quigley and Drewry, 1998), or suboptimalnutrition may predispose calves to disease.Management must strive to minimize exposureto high populations of disease-causing microbesand maximize the ability of the calf to ward offinfection. Success in these areas is oftendeceptively simple and centers on doing thebasic tasks well. Dairy producers, andagribusiness professionals that work withproducers, must not be mislead into the promiseof any “magic bullet” vaccine or nutritionalsupplement that will overcome poormanagement.

Nutrition and Health

Energy and protein supply

The biggest opportunity to improve calfhealth nutritionally may lie in improving theearly nutritional status of young dairy calves.Unequivocal data exist for the importance ofadequate energy and protein intakes in younganimals of many other species, including

humans. There is little reason to suspect thatdairy calves would be fundamentally differentin this regard.

Conventional heifer calf rearing schemesrely on restricted feeding of milk replacer toencourage early intake of starter. Typical liquidfeeding rates of 8 to 10% of body weight aremuch lower than ad libitum feeding rates.Calves allowed to suckle their dams typicallynurse between 6 and 10 times daily and consumebetween 16 and 24% of body weight as milk(Hafez and Lineweaver, 1968). This translatesto 16 to 24 lb, or 1.9 to 2.8 gal, for a 100-lb calf.Since whole milk contains about 12.5% solids,nursing calves would consume 2 to 3 lb of drymilk solids daily and could gain from 2.1 to 3lb/day. Recent experiments by Dan Weary’sresearch group at the University of BritishColumbia have demonstrated that calves of“modern” Canadian Holstein genetics consumedup to 20 lb of milk daily by day 5 of life, andover 22 lb/day by the fourth week of life(Appleby et al., 2001; Jasper and Weary, 2002).Considering standard practice in the beef cattleindustry, these facts should not be surprising.Consequently, what is currently termed“accelerated calf growth” probably is morecorrectly thought of as biologically “normal”growth for young calves.

The focus on the restricted feeding of theliquid diet is to encourage early development ofcalf starter intake. When management of earlyweaning programs is excellent, gains of 2 lb/day or more are possible by four weeks of age(Kertz et al., 1979). However, nutrient suppliesduring the first two weeks of life before starterintake becomes appreciable are very low relativeto voluntary consumption rates. Consequently,until starter intake begins to increase late in thesecond week of life (under the best managementin early weaning systems), calves receive enoughnutrients from milk or milk replacer only tocover maintenance and low rates of growth.


17

These concepts are illustrated in Table1, in which nutrient requirements as specifiedby the National Research Council (NRC, 2001)are shown. The overall feeding rate determinesenergy intake and sets limits on the growthpossible. Note that as average daily gain (ADG)increases, the required metabolizable energy(ME) intake also increases and more milkreplacer powder must be fed to provide thatenergy. The basis that this idea holds true inpractice is demonstrated by recent growthexperiments in which calf growth rates increasedin direct proportion to the amount of milk ormilk replacer fed (Bartlett et al., 2001; Diaz etal., 2001; Jasper and Weary, 2002). Calvesrequire over 0.8 lb of milk replacer powder justto support maintenance; consequently, standardrestricted feeding rates that provide only 1 lb ofmilk replacer powder provides only enoughenergy to support maintenance and about 0.25lb of body weight gain daily. It must be pointedout that these requirements assume that calvesare housed under thermoneutral conditions; ifcalves are subject to either heat stress or coldstress, energy intake may be insufficient evenfor maintenance.

The amount of protein required is largelydriven by the rate of growth, becausemaintenance requirements are small and intheory could be met with as little as 8.3% crudeprotein (CP) in milk replacer (NRC, 2001; Table1). Tissue deposition requires an average of 30grams of nitrogen per kgilogram of liveweightgained, or 187.5 g CP per kilogram of ADG(NRC, 2001). Consequently, the amount of CPrequired by the calf increases as it is fed moreenergy and rates of gain increase (Table 1). Notein Table 1 that the content of CP needed in milkreplacer approaches a maximum in the range of28% CP.

Also note in Table 1 that the calculatedcontent of CP in milk replacer needed for calvesgaining 0.5 lb/day is about 18%. On the basisof calculated protein required for maintenanceand growth of calves fed typically recommended

amounts of milk replacer (1 lb/day of powder or10% of BW), 18% CP likely is sufficient to meetabsolute requirements for maintenance andgrowth. However, increasing milk replacerprotein content results in marked increases inrate of gain and efficiency of gain (Donnelly andHutton, 1976a,b; Bartlett et al., 2001; Blome etal., 2003).

Effects of energy and protein supply on healthand immune status

What is the evidence that inadequatenutrition during early life decreases resistanceto disease and compromises health and wellbeing of calves? Williams et al. (1981)compared calves fed two amounts of milkreplacer solids (1.3 and either 0.7 or 0.9 lb/day)with either ad libitum or restricted access to calfstarter. Calves fed the higher amount of milkreplacer with ad libitum access to starter had thegreatest ADG and least mortality. Other studieshave shown that inadequate nutrition results inimpaired immune responses in young calves.Griebel et al. (1987) fed neonatal calves eitherbelow maintenance or above maintenanceintakes of milk replacer. Calves fed belowmaintenance lost BW; lymphocytes isolatedfrom these calves had decreased proliferativeresponses compared with adequately fed calves.Malnourished calves had lower primary antibodyresponse to administration of K99 antigen.Pollock et al. (1993, 1994) compared effects ofweaning age (5, 9, or 13 wk of age) and twolevels of nutrition (0.9 lb or 2.2 lb/day of milkreplacer powder). Weaning at 5 wk resulted incompromised lymphocyte responses (cellularimmunity) at 10 wk of age. The higher level ofnutrition, which was approximately twicemaintenance, resulted in improved responses ofcell-mediated immunity and decreased skinresponses to antigen (Pollock et al., 1993). Incontrast, the high level of nutrition resulted indecreased antibody titers to specific antigens,without changing total Ig concentration in serum(Pollock et al., 1994). More recent evidencedemonstrated that feeding for greater (normal)


18

growth increased nitric oxide and interferon-γproduction by isolated mononuclear leukocytescompared with cells from conventionally fedcalves (Nonnecke et al., 2000).

One way in which improved neonatalnutritional status might be expected to impactthe immune system is via growth hormone (GH)and the insulin-like growth factors (IGF). Theseanabolic hormones play a direct role inintegrating the growth, maintenance, repair, andfunction of the immune system (Clark, 1997).Lymphocytes express receptors for both GH andIGF-I. In rodents, IGF-I causes growth andmaturation of B cells, increases size of thethymus and spleen, and increases antibodyproduction by B cells (see review by Clark,1997). Consequently, assuming that responsesin calves are similar, increased concentrationsof IGF-I resulting from improved nutrition mightbe expected to enhance immunocompetence incalves.

Smith et al. (2002) recently reported thatcalves fed on an accelerated growth scheme hadgreater concentrations of IGF-I than did calvesfed milk replacer at a conventional (restricted)rate. Plasma from calves in both groups showedincreased IGF-I concentrations in response toinjection of bovine somatotropin at 5 wk of age,but calves on the higher plane of nutrition alsoresponded with increased growth rates. In ourown experiments, IGF-I in plasma (analyzed byM. J. VandeHaar, Michigan State University)was increased by greater feeding rates andincreased linearly as CP was increased (datanot shown). These results clearly show afunctional IGF-I system in young calves that isresponsive to early nutritional status. Therelationship of the enhanced IGF-1 status toimmune function remains to be determined.

A major argument in favor of restrictedliquid feeding and early weaning has been thatscouring is decreased. In most species, fecalconsistency clearly becomes less fluid as dry feedis consumed, primarily as a result of the bulking

effect of dietary fiber. However, merely feedingmore milk or more of a high-quality milkreplacer does not cause scouring (Mylrea, 1966;Huber et al., 1984; Nocek and Braund, 1986;Appleby et al., 2001; Diaz et al., 2001; Wearyand Jasper, 2002). The occurrence of calf scours,unless a poor-quality milk replacer containingdamaged ingredients is fed, depends more onthe load of pathogenic microorganisms in thecalf’s environment (Roy, 1980). Our ownexperiences with calves fed milk replacer at upto 18% of BW indicates that average fecal scoresare only slightly (nonsignificant) increased butthat the number of days with more fluid fecalscores is increased (Bartlett et al., 2000,unpublished data). Feeding milk replacer resultsin more fluid feces than feeding similar amountsof whole milk, regardless of the composition ofthe milk replacer (Bartlett et al., 2000,unpublished data).

Lessons from other species also shouldbe useful for improving our management of dairycalves. The degree of immune challengepresented by the environment can have markedeffects on growth and feed efficiency. Forexample, pigs subjected to challenge withbacterial lipopolysaccharide had decreasedgrowth and feed efficiency compared withcontrol pigs; growth for challenged pigs waslower than pair-fed littermates, indicating thatdecreased feed intake accounted for only about2/3 of the growth depression (Dritz et al., 1996).Effects of immune challenge did not interactwith diet complexity (Dritz et al., 1996). Pigsraised in a clean environment with minimalexposure to pathogens had greater rates of gain,improved feed efficiencies, and a greater lean-to-fat ratio than pigs raised in the presence of ahigh degree of pathogen exposure (Williams etal., 1997). Such responses are probablyattributable both to increased nutrient demandsby the immune system and to the anti-growtheffects of cytokines produced by the activatedimmune system. Klasing and Calvert (1999)calculated that up to 60% of the impaired growthof chicks during an intense immune response


19

could be explained by known processes. Almost7% of lysine intake was used by immuneprocesses during a lipopolysaccharide challengein chicks (Klasing and Calvert, 1999).Consequently, nutrient demands of the immunesystem may compete directly with growthprocesses for a limited supply of nutrients.

A fascinating discussion of thepartitioning of nutrients between maintenance,immune function, and growth or otherproductive functions has been provided byHoudijk et al. (2001). In particular, thepossibility that inadequate protein supply maylimit function of the immune system in calvesis intriguing. The potential importance of proteinis suggested indirectly from our ownexperiments (Bartlett et al., 2001; Blome et al.,2003) in which increasing protein content ofisocaloric milk replacers markedly increasedADG and efficiency of gain. For example, inlimit-fed calves (12% of body weight as liquid)offered isocaloric milk replacers in which theprotein content increased from 18% to 26%,ADG (from 0.84 to 1.36 lb/day) and efficiencyof gain (from 0.51 to 0.78 lb gain/lb feed)increased linearly at similar overall energyretention (Blome et al., 2003).

Vitamins and trace minerals

Many of the vitamins and trace mineralscan impact function of the immune system incalves. For a detailed discussion of these effectsin dairy cattle, readers are referred to the NRC(2001). Vitamin E deserves special mentionhere. A series of experiments by Jim Morrilland colleagues at Kansas State University(Reddy et al., 1986, 1987a,b; Eicher-Pruiett etal., 1992; Eicher et al., 1994) has built aconvincing argument that vitamin E is beneficialto calf health. The latest NRC committeemodestly raised the requirement for vitamin Efrom 18 IU/lb to 23 IU/lb of dietary dry matter(NRC, 2001). The committee declined to raisethe vitamin E requirement further in the absenceof large clinical studies demonstrating health

benefits. Most manufacturers of milk replacersroutinely supplement with much higher amountsof vitamin E.

Type of Liquid Feed

Options for feeding after colostruminclude whole milk, excess colostrum andtransition milk, waste or discard milk, and milkreplacer. Pros and cons associated with eachfeed type have been discussed in detail elsewhere(Davis and Drackley, 1998; Drackley, 1999).Many producers successfully use a pool of allnon-saleable milk to feed calves. Concernsabout bacterial load have resulted in increasesin on-farm pasteurization on larger farms.California studies (Jamaluddin et al., 1996a,b)showed that pasteurization increased growthrates of calves and that calves fed pasteurizednon-saleable milk were worth $8.13 more thancalves fed non-pasteurized milk. While properpasteurization is effective in inactivating thecausative bacteria for Johne’s disease (Stabel,2001) and Mycoplasma mastitis (Butler et al.,2000), it will not destroy all viruses. Producersin strict biosecurity programs should feed milkreplacer.

Milk replacers are fed on a majority ofU.S. dairy farms (Heinrichs et al., 1995). High-quality milk replacers are excellent liquid feedsfor young calves. Because of the high cost ofdried skim milk, the all-milk protein milkreplacers used currently are based almostexclusively on whey proteins (whey proteinconcentrate, dried whey, and delactosed whey).Milk replacers based on whey proteins aredigested and utilized at least as efficiently as theskim milk proteins used in earlier milk replacerformulations (Terosky et al., 1997; Lammers etal., 1998). Reports of poor calf performance onmilk replacer often are attributable to selectionof an inappropriate or poor-quality milk replacer,to underfeeding the calf, or to an underlyingdisease or sanitation problem. Milk replaceralmost always will be a cheaper feed for youngcalves than saleable whole milk. Although more


20

expensive than surplus colostrum, transitionmilk, or waste milk, milk replacers haveadvantages in consistency of product from dayto day, ease and flexibility of storage, and diseasecontrol. An in-depth discussion of themanufacture and use of milk replacers can befound elsewhere (Davis and Drackley, 1998).

Waste or discard milk and excesscolostrum/transition milk are often thought ofas “free feeds”. However, if waste milk werenot being produced, then the “free milk” wouldbe receiving the milk sale price. Thus, there issignificant “opportunity cost” associated withexcessive dumping of milk, which emphasizesone of the many reasons for mastitis control toincrease profits to dairy producers.Nevertheless, nearly all farms will have somewaste milk available at times. On the other hand,excess colostrum and transition milk has noother use – it is truly a “free feed” except foradditional labor costs and costs associated withstorage. However, producers must keep in mindthe disease control considerations of feeding anywaste milk or colostrum/transition milk, whichcould significantly increase the “true” costs ofusing these feeds.

Many producers will use whatever non-saleable milk is available each day to feed calves,whether excess colostrum, transition milk, ordiscard milk. This practice results in the calfreceiving a diet that varies considerably incomposition from day to day. Such variabilityreportedly does not affect the incidence orseverity of scouring or overall rates of gain(Foley and Otterby, 1978). Even frequentchanges between sources of colostrum or wastemilk and milk replacer did not affect calvesadversely in several earlier studies (Applemanand Owen, 1975). However, maintaining asmuch consistency as possible in the diet foryoung calves minimizes chances for digestiveupsets. This may be particularly important whencalves are raised under conditions of increasedstress, such as cold or wet weather or duringoutbreaks of disease.

Effects of Additives and Supplements on CalfHealth

General perspectives

As mentioned in the introduction, thelikelihood of finding a “magic bullet”supplement for improved health in calves isremote. Nevertheless, supplements may behelpful in improving health or growth of calves.One of the problems in evaluating the currentliterature on such additives is that most havebeen tested in calves fed at restricted intake, andthus have been more likely limited by energyand protein intakes than by the ingredients inthe additives. The usefulness of additives maybe enhanced if tested against the background ofmore adequate nutrition. For many of theseadditives, biologically significant increments inperformance or improved health may be toosmall to detect statistically in many experimentsbecause of inadequate numbers of calves.

It also is possible that several of thefollowing components may each provide smallimprovements that will be additive, such that aproduct combining several different types ofcompounds might show substantial efficacy.One hope is that such combinations might beable to compensate for antibiotics if use inlivestock feeding is eventually banned in the U.S.For example, a product containing probiotics,fructooligosaccharides, and alliumsupplemented to milk replacer withoutantibiotics provided similar calf performanceand health as a medicated milk replacer(Donovan et al., 2002).

Antibiotics

The use of antibiotics in all animalproduction is coming under increasing criticism.However, antibiotics typically added to“medicated” milk replacers (primarilyoxytetracycline and neomycin) still are clearlyeffective in improving performance and healthof calves (Morrill et al., 1977b; Quigley et al.,


21

1997; Davis and Drackley, 1998). For producerswith excellent sanitation and good nutritionprograms, medicated milk replacers probably arenot necessary. However, under less than optimalmanagement, or where producers are bringingcalves from other farms or unknown sources ontheir farm, medicated milk replacers will almostalways be beneficial (Davis and Drackley, 1998).

Coccidiostats

Coccidiosis is a common clincial orsubclinical cause of poor performance, illness,and economic loss in dairy calves. Use ofcoccidostats (decoquinate or lasalocid) in milkor milk replacer has been shown experimentallyto lessen the effects of coccidiosis and improvegains (Bauer et al., 1992; Webb et al., 1992;Quigley et al., 1997a). While use of coccidostatsin starters has been widely recommended andadopted, calves may not consume sufficientamounts of starter before 4 wk to preventinfection during the first 2 to 3 weeks of life(Quigley et al., 1997a).

Prebiotics and Probiotics

Prebiotics are live bacterial culturesdesigned to increase colonization of the digestivetract with species that produce favorable effectsand compete with pathogenic microorganismsfor nutrients and attachment sites. Variousprebiotic bacterial cultures have been examinedand promoted for use in dairy calves. Acombination of Bifidobacterium pseudolongumand Lactobacillus adidophilus in the diet of pre-weaned calves increased body weight gain anddecreased scouring (Abe et al., 1995). Calvesfed a probiotic (composition not reported)without antibiotics had growth rates and healthmeasures similar to that of calves fed a similarmilk replacer with antibiotics (Morrill et al.,1995). Small but generally positive effects frombacterial probiotics have been noted in otherstudies (Jenny et al., 1991; Higginbotham andBath, 1993; Cruywagen et al., 1996).

Prebiotics are compounds that areindigestible by the host animal’s digestiveenzymes but are usable by microorganisms.These compounds provide substrate for growthof “beneficial” bacteria in the intestine orcompete with pathogenic bacteria for intestinalattachment sites. Complex carbohydrates suchas oligofructose, mannanoligosaccharides, andothers have shown promise in young calves aswell as other species (Webb et al., 1992; Kaufoldet al., 2000; Flickinger and Fahey, 2002). Calvesfed the trisaccharide galactosyl-lactose in milkreplacer had greater weight gains and fewer daysscouring than control calves (Quigley et al.,1997b).

Plasma proteins and immunoglobulins

Plasma protein products have beeninvestigated for their potential to increase growthand health of young calves as they do in youngpigs (e.g., Hansen et al., 1993). Calves fed milkreplacers containing plasma protein had similar(Quigley and Bernard, 1996) or improvedperformance (Morrill et al., 1995; Quigley et al.,2002) relative to controls. Supplements preparedfrom serum or immunoglobulin fractions alsomay have promise for improving calf health(Quigley and Drew, 2000; Quigley et al., 2002).

Other additives

Many other compounds have beenpurported to have health benefits. A recent studydocumented improvements in health and growthfrom addition of fresh or autoclaved rumen fluidto young calves’ diets (Muscato et al., 2002).The mode of action for the effect is uncertainbut may be related to stimulation of the immunesystem by bacterial polysaccharides present inthe rumen fluid extracts. There is considerableinterest in other “neutraceutical” compounds forcalves, such as garlic extracts or other herbalproducts. For example, allicin, a component ofgarlic, has antimicrobial properties against avariety of microbial classes. A clinical trialdemonstrated that allicin had little effect on the


22

infection by Cryptosporidia in young calves(Olson et al., 1998). There are relatively fewreports in refereed scientific journals on theefficacy of herbal products in calves.

References

Abe, F., N. Ishibashi, and S. Shimamura. 1995.Effect of administration of bifidobacteria andlactic acid bacteria to newborn calves andpiglets. J. Dairy Sci. 78:2838-2846.

Appleby, M.C., D.M. Weary, and B. Chua. 2001.Performance and feeding behaviour of calves onad libitum milk from artificial teats. Appl. Anim.Behav. Sci. 74:191-201.

Appleman, R.D., and F.G. Owen. 1975.Breeding, housing, and feeding management. J.Dairy Sci. 58:447-464.

Bartlett, K.S., F.K. McKeith, and J.K. Drackley.2001. Effects of feeding rate and proteinconcentration in milk replacers on growth andbody composition of Holstein calves. J. DairySci. 84:1560. (Abstr.)

Bauer, K.J., D.J. Schauff, G.C. McCoy, and J.H.Clark. 1992. Effect of feeding decoquinate inreplacer and starter to growing calves. J. DairySci. 75(Suppl. 1):268. (Abstr.)

Blome, R.M., J.K. Drackley, F.K. McKeith, M.F.Hutjens, and G.C. McCoy. 2003. Growth,nutrient utilization, and body composition ofdairy calves fed milk replacers containingdifferent amounts of protein. J. Anim. Sci. 81:(inpress).

Butler, J.A., S.A. Sickles, C.J. Johanns, and R.F.Rosenbusch. 2000. Pasteurization of discardmycoplasma mastitic milk used to feed calves:thermal effects on various mycoplasma. J. DairySci. 83: 2285-2288.

Clark, R. 1997. The somatogenic hormonesand insulin-like growth factor-1: Stimulators oflymphopoiesis and immune function. EndocrineRev. 18:157-179.

Cruywagen, C.W., I. Jordaan, and L. Venter.1996. Effect of Lactobacillus acidophilussupplementation of milk replacer on preweaningperformance of calves. J. Dairy Sci. 79:483-486.

Davis, C.L., and J.K. Drackley. 1998. TheDevelopment, Nutrition, and Management of theYoung Calf. Iowa State University Press, Ames,IA.

Diaz, M.C., M.E. Van Amburgh, J.M. Smith,J.M. Kelsey, and E.L. Hutten. 2001.Composition of growth of Holstein calves fedmilk replacer from birth to 105-kilogram bodyweight. J. Dairy Sci. 84:830-842.

Donnelly, P.E., and J.B. Hutton. 1976a. Effectsof dietary protein and energy on growth ofFriesian bull calves. I. Food intake, growth, andprotein requirements. N.Z. J. Agric. Res.19:289-227.

Donnelly, P.E., and J.B. Hutton. 1976b. Effectsof dietary protein and energy on growth ofFriesian bull calves. II. Effects of level of feedintake and dietary protein content on bodycomposition. N.Z. J. Agric. Res. 19:409-414.

Donovan, D.C., S.T. Franklin, C.C.L. Chase, andA.R. Hippen. 2002. Growth and health ofHolstein calves fed milk replacers supplementedwith antibiotics or Enteroguard. J. Dairy Sci.85:947-950.

Drackley, J.K. 1999. Critical evaluation offeeding options for replacement calves. Pages141-152 in Advances in Dairy Technology, vol.11, J. Kennelly, ed. Proc. Western CanadianDairy Seminar, Univ. Alberta, Edmonton, AB,Canada.


23

Dritz, S.S., K.Q. Owen, R.D. Goodband, J.L.Nelssen, M.D. Tokach, M.M. Chengappa, andF. Blecha. 1996. Influence oflipopolysaccharide-induced immune challengeand diet complexity on growth performance andacute-phase protein production in segregatedearly-weaned pigs. J. Anim. Sci. 74:1620-1628.

Eicher, S.D., J.L. Morrill, F. Blecha, C.G. Chitko-McKown, N.V. Anderson, and J.J. Higgins.1994. Leukocyte functions of young dairy calvesfed milk replacers supplemented with vitaminsA and E. J. Dairy Sci. 77:1399-1407.

Eicher-Pruiett, S.D., J.L. Morrill, T.G. Nagaraja,J.J. Higgins, N.V. Anderson, and P.G. Reddy.1992. Response of young dairy calves withlasalocid delivery varied in feed sources. J.Dairy Sci. 75:857-862.

Flickinger, E.A., and G.C. Fahey, Jr. 2002. Petfood and feed applications of inulin,oligofructose and other oligosaccharides. Br. J.Nutr. 87(Suppl. 2):S297-S300.

Foley, J. A., and D.E. Otterby. 1978. Availability,storage, treatment, composition, and feedingvalue of surplus colostrum: a review. J. DairySci. 61:1033-1060.

Griebel, P.J., M. Schoonderwoerd, and L.A.Babiuk. 1987. Ontogeny of the immuneresponse: effect of protein energy malnutritionin neonatal calves. Can. J. Vet. Res. 51:428-435.

Hafez, E.S.E., and L.A. Lineweaver. 1968.Suckling behaviour in natural and artificially fedneonate calves. Z. Tierpsychol. 25:187-198.

Hansen, J.A., J.L. Nelssen, R.D. Goodband, andT.L. Weeden. 1993. Evaluation of animalprotein supplements in diets of early-weanedpigs. J. Anim. Sci. 71:1853-1862.

Heinrichs, A.J., S.J. Wells, and W.C. Losinger.1995. A study of the use of milk replacers fordairy calves in the United States. J. Dairy Sci.78:2831-2837.

Higginbotham, G.E., and D.L. Bath. 1993.Evaluation of Lactobacillus fermentationcultures in calf feeding systems. J. Dairy Sci.76:615-620.

Houdijk, J.G.M., N.S. Jessop, and I. Kyriazakis.2001. Nutrient partitioning betweenreproductive and immune functions in animals.Proc. Nutr. Soc. 60:515-525.

Huber, J.T., A.G. Silva, O.F. Campos, and C.M.Mathieu. 1984. Influence of feeding differentamounts of milk on performance, health, andabsorption capability of baby calves. J. DairySci. 67:2957-2963.

Jamaluddin, A.A., T.E. Carpenter, D.W. Hird,and M.C. Thurmond. 1996a. Economics offeeding pasteurized colostrum and pasteurizedwaste milk to dairy calves. J. Am. Vet. Med.Assoc. 209:751-756.

Jamaluddin, A. A., D.W. Hird, M.C. Thurmond,and T.E. Carpenter. 1996b. Effect of preweaningfeeding of pasteurized and nonpasteurized milkon postweaning weight gain of heifer calves ona Californian dairy. Prev. Vet. Med. 28:91-99.

Jasper, J., and D.M. Weary. 2002. Effects of adlibitum milk intake on dairy calves. J. DairySci. 85:3054-3058.

Jenny, B.F., H.J. Vandijk, and J.A. Collins. 1991.Performance and fecal flora of calves fed aBacillus subtilis concentrate. J. Dairy Sci.74:1968-1973.

Kaufhold, J., H.M. Hammon, and J.W. Blum.2000. Fructo-oligosaccharide supplementation:Effects on metabolic, endocrine andhematological traits in veal calves. J. Vet. Med.Assoc. 47:17-29.


24

Kertz, A.F., L.R. Prewitt, and J.P. Everett, Jr.1979. An early weaning calf program:summarization and review. J. Dairy Sci.62:1835-1843.

Klasing, K.C., and C.C. Calvert. 1999. Thecare and feeding of an immune system: ananalysis of lysine needs. Pages 253-264 inProtein Metabolism and Nutrition. G.E. Loley,A. White, and J.C. MacRae, ed. EAAP Publ. No.96, Wageningen Press, Wageningen,Netherlands.

Lammers, B.P., A.J. Heinrichs, and A. Aydin.1998. The effect of whey protein concentrate ordried skim milk in milk replacer on calfperformance and blood metabolites. J. DairySci. 81:1940-1945.

Morrill, J.L., A.D. Dayton, and R. Mickelson.1977. Cultured milk and antibiotics for youngcalves. J. Dairy Sci. 60:1105-1109.

Morrill, J.L., J.M. Morrill, A.M. Feyerherm, andJ.F. Laster. 1995. Plasma proteins and aprobiotic as ingredients in milk replacer. J. DairySci. 78:902-907.

Muscato, T.V., L.O. Tedeschi, and J.B. Russell.2002. The effect of ruminal fluid preparationson the growth and health of newborn, milk-fedcalves. J. Dairy Sci. 85:648-656.

Mylrea, P.J. 1966. Digestion in young calvesfed whole milk ad lib. and its relationship to calfscours. Res. Vet. Sci. 7:407-416.

National Animal Health Monitoring System.1993. National Dairy Heifer EvaluationProject: Dairy Herd Management PracticesFocusing on Preweaned Heifers. USDA-APHISVeterinary Services, Ft. Collins, CO.

National Animal Health Monitoring System.2002. Dairy 2002. Part 1: Reference of DairyHealth and Management in the United States.USDA-APHIS Veterinary Services, Ft. Collins,CO.

National Research Council. 2001. NutrientRequirements of Dairy Cattle. 7th rev. ed.National Academy Press, Washington, DC.

Nocek, J.E., and D.G. Braund. 1986.Performance, health, and postweaning growthon calves fed cold, acidified milk replacer adlibitum. J. Dairy Sci. 69:1871-1883.

Nonnecke, B.J., M.E. Van Amburgh, M.R.Foote, J.M. Smith, and T.H. Elsasser. 2000.Effects of dietary energy and protein on theimmunological performance of milk replacer-fed Holstein bull calves. J. Dairy Sci. 83(Suppl.1):135. (Abstr.)

Olson, E.J., W.B. Epperson, D.H. Zeman, R.Fayer, and M.B. Hildreth. 1998. Effects of anallicin-based product on cryptosporidiosis inneonatal calves. J. Am. Vet. Med. Assoc.212:987-989.

Pollock, J.M., T.G. Rowan, J.B. Dixon, and S.D.Carter. 1994. Level of nutrition and age atweaning: effects on humoral immunity in youngcalves. Br. J. Nutr. 71:239-248.

Pollock, J.M., T.G. Rowan, J.B. Dixon, S.D.Carter, D. Spiller, and H. Warenius. 1993.Alteration of cellular immune responses bynutrition and weaning in calves. Res. Vet. Sci.55:298-306.

Quigley, III, J.D., and J.K. Bernard. 1996. Milkreplacers with or without animal plasma for dairycalves. J. Dairy Sci. 79:1881-1884.


25

Quigley, III, J.D., and M.D. Drew. 2000. Effectsof oral antibiotics or IgG on survival, health andgrowth in dairy calves challenged withEscherichia coli. Food Agric. Immunol. 12:311-318.

Quigley, III, J.D., and J.J. Drewry. 1998.Nutrient and immunity transfer from cow to calfpre- and postcalving. J. Dairy Sci. 81:2779-2790.

Quigley, III, J.D., J.J. Drewry, L. M. Murray, andS.J. Ivey. 1997a. Effects of lasalocid in milkreplacer or calf starter on health and performanceof calves challenged with Eimeria species. J.Dairy Sci. 80:1751-1754.

Quigley, III, J.D., J.J. Drewry, L. M. Murray, andS.J. Ivey. 1997b. Body weight gain, feedefficiency, and fecal scores of dairy calves inresponse to galactosyl-lactose or antibiotics inmilk replacers. J. Dairy Sci. 80:1751-1754.

Quigley, III, J.D., C.J. Kost, and T.A. Wolfe.2002. Effects of spray-dried animal plasma inmilk replacers or additives containing serum andoligosaccharides on growth and health of calves.J. Dairy Sci. 85:413-421.

Reddy, P.G., J.L. Morrill, and R.A. Frey. 1987a.Vitamin E requirements of dairy calves. J. DairySci. 70:123-129.

Reddy, P.G., J.L. Morrill, H.C. Minocha, M.B.Morrill, A.D. Dayton, and R.A. Frey. 1986.Effect of supplemental vitamin E on the immunesystem of calves. J. Dairy Sci. 69:164-171.

Reddy, P.G., J.L. Morrill, H.C. Minocha, andJ.S. Stevenson. 1987b. Vitamin E isimmunostimulatory in calves. J. Dairy Sci.70:993-999.

Roy, J.H.B. 1980. Factors affectingsusceptibility of calves to disease. J. Dairy Sci.63:650-664.

Smith, J.M., M.E. Van Amburgh, M.C. Diaz,M.C. Lucy, and D.E. Bauman. 2002. Effect ofnutrient intake on the development of thesomatotropic axis and its responsiveness to GHin Holstein bull calves. J. Anim. Sci. 80:1528-1537.

Stabel, J.A. 2001. On-farm batch pasteurizationdestroys Mycobacterium paratuberculosis inwaste milk. J. Dairy Sci. 84:524-527.

Terosky, T.L., A.J. Heinrichs, and L.L. Wilson.1997. A comparison of milk protein sources indiets of calves up to eight weeks of age. J. DairySci. 80:2977-2983.

Webb, P.R., D.W. Kellogg, M.W. McGahee, andZ.B. Johnson. 1992. Addition offructooligosaccharide (FOS) and sodiumdiacetate (SD) plus decoquinate (D) to milkreplacer and starter grain fed to Holstein calves.J. Dairy Sci. 75(Suppl. 1):300. (Abstr.)

Williams, P.E.V., D. Day, A.M. Raven, and J.A.McLean. 1981. The effect of climatic housingand level of nutrition on the performance ofcalves. Anim. Prod. 32:133-141.

Williams, N.H., T.S. Stahly, and D.R.Zimmerman. 1997. Effect of level of chronicimmune system activation on the growth anddietary lysine needs of pigs fed from 6 to 112kg. J. Anim. Sci. 75:2481-2496.


26

Table 1. Effect of rate of body weight gain with constant initial body weight (100 lb) on proteinrequirements of pre-weaned dairy calves (adapted from National Research Council, 2001).1

Rate of gain ME ADP Required DMI2 CP Required (lb/day) (Mcal/day) (g/day) (lb/day) (% of DM)

0 1748 28 0.84 8.3 0.50 2296 82 1.11 18.1 1.00 3008 136 1.45 22.9 1.50 3798 189 1.83 25.3 2.00 4643 243 2.24 26.6 2.50 5532 297 2.67 27.2

1ME = metabolizable energy, ADP = apparently digestible protein, DMI = dry matter intake, CP = crudeprotein, and DM = dry matter.2Amount of milk replacer DM containing 2075 kcal ME/lb DM needed to meet ME requirements.


27

Effects of Mycotoxins on Ruminal Bacteria and Animal Performance

John A. Doerr1

College of Agriculture and Natural ResourcesUniversity of Maryland

Abstract

Dairy producers have a difficult taskwhen attempting to assess the possible exposureof their animals to mycotoxins. Informationgleaned from the scientific literature suggestssigns and symptoms for toxins which mayactually not be likely to appear in a herd. Also,that same literature suggests concentrations atwhich the toxin exerts its effects. Thoseconcentrations may have little meaning in aproduction setting in which multiplemycotoxins, as well as other production stresses,are at work. We do know, however, thatmycotoxins do pass through the dairy cow’srumen intact, although clearly some microbialmetabolism can and does take place. At the sametime, the mycotoxins may produce adverseaffects on key rumen bacteria. Therefore, wecannot count on rumen microbiology to protectthe dairy animal. Predicting signs which mightoccur in the herd exposed to mycotoxins is alsofutile. Such signs are generally not unique toparticular mycotoxins if to toxic compounds atall. Changes in productivity, in reproductivecapacity, in general health, behavior, etc. mayall signal the presence of mycotoxins, althoughother causes may be equally to blame. Thepresent report discusses some of these factorsand suggests the approach dairy farmers mayfind most productive in dealing with the issueof mycotoxin contamination on their farms.

Introduction

Case study reports and research trials canprovide both valuable and misleading cluesabout the effects of mycotoxins in ruminants. Aherd experienced a decrease in breedingefficiency over an extended period (~ fivemonths), with reduced birth weights, unhealthycalves, increased mastitis, prolapsed rectum,decreased feed consumption, and other signs.Corn contaminated with aflatoxin at up to120ppb had been included in the ration. Once thecontaminated material was found and removed,milk production rose by 28% (Guthrie, 1979).Choudary et al. (1998) reported that as little as10 ppb of aflatoxin B1 (AFB) resulted in asignificant reduction in feed consumption. Alsoresponding in a dose-dependent fashion, rumenmotility declined in dairy cows receiving 200 to800 ppb AFB (Cook et al., 1986). Of course,many producers throughout the U.S. and otherregulated production areas understand the lossthat can occur when violative levels of aflatoxinM1 are discovered in milk.

Calves consuming the cyto-necrotic T-2toxin (at 10 ppm and above) developed ulcersof the abomasum and sloughing of rumenpapillae (Cheeke, 1998). Goats fed 95 ppmfumonisin showed no overt signs of toxicosisalthough sphingolipid tests revealed toxin-related pathology (Gurung, et al. 1998), but dairycows fed that toxin did not present eithercirculating fumonisin or alterations in

1Contact at: 0107 Symons Hall, University of Maryland, College Park, MD 20742-5551, (301) 405-7761, FAX (301) 405-8570, Email: [email protected]


28

sphinganine/sphingosine ratios (Prelusky et al.,1995). When fumonisin was administered toHolstein calves, liver and renal toxicity withmultiple signs resulted (Mathur et al., 2001);however, these were young calves that had notyet developed rumen function. So, are fungaltoxins a threat to dairy operations?

Many have suggested that the bacteriaof the bovine rumen are exceptionally efficientat degrading dietary mycotoxins. Oncedestroyed, therefore, there is no opportunity forthese compounds to be absorbed and exert theirtoxic roles. Swanson et al. (1987) demonstratedthat when DAS (diacetoxyscirpenol), DON(deoxynivalenol), and T-2 toxin were incubatedwith rumen fluid, all three were rapidlymetabolized by the microflora. However, endproducts of that ‘degradation’ process wereidentified. Many of them, such asmonoacetoxyscirpenol (MAS) are also toxiccompounds. Earlier, Kiessling et al. (1984) hadreported that rumen contents (fluid andmicroflora) would metabolize ochratoxin,zearlanenone, DAS, and T-2 toxin but not DONor AFB (in vitro).

Dairy farmers, therefore, face somesignificant dilemmas. First, some of the reportedsigns are relatively non-descript and do notconsistently pertain to specific mycotoxinexposures. Diarrhea, lowered feed consumption,mastitis, estrus errors, etc. may all be part ofone or more mycotoxins’ impact on the dairyanimal; however, none are unique to fungaltoxins and most may occur more often as a resultof other factors. Similarly, if one takes theresearch on ruminal degradation of mycotoxinsat first blush, it would seem that ruminants areuniquely protected against these fungalmetabolites. How, then, can it be that staggers,slobbers, abortion, loss of production,hypothermia, etc. can all be induced by a dietaryapplication of certain toxins at levels consistentwith what is found naturally under farmconditions?

Factors Associated with Rumen Function

Mycotoxin metabolism by ruminal bacteria

The literature on microbial degradationof various mycotoxins by bacteria is abundant,and begins, generally, with evaluation ofcommon soil microorganisms. However, ofinterest are those studies dealing with eithermixtures or isolated species of organisms takenfrom biological sources. Galtier and Alvinerie(1976) isolated cecal contents from rats anddemonstrated that ochratoxin A could bedegraded to the non-toxic alpha form by themicroflora but not by the supernatant fluid of acentrifigal separation. Thus, the live bacteriawere considered responsible for the activity.They also showed that bovine or ovine rumenfluid metabolized ochratoxin A but at timesesterified it to ochratoxin C which is just as toxicas A. The rumen fluid, however, was obtainedat a slaughter house; fluid obtained viaesophagus did not convert ochratoxin A to C anddid hydrolize it to the alpha form (Galtier andAlvinerie, 1976). They suggested that this effectwas mediated by protozoa rather than bacteria.This also, then, suggests that in vitro studies maybe influenced greatly by the means by whichruminal fluid/contents are obtained. Post-mortem versus esophageal aspirates versusfistula samples must be considered very differenttest materials. He et al. (1992) suggested fromtheir studies with soil, chicken and swineintestinal, and rumen fluid bacteria that chemicalmodification of DON was influenced by severalconditions. Thirty-five percent of the originalDON was metabolized by rumen fluid; however,lowering the pH of the test medium fullyinhibited this metabolism. Further, theydetermined that while deepoxy DON (the de-epoxidation product) was the major metabolite,it was highly dependent on energy availability(He et al., 1992). Does this have implicationsfor studies in which rumen fluid and microflorahave been separated from ingesta before beingused for metabolism tests?


29

Xiao et al. (1991) measured hydrolysisof ochratoxin A by ovine rumen. Both in vitro(hydrolysis product) and in vivo (disappearancehalf-life) metabolism of ochratoxin A were foundto be influenced by the diet of the sheep. Grain-fed sheep did not metabolize the toxin nearly sowell as those fed hay. Their explanation wassimple and logical: differences in diet result indifferences in the microbial profile of the rumen.Thus, the question of whether a toxin may bedegraded or not, much less the apparent rate orefficiency of that degradation, is subject to thespecifics of the microflora resident at the timethe rumen material is collected. In turn, that isstrongly driven by what the donor animal hasconsumed prior to that sampling. Muller et al.(2001) have recently confirmed this effect usingcows fed diets with grass, grass silage or hay,and different amounts of a concentrate (barley/soymeal). In addition to dietary effects, they alsofound that the animal, itself, affected the results.

Many studies could be cited. However,the essential points are that 1) many bacteria andother microorganisms are capable of assimilatingand metabolizing mycotoxins (or otherxenobiotics). These include species found inthe normal, functional microbiota of the bovinerumen. At the same time, the relative efficienciesof such organisms to achieve full success undera wide ranging variety of conditions andcircumstances are very unclear. It is importantnot to assume that metabolism of a compoundmeans that all of that compound is nowaccounted for. Many reports indicate that onlysome percentage of the total compound testedis really changed. 2) Bacteria found tometabolize a foreign compound under one setof experimental conditions may not do so underanother. 3) In vitro is just that! By its verynature, the test is artificial and subject to manyinfluences not accounted for in all studies. Thatis not to say that in vitro work is not a goodidea; much useful information is provided, butdirect extrapolation to the effects that mightoccur in the intact rumen of a production unitanimal are, at best, potentially wrong, and, at

worst, dangerous to the extent that othersinterpret such results as real.

Mycotoxin effects on ruminal bacteria

While it is important to understand thatat least some change in ingested mycotoxins isprobable in the rumen, it is of equal importanceto understand that the reverse is true. Rumenmicroorganisms can be adversely impacted bymycotoxins. Periodically, there is interest in thecondition known as salivary syndrome (orslobbers) caused by infestation of clover byRhizoctonia leguminicola and its resultant toxin,slaframine. Froetschel et al. (1986) showed thatrumen contractions decreased from 20 to 78%with the introduction of slaframine into therumen of sheep and cows. Later, the same group(Freotschel et al., 1987) reported that totalruminal fluid volume and outflow increased by25% or more in response to slaframine. Ruminaldigestion of DM, ADF, and starch declined whilepostruminal digestion of those same componentsincreased (Froetschel et al., 1989). Bird et al.(1993), working with sheep, obtained similarresults with respect to slaframine-induced fluidvolumes and reported an increase in cellulolyticbacteria; however, their study had the additionalfactors of diet variables (low-quality hay withor without cottonseed meal) and found the mealsupplementation essential to counter-act weightloss associated with the toxin.

Escoula (1992) found that patulin fromlaboratory cultures reduced acetate productionwhen added to in vitro rumen fluid. Further,protein synthesis was inhibited in that samesystem by the toxin in a dose-dependent fashion.Patulin is most often associated with bruisedfruits, especially apples; therefore, it is unlikelyto be of major significance to most dairy herdsexcept in areas where apple pressing residues(e.g., for fruit wines or ciders) are used in animalrations. Earlier, reference was made to Cook’sstudy (Cook et al., 1986) involving graded dosesof AFB and slowing of rumen motility.However, they also noted that because of that


30

change in motility, the time required to eliminateAFB from rumen contents increased, a featurethey found helpful as a diagnostic aid. However,it must also be considered that two additionalcircumstances may occur because of this longerin rumen residence time for the xenobiotic. Onthe one hand, there is greater opportunity forcapable bacteria to metabolize more of the toxin.On the other hand, however, that same time mayallow AFB or other toxins to have a greaternegative impact on specific members of therumen microflora.

Arguably, dairy cows consume moreFusarium toxins than aflatoxins. Recently, Mayet al. (2000) published data indicating thatfusaric acid was inhibitory to two differentrumen organisms, Methanobrevibacterruminantium (an end-of-fermentation methanegenerator) and Ruminococcus albus (an onset-of-fermentation cellulolytic bacterium). Lowlevels of fusaric acid (e.g., ~7 ppm) weresufficient to affect microbial growth; at highlevels (e.g., ~ 60 ppm), growth stoppedaltogether. Further, upon removal of the toxin,neither organism recovered. These two bacteriarepresent quite different evolutionary positions.R. albus is a true prokaryote; M. ruminantiumhas both pro- and eucaryotic biochemistry. Forfusaric acid to inhibit both suggests that it maywork in more than one way mechanistically; thatis, it expressed its toxicity in one by a routedifferent than in the other. If this is proven true,it presents a new dilemma in that multiple modesof actions increase the likelihood and avenuesby which this toxin might interact with othermycotoxins. Addition of DON to tests in thiswork (May et al., 2000), if anything, spared thefusaric acid effects on the two bacteria. Can wepredict a similar response if zearalenone, orAFB, or slaframine is present?

Mycotoxin Effects on Dairy Cattle

Field cases generally present with non-descript signs, including reduced feed

consumption and milk production, lower fat and/or protein in milk, reproductive ‘errors’ (alteredcycles, etc.), and to more dire conditionsincluding abortion, morbidity, and death. Asfarmers note the onset of such signs, there isgenerally a systematic attempt to resolve theunderlying cause(s) and determine a solution.In acute mycotoxicotic episodes, by the time theveterinarian, nutritionist, breeder, etc. havecompleted examination, the particular feedstuffresponsible may well be gone. More typically,however, chronic exposure to multiplemycotoxins underlies the problem with largeanimals. Not too many years ago, I might havelooked first at grain supplements, cottonseedmeal, and other by-products used in feeding asthe first choice of possible contaminatedproducts. That, of course, stemmed from thecommon misconception that since ensiledmaterials are in an anaerobic state, they cannotsupport active mold growth (or mycotoxinproduction). However, silage is most often thefirst place we should look.

Points to keep in mind about silage aresimply these: 1) structures in which ensilingtakes place and in which the product is storeduntil fed are not necessarily as air-tight as wewould like to suppose. 2) Some of themycotoxigenic molds are micro-aerophilic; thatis, they grow under very low oxygen tension.Ensiling may alter conditions to reduce somegenera (e.g., Aspergillus) but produce acompetitive advantage for others (e.g.,Penicillium). 3) Some molds, again often in thegenus Penicillium, actively metabolize lowmolecular weight organic acids (e.g., propionicacid) that may be used initially to defeat moldsuntil anaerobiosis is achieved or that may formas a result of bacterial fermentation action duringensilation (e.g., acetic, et al.). 4) The mechanicalaction of cutting/removing silage for feeding canintroduce fresh air into the next ‘segment’ ofsilage. For toxigenic molds that are inactiveprior to that new infusion of oxygen, the responsetends to be rapid – significant amounts of toxins


31

may be synthesized within 24 to 48 hours. Atthe next feeding, cows are presented with newlycontaminated silage.

Analysis of silage is not an easy task.Obtaining representation in the sample andappropriately preparing the matrix for extractionare both major hurdles for the dairy farmer thatother animal producers (e.g., poultry) do notface. The coarse, highly heterogeneous natureof silage requires special preparation for testing.Molds, while often seen as surface contaminants,tend not to be surface metabolizers in feedstuffs.The corn kernel represents a good example. Weoften see mold spores or even visible mold onthe surface; however, few molds gain much fromthe pericarp. Instead, they penetrate andmetabolize the starch portion of the kernel. Ifone tries to extract such corn with a relativelybenign solvent system, such as aqueousmethanol (which is the extraction solventrequired for most rapid antibody-based tests),much of the mycotoxin will never be extractedsince the waxy exterior of the kernel retardspenetration by the solvent. In order for the testto work, kernels must be ground to a very fineparticle size (generally small enough to pass a20-mesh sieve). One can make the sameprovisions for silage, except with much greaterdifficulty than with corn. Since virtually all ofthe toxins of importance are relatively heatstable, silage samples can be dried and thenground through a conventional laboratorymilling system. However, for those attemptingon-farm analyses, maintaining a heavy dutyblender with a sharp blade assembly can allowone to comminute the sample within theextraction solvent system. The key to successhere, which means getting sufficient amounts ofanalytes extracted that some reasonable estimateof actual condition of the feedstuff is possible,is processing an adequate amount of sample inenough solvent to do the job. That may wellmean ‘scaling up’ from what a company sellingrapid tests suggests.

Farmers reporting signs and havinggotten beyond the issues of infectious disease,nutritional error, etc. are then faced with how tointerpret a sample analysis. A typical result mayshow a silage sample containing 600 ppbzearalenone, 1.8 ppm DON, 26 ppb aflatoxin,positive for fumonisin, and perhaps a little DASin the mix for good measure! Cows are off feed,milk production is down, and somatic cell countsare up. Which sign goes with which mycotoxin?Or do any of the problems match any of thetoxins at all?

This report opened with the concept thatboth valuable and misleading informationabounds. Research reports on controlled singlemycotoxin experiments in a few animalsgenerally do a very good job with the descriptivetoxicology of a particular mycotoxin under aspecial set of experimental conditions. One canlearn valuable information about how the toxinworks and what may be the results of exposureto that toxin. At the same time, the test animalsin most research trials are relatively protectedin that an investigator is trying to isolate theirtreatments as the sole causes of ultimate results.So, the kinds of management, nutritional,production, behavioral, environmental, etc.pressures faced by the animal on the farm areless a factor. Yet, these all impact how thatanimal responds to a single mycotoxin. The realproblem is what we find in that typical casementioned just above - multiple toxins. Howdo we interpret such a result?

First, we must allow for the fact thatmany of our tests are precise but not necessarilyaccurate in the field. So, the absolute values inppm or ppb provide a frame of reference but nota number that we should be overly enthusedabout. Second, the finding of one or more toxinstells us that conditions have favored the activegrowth and metabolism of toxigenic molds.Since there are hundreds of species and severalhundred different mycotoxins, our positive testsshould alert us to the real possibility that ourfeedstuffs contain much more than those four


32

or five test results tell us. Third, the toxins weselect for testing are often those which, whileimportant themselves, are relatively easy to assayand may reflect a “family” of related chemicalcompounds. Thus, measuring DON not onlytells us about toxin specifically, but also indicatesthat we should expect other mycotoxins of thetrichothecene class to be present. Finally, andmost importantly, the finding of more than onetoxin opens the sure possibility that chemicalinteractions are at work.

The idea that mycotoxins can potentiateor antagonize the action of other mycotoxins isno longer a supposition. It has been provenextensively in the literature. However, even withthose kinds of scientific studies, we have onlylooked at a limited number of interactioncombinations (e.g., two or maybe threemycotoxins at a time). We have little ability topredict what might happen when four or five ormore toxins are present together in the animal’sration. To further complicate this issue, we noware beginning to understand that a given animalmay react to two mycotoxins quite differentlydepending on which toxin gets in first!Gelderblom et al. (2002) have studied thisphenomena in rat liver using fumonisin andaflatoxin. Ultimate pathology was shown todepend on whether the samples were treated withfumonisin first or second. In short, if aflatoxingot to the liver first and caused damage, thenthe response to the fumonsin changed quitedrastically compared to liver treated withfumonisin first (Gelderblom et al., 2002).

Finally, low level chronic exposure tomycotoxins may fail to induce even a single signthat can truly be associated with any givenmycotoxin. However, almost universally,mycotoxins impact immune function in animals.Chronic exposure, then, often leads to anincrease in opportunistic pathogens gaining afoothold in the animal. Subsequently, we findsymptoms associated with infectious diseaserather than with an intoxicant. Predicting specificsigns, therefore, becomes relatively futile. The

best advice to producers, therefore, is to considerthat whenever production and health decline ina herd, mycotoxins should be among the factorsconsidered, and steps should be taken to reduceor eliminate those toxins.

References

Bird, A.R., W.J. Croom, Jr., J.V. Bailey, B.M.O’Sullivan, W.M Hagler, Jr., G.L. Gordon, andP.R. Martin. 1993. Tropical pasture hayutilization with slaframine and cottonseed meal:ruminal characteristics and digesta passage inwethers. J. Anim. Sci. 71:1634-1640.

Cheeke, P.R. 1998. Mycotoxins in cereal grainsand supplements. In Natural Toxicants in Feeds,Forages, and Poisonous Plants (P.R. Cheeke,Ed.). Interstate Publishers, Inc. Danville, IL, pp.243-274.

Choudhary, P.L., R.S. Sharma, V.N. Borkhataria,and M.C. Desai. 1998. Effect of feedingaflatoxin B1 on feed consumptions throughnaturally contaminated feeds. Ind. J. Anim. Sci.68:400-401.

Cook, W.O., J.L. Richard, G.D. Osweiller, andD.W. Trampel. 1986. Clinical and pathologicchanges in acute bovine aflatoxicosis: rumenmotility and tissue and fluid concentrations ofaflatoxins B1 and M1. Am . J. Vet. Res. 47:1817-1825.

Escoula, L. 1992. Patulin production byPenicillium granulatum and inhibition ofruminal flora. J. Environ. Pathol. Toxicol. Oncol.11:45-48.

Froetschel, M.A., H.E. Amos, J.J. Evans, W.J.Croom, Jr., and W.M. Hagler, Jr.,. 1989. Effectsof a salivary stimulant, slaframine, on ruminalfermentation, bacterial protein synthesis anddigestion in frequently fed steers. J. Anim. Sci.67: 827-834.


33

Froetschel, M.A., W.J. Croom, Jr., W.M. Hagler,Jr., R. Argenzio, J. Liacos, and H.P. Broquist.1987. Effects of slaframine on ruminantdigestive function: liquid turnover rate andfermentation patterns in sheep and cattle. J.Anim. Sci. 64:1241-1248.

Froetschel, M.A., W.J. Croom, Jr., W.M. Hagler,Jr., R. Argenzio, J. Liacos, and H.P. Broquist.1986. Effects of slaframine on ruminantdigestive function: ruminal motility in sheep andcattle. J. Anim. Sci. 63:1502-1508.

Galtier, P. and M. Alvinerie. 1976. In vitrotransformation of ochratoxin A by animalmicrobial floras. Ann. Rech. Vet. 7:91-98.

Gelderblom, W.C., W.F. Marasas, S. Lebepe-Mazur, S. Swanevelder, C.J. Vessey, and Pde L.Hall. 2002. Interaction of fumonisin B(1) andaflatoxin B(1) in a short-term carcinogenesismodel in rat liver. Toxicol. 171:161-173.

Gurung, N.K., D.L. Rankins, Jr., R.A. Shelby,and S. Goel. 1998. Effect of fumonisin B1-contaminated feeds on weaning angora goats.J. Anim. Sci. 76:2863-2870.

Guthrie, L.D. 1979. Effects of aflatoxin in cornon production and reproduction in dairy cattle.J. Dairy Sci. 62:134-136.

He, P., L.G. Young, and C. Forsberg. 1992.Microbial transformation of deoxynivalenol(vomitoxin). Appl. Environ. Microbiol. 58:3857-3863.

Kiessling, K.H., H. Pettersson, K. Sandholm,and M. Olsen. 1984. Metabolism of aflatoxin,ochratoxin, zearalenone, and threetrichothecenes by intact rumen fluid, rumenprotozoa, and rumen bacteria. Appl. Environ.Microbiol. 47:1070-1073.

Mathur, S., P.D. Constable, R.M. Eppley, A.L.Waggoner, M.E. Tumbleson, and W.M. Haschek.2001. Fumonisin B(1) is hepatotoxic andneprotoxic in milk-fed calves. Toxicol. Sci.60:385-396.

May, H.D., Q. Wu, and C.K. Blake. 2000.Effects of the Fusarium spp. Mycotoxins fusaricacid and deoxynivalenol on the growth ofRuminococcus albus and Methanobrevibacterruminantium. Can. J. Microbiol. 46:692-699.

Muller, H.M., K. Muller, and H. Steingass. 2001.Effect of feeding regime on the metabolism ofochratoxin A during the in vitro incubation inbuffered rumen fluid from cows. Arch.Tierernahr. 54:265-279.

Prelusky, D.B., M.E. Savard, and H.L. Trenholm.1995. Pilot study on the plasmapharmacokinetics of fumonisin B1 in cowsfollowing a single dose by oral gavage orintravenous administration. Nat. Toxins 3:389-394.

Swanson, S.P., J. Nicoletti, H.D. Rood, Jr., W.B.Buck, L.M. Cote, and T. Yoshizawa. 1987.Metabolism of three trichothecene mycotoxins,T-2 toxin, diacetoxyscirpenol anddeoxynivalenol, by bovine rumenmicroorganisms. J. Chromatogr. 414:335-342.

Xiao, H., R.R. Marquardt, A.A. Frohlich, G.D.Phillips, and T.G. Vitti. 1991. Effect of a hay anda grain diet on the rate of hydrolysis ofochratoxin A in the rumen of sheep. J. Anim.Sci. 69:3706-3714.


34


35

Abstract

Prostaglandins are active substancessynthesized from dietary fat and involved inreproduction. There are two main pathways usedto synthesize prostaglandins. One is used bymost dietary fat (e. g. corn and soybean, sourcesof omega-6 fatty acids) and leads to series 1 and2 prostaglandins, while the other one is morespecific to fish products and flax oil (sources ofomega-3 fatty acids) and leads to series 3prostaglandins. Thus, depending on the pathwayused for prostaglandins synthesis, the type androle of the resulting prostaglandins will differ.Fatty acid composition of the cell wallmembrane is modified by dietary fatty acids,which would alter the function of reproductivetissues. This suggests that fatty acids fromdifferent metabolic pathways have differenteffects on reproduction of the dairy cow. In fact,dietary omega-3 fatty acids can decrease series2 prostaglandins synthesis by different actions,which include decreasing the availability of theprecursor arachidonic acid, increasing theconcentration of fatty acids that compete witharachidonic acid for series 2 prostaglandins, andinhibiting prostaglandins synthase. Lowersecretion of series 2 prostaglandins usuallyimproves reproduction of cattle. This can beachieved through dietary supplementation withfish oil or linseed oil as they are major inhibitorsof desaturation and elongation, leading toarachidonic acid formation.

Effects of Dietary Fat on Reproduction

Hélène V. Petit1

Dairy and Swine Research and DevelopmentCentre Agriculture and Agri-Food Canada

Introduction

Recently, there has been a great deal ofinterest in feeding fat to dairy cows in order toincrease energy density of the diet and improvereproduction. It is known that cows fedsupplemental fat may experience improvedenergy balance and begin to cycle sooner becauseof enhanced follicular growth and development(Grummer and Carroll, 1991). However, Lucyet al. (1992) suggested that it was fatty acids,and not the additional energy provided by thefatty acids, that stimulated ovarian function.Recently, new information has been publishedthat demonstrates that the type of dietary fattyacids is important, as individual fatty acids donot have the same effects on reproduction of thedairy cow.

Fatty Acid Terminology

A fatty acid molecule is shaped like acaterpillar with two different ends: a methylgroup and a water-soluble end that is thecarboxyl end. There are different families of fattyacids in feed: omega-3, omega-6, omega-7, andomega-9. The most common numbering systemis called the omega system. This system numberscarbon atoms in sequence, starting from themethyl end. The other commonly used system,called the delta (d) system, starts at the acid endand numbers the carbon atoms in reversedirection.

1Contact at: P.O. Box 90, Lennoxville, QC J1M 1Z3,Canada, (819) 565-9171, Ext. 210, FAX (819) 564-5507, Email:[email protected]


36

The omega-7 family of fatty acids issynthetised from palmitic acid (C16:0), whilethe omega-9 fatty acid family is synthetised fromstearic acid (C18:0) via oleic acid (C18:1, Figure1). These two families are not consideredessential as they are produced in the body. Theomega-3 and omega-6 fatty acids are essentialbecause both are vital to health but cannot bemade by our cells and must, therefore, beprovided by foods.

Linoleic acid (C18:2) belongs to theomega-6 family, while linolenic acid (C18:3)belongs to the omega-3 family (Figure 2). Thesystem used to name fatty acids considers thenumber of carbons in the chain (e.g. 18 forlinoleic acid), the number of double bonds inthe chain (2 for linoleic acid), and where in thechain the first double bond is located from themethyl end (1st double bond between carbons 6and 7 for linoleic acid): C18:2.

Sources of Fatty Acids

The main sources of short chain fattyacids are cottonseed and palm oils. All sourcesof fat contain long chain fatty acids. The mainsources of linolenic acid (C18:3T3) are flaxseed,hemp, canola, soybean, nuts, and dark greenforages. Ryegrass silage contains as much as60% of linolenic acid as a percentage of totalfatty acids (Dewhurst and King, 1998), whichwould encourage the use of high forage systemsto increase dietary linolenic acid content.Omega-3 fatty acids are found also in cold waterand salt water fish (salmon, trout, makerel, andsardines). The main sources of linoleic acid(C18:2ω6) are sunflower seed, safflower, hemp,soybean, nuts, pumpkin seeds, sesame seeds, andflaxseed. Gamma-linolenic acid (C18:3ω6) is foundin evening primose oil, grape seeds, and borage.Dihomogamma-linolenic acid (C20:3ω6) is foundin maternal milk, while arachidonic acid (C20:4ω6)occurs mainly in meat and animal products. Oleicacid (C18:1) is found in olive, almond, avocado,peanut, pecan, cashew, macadamia nut, and butter.Omega 7 in the form of palmitoleic acid (C16:1) is

found in tropical oils (coconut and palm).Composition in C18 fatty acids of some ediblevegetable oils is reported in Table 1.

Fatty Acids and Fertility

Supplementary fats are likely to affectfertility because fatty acids are the precursorsboth of prostaglandins (PG) and, via cholesterol,the steroid hormones. In general, feedingsupplemental fat, such as calcium soaps of longchain fatty acids, fish meal, and tallow, increasesconception rates. However, a loweredconception rate at first service has been reportedwhen there was a paralleled increase in milkproduction (range of 4.8 to 9.9 lb/day; 2.2 to 4.5kg/day). Thatcher and Staples (2000) wrote anexcellent review on the subject. There are twomain families of essential fatty acids, omega-3and omega-6 fatty acids, that could affectfertility. The main source of omega-6 fatty acidsis dietary linoleic acid (C18:2n-6) and this isconverted to arachidonic acid (C20:4n-6), whichinter alia is the precursor of the dienoic (2 series)PG, such as PGF2α. The same elongase anddesaturase enzymes also convert the main dietaryomega-3 fatty acids (α-linolenic acid; C18:3n-3) to eicosapentaenoic acid (EPA; C20:5n-3),the precursor of the trienoic (3-series) PG, suchas PGF3α (Abayasekara and Wathes, 1999).Competition between omega-3 and omega-6precursors for desaturation and elongation aswell as at the site of PG synthetase means thatincreasing the supply of omega-3 fatty acids willdecrease production of dienoic PG (Barnouinand Chassagne, 1991). In many cases, thetrienoic PG have lower biological activity thanthe corresponding dienoic PG (Fly and Johnston,1990), and this may directly affect aspects offertility. For example, treatments that reduceovarian and endometrial synthesis of PGF2α, atthe expense of PGF3α, may contribute to areduction in embryonic mortality (Mattos et al.,2000). There is some evidence for differenteffects of α-linolenic acid and the omega-3 fattyacids from fish oil (EPA and docosahexaenoicacid (DHA), C22 :6n-3) on eicosanoid (interleukin)


37

synthesis, perhaps because of differences in the wayin which these fatty acids are incorporated into cellmembranes (Wu et al., 1996).

Supplementary fats can also reduce the totalsynthesis of PG by affecting the activity of PGsynthase (Thatcher et al., 1995). Diets rich inlinoleic acid (C18:2) increase arachidonic acidconcentration (C20:4) in tissues, and diets richin linolenic acid (C18:3) increase concentrationof eicosapentaenoic acid (C20:5) (Béréziat,1978). Moreover, eicosapentaenoic acid (C20:5)is a competitive inhibitor of the enzyme complexinvolved in the synthesis of PG from arachidonicacid (C20:4) (Leat and Northrop, 1979).Therefore, this would suggest that a diet with alow linoleic to linolenic acid ratio (C18:2:C18:3,omega-6:omega-3) could decrease PG secretionor PG activity as suggested by Barnouin andChassagne (1991), which would thus haveimportant effects on reproduction and immunityin the dairy cow.

Prostaglandin Synthesis

There are two main pathways used tosynthesize PG (Figure 3); one is used by mostdietary fat (e.g. corn and soybean, sources ofomega-6 fatty acids) and leads to series 1 and 2PG, while the other one is more specific to fishproducts and flax (sources of omega-3 fattyacids) and leads to series 3 PG. Thus, dependingon the pathway used for PG synthesis, the typeand role of the resulting PG will differ. The PGof series 2 are important at calving; they increaseplatelet agglutination and blood clot formation,and they increase salt retention in the kidneys,water retention, and blood pressure. The PG ofseries 2 also cause inflammation, which lead totheir role of “bad guys” among the different PGseries. The PG of series 1 improve the immunesystem of T cells, prevent platelet agglutinationand heart attack, contribute to removal of excessNa and water by the kidneys, decrease theinflammatory response, contribute in controllingarthritis, and contribute to decreasing cholesterolproduction. The PG of the series 3 have a very

weak platelet agglutination power and they preventfabrication of PG of the series 2; they also preventheart attack, water retention, and inflammation. ThePG of the series 1 and 3 are thus considered as“good guys” contrary to those of the series 2.

Some polyunsaturated fatty acids(PUFA) can serve as a substrate for the synthesisof PGF2α. These include cis-linoleic acid (C18:2)that is commonly found in natural fat sources. Itcan be desaturated and elongated to formarachidonic acid, which serves as an immediateprecursor for the series 2 PG of which PGF2α isa key member. Key regulatory enzymes for theseconversions include ∆6 desaturase andcyclooxygenase. Fatty acids also can inhibit PGsynthesis by competitive inhibition with thesekey enzymes. The EPA and docosahexanoic acid(C22:6) have been shown to inhibitcyclooxygenase activity, which is an enzymeinvolved in the synthesis of PGF2α.

Fatty Acids, Cholesterol, and Progesterone

Cholesterol serves as a precursor for thesynthesis of progesterone by ovarian luteal cells.Secretion of progesterone is the main functionof the corpus luteum. Progesterone not onlyprepares the uterus for implantation of theembryo but also helps maintain pregnancy byproviding nourishment to the conceptus. Thesuccessful establishment and maintenance ofpregnancy (before day 16 post AI) requires themaintenance of progesterone secretion throughthe critical period of the maternal recognitionof pregnancy when luteolysis occurs in the non-pregnant animal (Lamming and Royal, 2001).Between 25 and 55% of mammalian embryosdie in early gestation. Increased concentrationsof plasma progesterone have been associatedwith improved conception rates of lactatingruminants. Similarly, progesterone concentrationprior to AI has been associated with greaterfertility. In a field study involving 426 lactatingdairy cows, blood was sampled at 58 dayspostpartum for multiparous cows and 72 days forprimiparous cows and then analyzed for


38

progesterone. Cows were bred approximately 3days later in a synchronized estrus scheme.Conception rate increased 1.44% for every 1 ng/ml increase in plasma progesterone (r2 = 0.11,Staples et al., 1997). The recovery of embryos 7days after estrus increased as plasmaprogesterone concentration increased just priorto AI (Britt et al., 1996). In either association,dietary fat, which stimulates ovarian cyclicityor corpus luteum function, would contribute toincreased fertility. Increased progesteronesuggests that luteal function is enhanced bydietary fat. Dynamics of maternal progesteronesecretion also appear important for conceptusdevelopment and secretion of interferon-τ, whichis secreted by the embryo for gestationrecognition by the mother.

It has been suggested that improvedconception rate could be a result of increasedconcentrations in plasma cholesterol (Spicer etal., 1993), although this hypothesis was notsupported by our results. In fact, cows fedformaldehyde-treated flaxseed had lower plasmacholesterol concentration and better conceptionrates than those fed Megalac® (Church andDwight Co., Inc., Princeton, NJ) (Petit et al.,2001). Other studies have reported norelationship between cholesterol concentrationsin blood and reproductive measures (Fergusonet al., 1990).

The fatty acid profile of the dietary fatmay influence the propensity of animals toincrease plasma progesterone. Mature ewes wereinfused intravenously with saline, soybean oil,or olive oil for 5 hours on days 9 through 13 ofan estrous cycle (Burke et al., 1996). Serumcholesterol was increased by fat infusates, andolive oil was more effective than soybean oil(127, 141, and 153 mg/dl of serum cholesterolfor saline, soybean oil, and olive oil,respectively). However, soybean oil infusionresulted in greater progesterone response thandid infusion of olive oil at 2.5 hours postinfusion.Therefore, the greatest concentration of serum

cholesterol did not coincide with the greatestconcentration of serum progesterone.

Fatty Acids and Prostaglandin Secretion

It is known that there is a negativerelationship between concentration of PGF2α andthat of progesterone. For example, at calving,PGF2α concentration increases while that ofprogesterone decreases. Similarly, duringgestation, PGF2α concentration decreases andthat of progesterone increases. Progesterone issecreted by the corpus luteum and synthesizedby steroids. Therefore, an increase in PGF2αconcentration is paralleled with a decrease inprogesterone concentration and vice versa. Intheory, it could thus be possible to modulateconcentrations of PGF2α and progesterone bydifferent feeding strategies! In fact, in theexperiment we carried out in the UK, weobserved a tendency (P = 0.09) for greaterprogesterone concentration in the blood of cowsfed formaldehyde-treated flax compared to thosefed Megalac® (Petit et al., 2002). This may partlyexplain the greater gestation rate observed forcows fed formaldehyde-treated flax (87.5%)compared to those fed Megalac® (50.0%) in acompanion study (Petit et al., 2001).

Better conception rate for cows fedformaldehyde-treated flaxseed compared tothose fed Megalac® could result from differentPG synthesis. In fact, linolenic acid in flaxseeduses the eicosapentaenoic acid metabolicpathway, while fatty acids in Megalac® partlyuses the arachidonic acid pathway, (Cunnane,1995) and it is known that eicosapentaenoic acidinhibits PG synthesis (Spicer et al., 1993).Therefore, ingestion of linolenic acid containedin flaxseed could potentially inhibit PGF2"synthesis (Cunnane, 1995). Thatcher et al. (1997)has shown that PGF2α secretion is decreased indairy cows fed fish meal, which is a source ofomega 3 fatty acids. Moreover, feeding dietscontaining 2.6, 5.2, and 7.8% Menhaden fish mealto lactating dairy cows reduced uterine secretion ofPGF2α (Thatcher et al., 2001a). In fact, fish meal,


39

which would lead to eicosapentaenoic acid anddocosahexaenoic acid formation, has been shownto increase gestation rate of dairy cows and to altercorpus luteum regression as shown by greaterplasma concentrations of progesterone (Burke etal., 1997). This would agree with the tendencyobserved in one of our experiments (Petit et al.,2002) for greater milk progesteroneconcentration, expressed as the area under thecurve, for cows fed formaldehyde-treatedflaxseed compared to those fed Megalac®.However, it is not known if the greaterconception rate observed for cows fedformaldehyde-treated flaxseed in the experimentof Petit et al. (2001) was a result of a decreasein embryo mortality or better fertilization of theova as pregnancy was confirmed only once atday 45 post AI. More research is required todetermine the reasons for better conception ratefor cows fed a source rich in omega-3 fatty acids.The potential to improve reproduction of dairycows through dietary manipulation is an excitingconcept and needs to be further addressed.

Dietary PUFA can decrease PGF2αsynthesis by different actions, which includedecreasing the availability of the precursorarachidonic acid, increasing the concentrationof fatty acids that compete with arachidonic acidfor series 2 PG, and inhibiting PG synthase.Reduced availability of arachidonic acid in theuterine phospholipid membranes for conversionto series 2 PG can occur through a reduction inthe synthesis of arachidonic acid or throughdisplacement of existent arachidonic acid fromthe phospholipid membranes by other fatty acids.This can be achieved through dietarysupplementation with fish oil (rich in EPA andDHA) or linseed oil as they are major inhibitorsof desaturation and elongation in liver cells,leading to arachidonic acid formation (Bezardet al., 1994). Moreover, as there is a preferentialprocessing of n-3 fatty acids by ∆6 desaturase atthe expense of desaturation of n-6 fatty acids(Spreecher, 1981), feeding n-3 fatty acids wouldlead to a reduction in arachidonic acid formation.In summary, inhibition of PG secretion can be

achieved through: 1) reduced synthesis ofarachidonic acid by ∆6 and ∆5 desaturase enzymesnecessary for conversion of linoleic acid toarachidonic acid; 2) alteration in fatty acid profile(in favor of omega-3 fatty acids in membranephospholipids which may or may not be precursorsof other eicosanoids); 3) inhibition of synthesis andactivity of cyclooxygenase enzymes responsible forthe synthesis of PGF2α; and 4) inhibition of geneexpression involved in the synthesis of series 2 PG(Mattos et al., 2000).

Maternal Recognition of Pregnancy

The dialogue between the conceptus anduterine endometrium leads to maintenance of thecorpus luteum. The ability of embryonicinterferon-τ to inhibit uterine secretion of PGF2α iscritical to the establishment of pregnancy in cattle.Up to 40% of total embryonic losses are estimatedto occur beween day 8 and day 17 of pregnancy(Thatcher et al., 1994). This high proportion oflosses is coincident with the period of conceptusinhibition of uterine PGF2α secretion, suggesting thatsome loss may be occurring because certainconceptuses are unable to inhibit secretion of PGF2α.Future strategies to improve embryo survival duringthis critical period will be based on a throughunderstanding of the factors regulating “a bettercommunication between the embryo and the motherat the embryo interface”.

The success of early pregnancy in themated cow is dependant on the successfulmaternal recognition of pregnancy (Thatcher etal., 1995; Mann et al., 1999). To achieve this,the embryo must prevent the demise of thecorpus luteum by the timely production ofinterferon tau, the embryonic signal which actsto inhibit the development of the maternalluteolytic mechanism. Interferon tau acts locallyin the uterus to suppress the development ofoxytocin receptors in the endometrium and therebysuppresses the secretion of luteolytic episodes ofPGF2α generated by the binding of oxytocin to itsreceptors (Mann et al., 1999). It has been shownthat the pattern and level of ovarian steroid hormones


40

in early pregnancy can influence both embryodevelopment and survival and the timing and intensityof the mothers luteolytic drive. For example, lowprogesterone or high oestradiol concentrationsduring the luteal phase increase the strength of theluteolytic drive, while low post ovulatoryprogesterone concentrations result in retardedembryo development (Mann et al., 1999).Moreover, low estradiol concentrations may preventpremature regression of the corpus luteum andprevent early embryonic death (Staples et al., 1998).

Oxytocin induces release of PGF2α(Tysseling et al., 1998). Increased PG synthesisinduced by oxytocin during days 5 to 8 ofpregnancy reduced pregnancy rates of beef cowsat 30 days after AI from 80 to 30% (Lemaster etal., 1999). Treating cows concomitantly with aninhibitor of prostanoid synthesis neutralized theeffect of oxytocin and restored pregnancy ratesto 80%. The implication is that increased PGF2αsecretion during early pregnancy causesembryonic loss and supports the hypothesis thatreducing PGF2α during this period reducesembryonic loss and improves pregnancy rates.Arachidonic acid is the rate limiting fatty acidfor the synthesis of PGF2α via the action of PGF2αsynthase. The same enzymes also are capable ofprocessing other fatty acids, such as EPA, whichis the precursor for the synthesis of prostanoidsof the 3 series. Increased availability of EPA inmembrane phospholipids could displacearachidonic acid, leading to increased synthesisof prostanoids of the 3 series at the expense ofprostanoids of the 2 series, such as PGF2α.Prostanoids of the 3 series are less bioactive,and there appears to be no evidence for their rolein ruminant luteolysis. Gamma-linolenic acid(GLA, C18:3n6) and EPA have been shown toreduce the synthesis in vitro of PGF2α and PGE2(Graham et al., 1994).

Both EPA and interferon-τ inhibit secretionof PGF2α through different mechanisms. Interferon-τ, but not EPA, reduced levels of enzyme geneexpression (cyclooxygenase-2) and thus modulatesPGF2α production (Thatcher et al., 2001b). On the

other hand, EPA does not seem to affect enzymegene expression but would be involved incompetition of precursors for processing by thecyclooxygenase enzymes and regulation of enzymeactivity. The implication of these findings is thatsupplementation with inhibitory fatty acids, such asEPA, during early pregnancy by dietary orparenteral means may further enhance thesuppression of PGF2α secretion in concert with theaction of embryonic interferon-τ. Because asignificant proportion of bovine embryos are thoughtto be lost due to inadequate inhibition of uterinePGF2α secretion, further inhibition by exogenousmeans may result in increased embryo survival. Thishypothesis is supported by the findings of Burke etal. (1997), in which feeding lactating dairy cows oflow fertility a source of EPA and DHA in fish mealincreased pregnancy rates from 31.9 to 41.3%.

Fatty Acids, Parturition, and RetainedPlacenta

Parturition is a process that isaccompanied by the massive release of PG.Alterations of fatty acids in the endometriumhave been described in normal parturition andmanipulations of fatty acid content usedexperimentally to delay onset of parturition.Fatty acids of the omega-3 family have beenshown to affect uterine activity during parturitionin rats and sheep and to delay the onset ofparturition in humans (Olsen et al., 1992).Supplementing linolenic acid to a diet deficientin essential fatty acids resulted in an impairmentof parturition rates (Leat and Horthrop, 1979).This also occurred when fish oil was given torats as the major dietary essential fatty acidsource, and an inhibition of uterine synthesis ofPGE2 was detected (Leaver et al., 1986). In pre-term pregnant sheep, intravenous infusion of a20% omega-3 fatty acid emulsion resulted in a delayin the onset of induced labour and deliverycompared with a control group infused with anemulsion of soybean oil containing 7% omega-3 fatty acids.


41

Forages could also affect reproduction.Chassagne and Barnouin (1992) reported that cowsfed grass silage had lower blood PG concentrationsthan those fed corn silage. Grass silage had a greaterconcentration of linolenic acid and lowerconcentration of linoleic acid than corn silage. As aresult, the incidence of retained placenta in cowsfed grass silage was higher than in cows fed cornsilage. As linolenic acid is an inhibitor of PGsecretion, a high linolenic to linoleic acid ratio (grasssilage), therefore, could result in retained placenta.Kemp et al. (1998) reported that cows requiringmore time to expulse their placenta also had lowerPG metabolite blood concentrations at calving.However, in their experiment, the linolenic to linoleicacid ratio (C18:3:C18:2) had no effect on deliverytime of placenta, probably because the differencein the linolenic to linoleic acid ratio between flaxseedand sunflower seed based diets was not large. Thiswould suggest that fatty acid composition of foragesand diets could have important effects on cowreproduction. So far, we have observed (Benchaarand Petit, unpublished results) that a grass silagebased diet decreases the omega-6 to omega-3 fattyacid ratio in milk compared to a corn silage baseddiet; we have no data, however, on the effects ofthese two diets on reproduction parameters.

Fatty Acids and Reproduction Function

In theory, feeding omega-3 fatty acidswould delay the return to cyclicity after calvingdue to a decrease in the synthesis of series 2 PG,which could increase the number of days to firstservice. Synthesis of series 2 PG is required aftercalving for uterine involution, which leads tothe return of normal cyclicity. On the other hand,feeding omega-3 fatty acids would improvematernal recognition and thus decrease embryomortality. Taken altogether, this would stronglysuggest that feeding omega-3 fatty acids woulddelay the return to cyclicity but lead to a bettergestation rate when cows are bred. We arecurrently conducting an experiment to study thishypothesis and our preliminary results show thatcows fed omega-3 fatty acids have no embryo

mortality compared to those fed micronizedsoybeans or calcium salts of palm oil (Petit andTwagiramungu, 2002).

Therefore, feeding omega-3 fatty acidsshould improve the overall reproductive functionof cows as a result of better gestation rate,decreased embryo mortality, and decreasedservice per conception. However, we still needto do more research on this topic as there arealmost no published data regarding the effectsof specific fatty acids on the overall reproductionof dairy cows. There is a specific need to developdifferent feeding strategies according to thereproductive stage of cows; fatty acids requiredfor better maternal recognition (omega-3) won’tbe necessary the same as those required for easiercalving (omega-6). We should be balancing dietsfor specific fatty acids for optimum reproductionperformance. The only problem is thatpolyunsaturated fatty acids (e.g. omega-3 andomega-6) are biohydrogenated by rumenmicrobes and these fatty acids must bypass therumen to have any effect on reproduction. Thesefatty acids must therefore be protected againstthe attack of rumen microbes, but they mustremain digestible in the intestine and this is evenmore important for free oils. Oils contained infish meal (EPA and DHA) partially escapebiohydrogenation in the rumen (Ashes et al.,1992).

Conclusions

In a practical manner, we could summarizefour possible strategies to improve reproductionof the cow:

1) Generate a larger corpus luteum: it isknown that a larger corpus luteum will secretemore progesterone and this may have a positiveeffect on pregnancy recognition and consequentlypregnancy rates. Feeding flaxseed has increasedcorpus luteum diameter and progesteroneconcentration in dairy cows (Petit et al., 2002).


42

2) Increase progesterone concentration: agreater progesterone concentration leads to a bettermaternal recognition of pregnancy (Staples et al.,1997). Feeding omega 3 fatty acids usuallyincreases progesterone concentration.

3) Increase series 3 prostaglandin secretion:feeding fish meal (Burke et al., 1997) andflaxseed (Petit et al., 2002) inhibit PGF2αsynthesis as both will lead to the synthesis ofseries 3 prostaglandins. Competition for thesame key enzymes will lead to a lower synthesisof PGF2α.

4) Inhibition of cyclooxygenase activity:Cyclooxygenase is the enzyme leading to thesynthesis of PGF2α. Eicosapentaenoic anddocosahexanoic acids have been shown toinhibit cyclooxygenase activity. Highconcentrations of 20-carbon fatty acids (such asEPA, C20:5) other than arachidonic acid(C20:4n6), can compete with arachidonic acidfor active sites of prostaglandin-endoperoxidesynthase complex, therefore, reducing theconversion of arachidonic acid to the series 2PG (Weber and Sellmayer, 1990).

References

Abayasekara, D.R.E., and D.C. Wathes. 1999.Effects of altering dietary fatty acid compositionon prostaglandin synthesis and fertility.Prostaglandins, Leukotrienes and Essential FattyAcids 61:275-287.

Ashes, J.R., B.D. Sieber, S.K. Gulati, A.Z.Cuthbertson, and T.W. Scott. 1992. Incorporationof n-3 fatty acids of fish oil into tissue and serumlipids of ruminants. Lipids 27:629-631.

Barnouin, J., and M. Chassagne. 1991. Anaetiological hypothesis for the nutrition-inducedassociation between retained placenta and milkfever in the dairy cows. Ann. Rech. Vet. 22:331-343.

Béréziat, G. 1978. Voies métaboliques et régulationsde la biosynthèse des prostaglandines etthromboxanes. Rev. Fr. Corps Gras 25:463-473.

Bezard, J., J.P. Blond, A. Bernard, and P. Clouet.1994. the metabolism and availability of essentialfatty acids in animal and human tissues. Reprod.Nutr. Dev. 34:539-568.

Britt, J.H., D.W. Shaw, S.P. Washburn, and V.S.Hedgpeth. 1996. Endogenous progesteroneduring the luteal phase before inseminationinfluences embryo recovery in lactating dairycows. J. Anim. Sci. 74 (Suppl. 1):225 (Abstr.).

Burke, J.M., D.J. Carroll, K.E. Rowe, W.W.Thatcher, and F. Stormshak. 1996. Intravuscularinfusion of lipid into ewes stimulates productionof progesterone and prostaglandin. Biol. Reprod.55:169-175.

Burke, J.M., C.R. Staples, C.A. Risco, R.L. deLaSota, and W.W. Thatcher. 1997. Effect ofruminant grade Menhaden fish meal onreproductive and productive performance oflactating dairy cows. J. Dairy Sci. 80:3386-3398.

Chassagne, M., and J. Barnouin. 1992.Circulating PGF2α and nutritional parameters atparturition in dairy cows with and withoutretained placenta: relation to prepartum diet.Theriogenology 38:407-418.

Cunnane, S.C. 1995. Metabolism and function ofα-linolenic acid in humans. Pages 99-127 inFlaxseed in human nutrition. S.C. Cunnane andL.U. Thompson, eds. AOCS Press, Champaign,IL.

Dewhurst, R.J., and P.J. King. 1998. Effects ofextended wilting, shading and chemicaladditives on the fatty acids in laboratory grasssilages. Grass For. Sci. 53: 219-224.

Erasmus, U. 1993. Fats that heal fats that kill. 9thEd., 456 pages, Alive books, Burnaby, BC.


43

Ferguson, J.D., D. Sklan, W.V. Chalupa, and D.S.Kronfeld. 1990. Effects of hard fats on in vitro andin vivo rumen fermentation, milk production, andreproduction in dairy cows. J. Dairy Sci. 73: 2864-2879.

Fly, A.D., and P.V. Johnston. 1990. Tissue fattyacid composition, prostaglandin synthesis, andantibody production in rats fed corn, soybean,or low erucic acid rapeseed oil (canola oil). Nutr.Res. 10:1299-1310.

Graham, J., S. Franks, and R. C. Bonney. 1994.In vivo and in vitro effects of (-linolenic acidand eicosapentaenoic acid (C20:5, n-3) onprostaglandin production and arachidonic acid(C20:4, n-6) uptake by hyman endometriumprostalgandins Leukotrienes and Essential FattyAcids 50:321-329.

Grummer, R.R. and D.J. Carroll. 1991. Effectsof dietary fat on metabolic disorders andreproductive performance of dairy cattle. J.Anim. Sci. 69:3838-3852.

Kemp, B., N.M. Soede, M. Kankofer, M. Bevers,M.A.M. Taverne, T. Wensing, and J.P.T.M.Noordhuizen. 1998. Influence of linoleic/linolenic acid ratio in the diet of periparturientcattle on plasma concentrations of PGF2αmetabolite and placental expulsion rate.Theriogenology 49:571-580.

Lamming, G.E., and M.D. Royal. 2001. Ovarianhormone patterns and subfertility in dairy cows. Occ.Publ. Br. Soc. Anim. Sci. No. 26 (Vol. 1):105-118.

Leat, W.M.F., and C.A. Northrop. 1979.Supplementation with linolenic acid of a dietdeficient in essential fatty acids results in impairedparturition in rats. J. Physiol. 290:37P.

Leaver, H.A., F.D.C. Lytton, H. Dyson, M.L.Watson, and D.J. Mellor. 1986. The effect of dietaryo3 and 6 polyunsaturated fatty acids on gestation,parturition and prostaglandin in intrauterine tissuesand the kidney. J. Lip. Res. 25:143-146.

Lemaster, J.W., R.C. Seals, F.M. Hopkins, andF.N. Schrick. 1999. Effects of administration ofoxytocin on embryonic survival in progestogensupplemented cattle. Prostaglandins Other LipidMediat. 57:259-268.

Lucy, M.C., J.D. Savio, L. Badinga, R.L. De laSota, and W.W. Thatcher. 1992. Factors thataffect ovarian follicular dynamics in cattle. J.Anim. Sci. 70:3615-3626.

Mann, G.E., G.E. Lamming, R.S. Robinson, andD.C. Wathers. 1999. The regulation ofinterferon-tau production and uterine hormonereceptors during early pregnancy. J. Reprod. Fert.54 (Suppl.):317-328.

Mattos, R., C.R. Staples, and W.W. Thatcher.2000. Effects of dietary acids on reproductionin ruminants. Rev. Reprod. 5:38-45.

Olsen, S.F., J.D. Sorensen, N.J. Secher, M.Hedegaard, T.B. Henriksen, H.S. Hansen, andA. Grant. 1992. Randomised controlled trial ofeffect of fish-oil supplementation on pregnancyduration. Lancet 25:1003-1007.

Petit, H.V., R.J. Dewhurst, J.G. Proulx, M. Khalid,W. Haresign, and H. Twagiramungu. 2001. Milkproduction, milk composition, and reproductivefunction of dairy cows fed different fats. Can. J.Anim. Sci. 81:263-271.

Petit, H.V., R.J. Dewhurst, N.D. Scollan, J.G.Proulx, M. Khalid, W. Haresign, H.Twagiramungu, and G.E. Mann. 2002. Milkproduction and composition, ovarian function,and prostaglandin secretion of dairy cows fedomega-3 fats. J. Dairy Sci. 85:889-899.

Petit, H.V., and H. Twagiramungu. 2002.Reproduction of dairy cows fed flaxseed, Megalac®

or micronized soybeans. J. Anim. Sci. 80 (Suppl.1)/J. Dairy Sci. 85 (Suppl. 1):312.


44

Spicer, L.J., E. Alpizar, and S.E. Echternkamp.1993. Effects of insulin, insulin-like growth factor I,and gonadotropins on bovine granulosa cellproliferation, progesterone production, estradiolproduction, and(or) insulin-like growth factor Iproduction in vitro. J. Anim. Sci. 71: 1232-1241.

Spreecher, H. 1981. Biochemistry of essentialfatty acids. Prog. Lip. Res. 20:13-22.

Staples, C. R., J.M. Burke, and W.W. Thatcher.1998. Influence of supplemental fats onreproductive tissues and performance oflactating cows. J. Dairy Sci. 81:856-871.

Staples, C. R., W.W. Thatcher, and J.M. Burke.1997. Influences of dietary energy, fat, andprotein on reproductive performance of lactatingdairy cows. Pages 204-221 in Proc. IX Int. Conf.on Prod. Dis. Farm Anim. Ferdinand EnkeVerlag, Stuttgart, Germany.

Thatcher, W.W., M. Binelli, D. Arnold, R.Mattos, L. Badinga, F. Moreira, C.R. Staples andA. Guzeloglu. 2001a. Endocrine andphysiological events from ovulation toestablishment of pregnancy in cattle. Occ. Publ.Br. Soc. Anim. Sci. No. 26 (vol. 1):81-92.

Thatcher, W.W., M. Binelli, J. Burke, C.R. Staples,J.D. Ambrose, and S. Coelho. 1997. Antiluteolyticsignals between the conceptus and endometrium.Theriogenology 47:131-140.

Thatcher, W.W., A. Guzeloglu, R. Mattos, M.Binelli, T.R. Hansen, and J.K. Pru. 2001b.Uterine-conceptus interactions and reproductivefailure in cattle. Theriogenology 56:1435-1450.

Thatcher, W.W., M.D. Meyer, and G. Danet-Desnoyers. 1995. Maternal recognition ofpregnancy. J. Reprod. Fert. 49 (Suppl.):15-28.

Thatcher, W.W., and C.R. Staples. 2000. Effectsof dietary fat supplementation on reproduction inlactating dairy cows. Advances in Dairy Technology12:213-232.

Thatcher, W.W., C.R. Staples, G. Danet-Desnoyers, B. Oldick, and E.P. Schmitt. 1994.Embryo health and mortality in sheep and cattle.J. Anim. Sci. 72 (Suppl. 3):16-30.

Tysseling, K.A., W.W. Thatcher, F.W. Bazer, P.J.Hansen, and M.A. Mirando. 1998. Mechanismsregulating prostaglandin F2 alpha secretion fromthe bovine endometrium. J. Dairy Sci. 81:382-389.

Weber, P.C., and A. Sellmayer. 1990.Modification of the eicosanoid system and cellsignaling by precursor fatty acids. Adv.Prostaglandin Thromboxane Leukotriene Res.21:217-224.

Wu, D., S.N. Meydani, M. Meydani, M.G.Hayek, P. Huth, and R.J. Nicolosi. 1996.Immunologic effects of marine- and plant-derived n-3 polyunsaturated fatty acids innonhuman primates. Am. J. Clin. Nutr. 63:273-280.


45

Table 1. Comparison of major fatty acids in some edible oils (w/w% fatty acids)1.

1Adapted from Erasmus (1993).2Calcium salts of fatty acids. Church and Dwight Co., Inc., Princeton, NJ.

OIL C18:0 C18:1 C18:2 C18:3

Peanut 2 47 32 0

Canola 2 64 19 8

Safflower 2 12 77 0

Cottonseed 25 21 50 0

Linseed (flax) 4 19 14 58

Corn 2 25 60 1

Tallow 15 41 8 1

Fishmeal (10 to 12 % fat) 2 25 4 45

Hi Linolenic Ryegrass 6 4 14 43

Olive 2 76 8 0

Palm 4 39 10 1

Sesame 2 42 45 0

Soybean 4 24 53 7

Sunflower 5 20 69 0

Megalac®2 3.5 32.3 7.8 0.3


46

Omega-6 fatty acids Omega-3 fatty acids

C18:2n6 Linoleic acid C18:3n3 α-Linolenic acid

C18:3n6 γ-Linolenic acid C18:4n3 Stearidonic acid

C20:3n6 Dihomo−γ−Linolenic acid C20:4n3 Eicosatetraenoic acid

C20:4n6 Arachidonic acid C20:5n3 Eicosapentaenoic acid

∆6 Desaturase

Elongase

∆5 Desaturase

Figure 2. Schematic pathway of omega-6 and omega-3 fatty acid synthesis.

Figure 1. Schematic pathway of omega-7 and omega-9 fatty acid synthesis.

P a lm itic A c idC 16 :0

S te a ric A c idC 1 8 :0

O m e ga 7

O m e g a 9O le ic A c idC 1 8 :1

C 1 4 :0


47

Omega-6 fatty acids Omega-3 fatty acids

C18:2n6 Linoleic acid C18:3n3 α-Linolenic acid

C18:3n6 γ-Linolenic acid C18:4n3 Stearidonic acid

C20:3n6 Dihomo−γ−Linolenic acid C20:4n3 Eicosatetraenoic acid

C20:4n6 Arachidonic acid C20:5n3 Eicosapentaenoic acid

Series 1Prostaglandins



Prostaglandins H2

cyclooxygenase

PGF2α synthase

prostaglandin synthase

Figure 3. Metabolic pathway of series 1, 2, and 3 prostaglandins.


48


49

Concentrations of Carbohydrates for Close-up Rations

John E. Shirley1 and Aaron F. Park

1Contact at: Kansas State University, 135 Call Hall, Manhattan, KS 66506-1600, (785) 532-1218, FAX (785) 532-5681,Email: [email protected]

Department of Animal Sciences and IndustryKansas State University

Abstract

The major ingredient in dairy cow diets,regardless of physiological state, is carbohydrate.Altering the type and physical form of thecarbohydrate fraction can vary the energeticvalue of diets for dairy cows. A close-up dietshould not be formulated as an entity unto itself,but as a bridge between a low and high-energydiet. Therefore, it should retain some of thecharacteristics of both the far-off and lactationdiets. The ultimate success of a transition cownutrition and management program is a lactationcharacterized by high milk and component yieldsand an absence of ruminal, metabolic, mammarygland, and reproductive disorders. Therefore,close-up diets should encourage ruminaladaptation to subsequent lactation diets, preventmetabolic disorders, and minimize tissuemobilization prior to parturition. Rumenbacteria, protozoa, and fungi are sensitive to newdiet ingredients and the amount of substrateavailable (DM intake) thus, adequate timeshould be allocated to close-up diet exposureprior to parturition. The total carbohydratecontent of typical far-off, close-up, and lactationdiets approximates 78, 74, and 68% of totaldietary DM, respectively. The percentage of thecarbohydrate fraction allocated to neutraldetergent fiber (NDF) and non-fibercarbohydrate (NFC) in far-off, close-up, andlactation diets approximates 55, 45; 47, 53; and40, 60%, respectively. The net effect of thesechanges is an increase in estimated net energy

but the actual energy value per unit will varywith DM intake.

Introduction

The primary purpose of a close-up dietis to initiate ruminal changes beneficial to thecow after parturition when her nutritionaldemands dramatically increase. The far-off drycow diet is generally high in NDF, moderate tolow in NFC, and low in crude protein (CP)whereas, the lactation diet is relatively low inNDF, high in NFC, and high in CP. The close-up diet serves as a bridge between far-off andlactation diets thus, the primary changes informulation should be the type of carbohydrateand, possibly, level of CP. The most appropriatechanges to make are not clearly understood, buta recent study by Park et al. (2002b) suggeststhat the NFC content of the close-up diet shouldbe above 32% of DM. Holcomb et al. (2001)fed prepartum diets containing high or lowforage and reported no effect of forage level onpostpartum performance, but their dietscontained less than 32% NFC (25 and 30%,respectively). Hayirli et al. (2002) used data sets(49 diets) from 10 universities to evaluate animaland feed factors that affect feed intake duringthe prefresh transition period and demonstratedthat dietary NDF accounted for 15.3% of thevariation in DM intake of Holstein cows.Dietary NDF was negatively related, while NFCwas positively related, to DM intake. Their(Hayirli et al., 2002) diets ranged in NDF content


50

from 28.0 to 62.2% and NFC content from 10.5to 46.2% of DM. We would expect that DMdigestibility would increase as we replace NDFwith NFC up to a point and then decrease whenthe rumen environment is compromised.Confounded with the issue of the mostappropriate dietary NDF to NFC ratio is thedigestibility of the NDF fraction. Varga (2002)suggests that highly fermentable non-forage fibersources offer a means of stimulating intake inthe prefresh cow by increasing rate of passagethrough the rumen. Ferdinand et al. (2002)improved DM intake by replacing a portion ofthe alfalfa hay, corn silage, and corn grain witheither wet corn gluten feed or a soy hull-cornsteep liquor blend in diets fed to early lactationcows. The obvious advantage of highlyfermentable non-forage fiber sources is that theyincrease nutrient delivery to the cow withoutincreasing the risk of acidosis.

The basic question left unanswered in theliterature is how rumen microbes react to dietarychanges dictated by the two-tier dry cow feedingprogram. This presentation will discuss ruminaladaptation to changes in diet and DM intakeduring the transition from far-off to close-up drythrough early lactation and some factors toconsider regarding the carbohydrate fractionswhen formulating close-up diets.

Composition of far-off, close-up, and lactationdiets

Formulations that have been successfulin our program for far-off, close-up, and lactationdiets are shown in Table 1 and the chemicalcomposition in Table 2. These diets were offeredas a total mixed ration (TMR) and fed for adlibitum intake. The major component changesbetween the far-off and close-up diets were theintroduction of alfalfa hay, a decrease in prairiehay and corn grain, and an increase in corn silageand soybean meal.

These changes increased CP, NFC, andestimated NEL and decreased NDF as a

percentage of total diet DM. Table 3 provides adifferent perspective of these diets. Totalcarbohydrates decrease as we transition from thefar-off to lactation diet and a major shift incarbohydrate fractions occurs, NDF decreases,and NFC increases. The amount of corn grainwas reduced in the close-up compared to the far-off diet because the amount of corn silageincreased. A comparison of TMR particle sizeusing the Penn State Particle Separator(Lammers et al., 1996) is shown in Table 4. Theparticle size is slightly less in the close-upcompared to the far-off diet; the major changein particle size occurs in the lactation diet.

Intake and digestive characteristics of far-off,close-up, and lactation diets

Dry matter intake (Table 5) averaged2.3% of BW during the far-off and early close-up periods, then decreased to 1.97% of BWduring the week before calving when the abovediets were fed ad libitum. Actual intakes rangedfrom 15 to 20% higher than predicted by thedairy NRC (2001) for dry cows and cows duringearly lactation but were within 1% of actual at90 days in milk. Dry matter intake increasedslightly (31 to 32 lb/day) when cows wereswitched from the far-off to the close-up diet,but rumen fill decreased because solids passagerate increased (Tables 5, 6, and 7). These eventswere predictable because the close-up diet wasformulated to improve digestibility. Thedigestibility characteristics of the diets duringthe transition period and through 90 days in milkare in Table 8. Dry matter, organic matter, CP,and NDF digestion increased as DM intakedecreased during the close-up period.Digestibility of these major dietary componentswas less on day 20 of lactation than on day 6,likely due to the rapid increase in DM intake,and then gradually increased through day 90 oflactation. Solids and liquid passage rates wererelatively constant between day 20 and 48, thengradually declined through day 90 of lactation.These results support the concept that the close-up diet prepared the ruminal microbes for the


51

lactation diet but not for the substantial increasein DM intake. This concept is further supportedby our inability to detect lactic acid in ruminalsamples throughout the study. The concentrationof total ruminal volatile fatty acids (Table 9)gradually increased from day 51 prepartum (far-off diet) through day 34 postpartum, peaked onday 48 postpartum, and then gradually decreasedthrough day 90 of lactation. Concentrations ofacetate, propionate, and butyrate mimicked thepattern of total volatile fatty acids, while theconcentrations of valerate, isobutyrate, andisovalerate tended to peak by day 20 postpartum(Table 10). These results are consistent with thepattern of intake, DM fill, organic matter fill,and solids passage rate and retention time.

General comments about close-up diets

Properly formulated close-up diets canreduce the incidence of postpartum metabolicproblems and prepare the rumen for lactationdiets; however, it is not a magic formula thatcan cure problems created by overly fat cowsand low quality or improperly formulatedlactation diets. In our view, cows should enterthe close-up pen at a body condition score of2.75 to 3.0 and gain 0.25 points prior to calving.To accomplish this, cows should enter the close-up pen approximately 28 days prior to expectedcalving date to insure that they are exposed tothe close-up diet for a minimum of 14, buthopefully 21 days. Park et al. (2000) reportedthat body condition score at the beginning of theclose-up period (day 31 prepartum) wasnegatively related to prepartum and postpartumDM intake (first 90 days) and milk yield duringthe first 90 days of lactation but not significantlyrelated to complete lactation milk yield. Thesedata support the concept that fatter cows eat lessthan thinner cows and suggest that energy gainedfrom fat mobilization during early lactation doesnot offset the decrease in DM intake experiencedby fatter cows in support of milk production.Hayirli et al. (2002) also reported a negativerelationship between body condition score andprepartum DM intake.

The 2001 Dairy NRC recommends thatclose-up diets contain 0.73 Mcal NEL/lb ofdietary DM. The close-up diet discussed in thispresentation contained 0.71 Mcal NEL/lb of DMbased on the summation of values fromindividual feedstuffs and 0.74 NEL/lb dry matterbased on NRC 2001. A significant portion ofthe NEL in our close-up diet is due to the CPcontent (15.5% of DM), so the recommendedNEL value can be achieved without a substantialincrease in the rapidly fermentable carbohydratefraction.

Summary

The primary purpose of a close-up dietis to initiate ruminal changes beneficial to thecow after parturition when her nutritionaldemands dramatically increase. Rumen bacteria,protozoa, and fungi are sensitive to new dietingredients and the amount of substrate availablethus, adequate time should be allocated to close-up diet exposure prior to parturition. Alteringthe type and physical form of the carbohydratefraction can vary the energetic value of diets.Close-up diets containing 35% NFC and 14.5%CP should stimulate fermentation sufficiently toprepare the rumen for the lactation diet. Properlyformulated close-up diets can reduce but cannotcompletely solve the problem of over fat cowsand low quality feedstuffs.

References

Ferdinand, E.E., J.E. Shirley, E.C. Titgemeyer,J.M. DeFrain, and A.F. Park. 2002. The effectof feeding wet corn gluten feed and a rawsoybean hull-corn steep liquor pellet on theperformance of lactating dairy cows. J. DairySci. 85 (Suppl 1):69. (Abstr.)

Hayirli, A., R.R. Grummer, E.V. Wordheim, andP.J. Crump. 2002. Animal and dietary factorsaffecting feed intake during the prefreshtransition period in Holsteins. J. Dairy Sci.85:3430-3443.


52

Holcomb, C.S., H.H. VanHorn, H.H. Head, M.B.Hall, and C.J. Wilcox. 2001. Effects ofprepartum dry matter intake and foragepercentage on postpartum performance oflactating dairy cows. J. Dairy Sci. 84:2051-2058.

Lammers, B.P., D.R. Buckmaster, and A.J.Heinricks. 1996. A simple method for theanalysis of particle sizes of forages and totalmixed rations. J. Dairy Sci. 79:922-928.

National Research Council. 2001. Nutrientrequirement of dairy cattle. 7th rev. ed. NationalAcademy Press, Washington DC.

Park, A.F., J.E. Shirley, M.J. Meyer, M.J.VanBaale, and E.C. Titgemeyer. 2000.Relationships of body weight and condition ofHolstein cows with performance traits during theperiparturient period. J. Dairy Sci. 83(Suppl.1):253. (Abstr.)

Park, A.F., J.E. Shirley, J.M. DeFrain, E.C.Titegemeyer, E.E. Ferdinand, R.C. Cochran,D.G. Schmidt, S.E. Ives, and T.G. Nagaraja.2001. Changes in rumen capacity during theperiparturient period in dairy cows. J. Dairy Sci.84 (Suppl. 1):339. (Abstr.)

Park, A.F., J.E. Shirley, E.C. Titgemeyer, M.J.Meyer, M.J. VanBaale, and M.J. VandeHaar.2002a. Effect of protein level in prepartum dietson metabolism and performance of dairy cows.J. Dairy Sci. 85:1815-1828.

Park, A.F., J.E. Shirley, E.C. Titgemeyer, R.C.Cochran, J.M. DeFrain, E.E. Ferdinand and T.G.Nagaraja. 2002b. Characterization of ruminalfermentation in transition dairy cows. J. DairySci. 85 (Suppl 1): 186. (Abstr.)

Varga, G.A. 2002. Feeding managementstrategies for dry cows. In: SouthwestDairyman. October issue, pg. 6.


53

Table 1. Diet ingredients (% of DM; Adapted from Park et al., 2001).Diets

Item Far-off Close-up Lactation

Alfalfa hay ¯ 15.0 30.0Prairie hay 48.4 20.0 ¯Corn silage 19.8 30.0 15.0Corn grain 22.4 18.7 32.0Whole cottonseed ¯ ¯ 9.3Fishmeal ¯ ¯ 1.3Expeller soybean meal ¯ 9.4 3.348% soybean meal 8.4 4.4 4.4Wet corn gluten feed ¯ ¯ ¯Molasses ¯ ¯ 1.0Limestone 0.06 0.60 1.36Dicalcium phosphate 0.40 0.74 0.88Sodium bicarbonate ¯ ¯ 0.75Trace mineral salt1 0.34 0.50 0.32Magnesium oxide ¯ 0.50 0.21Vitamin A,D,E2 0.11 0.12 0.13Sodium selenite premix3 0.02 0.04 0.01

1Composition: not less than 95.5% NaCl, 0.24% Mn, 0.24% Fe, 0.05% Mg, 0.032% Cu, 0.032% Zn,0.007% I, and 0.004% Co.2Contributed 4912 IU vitamin A, 2358 IU vitamin D, and 24 IU vitamin E per kg diet DM.3Contributed 0.06 mg Se per kg diet DM.


54

Table 2. Chemical characteristics of diets (adapted from Park et al., 2001).

DietsItem Far-off Close-up Lactation

DM, % 74.5 70.3 75.0CP, % 11.5 15.6 18.4Soluble protein, % of CP 25.2 25.2 31.3RDP, % of DM1 7.3 10.3 11.7RUP, % of DM1 4.2 5.3 6.7ADF, % 25.2 22.0 18.2NDF, % 42.9 34.4 27.0Non-fiber carbohydrate, %2 35.2 39.1 40.4NEL, Mcal/lb3 0.70 0.74 0.73Ether extract, % 3.8 3.5 5.6Ash, % 6.7 7.4 8.4TDN, % 67.0 69.1 72.3Calcium, % 0.5 0.8 1.5Phosphorus, % 0.4 0.5 0.7Magnesium, % 0.2 0.4 0.3Potassium, % 1.2 1.5 1.5Sodium, % 0.1 0.2 0.3Sulfur, % 0.1 0.2 0.2

1Based on feed analysis from Dairy Herd Improvement Forage Testing Laboratory (Ithaca, NY).2Calculated based on DHI formula represented by 100- [(crude protein + (NDF – NDICP) + etherextract + ash)].3Calculated based on NRC, 2001. Estimates of NEL values from summation of individual ingredients(0.66, 0.71, and 0.77 for the far-off, close-up, and lactation diets, respectively).

Table 3. Carbohydrate fractions in dry and lactating cow diets (adapted from Park et al., 2001).

Item Far-off Close-up Lactation

CP, % of DM 11.5 15.6 18.4Ether extract, % of DM 3.8 3.5 5.6Ash, % of DM 6.7 7.4 8.4Carbohydrates, % of DM 78 73.5 67.6NDF, % of CHO1 55 47 40NFC, % of CHO1 45 53 601CHO = carbohydrate.


55

Table 4. Particle size characteristics of diets (adapted from Park et al., 2001).

DietsItem Far-off Close-up Lactation

Total mixed ration particle size, %1

> 19.0 mm 34.1 ± 12.6 30.2 ± 11.8 11.2 ± 4.6 8.0 to 19.00 mm 20.1 ± 01.3 19.1 ± 05.2 24.9 ± 3.3 < 8.0 mm 45.8 ± 12.9 50.7 ± 08.0 63.9 ± 3.41Particle size determined by the Penn State Particle Separator (Lammers et al., 1996), as-fed basis;mean ± SD.

Table 5. Performance characteristics (adapted from Park et al., 2001)1.

NRC DMI DMI BW DMI EB

Day Diet F:C (lb/day) (lb/day) (lb) (% of BW) BCS (Mcal/day)

-51 Far-off 70 : 30 31.1 25.4 1365 2.29 2.75 8.5-23 Close-up 65 : 35 32.2a 27.1 1417 2.29d 2.75a 10.3 -9 Close-up 65 : 35 28.2ab 25.4a 1465d 1.97ad 2.94ab 7.3 6 Lactation 45 : 55 36.2b 49.4a 1312d 2.70a 2.50b (2.0)20 Lactation 45 : 55 45.2 36.4 1293 3.55 2.56 8.334 Lactation 45 : 55 53.6 39.5 1263 4.29 2.38 14.948 Lactation 45 : 55 56.2 41.5 1284 4.43 2.25 17.062 Lactation 45 : 55 56.9 42.6 1249 4.58 2.19 19.176 Lactation 45 : 55 57.8 51.8 1179 4.93 2.13c 11.390 Lactation 45 : 55 58.2 57.1 1211 4.81 2.38c 7.4SEM 2.0 8.6 144 0.20 0.18 5.9

abcLike superscripts within column differ by (P < 0.05).dLike superscripts within column tend to differ by (P < 0.10).1FC=Forage to concentrate ratio, DMI = dry matter intake, NRC = National Research Council(2001), BW = body weight, BCS=body condition score, EB = energy balance, and SEM = standarderror of mean.


56

Table 6. Rumen capacity and fill (adapted from Park et al., 2001).1

Total fill DM fill OM fill Liquid fill WHCDay Diet F:C (lb/day) (lb/day) (lb/day) (lb/day) (L)

-51 Far-off 70 : 30 132.5c 13.6c 11.7c 118.9 124.6-23 Close-up 65 : 35 117.3c 10.7ac 8.8ac 106.5 139.6 -9 Close-up 65 : 35 111.0a 13.4ab 11.5ab 97.6a 143.1c

6 Lactation 45 : 55 138.9a 19.3b 16.5b 119.5a 159.8c

20 Lactation 45 : 55 134.7 19.4 16.8 115.3 148.7 34 Lactation 45 : 55 152.5 21.7 18.5 130.8 170.0 48 Lactation 45 : 55 146.4 21.9 19.0 124.5 164.3 62 Lactation 45 : 55 162.9 24.9 21.4 137.9 160.3 76 Lactation 45 : 55 162.1 25.0 21.2 137.1 169.7 90 Lactation 45 : 55 170.6 25.1 21.8 145.5 171.4SEM 9.5 1.7 1.5 8.2 9.4abLike superscripts within column differ by (P < 0.05).cTended to be different (P < 0.1).1FC = Forage to concentrate ratio, DM = dry matter, OM = organic matter, WHC = water holdingcapacity, and SEM = standard error of mean.

Table 7. Passage rate and retention time (adapted from Park et al., 2002a).1

Solids Solid retention Liquid Liquid retentionDay Diet F:C (% / h) time (h) (% / h) time (h)

-51 Far-off 70 : 30 5.0a 20.7 12.6 8.2-23 Close-up 65 : 35 6.5ab 17.4a 12.0 8.3 -9 Close-up 65 : 35 4.5bc 22.3ab 12.2 8.2 6 Lactation 45 : 55 3.3c 31.0b 13.2 7.6 20 Lactation 45 : 55 3.8 27.2 14.4 7.0 34 Lactation 45 : 55 4.0 25.9 13.1 7.8 48 Lactation 45 : 55 4.4 23.4 14.1 7.2 62 Lactation 45 : 55 4.0 25.8 11.8 8.6 76 Lactation 45 : 55 3.8d 27.0c 12.7 8.3 90 Lactation 45 : 55 3.0d 35.2c 11.8 8.8SEM 0.5 2.8 1.2 0.9abcdLike superscripts within column differ by (P < 0.05).1FC = Forage to concentrate ratio and SEM = standard error of mean.


57

Table 8. Digestibility characteristics (adapted from Park et al., 2002a).1

PercentageDay Diet F:C DM OM CP NDF ADF Starch

-51 Far-off 70 : 30 45.9a 49.1a 42.9a 45.2 45.5e 100.0-23 Close-up 65 : 35 50.4 ab 54.2ab 52.2ab 49.5 50.5ae 99.4 -9 Close-up 65 : 35 54.5bc 58.0b 65.8b 48.0a 44.5ab 94.9e

6 Lactation 45 : 55 58.3c 60.1 62.4 62.1a 63.0b 89.2e

20 Lactation 45 : 55 50.7 53.0 52.8 58.4 57.1 80.9 34 Lactation 45 : 55 55.3 58.1 59.0 52.2 40.7 85.2 48 Lactation 45 : 55 53.2 55.4 55.8 55.9 42.8 80.7 62 Lactation 45 : 55 52.1 54.5 56.0 58.5 43.1 76.0 76 Lactation 45 : 55 55.4d 57.4c 55.9c 59.5 48.0c 81.3 90 Lactation 45 : 55 64.1d 66.1c 66.2c 59.4 56.4c 83.0SEM 1.4 1.3 1.6 1.8 1.7 2.4abcdLike superscripts within column differ by (P < 0.05).eTended to be different (P < 0.1).1FC = Forage to concentrate ratio, DM = dry matter, OM = organic matter, CP = crude protein, NDF= neutral detergent fiber, ADF = acid detergent fiber, and SEM = standard error of mean.

Table 9. Fermentation products and pH (adapted from Park et al., 2002b).1

Concentration, mMDay Diet F:C pH A:P ratio TVFA peptides FAA ammonia

-51 Far-off 70 : 30 6.59 3.88a 87.4 — 0.5 12.3-23 Close-up 65 : 35 6.59 3.46ab 94.1 0.2 1.0 9.6 -9 Close-up 65 : 35 6.65a 4.28bc 93.8a 0.2a 0.5a 6.9a

6 Lactation 45 : 55 6.28a 3.52c 109.6a 2.0a 2.6a 12.3a

20 Lactation 45 : 55 6.05 3.47 124.2 2.2 1.8 6.1 34 Lactation 45 : 55 6.29 3.84 123.7b 2.4 1.2 7.9 48 Lactation 45 : 55 6.15 3.96 153.2b 2.5 2.9 9.3 62 Lactation 45 : 55 6.14 3.26 131.0 1.1 1.8 7.1 76 Lactation 45 : 55 6.11 3.17 125.9c 1.4b 2.9b 9.8 90 Lactation 45 : 55 6.20 3.18 110.2c 1.2b 1.6b 9.5SEM 0.09 0.16 5.9 0.3 0.6 1.9abc Like superscripts within column differ by (P < 0.05).1FC = Forage to concentrate ratio, A:P = acetate:propionate, TVFA = total volatile fatty acids, FAA =free amino acids, and SEM = standard error of mean.


58

Table 10. Ruminal volatile fatty acids (adapted from Park et al., 2002a).1

Concentration, mMDay Diet F:C acetate propionate butyrate valerate isobutyrate isovalerate

-51 Far-off 70 : 30 60.5 15.7 9.5 0.9a 1.1 1.1a

-23 Close-up 65 : 35 63.9 18.6 9.7 1.2a 1.4 1.7a

-9 Close-up 65 : 35 65.0 15.3a 9.2a 1.1b 1.3a 1.5b

6 Lactation 45 : 55 73.1 21.5a 12.5a 1.7b 1.6a 1.9b

20 Lactation 45 : 55 79.6 23.1 14.3 2.1 1.9 2.5 34 Lactation 45 : 55 80.1 20.8b 13.5 1.7c 1.7 2.1 48 Lactation 45 : 55 106.7 27.2 14.1 2.1 1.3 2.0 62 Lactation 45 : 55 84.1 26.0b 15.3 2.1c 1.5 2.2 76 Lactation 45 : 55 79.7a 25.6c 14.1b 1.9d 1.5b 1.8c

90 Lactation 45 : 55 69.1a 21.9c 12.3b 1.6d 1.8b 2.1c

SEM 3.9 1.7 0.6 0.1 0.1 0.1

abcdLike superscripts within column differ by (P < 0.05).1FC = Forage to concentrate ratio and SEM = standard error of mean.


59

Soluble Carbohydrates for Lactating Dairy Cows

Gabriella A. Varga1

Department of Dairy and Animal ScienceThe Pennsylvania State University

Abstract

A major challenge in ration formulationis to provide an adequate amount of solublecarbohydrates in the diet to maintain highproduction, while at the same time assuringprevention of ruminal acidosis. The solublecarbohydrates include organic acids, sugars,starch, and soluble fiber, such as pecticsubstances, and are defined as non NDFcarbohydrates (NFC). It has been suggested torestrict NFC to 32 to 38 % of ration dry matter(DM) when carbohydrates are derived primarilyfrom sugar or starch or 38 to 42% when othersoluble carbohydrates make up the ration. Thecontribution of various NFC is critical as theydiffer in their rate and extent of fermentation inthe rumen and can have variable effects onrumen pH. Sugar and starch may ferment tolactic acid, while pectic substances fermentrapidly to acetate and lactate, but theirfermentation is depressed at lower pH. Tominimize the risk of ruminal acidosis andmaximize the provision of rapidly fermentableNFC in the ration of dairy cows, a balance ofthese carbohydrate sources is recommended.

Introduction

Carbohydrate nutrition of the highproducing dairy cow is of extreme importancefor the achievement of optimum performancebut is perhaps the most difficult nutrient toadequately balance in the ration. Part of the

1Contact at: 324 Henning Building, University Park, PA 16802, (814) 863-4195, FAX (814) 865-7442, Email:[email protected]

problem is that there is an inadequate descriptionof the carbohydrate content of various feedingredients and partly because of limitations forutilization of certain materials residing at thefarm. In order for high producing dairy cattle tomeet their high energy demands, diets mustconsist of large quantities of concentrates andhigh quality forages containing relatively lowamounts of fiber. However, to maintain normalrumen function and milk fat percentage, a largeportion of the fiber needs to come from forage.Sugars, starches, and other reservecarbohydrates, such as galactans and pectin,make up the highly digestible NFC and are theprimary dietary sources of energy for the highproducing dairy cow. Ruminant diets vary inNFC, and this can have a major impact on theend products of fermentation in the ruminalenvironment. These soluble carbohydrates arehighly digestible and ruminal fermentationvaries greatly with type and processing method.Understanding how these fermentablecarbohydrate sources fit in the ration and howthey differ in the nutrients they supply to theanimal, will offer a better sense of how weshould consider them in ration formulation.

NFC

Dietary NFC are generally categorizedinto organic acids, sugars (mono- andoligosaccharides), starch, and neutral detergent-soluble fiber (Hall and Van Horn, 2001; Figure1). Organic acids from fermented feeds may be


60

utilized by the animal but do not supportappreciable microbial growth. Sugars fermentquite rapidly in the rumen environment and alsocan be digested by mammalian enzymes. Starchmay be digested both by ruminalmicroorganisms and by digestive enzymes postruminally. However, there can be great variationin rate of ruminal fermentation or digestiondepending on the processing, storage method,or plant source of the starch. Neutral detergent-soluble fiber includes pectic substances, beta-glucans, fructans, and other non-starchpolysaccharides not included in neutral detergentfiber (NDF). Generally, most soluble fibersources tend to ferment very rapidly (20 to 40%/h), while others, such as soyhulls, are digestedat a much slower rate. Soluble fiber cannot bedigested by mammalian enzymes and must befermented by ruminal microbes to be utilizedby the animal. Ruminal fermentation of sugar,starch, and fructan may yield lactic acid and tendto yield more propionate than acetate. Pectins,which usually predominate in soluble fiber, tendto yield more acetate. Other than organic acids,the various soluble carbohydrates have beenconsidered to give similar yields of microbialprotein when pH is relatively neutral andfermentation rates are similar (Hall and VanHorn, 2001). Table 1 provides a list of somefeedstuffs, their particular NFC make-up, andthe potential end products formed during ruminalfermentation.

Organic acids

Plant organic acids include citrate,malate, quinate, succinate, fumarate, oxalate,shikimate, trans-aconitate, and malonate amongothers. The range of organic acids present inforage has made complete analysis for themdifficult (Hall, 2000). Total levels in plant DMreported in cool season grasses are 1.3 to 4.5%,5.8 to 9.8% for alfalfa, 2.8 to 3.8% in red clover,and 3.0 to 3.5% in white clover (Dijkshoorn,1973). Alfalfa varieties may contain 2.9 to 7.5%malate (Callaway et al., 1997), while citrus peelcontains 3 to 4% of DM as organic acids

(Vandercook, 1977). The volatile fatty acids(VFA) are the fermentation acids in feedstuffs,such as silages, and contribute to the energy thatthe cow receives from the fermented feed butprovide little or no fermentable carbohydratefor the rumen (Hall, 2000). However, as it relatesto the animal, VFA can contribute as much as150 to 300 g of absorbed amino acids per day,which is quite significant (Fox et al., 1992).

Sugars

Glucose, fructose, and sucrose are thepredominant low molecular weightcarbohydrates in forages. The water-soluble cellcontents are reported to account for 1 to 3% offorage DM for the simple sugars and 2 to 8%for sucrose in temperate forages (Smith, 1973).Some of the by-products from the food industry,such as bakery and cereal, can vary widely insugar content. Candy, donuts, and sugar-basedcereals will be higher in sugar than breadproducts or bran type cereals. Analysis of 17samples of citrus pulp showed an average sugarcontent of 25.9%, with a range of 19 to 31%.Beet pulp is lower in sugar (12%) than citruspulp (Aldrich, 2001). Beet molasses is generally40% sugar on an as-is basis, while whey isapproximately 70% sugar (lactose). Almondhulls contain 19 to 34% soluble sugars (Aguilaret al., 1984), varying by variety. Other feedstuffsand their sugar composition are provided inTable 2.

Simple sugars and oligosaccharides arethe most rapidly fermented of carbohydratesources by ruminal microorganisms. Thefermentation of sugars is similar to that of starchin that both can ferment to lactic acid. Endproducts of sucrose fermentation are dependenton other carbohydrate sources in the diet thatmay contribute to variability in pH. Table 3provides the proportion of carbohydratesconverted to substrate based on all forage versus50% forage rations. In the study conducted byStrobel and Russell (1986), fermentation ofsucrose by rumen microbes resulted in similar


61

concentrations of microbial protein, acetate, andpropionate as compared to starch, but morebutyrate and lactate at pH 6.7. However, at amore acidic pH (5.5), fermentation of sucroseproduced more lactate than did starch, andmicrobial protein yield from sucrose wasreduced by 34%.

An in vitro study with mixed ruminalmicrobes examined the effects of different levelsof sucrose (65, 130, and 195 mg) fermented withbermudagrass NDF (130 mg) on microbialproducts (Hall, 2002). The maximal yield ofmicrobial protein increased linearly withincreasing sucrose, and the efficiency ofproduction decreased linearly from 0.32 to 0.23(mg of microbial CP / mg of sucrose). Dextranyield peaked at four hours of fermentation andgradually declined (Table 4). Reports ofinefficiency of CP utilization in diets containingmolasses compared to those containing corn(Bell et al., 1953) may be related to sucrose-utilizing microbes storing dextran to use formaintenance rather than growth. Concentrationsof all the VFA and lactate increased linearly (P< 0.01) with increasing sucrose (Hall, 2002)(Table 5). It appears that yields of microbialproducts (dextran, microbial protein, and typesof organic acids) usable by the animal to meetnutrient requirements may change with changinglevel of sucrose. In studies conducted by Khaliliand Huhtanen (1991), and Huhtanen and Khalili(1991) sucrose and molasses fed at high levelsdecreased ruminal fiber digestion. Martin et al.(1981) demonstrated, however, that increasingthe amount of protein in molasses-supplementedrations can significantly improve the digestibilityof fiber. The relationship between fiberdigestibility and protein supplementation onmolasses diets may be related to a competitionfor ammonia nitrogen between fiber and NFCfermenting bacteria (Jones et al., 1998).Adequate nitrogen must be supplemented to therumen to avoid starving fiber digesters,especially if rapidly growing NFC bacteria arescavenging available nitrogen.

Few studies have examined the additionof sucrose to lactating cow diets, and fieldobservations have been mixed. Corn silage canvary from 1 to 5% sugar on a DM basis. Haycrops are generally higher than ensiled haylages.Therefore ensiled feeds, depending on theirquality and fermentation process, can vary insugar content. The response to sugarsupplementation could vary based on the amountof total sugar needed in the diet and the amountalready present in forages. Optimum sugar levelshave not been well researched and certainly willdepend as previously mentioned on the amountof readily available protein in the diet. Animalresponses to added sugar may be highest whenthe basal diet contains 4% sugar or less.However, the challenge is to have sugar analysesavailable on all ingredients prior to sugaraddition. Two studies in which sucrose wassubstituted for starch in lactating dairy cowrations suggest that sucrose increases butterfatyield, but other results are varied. In diets wheresucrose was substituted for corn starch (0 to7.5% of dietary DM, diet NFC ~ 43% of DM;Broderick et al., 2000), there were increases inDM intake, milk fat content, and fat yield. Fat-corrected milk production tended to increase(Table 6). In terms of feed efficiency, milk /DM intake decreased from 1.60 to 1.52, and theconversion of ration nitrogen to milk protein Ndeclined linearly with increasing substitution ofsucrose for starch (from ~0.31 to ~0.29). Acontinuous culture study was conducted tocompliment this lactating cow study (Varga etal., 2001). Replacing corn starch with sugar inthe diet altered fermentation of ruminal endproducts, such that molar proportions of butyratewere increased and may explain the increase inmilk fat percentage in the study of Broderick etal. (2000). In addition, at the highest level ofsugar inclusion, fiber digestibility was enhancedsix percentage units. Increased bacterial nitrogenflow as a percentage of total nitrogen flowingas sugar replaced corn on the diet, may alsoexplain the increase in milk yield and trend forthe increase in milk protein observed in thecompanion lactating cow trial.


62

In a study conducted by Nombekela andMurphy (1995), sucrose was substituted for cornmeal at 1.5% of ration DM. Intake, milk yield,and fat-corrected milk yield did not change, butmilk fat yield increased from 2.12 to 2.14 lb/day, and milk protein decreased from 3.51 to3.28%. The lower milk protein percentage mayhave been the result of reduced microbial proteinproduction due to a decrease in ruminal pH andthe negative effects on fiber digestibility.Therefore, for sugar to be effective in lactatingcow rations, basal sugar levels need to be known,adequate rumen degradable protein needs to befed, and efficiency of microbial CP is likelydependent on level of sucrose and other solublecarbohydrates in the diet.

Varga and Whitsel (1991) demonstratedthat greater than 95% of the variation in rate andextent of fiber degradation could be explainedby the proportion of sugars in the diet. Inaddition, the carbohydrate level at which the dietis formulated is important. For example, if aration is formulated to contain 25% NDF on aDM basis, its relationship to NFC will bedifferent than when formulated to contain 30%NDF. This in part is related to the effect of thefermentability of the carbohydrate source andits effects on ruminal pH and ultimately fiberdigestion in the rumen.

Starch

Fermentation of starch by rumenmicrobes has a variety of similarities to that ofsugars. Starch may ferment to lactate (Strobeland Russell, 1986) and tends to produce a loweracetate to propionate ratio than cell wallcarbohydrates. Although starch-fermentingbacteria are more tolerant of acidic conditionsthan are fiber digesters, growth of starch-digesting microbes declines as pH declines(Therion et al., 1997). Strobel and Russell (1986)demonstrated that the yield of microbial proteindecreased by 35% when starch was fermentedby mixed rumen microbes at pH 5.8 versus 6.7.

The rate and extent of starch digestionis affected by a variety of factors. Particle size,grain type, steam flaking, and preservationmethod (dry or ensiled) all affect the availabilityof starch. In feeds such as corn, the smoothcovering of the seed offers the first barrier todigestion, and the protein matrix that surroundsthe starch granules the second (Kotarski et al.,1992). For whole grain, approximately 30% maypass undigested into the manure in cattle(Orskov, 1986). But as particle size in whole corndecreases, ruminal starch disappearancegenerally increases (Gaylean et al., 1981).Processing methods which disrupt the proteinmatrix around the starch granules have beenshown to increase grain digestibility (Lykos andVarga, 1995). Subjecting starch to heat andmoisture gelatinizes it, destroying the crystallinestructure of starch granules and increasingdigestibility in the rumen and total tract (Lykoset al., 1997). It has been suggested that overallmetabolizable energy yield to the cow is bestwhen starch is fermented in the rumen, due topossible limitations on its digestion in the smallintestine (Huntington, 1997). However, ifdigestion of starch in the small intestine wereenhanced, such as with more post ruminalprotein supply, it may be possible to improvethe animal’s capture of glucose from starch(Huntington, 1997).

In order to avoid acidosis and othermetabolic problems, the maximum level of truesugars plus starches should not exceed 30% ofthe ration DM. This has resulted in the followingguidelines: 32 to 38 % of DM as NFC when thedietary ingredients are high in sugar and starch,such as barley, corn, and corn silage; 38 to 42%of DM as NFC when the forage is all hay cropsilage and contains by-products such as corngluten feed and soyhulls. Results of studies byBatajoo and Shaver (1994) are generallysupportive of these recommendations. Theyconcluded that, for cows producing over 88 lb/day of milk, the diet should contain more than30% NFC, but they found little benefit of a 42%over 36% NFC diet. Nocek and Russell (1988)


63

suggested that 40% dietary NFC was optimal indiets for lactating cows from an evaluation ofdiets based on alfalfa silage, corn silage, and50:50 alfalfa:corn silage; dietary NFC rangedfrom 30 to 46%. Hoover and Miller (1991)regressed data from Nocek and Russell (1988)and demonstrated that when dietary NFC wasgreater than 45 to 50% or less than 25 to 30%,milk production was decreased. Alteration ofdietary NFC has been shown to influenceruminal fermentation patterns, total tractdigestion of fiber, and milk fat percentage(Sutton et al., 1987; Sievert and Shaver, 1993).

Hoover (West Virginia University,unpublished data), measured slow and fast sugarand starch (NSC) digestion of various nonforagefiber sources in situ and found quite a range inthe rate of digestion (Kd) of these fractions. Forexample, the fast pool for beet pulp, representing72% of the 15.9% total NSC as a percentage ofDM, had a Kd of 429%/h, while brewer’s grains(18.4% NSC as a percentage of DM) had a Kdfor the fast pool (65% of NSC) of only 66%/h.The variation in composition and rate of the NSCfraction of various nonforage fiber sources cangreatly affect overall fiber digestion throughchanges in rate of passage and amounts ofindigestible residue remaining in the rumen.

In addition to total starch level, the rateand extent of ruminal starch digestion also mayaffect the amount of a particular starch sourcethat can safely be added to a diet. The rate atwhich NFC is fermented and the retention timeof the NFC in the rumen determine ruminalfermentation of NFC. Rate of fermentation ofstarch varies extensively by type of grain andprocessing. Herrera-Saldana et al. (1990) rankedthe degradability of starch from various sourcesas follows: oats > wheat > barley > corn > milo.Processing methods, such as fine grinding andsteam flaking, also may alter ruminal availabilityof starch. Lykos and Varga (1995) demonstratedthat effective degradability of starch in situ forcracked corn, fine ground corn, and steam flakedcorn was 44.4, 64.5, and 75.4%, respectively. In

addition, the effective degradability of starch wasincreased for ground versus cracked soybeanswhether raw or roasted. Most grain processingmethods increase both rate of starchfermentation and ruminal starch digestibility. Itis important to note that grinding increases bothrate of digestion and rate of passage, which havecounteractive effects on ruminal digestibility(Gaylean et al., 1981). Animal characteristics andlevel of intake affect rate of passage. Therefore,fine grinding may have less effect on ruminalstarch digestibility at higher levels of intake, suchas dairy cattle in early lactation, than for growingor fattening animals, dry cows, or dairy cattle inlate lactation.

Results of lactation studies comparingstarch sources with differing digestibilities aresomewhat inconsistent. Herrera-Saldana andHuber (1989) reported higher milk productionwith a barley-cottonseed meal diet than with amilo-cottonseed meal diet, while McCarthy etal. (1989) and Casper et al. (1990) found milkproduction to be higher in diets containing cornthan those containing barley. Wilkerson andGlenn (1997) demonstrated increased yield ofmilk for cows fed high moisture corn versus drycorn (91.7 vs. 87.3 lb/d) and ground corn versusrolled corn (92.0 versus 87.1 lb/d). Lykos et al.(1997) demonstrated that increasing rate of totalNSC digestion from 6 to 7.9 %/h significantlyincreased milk yield 5.5 lb/day and protein yield130 g/day. Aldrich et al. (1993) observed lower4% FCM when diets high in rapidly fermentableNSC were fed to lactating cows in early lactation.Seymour et al. (1993) demonstrated that cowsfed with partially pelleted steam flaked grainproduced less milk compared those fed pelletedgrain. Diets with increased ruminally degradedstarch did not affect milk yield or FCM in otherstudies (Oliveira et al., 1995). Some of thevariation in results of type and amount of NFCin the diets for high producing lactating cowsmay be related to: 1) effects of rapidly degradablestarch on ruminal digestion of fiber, which candecrease the differences between diets relativeto total carbohydrate digestion; 2) level to which


64

NFC replaces fiber in the diet, as this can effectrumination and saliva production; 3) site ofstarch digestion; 4) level of DM intake andphysiological state of the animal; and 5)conservation and processing methods used toalter rate and extent of NFC digestion.

Nocek and Russell (1988) conductedstarch analysis on a variety of grain and foragesources and the relationship between starch andNDF content was r = -0.68 (P > 0.10). In orderto identify relationships between chemicalcarbohydrate measures and productionperformance in lactating dairy cows, Nocek(1988) assembled a database representing 14studies consisting of 62 diets. Milk yieldaveraged 70 lb/day and DM intake averaged 46lb/d. Neutral detergent fiber and starch intakeswere poorly correlated (r = 0.36, P > 0.10), whilestarch was highly correlated to NElconcentration of the diet (r = 0.85, P < 0.01).Starch intake was highly correlated to milk yield(r = 0.60, P < 0.01), DM intake, (r = 0.85, P <0.01), and protein yield (r = 0.71, P < 0.01). Inthese studies, NDF was not correlated to milkor protein yield; however, a significantrelationship between NDF intake and DM intakewas observed (r = 0.69, P < 0.01). The studiesevaluated by Nocek et al. (1988) were mainlydiets where the starch sources were primarilybarley or corn and with or without processingimposed on these two starch sources. Varga andKononoff (1999) evaluated 16 studies from 1992through 1998 where byproduct feeds were alsoincluded as well as processed feed sources, suchas corn and barley. The relationship betweenNFC and milk yield was poor and tended to becurvilinear. More information is needed tounderstand the composition of the NFC thatcontributes to the dietary DM as more than likelyrate of degradation has a major influence on itsutilization by the animal and ultimately milkyield (Beauchemin et al., 1997; Lykos et al.,1997). More direct measures, such as starch,sugar, pectin, glucans, and volatile fatty acids,are needed to better formulate rations for dairycows. Table 7 demonstrates how much feeds can

vary based on not only their concentration ofNDF, starch, and sugars, but in particular, theamount digested. Data of this nature, along withrate of degradation, are needed to betterformulate rations for lactating cows.

To emphasize the need for moreinformation of feeds for soluble carbohydrates,as well as digestibility, 35 corn silage samplesin the Capital region of PA were evaluated forvarious nutrient contents and digestibilityestimates (Table 8). The starch content of thesecorn silage samples was not that different;however, starch digestibility varied by almost20% units. This demonstrates that withoutknowing the extent of starch digestion, a starchvalue on its own may not be that meaningful.

Fructans

Fructans (fructosans) are water-solublechains of fructose found in the cell contents ofplants. They are the principal storagecarbohydrates of temperate cool season grasses.Depending on the species and environmentalconditions, temperate grass forage has beenreported to contain less than 1% and up to 30%fructan (Smith, 1973). Although mammals canutilize fructose, they do not have the enzymesto digest fructans (Nilsson et al., 1988). In therumen, both bacteria and protozoa fermentfructan (Ziolecki et al., 1992). Fructans can befermented to lactic acid during ensiling (Mullerand Lier, 1994) and in the rumen (Ziolecki etal., 1992). Additionally, rumen microbes candegrade fructan and store it as starch and utilizeit at a later time when other nutrients are nolonger available.

β-Glucans

The β-glucans are found in theendosperm and cell walls of grasses (Van Soest,1994). Barley and oats are major sources of β-glucans, containing from 4 to 12% by weight.Just as mammals cannot digest cellulose, theyalso cannot digest β-glucans. However, this


65

carbohydrate appears to be very rapidlyfermented in the rumen. In steers fed barley,ruminal in sacco disappearance of β-glucansvaried by lot of barley but was between 61.4and 70.4% of DM at time 0 and 93.8 to 96.2%of DM by 8 hours. Disappearance of b-glucanwas greater from dry rolled than from steamrolled barley (Engstrom et al., 1992).

Pectic Substances

Pectic substances are found chiefly in themiddle lamella of the plant cell wall and havebeen defined as non-starch polysaccharidessoluble in water, in chemicals which removedivalent cations (eg., Ca++, Mg++), and in diluteacids or bases that break covalent bonds (Doner,1986). In terms of sugar composition, theycontain a galacturonic acid backbone withrhamnose inserts, plus neutral sugar side chainsmade up largely of arabinose and galactose(Jarvis, 1984). There are differences amongplants and plant parts in the content andcomposition of pectic substances (Hall, 2000).Because they are very complex carbohydrates,analysis for pectic substances has been difficult,and there are few values available (Hall, 2000).By-products, such as citrus pulp, sugar beet pulp,and soybean hulls, are reported to contain 29,33.7, and 20% pectin, respectively. Amongforages, grasses are low in pectic substances (2to 5%), while legumes contain higher quantities(7 to 14%) (Chesson and Monro, 1982).

Pectins are rapidly and extensivelydegraded by rumen microbes. In vitrofermentation rates of 30 to 40% per hour forpectin have been reported (Hatfield and Weimer,1995). Pectin fermentation tends to produce highacetate to propionate ratios and relatively littleor no lactate. However, the organic acidcontribution from the fermentation of pecticsubstances depends on its sugar composition(Titgemeyer et al., 1992). The yield of microbesfrom pectin or pectic substances is not differentfrom starch (Strobel and Russell, 1986).However, the fermentation of pectin at low pH

(5.8) is reduced, resulting in a lower extent ofdegradation and up to 70% less microbial proteinproduced. A more acidic ruminal pH translatesinto decreased amounts of pectin fermented inthe rumen (Ben-Ghedalia et al., 1989) and adecrease in the amounts of microbial protein andVFA available to the animal. Unlike starch, thecow cannot digest pectic substances that escapethe rumen.

Lactating cows fed diets formulated tocontain a greater proportion of by-product feeds(citrus pulp and beet pulp), as compared to thosecontaining more starch (from corn products), hadlower intakes (Solomon et al., 2000), decreasedmilk protein percentage and yield (Mansfield etal., 1994; Solomon et al., 2000; Leiva et al.,2001), and increased milkfat percentage(Mansfield et al., 1994; Leiva et al., 2001).Mertens et al. (1994) compared the efficacy ofcarbohydrate sources in utilizing alfalfa silagenon-protein nitrogen (NPN) with cows fed dietscontaining 19% citrus pulp plus 19% highmoisture shell corn, or 39% high moisture shellcorn (% of dietary DM), with or withoutsupplementation with expeller soybean meal.The cows on the citrus diets showed greater milkand protein yield responses to supplementalrumen escape protein than did cows on the highmoisture corn diets, suggesting a poorer efficacyof NPN utilization with citrus.

It appears that altering the proportionsof sugars, starch, and soluble fiber in rations canalter intake, milk yield and composition, andfeed efficiency. In a study designed to examinethe microbial yield from the different NFC,isolated bermudagrass NDF (iNDF), and blendsof sucrose, citrus pectin, or corn starch and iNDFwere fermented in vitro (Hall and Herejk, 2001).The results show that the maximal yield of CPwas greatest from the starch fermentation (32.5mg), and lower, but similar, between pectin (28.1mg) and sucrose (27.6 mg). If the lower yieldsof CP from the fermentations of sucrose andpectin translate to reduced amino acid amountsavailable to the cow, it could explain the


66

reductions in milk protein in the animal feedingtrials. For pectin, it could explain the lowerefficacy of NPN utilization and the response tofeeding rumen escape protein.

Summary

To minimize the risk of ruminal acidosisand optimize the provision of rapidly fermentedcarbohydrate in the ration of dairy cows, abalance of NFC types is recommended. The2001 NRC recommended making adjustmentsfor NFC in the ration along with NDF and forageNDF as a percentage of DM (Table 9).Suggestions for target levels of NFC to beincluded in rations (% of DM) are: 5% sugars,10% soluble fiber, and 20% starch (Hall, 2000).If pectin/soluble fiber sources and sugars areincluded in rations as the predominant solublecarbohydrate sources, then additional rumenundegradable protein may be recommended.Starch appears to offer the greatest potential tooptimize microbial protein synthesis; however,feeding high levels of starch can lead to ruminalacidosis and digestive upset. Information isneeded to determine what extent sugars andstarches are interchangeable to deliver glucoseto the small intestine and what proportions ofsoluble fiber, sugars, and total effective NDF toinclude in rations to prevent digestive upsets.

References

Aguilar, A.A., N.E. Smith, and R.L. Baldwin.1984. Nutritional value of almond hulls for dairycows. J. Dairy Sci. 67:97-103.

Aldrich, M.M., L.D. Muller, G.A. Varga, andL.C. Griel Jr.. 1993. Nonstructural carbohydrateand protein effects on rumen fermentation,nutrient flow, and performance of dairy cows.J. Dairy Sci. 76:1091-1110.

Aldrich, J. 2002. Carbohydrates for lactatingcows. Akey Newsletter, August issue.

Batajoo, K.K. and R.D. Shaver. 1994. Impactof nonfiber carbohydrate on intake, digestion,and milk production by dairy cows. J. DairySci. 77:1580-1587.

Beauchemin,K.A., L.M. Rode, W.Z. Yang.1997.Effects of nonstructural carbohydrates andsource of cereal grain in high concentrate dietsof dairy cows J. Dairy Sci. 80:1640-1650.

Bell, M.C., W.D. Gallup, and C.K. Whitehair.1953. Value of urea nitrogen in rationscontaining different carbohydrate feeds. J. Anim.Sci. 12:787-797.

Ben-Ghedalia, D., E. Yosef, J. Miron, and Y. Est.1989. The effects of starch- and pectin-rich dietson quantitative aspects of digestion in sheep.Anim. Feed Sci. Technol. 24:289.

Broderick, G.A., N.D. Luchini, W.J. Smith, S.Reynal, G.A. Varga, and V.A. Ishler. 2000. Effectof replacing dietary starch with sucrose on milkproduction in lactating dairy cows. J. Dairy Sci.83(Suppl. 1):248 (Abstr.). Callaway, T.R., S.A. Martin, J.L. Wampler, N.S.Hill, and G.M. Hill. 1997. Malate content offorage varieties commonly fed to cattle. J. DairySci. 80:1651-1655.

Casper, D.P., D.J. Schingoethe, and W.A.Eisenbeisz. 1990. Response of early lactationcows to diets that vary in ruminal degradabilityof carbohydrates and amount of fat. J. DairySci.73:425-444.

Chesson, A. and J. A. Monro. 1982. Legumepectic substances and their degradation in theovine rumen. J. Sci. Food Agric. 33:852.

Dijkshoorn, W. 1973. Organic acids, and theirrole in ion uptake. Page 163 in Chemistry andBiochemistry of Herbage. Butler, G. W. and R.W. Bailey, eds. Academic Press, London andNew York.


67

Doner, L.W. 1986. Analytical methods fordetermining pectin composition. Page 13. inChemistry and Function of Pectins. M.L.Fishman, and J.J. Jen, eds. American ChemicalSociety, Washington, DC.

Engstrom, D.F., G.W. Mathison, and L.A.Goonewardene. 1992. Effect of B-glucan, starch,and fibre content and steam vs. dry rolling ofbarley grain on its degradability and utilisationby steers. Anim. Feed Sci. Technol. 37:33.

Fox, D.G., C.J. Sniffen, J.D. O’Connor, J.B.Russell, P.J.A. Van Soest. 1992. Netcarbohydrate and protein system for evaluatingcattle diets: III. Cattle requirements and dietadequacy. J. Anim. Sci. 70:3578-3596.

Galyean, M. L., D. G. Wagner, and F. N. Owens.1981. Dry matter and starch disappearance ofcorn and sorghum as influenced by particle sizeand processing. J. Dairy Sci. 64:1804-1812.

Hall, M.B. 2000. Neutral detergent-solublecarbohydrate: nutritional characteristics. Univ.Fla. Bull. 339, p 3-18.

Hall, M.B. 2002. How should we formulate fornon-NDF carbohydrates? Proc.12th Ann. Fla.Rum. Nutr. Symp. pp 44-49.

Hall, M.B. and C. Herejk. 2001. Differences inyields of microbial crude protein from in vitrofermentation of carbohydrates. J. Dairy Sci.84:2486-2493.

Hall, M.B. and H.H. Van Horn. 2002. Howshould we formulate for non-NDFcarbohydrates? Proc. 12th Ann. Fla. Rum. Nutr.Symp. pp 44-49. Hatfield, R.D. and P.J. Weimer. 1995.Degradation characteristics of isolated and in situcell wall lucerne pectic polysaccharides bymixed ruminal microbes. J. Sci. Food Agric.69:185.

Herrera-Saldana, R.E. and J.T. Huber. 1989.Influence of varying protein and starchdegradabilities on performance of lactating cowsJ. Dairy Sci. 72:1477-1483.

Herrera-Saldana, R., R. Gomez-Alarcon, M.Torabi, and J.T. Huber. 1990. Influence ofsynchronizing protein and starch degradation inthe rumen on nutrient utilization and microbialprotein synthesis. J. Dairy Sci. 73:142-148.

Hoover, W.H. and T.K. Miller. 1990. Balancingdairy cattle rations for total carbohydrates. Pacif.Anim. Nutr. Conf. p 173.

Hoover, W.H. and T.K. Miller. 1991.Carbohydrate-protein considerations in rationformulation. Proceedings Feed Dealer Seminars,NH. Huhtanen, P. and H. Khalili. 1991. Sucrosesupplements in cattle given grass silage-baseddiet. 3. Rumen pool size and digestion kinetics.Anim. Feed Sci. Technol. 33:275-287.

Huntington, G.B. 1997. Starch utilization byruminants: from basics to the bunk. J. Anim. Sci.75:852.

Jarvis, M. C. 1984. Structure and properties ofpectin gels in plant cell walls. Plant,Cell andEnvironment 7:153. Jones, D. F., W. H. Hoover, and T. K. MillerWebster. 1998. Effects of concentrations ofpeptides on microbial metabolism in continuousculture. J. Anim. Sci. 76:611-616. Khalili, H. and P. Huhtanen. 1991. Sucrosesupplements in cattle given grass silage-baseddiet. 2. Digestion of cell wall carbohydrates.Anim. Feed Sci. Technol. 33:263-273.

Kotarski, S. F., R. D. Waniska, and K. K. Thurn.1992. Starch hydrolysis by the ruminalmicroflora. J. Nutr. 122:178.


68

Leiva, E., M. B. Hall, and H. H. Van Horn. 2000.Performance of dairy cattle fed citrus pulp orcorn products as sources of neutral detergent-soluble carbohydrates. J. Dairy Sci. 83:2866-2875.

Lykos, T. and G.A. Varga. 1995. Effects ofprocessing method on degradationcharacteristics of protein and carbohydratesources in situ. J. Dairy Sci. 78:1789-1801.

Lykos, T., G.A. Varga, and D. Casper. 1997.Varying degradation rates of total nonstructuralcarbohydrates: Effects on ruminal fermentation,blood metabolites, and milk production andcomposition in high producing Holstein cows.J. Dairy Sci. 80:3341-3355.

Mansfield, H.R., M.D. Stern, and D.E. Otterby.1994. Effects of beet pulp and animal by-products on milk yield and in vitro fermentationby rumen microorganisms. J. Dairy Sci. 77:205-216.

Martin, L.C., C.B. Ammerman, P.R. Henry, andP.E. Loggins. 1981. Effect of level and form ofsupplemental energy and nitrogen utilization oflow quality roughage by sheep. J. Anim. Sci.53:479-488. McCarthy, R.D., Jr., T.H. Klusmeyer, J.L. Vicini,and J.H. Clark. 1989. Effects of source ofprotein and carbohydrate on ruminalfermentation and passage of nutrients to thesmall intestine of lactating cows. J. Dairy Sci.72:2002-2016.

Mertens, D. R., G. A. Broderick, and R. Simons.1994. Efficacy of carbohydrate sources forimproving utilization of N in alfalfa silage. J.Dairy Sci. 77(Suppl. 1):240. (Abstr.)

Muller, M. and D. Lier. 1994. Fermentation offructans by epiphytic lactic acid bacteria. J. Appl.Bacteriol. 76:406.

Nilsson, U., R. Oste, M. Jagerstad, and D.Birkhed. 1988. Cereal fructans: in vitro and invivo studies on availability in rats and humans.J. Nutr. 118:1325.

Nocek, J. E. 1988. Feeding management toenhance carbohydrate metabolism in ruminants.Cornell Nutr. Conf. Ithaca, NY.

Nocek, J. E. and J. B. Russell. 1988. Proteinand energy as an integrated system. Relationshipof ruminal protein and carbohydrate availabilityto microbial synthesis and milk production. J.Dairy Sci. 71:2070-2107.

Nombekela, S. W. and M. R. Murphy. 1995.Sucrose supplementation and feed intake of dairycows in early lactation. J. Dairy Sci. 78:880-885.

Oliveira, J. S., J. T. Huber, J. M. Simas, C. B.Theurer, and R. S. Swingle. 1995. Effect ofsorghum grain processing on site and extent ofdigestion of starch in lactating dairy cows. J.Dairy Sci. 78:1318-1327.

Orskov, E. R. 1986. Starch digestion andutilization in ruminants. J. Anim. Sci. 63:1624.

Seymour, W. M., J. Siciliano-Jones, J. E. Nocek,and J. E. English. 1993. Effect of grain typeand processing method on lactationalperformance. J.Dairy Sci. 76:308.

Sievert, S. J. and R. D. Shaver. 1993.Carbohydrate and Aspergillus oryzae effects onintake, digestion, and milk production by dairycows. J.Dairy Sci. 76:245

Smith, D. 1973. The nonstructuralcarbohydrates. Page 105 in Chemistry andBiochemistry of Herbage. Butler, G.W. and R.W.Bailey, eds. Academic Press, London.


69

Solomon, R., L. E. Chase, D. Ben-Ghedalia, andD. E. Bauman. 2000. The effect of nonstructuralcarbohydrate and addition of full fat extrudedsoybeans on the concentration of conjugatedlinoleic acid in the milk fat of dairy cows. J.Dairy Sci. 83:1322-1329. Strobel, H. J. and J. B. Russell. 1986. Effect ofpH and energy spilling on bacterial proteinsynthesis by carbohydrate-limited cultures ofmixed rumen bacteria. J. Dairy Sci. 69:2941-2947.

Sutton, J.D., J.A. Bines, S.V. Morant, D.J.Napper, and D.I.A.Givens. 1987. Comparisonof starchy and fibrous concentrates for milkproduction, energy utilization and hay intake byFriesian cows. J. Agric. Sci.Camb.109:375-386.

Therion, J. J., A. Kistner, and J. H. Kornelius.1997. Effect of pH on growth rates of rumenamylolytic and lactilytic bacteria. Appl. Environ.Microbiol. 44:428.

Titgemeyer, E. C., L. D. Bourquin, and G. C.Fahey, Jr. 1992. Disappearance of cell wallmonomeric components from fractionschemically isolated from alfalfa leaves and stemsfollowing in-situ ruminal digestion. J. Sci. FoodAgric. 58:451.

Vandercook, C.E. 1977. Organic acids. Page 208in Citrus Science and Technology: Nutrition,anatomy, chemical composition andbioregulation. Nagy, S., P. E. Shaw, and M. K.Veldhuis, eds. AVI Publishing Company, Inc.Westport, Conn.

Van Soest, P.J. 1994. Nutritional Ecology of theRuminant, 2nd Ed. Cornell University Press,Ithaca, NY.

Varga, G.A., T.W. Cassidy, V.A. Ishler, X.Markantonatos, N.D. Luchini, and G.A.Broderick. 2001. Effect of replacing dietarystarch with sucrose on nutrient utilization byruminal microorganisms during continuousculture fermentation. J. Dairy Sci.84(Suppl.1):290. (Abstr.)

Varga, G.A. and P. Kononoff. 1999. Dairyrationing using structural and nonstructuralcarbohydrates: from theory to practice.Southwest Nutr. Conf, AZ. pgs 77-91.

Varga, G.A. and T.J. Whitsel. 1991. Effect ofnonstructural to structural carbohydrate ratio onrate and extent of nutrient utilization in situ.Anim. Feed Sci. Tech 32:275-286.

Wilkerson, V. A. and B. P. Glenn. 1997. Energybalance in early lactation Holstein cows fed corngrains harvested and processed differently. J.Dairy Sci. 80:2487-2496.

Ziolecki, A., W. Guczynska, and M.Wojciechowicz. 1992. Some rumen bacteriadegrading fructan. Letters in AppliedMicrobiology 15:244.


70

Table 1. Neutral detergent soluble carbohydrate of selected feedstuffs and their fermentation endproducts.

Feed Major NFC Fractions1 End-products

Alfalfa hay Pectin and sugar Acetate and LactateAlfalfa silage Pectin, sugar, and organic acids Acetate, Lactate, ButyrateCorn silage Starch and organic acids LactateGrass hay Fructan and sugar LactateGrass silage Organic acids and sugar LactateCorn grain Starch LactateBarley grain Starch and b-glucans Lactate and AcetateWheat grain Starch and b-glucans Lactate and AcetateHigh moisture corn Starch and organic acids LactateHominy Starch LactateMolasses Sugar LactateWhey Sugar LactateBeet pulp Pectin AcetateCitrus pulp Pectin and sugar Acetate and LactateAlmond hulls Pectin and sugar Acetate and LactateSoy hulls Pectin AcetateWheat midds Starch Lactate

Table 2. Composition of the neutral detergent soluble carbohydrates (NDSC) of selected feedstuffs.1

Feedstuff sugar starch pectin VFA2

----------------------% of NDSC------------------------Alfalfa silage 0 24.5 33.0 42.5Grass hay 35.4 15.2 49.4 0Corn silage 0 71.3 0 28.7Barley 9.1 81.7 9.2 0Corn grain 20.9 80.0 0 0Beet pulp 33.7 1.8 64.5 0Soyhulls 18.8 18.8 62.4 0Soybean meal, 48% CP 28.2 28.2 43.6 0

1Adapted from Hoover and Miller (1990).2VFA = volatile fatty acids.

1NFC = nonfiber carbohydrates.


71

Table 3. Estimated rumen fermentation characteristics.1

Proportion of carbohydrates converted to substrate

Diet2 Acetate Propionate Butyrate Valerate

Soluble CHO3 R 0.69 0.20 0.10 0C 0.45 0.21 0.30 0.04

Starch R 0.59 0.14 0.20 0.06C 0.40 0.30 0.20 0.10

1From Van Soest, 1994.2R codes roughage diets: C codes diets containing more than 50% of a cereal-based concentrate diet.3Soluble carbohydrate (CHO) fraction includes organic acids and pectin.

Table 4. Least squares means of dextran yield at fermentation hour 4 and its proportion of sucroseinitially added to the vial (Hall, 2002).

Sucrose, mg1 Dextran, mg SE2 For Hour 4, Dextran/Initial Sucrose In Vial

65 9.3 1.59 0.14130 11.8 1.59 0.09195 15.0 1.59 0.08

1Milligrams of sucrose added per fermentation vial.2Standard error.

Table 5. Least squares means ± standard error of organic acid production at the hour of their peakyields (millimolar in 32 milliliters) (Hall, 2002).

TotalSucrose, Organic Acids Acetate Propionate Butyrate Lactate mg1 24 hour 24 hour 24 hour 24 hour 4 hour

65 25.8 ± 1.9 15.9 ± 1.2 7.4 ± 0.6 2.4 ± 0.2 0.12 ± 0.83 130 39.6 ± 1.9 22.7 ± 1.2 12.8 ± 0.6 4.0 ± 0.2 1.75 ± 0.83 195 53.8 ± 1.9 29.9 ± 1.2 17.6 ± 0.6 6.0 ± 0.2 4.82 ± 0.83

1Milligrams of sucrose added per fermentation vial.


72

Table 6. Changes in milk yield and composition with changes in sucrose and starch supplementation(Broderick et al., 2000).

Milk, Milk Fat, Milk Protein, FCM,Sucrose, % Starch, % DM intake, lb lb/day lb/day lb/day lb/day

0 7.5 54.0 85.8 3.24 2.73 89.3 2.5 5.0 56.4 89.1 3.37 2.82 93.0 5.0 2.5 57.3 88.2 3.64 2.84 96.8 7.5 0 57.3 86.9 3.57 2.82 95.2

Table 7. Fermentability of NDF, starch, and sugar for various feedstuffs (Hoover and Miller, 1990).

Ingredient NDF, % NDF dig., % Sugar + starch, % Sugar + starch dig., %

Beet pulp 48 60 12.8 70Hominy 23 45 54 65Soyhulls 66 56 5.3 70Wheat bran 33 58 47 80Wheat midds 42 60 32 80

Table 8. Starch, starch digestibility, and DM comparison of corn silage samples (n = 35).

Starch, % Starch digestibility, % Dry matter, %

34.4 75.0 41.8 33.5 74.6 42.0 27.2 78.3 39.8 31.2 91.2 34.7 31.2 93.5 30.7 30.4 88.3 33.8


73

Table 9. Minimum concentrations (% of DM) of total and forage NDF and maximumconcentrations (% of DM) of NFC (NRC, 2001).1

Min forage NDF Min NDF Max NFC Min ADF

19 25 44 1718 27 42 1817 29 40 1916 31 38 2015 33 36 21

1Values in this table are based on the assumption that actual feed composition has been measured.NFC = non fiber carbohydrates.

Figure 1. Plant carbohydrate fractions. The NDSC contain the carbohydrates listed and others of theappropriate solubility. ADF = acid detergent fiber, glucans = glucans, mono- + oligosaccharides =“sugars”, NDF = neutral detergent fiber, NDSC = neutral detergent-soluble carbohydrates, and NDSF =neutral detergent-soluble fiber. From Hall, 2000.

Plant Carbohydrates

CellContents

CellWall

HemicellulosesPecticSubstances

β -glucans

FructansStarchesMono+Oligo-saccharides

OrganicAcids

Cellulose

ADF

NDSF

NDSC

Galactans

Non-Starch Polysaccharides

NDF


74


75

New Developments in TMR Particle Size Measurement

P.J. Kononoff1 and A.J. Heinrichs2

1Renaissance Nutrition, Inc., Roaring Spring, PA2Department of Dairy and Animal Science,

The Pennsylvania State University, University Park, PA

Abstract

The Penn State TMR and Forage ParticleSeparator (PSPS), has become a routinetechnique used in forage and TMR particle sizeevaluation. The objective of this paper is tooutline the procedure to estimate particle sizeusing the PSPS and to briefly describe how itmay be used to further understand rumenfermentation and feeding behavior. Althoughproven useful, a large proportion of smaller TMRparticles may pass through both sieves of theoriginal device. Recently, a smaller sievemeasuring 1.18-mm has been added to improveaccuracy of estimating the smaller particlefraction. The PSPS is also a useful tool tomonitor sorting behavior of dairy cattle and canbe used to troubleshoot a variety of feeding,digestion, metabolic, and production-relatedproblems. Plotting the particle size distributionof both TMR and refusals using a newlydeveloped spreadsheet can also be useful inevaluating the degree of sorting behaviorexhibited by dairy cows.

Introduction

Since its introduction in 1996, ThePSPS, has become a routine technique used inforage and TMR particle size evaluation. Therapid acceptance of this method is due to thesimplicity of the procedure, low cost of analysis,and rapid determination of results. The PSPSwas originally constructed out of two sieves and

a bottom pan. Apertures of the two sievesmeasured 19.0 and 8.0-mm and had thicknessof 12.2 and 6.4-mm. With its simple constructionand size, the PSPS sieving method can befrequently implemented on the farm and used atthe time of harvest or feeding to determineparticle size of forages or TMR (Lammers et al.,1996). Even though the apparatus has beenwidely accepted and particle size measurementsusing the PSPS are now commonly reported inthe scientific literature, the TMR typicallycontain 40 to 60 % concentrate, most of whichpasses through the smallest sieve measuring 8.0-mm sieve. Recently, a smaller sieve measuring1.18-mm has been added to improve accuracyof estimating the smaller particle fraction. Thepurpose of this paper is to outline some of therecent developments and understandings in theTMR particle size measurement as estimated bythe PSPS.

Using The Penn State TMR and ForageParticle Separator

Sieving Frequency

The PSPS must be operated correctly inorder to achieve accurate results. During ourfield experience, we have observed that the rateof shaking, or frequency of moving the device,sometimes differs between users. Because thefrequency of shaking will influence particlemovement over enclosed sieves and willultimately affect particle size estimates, we have

1Contact at: Post Office Box 229, 481 Frederick Rd., Roaring Spring, PA 16673, (814) 793–2113, FAX (814) 793- 0037,Email: [email protected]


76

recently tested the effect of shaking frequencyon TMR particle size measurement (Kononoffet al., 2003). The unit used to express frequencyof movement is a Hertz (Hz), and one Hertz isequal to the number of full forwards andbackward movements within one second. Forexample, if the PSPS is shaken at a frequencyso that a total of 66 full movements (forwardand back) would occur within 60 seconds, themeasured frequency would be 1.1 Hz. In theexperiment, three frequencies, 0.9, 1.1 and 1.6Hz (slow, medium, and fast), were used todetermine the effect of sieving frequency onTMR particle size estimation.

Results of the experiment are listed inTable 1 and indicate that reducing sievingfrequency below 1.1 to 0.9 Hz resulted insignificantly more material being retained on the19.0-mm sieve and less on the 8.0 and 1.18-mmsieves. Consequently, these results yield a meanparticle length (MPL) that was significantlygreater when estimated at 0.9 Hz compared to1.1 Hz. In contrast, increasing sieving frequencyfrom 1.1 to 1.6 Hz did not result in significantdifferences in particle size measurements. Werecommend the PSPS to be shaken at 1.1 Hz(66 cycles/min) or greater. Practically, it isimportant to understand that one cannot shakethe PSPS too fast; however, movements whichare too slow may result is very different estimatesof particle size. It is recommended that operatorsof the device calibrate the frequency ofmovement over a distance of 17 cm for aspecified number of times. The number of fullmovements divided by time in seconds resultsin a frequency value that can be compared tothis recommendation.

Effect of Moisture on Particle SizeMeasurement

We know that sample moisture contentmay affect sieving properties, but it is notpractical to recommend that particle sizeestimation be conducted at a standardizedmoisture content (Finner et al., 1978). The PSPS

is designed to describe particle size of the feedoffered to the animal, thus samples should notbe chemically or physically altered beforesieving. Because sample moisture loss mayoccur during storage or transport, a study wascarried out to determine the effects of foragemoisture on particle size measurement, asestimated by the PSPS. In the experiment,samples of both alfalfa haylage and corn silagewere placed in a large forced air oven set at 55°C. Five times over a 48 hour period, samplesof different moisture content were removed fromthe oven and analyzed for particle size using thePSPS.

Tables 2 and 3 outline the effects offorage moisture content on particle sizeestimated for both alfalfa haylage and cornsilage. For alfalfa haylage, oven drying timesof 0, 2, 6, 12, and 48 hours resulted in moistureconcentrations of 57.4, 35.6, 10.4, 2.5, and 0.0%, respectively. Similarly, for corn silage, ovendrying times of 0, 3, 6, 18, and 48 hours resultedin moisture concentrations of 58.0, 34.4, 14.6,3.5, and 0.0 %, respectively. For alfalfa haylagesamples, particle size measurements were notsignificantly different between 57.4 and 35.6 %moisture, indicating that moisture loss insamples within this range will not affect particlesize measurements. Conversely, for corn silage,the amount of particle mass < 1.18-mm wassignificantly different between 58.0 and 34.4 %moisture and resulted in a small but significantdifference in MPL. These results suggest thatmoisture loss from corn silage may affect particlesize results; but these differences, whenobserved, are small. For alfalfa haylage,compared to 57.4 and 35.6% moisture, theamount of material > 19.0-mm was significantlylower in samples containing 10.4, 2.5, and 0.0%moisture, but this difference was not observedfor corn silage as most material > 19.0-mmcontained cob particles for which sizemeasurement appeared to be unaffected bymoisture content. For both forages, amount ofmaterial < 1.18-mm was greatest at 0.0 %moisture content, while MPL decreased with


77

decreasing moisture content. These results aresimilar to Finner et al. (1978) who suggestedthat completely drying a sample results inshattering of particles and further size reductionduring the sieving process. Differences in sievingresults associated with sample drying may havebeen due to brittle particles that shatter duringshaking or to decreased adhesion of smallparticles to larger ones when materials are dry.

Because it would be impractical torecommend a constant sample moisture formeasuring forage or TMR particle size duringfield measurement, it is advantageous to knowthat slight losses of moisture have only limitedeffects on measurements according to themoisture range of this study. It is recommendedthat samples be analyzed in the same physicalform as that fed to the animal and moisture lossin samples should be minimized. Based on ourresults, only small differences result whensample moisture loss is as much as 40 % of theoriginal sample.

Particle Size Recommendations

Table 4 contains the original particle sizerecommendations of corn silage, haylage, andTMR according to the Penn State TechnicalBulletin, DAS 96-20 (Heinrichs, 1996).Although proven useful, the basis for therecommendations was simple surveys and fieldobservations. More recently, a number ofexperiments have been conducted that haveattempted to further understand the effects offorage and TMR particle sizes, thus theserecommendations have been revised and arelisted in Table 5 and published in the newlyrevised Penn State Technical Bulletin, DAS 02-42 (Heinrichs and Kononoff, 2002).

TMR Particle Size and Chewing Activity

It is widely understood that dairy cattlerequire fiber in a coarse physical form becauseit is effective in stimulating chewing activity,saliva secretion, and maintaining normal rumen

health and function. As a result of thisunderstanding, many studies have attempted tomanipulate forage particle size to increasechewing activity and thereby increase thebuffering capability in the rumen.

Recently, Krause et al. (2002a,b)demonstrated that feeding an alfalfa silage basedTMR with approximately 23% greater than 19.0-mm, increased total chewing time 7.5 min/kg ofDM consumed compared to a TMR whichcontained less than 1% greater than 19.0-mm.Similarly, Kononoff and Heinrichs (2003a) notedthat increasing particle size of an alfalfa haylagebased TMR from 3 to 31% of the TMR greaterthan 19.0-mm increased chewing activity 6.7min/kg of DM consumed. Taken together, thesestudies support the suggestion that the proportionof material retained on the top sieve of the PSPSis positively correlated to chewing activity(Krause et al., 2002b).

Differences in total time spent eating andruminating are not always observed whenfeeding corn silage of different particle size. Balet al. (2000) observed that fine cut length reducedeating activity but did not affect time spentruminating; however, this has been reported byothers (Schwab et al., 2002). In studies in whichfeed refusals were chemically analyzed (Bal etal., 2000; Kononoff and Heinrichs, 2003b), dietsof longer particle size resulted in a higherconcentration of NDF, indicating greatertendency for sorting. Based on these and otherstudies, it is apparent that increasing particle sizeof a TMR can increase chewing activity;however , if the TMR is too long, this maypromote sorting behavior.

TMR Particle Size and Rumen pH

Nutritionists are often concerned aboutrumen pH because when pH levels fall below6.0, fiber digestion may be impeded (Russelland Wilson, 1996). Generally speaking, rumenpH is a function of lactic acid and VFAproduction and is buffered by saliva (Maekawa


78

et al., 2002). Because of this idea, it is commonpractice to feed diets of longer particle size andgreater amounts of effective fiber so that salviaproduction is stimulated. In support of thisunderstanding, Krause et al. (2002b) noted thatintake of particles > 19.0-mm was negativelycorrelated to the amount of time rumen pH spentbelow 5.8. However, we also know that forageshould not be harvested at excessively longlengths. Kononoff and Heinrichs (2003a)demonstrated that if diets are too long, eatingpatterns may be affected and may result in lowermean rumen pH even though chewing activitymay be increased. In this study, increasing theproportion of particles > 19.0-mm from 3 to 12%did result in increased chewing activity andrumen pH. However, increasing the proportionof particles to 31% resulted in depressions inmean rumen pH, which was thought to be a resultof changes in eating behavior and increasedsorting.

In light of these studies, it is apparentthat particle size can affect rumen pH; however,differences in particle size do not always resultin differences in mean rumen pH. Whenevaluating a diet to determine a possible risk ofsubclinical acidosis, we should review levels offiber and non-structural carbohydrates, alongwith their associated fermentability (Yang et al.,2001a).

New Particle Size Recommendations

Our original particle sizerecommendations have been revised, and theamount of TMR > 19.0 mm has been reducedfrom the original recommendation. This isbecause literature demonstrates that when rationsare correctly balanced for both fiber andnonstructural carbohydrates, no severedeleterious effects on rumen pH from reducedchewing activity are observed when rationscontained less than 6% of the particles > 19.0mm. Although many rations are commonly fedwith longer particles, there is increasing evidencethat these rations may result in increased sorting.

The recommended amount of materialbetween 8.0 and 19.0 mm was maintained andreflected the variation commonly observed oncommercial dairy farms (Heinrichs et al., 1999).As a result of the modification of the PSPS,particle size recommendations of material < 8.0mm is now further partitioned into the proportionbetween 1.18 and 8.0 mm and the proportion <1.18 mm. Given that no severe negative effectshave been observed in animals consumingrations which contained 30 to 50 % of the ration1.18 to 8.0 mm, our recommendations arelimited to that range (Kononoff and Heinrichs,2003 a,b).

Few published studies have included thenew sieve in particle size measurements;however, it is generally understood that particlesretained on a 1.18-mm sieve pass out of therumen slower than those not retained (Poppi etal., 1985). Recently, using a different method ofdry sieving, Yang et al. (2001b) demonstratedthat rumen outflow rate of particles less than1.18-mm averaged 5.57%/h compared to thoseretained on a 3.35-mm sieve, which had anaverage outflow rate of 1.75%/h. Thus, theadditional sieve should prove useful in furtherunderstanding of factors affecting rate ofpassage. Although the proportion of material <1.18 mm was not different within two publishedexperiments, no negative effects were observedwhen as much as 20% of the material was lessthan 1.18 mm (Kononoff and Heinrichs, 2003a,b). We are thus currently limited to recommendthat TMR contain no more than 20% of thematerial < 1.18 mm. This recommendation mayrequire further refinement based on the relativeamounts of soluble and degradable carbohydratecomponents in the TMR.

Understanding Eating Behavior by Using thePSPS

Rations That Promote Sorting Behavior

Particle size analysis is only part of thestory. Dairy rations must be balanced for proper


79

nutrient intake, and as we all know, there is nostandard ration. Given a ration, which appearsto be adequate in terms of its chemicalcomposition, particle size results can aid in ourunderstanding of how nutrients contained in theration will be consumed by the animal. Furtherunderstanding of how particle size affectsfeeding behavior is warranted because if theingestion of feed is more or less uniform, rumenfermentation will proceed with limitedfluctuation in acid production. Alternatively, ifingestion is selective, fluctuations in ruminalacid production may result in imbalance, leadingto abnormal or pathological conditions in therumen (Van Soest, 1994). As a result of this,when attempting to identify limits to production,particle size evaluation should be included inour trouble shooting procedures.

Lets look at an experimental situation inwhich lactating Holstein cows were fed one oftwo different rations that were chemicallyidentical but one considered “long” and the othershorter but adequate according to our currentguidelines. As presented in Table 6, the longration contained 16 % of the particles greaterthat 19.0-mm while the short ration contained7% of the particles > 19.0-mm. In both rations,the proportions of particles less that 1.18-mmwere similar, approximately 4%. Thecumulative percentage of material that fallsthrough each sieve (cumulative percentundersize) is also presented in Table 6 and willbe used for graphical purposes to be illustratedlater. What can these measurements tell us aboutpossible animal feeding behavior? In order fora ration, which looks good on paper, to work, itmust be consumed. There is increasingunderstanding that the amount of particles >19.0mm is correlated with eating time and chewingbehavior (Johnson et al., 2003). It is likely thatthis relationship is not perfectly linear, but asration particle size increases, total chewing timeshould also increase, however, only up to a point.Past this point, animals may then exhibitselective tendencies against longer, high fibercontaining particles. This is cause for concern

because the ration was balanced withunderstanding that the animals would consumethese longer particles.

Refusal Particle Size and Fiber Content

Now lets evaluate the particle size of theremaining refusal after the animals have had a24-hour access to the feed. Remember, we wantthe refusals to be similar to the feed originallyfed, if its drastically different (> 5% variation),animals may have sorted out the ration andconsumed something different that what weoriginally assumed when we balanced the ration.As seen in Table 7, the particle size of the longTMR is much longer that originally offered.From these observations, we can see that animalsconsuming the long ration consumed more grainas 1% of the DM was less than 1.18-mm,compared to 5% when animals consumed theshort ration.

To support our observation that animalsconsuming the long ration exhibited an increasedsorting tendancy, we can evaluate the NDFcontent of the feed remaining in the bunk at 8,16, and 24 hours after feeding. In our currentexample, the concentration of the original TMRfor both long and short TMR was identical, 33%of DM. Data presented in Table 8 demonstratethat the NDF contents of the refusals at 24 hoursafter feeding for animals consuming the longTMR were 36 and 44% for the short and longTMR, respectively. After 24 hours for the longTMR remaining in the bunk, it was 11percentage units greater in NDF than thatoriginally fed, indicating considerable sortingactivity. In comparison, the NDF content of theshort feed remaining in the bunk was measuredat 34, 36, and 36% at 8, 16, and 24 hours afterfeeding. This was very similar to the rationcontaining 33% NDF, indicating that very littlesorting activity occurred. These data supportour idea that particle size may affect how cowsconsume the ration and that feeding a ration thatis too long or coarse may result in large amountsof sorting activity.


80

Using a Spread Sheet to Compare the TMRand Refusals

Additional tools and guidelines to beused in the evaluation of particle size arebecoming available. A simple spreadsheet thatperforms all of the calculations, while alsographing the results, can be downloaded fromthe Penn State Dairy Nutrition website http://www.das.psu.edu/dcn/catforg/particle. Thisspreadsheet can be used to gather information,while at the same time aid in furtherunderstanding feeding behavior. The sizedistribution of forage and TMR particles doesnot follow a normal distribution pattern;however, it can be plotted as a straight-linedistribution using lognormal graphing paper.When plotting the data, sieve size is plotted onthe horizontal or X-axis, while the cumulativepercentage of material that falls below each sieveis entered on the vertical, or Y-axis. Once enteredinto the PSPS spreadsheet, these values are thenplotted, and a best-fit line is drawn between eachof the three points. For example in Figure 1,approximately 16% of the feed is more than19.0-mm (0.75 inches), and as a result, 84% ofthe material is undersized. With 50% beingretained on the 8.0-mm (0.31 inches) sieve, acumulative amount of 34% falls below 8.0-mm.Lastly, 23% is retained on the 1.18-mm sieve,and as a result, only 5% now falls less than 1.18-mm (0.05 inches).

Figures 1 and 2 illustrate the differencesin particle size between the originally fed TMRand the refusals. In both figures, the originalTMR particle size is plotted with dark pointsand a best-fit line is drawn in the same color(upper line). Refusal particle size is plotted withlighter points and a best-fit line is drawn in thesame color (lower line). Generally, we woulddesire that each point remain in, or close, to therecommended rectangles and that the associatedbest-fit lines be very close together. When rationsin which little sorting has occurred, such as thatillustrated in Figure 1, points falling on the sameX-axis will be close together. In rations that are

severely sorted, as is Figure 2, the lighter refusalline is much lower that that of the original TMRline. To determine if sorting has occurred,compare the distance between the two points ateach particle size (i.e. 1.18, 8.0, and 19.0-mm).In rations which result in a high degree of sorting,a greater spread between points will be observed.

Rations typically contain 40 to 60%concentrate, most of which passes through thesieve measuring 8.0-mm. Thus, the additionalsieve will allow us to more precisely describeparticle size by partitioning this fraction withthe smaller sieve. Secondly, in Figure 1, thedistance between the original TMR and refusalat >19.0-mm illustrates that some sorting of thematerial > 19.0-mm, but the proportion ofparticles < 1.18-mm was almost identical. Thisinformation indicated that although some sortingactivity occurred, especially with respect to thoseparticles > 19.0-mm, the proportion of grain inthe TMR and refusal was similar. In comparison,Figure 2 illustrates that the proportion ofparticles of all size ranges were different,including those less than < 1.18-mm. Thus byincluding the additional sieve into ourevaluation, we can more clearly see that thelonger ration resulted in even greater selectivetendency for the smaller grain particles.

Summary

The PSPS is a tool to quantitativelyestimate forage and TMR particle size. As withany diagnostic tool, users of the PSPS mustensure correct operation. Because differences inthe rate of shaking, or frequency of moving thedevice, may affect particle size estimations, wespecify that the PSPS be shaken at a frequencyof 1.1 Hz or greater (66 cycles/min) with a strokelength of 17 cm. Recently, we have alsoinvestigated the effects of sample moisture onmeasurements and determined that smallamounts of moisture loss from collected samplesmay affect particle size results, but thesedifferences, when observed, are small.Conversely, completely drying a sample resulted


81

in large differences in particle size results. As aresult of these observations, we recommend thatsamples be analyzed in the same form as fed tothe animal, but precautions should be made sothat moisture loss is minimized. In addition tothese analytical recommendations, we havediscussed the importance of particle sizeevaluation and outlined how the additional sieveand available spreadsheet may lead to furtherunderstanding of rumen fermentation andfeeding behavior.

References

ASAE. 2001. Method of determining andexpressing particle size of chopped foragematerials by sieving. In Standards. Am. Soc.Agric. Eng., St. Joseph, MI.

Bal, M.A., R.D. Shaver, A.G. Jirovec, K.J.Shinners, and J.G. Coors. 2000. Crop processingand chop length of corn silage: Effects on intake,digestion, and milk production by dairy cows.J. Dairy Sci. 83: 1264-1273.

Finner, M.F., J.E. Hardzinski, and L.L Pagel.1978. Evaluating particle length of choppedforages. ASAE paper No. 78-1047. Am. Soc. Ag.Eng., St. Joseph, MI.

Heinrichs, A.J. 1996. Evaluating particle size offorages and TMRs using the Penn State ParticleSize Separator. Technical Bulletin of ThePennsylvania State University, College ofAgriculture Science, Cooperative Extension.DAS 96-20.

Heinrichs, A.J., D.R. Buckmaster, and B.P.Lammers. 1999. Processing, mixing, and particlesize reduction of forages for dairy cattle. J. Anim.Sci. 77: 180-186.

Heinrichs, A.J., and P.J. Kononoff. 2002.Evaluating particle size of forages and TMRsusing the new Penn State Forage ParticleSeparator. Technical Bulletin of ThePennsylvania State University, College ofAgriculture Science, Cooperative Extension.DAS 02-42.

Johnson, L.M., J.H. Harrison, D. Davidson, W.C.Mahanna, and K. Shinners. 2003. Corn silagemanagement: effect of hybrid, chop length, andmechanical processing on digestion and energycontent. J. Dairy. Sci. 86: 208-231.

Kononoff, P.J., and A.J. Heinrichs. 2003a. Theeffect of decreasing alfalfa haylage particle sizeon effective fiber values and ruminalfermentation. J. Dairy Sci. In Press.

Kononoff, P.J., and A.J. Heinrichs. 2003b. Theeffect of decreasing corn silage particle size andthe inclusion of cottonseed hulls on effectivefiber values and ruminal fermentation. J. DairySci. In Press.

Kononoff, P.J., A.J. Heinrichs, and D. ABuckmaster. 2003. Modification of the PennState forage and TMR separator and the effectsof moisture content on its measurements. J.Dairy Sci. In Press.

Krause, K.M., D.K. Combs, and K.A.Beauchemin. 2002a. Effects of forage particlesize and grain fermentability in midlactationcows. I. Milk production and diet digestibility.J. Dairy Sci. 85: 1936-1946.

Krause, K.M., D.K. Combs, and K.A.Beauchemin. 2002b. Effects of forage particlesize and grain fermentability in midlactationcows. II. Ruminal pH and Chewing activity. J.Dairy Sci. 85: 1947-1957.

Lammers, B.P., D.R. Buckmaster, and A.J.Heinrichs. 1996. A simplified method for theanalysis of particle sizes of forage and totalmixed rations. J. Dairy Sci. 79: 922-928.


82

Maekawa, M., K.A. Beauchemin, and D.A.Christensen. 2002. Effect of concentrate leveland feeding management on chewing activities,saliva production, and ruminal pH of lactatingdairy cows. J. Dairy Sci. 85: 1165–1175.

Poppi, D.P., R.E. Hendrickson, and D.J. Minson.1985. The relative resistance to escape of leafand stem particles from the rumen of cattle. J.Argric. Sci. 105: 9-14.

Russell, J.B., and D.B. Wilson. 1996. Why areruminal cellulolytic bacteria unable to digestcellulose at low pH? J. Dairy Sci. 79: 1503-1509.

Schwab, E.C., R.D. Shaver, K.J. Shinners, J.G.Lauer, and J.G. Coors. 2002. Processing andchop length effects in brown-midrib corn silageon intake, digestion, and milk production bydairy cows. J. Dairy Sci. 85: 613-623.

Van Soest, P.J. 1994. Nutritional Ecology of theRuminant, 2nd Edition. Comstock PublishingAssociated, a division of Cornell UniversityPress, Ithaca, NY.

Yang, W.Z., K.A. Beauchemin, and L.A. Rode.2001a. Effects of grain processing, forage toconcentrate ration, and forage particle size onrumen pH and digestion by dairy cattle. J. DairySci. 84: 2203-2216.

Yang, W.Z., K.A. Beauchemin, and L.A. Rode.2001b. Barley processing, forage:concentrateration, and forage particle size on chewing anddigesta passage in lactating cows. J. Dairy Sci.84: 2709-2720.

SEM


83

Table 1. Effects of sieving frequency on particle size measurements of TMR1 samples as measured bythe modified Penn State Particle Separator.

Frequency (Hz)2

Particle size (mm) 0.9 1.1 1.6 P - value

TMR1,3

> 19.0 40.9a 6.4b 6.9b 0.02 19.0 – 8.0 24.6b 42.9a 43.8a 0.05 8.0 – 1.18 31.5 36.7 35.3 0.31 < 1.18 3.0b 14.0a 14.0a 0.001 MPL (mm)4 11.2a 5.8b 5.7b 0.011TMR containing 50:50 forage to concentrate ratio and 9.5% grass hay, 25.3% corn silage, and 14.6%alfalfa haylage (% of DM).2Means in the same row with different superscripts differ (P < 0.05).346.0 + 1.6 % moisture.4MPL = geometric mean length as calculated by the ASAE (2001).

Table 2. Effects of alfalfa haylage moisture content on particle size measurements according to thePenn State Particle Separator shaken at 1.2 Hz with a stroke length of 17 cm.

Particle size Percent Moisture1

(mm) 57.4 35.6 10.4 2.5 0 SEM2 P - value

> 19.0 61.5a 63.0a 45.2b 40.3b 27.5c 2.87 < 0.00119.0 – 8.0 25.3c 24.4c 35.4b 37.3a,b 44.5a 2.25 < 0.0018.0 – 1.18 11.3d 10.6d 15.1c 18.0b 22.6a 0.69 < 0.001< 1.18 1.9c 2.1c 4.3b 4.4b 5.4a 0.27 < 0.001

MPL (mm)3 17.7a 17.9a 13.7b 12.6b 10.3c 0.54 < 0.001

1Means in the same row with different superscripts differ (P < 0.05).2SEM = standard of mean.3MPL = geometric mean length as calculated by the ASAE (2001).

M = sta


84

Table 3. Effects of corn silage moisture content on particle size measurements according to the PennState Particle Separator shaken at 1.2 Hz with a stroke length of 17 cm.

Particle size Percent Moisture1

(mm) 58.0 34.4 14.6 3.47 0 SEM2 P - value

> 19.0 14.3 11.0 9.5 9.6 12.9 2.16 0.3219.0 – 8.0 74.0a 74.5a 73.2a 70.4a 52.3b 2.08 < 0.001 8.0 – 1.18 11.4d 13.1c,d 15.4b,c 18.0b 31.5a 1.36 < 0.001 < 1.18 0.23d 1.36c 2.0b 2.0b 3.4a 0.16 < 0.001

MPL (mm)3 12.1a 11.2b 10.6b,c 10.2c 8.62d 0.33 < 0.0011Means in the same row with different superscripts differ (P < 0.05).2SEM = standard error of mean.3MPL = geometric mean length as calculated by the ASAE (2001).

Table 4. Original recommended forage and TMR particle size recommendations as measured by thePenn State Particle Separator (Heinrichs, 1996).

Type Corn Silage1 Haylage2 TMR

Percentage retained > 19.0 mm 2 - 4 15 - 25 6 - 10 19.0 – 8.0 mm 40 - 50 30 - 40 30 - 50 8.0 – 1.18 mm 40 - 50 40 - 50 40 - 601This recommendation applies if corn silage is not the sole forage used in the TMR; if cornsilage ischopped and rolled, 10 to 15% may be > 19.0-mm.2This recommendation applies if haylage is contained in a bunker silo; if it is in a sealed silo, 10 to 15%may be > 19.0-mm.

Table 5. Forage and TMR particle size recommendations1 based on two experiments that fed eitheralfalfa haylage or corn silage, with or without cottonseed hulls, to cows in early lactation.

TypeCorn Silage Haylage TMR

Sieve Size % of DM retained > 19.0 mm 5 ± 3 15 ± 5 5 ± 3 19.0 – 8.0 mm 55 ± 10 60 ± 15 40 ± 10 8.0 – 1.18 mm 40 ± 10 30 ± 10 40 ± 10 < 1.18 mm < 5 < 5 < 20

MPL (mm)2 8 ± 2 10 ± 2 5 ± 21As measured by the Penn State Particle Separator (Heirichs and Kononoff, 2002).2MPL = geometric mean length as calculated by the ASAE (2001).

S


85

Table 6. Particle size measurements of one short and one long corn silage based TMR.

Short TMR Long TMRCumulative Cumulative

Sieve Particle Size Percent Percent Percent PercentNumber (mm) Retained Undersized Retained Undersized

1 > 19.0 7 93 16 84 2 19.0 – 8.0 56 37 50 34 3 8.0 – 1.18 34 4 30 4 4 < 1.18 4 - 4 -

Table 7. Particle size measurements of refusals for animals fed one short and one long corn silagebased TMR.

Short TMR Long TMRCumulative Cumulative

Sieve Particle Size Percent Percent Percent PercentNumber (mm) Retained Undersized Retained Undersized

1 > 19.0 25 75 60 40 2 19.0 – 8.0 40 35 24 16 3 8.0 – 1.18 31 4 15 1 4 < 1.18 5 - 1 -

Table 8. The effect of feeding two rations of different particle size on the NDF concentration ofrefusals.

Hours After Feeding Short TMR Long TMR % NDF (DM basis)

0 33 33 8 34 3516 36 4124 36 44


86

Figure 1. Particle size distribution of short TMR(upper line) and refusals (lower line) as measuredby the Penn State Particle Separator.

Figure 2. Particle size distribution of long TMR(upper line) and refusals (lower line) as measuredby the Penn State Particle Separator.

Particle Size Analysis

99.9

99

95

90

80

70

60

50

40

30

20

10

5

2

0.1

1

0.750.310.070.01 0.1 1 10

Particle size (inches)

Cum

ulat

ive

Perc

enta

ge U

nder

size

d

targetsample 1sample 2

DAIRY& ANIMAL SCIENCE TMR

Particle Size Analysis

99.9

99

95

90

80

70

60

50

40

30

20

10

5

2

0.1

1

0.750.310.070.01 0.1 1 10

Particle size (inches)

Cum

ulat

ive

Perc

enta

ge U

nder

size

d

targetsample 1sample 2

DAIRY& ANIMAL SCIENCETMR


87

Problems with Sorting in Total Mixed Rations

Lou Armentano 1and Claudia Leonardi

1Contact at: 1675 Observatory Dr., Madison, WI 53706-1284, (608) 263-3490, FAX (608) 263-9412, Email:[email protected]

Department of Dairy ScienceUniversity of Wisconsin

Abstract

Cows need to consume physicallyeffective fiber to maintain health. In order forthis to happen, cows must be offered anadequately coarse diet and cows must consumethe coarse fractions of the diet. A cow’s nose isabout 5 inches wide; she can use it to push awayfeed particles that are a few inches long,especially on a dry diet. Some cows seem to sortextremely against these very long particles.Medium length fibers probably contribute tophysically effective fiber and are much less likelyto be sorted. Therefore, to insure an adequateintake of physically effective fiber by all cowsin the herd, moist diets with intermediate particlesize should be best. Intermediate particles arethe finer particles on the top Penn State screenand some large percentage of the particles onthe middle screen. One solution is well-managedhay crop silage.

Introduction

This paper focuses on how we make surethat cows consume an adequate amount ofphysically effective fiber. For some people, theterm physically effective fiber seems redundant,because to them, effective fiber means the samething. Physically effective fiber is used toemphasize that adding fiber works in both aphysical and chemical way. Coarse fiber

provides both important physical and chemicalproperties, and fine fiber provides only thechemical properties.

Physically effective fiber is essential toprovide rumen fill and prevent abomasaldisplacement. Physically effective fiber formsa thick rumen mat that slows passage of smallerfibrous feed particles and increases theirdigestion. Finally, physically effective fiberenhances rumination and salivation and providesa better buffer source for the rumen. The latteris very important for cows eating large quantitiesof feed, much of which is rapidly fermented toorganic acids in the rumen.

Nutritionists know that there are severaldiets on the farm: the one they told their clientsto feed, the one actually being offered to thecows, and the one that the cow actuallyconsumes. The last category also must includevariations imposed by each cow on her own diet.This cow variation becomes an important factorin delivering the minimum amount of adequatephysically effective fiber to each cow (or at leastmost cows) in the herd, which is different thanoffering a diet with adequate physical fiber andalso different than having the average cowconsume a diet with adequate physicallyeffective fiber.


88

Basic Cow Behavior and Variation

Our first experiment on sorting(Leonardi and Armentano, 2003) sought toexamine how cows vary in their sorting behaviorand to explore other issues that we thought mightbe important. The design used 24 cows so thatindividual cow behavior could be measured. Sixdiets were fed in a Latin square design, and wemeasured sorting on several days within each ofthe six periods. Both of these features allowedus to distinguish cow effects from simple day-to-day variation.

Sorting was measured by determiningthe physical distribution of the feed refused byeach individually housed cow. Actual as-fedintake of each screen (total offered – amount inrefusals) was calculated directly. This numberwas divided by the predicted as-fed intake forthat same screen, where predicted intake for ascreen is the as-fed intake for that cow (totaloffered – total refused) multiplied by the as feddistribution of the total mixed ration (TMR) onthat screen. So the predicted intake is the ‘valueon paper’ for the diet and the numbers we showare the actual cow intakes of each screen,presented as a percentage of the predicted. Whenexpressing data this way, screens with less than100% are being sorted against and screens morethan 100% are being selectively consumed. Ifone screen is less than 100%, some other screenwill have to be more than 100%. In general,screens with a small amount on them (like thecoarsest screen) can deviate from 100% morethan screens with a lot of material on them, andit is important to remember this to avoid over-interpreting the data.

Figure 1 shows the behavior of theindividual cows averaged across all sixtreatments (more about treatment effects will bediscussed later). Note that the bottom line onFigure 1 represents the top screen from the 6-screen Wisconsin separator. In comparing thisto the Penn State screens used in the field,remember that the top screen of the Penn State

shaker is roughly equivalent to the top twoscreens of the Wisconsin separator. Another wayof saying this is that the longest particles thatwe have labeled as Y1 in Figure 1 are the coarserparticles on the Penn State top screen, Y2represents the finer particles on the Penn Statetop screen, and Y3 is roughly equivalent to thesecond (middle) Penn State box.

The most important thing to note fromFigure 1 is that for those cows that did sort, theyalmost always sorted against the very longestparticles, much less against the next coarsestparticles, and selectively consumed the particlesfrom the finest screen and pan. Although somecows (such as cow 3930 and 4178) showed verylittle sorting, and one cow (4454) showed reversesorting, this trend is pretty strong. Also, therewas a great deal of variation in how muchindividual cows sorted against the top screen.Cows 3894, 4156, 4454, and 4460 were repeatoffenders, tending to sort more than the othercows across treatments.

Forage Effects

Figure 2 shows the same data averagedacross cows for each of the six treatments(20HQC through 40LQC). If you ignore whatthe treatments are for now, you can see moresorting occurred against the longest screens andthe effect is rather linear, with sorting againstlong, no sorting in the middle, and preferentialconsumption of the finest material. The HQ andLQ refer to the quality of hay in the diet, wherethe lower quality hay (LQ) was 44.6% NDFand the higher quality hay (HQ) was 34.5%NDF. Diets contained 20% hay and 20% haylage(the diets marked 20), and the hay was eitherleft long (HQL and LQL) or chopped (HQC orLQC) by overmixing in a horizontal auger TMRmixer. The 40HQC and 40HQL diets were 40%hay with no haylage.

The first surprise is that quality did nothave an effect on sorting. We thought cowswould be more likely to sort against lower quality


89

hay. It didn’t happen. That doesn’t mean that itwill never happen, and it is important to notethat the hay was not moldy or unsound (or evenall that high in NDF). Also, when a low NDFhay is fed, NDF tends to be distributedthroughout the screens somewhat more evenly,especially if non-forage sources of NDF arebrought in to increase dietary NDF. In that case,even if sorting doesn’t change, the effect ofsorting on NDF intake might. But basically, wedidn’t see high hay quality as a solution tosorting.

You can see that the treatments 40HQCand 40LQC cause the most sorting. These arevery dry diets (90% DM) with no silage, whichmost nutritionists have probably learned toavoid. Comparing the particle size distributionof the hay, chopped hay, and haylage used inthese diets (Figure 3), we can see that haychopped by processing for 15 minutes in a TMRmixer (middle set of bars) results in a bimodalor ‘dumbell’ distribution with a lot of very longfibers but also a lot of fines. Haylage on theother hand, produces something more like a bellshaped curve with more particles on the Y2 andY3 screens. The distribution and/or the DMcontent of the 40% chopped hay diets (40HQCand 40LQC) made these diets very sortable. Themixed rations for these dietary treatments(Figure 4) had roughly 15% on the Y1 and Y2screens combined, but the chopped hay diets hadmore fines than the diets with haylage and lessmaterial on the middle screens.

What happens when we take a sortablediet like these last two and combine them with apicky cow (like cow 4454) can be seen in Figure5. The treatment effects of Figure 2 are greatlyexaggerated in this cow. Look closely at the lastset of bars and you will see that this cow atenone of the longest particles on diet 40LQC(40% lower quality chopped hay with nohaylage). This was based on observations of twoconsecutive days, starting five days after firstoffering the diet. Its possible that she didn’t sortas badly on other days, and she certainly didn’t

sort any worse. But, this cow was obviouslygetting no physically effective fiber from thelongest particles on these two days. In fact, thisextreme sorting reduced the mean particle sizefrom 0.13 inches (3.35 mm) to 0.09 inches (2.26mm), or a reduction of roughly one third. Later,we’ll show how different particle sizedistributions can be affected differently whenassigning similar levels of sorting on eachscreen.

Importance of ‘Top’ versus ‘Middle’ Screens

Recommendations generally call for aminimum percentage (7 to 8% is a commonrecommendation) of a TMR on the top PennState screen. This is obtained by using forageswith even higher percentage on the top screenand mixers that don’t reduce this value muchduring mixing. Its important to realize that thePenn State screens don’t really allow calculatingmean particle size, and even the cumbersomeWisconsin screens only ‘estimate’ particle size.Also, we rarely account for DM and never fibercontent on the screens. In general, within aforage type, the mean particle size is going tobe highly correlated with the fraction on the topPenn State screen. For example, a coarse haylagewill have more on the top screen and less on thepan. However, as in the case of chopped hayversus haylage or long hay in Figure 3, thiscorrelation may not hold well across differentforage types. Therefore, although we makerecommendations based on percentage on thetop screen, it may well be that a similar meanparticle size obtained by a large percentage onthe middle screens may perform similarly to thesame mean particle size achieved with morelonger particles. What may be different is thatthe latter diet may be more susceptible to sorting,and therefore, it will not result in delivery to thecows rumen what appeared on paper.

We conducted an experiment usingoatlage processed to achieve different particlelengths (Leonardi et al., 2001). The goal was toobtain two TMR with equal mean particle sizes


90

but with one diet getting more of its meanparticle size from very long particles and theother from more intermediate fibers. How wedid this is shown in Figure 6. Both silages wereprocessed in the field to provide long or mediumparticle length. Then, the long oat silage wasfinely chopped just prior to feeding. Thisprovided three particle lengths for diets that wenamed long oat silage (LOS), medium oat silage(MOS), and fine from long oat silage (FLOS).We also made a fine silage daily from themedium oat silage (FMOS) to compare toFLOS. The last comparison was just to makesure there weren’t any effects due to ensiling (andthere weren’t). The last, and most importanttreatment, was obtained by mixing long oatsilage with chopped oat silage to make a mixedoat silage. The TMR with this oat silage wascalled LFLOS to represent a mixture of long oatsilage and fine (from long) oat silage. Theimportant comparison is between the MOS andthe mixed fine and long oat silage diets.

The mean particle size of the diets andtheir particle distribution are shown in Figure 7.Note that the MOS and LFLOS diets have aboutthe same mean particle size in the diets offered(5.19 and 5.39 mm, respectively). Also, notethat the LFLOS mixed diet has more longparticles (Y1 and Y2) and less medium particles(Y3 and Y4) compared to the medium MOSdiet. In Figure 8, rumination is plotted againstmean particle size of the diets as consumed (thatis, after taking into account each cows sorting).Mean particle size does a rather good job ofpredicting rumination, an indicator of physicallyeffective fiber content of the diet. Although notshown, the percentage of feed on the top twoscreens also would be an indicator of physicallyeffective fiber. My interpretation of this trial isthat the medium particles probably do contributeto the physically effective fiber in the diet, butthis is based more on the inability to showsomething special about the long fibers.

Extrapolating From Our Observations

If the assumption that mean particle sizeis a good estimator of the physical effectivenessof the diet is true, two things must follow. First,the value of material on the top and middlescreens toward physical effectiveness is simplya function of how they affect mean particle size,something that we can estimate mathematically,at least for well characterized particle sizedistributions. Secondly, we can also estimatethe impact of sorting on different types of dietsto estimate the mean particle size consumed.

Based on their contribution to meanparticle size, it is possible to roughly estimatethat 5% on the Y1 screen is equal to 7% on theY2 screen or 10% on the Y3 screen. In thiscalculation, the ‘excess’ was coming from theY5 screen. So, one way to look at this is that ittakes about twice as much increase in the middlescreen to offset a decrease in the coarsestparticles in the as-fed diets. But, what if a coweats most of the middle screen particles andextensively sorts the very coarse particles? Is itpossible that, at least for the extreme sortingcows, that the diets with mean particle sizecoming from the middle screen will actuallydeliver more physically effective fiber to the cowthan a diet with similar mean particle sizecoming from long fibers?

We did this calculation on two‘imaginary’ diets shown in Figure 9. Note thatTMR 2 has 6.6% on Y1 versus 0% for TMR 1,but TMR 1 compensates by having 28% on Y3versus 15% for TMR 2 ( a difference of 13%).We applied sorting coefficients of the cow eating60% of Y1 offered, 80% of Y2, and 90% of Y3.For TMR 1, this was compensated for by 105%consumption of Y5 and 110% consumption ofPan and for TMR 2 by 108% consumption ofY5 and 115% consumption of pan. It is easy tosee that similar sorting has a larger effect onTMR 2, reducing the mean particle size from4.35 to 3.72 mm (about 15%) compared to 3.99for TMR 1 (8% reduction). Also, notice that


91

the change in mean particle size is fairly large ifyou compare it to the wide range achieved inthe oat silage trial shown in Figure 8. Clearly,by looking at the degree of sorting expressed bysome cows on diets very susceptible to sorting,the changes could be much more extreme.

Note that we applied the same sorting tothese two diets. Sorting in the oatlage trial isshown in Figure 10. Sorting was significantlyincreased as mean particle size increased.Obviously, increasing mean particle size offeredwill still increase mean particle size consumed,just not as much as the paper diets tell you. Therewas also a significant difference in the patternof sorting caused by the particle size distribution.

Addition of Water

It is a common practice to add water todry diets fed to dairy cattle. Our first experimentcompared hay to haylage to achieve differentparticle size distributions, and changes in sortingcould be attributed to either distribution or DM.In the oatlage trial (Leonardi et al., 2001) therewas some evidence that distribution had an effectand moisture was not a factor in this experimentas all diets were about 48% DM. What aboutthe effect of moisture, independent of particlesize distribution? We examined this by usingchopped hay in diets with and without addedwater (Leonardi et al., 2002). Realize that whenwater is added to diets, feeds will cling together,increasing the apparent particle length of thediets. We wouldn’t expect this artificial increasein mean particle size to have any effect on thephysical effectivenss of the diet offered;however, it could affect sorting. In these trials,we measured sorting using the same as-fedshaking apparatus as before, but after shaking,we determined the DM. We did not use wetsieving of the diets but that probably would havebeen a better approach to determine true meanparticle size.

Dry matter content of the diet wasreduced from 80.2% to 64.3% by adding water.

There was some statistically significantreduction in sorting due to water addition (Figure11). We added as much water as we couldwithout making the rations sloppy, andpresumably on the farm, much less water wouldbe added. Perhaps even more importantly, whenindividual cow sorting was measured, both thedry diets and the wet diets had some cows thatsorted extremely against the longest particles ontwo consecutive days (Figure 12). It isinteresting that they weren’t the same cows inthis switchover trial. The net result is that wateraddition was slightly helpful in reducingsorting, but some cows still sort strongly againstthe coarsest particles.

Summary

The most important observations toremember are that most of the impact on sortingis on the coarsest particles or the top Y1 screenof the Wisconsin separator. This screencorrelates to the coarsest particles on the PennState top screen. The Y2 and Y3 screens aremuch less likely to be sorted against andtherefore deliver somewhat less physicallyeffective fiber but to a higher percentage of cows.

Using dry hay or straw processed toprovide very long particles can increase the meanparticle size of the diet with a small amount ofDM. However, the use of this long fiber is morelikely to result in the diet on paperunderestimating the mean particle size of the dietconsumed, especially for some cows. The drynature of these materials probably makes thiseven worse, and adding water does not cure theproblem.

No one has yet established whether cowsthat sort excessively against long fibers are moreprone to the physiological problems associatedwith offering diets low in physically effectivefiber. Also, we are not aware of evidence to showthat providing adequate mean particle size fromlarge amounts of medium sized fibers is betterthan providing the same particle length from a


92

‘bimodal’ mixture of short and long fiber. Therecommendations that follow are what weperceive as common sense interpretations of thedata that we have with dietary effects on sorting,on individual cow behavior, and on the cleartendency for some cows to sort worst againstthe very long fibers.

Making sure that a cow is offeredadequate amounts of physically effective fiberis not rocket science. This can usually be doneby simple good management practices whenharvesting and ensiling silage and haylage andby avoiding overmixing of diets. This seems tobe the cheapest and most reliable way to promoteadequate consumption of physically effectivefiber by dairy cows. Adding coarsely choppedor long hay is another common way of addingphysically effective fiber. While this certainlyworks on the diet offered, and your Penn statebox numbers will look good on paper, it willonly work on the cow that uniformly eats thediet.

To paraphrase Abe Lincoln, you can getsome of the cows to eat all their long particlessome of the time, but you cannot get all of thecows to eat all of the long particles all of thetime. In other words, some cows sort. We areless concerned about the average amount ofsorting in the herd being fed diets with normalmoisture than we are about the extreme sortingthat can be done by a few cows. Adding verylong fiber to the diet (or free choice feeding oflong hay) raises the average physical fiber intakeof the herd, but is adding this ‘avoidable’ fiberthe best way to protect cows that are extremelypicky? Probably not.

References

Leonardi, C. and L.E. Armentano. 2003. Effectof quantity, quality and length of alfalfa hay onselective consumption by dairy cows. J. DairySci. 86:557-564.

Leonardi, C., L.E. Armentano, and K.J. Shinners.2001. Effect of different particle size distributionof oat silage on feeding behavior and productiveperformance of dairy cattle. J. Dairy Sci. 84(Suppl. 1):199. (Abstr.)

Leonardi, C., F. Giannico, and L.E. Armentano.2002. Effect of water addition on selectiveconsumption (sorting) of dry diets by dairy cattle.J. Dairy Sci. 85 (Suppl. 1):62. (Abstr.)


93

020406080

100120

20HQC 20HQL 20LQC 20LQL 40HQC 40LQC

Treatments

%in

take

/ the

oret

ical

inta

ke Y1Y2Y3Y4Y5PAN

020406080

100120


Treatments

%in

take

/ the

oret

ical

inta

ke Y1Y2Y3Y4Y5PAN

Figure 1. Sorting of different sized particles (Y1 = longest, Pan = shortest) by individual cows.Numbers less than 100% indicate particles are sorted against, numbers greater than 100% means thatparticles are preferentially consumed. These data points were averaged across 12 days, 2 days eachfrom six different treatments. Reprinted from Leonardi and Armentano, 2003.

Figure 2. Sorting plotted by treatment for sorting of each particle size from longest (Y1) to shortest(Pan). Treatments are 20HQC = 20% high quality chopped hay, 20HQL = 20% high quality long hay,20LQC = 20% low quality chopped hay, 20LQL = 20% low quality long hay, 40HQC = 40% highquality chopped hay, and 40LQC = 40% low quality chopped hay. Reprinted from Leonardi andArmentano, 2003.

50

60

70

80

90

100

110

120

3816

3894

3930

3968

4075

4156

4178

4180

4202

4273

4275

4327

4422

4424

4438

4442

4444

4448

4454

4459

4460

4466

4469

4470

Cows

% in

take

/ th

eore

tical

inta

ke

Y1Y2Y3Y4Y5PAN

No sorting No sorting

High sorting

50

60

70

80

90

100

110

120

3816

3894

3930

3968

4075

4156

4178

4180

4202

4273

4275

4327

4422

4424

4438

4442

4444

4448

4454

4459

4460

4466

4469

4470

Cows

% in

take

/ th

eore

tical

inta

ke

Y1Y2Y3Y4Y5PAN

No sorting No sorting

High sorting


94

Figure 3. Particle size distribution of high quality long hay (HQL), high quality chopped hay(HQC) and haylage. Particle size shown from longest (Y1) to shortest (Pan). Taken from Leonardiand Armentano, 2003.

Figure 4. Particle size distribution of total mixed rations from longest (Y1) to shortest (Pan). Treat-ments are 20HQC = 20% high quality chopped hay, 20 HQL = 2-% high quality long hay, 20LQC =20% low quality chopped hay, 20LQL=20% low quality long hay, 40HQC=40% high qualitychopped hay, and 40LQC=40% low quality chopped hay. Reprinted from Leonardi and Armentano,2003.

0

10

20

30

40

50

60

HQL hay HQC hay Haylage

% o

n ea

ch sc

reen YI

Y2Y3Y4Y5Pan

Long HayChopped Hay

0

10

20

30

40

50

60

HQL hay HQC hay Haylage

% o

n ea

ch sc

reen YI

Y2Y3Y4Y5Pan

Long HayChopped Hay

0

5

10

15

20

25

30

35

40


% o

n ea

ch sc

reen Y1

Y2Y3Y4Y5Pan


95

Figure 5. Sorting by a single cow (4454) of the six diets. Data represents sorting on twoconsecutive days. Treatments are 20HQC = 20% high quality chopped hay, 20HQL = 2-% highquality long hay, 20LQC = 2-% low quality chopped hay, 20LQL = 20% low quality long hay,40HQC = 40% high quality chopped hay, and 40LQC = 40% low quality chopped hay. Taken fromLeonardi and Armentano, 2003.

Figure 6. Explanation of how oat silage diets with different particle lengths were obtained (Leonardiet al., 2001).

0

20

40

60

80

100

120

140


Treatments

% in

take

/ the

oret

ical

inta

keY1Y2Y3Y4Y5PAN

0

20

40

60

80

100

120

140


Treatments

% in

take

/ the

oret

ical

inta

keY1Y2Y3Y4Y5PAN

Oats in field

MediumLong Fermented in separate bags

Daily processingprior to feeding

Fine Oat Silage (FMOS)Fine Oat Silage (FLOS)

No further processing

Long Oat Silage(LOS)

Medium Oat Silage(MOS)

Mixed Fine and Long Oat

Silage (LFLOS)

Oats in field


Oats in field













Silage (LFLOS)


Silage (LFLOS)


96

Figure 7. Particle size distribution and mean particle size (MPS, mm) of total mixed rations(Leonardi et al., 2001). FMOS = fine from medium ensiled oat silage, FLOS = fine from longensiled oat silage, MOS = medium ensiled oat silage, LFLOS = mixed long and fine from longensiled oat silage mixed to give particle size similar to MOS, and LOS = long ensiled oat silage.

24

25

26

27

28

29

4 4.5 5 5.5 6 6.5

Mean Particle Size Consumed (mm)

Rum

inat

ing

(min

/kg

DM

I)

Mixed (LFLOS)

Medium (MOS)

24

25

26

27

28

29

4 4.5 5 5.5 6 6.5

Mean Particle Size Consumed (mm)

Rum

inat

ing

(min

/kg

DM

I)

Mixed (LFLOS)

Medium (MOS)

Figure 8. Relationship of ruminating time to mean particle size of diets consumed (Leonardi et al.,2001). Particle size of diets has been adjusted to reflect sorting. LFLOS = long from fine plus longoat silage, and MOS = medium oat silage. Rumination is proportional to mean particle size(P < 0.01) and no difference exists between medium (MOS) or mixed (LFLOS, P = 0.46).

0

5

10

15

20

25

30

35

F M O S F LO S M O S LF LO S LO S

Y 1Y 2Y 3Y 4Y 5P an

4.35 4.46 5.19 5.39 6.68

FINE MEDIUM LONG

MPS

0

5

10

15

20

25

30

35


Y 1Y 2Y 3Y 4Y 5P an

4.35 4.46 5.19 5.39 6.68

FINE MEDIUM LONG

0

5

10

15

20

25

30

35


Y 1Y 2Y 3Y 4Y 5P an

4.35 4.46 5.19 5.39 6.68

FINE MEDIUM LONG

4.35 4.464.35 4.46 5.19 5.395.19 5.39 6.68

FINE MEDIUM LONG

MPSMPS


97

0%

10%

20%

30%

40%

50%

TMR 1Offered

TMR 2Offered

TMR 1Consumed

TMR 2Consumed

Y1Y2Y3Y4Y5Pan

MPS = 4.35 4.35 3.99 3.72

Figure 9. Effect of sorting of two different total mixed rations (TMR) with equal as fed mean particlesize (MPS). Sorting was consuming 60% of top screen (Y1), 80% of second screen (Y2), 90% ofscreen Y3, 100% of screen Y4, 105 to 108% of screen Y5, and 110 to 115% of pan.

Figure 10. Sorting in the oat silage trial (Leonardi et al., 2001). Y1=longest particles throughpan=shortest. FMOS = fine from medium ensiled oat silage, FLS = fine from long ensiled oat silage,MOS = medium ensiled oat silage, LFLOS=mixed long and fine from long ensiled oat silage mixedto give particle size similar to MOS, and LOS = long ensiled oat silage.

Sorting

80

85

90

95

100

105

110

FMOS FLOS MOS LFLOS LOS

PANY5Y4Y3Y2Y1

NO SORTING

%

NO SORTINGNO SORTING

%


98

Figure 11. Effect of adding water on sorting of particles from longest (Y1) to shortest (pan)(Leonardi et al., 2002).* = P < 0.05.** = P < 0.01.*** = P < 0.001.

0

20

40

60

80

100

120

Y1 Y2 Y3 Y4 Y5 pan

DRY WET

Screens

% (

DM

bas

i s) *

******

0

20

40

60

80

100

120

Y1 Y2 Y3 Y4 Y5 pan

DRY WET

Screens

% (

DM

bas

i s) *

******


99

Figure 12. Sorting by individual cows when fed the wet or dry diets (Leonardi et al., 2002). Y1 =longest particles through pan = shortest.

0

20

40

60

80

100

120

3890

3919

40774078

4088

4130

41484155

4161

4218

4279

4366437

5439

8442

044

794490

4535

Y1 Y2 Y3 Y4 Y5 PANSo

rtin

g, %

DM

bas

is

Cows #

DRY diet

0

20

40

60

80

100

120

3890

3919

40774078

4088

4130

41484155

4161

4218

4279

4366437

5439

8442

044

794490

4535

Y1 Y2 Y3 Y4 Y5 PANSo

rtin

g, %

DM

bas

is

Cows #

DRY diet

0

20

40

60

80

100

120

38903919

40774078

40884130

41484155

41614218

42794366

43754398

44204479

44904535

Y1 Y2 Y3 Y4 Y 5 PA N

Sort

ing,

% D

M b

asis

Cow s #

W ET diet

0

20

40

60

80

100

120

38903919

40774078

40884130

41484155

41614218

42794366

43754398

44204479

44904535

Y1 Y2 Y3 Y4 Y 5 PA N

Sort

ing,

% D

M b

asis

Cow s #

W ET diet


100


101

Forage Alternatives

Michael S. Allen1 and Jennifer A. Voelker

1Contact at: 2265G Anthony Hall, Department of Animal Science, Michigan State University, East Lansing, MI 48824-1225, (517) 432-1386, FAX (517) 432-0147, Email: [email protected]

Department of Animal ScienceMichigan State University

Abstract

Forage alternatives can be used to reducethe filling effects of diets and to stretch availableforages when their supply or quality is limited.Dietary space vacated by forage is filledprimarily by high fiber byproduct feeds or bycereal grains; the extent to which they cansubstitute for forage depends primarily on theirdigestion characteristics - they should be slowlyfermented with high total tract digestibility. Dietfermentability must be limited to prevent excessruminal acid production. Adequate coarse fiberparticles should be included in the diet forformation of the rumen mat to entrap and retainsmall potentially fermentable fiber particles,increasing their digestibility.

Introduction

Forage availability is sometimes limitedbecause of poor growing conditions or becauseof insufficient land base on individual farms. Inaddition, forage fiber can limit feed intake,particularly for high producing cows, and foragequality is highly variable. Therefore there iscontinual interest in alternatives to forages indiets for dairy cows. Decreasing the foragecontent of diets requires careful selection ofingredients for use in the dietary space created.Because cereal grains must be limited in dietsto avoid ruminal acidosis, byproduct feeds withhigh fiber content are the most commonalternatives to forages and can be used to replace

some, but not all, forage in diets of lactatingcows. These byproduct feeds are referred to asnon-forage fiber sources (NFFS) to distinguishthem from other byproduct feeds. However,cereal grains can also be used in place of foragesto some extent if fermentability of starch isreduced. Forage alternatives should beconsidered based on their ability to functionallysubstitute for forage. This paper discusses howforages function in diets of dairy cows andfactors to consider when selecting foragealternatives.

Forage function

Forages are unique compared to otherdietary ingredients because they provide longfibrous particles that are retained in the rumenlonger, and tend to ferment more slowly, thansmaller feed particles. This provides a consistentsource of fuels to microbes in the rumen, as wellas a basal supply of fuels to the liver andmammary gland over time, allowing greater milkyield. Some long fibrous particles are necessaryfor formation of the rumen mat, which entrapssmall particles, increasing their ruminaldigestibility. An adequate mass of digesta in therumen is required to stimulate chewing, whichincreases secretion of salivary buffers.

A high basal fuel supply from fiberfermentation in the rumen is highly desirable anddependent on the digestion characteristics of thefiber source. Forage fiber should have a high


102

turnover rate from both digestion and passageto maximize fuel supply, while minimizing thefilling effect of the fiber over time. Increasedpassage of particulate matter is also expected toincrease efficiency of microbial proteinproduction by increasing passage of attachedmicrobes before lysis occurs. However, foragesvary considerably in their digestioncharacteristics, and it is often desirable to findalternatives for forages if their quality is poor.

Forages also dilute rapidly fermentablefeeds, such as cereal grains in diets, thuspreventing excessive ruminal fermentation.However, this function is not unique to forages,and the ability of feeds to slow ruminalfermentation while providing high total tractdigestibility determines their value as a foragealternative. The concentration of forage in dietscan often be decreased while maintaining orenhancing feed intake, milk yield, and health.The goal is to provide cows low-fill, highlyfermentable diets that result in consistentfermentation over time.

Filling dietary space vacated by forages

Non-forage fiber sources

The primary advantage of NFFS is thatthey generally have higher energy concentrationsand are much less filling than forages; feedintake is limited to a greater extent when theneutral detergent fiber (NDF) concentration ofthe diet increases from forage compared to NFFS(Allen, 2000). They are commonly used asforage alternatives and vary widely in cost,availability, and chemical composition acrosstype and location, and over time. Compared tocereal grains, NFFS are less likely to result inruminal acidosis because they generally fermentmore slowly, are not fermented to lactic acid,and because their rate of fermentation slows asruminal pH declines. Like forages, NFFS oftenresult in decreased propionate and increasedacetate production in the rumen when substitutedfor cereal grains (Harvatine at al., 2002;

Ipharraguerre et al., 2002; Voelker and Allen,2002). A slower rate of fermentation anddecreased propionate production might allowincreased meal size and greater feed intake whenthey are limited by absorbed propionate (Allen,2000). However, substitution of NFFS for forageusually decreases the ratio of acetate topropionate concentrations in the rumen (Weidnerand Grant, 1994; Clark and Armentano, 1997;Mowrey et al., 1999), as well as ruminal pH(Weidner and Grant, 1994; Harvatine et al.,2002). Wisconsin researchers reported thatNFFS were only about one half as effective asalfalfa NDF at elevating milk fat concentrationand were less effective at stimulating chewingactivity when substituted for grain in low foragediets (Swain and Armentano, 1994). AnotherWisconsin experiment found that NFFS wereonly 27% as effective at increasing milk fatconcentration as alfalfa silage NDF (Pereira etal., 1999). Therefore, when replacing foragewith NFFS, all non-forage components must beconsidered for their effects on diet fermentability.

The fiber concentration (NDF andsoluble fiber) of most NFFS are in the rangeobserved for most forages (40 to 60% on a DMbasis), but some NFFS have NDF concentrationsexceeding 75% (oat hulls, cottonseed hulls, andground corn cobs). The primary difference inthe fiber of NFFS compared to forages is thatparticle size of NFFS is smaller, so they cannotprovide long fibrous particles to form a rumenmat. While some NFFS, such as cottonseed hullsand oat hulls, are poor sources of energy andprotein and function only to dilute rapidlyfermentable grains, most NFFS are good sourcesof fermentable fiber. Some have additional valuebecause they provide fatty acids and(or)ruminally undegraded protein. Soyhulls and beetpulp are excellent sources of fermentable fiberbut have low protein and fatty acidconcentrations. Crude protein concentrations ofdistiller’s grains, brewer’s grains, corn glutenfeed, and whole linted cottonseeds are higherthan for most forages, ranging from 24 to 30%of DM. Whole cottonseeds, and to a lesser extent


103

distiller’s grains, provide additional energy asfatty acids. The extent to which each NFFSshould be used as a forage alternative dependson its availability and price relative to otherenergy and protein sources, as well as its abilityto substitute functionally for forage.

Recent research with lactating cowssuggests that NFFS not only reduce the amountof starch fermented in the rumen whensubstituted for grain but also reduce the ruminaldigestibility of the remaining starch. Illinoisresearchers substituted soyhulls for dry groundcorn at 0, 10, 20, 30, and 40% of dietary DMand reported that ruminal digestibility of non-structural carbohydrates tended to decreaselinearly from nearly 30% to less than 5% withoutreducing ruminal or total tract digestibility oforganic matter (Ipharraguerre et al., 2002). Werecently reported similar effects on ruminalstarch digestibility when pelleted beet pulp wassubstituted for high moisture corn at 0, 6, 12,and 24% of dietary DM (Voelker and Allen,2002). The amount of starch truly digested inthe rumen decreased dramatically from 8.4 to1.5 lb/day, partly because of the expectedreduction in starch intake from lower dietarystarch concentration but also from an unexpectedreduction in true ruminal starch digestibilityfrom 47 to 17% as beet pulp replaced highmoisture corn in diets. There was no effect oftreatment on total tract starch digestibility,despite this large reduction in ruminal starchdigestion because of compensatory postruminalstarch digestion. In addition, true digestibilityof organic matter in the rumen was not affectedby treatment because of increased NDFdigestibility; this increase resulted in increasedapparent total tract digestibility of organic matteras beet pulp replaced corn in the diet. Thedramatic reduction in ruminal starch digestibilitywas because rate of ruminal starch digestiondecreased from 11.3 to 1.7%/h and rate of starchpassage from the rumen increased from 15.9 to23.5 %/h. It is not known if other NFFS havethe same effects on rate of digestion and passagefrom the rumen, so this should be investigated

further because it has important implications forlimiting ruminal starch digestion. It might notbe desirable to reduce ruminal starch digestibilityto the extent observed with the highestsubstitution rates of beet pulp or soyhulls, butthe linear responses obtained in these twoexperiments suggest that NFFS can be used atlower substitution rates to manipulate ruminalstarch fermentability and site of starch digestion.Therefore, substitution of NFFS for both forageand grain might be necessary to limit dietfermentability as the forage content of the dietdecreases.

A limitation of NFFS as a foragealternative is that particle length is not adequatefor formation of the rumen mat. In addition,passage rate is generally greater for NFFS NDFthan for forage NDF (Firkins, 1997), and ruminaldigestibility can decrease if poor matdevelopment reduces their retention time in therumen. If adequate long particles are providedby coarse forage in the diet, NFFS can contributeto mat formation because they tend to be buoyantand become entrapped. Another limitation ofNFFS is that, compared to cereal grains, NFFSprovide fewer glucogenic precursors, which canbe limiting for milk production, particularly forhigh producing cows.

Cereal grains

Although one doesn’t normally think ofcereal grains as forage alternatives, they can alsobe used in the dietary space vacated by forage.Cereal grains contribute less to digesta mass inthe rumen than NFFS because their retentiontime and water holding capacity are lower. Theextent to which a cereal grain can replace foragein the diet without reducing feed intake orcausing ruminal acidosis depends on itsdigestion characteristics. Cereal grains withmoderate ruminal fermentability and highwhole-tract digestibility, such as dry groundcorn, are desirable to include in diets in place offorage. They are less filling than forages andhave higher energy concentrations. They have


104

an advantage over NFFS because moreglucogenic precursors are provided from starchdigestion compared to fiber digestion. Highlyfermentable starch sources, such as finely groundhigh moisture corn, barley, wheat, and bakerywaste, should be avoided to prevent excessivefermentation. Starch sources with low total tractstarch digestibility, such as unprocessed sorghumor coarsely cracked corn, should also be avoidedbecause they reduce dietary energy density.

Reduced ruminal fermentability andpropionate production might result in increasedmeal size and feed intake. A recent experimentfrom our laboratory showed that meal sizeincreased from 4.2 to 5.1 lb and feed intakeincreased from 45.8 to 49.5 lb/day, as dry groundcorn replaced high moisture corn in high starchdiets fed to lactating dairy cows (Oba and Allen,2003a). True ruminal starch digestibilitydecreased from 71% for high moisture corn to47% for dry ground corn with no effect oftreatment on total tract starch digestibility (95%;Oba and Allen, 2003b). Although milk yieldwas similar for the two treatments, fat-correctedmilk yield tended to be 6.6 lb/day higher for dryground corn compared to high moisture corntreatment because of higher milk fatconcentration. When forage is replaced withcereal grains, those grains with slower ruminalfermentation are likely to improve feed intakeand milk production, depending on theirdigestion characteristics and other dietarycharacteristics as discussed below.

Considerations for feeding low-forage diets

The forage concentration of diets can bereduced by forage alternatives in manysituations; in some, forage alternatives canreplace a large fraction of the forage, allowingvery low forage diets to be fed. Ohio researcherssuggested that diets with forage NDF contentsas low as 9 to 11% with whole cottonseeds and14 to 16% without whole cottonseeds can be fedto mid-lactation cows when non-structuralcarbohydrates are limited to 30% of DM (Slater

et al., 2000). Nebraska researchers reported thata wet corn milling product (40% NDF, 23%crude protein) has the potential to replace all ofthe concentrate and up to 45% of the forage inthe diet for lactating cows (Boddugari et al.,2001). The forage content of diets can beminimized by considering the followingrecommendations.

Include forages with high fiber concentrationsand long particles

Forages vary greatly in fiberconcentration and particle size, and long fibrousparticles are what distinguish them from otherfeed ingredients. Therefore, the forage contentof diets can be reduced if forage with higher fiberconcentrations and longer particles are used. Forinstance, a diet formulated using a forage with50% NDF can contain 20% less forage toprovide the same forage NDF concentrationcompared to a forage with 40% NDF, becausemore fiber is provided per unit of forage weight.However, it is also important to considerdigestibility of forage NDF; increasing forageNDF by delaying harvest will result in decreasedrate and extent of NDF digestion, which willdecrease the basal energy supply to microbes andto the animal. Across cuttings and maturities,NDF concentration is poorly related to digestioncharacteristics of NDF for alfalfa (Allen andOba, 1996), allowing opportunity to purchaseor select forage with both high NDF content andhigh NDF digestibility. Corn hybrids with highNDF and high NDF digestibility exist and canbe selected when land is limited for forageproduction. Forages can be compared for NDFdigestibility by in vitro rumen fermentation orby evaluating the lignin concentration as apercentage of NDF; within a forage type,digestibility of NDF decreases as lignificationof NDF increases (Allen, 2001).

Less forage is also needed in the diet ifit is long or coarsely chopped. Fibrous particlesof finely chopped alfalfa or corn silage are notas effective at forming a rumen mat as are longer


105

particles from coarsely chopped forage.Addition of coarse fiber as chopped hayimproved digestion of soyhulls in severalexperiments (Grant, 1997); chopped hayincreased rumen mat consistency and ruminationtime and decreased passage rate of wet corngluten feed in another experiment (Allen andGrant, 2000). The length and concentration oflong particles required in diets is not constantand depends on their digestion characteristics,as well as characteristics of other dietaryingredients.

Avoid rapidly fermented feeds

Rapidly fermented feeds containingsugars and starch should be limited. It is veryimportant to consider the fermentability of alldietary ingredients when feeding low foragediets because most forage alternatives are morefermentable than forages and are also lesseffective at stimulating chewing. Limitingrapidly fermented feeds, such as finely groundhigh moisture corn and molasses, will reducethe risk of acidosis.

Manage to avoid slug feeding

Diets with less coarse fiber require lesschewing during eating and can be consumedmore quickly, resulting in larger meals beforesatiety occurs. Although this might be desirableto increase feed intake in some situations, it canalso result in ruminal acidosis if fermentationacids cannot be absorbed quickly enough andthe buffering capacity of ruminal contents isgreatly exceeded. Overcrowding and feedingto an empty bunk increases competition amonganimals and encourages slug-feeding and shouldbe avoided when feeding minimum forage diets.

Include dietary buffers

Although efforts should be made toreduce fermentability by limiting highlyfermentable feeds and to reduce slug-feeding byproviding adequate bunk space, inclusion of

buffers makes sense as another way to reducethe risk of acidosis.

Limit NFFS with high fat content

Concentration and availability ofpolyunsaturated fatty acids might limit inclusionrates for some NFFS, such as whole cottonseedsand distiller’s grain, because of potential milkfat depression from the production of trans fattyacids by ruminal fermentation. Limitationsdepend on fatty acid concentration andcomposition of the NFFS and other dietaryingredients.

Conclusion

Numerous strategies exist to reducedietary forage content. Some coarse forage isneeded in diets for retention of small particlesin the rumen, increasing their digestibility. Whenreplacing forage with other feeds, it is importantto consider digestion characteristics of not onlythe forage alternative but of all dietarycomponents.

References

Allen, M.S. 2000. Effects of diet on short-termregulation of feed intake by lactating dairy cattle.J. Dairy Sci. 83:1598-1624.

Allen, M.S. 2001. Dietary factors that affect drymatter intake. Proc. Tri State Dairy NutritionConference, The Ohio State University,Columbus. pgs. 45-58.

Allen, D.M. and R.J. Grant. 2000. Interactionsbetween forage and wet corn gluten feed assources of fiber in diets for lactating dairy cows.J. Dairy Sci. 83:322-331.

Allen, M.S. and M. Oba. 1996. Fiberdigestibility of forages. Proc. 57th MinnesotaNutrition Conf. and Protiva TechnicalSymposium, Extension Special Programs, Univ.of Minnesota, St. Paul, MN. pgs. 151-171.


106

Boddugari, K., R.J. Grant, R. Stock, and M.Lewis. 2001. Maximal replacement of forageand concentrate with a new wet corn millingproduct for lactating dairy cows. J. Dairy Sci.84:873-884.

Clark, P.W. and L.E. Armentano. 1997.Replacement of alfalfa neutral detergent fiberwith a combination of nonforage fiber sources.J. Dairy Sci. 80:675-680.

Firkins, J.L. 1997. Effects of feeding nonforagefiber sources on site of fiber digestion. J. DairySci. 80:1426-1437.

Grant, R.J. 1997. Interactions among foragesand nonforage fiber sources. J. Dairy Sci.80:1438-1446.

Harvatine, D.I., J.L. Firkins, and M.L. Eastridge.2002. Whole linted cottonseed as a foragesubstitute fed with ground or steam-flaked corn:digestibility and performance.J. Dairy Sci. 85:1976-1987.

Ipharraguerre, I.R., Z. Shabi, J.H. Clark, andD.E. Freeman. 2002. Ruminal fermentation andnutrient digestion by dairy cows fed varyingamounts of soyhulls as a replacement for corngrain. J. Dairy Sci. 85:2890-2904.

Mowrey, A., M.R. Ellersieck, and J.N. Spain.1999. Effect of fibrous by-products onproduction and ruminal fermentation in lactatingdairy cows. J. Dairy Sci. 82:2709-2915.

Oba, M. and M.S. Allen. 2003a. Effects of corngrain conservation method on feeding behaviorand productivity of lactating dairy cows at twodietary starch concentrations. J. Dairy Sci.86:174-183.

Oba, M. and M.S. Allen. 2003b. Effects of corngrain conservation method on ruminal digestionkinetics for lactating dairy cows at two dietarystarch concentrations. J. Dairy Sci. 86:184-194.

Pereira, M.N., E.F. Garrett, G.R. Oetzel, and L.E.Armentano. 1999. Partial replacement of foragewith nonforage fiber sources in lactating cowdiets. I. Performance and health. J. Dairy Sci.82:2716-2730.

Slater, A.L., M.L. Eastridge, J.L. Firkins, andL.J. Bidinger. 2000. Effects of starch source andlevel of forage neutral detergent fiber onperformance by dairy cows. J. Dairy Sci. 83:313-321.

Swain, S.M. and L.E. Armentano. 1994.Quantitative evaluation of fiber from nonforagesources used to replace alfalfa silage. J. DairySci. 77:2318-2331.

Voelker, J.A. and M.S. Allen. 2002. Effects oflevel of pelleted beet pulp substituted for high-moisture corn on rumen digestion kinetics andmicrobial protein efficiency in lactating dairycows. J. Dairy Sci. 85(Suppl. 1):404. (Abstr.)

Weidner, S.J. and R.J. Grant. 1994. Alteredruminal mat consistency by high percentages ofsoybean hulls fed to lactating dairy cows. J.Dairy Sci. 77:522-532.


107

Nutritional Considerations for Jersey Cows

Christopher K. Reynolds1

Department of Animal SciencesThe Ohio State University

Abstract

There are numerous suggestions in thepopular press that the nutritional requirementsand management of Jersey cows can not be basedsolely on recommendations for Holstein cows.However, in many cases, these recommendationsare based more on anecdotal evidence andpractical experience than on results of studiesor data integrations designed to test breeddifferences. Many of these issues need to beclarified. For example, the high milk fat contentof the Jersey is often the basis of suggestionsthat Jersey cows have a higher fiber requirementthan Holstein cows, but other recommendationssuggest that Jersey cows can tolerate higherdietary starch levels than Holstein cows andshould be fed less fiber to achieve maximalintake. The Jersey is often purported to have ahigher dry matter (DM) intake capacity perkilogram of body weight (BW) than the Holstein,but in many cases, this is related to differencesin milk energy output per kilogram of BW orBW0.75, which will increase appetite. However,there are data suggesting a higher rate of passagefor the Jersey, which could allow greater intake,as well as affecting starch and NDF digestion.It is generally accepted that Jersey cows are moreprone to milk fever, perhaps due to differencein vitamin D receptors; therefore, it has beensuggested that their target dietary cation:aniondifference be lower during transition. There arenumerous reports in the literature that dry Jerseycattle have a higher maintenance energy

requirement than Holstein cattle, but in acalorimetry trial at Beltsville, MD, there waslittle difference in the energy metabolism oflactating Jersey and Holstein cows when the datawere expressed on a BW0.75 basis. There was atendency for the Jersey cows to partition moremetabolizable energy towards milk than bodytissue, but on a BW0.75 basis, the incremental useof metabolizable energy for net energy wasremarkably similar for the two breeds.

Introduction

A brief perusal of web sites and thepopular press articles provides immediateevidence the Jersey breed has a passionate andenthusiastic following. Numerous surveys ofherd records and data bases suggest a long listof advantages compared to other breeds, suchas easier calving, reduced mastitis and lameness,and greater longevity. They are purported tohave a good ‘temperament’ and be easy tohandle, be ‘easy’ on pastures (they weigh less)and need less acreage, and are believed to bemore heat tolerant than other breeds. Ofparticular relevance for the present paper, theirmilk has a higher milk fat and thus energycontent, as well as higher protein content andmanufacturing quality (CAS, 1978). Therefore,Jersey milk has more value and Jersey cowsoften produce more milk energy per kilogramof body weight than her larger cousins. Thishigh milk energy output is generallyaccompanied by a greater DM intake per

1Contact at: OARDC, 1680 Madison Ave., Wooster, OH 44676, (330) 263-3793, FAX (330) 263-3949, Email:[email protected]


108

kilogram of BW than for the Holstein and otherlarge frame breeds. Other than a lower yieldper cow, there are few ‘howevers’, but Jerseybull calves typically have less value and thebreed tends to have more susceptibility to milkfever. In addition, they are reported to have ahigher maintenance energy requirement.

What are the nutritional issues for theJersey? Is she simply a mini-Holstein or shouldshe be given special considerations whenformulating rations? Evidence suggests thatthere are certain nuances to the Jersey that shouldbe considered, but the experimental evidenceforming the basis of Jersey ‘guidelines’ is oftensketchy, anecdotal, or based on practicalexperience, thus harder to document. Issues tobe considered in the present paper include:

1. Feed efficiency of Jersey cows compared toother breeds,

2. The potential effect of high intake levels ondigestibility,

3. Effective fibre requirements for high milkfat synthesis,

4. Susceptibility to milk fever and transitionmanagement,

5. The impact of high milk energy yield onenergy metabolism, and

6. Do Jerseys have a higher maintenance energyrequirement?

Feed Efficiency

Whilst Jersey cows produce less milkthan Holstein cows, relative to BW or‘metabolic’ BW (BW0.75, or MBW), milk energyoutput is often higher in the Jersey. Thesecomparisons are obviously affected by thegenetic potential of the individuals, asconsiderable variation in potential yield existswithin breeds as well as between breeds. Forexample, the recent highest 365-day productiondata available on the internet were 38,570 and75,275 lb for a Jersey and Holstein, respectively.Assuming a BW of 425 kg (935 lb) for the Jerseyand using the 800 kg (1760 lb) BW reported for

the Holstein, the milk yield per kilogram of BWis virtually the same. However, fat content ofthe Jersey milk was more than double that ofthe Holstein, while milk protein content wasincreased by half in the Jersey. Thus, on a solids-corrected or energy basis, the yield per kilogramof BW is higher for the Jersey. Relative to MBW,which relates to surface area and relative heatdissipation, the energy-corrected milk yield ofthe two breeds is more similar.

In a summary of the U.K. dairy industry(CAS, 1978), the feed efficiency of Jersey cowsand Friesian cows were compared based on milksolids output relative to total metabolizableenergy intake required for maintenance andproduction. On this basis, there was a slightadvantage for the Jersey. However, whencompared on the basis of the maintenancerequirement of the Friesian and the currentmarket value of the milk, the net value of Jerseysolids output was greater than for the Friesian.In another comparison, Oldenbroek (1988)compared the use of feed net energy for milkenergy output in the Danish Jersey and threelarge frame Dutch breeds fed either all roughageor 50% concentrate diets. In this comparison,which used relatively low yielding cows, theJersey produced more lactation energy fromforage. In addition, the Jersey produced moremilk energy relative to maintenance energyrequirements, regardless of diet. However, inboth cases, the comparisons are based on acommon maintenance requirement per kilogramof BW, and there are a number of reports in theliterature suggesting the Jersey has a highermaintenance energy requirement than theHolstein (e.g. Solis et al., 1988).

Intake and Digestibility

In the study by Oldenbroek (1988), theDM intake of the Danish Jersey studied was 3.8to 3.9 % of BW, whilst that of the relatively lowyielding large frame cows was only 3 to 3.2%of BW. Relative to the larger cows, intake ofthe Jersey was 83%, whilst BW of the Jersey


109

was roughly 75%. Thus, even in low yieldingJersey cows, intake can not be predicted fromBW using equations developed for Holsteincows (Holter et al., 1997). It is not uncommonfor intakes equalling 5% of BW to be achievedin Jersey cows. As increased intake depressesthe digestibility of many dairy rations, oneconcern is that diet digestion may be depressedto a greater extent in the Jersey than assumedfor the Holstein. However, in a study employinga limited number of observations, rate of passagemeasured in Jersey cows was higher than forFriesian cows, but diet digestion was notdifferent for the two breeds (Ingvarsten andWiesbjerg, 1993). Faster passage rate may be aconsequence of higher intake levels driven bygreater milk energy output. On the other hand,faster passage rate may allow greater intake andbe a consequence of more efficient masticationand rumination in the smaller breed (Dürst etal., 1993). Alternatively, faster passage rate maysimply reflect a genetic difference related to therumen size and anatomy of two breeds.

Milk Fat Composition and Ration FiberContent

The higher fat content of Jersey milkrequires more precursors for milk fat synthesis.Comparisons of the milk fatty acid compositionof Jersey and Holstein milk (Beaulieu andPalmquist, 1995; Bitman et al., 1996; Drackleyet al., 2001) revealed that Jersey milk fat had ahigher content of shorter chain fatty acids, butthere was little difference in total long chain fattyacid content. However, differences in theresponse of C18:0 and C18:1 to treatments didsuggest that mammary activity of “-9 desaturasewas lower in Jersey compared to Holstein cows(Beaulieu and Palmquist, 1995; Drackley et al.,2001). This enzyme converts C18:0 (stearicacid) to cis-9 C18:1 (oleic acid) and trans-11C18:1 (vaccenic acid) to cis-9, trans-11 C18:2(the predominant conjugated linoleic acid inmilk) in the mammary gland. Unlike the longchain fatty acids which are derived from blood,the mammary gland must synthesize the short

chain fatty acids from acetate and β-OH-butyrate. Thus, higher milk fat in Jersey milk isin part accomplished by greater endogenous fatsynthesis by the mammary gland, which requireslipogenic volatile fatty acids (VFA) from therumen. The proportion of the lipogenic VFAabsorbed is promoted by forage digestion;therefore, recommendations in the popular pressfrequently stress the need for ‘effective fiber’ inJersey rations (e.g. Macmillan, 1996 citingMcCollough). While this logic is sound,supporting data are not evident.

In contrast, other popular press articles(e.g. Etchebarne, 1999; Calvin Covington,personal communication) state the opposite.Based on observations working with Californiaherds, Etchebarne (1999) suggests that Jerseycows can, and should, be fed diets containingless fiber to insure that their maximal DM intakecan be achieved. The author asserts that lowerfiber diets can be fed to the Jersey without thedrop in milk fat content and health problemsarising from rumen acidosis that would occur ifthe same diet was fed to the Holstein, implyinga greater tolerance to high levels of dietarystarch. Evidence in the scientific literatureappears to support the latter premise. Researchwith low-forage diets designed to cause milk fatdepression observed less depression in Jerseythan Holstein cows (Jorgensen and Schultz,1965; Annison et al., 1974), but the studies werenot designed to test breed differences. In morerecent studies, variations in source and contentof neutral detergent fiber (NDF) andnonstructural carbohydrate (NSC) of rations fedto Jersey cattle were not associated withdifferences in milk fat content or yield (Elliot etal., 1995; Harmison et al., 1997; West et al.,1997; Kauffman and St-Pierre, 2001). There wasalso no effect on milk fat percentage of Holsteincows when soybean hulls replaced ground cornin the study of Kauffman and St-Pierrre (2001),but the low NDF level was not extreme. Westet al. (1997) observed a depression in DM withincreased inclusion of Bermuda grass hay inrations fed to Jersey cows. This depression in


110

DM intake was not observed in Holstein cows,supporting the concept that, depending on fibersource, Jersey cows may need lower NDF levelsin their ration to achieve maximal DM intake(Etchebarne, 1999). Further confusing the issue,milk fat percentage was decreased when theproportion of alfalfa and Bermuda grass hay wasincreased in rations fed to Jersey cows (West etal., 1997). Obviously, these issues needclarification.

Milk Fever and Transition Diets

There are a number of studies whichhave documented a higher incidence of milkfever in the Jersey (Horst et al., 1997). This hasbeen suggested be related to a reduced numberof 1,25(OH)2D3 (the active form of vitamin D)receptors in the intestine which are responsiblefor increasing calcium absorption in earlylactation (via increased calcium binding proteinactivity). The activation of vitamin D isstimulated by parathyroid hormone, but theeffectiveness of parathyroid hormone is inhibitedas blood base levels increase. Current thinkingis that cationic rations increase the incidence ofmilk fever by limiting the ability of parathyroidhormone to up-regulate calcium absorption viavitamin D receptors in the gut. For this reason,it has been recommended that target urine pHlevels when adjusting the cation-anion differenceof transition rations be lower for Jersey (5.8 to6.2) than for Holstein cows (6.2 to 6.7). Thehigher acid load may compensate for lowervitamin D receptor numbers in the gut of theJersey, but acidic diets may have negative effectson metabolism and health, thus theserecommendations may also need furtherclarification.

Energy Metabolism

Do Jersey cows have a highermaintenance energy requirement? As theefficiency of milk energy synthesis increaseswith milk fat content, is the efficiency ofmetabolizable energy (ME) use for lactation

greater for Jersey cows? Do high levels of intakeaffect digestion or influence gaseous or heatenergy loss? Limited calorimetric comparisonsof the two breeds from the first half of the centurysuggested little difference between the Jersey andHolstein in terms of energy metabolism(Ritzman and Benedict, 1938; Brody, 1945), butstudies in nonlactating animals showed a muchhigher maintenance requirement for the Jersey(Solis et al., 1988).

A study comparing the energymetabolism of Holstein and Jersey cows wasconducted at the USDA Energy Metabolism Unitin Beltsville, MD in cooperation with the USDAARS Experiment Station (a Jersey herd) inLewisburg, TN and supported in part by theAmerican Jersey Cattle Association (Tyrrell etal., 1990 and 1991; Bitman et al., 1996). Thestudy compared the energy metabolism of Jerseyand Holstein cows fed total mixed rationswithout or with supplemental fat fromcottonseed (Table 1). Three measurements ofenergy balance were obtained in eight lactatingcows of each breed at 7, 23, and 39 weeks aftercalving. In addition, measurements wereobtained on four dry cows of each breed todetermine basal levels of digestion andmetabolism. The relative yield potential of thetwo breeds was similar, as fat-corrected milkyield per kilogram of MBW at peak lactationand DM and energy intakes per kilogram of0MBW throughout the study were not differentbetween breeds. Surprisingly, digestibility ofenergy was higher in the Jersey. This wascountered by slight increases in urine andmethane energy loss, such that ME intake wasnot different. The higher milk fat content of theJersey was associated with numerically highermilk energy (as a percentage of intake energy),while tissue energy retention was numericallylower such that energy balance was not differentbetween breeds. On a MBW basis, theincremental relationship between ME intake andnet energy (NE) for lactation (milk energycorrected for tissue energy loss or gain) wassimilar between the two breeds. However, the


111

breeds differed slightly in the partitioning of MEbetween milk and body tissue. A similarconclusion was reached when measurements ofthe energy metabolism of lactating beef cattlewas compared to previous measurements fromHolstein cattle (Reynolds and Tyrrell, 2000).

Based on the regression of NE on ME,the data obtained from the lactating cows didnot suggest any difference in the maintenanceenergy requirements of lactating Jersey andHolstein cows. However, the validity of usingof the regression of NE on ME to estimatemaintenance energy requirements is oftenquestioned. In feeding the dry cows for thisstudy, the maintenance requirement of Holsteincows was used to estimate energy requirementsfor maintenance of the dry Jersey cows (120 kcal/kg MBW). This level of feeding was quicklyincreased by 30% to slow dramatic weight lossin the Jersey cows. The energy balances in Table1 show the higher ME intake required to avoidtissue energy loss in the Jersey cows. Reasonsfor this difference are not certain, but may inpart be due to differences in body compositionas the Jersey tended to be leaner, and numerousstudies have shown that maintenancerequirements decrease with increasing body fatcomposition.

Conclusions

1. Dry matter intake per kilogram of BW isoften higher in Jersey cows, reflectinggreater milk energy yield relative to BW.This may be accompanied by faster passagerate of digesta through the gut, which couldbe a consequence of greater intake.Alternatively, more effective mastication andrumination in Jersey cows may increasepassage rate and allow more intake.Regardless, in a limited number of studies,higher relative intake in Jersey cows doesnot appear to reduce diet DM digestibilitycompared to Holstein cows.

2. The effective fiber requirements of Jerseycows have been suggested to be morestringent than that of Holstein cows.Alternatively, the Jersey is suggested to bemore tolerant of high dietary starch levels.These considerations may be related todifferences in rate of passage between Jerseyand Holstein cows and need clarification.

3. Maintenance energy requirements are higherin dry Jersey cows compared to Holsteincows. Reasons for this difference are notapparent, but body composition may play arole.

4. In a calorimetric comparison of the energymetabolism for Jersey and Holstein cows,the Jersey used slightly more dietary energyfor milk and less for tissue, but on a MBWbasis, the use of ME for NE was remarkablysimilar for the two breeds.

References

Annison, E.F., R. Bickerstaffe, and J.L. Linzell.1974. Glucose and fatty acid metabolism incows producing milk of low fat content. J. Agric.Sci. (Camb.) 82:87-95.

Beaulieu, A.D., and D.L. Palmquist. 1995.Differential effects of high fat diets on fatty acidcomposition in milk of Jersey and Holstein cows.J. Dairy Sci. 78:1336-1344.

Bitman, J., D.L. Wood, R.H. Miller, H.F. Tyrrell,C.K. Reynolds, and H.D. Baxter. 1996.Comparison of milk and blood lipids in Jerseyand Holstein cows fed total mixed rations withor without whole cottonseed. J. Dairy Sci.79:1596-1602.

Brody, S.A. 1945. Bioenergetics and Growth.Hafner Publ., New York.

CAS. 1978. Strategy for the UK Dairy Industry.Centre for Agricultural Strategy. Report 4, TheUniversity of Reading, England.


112

Drackley, J.K., A.D. Beaulieu, and J.P. Elliott.2001. Responses of milk fat composition todietary fat or nonstructural carbohydrates inHolstein and Jersey cows. J. Dairy Sci. 84:1231-1237.

Dürst, B., M. Senn, and W. Langhans. 1993.Eating patterns of lactating dairy cows of threedifferent breeds fed grass ad lib. Physiology andBehaviour 54:625-631.

Elliott, J.P., J.K. Drackley, G.C. Fahey, Jr., andR.D. Shanks. 1995. Utilization of supplementalfat by dairy cows fed diets varying in content ofnon-structural carbohydrates. J. Dairy Sci.78:1512-1525.

Etchebarne, M.A. 1999. Producing milk withHosteins and Jerseys. The Western Dairyman,August, pp 10-11.

Harmison, B., M.L. Eastridge, and J.L. Firkins.1997. Effect of percentage of dietary forageneutral detergent fiber and source of starch onperformance of lactating Jersey cows. J. DairySci. 80:905-911.

Holter, J.B., J.W. West, and M.L. McGilliard.1997. Predicting ad libitum dry matter intakeand yield of Holstein cows. J. Dairy Sci.80:2188-2199.

Horst, R.L., J.P. Goff, T.A. Reinhardt, and D.R.Buxton. 1997. Strategies for preventing milkfever in dairy cattle. J. Dairy Sci. 80:1269-1280.

Ingvarsten, K.L. and M.R. Weisbjerg. 1993.Jersey cows have a higher feed intake capacityand higher rate of passage than Friesian cows.Arch. Tierz. 36:495-498.

Jorgensen, N. A., and L. H. Schultz. 1965. Rationeffects on rumen acids, ketogenesis, and milkcomposition. II. Restricted roughage feeding. J.Dairy Sci. 48:1040-1045.

Kauffman, A.J., and N.R. St-Pierre. 2001. Therelationship of milk urea nitrogen to urinenitrogen excretion in Holstein and Jersey cows.J. Dairy Sci. 84:2284-2294.

Macmillan, S. 1996. Learning to feed Jerseys.Dairy Farmer, March 5, pp 4-5.

Oldenbroek, J.K. 1988. The performance ofJersey cows and cows of larger dairy breeds ontwo complete diets with different roughagecontents. Livestock Prod. Sci. 18:1-17.

Reynolds, C.K., and H.F. Tyrrell. 2000. Energymetabolism in lactating beef heifers. J. Anim.Sci. 78:2696-2705.

Ritzman, E.G. and F.G. Benedict. 1938.Nutritional Physiology of the Adult Ruminant.Carnegie Institute, Washington, DC.

Solis, J.C., F.M. Beyers, G.T. Schelling, C.R.Long, and L.W. Greene. 1988. Maintenancerequirements and energetic efficiency of cowsof different breed types. J. Anim. Sci. 66:764-773.

Tyrrell, H.F., C.K. Reynolds, and H.D. Baxter.1990. Energy metabolism of Jersey and Holsteincows fed total mixed diets with or without wholecottonseed. J. Dairy Sci. 73(Suppl. 1):192.(Abstr.)

Tyrrell, H.F., C.K. Reynolds, and H.D. Baxter.1991. Effect of dietary fat from wholecottonseed on energy metabolism of lactatingJersey and Holstein cows. J. Dairy Sci.74(Suppl. 1):250. (Abstr.)

West, J.W., G.M. Hill, R.N. Gates, and B.G.Mullinix. 1997. Effects of dietary forage sourceand amount of forage addition on intake, milkyield, and digestion for lactating dairy cows. J.Dairy Sci. 80:1656-1665.


113

Table 1. Production and energy metabolism in Jersey and Holstein cattle1.

Jersey HolsteinLactating Dry Lactating Dry Breed effect

Average performance

DM intake (kg/day) 15.9 4.8 22.0 6.4 P < 0.01

Body weight (kg) 402 364 636 738 P < 0.01

Milk yield (kg/day) 21.1 - 32.9 - P < 0.01

Milk fat (%) 5.10 - 3.86 - P < 0.01

Milk protein (%) 3.68 - 3.22 - P < 0.01

Energy intake

Intake (Mcal/day) 304 72.6 100.2 - P < 0.01

Intake (Kcal/BW0.75) 3.41 814 799 - NS

Energy metabolism (% of ingested energy)

Digested 65.2 - 64.1 - P < 0.10

Metabolizable 56.3 - 55.4 - NS

Milk 23.6 - 21.6 - P < 0.10

Tissue 0.8 - 2.0 - NS

Energy balance 24.3 - 23.6 - NS

Energy metabolism (Kcal/BW0.75)

Intake - 261 - 205 -

Digested - 183 - 138 -

Metabolized - 151 - 113 -

Tissue - -0.6 - -1.5 -1Tyrrell et al. (1991) and unpublished observations.


114


115

Managing Feed Costs During Low Milk Prices

David T. Galligan1

University of PennsylvaniaSchool of Veterinary Medicine

Introduction

Producers all around the world areinterested in making a decent return on theirinvestment in dairy production. Milk prices havesteadily decreased over time and producers mustmake the appropriate decisions to maintain orimprove profitability. An immediate responseto this type of scenario is to control input costs.However, there are several ways (someappropriate and others not) in which “cost”should be controlled. Furthermore, producersand consultants often question the profitabilityof higher production and suggest that lowerproduction levels are more profitable based onobservational analysis. The purpose of thispresentation is to review these concepts withspecific emphasis on feed cost.

Controlling Cost

There are several ways in which cost canbe controlled. One strategy might involvefinding the same inputs at cheaper sources – wecall this cost efficiency control. Another mightinvolve selecting a “different level of input” andtherefore indirectly selecting a different level ofproduction as a target. In this second strategy,overall feed cost are indeed controlled, but theproduction level also changes as a consequenceof the decision choice. Each of these strategieswill be described and discussed.

Cost Efficiency Control

The basic assumption of the strategy toimprove feed cost efficiency is that theproduction response to inputs is fixed and oneis merely finding a cheaper source of nutrients.As input cost per unit is reduced and/or productvalue per unit increased, profits will increase.Ferguson et al. (1987) and Galligan et al. (1990)reported an approximate 15% feed cost savingson herds subscribing to a routine rationformulation program. These relatively smallherds found cheaper sources of grain inputsthrough the selection of commodity feeds.

Several methods have emerged to helpidentify “efficient” sources of feed inputs:

a) Petersen method – The basic assumption ofthis methodology is that the value of a feedingredient is the sum of the economic valuesascribed to its nutrient content. Often energyand protein are used as the major nutrients todetermine a feed’s value. Economic estimatesof the value of energy and protein are estimatedfrom the value of soybean meal and shelled corn,feeds that are commonly purchased (Petersen,1932). However a limitation to thismethodology is that it does not ascribe value to“nutrient density”.

b) Regression methods are extensions of thePetersen method. With these methods, morethan two feeds can be used as references to

1Contact at: 382 West Street Road, Kennett Square, PA 19348-1692, (610) 444-5800, FAX (610) 925-8123, Email:[email protected]


116

estimate the nutrient values. Secondly, multiplenutrients can used to establish value of a feedingredient (St-Pierre and Glamocic, 2000).

c) Linear programming methods are perhaps themost robust and widely used and are found inration formulation programs (e.g. Spartan andDairyLP). This method establishes the value ofa feed based a number of dimensions(nutritional, feed availability, feed ratios, etc) andcalculates a “reduced cost” for each potentialfeed ingredient. This reduced cost value is basedon the hypothesized biological response (ie. thebiology described as nutrient constraints in thelinear program of the ration formulationsoftware). Secondly, non-nutrient constraints,such as the availability of certain feed ingredientsor feed ratios, are considered as well. Lastly,the nutrient profile of a feed or “the fit” of afeed into a ration is also considered in thedetermination of these values. Reduced costestimates for feed ingredients, indicating thechange in feed cost that could occur before adifferent ration is used, is optimal.

d) Non-linear programming methods (e.g. CPMdairy) do not have “reduced cost” as directoutputs. Users must vary parameters of interestand see the changes in the response function(ration cost) to estimate reduced cost.

Other options to “control feed cost”occur when herds increase in capacity and areable to bulk purchase feed ingredients at reducedrates. Small producers can organize intopurchasing groups to also capture some of theseeconomies of scale benefits.

Feed inventory management is yetanother method to control feed cost by reducinginventory cost. These costs are scale dependent,with larger herds enjoying greater benefits fromthese strategies.

In short, cost efficiency should alwaysbe sought after regardless of the price of milk.

One has to be careful that cost control is not doneto the point where production is compromised.

Controlling the level of input to also controlproduction

There are two approaches to control thelevel of input: a) lower the level of productionfor which the ration is balanced for and therebyreduce the input feed cost, and b) introduceadditional “groups” of cows to control the degreeof over and underfeeding and thereby controloverall feed cost.

Behind both of these approaches is thequestion as to whether the production ofadditional milk is profitable and what is the milkresponse of cows to over and under feeding.One has to be careful to use marginal cost ofproduction rather then average cost of productionto evaluate this question. Production curvetheory tells us that optimal production occurswhen the marginal cost of production equals themarginal return of the product produced.Production below and above this level will yielda lower level of profit. What is the marginalcost of production and how does it differ fromaverage cost of production?

In any production system, there are fixedand variable inputs associated with making aproduct. Fixed costs are those that will notchange with a decision choice or an increase ina variable input and thus can be ignored in thedecision process. However, fixed cost areincluded along with variable cost in average costcalculations. It is this inclusion of “fixed cost”that makes average cost of productionproblematic in tactical decision making (ie.,should I feed for a higher production or not). Inthe context of nutrition and at the cow level, themaintenance cost of the animal is essentially“fixed” – it is the same whether she is producing1 kg of milk or 50 kg. Under U.S. conditions ofproduction, this cost is approximately $1.25 to$1.50 (slightly lower then feed cost per day of adry cow). In Figure 1, there are two levels of


117

production and the corresponding ration cost,which would include the maintenance feed costplus the feed cost associated with production(levels are figurative to facilitate theunderstanding of the concepts). Ration costdivided by the level of milk yield gives anaverage feed cost per pound of milk produced,then multiplied by 100 provides the commonlyused feed cost per cwt of milk. The profit(income over feed cost) is also calculated. Whilethe average cost per cwt for the 90 lb/day of milkproduction is considerably higher then the 80lb/day of production, the higher production levelyields a greater profit ($7.25 vs. $7.00).Controlling average feed cost per cwt would notlead to profitable production.

What is important is the cost of“marginal” production. The “marginal” cost ofproducing at 90 lb/day versus 80 lb/day is $4.00-$3.00 or $1.00 per 10 lb of milk. The marginalrevenue of 90 lb versus 80 lb per day of milk isthe value of 10 lb of milk or $12.50/cwt x 10 =$1.25. The marginal profit is therefore $0.25for the additional 10 lb of milk. As long as themarginal cost is less than the marginal revenue,higher profits will be realized by increasedproduction.

What are typical marginal costestimates? One hundred rations (amount andtype of feeds) and corresponding milk levelswere recorded from nutritional studies publishedin the Journal of Dairy Science. For moderatelevels of milk production 60 to 80 lb/day, feedcost was estimated at about $0.03/lb of milk(Figure 2). As production increases above 80lb/day of milk, the marginal cost of productionmay increase due to: a) the need to use morecostly feed inputs (feeds with higher nutrientdensities), b) increase in passage rate relative todigestion, and c) an increased use of body weightas a nutrient source. However, at moderate levelsof herd production, the feed cost of $0.03/lb ofmilk can be viewed as relatively constant.

Body weight as a nutrient can have adramatic effect on the marginal cost ofproduction. One could argue that at the higherlevels of production (weight loss is occurring tosome degree), the true ration cost (current + thecost of weight loss) are actually higher thanreported. Conversely, at the lower levels ofproduction, the ration costs are lower in thatsome of the nutrients are used for weight gain.These issues taken together would make theestimate of the marginal cost of production($.034/lb) a low figure.

Another approach to looking at themarginal cost of production is to investigate aseries of rations formulated for different levelsof production and then calculating the marginalchanges in cost. In Table 1, rations wereformulated using CPM Dairy as a rationformulation program. Three sets of rations basedon different forages (50% corn silage/50%haylage, 100% haylage, and 100% corn silage)were formulated for increasing levels ofproduction. Marginal cost per pound of milkwere calculated as outlined above.

In general the 50/50 forage based rationhad the cheapest marginal cost of production.As production increases, the marginal cost ofproduction increases from $0.03/lb to higherlevels, reflecting the use of more expensiveingredients, as well as the effects of an increasingpassage rate on the digestibility of feeds. Thedecrease in marginal cost at the highest level ofproduction reflects inconsistent attributes ofproduction functions and which reflects changesin nutritional programming constraints (i.e.,relaxing constraints) to obtained solutions(perhaps these changes are “arbitrary”).

From a feeding perspective, it appearsthat higher production levels will yield higherprofits. Higher production is often associatedwith “disease concerns or reproductiveinefficiency”, and thus, producers claim that inaddition to feed cost, other costs increase, thusinflating the marginal cost of production. A


118

review of the epidemiological literature (Fetrow,2002) does not see this causal relationship ofhigher production and disease, and the effectsof high production on reproduction appear to besmall and controllable (Grohn and Rajala-Schultz, 2000). These associations of productionand disease/reproductive inefficiency areperceived associations and are due to the greaterconcern bestowed on high producing cowscompared to low production cows (i.e., a sickhigh producing cow is more likely to be seen bya vet than a low producing cow). More attemptsat rebreeding, higher producing cows than lowerproducing animals, as well as more health caredollars spent on diseased animals, fosters thisperceived casual association.

Summary

To be profitable in dairy production,producers must understand the economic basisof changes in their production system. All herds,at all times, should practice cost controlefficiency as long as production is not impaired.For the economic evaluation of a given decision,they must know the marginal changes in costand revenue to make sound choices. In general,a marginal cost of higher production at $0.03 to$0.06/lb of milk can be used as an industryestimate. At higher levels of production, thismarginal cost should theoretically increase. Thepoint where the marginal cost of milk productionequals the marginal return of production willdefine the limit of production that maximizesprofit (income over feed cost).

References:

Ferguson, J.D., D.T. Galligan, C.F. Ramberg, andD.S. Kronfeld. 1987. Veterinary nutritionaladvisory services to dairy farms. Compend.Contin. Educ. Pract. Vet. 9:F192-F200.

Fetrow, J.P. 2002. Monsanto Dairy ScienceSymposium. St. Louis, MO. April 19-21.

Galligan, D.T., C.R. Curtis, C.F. Ramberg. 1990.Economic value of nutritional services for dairyherds. J. Dairy Sci. 73:(Suppl. 1):222. (Abstr.)

Galligan, D.T., and J.D. Ferguson. 1991. Newapproaches to nutritional monitoring. Vet. Clin.North Am. Food Anim. Pract. 7:473-482.

Grohn, Y.T., and P.G. Rajala-Schultz. 2000.Epidemiology of reproductive performance indairy cows. Anim. Reprod. Sci. 60-61:605-614.

Petersen, J. 1932. A formula for evaluating feedson the basis of digestible nutrients. J. Dairy Sci.15:293-297.

St-Pierre, N.R., and D. Glamocic. 2000.Estimating unit cost of nutrients from marketprices of feedstuffs. J. Dairy Sci. 83:1402-1411.


119

Feed Cost/Cow/day Marginal Cost of Production ($/lb milk)Milk, lb/day CS/Haylage Haylage Corn Silage CS/Haylage Haylage Corn Silage

40 $2.51 $2.46 $2.07 $0.00 50 $2.77 $2.76 $2.39 $0.03 $0.03 $0.03 60 $3.06 $3.32 $2.76 $0.03 $0.06 $0.04 70 $3.52 $3.74 $3.24 $0.05 $0.04 $0.05 80 $3.92 $4.22 $3.75 $0.04 $0.05 $0.05 90 $4.31 $4.81 $4.17 $0.04 $0.06 $0.04 100 $4.62 $5.02 $4.66 $0.03 $0.02 $0.05 110 $4.95 $5.29 $5.15 $0.03 $0.03 $0.05

Table 1. Feed costs and marginal cost of production as affected by milk production and type offorage.1

Figure 1. Calculations to demonstrate that average feed cost per cwt (or lb or per kg) of milk can bemisleading.

1Rations were formulated using CPM Dairy. Rations consisted of 50% cornsilage (CS) and 50% haylage,100% haylage, or 100% corn silage, respectively.

Milk is valued at $12.50/cwt.Profit = income over feed cost.

What level of production is best: 90 lb.

$7.00

$7.25

$3.75

$4.44

$3.00

$4.00

80

90

Profit($/day)

Feed Cost/cwt

Feed Cost($/day)

Milk Yield(lb/day)


120

Figure 2. Relationship between 4% fat-corrected milk yield and ration cost (100 rations used).

y = 0.0338x - 0.2197R2 = 0.3226

00.5

11.5

22.5

33.5

4

0 20 40 60 80 100Milk (lb)

Rat

ion

$/d Ration Cost

Linear (Ration Cost)


121

Managing N, P, and K for the Dairy Industry

Wendy J. Powers1

Department of Animal ScienceIowa State University

Abstract

Challenges in meeting environmentalobjectives continue to dictate that new andinnovative strategies be developed, such thateconomically viable practices are implemented.Nutritional regimens drive the input side ofnutrient management. Therefore, adoption offeeding practices that minimize excretion arenecessary. Minimizing over-feeding ofphosphorus has been demonstrated to reduce Pexcretion. Similar findings have been observedwhen protein feeding is tailored to not exceedanimal needs. Current regulations enhance thechallenge of utilizing co-product feeds whileminimizing nutrient excretions. Yet, regulationscurrently in place in some states, with theopportunity to be adopted in others, require thatwe address the challenges now and that we notexpect a post-excretion solution.

Introduction

Environmental issues are not going awayany time soon. Greater emphasis on animalproduction impacts on the environment hasresulted in greater scrutiny of practices andmethods to mediate impacts. Historicalapproaches have addressed post-excretionstrategies, while more recently, the role ofnutrition has received an increasing degree ofattention. As we continue to minimize thepollution potential of animal agriculture, thenutritionist will have greater involvement in an

effort to develop pre-excretion solutions. We’veseen the signs for this in the EPA-USDA AnimalFeeding Operations (AFO) Strategy, wherebythe concept of a Comprehensive NutrientManagement Plan was introduced. These plansincorporated the option of including feedmanagement as a component of a farm’senvironmental plan. To take this a step further,the Natural Resources Conservation Service(NRCS) is currently developing a ConservationPractice Standard for feed management. Onceadopted, the standard becomes part of NRCS’stechnical manual and will further elevate theattention nutrition strategies receive,incorporating the nutritionist as a standardmember of the nutrient management team.

Regulatory update on nitrogen andphosphorus

On December 16, 2002, much of thelivestock sector breathed a sigh of relief. Thenew Concentrated Animal Feeding Operation’s(CAFO) regulations, a revision of the CleanWater Act, had been signed and appeared to bequite fair, leaving much of the details up toindividual states to work through. The final ruleprovided state flexibility by:

• Allowing states to tailor their permitprogram to address specific state needs,taking into account the size, location, andenvironmental risk posed by individualoperations,

1Contact at: 109 Kildee Hall, Ames, IA 50011, (515) 294-1635, FAX (515) 294-5698, Email: [email protected]


122

• Authorizing states to determine that specificCAFO have no potential to discharge underany circumstances, thereby not needingpermits,

• Providing flexibility for states to tailornutrient managements, and

• Authorizing states to adopt alternativeperformance standards to promoteinnovative technologies.

One of the greatest concerns over the twoyears during which time the revision processensued was how land application of animalmanure would be handled. The final rule requires“CAFO to develop and submit a certified PermitNutrient Plan, which would be reviewedannually, recertified every five years, and wouldhave limited manure spreading on all land ownedor under the operational control of the CAFO tothe nitrogen-based rate, unless soil or other fieldconditions at the CAFO warranted limiting theapplication rate to the more stringentphosphorus-based rate” (Federal Register, Vol.68, No. 29, February 12, 2003). However,Michigan and Ohio have been using P-basedplanning practices for a number of years, andIndiana considers P-based nutrient planning ona case-by-case basis.

A couple of changes to the previousCAFO rule do affect the dairy industry. One isthe inclusion of stand-alone heifer operations inthe definition of a CAFO. Those operations withgreater than 1000 head are designated as CAFOand must apply for a permit. The new rule alsoeliminates the 25-year, 24-hour storm eventexemption that previously allowed operationsthat only discharged during such storm eventsfrom applying for a National Pollutant DischargeElimination System (NPDES) permit. Thischange means that all operations of CAFO sizemust apply for a permit regardless of whetheror not they only discharge under 25-year, 24-hour storm conditions. As a result, a greaternumber of operations will be permitted.However, the containment standard for a 25-year, 24-hour storm event is retained. Finally,

the new rule clarifies that runoff from land as aresult of manure application from a CAFO is adischarge and therefore subject to NPDESpermit requirements.

Many state Natural ResourceConservation Services (NRCS) have developeda Conservation Practice Standard 590, oftenreferred to as the P Index. The P Index is a riskassessment tool that can be implemented as amechanism for determining manure applicationrates. However, the P Index, itself, does notdesignate application rates. Rather, it is up tothe state to take the calculated risk factors anddevelop a scaling system whereby operationsthat fall within a range of risks are allowed toapply manure at a state-determined applicationrate. While the EPA’s new CAFO regulationsdo not require P-based manure application oruse of the P Index, it is likely that states with aP Index developed will continue to proceed withimplementation of the index and other states willcontinue to develop one. Therefore, asnutritionists, we need to continue to be aware ofthe challenges in maintaining a nutrient balanceand of the implications that nutritional strategieshave on environmental risk.

Nutritional management of phosphorus

Recently, a number of studies havereported that overfeeding dietary P, by as muchas 50%, is a common practice (Beede andDavidson, 1999; Wu et al., 2001). Beede andDavidson (1999) discussed several reasons foroverfeeding, including addition of safetymargins beyond the recommended dietary P inorder to adjust for inaccuracies withrecommendations or uncertainty with theproportion of dietary P that is available.Improved milk yield and reproductiveperformance were discussed as reasons foroverfeeding as well. Beede and Davidson (1999)go on to establish that each of these reasons is amisconception. Trials have demonstrated thatsuch practices result in quadratic increases in Pexcretion (Morse et al., 1992). Morse et al.


123

(1992) observed an increase in fecal P excretionof 48.6% when lactating cows were fed dietscontaining 36.5% greater P. Knowlton andHerbein (2001) observed linear increases in fecaland urine P excretion with increasing dietary Pfrom 0.34% to 0.51% to 0.67% dietary P.Apparent digestibility of P in early lactationHolsteins decreased quadratically withincreasing dietary P (Knowlton and Herbein,2001). Wu et al. (2000) found a 23% reductionin fecal P excretion by reducing the dietary Pcontent from 0.49 to 0.40% for mature lactatingHolstein cows.

The interest in feeding co-products todairy cattle and other livestock has surged in thepast couple of years as a result of increasedinterest in ethanol production in response tofossil fuel costs. Particularly in the Midwest,ethanol co-products have become aneconomical, available feed source. However,corn co-products are higher in phosphorus thancorn and soybean meal (Table 1). While, in manycases, these feeds are less expensive than cornand soybeans, thereby reducing diet cost, costof manure application to more acres should beconsidered when determining the economics offeed sources. Co-products serve as an importantfeed source for the animal industry. Producersand nutritionists need to be aware of productnutrient content and the availability of nutrientsin order to appropriately manage excretednutrients. Variability in feed source, bothbetween processors and within a processor,necessitates that each batch of feed be analyzedand ration formulations adjusted accordingly inorder to minimize overfeeding of nutrients.Using a feedlot example, it becomes obvious thatappropriately managing manure nutrientsrequires some forward thinking. Table 2illustrates calculated land needs to maintain aphosphorus balance between P excreted and Pharvested by corn silage. The calculations arebased on research data collected by Wu et al.(2000) and that estimated from a simulation thatincorporates corn co-products in the diet andestimates excretions based on the relationship

between intake and excretion observed by Wuet al. (2000). The simulation shows thatincorporation of corn co-products does increasethe land needed to maintain soil P. However, Pexcretion of the example diet that includes thecorn co-product remains below that observed byWu et al. (2000) when mineral P was includedin the diet beyond that recommended by NRC(2001). In the examples provided, it is assumedthat manure is applied only on corn silage.Therefore, more silage is grown than thatconsumed by the animal (20% of dietary DMI)in a year’s time. In reality, producers feeding bothcorn silage and alfalfa silage may grow thealfalfa silage as well, in which case, the cornsilage grown would match feed inventory needsand alfalfa fields would utilize the remainingmanure. One factor that can’t be overstated isthat as we move towards P-based nutrientmanagement planning, the benefit of havingmanure handling systems designed to retain Nrather than lose N to the atmosphere throughvolatilization becomes obvious. In all threeexamples provided in Table 2, supplemental Nis needed to grow the indicated quantity of cornsilage. So, while in the past, manure handlingsystems sought to lose N in order to reduce theland needed to spread manure nutrients usingN-based nutrient planning, volatilizationbecomes a costly process when managingnutrients under P-based planning.

What about K?

Potassium is not an environmentalconcern from a human health perspective.Rather, soils rich in K can contribute to highforage K concentrations as a result of luxuryconsumption. The result can be grass tetany forthe animal that consumes the forage. The dairycow’s minimum requirement for K is 0.90 to1.0% of the ration DM (NRC, 2001). Themaximum tolerable level is about 3.0%.Because more K is lost through sweat and saliva,supplemental K can help to alleviate symptomsof heat stress. Research results have beenvariable, but increasing dietary K levels to 1.5


124

or 1.6% of ration DM during periods of heatstress may be beneficial to the cow. ExcessiveK intake can lead to udder edema in fresh cows,greater incidence of retained placenta, andgreater risk of displaced abomasum. ExcessiveK intake decreases Mg absorption, decreasesfeed intake and milk production, increases waterintake, and increases urine output. High levelsof dietary K during the dry period can predisposethe fresh cow to milk fever, DA, uterineproblems, and other metabolic disorders.Potassium supplementation is seldom needed inforage-based diets because most forages containhigh concentrations of K. As with most nutrients,feed sampling and analysis is recommended toavoid over-supplementing diets with mineralsources of K. In the case of K, forages take upfar more than needed for maximum DM yieldsper acre. Plant levels of 2.0 to 3.0% are adequatefor plant growth, but 6.0% K in grass silage hasbeen reported. Excessive feeding of K putssubstantially more K into the environment thannecessary.

What’s ahead?

In the past decade, most of the emphasisplaced on environmental decisions has beenbased on water quality concerns. However,nuisance issues, primarily odor, are often thetrigger that stimulates controversy regardinganimal production. It is no surprise, then, thatwe are seeing greater emphasis placed on airemissions from animal agriculture. Some stateshave implemented odor regulations, while otherstates now have standards for specific gases orclasses of gaseous compounds. For example, in1997, the California South Coast Air QualityManagement District (SCAQMD) establishedemission reduction goals for animal agriculturewithin their area in response to non-attainmentof PM10 (particulate matter) and ozone standards(CM#99WST-01 Appendix B: New and RevisedStationary Source Control Measures). TheSCAQMD has set a goal of 30% reduction ofvolatile organic carbon emissions from livestockwaste by 2006 and 50% reduction of ammonia

emissions from dairy operations by 2006. Oneof the primary mechanisms for reaching thestated ammonia goal is relocation of the dairyindustry out of the area. However, if the targetsare not met by January 1, 2004, dairy and otherlivestock facilities still located in the Basin willbe subject to ammonia controls. Possiblecontrols that were indicated by SCAQMDinclude reduced nitrogen feeding, promotion ofaerobic manure storage conditions including useof enzymes and microbial products, off-sitecomposting, and chemical alteration of manurepH. With the pending revision of the Clean AirAct, it is foreseeable that restrictions mayincrease at the Federal level as well.

Particulate matter poses perhaps thegreatest challenge for animal agriculture. Directemission sources of PM10, the coarseparticulates, arise from primarily combustionprocesses (Figure 1; EPA, 1998). Directemissions of PM2.5, respirable particulates, arealso primarily the result of combustion processes(Figure 2; EPA, 1998). In addition to directemissions, secondary processes whereby SOx,NOx, and NH3 react in the atmosphere to formammonium sulfate and ammonium nitrate fineparticles, contribute to as much as half of thePM2.5 measured in the U.S. The EPA estimatesthat 86% of the national ammonia emissions arefrom miscellaneous sources that includelivestock and fertilizer (Figure 3; EPA, 1998).Livestock agriculture accounted for 83% of allemissions in the miscellaneous category withfertilizer application comprising the remainder.

The EPA already regulates manycombustion processes through industrial andutility permitting processes. Area sources are notlikely to be targeted for future regulation becausethey are non-point source emitters. To addressparticulate emission reductions, animalagriculture is a likely target due to itscontribution to ammonia emissions which, inturn, can contribute up to half of particulateconcentrations and emissions. Animalagriculture must be taking steps to address


125

ammonia production and emission. In 1984,legislation was enacted in the Netherlands toreduce 2000 ammonia emissions by 50% relativeto 1980 emissions (Lenis and Jongbloed, 1999).Such requirements here in the U.S. coulddramatically change animal agriculture, bothstructure and management.

Nutritional strategies currently availableto reduce ammonia emissions include loweringdietary protein levels, changing the ratio betweenurinary nitrogen and fecal nitrogen, reducingurine pH, inhibiting urea degradation, andbinding ammonia. Many of these examples havebeen tested in swine and poultry research;however, the principles should apply to the dairyanimal as well. As the issue of reducing ammoniaemissions travels across the Atlantic and theU.S., additional innovative strategies will requiredevelopment and testing. Furthermore, we couldsee the cost to stay in business change as well.

Reduced dietary protein

Minimization of nitrogen excretion isthe most obvious method to curb ammoniaemissions. By reducing the available substrate,less ammonia will be formed and volatilized.Endogenous nitrogen losses are unavoidable.However, steps should be taken to minimizeoverfeeding of protein and amino acids beyondanimal needs. In Holstein heifers, James et al.(1999) observed a 28% reduction in ammoniaemissions when N intake was reduced 14%.Total N excretion was reduced only 20%.Tomlinson et al. (1996) reported an approximate20% reduction in N excretion associated with a3% reduction in dietary CP for lactating dairycows. A greater number of examples of thisstrategy are found in swine and poultry research.Reduced crude protein (CP) in swine grower andfinisher diets, supplemented with lysine,methionine, tryptophan, and threonine, resultedin reduced nitrogen content of the manure andreduced ammonia emission rates from themanure (Hobbs et al., 1998). The reduced CPdiets contained 16.1% CP in the grower phase

and 14.0% CP in the finisher phase compared to20.8% and 18.9% CP in the control diets,respectively. Ammonia emission rates werereduced 74% and 65% in the grower and finisherphases, respectively, as a result of feeding thereduced CP diets. In a review, Lenis andJongbloed (1999) cited work that observed a15% reduction in N excretion and 17% reductionin ammonia emission following implementationof a 2-phase feeding plan. Greater reductions arepossible. Reducing CP content of broiler dietsresulted in decreased litter N content but nosignificant differences in ammonia concentrationin the house (Ferguson et al., 1998). However,analyzed diets showed that CP content had beendecreased by less than 2 percentage units. The13.3% decrease in N intake did correspond to18.2% reduction in litter N content. Elwingerand Svensson (1996) fed broilers dietscontaining 18, 20 or 22% CP and measuredammonia emissions from the litter bed. Total Nlosses in the houses averaged 18 to 20% of totalN input. A linear trend of increasing ammoniaemission with increasing dietary CP content wasobserved. Increasing dietary CP significantlyincreased calculated N losses as well. Generally,as a guide, for each 1% reduction in dietary CPestimated, ammonia losses are reduced by 10%in swine and poultry (Sutton et al., 1997; Kayand Lee, 1997; Blair et al., 1995; Jacob et al.,1994; Aarnink et al., 1993). Similarly, Kerr(1995) concluded that, in swine, N excretion isreduced by approximately 8.4% for each onepercentage unit reduction in dietary CP. Jameset al. (1999) and Tomlinson et al. (1996) showedsimilar findings in the dairy animal. However,current work demonstrates that as animals arefed closer to true N requirements, furtherreductions in dietary CP may result in lesspronounced reduction in N excretion andammonia losses. Thus, this issue is really, howclose to requirements can we economicallyachieve? Regardless, for the ruminant, wecontinue to be challenged with trying to moreprecisely define the N and amino acidrequirements of the rumen and lowergastrointestinal tract, separately, in order to better


126

feed the animal as a whole. Restrictions onammonia emissions add urgency to these needs.

Fecal and urinary nitrogen ratios

Inclusion of fermentable carbohydratesinto swine diets has been shown to shift Nexcretion from urine to feces (Canh et al.,1997b). Fecal nitrogen is less easily degradedto ammonia. Hindgut fermentation can increasefecal nitrogen while decreasing urinary nitrogen.Inclusion of sugar beet pulp into grow-finishingdiets at 5, 10, and 15% resulted in a reduction ofammonia emission by14% for each 5% increasein sugar beet pulp inclusion (Canh et al., 1998).Slurry pH decreased by 0.4 to 0.5 units with eachincremental increase in dietary sugar beet pulpcontent. Tomlinson et al. (1996) illustrated thatcows fed calcium soaps of long chain fatty acids(Ca-LCFA) excreted greater urine N (g/day) thanthose fed control diets (P = 0.0006). The authorsattribute this to reduced microbial proteinsynthesis in the Ca-LCFA diets with concomitantincrease in ammonia conversion to urea andexcretion in the urine. Fecal N (g/day) wasnumerically less in cows fed the Ca-LCFA diets,demonstrating that a shift in N excretion patternsmay have occurred. While this study was notdesigned to examine strategies to shift Nexcretion in the dairy animal, future studies areneeded to investigate dietary manipulation of Nexcretion to promote reduced ammoniaemissions. No literature was found to confirmor refute the theory that as we push ruminantstowards higher grain diets, we are increasingammonia emissions. However, the need to testthis theory is evident. The economics of shiftingthe high producing dairy cow to a diet with agreater proportion of forage needs considerationas part of a whole farm scenario analysis.

pH manipulation of manure by diet

Sutton et al. (1997) observed a reductionin swine manure pH when 5% cellulose wasadded to the diet. Canh et al. (1997a) observedreduced ammonia emissions of 26 to 53% by

including Ca-salts up to dietary Ca levels of 7to 10 g/kg. Resulting reductions in urinary pHranged from 1.6 to 1.8 units. Hendricks et al.(1997) observed a 37% reduction in ammoniaemissions following feeding calcium benzoateto grow-finishing pigs. Kim et al. (2000)observed a 30% reduction in ammonia emissionsassociated with growing pig diets containing acombination of phosphoric acid and calciumsulfate and lesser reduction in emission (17%)when diets contained a combination ofmonocalcium phosphate, calcium sulfate, andcalcium chloride. Relative to the control diets,no ammonia emission reductions were observedwhen diets contained a combination ofmonocalcium phosphate and calcium sulfate,despite significantly reduced urinary pH in theseanimals. This may relate to the buffering capacityin animals fed these diets. Although no workwas found that tested pH manipulation of theruminant animal, investigation of such as anapproach is worthy of consideration. However,the buffering capacity of the rumen may precludefavorable results.

Ammonia binding, urease inhibition

Several feed additives claim to reducenitrogen excretion and ammonia emissionpotential by binding ammonia or inhibitingurease. Yucca plant extracts have been evaluatedby Powers et al. (1999) finding that ammoniaconcentrations in stored dairy manure werereduced. The product fed claimed that ureaseinhibition was the mode of action. Amon et al.(1995) fed a De-Odorase®, a yucca extract, tofattening pigs and observed reduced ammoniaconcentrations in the feeding rooms over a 7-wk period. Ammonia concentration was reduced,on average, 26% in rooms where the extract wasfed. Similarly, ammonia emission was reduced26% in the study. Alltech, Inc. (Nicholasville,KY), the manufacturer of De-Odorase® , believesthat the mode of action is through binding ofammonia rather than urease inhibition. Dietaryinclusion of clinoptilolite and other clay mineralsto reduce ammonia emissions has resulted in


127

variable findings. Kithome et al. (1999)observed that application of a layer of 38%zeolite placed on the surface of the compostingpoultry manure reduced NH3 losses by 44%.Amon et al. (1997) observed greater ammoniaconcentration and emission when clinoptilolitewas used in broiler houses. However, use ofzeolites, as post-excretion amendments ordietary constituents, have long been studied fortheir effectiveness in binding ammonia.Continued investigation into binding andinhibition strategies is warranted as is delineationof the mechanism of effective strategies.

Summary

Odor remains the trigger that spursneighbors into filing complaints against animalproduction facilities. However, where previouslyneighbors viewed odor as a nuisance, there isincreasing concern by residents that the odorcauses negative health effects. So, the motivationto restrict emissions from animal operations hasgained momentum. This is combined, then, withEPA’s own efforts to meet the National AmbientAir Quality Standards that were establishedunder the 1997 Clean Air Act Amendments, aswell as EPA’s interest in adding PM2.5(particulate matter with an aerodynamicdiameter of 2.5 microns or greater) to the list ofNAAQS pollutants. Because ammonia serves asthe nucleus for PM2.5 formation, and because thefindings of EPA’s monitoring studies suggest thatanimal agriculture is the large contributor toammonia emissions, we need to continue totarget N feeding practices from the standpointof air quality-related issues, while continuing toaddress P due to water quality-related issues.Non-traditional approaches will be necessary.Beyond nutrition, we need to be thinking aboutthings such as manipulating the gene thatexpresses urease production, the role of beta-agonists in environmental plans, and othermechanisms to improve nutrient utilization,including feed processing. As the challengesbecome greater, so must the breadth of expertiseinvolved in developing solutions.

References

Aarnink, A.J.A., P. Hoeksma, and E.N.J.Ouwerkerk. 1993. Factors affecting ammoniumconcentration in slurry from fattening pigs. Pages413-420. In: Proc. of the First InternationalSymposium on Nitrogen Flow in Pig Productionand Environmental Consequences. EAAPPublications No. 69. Pudoc, Wageningen, TheNetherlands.

Amon, M., M. Dobeic, T.H. Misselbrook, B.F.Pain, V.R. Phillips, and R.W. Sneath. 1995. Afarm-scale study on the use of De-Odorase forreducing odour and ammonia emissions fromintensive fattening piggeries. Bioresource Tech.51:163-169.

Amon, M., M. Dobeic, R.W. Sneath, V.R.Phillips, T.H. Misselbrook, and B.F. Pain. 1997.A farm-scale study on the use of clinoptilolitezeolite and de-odorase for reducing odour andammonia emissions from broiler houses.Bioresource Technol. 61:229-237.

Beede, D. K. and J. A. Davidson. 1999.Phosphorus: Nutritional Management for Y2Kand Beyond. Tri-state Dairy NutritionConference, Fort Wayne, IN. The Ohio StateUniversity, Columbus.

Blair, B., J. Jacob, and T.Scott. 1995. Reducingmanure pollution. Poultry Highlights. ResearchReport, Page 2. Department of Animal Science,The University of British Columbia.

Canh, T.T., A.J.A. Aarnink, Z. Mroz, and A.W.Jongbloed. 1997a. Influence of dietary calciumsalts and electrolyte balance on the urinary pH,slurry pH and ammonia volatilization from slurryof growing finishing pigs. J. Anim. Sci.75(Suppl 1):1989. (Abstr.)


128

Canh, T.T., M.W.A. Verstegen, A.J.A. Aarnink,and J.W. Schrama. 1997b. Influence of dietaryfactors on nitrogen partitioning and compositionof urine and feces of fattening pigs. J. Anim.Sci. 75:700-706.

Canh, T.T., A.J. Sutton, A.J.A. Aarnink, J.W.Schrama, M.W.A. Verstegen, and G.C.M.Bakker. 1998. Dietary carbohydrates alter fecalcomposition and pH and ammonia emissionfrom slurry of growing pigs. J. Anim. Sci.76:1887-1895.

Elwinger, K. and L. Svensson. 1996. Effect ofdietary protein content, litter and drinker typeon ammonia emission from broiler houses.Agric. Eng. Res. 64:197-208.

EPA, 1998. National air pollutant emissiontrends, 1990-1998.

Ferguson, N.S., R.S. Gates, J.L. Taraba, A.H.Cantor, A.J. Pescatore, M.L. Straw, M.J. Ford,and D.J. Burnham. 1998. The effect of dietaryprotein and phosphorus on ammoniaconcentration and litter composition in broilers.Poultry Sci. 77:1085-1093.

Hendriks, G.L., M.G.M. Vrielink, and C.M.C.van der Peet-Schwering. 1997. Reducingammonia emission from pig housing by addingacid salts to the feed. Pages 65-70. In:Proceeding of the Fifth International LivestockEnvironment Symposium, ASAE, St. Joseph,MI.

Hobbs, P.J., T.H. Misselbrok, and B.F. Pain.1998. Emission rates of odorous compoundsfrom pig slurries. J. Sci. Food Agric. 77:341-348.

Jacob, J.P., R. Blair, D.C. Benett, T. Scott, andR. Newbery. 1994. The effect of dietary proteinand amino acid levels during the grower phaseon nitrogen excretion of broiler chickens. Page137. In: Proceedings of the 29th PacificNorthwest Animal Nutrition Conference.Vancouver, B.C. Canada.

James, T., D. Meyer, E. Esparaza, E.J. DePeters,and H. Perez-Monti. 1999. Effects of dietarynitrogen manipulation on ammoniavolatilization from manure from Holsteinheifers. J. Dairy Sci. 82:2430-2439.

Kay, R.M. and P.A. Lee. 1997. Ammoniaemission from pig buildings and characteristicsof slurry produced by pigs offered low crudeprotein diets. Pages 253-259. In: InternationalSymposium on Ammonia and Odour Controlfrom Animal Production Facilities. Rosmalen,The Netherlands.

Kerr, B.J. 1995. Nutritional strategies for wastereduction management. Page 47 In: Newhorizons in animal nutrition and health. TheInstitute of Nutrition, North Carolina StateUniversity.

Kim, I., D. Kim, and T. van Kempen, 2000.Effects of different sources of phosphorus andcalcium on urine pH and ammonia emission. J.Anim. Sci. 78(Suppl. 2):69. (Abstr.)

Kithome, M., J.W. Paul, and A.A. Bomke. 1999.Reducing nitrogen losses during simulatedcomposting of poultry manure using adsorbentsor chemical amendments. J. Environ. Qual.28:194-201.

Lenis, N.P. and A.W. Jongbloed. 1999. Newtechnologies in low pollution swine diets: dietmanipulation and use of synthetic amino acids,phytase and phase feeding for reduction ofnitrogen and phosphorus excretion and ammoniaemissions – review. Asian-Aus. J. Anim. Sci.12:305-327.


129

Knowlton, K.F. and J. H. Herbein. 2001.Phosphorus partitioning during early lactationin dairy cows fed diets varying in phosphoruscontent. J. Dairy. Sci. 85(5):1227-1236.

Morse, D., H. H. Head, C.J. Wilcox, H.H. VanHorn, C.D. Hissem, and B. Harris, Jr. 1992.Effects of concentration of dietary phosphoruson amount and route of excretion. J. Dairy Sci.75: 3039-3049.

National Research Council. 2001. NutrientRequirements of Dairy Cattle. 7th rev. ed. Natl.Acad. Sci., Washington, DC.

Powers, W.J., H.H. Van Horn, A.C. Wilkie, C.J.Wilcox, and R.A. Nordstedt. 1999. Effects ofanaerobic digestion and additives to effluent orcattle feed on odors and odorant concentrations.J. Anim. Sci. 77:1412-1421.

Sutton, A.J., K.B. Kephart, J.A. Patterson, R.Mumma, D.T. Kelly, E. Bogus, B.S. Don, D.D.Jones, and A.J. Heber. 1997. Dietarymanipulation to reduce ammonia and odorouscompounds in excreta and anaerobic manurestorage. Pages 245-252. In: InternationalSymposium on Ammonia and Odour Controlfrom Animal Production Facilities, Rosmalen,The Netherlands.

Tomlinson, A. P., W.J. Powers, H.H. Van-Horn,R.A. Nordstedt, and C.J. Wilcox. 1996. Dietaryprotein effects on nitrogen excretion and manurecharacteristics of lactating cows. Trans of theASAE 39:1441-1448.

Wu, Z., L.D. Satter, A.J. Blohowlak, R.H.Stauffacher, and J.H. Wilson. 2001. Milkproduction, estimated phosphorus excretion, andbone characteristics of dairy cows fed differentamounts of phosphorus for two or three years. JDairy Sci. 84(7):1738-1748.

Wu, Z., L.D. Satter, and R. Sojo. 2000. Milkproduction, reproductive performance, and fecalexcretion of phosphorus by dairy cows fed threeamounts of phosphorus. J. Dairy Sci.83(5):1028-1041.


130

Table 1. Phosphorus content of typical feedlot feed ingredients.

Feed P content (% of DM)

Soybean meal 0.70Corn 0.35Corn gluten feed 0.90Distillers grains with solubles 0.70Condensed corn distillers solubles 1.00

Table 2. Phosphorus balance for lactating cow diets with and without corn co-products.

Wu et al. (2000), Wu et al. (2000),0.31% P 0.49% P Co-product diet

Ingredient, %Alfalfa silage 30.00 30.00 30.00Corn silage 20.00 20.00 20.00High moisture ear corn 25.50 25.50 12.00Soybeans, roasted 10.00 10.00 3.00Distillers dried grains w/ solubles 0.00 0.00 23.50Bloodmeal 2.00 2.00 2.00Molasses 5.00 5.00 5.00Beet pulp 7.00 6.16 4.00Salt 0.40 0.40 0.40Mineral and vitamin mix 0.10 0.10 0.10Monosodium phosphate 0.00 0.84 0.00Diet compositionCP, % 17.8 17.8 17.8Ca, % 0.65 0.65 0.65P, % 0.31 0.49 0.39NEL, Mcal/kg 1.65 1.65 1.65Nutrient management factorsDMI, kg/day 23.0 23.4 23.0P intake, g/day 72.8 111.8 89.7Fecal P excretion, g/day 39.9 68.9 54.0Fecal P, kg/year 14.6 25.1 19.1Corn hectares (P-basis)1,2 0.56 0.96 0.73Silage consumed, ha/yr (30% DM) 0.17 0.17 0.17Silage for export, ha3 0.39 0.79 0.56

11.75 g P2O5 (0.7777 g P) needed to grow 1 kg of corn silage. Assumed silage yield of 33,695 kg/ha(14.8 T/acre). Therefore, 26.2 kg P needed per hectare of corn silage grown (1 hectare = 2.5 acres).2Assuming harvested at 30% DM, DM yield equals 10,109 kg/ha (23.1 lb/acre).3Silage available for feeding to other animals (in addition to the lactating cows) or for sale.


131

Figure 1. Directly emitted national particulate matter (PM10) emissions by principal source categoriesfor nonfugitive dust sources (EPA, 1998).

Figure 2. Directly emitted national particulate matter (PM2.5) emissions by principal source categoriesfor nonfugitive dust sources (EPA, 1998).

9%6%

8%

14%

12%

32%

19%

Other IndustrialProcesses - 9%

Fuel Combustion -Industrial - 6%

Fuel Combustion -Electrical Utility - 8%

Fuel Combustion -Other - 14%

All Other - 12%

Area SourceCombustion - 32%

On-Road and Non-Road Engines andVehicles - 19%

6%5%

6%

16%

9%37%

21%

Other IndustrialProcesses - 6%

Fuel Combustion -Industrial - 5%

Fuel Combustion -Electrical Utility - 6%

Fuel Combustion -Other - 16%

All Other - 9%

Area SourceCombustion - 37%

On-Road and Non-Road Engines andVehicles - 21%


132

Figure 3. National ammonia emissions by principal source categories (EPA, 1998).

5%4%

3%

2%

86%

On-Road and Non-Road Engines andVehicles - 5%All Other - 4%

Chemical & AlliedProduct Mfg. - 3%

Waste Disposal &Recycling - 2%

Misc. (includeslivestock andfertilizer) - 86%


133

Practicalities of Balancing Diets for Amino Acids

Essi Evans1

Essi Evans Technical Advisory Services, Inc.

Abstract

Formulating diets to meet the amino acidrequirements of cows is not widely practiced,even though the concept is not new. Where theprocedure is accepted because of reasonablyaccurate formulation tools, the benefits includeconsistent production performance along withthe potential to reduce nitrogen wastage. Aminoacid nutrition for ruminants is somewhat morecomplicated than for monogastrics due to theinvolvement of rumen microbes. However,microbes provide the greatest potential formeeting requirements and reducing nitrogenlosses. This paper provides a practical approachto understanding amino acid nutrition of dairycows.

Why Balance Diets for Amino Acids?

There is an urgent need to optimizeprotein and amino acid use in dairy cows. Excessnitrogen (N), produced by poorly formulatingor over formulating the protein portion of thediet, is a burden to both the environment and tothe cow. Many producers are following strictguidelines that limit manure application. At thesame time, there is pressure to reduce the use ofanimal byproduct feeds, often resulting in theneed to increase ration protein. On the other sideof the equation, cows utilize additional energywhen converting excess N and amino acids tourea for elimination. Diets need to be formulatedto reduce oxidation of amino acids to the extent

possible by feeding the correct amounts of aminoacids whenever possible and whenevereconomical.

Another consideration for usingavailable amino acid technology revolves aroundanimal health. Amino acids are key componentsof proteins required for the production ofenzymes, immunoglobins, some hormones,muscle, and milk. Amino acids contribute to theformation of glucose, which can be in shortsupply, particularly in early lactation. When thefeed fails to supply sufficient amino acids, netcatabolism of tissues occurs in order to supplyamino acids for the most critical functions.Insuring that the correct amounts of amino acidsare available contributes to productiveperformance by supporting physical andphysiological conditions.

Although the benefits of using aminoacids is well understood, many feed formulatorsare lost in the apparent controversy over productsand programs available. The purpose of thispaper is to remove some of the mysticsurrounding this technology and to clear the pathfor greater application.

Sources of Amino Acids

In order to meet the cow’s need for aminoacids, there must be an understanding of not justthe needs, but also the supply and the availabilityof amino acids. The principle sources of amino

1Contact at: 64 Scogog St, Bowmanville, ON, Canada, L1C 3J1, (905) 623-7599, FAX (905) 623-6841, Email:[email protected]


134

acids are: microbial protein, escape feed protein,and protected amino acids.

Microbial protein

Microbial protein is by far the mostimportant source of protein available to the cow.The amino acid composition of microbial proteinhas a profile that more closely approximates thatof milk and somatic tissue than many of thevegetable protein sources commonly used(Table 1). It is advisable to maximize theproduction of microbial protein, as it is generallythe cheapest source of protein available.

The rumen provides roughly 100 to150 g of microbial protein/kg of DM consumed(Verbic, 2002). The higher levels are attainablewhen sufficient nutrients are supplied to supportthe growth of the microbes. Much like cows,rumen microbes are sensitive to the nutrientsupply available to them. Microbial productionis influenced by the availability of fermentablecarbohydrate (Ahvenjarvi et al., 2002), nitrogen(Verbic, 2002), and minerals (Broudiscou et al.,1999; Carneiro et al., 2000).

Most systems for predicting microbialprotein utilize some measurement of energy topredict the amount of protein that can beproduced in the rumen (INRA, 1988; AFRC,1992; NRC, 2001). However, microbial proteincannot be readily predicted from the amount ofenergy or fermentable carbohydrate supplied(Orskov, 1994; Dijkstra et al., 1998), except atsub-optimal input levels. Growth will beproportionately increased when supply to therumen is below optimum; however, an oversupply does not result in extra growth any morethan it does in the cow. In the cow, oversupplying feed simply results in greater bunklosses and increased cost. This may explain whymicrobial efficiency (microbes produced/carbohydrate assumed to be fermented) is oftenreduced with high grain diets.

Microbial growth and microbialefficiency is improved at higher turnover rates.Turnover rate has long been known to bedependent upon feed intake (Evans, 1981) andhigh growth rate results in the dilution ofmicrobial maintenance and improved efficiency(Meng et al., 1999). Therefore, insuring thatintake is maximized is an ideal system forreducing the need for escape protein fromingredients and gaining the most from the rumen.

The nutritional requirements of rumenmicrobes may differ from that of the host,particularly in the area of intermediarymetabolism. Microbial production may beenhanced or reduced with the addition ofminerals, such as sulfur (Carneiro et al., 2000)or phosphorus (Broudiscou et al., 1999). Theaddition of unprotected fat, particularly if thefat is unsaturated, can inhibit protozoa growthand fiber digestion (Oldick and Firkins, 2000).

Under certain conditions, microbialgrowth has been shown to be greater thananticipated. This may largely reflect the lack ofunderstanding of the needs of rumen microbes.Miller-Webster et al. (2002) showed that yeastculture improves the digestion capacity of mixedrumen microbes. Improvements in ammoniauptake at some times after feeding, likelyassociated with increased microbial production,were demonstrated by Ehjalberta et al. (1999).Likewise, other fermentation products have beenshown to result in enhanced microbialperformance (Hoover et al., 2001).

To insure that microbes get enoughsoluble nitrogen to support growth, they need atleast as much as what is incorporated intomicrobial protein, plus some extra to accountfor material that escapes from the rumen beforeit is converted to microbial protein. The amountlost will increase with rumen turnover rate.

If there is not enough nitrogen in therumen to support microbes, then feed will notbe digested properly. Often the manure will


135

appear stiff and fibrous. Bear in mind that there alsoare other causes for this to occur. Results ofexperiments where the flow of nitrogenouscomponents have been measured suggest thatthere is a general overage of ammonia nitrogenemanating from the rumen (Blouin et al., 1995).Saliva also contributes N to the rumen to assistin the support of microbial protein synthesis.Blood urea nitrogen is sometimes used as anindicator of excess soluble nitrogen. However,blood urea nitrogen levels can be high whenthere is also too much total protein in the diet.Excess escape protein will be deaminated, withurea the result. Therefore, using the potential formicrobial protein synthesis, based on intake andconsidering positive and negative modifiers, isa useful guideline to evaluating the adequacy ofthe soluble nitrogen available.

The amino acid compositions of rumenbacteria and rumen protozoa are fairly consistent.In spite of considerable changes in diet andsampling time, Martin et al. (1996) found nodifferences in the amino acid composition ofbacteria and protozoa measured at various timespost feeding. Prestlokken and Harstad (2001)reported no effects of diet on the compositionof bacteria and protozoa. Likewise, Rodriquezet al. (2000) found no differences in the profilesof bacteria that could be associated with levelof intake. The composition of bacteria is,however, different than protozoa (Martin et al.,1996; Korhoren et al., 2002).

Of the microbial protein produced,approximately 70% will be in the form of aminoacids (Korhoren et al., 2002). In other words, if100 g of microbial protein are synthesized perkilogram of feed DM, then 70 will be in theamino acid form. Evans and Patterson (1985)used a mixture of bacteria and protozoa (80:20)to derive an average amino acid pattern forrumen microbes. The proportion was based onolder data, suggesting that of the microbialprotein entering the gut in dairy cows consuminghigh grain diets, 20 to 25% originated fromprotozoa and 75 to 80% from bacteria. The origin

of the profile used in the popular CPM model(O’Connor et al., 1993) is based on the patternfound for bacteria.

There are two ways of providing resultsof amino acid analyses. Some laboratories reportresults from amino acid analyses on a molar,rather than a weight basis. It is important to beable to distinguish the difference. Conventionfor feed ingredients is to express amino acidpercentage on a weight and not a molar basis.

Escape protein

Escape protein is the portion of the totalfeed protein that is not fermented in the rumen.Because each protein source is made up ofseveral proteins with differing solubilities, theamino acid profile of the escape portion of theprotein will differ somewhat from the nativeprotein. Therefore, it is useful to be able topredict the profile for the escape portion. A goodsource of information is Dr. Peter Robinson’s’home page:

http://animalscience.ucdavis.edu/faculty/robinson/default.htm

Protected amino acids

Protected amino acids have beenavailable to the dairy industry for a considerablelength of time. However, the addition of aprotected amino acid will only be of benefit if itis limiting. The value of adding protected aminoacids can be determined with the use of a rationformulating program specifically designed toaddress the issue of the amino acid needs of thedairy cow. Several amino acid products arereadily available on the North American market(Table 2). A complete description of most ofthese has been provided by Lapierre et al. (2002).The potential advantage of being able to targetamino acids is to meet a specific need and loweroverage.


136

Assessing the Needs of the Dairy Cow

Once some control over the supply canbe exercised, it is important to be able to estimatethe amounts that should be supplied to the cow.There must be an awareness of the performancepotential to serve as a target, as well asknowledge of factors that impact this target.There must further be a procedure to assesstemporary or long-term responses to shortages.Then, diets can be formulated to permit the cowto achieve her expected potential.

The lactating dairy cow needs aminoacids for maintenance, growth, and reproduction,as well as for milk production. However, theamount needed for milk is generally much largerthan the rest. O’Connor et al. (1993) and Evansand Patterson (1985) provided estimates of theneeds for maintenance, growth, and productionbased on the efficiencies by which absorbedamino acids are used.

Studies of amino acid uptake to outputratios dating back as far as 1975 (Clark, 1975)have shown that total uptake by the mammarygland are consistently higher than output intomilk for arginine, lysine, and the branch-chainedamino acids. On the other hand, the mammaryuptake of the non-essential amino acids,glutamate, aspartate, and proline is considerablyless than the quantities in milk. Evans (1999)demonstrated that the uptake of total amino acidsis similar to output, indicating that theseparticular essential amino acids are needed tosynthesize the non-essential amino acids foundin milk. Because of this, the pattern of aminoacids needed for milk is different than thecomposition of milk (Table 3).

Amino acid ratios

There is a certain amount of resiliencebuilt into the supply and demand within theudder. Practical formulation systems allowneeds to be met to the extent possible, withoutelevating costs. Maintaining lysine and

methionine in a narrow range relationship to eachother, especially without concern for the remainingessential amino acids, has the potential to elevatecosts and provides no proven benefit. Althoughefficiency may be higher when a ratio is maintainedbetween all amino acids, the ability to supply themin a particular ratio may pose difficulties. As a result,one or more amino acids will be over supplied. Ifone of the amino acids found to be in excess happensto be lysine or methionine, then in order to maintaina constant ratio, more of the other amino acid wouldbe required. If needs are met, then furthermaintaining of a constant ratio will only resultin additional wastage.

McGuire (1998) established that themammary gland could alter the extent ofextraction of nutrients based on the relationshipbetween supply and need. This means that ifthere is an overage of one amino acid, there isno advantage to over formulating for another.Lacasse et al. (1996) demonstrated that secretoryepithelial cells in proximity to arterioles producenitric oxide from arginine. Nitric oxide is avasorelaxant, increasing local blood flow. Themammary gland appears to have the flexibilityto regulate its own supply of nutrients through acombination of altering blood flow rates andaltering percentage extracted to meet the targetrequirement (Maas et al., 1998).

The need for non-essential amino acids

Although systems, however crude, havebeen established to predict the utilization ofessential amino acids, little regard has been givento the use of non-essential amino acids, and howthey may impact the need for essential aminoacids. It must be borne in mind that, ifnonessential amino acids are not available, theywill be synthesized from essential amino acids,much as occurs in the mammary gland. As well,Kreb’s cycle intermediates contribute to theformation of non-essential amino acids.


137

The availability of non-essential amino acidsmay therefore be important in insuring the adequacyof the essential amino acid supply. Lobley et al.(2001) indicated that glutamine protects againstmethionine oxidation and that this amino acid canalter the need for methionine. As another example,glutamate is preferred over glucose as an energysubstrate for mucosa cells. Stoll et al. (1999)determined that 94% of the enteral glutamate, butonly 6% of the enteral glucose, was utilized by thegut mucosa. The gut requirement for glutamate mayresult in withdrawal of this amino acid from somatictissues and the need to synthesize glutamate fromarginine and proline.

The mammary gland and somatic tissueshave the capability of synthesizing non-essentialamino acids. However, it is important not toforget that the roles of these amino acids can beessential, and that as such, it may influence thesupply of essential amino acids. It is advisableto provide total amino acids at levels thatapproximate the quantity needed for proteinsynthesis.

Conclusions

Cows have amino acid requirements andrespond to a change in supply. Both the supplyof amino acids and the needs for amino acidscan be estimated with reasonable confidencewith the information currently available. Greateruse of amino acid balancing programs would beinstrumental in meeting performance potential.In addition, amino acid programs are helpful forwhole farm nutrient management considerationsand should be used to determine how to bestmeet the producers’ goals.

References

AFRC, 1992. Nutritional requirements of ruminantanimals. Protein Nutr. Abst. Rev. Series B 62:787-835.

Ahvenjarvi, S., A. Vanhatalo, and P. Huhtanen.2002. Supplementing barley or rapeseed mealto dairy cows fed grass-red clover silage. 1.Rumen degradability and microbial flow. J.Anim. Sci. 80:2176-2187.

Blouin, J.P., J.F. Bernier, G.E. Lobley, C.K.Reynolds, P. Debreuil, and H. Lapeirre. 1995.Effects of protein degradability on proteinmetabolism in dairy cows. J. Dairy Sci .78(Suppl. 1): 213. (Abstr.)

Broudiscou, L-P., Y. Papon, and A.F. Broudiscou.1999. Effects of minerals on feed degradationand protein synthesis by rumen microbes in adual effluent fermenter. Reprod. Nutr. Dev. 39:1999.

Carneiro, H., R. Puchala, F.N. Owens, T. Sahlu,K. Qi , and A.L. Goetsch. 2000. Effects of dietarysulfur level on amino acid concentrations inruminal bacteria of goats. Small Rum. Res.37:151-157.

Clark, J.H. 1975. Lactational responses topostruminal administration of proteins andamino acids. J. Dairy Sci. 58:1178-1197.

Degussa Corporation. 2001. AminoDat 2.0.Interactive Software. Degussa AG. Hanau-Wolfgang, Germany.

Dijkstra, J., J. France, and D.R. Davies. 1998.Different mathematical approaches to estimatingmicrobial protein supply in ruminants. J. DairySci 81:3370-3384.


138

Enjalberta, F., J.E. Garrett, R. Moncoulonc, C.Bayourthec, and P. Chicoteaud . 1999. Effects ofyeast culture (Saccharomyces cerevisiae) onruminal digestion in non-lactating dairy cows. Anim.Feed Sci. Technol. 76:195-206.

Evans, E. 1981. An evaluation of therelationships between dietary parameters andrumen solid turnover rate. Can J. Anim. Sci.61:91-96.

Evans, E. 1982. A dynamic model of crudeprotein digestion and utilization in ruminants.J. Anim. Sci. 55 (Suppl. 1):419. (Abstr.)

Evans, E. 1999. Criteria for assessing amino acidutilization for milk protein synthesis. Proc.Eastern Nutr. Conf., p. 113-123.

Evans, E. and R.J. Patterson. 1985. Use ofdynamic modelling seen as good way toformulate crude protein, amino acidrequirements for cattle diets. Feedstuffs 57(42):24-25.

Evans, E., S.A. Yorston, and D. Van Binnendyk.1993. Numerous factors affect milk proteinpercentage. Feedstuffs 65 (10):14-17.

INRA, 1988. Alimentation de bovins, ovins etcaprins. INRA, Paris.

Hoover, W.H., T.M. Miller, J. E. Nocek, andW.E. Julian. 2001. Interaction betweenFermenten or soybean meal and fermentabilityof carbohydrate source on microbial yield andefficiency in continuous culture. J. Dairy Sci.84(Suppl. 1):80 (Abstr.)

Korhoren, M., S. Ahvenjarvi, A. Vanhatalo, andP. Huhtanen. 2002. Supplementing barley orrapeseed meal to dairy cows fed grass-red cloversilage. II. Amino acid profile of microbialfractions. J. Anim. Sci. 80:2188-2196.

Lacasse, P., V.C. Farr, S.R. Davis, and C.G.Prosser. 1996. Local secretion of nitric oxideand the control of mammary blood flow. J. DairySci. 79: 1369-1374.

Lapierre, H., R. Berthiaume and L. Doepel.2002. Rumen protected amino acids: Why, whatand when? Proc. Maryland Nutr. Conf. p. 86-94.

Lobley, G.E., S.O. Hoskins, and C.J. McNeil.2001. Glutamine in animal science andproduction. Brit. J. Nutr. 131:2525S-2531S.

Maas, J.A., J. France, and B.W. McBride. 1998.Towards a mechanistic model of amino aciduptake and metabolism by the mammary glandof the lactating cow. A literature review. J. Anim.Feed Sci. 7:105-129.

Martin, C., L. Bernard, and B. Michalet-Doreau.1996. Influence of sampling time and diet onamino acid composition of protozoal andbacterial fractions from bovine ruminal contents.J. Anim. Sci. 74:1157-1163.

McGuire, M.A. 1998. The potential to enhancemilk protein. Proc. Pacific NW Anim. Nutr.Conf. pg. 157.

Meng, Q., M. S. Kerley, P.A. Ludden, and R. L.Belyea. 1999. Fermentation substrate anddilution rate interact to affect microbial growthand efficiency. J. Anim. Sci. 77:206-214.

Miller-Webster, T., W. H. Hoover, M. Holt, andJ. E. Nocek. 2002. Influence of yeast cultureon ruminal microbial metabolism in continuousculture J. Dairy Sci. 85:2009-2014.

NRC. 2001. Nutrient Requirements of DairyCattle. Seventh Revised Edition. NationalAcademies Press. Washington, DC.


139

O’Connor, J.D., C.J. Sniffen, D.G. Fox, and W.Chalupa. 1993. A net carbohydrate and proteinsystem for evaluating cattle diets. IV. Predictingamino acid adequacy. J. Anim. Sci. 71: 1298-1311.

Oldick, B.S., and J.L. Firkins. 2000. Effects ofdegree of fat saturation on fiber digestion andmicrobial protein synthesis when diets are fed12 times daily. J. Anim. Sci. 78:2412-2420.

Orskov, E.R. 1994. Recent advances inunderstanding of microbial transformations inruminants. Livestock Prod. Sci 39:53-60.

Prestlokken, E. and O.M. Harstad. 2001. Effectsof expanded-treating a barley based concentrateon ruminal fermentation, bacterial N synthesis,escape of dietary N and performance of dairycows. Anim. Feed Sci. Technol. 90: 227-246.

Rodriguez, C.A. , J. Gonzalez, M.R. Alvir, J.L.Repetto, C. Centeno, and F. Lamrani. 2000.Composition of bacteria harvested from liquidand solid fractions of the rumen of sheep asinfluenced by feed intake. Br. J. Nutr. 84:369-376.

Stoll, B., D.G. Burrin, J. Henry, H. Yu, F.Jahoor, and P.J. Reeds. 1999. Substrateoxidation by the portal drained viscera of fedpiglets. Am. J. Physiol: 277:E168-E175.

Van Soest, P. J., C. J. Sniffen, D.R. Mertens, D.G.Fox, P.H. Robinson, and U. Krishnamoorthy.1982. A net protein system for cattle:the rumensubmodel for nitrogen. Page 265 in ProteinRequirements for Cattle: Proceedings of anInternational Symposium (F.N. Owens, Ed.),Oklahoma State Publ. MP-109, Oklahoma StateUniversity.

Verbic, J. 2002. Factors affecting microbialprotein synthesis in the rumen with emphasison diets containing forages. Bericht 29.Viehwirtschaftliche Factgang, BAL,Gumpenstein April 24-25, pages 1-6.


140

Table 1. Amino acid profiles of selected items (g/100 g amino acids).

Amino Rumen Rumen Soybean Distillers GlutenAcid Bacteriaa Protozoaa Meal b Grainsb Mealb Caseinb

Arginine 4.8 4.6 7.3 3.7 3.0 3.7Histidine 2.1 1.8 2.6 2.2 1.9 3.1Isoleucine 5.2 6.0 4.5 3.5 3.9 5.2Leucine 8.2 8.1 7.6 10.4 16.0 9.6Lysine 7.1 10.2 6.0 2.3 1.6 8.1Methionine 2.0 1.7 1.3 1.8 2.4 3.0Phenylalanine 5.5 5.5 5.1 4.6 6.1 5.2Threonine 5.8 5,6 3.6 3.5 5.3 4.2Valine 5.6 5.3 4.8 4.8 4.4 6.6

Table 2. Companies providing protected amino acids or amino acid analogs in North America.

Company Amino Acid Provided

Adisseo Methionine, methionine analogChurch and Dwight Co., Inc. Methionine analogDegussa Corporation MethionineJefo Nutrition, Inc. Methionine, lysineNisso, Inc. MethionineNovus International, Inc. Methionine analog

Table 3. Uptake of amino acids by the mammary gland as compared to output in milk (g/100 g aminoacids).a

aKorhoren et al., 2002.bDegussa Corporation, 2001.

Amino Acid Total Uptake Mammary Output

Arginine 8.53 3.40Histidine 3.29 2.74Isoleucine 8.80 5.79Leucine 13.04 9.18Lysine 9.14 3.40Methionine 2.82 2.71Phenylalanine 4.51 4.75Threonine 4.76 3.72Valine 10.01 5.89a Evans, 1999

Documents

Proceedings Tri-State Dairy Nutrition Conference