Interspecific Differences in the Diel Vertical Migration of Marine Copepods: The Implications of Size, Color, and Morphology

Interspecific Differences in the Diel Vertical Migration of Marine Copepods: The Implicationsof Size, Color, and MorphologyAuthor(s): G. C. Hays, C. A. Proctor, A. W. G. John and A. J. WarnerSource: Limnology and Oceanography, Vol. 39, No. 7 (Nov., 1994), pp. 1621-1629Published by: American Society of Limnology and OceanographyStable URL: http://www.jstor.org/stable/2838198 .

Accessed: 15/06/2014 23:48

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

American Society of Limnology and Oceanography is collaborating with JSTOR to digitize, preserve andextend access to Limnology and Oceanography.

http://www.jstor.org

This content downloaded from 91.229.248.187 on Sun, 15 Jun 2014 23:48:21 PMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=limnoc

http://www.jstor.org/stable/2838198?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


Limnol. Oceanogr., 39(7),1994,1621-1629 ? 1994, by the American Society of Limnology and Oceanography, Inc.

Interspecific differences in the diel vertical migration of marine copepods: The implications of size, color, and morphology

G. C. Hays,' C. A. Proctor, A. W. G. John, and A. J. Warner The Sir Alister Hardy Foundation for Ocean Science, The Laboratory, Citadel Hill, Plymouth PLI 2PB, U.K.

Abstract Samples collected by continuous plankton recorders (CPRs) between 1948 and 1992 were used to

describe the diel vertical migration (DVM) behavior of 41 copepod taxa in the northeast Atlantic between 45 and 55?N and 11 and 31?W. A total of 13,622 samples, each representing - 18.5 km (10 nm) of tow, were analyzed. Since CPRs are towed in near-surface waters, taxa that exhibit DVM occur predominantly in samples taken at night. Larger taxa showed significantly stronger DVM, with body size explaining 47% of the intertaxa variation in DVM. For small taxa (< 1 mm wide) the residual variation in DVM was correlated with carotenoid pigment levels but not with body morphology, with more heavily pigmented taxa exhibiting DVM. For larger taxa (> 1 mm wide) the residual variation in DVM was correlated with body morphology but not with carotenoid pigment levels, with more elongate copepods not exhibiting DVM.

Diel vertical migration (DVM) occurs in a wide range of both marine and freshwater copepods as well as in a variety of other zooplankton, with the "normal" pattern being movement from greater depths during the day to shallower depths at night (Lampert 1989). The ubiquity of DVM in zooplankton communities has provoked considerable and ex- tended debate as to its functional significance (see Lampert 1989). Although it has been suggested that DVM may provide a metabolic advantage associated with the diel movement across a thermocline (McLaren 1963), this hypothesis is now generally refuted (Lampert 1989). Alternatively, Zaret and Suffern (1976) suggested that DVM may be an antipredator behavior, with populations moving to a greater depth during the day to reduce risk from visual predators.

A considerable amount of both field and experimental evidence has accumulated to sup- port this predator-evasion hypothesis. For example, Gliwicz (1986) reported that the

' Present address and address for correspondence: School of Ocean Sciences, University of Wales at Bangor, Menai Bridge, Gwynedd LL59 5EY, U.K. Acknowledgments

We thank all the staff who have analyzed CPR samples in the last 45 yr and Steve Coombs for comments on the manuscript. The CPR is funded by the Ministry of Agri- culture Fisheries and Food (U.K.), Department of the En- vironment (U.K.), Department of Fisheries and Oceans (Canada), NOAA/NSF (USA), the European Commission, and the Consortium (Netherlands).

amplitude of DVM was greater in the copepod populations of lakes where there were planktivorous fish rather than in fishless lakes and that the amplitude of DVM was correlated with the age of the fish population. These observations suggest that upon exposure to a high risk of visual predation there was natural selection for more migratory individuals and hence the gradual change from less to more migratory populations. Similarly, in a series of papers Frost and Bollens have reported that the extent of DVM in certain marine copepods may correlate with the abundance of planktivorous fish (cf. Bollens and Frost 1989). Recent experimental manipulations have shown that the change from a nonmigratory to a migratory mode can be rapid (< 1 h) upon the introduction of planktivorous fish (Neill 1990), illus- trating that population changes in DVM may also occur as a result of individual changes in behavior (i.e. as a rapid phenotypic response).

The evidence supporting the predator-evasion hypothesis is compelling in these exam- ples, but doubt remains over the ubiquity of this hypothesis to account in general for DVM (Lampert 1989). The generality of predator evasion as the ultimate cause of DVM in pelagic communities will only be derived slowly from autecological studies. What is ideally needed to test the generality of this hypothesis are synecological studies where differences in DVM within communities are examined and the causality of these differences established. If individuals do indeed migrate principally to lessen their risk of mortality. then intersDecific

1621



1622 Hays et al.

Species A Species B Species C

Sea surface

AL AL AL Tow depth

Fig. 1. Schematic representation showing the ability of the CPR to detect vertical migration. Species A does not undertake DVM and will occur equally in samples collected in the day and at night, and thus FDVM ; 0.5 (Eq. 1). Species B undertakes DVM and even though the amplitude of the daily vertical movement is small, FDVM

1.0. Species C also undertakes DVM and even though the amplitude of the daily vertical movement is large, like species B, FDVM - 1.0. The width of each diamond represents the abundance of the species: open diamond- day; closed diamond-night.

differences in the extent of DVM should be correlated with the susceptibility of different species to visual predation.

Collecting such multispecies data sets is, however, difficult. First, when examining DVM it is simpler to collect data for only a few species; second, if data are collected for a wide variety of species, such studies are usually con- ducted intensively over short spatial and tem- poral scales and thus cannot provide a description of the average or typical migration tendencies. However, one data set in which a large range of species has been identified, with samples also being collected in an extensive range of locations and in different seasons and years, is that provided by the continuous plankton recorder (CPR) survey. In this paper we used data collected by the CPR survey to quantify and investigate the causality of interspecific differences in DVM within a copepod community.

Methods Samples were collected with CPRs towed at

high speed (1 1-33 km h- 1) and in near-surface waters (mean depth, 6.5 m) (Hays and Warner 1993) from ships of opportunity. Plankton were retained inside recorders on a continually moving band of silk mesh of 270-,um nominal aperture and preserved in situ with formaldehyde (Hays 1994). After each tow this silk

600 N

326 2211 _0 1340 2437 55? N 128 188 J323 869 1177 9<

298 3971401 568 1013 \ s 500 N 259 259 305 367 1181

145 202 218 281 639 r 45 N

450 N . ffi i ~400 N

50? W 30? W 10? W 10? E

Fig. 2. The study region (shaded) in the northeast At- lantic. The numbers represent the total number of samples collected between 1948 and 1992 in each 20 latitude by 40 longitude area.

roll was unwound, cut into sections corre- sponding to -18.5 km (10 nm) of tow and then the abundance of various copepod taxa was quantified in a routine manner (Colebrook 1960) for alternate sections. Some copepod taxa were identified to species and others only to genus.

The time and location of the CPR deploy- ment and retrieval was recorded for each tow, from which the location and local time for the midpoint of each sample was calculated. Sam- ples were categorized as night or day samples, with night being defined as the time when the sun was below the horizon.

Since CPRs sample at a fixed depth near the surface, taxa that undertake DVM should be absent in samples taken during the day but present in samples taken at night. However, differences in the distance (i.e. amplitude) that species have migrated to reach the surface will not be revealed. For example, a species that migrates during the day to just below the depth of sampling will record the same index of vertical migration as a species that migrates much deeper (Fig. 1). Thus we cannot determine the amplitude of DVM from CPR samples. In- stead we calculated the frequency of DVM (FDVM) as

No. of night samples

FDVM recording the species (1) total No. of samples recording the species (n)

and thus assuming a normal approximation of the binomial distribution,

SD = [FDVM(1 - FDVM)1/n]0?5. (2)



Diel vertical migration 1623

Table 1. The number of samples that recorded each taxa, the measured prosome length-to-width ratio (L: W), and the calculated prosome width.

No. of occur- Width

Taxon rences L: W (mm)

Calanoida Acartia spp. 4,638 3.03 0.33 Aetideus armatus 53 2.63 0.68 Calanoides carinatus 89 3.20 0.75 Calanus helgolandicus 2,300 3.46 0.81 Calanusfinmarchicus 1,370 3.15 0.95 Calanus tenuicornis 127 2.80 0.64 Calocalanus spp. 179 2.80 0.39 Candaciaarmata 309 2.60 1.00 Centropages hamatus 24 2.68 0.52 Centropages typicus 1,944 2.68 0.63 Clausocalanus spp. 1,287 2.75 0.47 Ctenocalanus vanus 30 2.80 0.39 Eucalanus crassus 217 3.28 0.95 Eucalanus elongatus 98 4.00 1.55 Euchaeta acuta 470 2.80 1.36 Euchaeta gracilis 24 2.85 2.35 Euchaeta hebes 78 2.80 1.04 Euchaeta marina 8 2.80 1.18 Euchaeta norvegica 786 2.80 2.68 Euchirella rostrata 341 2.64 1.14 Euchirella messinensis 4 2.70 2.00 Heterorhabdus papilliger 124 2.60 0.77 Heterorhabdus norvegicus 29 2.60 1.15 Lucicutia spp. 52 2.80 0.54 Mecynocera clausi 141 2.80 0.36 Metridia lucens 3,472 2.76 1.05 Nannocalanus minor 808 2.70 0.74 Neocalanus gracilis 519 3.00 1.07 Pleuromamma abdominalis 760 2.73 1.28 Pleuromamma borealis 1,350 2.73 0.73 Pleuromamma gracilis 1,444 2.73 0.73 Pleuromamma robusta 2,302 2.73 1.46 Pleuromamma xiphias 124 2.73 1.68 Pseudocalanus elongatus 693 3.16 0.44 Rhincalanus nasutus 458 4.40 1.14 Scolecithricella spp. 100 2.70 0.52 Temora longicornis 68 2.10 0.62 Undeuchaeta major 16 2.80 1.71 Undeuchaeta plumosa 493 2.79 1.33

Cyclopoida Corycaeus anglicus 251 2.54 0.43 Oithona spp. 4,436 2.93 0.27

Due to the nature of the CPR sampling (Fig. 1), the range of values recorded for FDVM would not be expected to have a continuous distribution. Rather, a bimodal distribution would be expected, with values of -0.5 representing species that do not migrate (assuming the sampling intensity is the same both day and night) and values of 1.0 representing species that do migrate.

12 -

10- x 0

O- _

.0

z 2

0 0.4 0.5 0.6 0.7 0.8 0.9 1.0

FDVM Fig. 3. Number of taxa showing differing FDVM (from

Eq. 1).

The length of different copepod species was taken from published sources (Rose 1933; Ver- voort 1963, 1965; Roe 1972). For taxa identified only to genus, the mean length of that genus was calculated. The ratio of prosome length to width was measured from published diagrams (Sars 1903; Rose 1933) and hence the width of the different taxa calculated. Ca- rotenoid pigment levels were taken from Fish- er et al. (1964), although pigment levels for all the taxa used from the CPR records were not available.

The area bounded by 45-55?N and 11-3 loW was chosen for analysis since this area has been well sampled by CPRs and a diverse size range of copepods are identified in the area (Ocean- ogr. Laboratory, Edinburgh 1973).

Results We analyzed 13,622 samples (representing

- 250,000 km of towing) collected in the study area between 1948 and 1992. Sampling intensity was not uniformly distributed in the study area, but instead tended to be more concen- trated closer to the continental shelf (Fig. 2). Ofthese samples, 6,610 were collected at night. Thus the expected FDVM for a species that did not migrate would be 6,610/13,622 - 0.485.

We identified 41 taxonomic groups of copepod, six to genus and 35 to species. These taxa with the number of samples in which each



1624 Hays et al.

1.0- -

0-9-4,S . *. *

0.8-

0.7-

"-0.6 .

0.5- 0.4 6 50 100 150 200 250

Carotenoid concentration

1.0

0.8- (b) > 0.7- U.

0.6-

0.5 - e 0.4

Prosome length/width

1.0

0.9 * *.

0.8 > 0.7

0.6 (c) 0.5 . .

*0 1000 300 0 500 Number of samples

Fig. 4. Relationships between FDVM (from Eq. 1) in the different copepod taxa and (a) the copepod carotenoid concentration (ug g-'), (b) the prosome length-to-width ratio, and (c) the number of samples in which each taxa was identified. For a straight line model: F1,22 = 0.8, P = 0.39; F, 39 = 2.4, P = 0.13, F1,39= 2.0, P = 0.17.

1.0 -

0.9

0.8

>

0.5 0.7 0.6 -

0.4I 0 1 2 3

Prosome width (mm) Fig. 5. Relationship between FDVM (from Eq. 1) (? 1

SD) in the different copepod taxa and prosome width. For the fitted step function F1, = 34.2, r2 = 0.47, P < 0.001. The point marked 1 is Rhincalanus nasutus and the point marked 2 is Eucalanus elongatus.

was identified and their measured prosome length-to-width ratio and hence their calculated widths are given in Table 1. Carotenoid pigment levels from Fisher et al. (1964) were available for 24 of these taxa.

There were two distinct groups of taxa with respect to FDVM (Fig. 3), with 21 of the species showing a strong tendency to migrate vertically (mean FDVM = 0.880, SD = 0.077) and 20 species showing little tendency to migrate (mean FDVM = 0.485, SD = 0.059).

The tendency to migrate vertically was independent of carotenoid levels, the prosome length-to-width ratio, or the number of samples in which that taxa was identified (Fig. 4). However, larger copepods showed a greater FDVM than smaller species (Fig. 5). We visually fitted a step function to this relationship, assuming that for copepods < 1.0 mm wide FDVM was 0.485 (from the ratio of night samples to total samples) and for copepods > 1.0 mm wide FDVM was 0.9 (from the mode of Fig. 3). This step function explained 47% of the variability in the taxonomic differences in FDVM.

There were, however, several taxa that di- verged markedly from the modeled step function. These anomalous taxa were not all of a similar size, nor was the pattern of their de-




0.6

o ~** o 0.3 E * E o 00. S e 6-. 0.0

a -0.3 *

-0.6 , 0 1 2 3

Prosome width (mm) Fig. 6. The deviations of FDVM (from Eq. 1) from

that predicted by the model based on prosome width (solid line on Fig. 5) plotted against the prosome width of the different copepod taxa.

viation (i.e. positive or negative) from the model the same (Fig. 6). The six most anomalous taxa were Pleuromamma borealis, Pleu- romamma gracilis, Heterorhabdus papilliger, and Lucicutia spp., which all had a large FDVM despite their small size, and Rhincalanus nasutus and Eucalanus elongatus, which did not have a large FDVM despite their large size.

The poor fit of the model may have been due to the inaccuracy with which FDVM was estimated for each taxa. To investigate this, we assumed that the values for FDVM had a perfect fit (i.e. r2 = 1.0) with the model, and then we used the known SD (Eq. 2) to randomly assign an error to the estimated FDVM for each taxa. We then calculated the r2 for the relationship between these realized values for FDVM and the model. We repeated this error simulation 100 times for each taxa. The resulting r2 values averaged 0.97 (SE = 0.001) (i.e. the error introduced by the inaccuracy in the estimate of FDVM was not the main reason for the comparatively poor fit of the model).

An analysis of the unexplained residual variation in vertical migration tendency was di- vided into an analysis of taxa < 1 mm wide (i.e. mainly nonvertical migrators) and of taxa > 1 mm wide (i.e. mainly vertical migrators).

1.0

i.02 3 0.9 3

2 0.8 /

U 0.6 - (a)

0.5 /

0.4 *0 50 ioo 1 Carotenoid concentration

1.0

0.9

0.8

a 0.7 (b) LL

0.6 -

0.5 * * * .

0.4.. 2.0 3.0 4.0 5.0


Fig. 7. FDVM (from Eq. 1) in copepods < 1 mm wide. [a.] FDVM in relation to carotenoid concentration (gg g- '). FDVM = 0.000051 [carotenoid concn]2 + 0.437 (F1,7 = 33.1, r2 = 0.83, P < 0.00 1). The point marked 1 is Het- erorhabdus papilliger, point 2 is Lucicutia spp., and point 3 is Pleuromamma borealis. [b.] FDVM in relation to the prosome length-to-width ratio (F1,22 = 1.4, P = 0.26).

For taxa < 1 mm wide the residual variation in FDVM was positively related to carotenoid pigment levels (Fig. 7a) but was independent of body morphology (Fig. 7b), i.e. the small copepods that migrated, such as P. borealis, H. papilliger, and Lucicutia spp., tended to be more heavily pigmented than small nonmigratory copepods. For taxa > 1 mm wide the residual variation in FDVM was independent of carotenoid pigment levels (Fig. 8a) but was negatively correlated with body morphology (Fig. 8b), i.e. the large copepods that did not migrate (R. nasutus and E. elongatus) were characterized as long and thin.



1626 Hays et al.

1.0 I

0.9 . (a) * 0

> 0.8-

X o.6 2 ?

0.5

04 *0 50 100 150 200 250 Carotenoid concentration

1.0 *

0.9 (b)

s 0.8 .

U- 0.6 1

0.5 -

0.4 . 2.5 3.0 3.5 4.0 4.5


Fig. 8. FDVM (from Eq. 1) in copepods > 1 mm wide. [a.] FDVM in relation carotenoid concentration (,g g-') (FI,13 = 1.4, P = 0.26). [b.] FDVM in relation to the prosome length-to-width ratio. FDVM = 1.51-0.221 (length/ width) (F1,15 = 47.9, r2 = 0.76, P < 0.001). As in Fig. 5, the point marked 1 is Rhincalanus nasutus and the point marked 2 is Eucalanus elongatus.

Discussion Samples collected by CPRs have tradition-

ally been used to examine seasonal, spatial, and long-term trends in plankton abundance. The potential for using the accumulated data set to look at DVM has, however, not been explored. Because CPRs are towed from ships of opportunity rather than dedicated research ships, the tow routes tend to be dictated by the changing merchant shipping lines. Thus uni- form sampling intensity over wide regions would not be expected and was not found in the sampled area. Sampling by CPRs may be

criticized in that the flow rate through the recorder is not measured and so absolute abun- dances of zooplankton cannot be reliably de- termined. This shortcoming will not, however, compromise the use of the CPR data set to examine diel vertical migration through, for example, calculation of FDVM (Eq. 1) because no assumptions are made concerning the absolute abundance of the different taxa. The expected bimodal distribution of values for FDVM resulting from the fixed sampling depth of the CPR was found (Fig. 3), with a value FDVM ` 0.485 representing taxa that did not migrate and a value FDVM ` 0.925 representing taxa that did migrate.

It has been repeatedly suggested that DVM functions to reduce the risk of mortality from visually orientating predators (cf. Lampert 1989). It would therefore be expected that those species with the greatest susceptibility to visual predation should show the greatest FDVM. The axiom of this argument is that zooplankton species differ in their susceptibility to visual predation and, indeed, such prey selection by zooplanktivores has been repeatedly dem- onstrated. In broad terms, species that are more readily perceived and have a poorer escape ability once they have been perceived have the greatest susceptibility to visually orientating predators. Prey perception by visually orientating predators may depend on prey size (Brooks and Dodson 1965) and prey color (Zaret and Kerfoot 1975), although the relative importance of these two factors is equivocal. For example, Brooks and Dodson found that after the introduction of planktivorous fish to a lake, the size distribution of the zooplankton community had changed to a predominance of smaller species, suggesting that the fish had selectively removed the larger species. Con- versely, Zaret and Kerfoot suggested that the planktivorous fish Melaniris chagresi selected the cladoceran Bosmina longirostris on the basis of the amount of pigmentation rather than body size.

The escape ability of zooplankters once they have been perceived will depend on how far away they themselves can perceive an ap- proaching predator, their speed and maneu- verability once they have perceived a predator, and their palatability once they have been caught. Interspecific differences have been




shown in the ability of copepods to avoid a slowly moving siphon intake, i.e. an index of the first and second factors above (Singarajah 1969), and interspecific escape ability differences have also been inferred from field observations (Drenner et al. 1978). However, the factors influencing these differences are un- known. At a higher taxonomic level (i.e. when comparing copepods, cladocerans, and roti- fers), Kerfoot et al. (1980) suggested that escape ability may be correlated with body morphology, with more streamlined taxa having the greater, and hence more effective, escape speed. In summary, zooplankton do differ in their susceptibility to visual predators and this may be due to differences in body size, color, and shape.

We used published information to derive body width, which we used as an index of body size, and, where available, published carotenoid pigment levels. Carotenoid pigments tend to be the major source of pigmentation in planktonic crustacea; thus carotenoid levels probably give a good indication of overall body color (Herring 1972). It may have been more informative to have measured both of these parameters ourselves, since regional variations might reduce the validity of applying literature values to our study area. However, neither of these parameters has been routinely measured during the CPR survey and indeed pigment levels probably cannot be derived from the samples since they lose color as a result of being preserved in formaldehyde.

When all the available taxa were compared, FDVM was not correlated with pigment concentration (Fig. 4a) but was significantly correlated with body size (Fig. 5), with DVM tending to occur in larger taxa. We would therefore suggest that the pattern of interspecific differences in DVM can be accounted for by the predator-evasion hypothesis. Further- more, we would suggest that copepod size rather than color is the principal factor influencing the probability of perception of copepods by visual predators. However, only 47% of the intertaxa variation in FDVM was explained by body size. There are several potential explanations for this observation. There may have been an inherent inadequacy of the way in which the step-function model was fitted. For example, we may have poorly estimated

the copepod size at the position of the step between vertical migrators and nonvertical migrators; alternatively, the upper and lower levels of the step function may have been poorly fitted. However, neither of these explanations is supported by the pattern of residuals (Fig. 6), i.e. copepods that differed markedly from the fitted model were not at a specific size range of the distribution. Some of the taxa were sampled with low incidence (Table 1). For Centropages hamatus, this reflected that the study area was outside the center of distribution for that taxa; for the other taxa (such as Ctenocalanus vanus, Euchaeta gracilis, Eu- chaeta marina, and Euchirella messinensis), this reflected their low overall abundance in CPR records (Oceanogr. Laboratory, Edin- burgh 1973). However, the poor fit of the model was not due to the inaccuracy with which FDVM was estimated as a consequence of the low incidence of certain species, because when we assumed the model and data had a perfect fit and then introduced the known error with which FDVM was estimated, the r2 between the model and the data dropped to only 0.97 (i.e. markedly higher than the observed r2 of 0.47).

Alternatively, the reason for the poor fit of the model may have been due to other factors (besides size) influencing the susceptibility of copepods to visual predation. For small copepods, we found that FDVM was significantly related to carotenoid pigment levels (Fig. 7a). We suggest that this observation is also con- sistent with the predator-evasion hypothesis. For small copepods that are transparent there may be no need to migrate vertically because small size and transparency combine to ensure that such copepods are not readily perceived by visual predators. However, increased color in a small copepod will increase its conspic- uousness and hence the risk of being consumed if it remains near the surface during the day. To avoid this risk necessitates the need to oc- cupy a darker, and hence deeper, depth during the day.

It has been suggested that carotenoid pigments may serve to reduce the risk of damage from visible blue light and UV light; hence in the absence of visually orientating predators, copepods may be heavily pigmented as a result of natural selection by solar radiation (Hair-



1628 Hays et al.

ston 1979). On this basis, carotenoid levels should be higher in species occurring closer to the surface during the day. Although this has been found in certain lakes (Hairston 1980), in the sea maximal carotenoid levels may occur at midwater depths (Herring 1972). Our results also cast doubt on the applicability of the light protection hypothesis in marine sys- tems, since we found that in small copepods DVM was restricted to those taxa having the highest levels of carotenoids, i.e. the taxa with high carotenoid levels moved out of the near- surface waters during the day, the opposite of that predicted from the light protection hypothesis. Therefore, alternative explanations (e.g. carotenoids may serve as a energy store, Ringelberg 1980) need to be found for the function of carotenoids in marine copepods.

For larger copepods there was no relationship between FDVM and carotenoid levels. This result is, however, expected given the fixed sampling depth of the CPR. Large copepods are more conspicuous than small copepods and hence may be selectively consumed by visual predators (Brooks and Dodson 1965). Large copepods would therefore be expected to migrate vertically, as was generally found to be the case (Fig. 5). However, while increased pigment levels might be expected to increase con- spicuousness, and hence the amplitude of DVM, this information would not be revealed by the CPR sampling (Fig. 1). The poor fit of the step-function model to explain FDVM in larger copepods was restricted to two species, E. elongatus and R. nasutus, which were large but did not occur predominantly in samples taken at night (Fig. 5). These two species are, however, characterized by having a markedly different morphology to the other taxa, being far more elongate (Fig. 8b). Given the influence of body morphology on escape speed (Kerfoot et al. 1980), we suggest that the tendency for E. elongatus and R. nasutus to remain in near- surface waters during the day may be related to their morphology facilitating a greater escape ability from visual predators. These two species may therefore be able to remain in near- surface waters during the day without suffering the high risk of mortality expected for copepods of a similar size.

In conclusion, our observations are consis- tent with the risk of predation from visual predators being the most important factor influencing taxonomic differences in DVM.

References BOLLENS, S. M., AND B. W. FROST. 1989. Predator-in-

duced diel vertical migration in a planktonic copepod. J. Plankton Res. 11: 1047-1065.

BROOKS, J. L., AND S. I. DODSON. 1965. Predation, body size, and composition of plankton. Science 150: 28- 35.

COLEBROOK, J. M. 1960. Continuous plankton records: Methods of analysis, 1950-1959. Bull. Mar. Ecol. 5: 51-64.

DRENNER, R. W., J. R. STRICKLER, AND W. J. O'BRIN. 1978. Capture probability: The role of zooplankter escape in the selective feeding of planktivorous fish. J. Fish. Res. Bd. Can. 35: 1370-1373.

FISHER, L. R., S. K. KON, AND S. Y. THOMSON. 1964. Vitamin A and carotenoids in certain invertebrates. J. Mar. Biol. Assoc. U.K. 44: 685-692.

GLIWIcz, Z. M. 1986. Predation and the evolution of vertical migration by zooplankton. Nature 320: 746- 748.

HAIRSTON, N. G., JR. 1979. The adaptive significance of color polymorphism in two species of Diaptomus (Co- pepoda). Limnol. Oceanogr. 24: 38-55.

1980. The vertical distribution of diaptomid copepods in relation to body pigmentation. Am. Soc. Limnol. Oceanogr. Symp. 3: 98-110. New England.

HAYS, G. C. 1994. Mesh selection and filtration efficiency of the continuous plankton recorder. J. Plankton Res. 16: 403-412.

, AND A. J. WARNER. 1993. Consistency of towing speed and sampling depth for the continuous plankton recorder. J. Mar. Biol. Assoc. U.K. 73: 967-970.

HERRING, P. J. 1972. Depth distribution of the carotenoid pigments and lipids of some oceanic animals. J. Mar. Biol. Assoc. U.K. 52: 179-189.

KERFOOT, W. C., D. L. KELLOGG, JR., AND J. R. STRICKLER. 1980. Visual observations of live zooplankters: Eva- sion, escape, and chemical dcfenses. Am. Soc. Lim- nol. Oceanogr. Spec. Symp. 3: 10-27. New England.

LAMPERT, W. 1989. The adaptive significance of diel vertical migration of zooplankton. Funct. Ecol. 3: 21- 27.

McLAREN, I. A. 1963. Effect of temperature on growth of zooplankton and the adaptive value of vertical migration. J. Fish. Res. Bd. Can. 20: 685-727.

NEILL, W. E. 1990. Induced vertical migration in copepods as a defence against invertebrate predation. Nature 345: 524-526.

OCEANoGRAPHIC LABORATORY, EDIBURGH. 1973. Con- tinuous plankton records: A plankton atlas ofthe North Atlantic and the North Sea. Bull. Mar. Ecol. 7:1-174.

RrNGELBERG, J. 1980. Aspects of the red pigmentation in zooplankton, especially copepods. Am. Soc. Lim- nol. Oceanogr. Spec. Symp. 3: 91-97. New England.

ROE, H. S. J. 1972. The vertical distributions and diurnal migrations of calanoid copepods collected on the Sond cruise, 1965. 1. The total population and general discussion. J. Mar. Biol. Assoc. U.K. 52: 277-314.

ROSE, M. 1933. Copepodes pelagiques. Faune France 26: 1-374.

SARS, G. 0. 1903. An account of the Crustacea of Nor- way. V. 4. Copepoda Calanoida. Bergen Museum.

S1NGA1AJAH, K. V. 1969. Escape reactions of zooplankton: The avoidance of a pursuing siphon tube. J. Exp. Mar. Biol. Ecol. 3: 171-178.




VERVOORT, W. 1963. Pelagic Copepoda. Part 1. Copep- oda Calanoida of the families Calanidae up to and including Euchaetidae. Atl. Rep. 7: 77-194.

1965. Pelagic Copepoda. Part 2. Copepoda Cal- anoida of the families Phaennidae up to and including Acartiidae, containing the description of a new species of Aetideidae. Atl. Rep. 8: 9-216.

ZARET, T. M., ANDW. C. KERFOOT. 1975. Fish predation on Bosmina longirostris: Body-size selection versus visibility selection. Ecology 56: 232-237.

, AND J. S. SuPPERN. 1976. Vertical migration in zooplankton as a predator avoidance mechanism. Limnol. Oceanogr. 21: 804-813.

Submitted: 20 January 1994 Accepted: 26 May 1994 Amended: 28 June 1994



Documents

Interspecific Differences in the Diel Vertical Migration of Marine Copepods: The Implications of Size, Color, and Morphology