tonic crustaceans by Cuvier in 1817 (Cuvier, 1834), hy- potheses explaining the phenomenon ...... and Metridia) off the south coast of Ireland. Proc. R. Ir. Acad.
MARINE BIOLOGY
Marine Biology 71, 23-31 (1982)
9 Springer-Verlag 1982
Effects of Age and Food Availability on Diel Vertical Migration of Calanus pacificus M. Huntley and E. R. Brooks Institute of Marine Resources, A-018, University of California, San Diego; La Jolla, California 92093, USA
Abstract
Age-specific differences in diel vertical migration behavior of Calanus pacificus were investigated in a 58 d (30 April26June, 1981) experiment in the Scripps Institution of Oceanography Deep Tank, La Jolla, California, USA; the experiment spanned three successive generations of copepods. The onset of vertical migration behavior occurred in the first feeding stage, Nauplius III. The amplitude of vertical migration gradually increased with age, becoming maximal in the late copepodite stages. Night depths remained constant with age while daytime depths increased. The migratory behavior of late copepodite stages was influenced by food availability. When phytoplankton was abundant and individual ingestion rates were high, copepodites performed high-amplitude migrations. As food availability declined, however, and the competition for food increased, migration amplitudes decreased and then ceased altogether so that copepodites remained in the relatively food-rich surface waters at all times. We suggest that hunger is the primary factor controlling vertical migration behavior.
Introduction
Since diel vertical migration was first recorded in planktonic crustaceans by Cuvier in 1817 (Cuvier, 1834), hypotheses explaining the phenomenon have included control by light (Clarke, 1934; Russell, 1934), temperature (Esterly, 1912; Russell, 1927), gravitation (Esterly, 1919) and pressure (Hardy and Paton, 1947; Rice, 1962). Endogenous rhythms (Harris, 1963; Enright and Hamner, 1967), avoidance of predation (Manteufel, 1959; Zaret and Suffern, 1976) and the requirement for feeding on phytoplankton (Weismann, I877; Hardy, 1956) have also been suggested as possible factors. One of the least understood aspects of vertical migration behavior is how it varies through the life-cycle of a species. The migratory behavior of Calanus finmarchicus
has probably been studied more often than that of any other zooplankton species, but usually only the late copepodite stages have been examined (Russell, 1928; Clarke, 1933, 1934; Gardiner, 1933; Farran, 1947). Where the vertical distribution of nauplii has been studied (Nicholls, 1933), no migration behavior has been noted. The overall conclusion drawn from these studies indicates that "the nauplii and eggs are found mainly near the surface" (Marshall and Orr, 1955) and that they, and the early copepodite stages (CI and CII), are non-migratory (Nicholls, 1933). The late copepodites do migrate and, of these, the CV and CVI stages have the most extensive migrations (Gardiner, 1933; Clarke, 1934). We sought to examine in more detail how vertical migration behavior varies through the life cycle of another species of Calanus. Our study of vertical migration used Calanus pacificus, whose known migratory behavior (Esterly, 1911; Enright and Honegger, 1977) is similar to that of C.finmarchicus. The Deep Tank at Scripps Institution of Oceanography, La Jolla, California (Balch et al., 1978) was used for our investigation for several reasons. First, we were able to sample at relatively small depth intervals. This is important since, if nauplii do migrate, they must do so over smaller depth intervals than have ordinarily been sampled (> 5 m). Second, turbulence, which could obscure the observation of small-amplitude migrations, is greatly reduced in the Deep Tank relative to the open ocean. The coefficient of eddy diffusivity in CEPEX (controlled environmental pollution experiment) bags, which are like the Deep Tank in that they are cylindrical enclosures of similar dimensions, was estimated to be about 2 orders of magnitude lower than in surrounding natural waters (Steele etal., 1977). Finally, the Deep Tank provided control over some factors which might otherwise affect vertical migration, since we could (1) create a permanent thermocIine, (2) establish phytoplankton-rich and phytoplankton-poor strata, and (3) exclude predators. We present here the results of a 58 d study of the vertical migration behavior of all developmental stages in 0025-3162/82/0071/0023/$ 01.80
24 3 successive cohorts of Calanus pacificus. Maintenance of constant food concentration was beyond our control, and our experiment thus produced some unanticipated results on the relationship between vertical migration and feeding.
Materials and Methods
Experimental Conditions Observations of vertical migration behavior were made on a reproducing population of Calanuspacificus held captive in the Scripps Institution of Oceanography Deep Tank. This experimental system, described in detail by Balch et al. (1978), is a 70 m% 10 m deep steel cylinder that is situated outdoors and thus exposed to an ambient cycle of light intensity. The tank has an external coating of polyurethane insulating material and an internal coating of black, non-toxic epoxy. Seawater for the system is pumped from a depth of several meters at the end of Scripps Pier. The Deep Tank was prepared for the experiment by mechanically scrubbing the inside walls with a dilute solution of sodium hypochlorite and then rinsing with freshwater. After repeating this procedure the tank was filled with seawater pumped through a diatomaceous earth filter (nominal pore size 5 Hm). A single aluminum cooling coil, connected to a cooling system using fresh water as a coolant, was placed around the internal circumference of the tank at a depth of 4 m. This arrangement permitted us to maintain a convectively-cooled layer below a surface thermocline, although not providing absolute control of the temperature. In two previous attempts to begin the experiment with exposure to full sunlight, phytoplankton production so exceeded zooplankton grazing pressure that the phytoplankton decomposed, creating anoxic conditions lethal to Calanus pacificus. We controlled phytoplankton production by covering the tank with neutral density screening, which reduced the surface light intensity to 15% ambient. An additional covering, a 100 m 2 sheet of clear plastic, was fixed at a height of several meters above the surface in order to prevent precipitation from entering the tank. The experiment commenced on 26 April and ended on 26 June 1981. We began by sorting ~ 3 400 live females of Calanus pacificus from oblique plankton hauls taken 2 km offshore in La Jolla Bay. Within 4 h after capture, sorted females were placed in 505 Hm Nitex mesh bags, immersed in three 200-liter polypropylene tubs containing Whatman GF/C-filtered seawater to which enough logphase Thalassiosira weissflogii culture had been added to turn the water visibly green. The females were permitted to lay eggs for a period of 3 d; the eggs (about 200/~m) fell through the Nitex mesh bags and settled on the tub bottoms. On 29 April, the Nitex mesh bags containing female C. pacificus were removed from the polypropylene tubs, leaving phytoplankton, eggs and early stage nauplii.
M. Huntley and E. R. Brooks: Vertical Migration of Calanuspacificus The contents of the tubs were then poured into the Deep Tank. To ensure continued availability of phytoplankton food to the developing nauplii we began, on 1 May, to add cells from 200-liter semi-continuous monocultures of T. weissflogii, Isochrysis galbana and Ditylum brightwelli; these cultures had approximate concentrations of 10 7, 105 and 104 cells ml-% respectively. The cultures were delivered at a rate of 1 liter h -1 into the Deep Tank at a depth of 0.3 m using peristaltic pumps; phytoplankton were added continuously throughout the experiment. The cultures were grown immediately adjacent to the tank, in IMR medium (Eppley et al., 1967) at ambient temperature and light intensity. Sampling A centrifugal pumping system simultaneously sampled the Deep Tank at depths of 0.2, 3, 5, 7 and 9 m (Fig. 1). The
SAMPLING SYSTEM
~ - - ~ - - \
(a)
PU
/
SAMPLING SYSTEM ,A
.------~A.-E
(b) DEEP TANK,,.
B
/ CEMENT PLATFORM
B(Z////~/////////////H/////////,//,;: Fig. 1. Deep Tank. (a) Top view of sampling system, which consists of 5 independently-operated sampling units, each with its own centrifugal pump; sampling pipes (3.81 cm i.d. PVC) lead from center of tank to collecting tubs; depth of sample intake (m) is shown for each tub. (b) Side view of one unit of sampling system; prior to taking sample, standing water in intake pipe was drained through valve (B); centrifugal pump (C) was then turned on, sucking sample through intake (A) and delivering it to 200-liter collecting tub (D); after excess sample drained through the overflow (E), the collecting tub contained 200 liters; a 5-liter sample for particulate analysis was removed through sampling valve (I); zooplankton sample was taken by opening valve (F) and allowing water to drain through a 103 Hm mesh net (G); temperature was measured as water drained through net and into outflow pipe (H)
M. Huntley and E. R. Brooks: Vertical Migration of Calanuspacificus 0.5 horsepower, 3 450 rpm, centrifugal pumps had semiopen, cast-iron impellers (Teel | capable of pumping about 200liters rain -1 through 3.81 cm inner diameter PVC sampling pipes. Tests with Calanuspacificus and live, mixed zooplankton showed that these crustaceans survived intact after passage through the pump. Two hundred liters were pumped from each of the 5 depths at 36 h intervals from 30 April to 26 June. Samples were collected either at mid-morning (10.00hrs Pacific Standard Time, PST) or mid-evening (22.00 hrs PST) by simultaneously filling the five 200-liter sampling tubs (Fig. 1). A 5-liter aliquot was removed via a side drain to provide samples for analysis of chlorophyll and phaeopigments. The remaining 195 liters were emptied through a 103#m mesh Nitex filter fixed by a quick-disconnect fitting to a drain in the tub bottom. Temperature of the filtrate was measured with a thermometer. The content of each filter was rinsed into a sample jar with GF/C-flltered seawater and preserved with 5 to 10% formalin. The entire sample collection process required 15 min. The 1 m 3 of seawater removed from the Deep Tank by sampling was replaced with seawater filtered through cartridges of 5 and 1 #m nominal pore size (Culligan| Chlorophyll and phaeopigment samples for each depth were prepared in triplicate by filtering 100 ml aliquots through G F / C filters; samples were analyzed by fluorimetry (Strickland and Parsons, 1972). The preserved zooplankton samples were analyzed for the abundance of each developmental stage from egg to adult; first and second stage nauplii (NI and NII) were grouped together as a single stage because of the difficulty in differentiating the two stages. Entire samples were counted except when the abundance of a given stage exceeded 500 sample-l; then only 10% of the sample was counted. The biomass of copepods (g C m 3) was calculated from measurements of carbon, nitrogen and dry weight made on preserved individuals from samples throughout the experiment; we based our correction for material lost in preservation on a comparison between Stage III, IV, V and VI copepodites frozen live and those preserved from 10 samples taken in mid-June (Huntley and Brooks, unpublished data).
25 creasing temperature (Fig. 3). As noted earlier, we were able to maintain a deep, isothermal layer below a surface thermocline throughout the experiment. However, we were unable to prevent the gradual warming caused by ambient weather conditions. Temperature at the surface (0.2 m) increased from 15 ~ at the beginning of the experiment to 21.7~ at the end. Below the thermocline (> 5 m) the temperature increased from 11.4 ~ to 17.3 ~ during the same period. The temperature at the surface fluctuated greatly as it increased; increases at other depths were more regular and uniform with time. Phytoplankton biomass was consistently greater in surface waters to a depth of about 3 m, as indicated by the isopleths of chlorophyll-a shown in Fig. 4. Chlorophyll values of > 1 #g 1-1 generally occurred only within the surface thermocline. However, two phytoplankton blooms (> 10 #g 1-1) occurred - one at the beginning of the ex-
GENERATION 2
~[_ 30'0 I OE25'0 I
F 0 15.0 .,-,, nO I0.0
I.d n O C) 5.0 _J 0
15
50
45
TIME
60
(d )
Fig. 2. Calanuspacificus.Total biomass ofcopepods during experiment, showing times at which the 3 successive cohorts occurred
._./0.2
22f
m
zol-
._/ / s m
Results
Three successive generations (egg-to-egg) of Calanus pacificus were produced during the experiment (Fig. 2). The first, second and third generations lasted, respectively, about 26, 20 and 16 d. The maximum biomass of each generation was, respectively, 2.2, 30.1 and 5.1 g C m-L The biomass peak of each generation occurred when the population was dominated by late-stage nauplii and earlystage copepodites. The first generation was a distinct cohort and was dominated by only 4 developmental stages at any given time. However, "smearing" of the cohort became more pronounced with each new generation, resulting in a progressively wider age distribution. The decreasing generation time in successive generations appears to have been primarily a function of in-
~1~ :::) 18 "
.
: ,/../x/'~" . _ _ ,,..,_
,\
o.hm
N ,4 -/"
0
I
I
]
I
I
I
I0
20
50
40
50
60
TIME
(d)
Fig. 3. Temperature (~ at all sampling depths during course of experiment. Temperatures were highest at 0.2 and 3 m; at depths > 5 m water was isothermal. Although temperatures at all depths increased gradually with time, the absolute temperature gradient with depth remained constant
26
M. Huntley and E.R. Brooks: Vertical Migration of Calanus pacificus oI
E -rl--'-
(3-
"' 6 i
5-5/xg 12 i01
o
i
I
I
I
15
50
45
6O
Fig. 4. Isopleths of chlorophyll a ~ug 1-1) during experiment. With exception of two blooms, one at beginning and one at end of the experiment, water below 3 m was low in chlorophyll a. Most of plant biomass occurred in surface water
TIME (d) stage in the depth layer H,. We then pooled, for each stage, the day values and the night values of Zcm for the entire experiment. The mean day and night depths of the center of mass (__ SE) for each developmental stage from egg to adult are shown in Fig. 5. The mean day and night Zcm's of eggs and Naupliar Stages I and II (NI, NII) were deep (7 m), and provide no evidence for vertical migration. However, the Zcm'S of NIII, which is the first-feeding stage of Calanus pacificus (Marshall and Orr, 1955), were significantly higher in the water column than those of the non-feeding stages. The night depth of NIII (3.84 m) was somewhat higher than its day depth (4.28 m), but the difference is not significant. As development proceeded through the next two naupliar stages (NIV and NV) the copepods moved even higher in the water column, during both the day and the night. From the fifth naupliar stage (NV) to maturity (CVI) the night depths of all stages were approximately the same (2 to 3 m). However, the day depths became progressively deeper with age; thus, the amplitude of the vertical migration increased with age. Apparent amplitudes of vertical migration for all developmental stages are given in Table 1. It is clear from the foregoing, for all feeding stages, that the mean distributions at night were concentrated in the
periment and one near the end - after which the phytoplankton sank, giving rise to higher chlorophyll values (1 to 3 #g 1-1) below the thermocline. The first bloom took place during the first generation of Calanus pacificus. It lasted approximately 7 d and reached a surface concentration of 13.4#g1-1 chlorophyll-a on Day6. The second bloom, again lasting for about 7 d, occurred between the second and third copepod generations. A surface concentration of 18.1 #g 1-1 chlorophyll-a was reached on Day 44. Chlorophyll values were lowest during the second copepod generation, when surface values fell to < 5/~g 1-1 and values below the thermocline were consistently < 1 pg 1-1. Our interpretation of vertical migration behavior is based on differences in the day/night distributions of Calanus pacificus. To determine the depth at which a given developmental stage was most concentrated on a given day, we calculated the depth of its center of mass (Zcm) according to the equation given by Vinogradov (1970, p. 36):
Zcm H/2"at + (H~ + H2/2)'a2 + . . . + (H1 + H2 +... + Hn 1-1-Hn/2)" an. al + a 2 + . . . + a n
where a n is the abundance (No. m -3) of the developmental
0
i
1
I
i
I
I
I
I
I
]
I
1.0
N
.~
5
2.0
4.0
6.0 o DAY (lO00h) 9 NIGHT (2200 h)
i0.01
1
I
I
EGG NI a NIl NI]I
I
I
I
I
I
I
I
I
I
N]E
NV
NVF
CT
C]I
CTFF
C'~"
C~"
C'gT
DEVELOPMENTAL STAGE
Fig. 5. Calanus pacificus. Depths of center of mass, Zcm, for each developmental stage of copepods during day and night. Values shown are overall means for entire experiment. Error bars are standard errors. Time, hrs, is Pacific Standard Time
M. Huntley and E. R. Brooks: Vertical Migration of Calanuspaeificus
Calanuspacificus. Mean day and night depths (m) of center of mass, and amplitudes of vertical migration, for each developmental stage Table 1.
Developmental stage NI + NII NIII NIV NV NVI CI CII CIII CIV CV CVt
Day depth
Night depth
Migration amplitude
(m)
(m)
(m)
7.37 4.28 3.30 2.68 2.97 3.68 4.49 4.72 5.00 4.84 5.37
7.31 3.84 2.95 2.36 2.33 2.93 2.78 2.46 2.37 2.07 2.93
0.06 0.44 0.35 0.32 0.64 0.75 1.71 2.26 2.63 2.77 2.44
relatively phytoplankton-rich surface waters, whereas the distributions by day tended to concentrate in phytoplankton-poor waters. This certainly appears to confirm the view that vertical migration is related to food availability and feeding activity (Hardy, 1956; Boyd et al., 1981). The foregoing results led us to ask how, if at all, changes in food availability in surface waters affect the vertical migration behavior, We pooled the values of Zcm for Copepodite Stages CII-CV, inclusive, on every sampling occasion. These developmental stages were deemed best for such an analysis because (a) their depth distribu-
27 tions on any given sampling day were not significantly different; (b) they have the greatest migration amplitude (Table 1); and (c) their overall mean depth distributions, day or night, are not significantly different (Fig. 5). We did not include CVI adults in this pooling procedure because, although their migration amplitude was approximately the same as that of the late copepodite stages, they tended to occur deeper in the water column during both day and night (Fig. 5). The mean day and night depths of the center of mass of the CII-CV copepodite stages (_+ SD) are shown in Fig. 6 as a function of time elapsed during the experiment. Most striking is the observation that, while the night distribution remained constant at 1.5 to 3.5 m, the day distribution changed significantly during the course of the experiment. During the first generation of Calanuspacificus, copepodites were concentrated at 6 to 7 m during the day and their migration amplitude was great. However, during the second generation, beginning at Day 30, the day depth began to decrease steadily. On Day 39 the day depth was at 2.5 m, not significantly different than the constant night depth. On this day, and on the 3 to 6 d immediately preceding or following, vertical migration behavior apparently ceased, During the third generation, the day distributions became progressively deeper and vertical migration resumed. The dispersion of the population about the depth of its center of mass is exemplified by the depth distribution of the CIII copepodite stage shown in Fig. 7, where we have
0
N
&
o~
