field data demonstrate the link between consumption rate of ... · web viewmunk p, larsson po,...
TRANSCRIPT
Does copepod size determine food consumption of particulate feeding fish?
Mikael van Deurs1, Marja Koski1 and Anna Rindorf1
1DTU Aqua National Institute of Aquatic Resources, Technical University of Denmark (DTU), Jægersborg Alle 1,
Charlottenlund Castle, 2920 Charlottenlund, Denmark.
ABSTRACT: Climate-induced reduction in mean copepod size, mainly driven by a decrease in
abundance of the large Calanus finmarchicus around 1987, has been linked to low survival of fish
larvae in the North Sea. However, to what extent this sort of reduction in copepod size has any
influence on adult particulate feeding fish is unknown. In the present study we investigated the
hypothesis that availability of the large copepods determines food consumption and growth
conditions of Lesser sandeel (Ammodytes marinus) in the North Sea. Analysis of stomach content
suggested that food consumption is higher for fish feeding on large copepods, and additional
calculations revealed how handling time limitation may provide part of the explanation for this
relationship. Comparing stomach data and zooplankton samples indicated that Lesser sandeel
actively target large copepods when these are available. Finally, we observed that the length of
Lesser sandeel began to decrease in the late 1980s, simultaneously with the C. finmarchicus decline.
Key words: predator-prey interactions; optimal foraging; prey selection; North Sea regime-shift;
climate change; sandeel; sand lance; forage fish; Calanus finmarchicus; handling time limitation
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
INTRODUCTION
Copepods play a key role as prey for pelagic fish larvae and zooplanktivorous fish. Hence, as a
changing climate modifies copepod communities (Planque and Taylor 1998; Möllmann et al. 2003;
Beaugrand 2004; Helaouët and Beaugrand 2009; Kjellerup et al. 2012), cascading effects through
the foodweb are likely to affect fish populations (Beaugrand et al. 2003; Möllmann et al. 2003;
Frederiksen et al. 2006; van Deurs et al. 2009).
Long-term copepod abundances in the North Sea show a decreasing trend for relatively large
Calanus finmarchicus in spring and an increasing one for the equally large C. helgolandicus in fall
(Planque & Fromentin 1996; Planque et al. 1997). The spring decline in abundance of C.
finmarchicus has been linked to low survival of cod larvae (Gadus morhua) and sandeel larvae
(Ammodytes marinus) (Beaugrand et al. 2003; van Deurs et al. 2009). Besides cod and sandeel
larvae, also juvenile and adult zooplanktivorous fishes in the North Sea feed on the calanoid
copepods (De Silva and Balbontin 1974; Last 1987; Maes et al. 2005; Raab et al. 2012), and the
literature report northward shifts in the feeding grounds of North Sea herring in response to changes
in the timing of C. finmarchicus (Corten 2000).
While C. finmarchicus has declined in abundance, overall annual production of calanoid copepods
is seemingly unchanged (Beaugrand et al. 2002) and the spring reduction in C. finmarchicus may
therefore to some extent have been compensated for by an increase in C. helgolandicus and smaller
calanoids. Although little attention has been dedicated to the long-term changes in abundance and
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
biomass of smaller calanoids, such as genus Temora, Pseudocalanus, Centropages, Paracalanus,
Oithona, Microsetella and Oncaea, these are all abundant in the North Sea (Nielsen et al. 1993).
Similarly to C. finmarchicus, some of these species also have their reproduction peak in spring
(Kiørboe & Nielsen 1994; Broekhuizen and McKenzie 1995).
The weights of calanoids in the North Sea are around four to six orders of magnitude smaller than
the weight of zooplanktivorous fishes. Consequently, for a particulate feeding fish that snatches one
prey at the time, handling time (here defined as the average time it takes to prepare attack, capture
and swallow each prey) may represent a potential bottleneck on food consumption when feeding at
high prey concentrations (Jeschke et al. 2002). However, since Calanus are as much as ten to
twenty times larger than the small calanoids mentioned above, the strength of handling time
limitation may be dependent on the species and size composition of copepods within in the prey
field.
Lesser sandeel is a particulate feeding zooplanktivorous fish throughout its life. They are abundant
in the Dogger area (central North Sea) on the southerly distribution limit of the large C.
finmarchicus, which suggests a particular susceptibility to fluctuations in the spring invasion of this
copepod species. The highly resident nature of sandeel renders them unable to move between areas
(Jensen et al. 2011) and hence they are likely to be more sensitive to changes in copepod abundance
than, for example, Atlantic herring that is known to redistribute according to food density
(Dragesund et al. 1997; Misund et al. 1998; Corten 2001). Furthermore, adult Lesser sandeel feed
mainly during spring (Macer 1966; Winslade 1974; MacLeod et al. 2007). From the point of view
of the sandeel, increasing C. helgolandicus concentrations in late summer are thus not likely to
compensate for a spring-time reductions in C. finmarchicus.
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
In the present study we examined the hypothesis that climate-driven reductions in copepod size
affect the growth of adult particulate feeding fish. If this hypothesis is valid, we would expect that
(i) fish on a diet of small copepods have diminished food consumption relative to fish on a diet of
large copepods, (ii) fish actively select for large copepods when these are available, and (iii) size at
age for particulate feeding fish in the North Sea have decreased simultaneously with the spring time
reduction in C. finmarchicus in the late 1980s. We investigated these assumptions by (1) using
stomach content of Lesser sandeel from the North Sea to analyze the relationship between prey size
and daily food consumption, (2) carrying out theoretical calculations to address the role of handling
time limitation, (3) comparing mesozooplankton data to the stomach content to shed light on prey
size preference and (4) investigating long-term trends in length at age for Lesser sandeel before and
after the late 1980s.
MATERIALS AND METHODS
Stomach analysis
Samples were collected from commercial fishing grounds in the Dogger area in 2006, 2007, 2009
and 2010 by fishermen. A total of 30 samples (containing between 16 and 86 sandeels each) were
used (Fig. 1). Each sample was taken by fishermen on commercial sandeel fishing vessels, frozen
on board the fishing vessel, and labeled with catch date, time of capture, and position. This enabled
sufficient spread in time and space and thereby a wide range of stomach contents. Fish from all 30
samples were measured (weight and length) and the stomachs were dissected, dapped on paper
towel, and weighed to the nearest mg. After weighing the stomachs, some were preserved in
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
alcohol. An overview of samples and number of alcohol preserved stomachs is presented in Table 1.
In order to get the wet weight of the stomach content, the weight of the empty stomach must be
subtracted from the total weight of the stomach. Therefore, 30 stomachs covering fish lengths (total
length) from 8 to 18 cm were emptied and the relationship between the wet weight of empty
stomachs (WES) [g] and fish length (FL) [cm] was established. A power law model provided the best
fit to data: WES = 9 × 10-5 × FL2.30 (R2 0.87).
The alcohol preserved stomach contents were spread evenly on Petri-dish with added water. A sub-
area of 4 cm2 in the middle of the petri-dish was photographed using Leica MZ12 stereo-microscope
with a 6.5 × magnification and a Leica DFC290 camera Image Pro Plus software was used to
determine the size of copepods by digitally measuring the length of all intact copepod prosomes
[mm ± 0.1]. Countable biomass constituted on average 45% of the stomach content, and of the
countable biomass 20% was available for counting in the 4 cm2 counting frame. A total of 4368
copepods were measured. Mean copepod size was calculated for each of the 30 samples giving each
stomach the same weight, independent of the number of measured prosomes. Due to the state of
degradation (e.g. lost appendixes and antennas) copepods could not be quantified per species.
In order to validate whether a 4 cm2 subarea provides a representative impression of the entire
stomach content, we spread out the content from 4 stomachs onto four Petri-dishes and divided each
dish into five zones of even size. Thereafter we measured all measurable prosomes within each of
the five zones and calculated the coefficients of variance. The average coefficient of variance was
0.056 (min = 0.032, max = 0.074).
The effect of copepod size on food consumption
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
Stomach analysis revealed that the variation in the size of ingested copepods was smaller within
samples than between samples. A bimodal distribution separated samples into two groups: those
with mean prosome length significantly above 1.3 mm and those with mean prosome length
significantly below 1.1 mm (see results). Based on this finding two feeding groups were defined:
sandeels from samples with mean copepod size <1.3 mm (group-1) and sandeels from samples with
mean copepod size >1.3 mm (group-2). We chose 1.3 mm over 1.1 mm as the threshold, since
among the common copepods in the North Sea only Calanus exceed 1.3 mm (Pitois et al. 2009)
plus it represented the local minimum when a density function was fitted to the bimodal
distribution.The effect of copepod size on consumption rate was subsequently investigated by
testing whether the mean consumption rate of sandeels belonging to group-1 differed significantly
from that of group-2. How this was tested is described below.
Assuming that gastric evacuation is an exponential process, we approximated the weight specific
consumption rate (C [g (g fish) -1 d-1]) for each fish using the following equation (e.g. Eggers 1977):
(1)
where S is the weight of the stomach content [g], W is the wet weight of the fish [g] and α(T) is the
gastric evacuation coefficient as a function of temperature (derived from van Deurs at al. 2010):
(2)
Modeled sea bottom temperature for each sample was provided by the Danish Meteorological
Institute based on predictions from the DMI-BSHcmod model (http://baltic.mersea.eu.org/). The
DMI-BSHcmod model is a 3 dimensional finite difference hydrodynamic model set up in spherical
coordinates. Calculations are done on an Arakawa C grid (Kleine 1994).
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
In order to estimate the effect of copepod size it was necessary also to account for other potential
predictors (i.e. fish size, year, season, latitude, longitude) and a random effect of sample. This was
achieved by estimating the parameters in a linear mixed effects model on the following form:
(3)
where and , , , , , and are parameters to be estimated, Lon is longitude, Lat
is latitude and indices CS, FS, y, s and sample denote copepod size group, fish size group, year,
season and sample, respectively. Copepod size was implemented as a class variable with two
discrete levels describing whether a fish belonged to group-1 or group-2. Fish size also had two
levels; sandeels <12 cm and sandeels >12 cm. The interaction effect between copepod size and fish
length was tested in order to determine whether the effect of copepod size depends on fish size.
Year and season had four levels (2006, 2007, 2009 and 2010) and two levels (April/May and June)
respectively. The p-values associated with parameter estimates for copepod size, fish size and the
interaction between copepod size and fish size were derived by comparing the full model with a
reduced model (likelihood ratio test).
Handling time limitation
To investigate if food consumption is potentially handling time limited, we calculated the average
time period available to handle (preparing attack, capture and swallow) each prey given a
hypothetical situation where search time is zero, all attacks are successful, and no time is spent on
other activities such as escaping predators. We refer to this time period as the maximum allowed
handling time per prey (MAH [s prey-1]). The assumption is that if MAH is sufficiently small,
handling time limitation is likely to affect food consumption. Mean MAH was calculated for each
fish as:
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
(4)
where C is derived from the equation (3), W is the wet weight of the fish [g], Pm is the copepod
individual wet weight [g], and τ is the daily foraging period [s]. Sandeels are visual feeders and
spend the night buried in the sand. The day length in May in the North Sea is roughly 15 hours.
However, as indicated by van Deurs et al. (2011) it takes at least one hour for the school to form in
the morning and another hour for the school to disintegrate after completing foraging. The τ-
parameter was therefore set to 46800 s (= 13 h). Calculations were made for each of the following
three scenarios of daily ration and copepod size: (A) Daily ration as observed in group-1 and
prosome length = 0.85 mm (equal to 0.000035 g), (B) Daily ration as observed in group-2 and
prosome length = 0.85 mm, and (C) Daily ration as observed in group-2 and prosome length = 1.75
mm (equal to 0.00044 g). Pm was derived from the length-weight relationships for Calanus and
Acartia; representing large and small copepods respectively (Uye 1982). For the dry weight to wet
weight conversion we applied a ratio of 5 adopted from Yamaguchi and Ikeda (2000).
Zooplankton samples
The inclusion of zooplankton samples was a post hoc decision taken based on the outcome of the
stomach analysis, showing that fish caught on a given position and date contained either small
copepods or large copepods, but never a mixture. Assuming that active selection for small copepods
is not an option; the simplest pair of mutually exclusive explanations is: (a) locations where
Calanus is common has relatively low concentrations of smaller calanoids (inverse correlation
between Calanus and smaller calanoids), or (b) sandeels avoid small calanoids when Calanus is
available (positive selection for larger copepods). The purpose of including zooplankton data was to
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
explicitly test explanation a. Since fish samples were collected onboard commercial fishing vessels,
parallel zooplankton samples could not be obtained. Instead we used a large mesozooplankton data-
set collected during the LIFECO program from late March to May in 2001 (Arendt et al. 2005) to
analyze the general covariation between small copepods (Acartia, Temora, Centropages,
Pseudocalanus, Oithona, Microcalanus, Oncaea, and Microsetella) and large copepods (Calanus).
Although no sampling stations were directly overlapping the stomach samples, all 62 samples used
overlapped with the continuum of sandeel habitat extending north-east from Dogger along the
southern boundary of the C. finmarchicus distribution (Fig. 1). Hence, they were found suitable for
testing the very general aspect of explanation a. Samples were typically taken from 50-70 m depth
to the surface (if the station was shallower, ca. 10 m above the bottom), using a submersible pump
connected to a 30 µm plankton net. The pump was towed vertically from 50-70 m to the surface
with an average rate of 15 cm s-1 and with a sampling volume of 1.3 m3 min-1. The samples were
preserved in 4% borax buffered formalin and counted from 1/4 to 1/64 fraction of the total sample
to achieve a total of >200 individuals for the dominant species (Jónasdóttir & Koski 2011). The
copepods were identified to the genus or species level, and the life-stages were grouped to nauplii,
copepodites (I-V), females and males, as well as placed into size categories at 50 µm intervals. The
average prosome length of each species was estimated for 5 stations per cruise combining the
copepodites and adults (not nauplii) by calculating the weighted mean length of each species. The
average length was converted to wet weight using length-weight conversions from the literature
(Cohen and Lough 1981; Breteler et al. 1982; Uye 1982) and a wet weight/dry weight ratio of 5 to
convert dry weight to wet weight (Yamaguchi and Ikeda 2000). In most cases length-weight
conversions at the genus level could be found in the literature. Exceptions were Microcalanus and
Oncaea, where the length-weight conversion for Oithona was applied. Lastly, we multiplied
weights and numbers and thereby derived the wet biomass [g m-2] of small and large copepods for
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
each sampling station. The spearman correlation was applied to test if the concentration of Calanus
was related to the concentration of small copepod species or vice versa
Trends in length at age of sandeels over time
Lesser sandeel time series of mean total length at age 1 and age 2 were analyzed to investigate if
periods of changing growth rates could be detected. In particular, we investigated whether the
growth rate of Lesser sandeel in the Dogger area slowed down around 1987, the year of the reported
change in the North Sea copepod community (Beaugrand et al. 2004). The time-series represent
length at age in June (1976-2011) from commercial samples.
A linear mixed effect model was applied to test if the period after 1987 differed from the period
prior to 1987 in terms of length at age:
where and and are constants within indices to be estimated in the model. With this
formulation, slope and intercept of a linear trend line are estimated for each combination of age and
period. The linear mixed effect model allowed us to account for correlation between length at age 1
and age 2 in a given year through the addition of a random effect . Finally, the effect of Period
was tested by comparing the presented model with a version without Period. Length at age trends
were held up against development in total stock biomass and fishing mortality going back to 1983
(ICES 2012), to allow us to discuss potential alternative explanations, such as density dependence
and fishing induced evolution of growth rates (Olsen et al. 2004).
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
RESULTS
Stomach analysis: Variation in the size of ingested copepods was much smaller within samples
than between samples. When samples was aligned and sorted according to mean prosome length a
bimodal distribution was revealed, that separated samples into two groups, those with mean
prosome length significantly above 1.3 mm and those with mean prosome length significantly
below 1.1 mm (Fig. 2, the use of “significantly” was based on the standard errors associated with
each sample). Only one sample (id 30) fell in between 1.1 and 1.3 mm. Each sample was
represented by only a subset of stomachs (Table 1). However, mean prosome length between
stomachs from the same sample varied little: For 88% of all stomachs analyzed, it applied that the
mean prosome length of copepods in a stomach was on the same side of 1.3 mm as the combined
mean prosome length of all stomachs from the sample to which this stomach belonged.
Furthermore, the coefficient of variance of prosome lengths within a given stomach was on average
0.22. Together these findings indicate that sandeels from the same sample had been foraging on the
same narrow size range of copepods, either small (<1.3 mm) or large (>1.3 mm) copepods. It
should, however, be noted that a statistically significant relationship existed between fish size and
prey size at the level of the sample (p < 0.01, mixed effect model with random sample effects). The
estimated coefficient of this relationship was 0.027, corresponding to an increase in prey size of
0.027 mm for each 1 cm increase of the fish.
The effect of copepod size on food consumption: Sandeels feeding on large
copepods (group-2) displayed on average 130% (for small sandeels) and 100% (for large sandeels)
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
higher stomach content weight compared to sandeels feeding on smaller copepods (group-1). This
pattern was unchanged after converting stomach content to weight specific food consumption (Fig.
3). Weight specific food consumption of group-1 sandeel was significantly lower than that of
group-2 sandeel (p = 0.001; Gaussian criteria of the model was satisfied and residuals were without
trend). This pattern was consistent among sampling years with an average consumption rate of
~0.010 (g g-1 fish) for group 1 and ~0.022 (g g-1 fish) for group-2 (Table 2). Further, weight-specific
consumption of large sandeels was significantly lower than the weight-specific consumption of
small sandeels (p = 0.04), but the interaction-effect between copepod size and fish length was
insignificant. In addition, the food consumption was influenced both by longitude and latitude (p =
0.01 and p = 0.04, respectively), whereas the season and year had no effect (p > 0.05).
Handling time limitation: Mean maximum allowed handling time (MAH) was approx. 10 seconds
for large sandeels in scenario A and for both small and large sandeels in scenario B. For small
sandeels in scenario A MAH was approx. 30 seconds, whereas it was well above 100 seconds for
both small and large sandeels in scenario C (Fig. 4).
Zooplankton data: Zooplankton samples from 2001 revealed that Calanus concentration within
sandeel feeding grounds was positively correlated to the concentration of smaller copepods (Fig. 5),
indicating that sandeels experience a prey field containing both small and large copepods. On all
sampling stations smaller copepods exceeded Calanus in terms of numbers, whereas the biomasses
were in the same order of magnitude.
Length at age: Length at age time-series for age 1 and age 2 Lesser sandeel in the Dogger area
indicated a turning-point around 1987, with increasing length at age in the period before 1987 and a
gradual decrease in the period after 1987 (Fig. 6). The effect of Period was significant (p = 0.02)
and data satisfied model assumptions as indicated by inspection of residuals.
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
Total stock biomass displayed large fluctuations between 1983 and 2002, but with no clear temporal
trend or apparent turning point around 1987. In 2003 stock biomass dropped to the lowest level
experienced in the entire time series followed by an increasing trend. The long-term pattern in
fishing mortality was also seemingly disconnected from the overall patterns in length at age.
Fishing mortality exceeded 1 in the period 1999 to 2004, for the remaining of the time period (going
back to 1983) fishing mortality fluctuated around a stable level between 0.5 and 0.8.
DISCUSSION
Feeding on larger copepods doubled the food consumption in lesser sandeel in the Dogger area.
Further calculations indicated that in a scenario where only small copepods are available, the
individual sandeel have on average roughly 10 seconds (one order of magnitude lower than for
large copepods) available for handling each prey item to achieve a daily ration equal to that
observed for sandeels feeding on large copepods (group-2). The upper limit of realistic handling
times reported for similar sized zooplanktivorous fish is also 10 seconds (Hjelm and Persson 2001),
indicating that very little time is left in this scenario for prey searching, unsuccessful attacks, and
escaping predators (either by burying into the sand or by forming temporary denser schools). In
consequence this suggests that consumption rate is inclined to becoming handling time limited
when only small copepods are available.
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
According to the strict conceptual understanding of handling time limitation (Jeschke et al. 2002)
only predators that digest prey items faster than they are handled are likely to be handling time
limited. From the evacuation rate applied in the present study (around 5% per hour) (van Deurs et
al. 2010) and a handling time of maximum 10 seconds it follows that sandeels handle prey faster
than they are digested. Hence, sandeels may not be handling time limited in a strict sense. However,
the longer sandeels spend foraging away from their sand refuge, the higher the risk of predation.
This leads to the prediction that as prey size decreases (and the reward per unit handling time
decreases) the sandeel thus become increasingly inclined to choose between exposing itself to a
high level of predation risk or consent to retreat to their sand refuge with less in their stomach (the
energy-predation trade off, see van Deurs et al. (2010)).
The weight specific daily ration of large sandeels was significantly less than for small sandeels.
This is in line with our calculations of maximum allowed handling time, which indicated that large
sandeels are more likely (lower MAH) to be handling time limited than small sandeels. However,
the interaction-effect between fish size and prey size on weight specific consumption was not
significant, which would be reasonable to expect in a situation where handling time constraints is
intensifying with fish size.
C. finmarchicus is the only abundant zooplankton genus in the North Sea in spring that reaches
prosome lengths >1.3 mm (Pitois et al. 2009). We therefore suggest that sandeels with mean food
item prosome length of >1.3 mm (group-2) were feeding mainly on Calanus by the time they were
caught. The springtime mesozooplankton data revealed that availability of Calanus is positively
correlated to the availability of smaller copepods, yet stomachs from sandeels of all sizes, contained
either small or large copepods, but never a mix. The simplest explanation for this observation is that
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
sandeels deliberately ignore smaller copepods when larger copepods are sufficiently abundant,
which is also in line with traditional optimal foraging theory stating that an optimal forager rank
prey preference according to energy gain per handling time (Stephens and Krebs 1986; Hart and
Ison 1991). However, this conclusion requires that sandeels containing only small copepods were
experiencing a prey field nearly deprived of Calanus (we will return to this later in the discussion).
Acting to blur the picture, the stomach content of the sandeels may also reflect the copepod
assemblage in its horizontal or vertical position. The zooplankton samples used in the present study
did not originate from the same location or date as the sandeel catch, and were integrated over the
entire water column. The relationship between large and small copepods could thus potentially be
different between areas, seasons or depths. Copepods are known to be patchily distributed, occupy
different layers of the water column and have different vertical migration behavior (Daro 1988). For
instance, in some areas of the North Sea small copepods with limited reserves have been observed
to mainly reside in the proximity of the chlorophyll maximum layer (Daro 1988; Koski et al. 2011),
while C. finmarchicus migrates between different layers (Daro 1988) or resides deeper down in the
water column (Jónasdóttir and Koski 2011). Although typically one would expect both small and
large copepods to be abundant at similar productive areas, we cannot exclude the possibility that
sandeel stomachs reflect the prey availability at their exact location rather than selective feeding on
larger copepods. Nevertheless, the presence of exclusively large or small copepods in the stomachs
suggest some degree of selective feeding, and according to optimal foraging theory (i.e. Stephens
and Krebs 1986; Hart and Ison 1991) it would be unexpected if that selective feeding would mainly
focus on small prey.
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
Why did some stomachs contain only small copepod species? Access to large copepods may be
determined by the overlap between sandeel habitats and the distribution of C. finmarchicus. C.
finmarchicus protrudes into the North Sea from the north during spring. The distribution finds a
southern boundary at latitude 55 to 56oN (Planque and Fromentin 1996) and thereby partially
overlaps with the important sandeel areas on Dogger Bank (Fig. 1). However, while the exact
position of the front in that area is dynamic and change over time (Munk et al. 1999), sandeels are
bound to their habitats (Jensen et al. 2011). Consequently, variation in the degree of overlap could
be an important factor affecting access to larger copepods, and may act independent of the overall
climatic changes and the general C. finmarchicus distribution in the North East Atlantic.
The time-series analysis indicated changing growth rates for Lesser sandeel in the Dogger area
around 1987, where also C. finmarchicus began to decline (Beaugrand et al. 2004). However, this
change was not characterized by a shift from a high plateau to a low plateau, as reported by
Beaugrand et al. (2003) for the abundance of C. finmarchicus. Instead it conformed to a turning-
point with gradually increasing length at age prior to 1987 and declining length at age in the period
after, consistent with reports for Lesser sandeel off the Scottish coast (Wanless et al. 2005;
Frederiksen et al. 2011). The gradual decline in length at age corresponds to an earlier CPR time-
series analysis (Planque and Fromentin 1996), showing a similar gradual decline in C. finmarchicus
after the mid-1980s rather than an abrupt change as suggested by Beaugrand et al. (2004). However,
the gradual increase in length at age prior to the mid-1980s is not represented in the trends in C.
finmarchicus development. It should however be noted that the CPR survey does not cover small-
scale changes in vertical distribution of different copepod species: for instance, C. finmarchicus
often resides much deeper than the CPR sampling depth (Jónasdóttir and Koski 2011), It is thus
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
unclear how evident long-term trends could one expect to observe between different trophic levels
occupying different and varying water layers using the present day sampling strategies.
In addition to the connection to Calanus finmarchicus, we searched for alternative explanations for
the changes in sandeel size. Based on long term patterns in total stock biomass and fishing
mortality, we investigated the role of density dependence and fishery induced evolution as drivers
of the observed decreasing length at age (Fig. 6). We found no reason to believe that either of these
were the responsible factor. There were, however, indications that a period of increasing length
after 2003 may be explained by relaxed food competition (density dependence) after a pronounced
stock decline in 2003.
Conclusion: (1) Food consumption is higher for Lesser sandeel dieting on larger copepods
compared to smaller copepods. (2) Handling time limitation provides a plausible causal explanation.
(3) Lesser sandeel actively target large copepods. (4) Length at age of Lesser sandeel began to
decrease in the late 1980s, simultaneously with the C. finmarchicus decline. Although, during other
periods sandeel growth-trends are inconsistent with patterns of C. finmarchicus concentrations
reported in the literature. Overall we found no strong arguments for rejecting the hypothesis that
climate driven reductions in copepod size will lead to diminished growth rates of Lesser sandeel.
In the present study we focused on food consumption measured as biomass. Energy density of
Calanus may on certain times of the year be double that of smaller short-lived non-overwintering
copepods in the North Sea (i.e. Corner and O´Hara 1986). Certainly a difference this large in food
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
quality will most likely also affect fish growth. However, the seasonal dynamics in lipid content and
energy density of relevant North Sea copepod species, experiencing the same environmental
conditions, is currently not available in the scientific literature. We suggest that a food quality study
represents a natural next step from the study presented here.
ACKNOWLEDGEMENT
This research has been supported by EU FP7 grant FACTS (Forage Fish Interactions), Grant
Agreement No. 244966. We like to acknowledge DMI and people (in particular Peter Munk) behind
the LIFECO data for providing modelled temperature data and mesozooplankton samples. We also
like to thank Peter Munk for reading and constructively commenting on the manuscript.
REFERENCES
Arendt KE, Jonasdottir SH, Hansen PJ, Gartner S (2005) Effects of dietary fatty acids on the reproductive success of the calanoid copepod Temora longicornis. Mar Biol 146:513-530
Beaugrand G (2004) The North Sea regime shift: evidence, causes, mechanisms and consequences. Prog Oceanogr 60:245-262
Beaugrand G, Brander KM, Lindley JA, Souissi S, Reid PC (2003) Plankton effect on cod recruitment in the North Sea. Nature 426:661-664
Beaugrand G, Ibanez F, Lindley JA, Reid PC (2002) Diversity of calanoid copepods in the North Atlantic and adjacent seas: species associations and biogeography. Marine Ecology-Progress Series 232:179-195
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398399
400401
402403
404405406
Breteler WCM, Fransz HG, Gonzalez SR (1982) Growth and development of four calanoid copepod species under experimental and natural conditions. Netherland J Sea Res 16:195-207
Broekhuizen N and McKenzie E (1995) Patterns of abundance for Calanus and smaller copepods in the north-sea - time-series decomposition of 2 cpr data sets. Marine Ecology-Progress Series 118:103-120
Cohen RE, Lough RG (1981) Length-Weight Relationships for Several Copepods Dominant in the Georges Bank-Gulf of Maine Area. J Northw Atl Fish Sci 2:47-52
Corner EDS and O´Hara SCM. (1986) The biological chemistry of marine copepods. Clarendon press, Oxford
Corten A (2000) A possible adaptation of herring feeding migrations to a change in timing of the Calanus finmarchicus season in the eastern North Sea. ICES J Mar Sci 57:1261-1270
Daro MH (1988) Migratory and Grazing Behavior of Copepods and Vertical-Distribution of Phytoplankton. Bull Mar Sci 43:710-729
De Silva SS, Balbontin F (1974) Laboratory studies on food intake, growth and food conversion of young herring, Clupea harengus (L.). J Fish Biol 6:645-658
Dragesund O, Johannessen A, Ulltang O (1997) Variation in migration and abundance of norwegian spring spawning herring (clupea harengus L.). Sarsia 82:97-105
Frederiksen M, Edwards M, Richardson AJ, Halliday NC, Wanless S (2006) From plankton to top predators: bottom-up control of a marine food web across four trophic levels. J Anim Ecol 75:1259-1268
Frederiksen M, Elston DA, Edwards M, Mann AD, Wanless S (2011) Mechanisms of long-term decline in size of lesser sandeels in the North Sea explored using a growth and phenology model. Mar Ecol Prog Ser 432:137-147
Hart P J.B, Ison S (1991). The influence of prey size and abundance, and individual phenotype on prey choice by the three-spined stickleback, Gasterosteus aculeatus L.. J Fish Biol 38: 359-372
Helaouët, P., Beaugrand, G., 2009. Physiology, ecological niches and species distribution. Ecosystems 12, 1235 - 1245.
Hjelm J, Persson L (2001) Size-dependent attack rate and handling capacity: inter-cohort competition in a zooplanktivorous fish. Oikos 95:520-532
ICES (2012) ICES WGNSSK Working Group on the assessment of demersal stocks in the North Sea and Skagerrak (www.ices.dk)
407408
409410411
412413
414415
416417
418419
420421
422423
424425426
427428429
430
431432433
434435
436437
438439
Jensen H, Rindorf A, Wright PJ, Mosegaard H (2011) Inferring the location and scale of mixing between habitat areas of lesser sandeel through information from the fishery. ICES J Mar Sci 68:43-51
Jeschke JM, Kopp M, Tollrian R (2002) Predator functional responses: Discriminating between handling and digesting prey. Ecol Monogr 72:95-112
Jonasdottir SH, Koski M (2011) Biological processes in the North Sea: comparison of Calanus helgolandicus and Calanus finmarchicus vertical distribution and production. J Plankton Res 33:85-103
Kjellerup S, Dunweber M, Swalethorp R, Nielsen TG, Moller EF, Markager S, Hansen BW (2012) Effects of a future warmer ocean on the coexisting copepods Calanus finmarchicus and C-glacialis in Disko Bay, western Greenland. Marine Ecology-Progress Series 447:87-U132
Koski M, Jonasdottir SH, Bagoien E (2011) Biological processes in the North Sea: vertical distribution and reproduction of neritic copepods in relation to environmental factors. J Plankton Res 33:63-84
Last JM (1987) The food of immature sprat (Sprattus sprattus (L.)) and herring (Clupea harengus L.) in coastal waters of the North Sea. Journal du Conseil - Conseil International pour l'Exploration de la Mer 44:73-79
Macer CT (1966) Sandeels (Ammodytidae) in the south-western North Sea: Their biology and fishery. MAFF Fishery Invest London ser II 24:1-55
MacLeod CD, Santos MB, Reid RJ, Scott BE, Pierce GJ (2007) Linking sandeel consumption and the likelihood of starvation in harbour porpoises in the Scottish north sea: Could climate change mean more starving porpoises? Biology Letters 3:185-188
Maes J, Tackx M, Soetaert K (2005) The predation impact of juvenile herring Clupea harengus and sprat Sprattus sprattus on estuarine zooplankton. Hydrobiologia 540:225-235
McKenzie N, Broekhuizen E (1995) Patterns of abundance for Calanus and smaller copepods in the North Sea: time series decomposition of two CPR data sets. Mar Ecol Prog Ser 118: 103-120
Misund O, Vilhjalmsson H, Jakupsstovu S, Rottingen I, Belikov S, Asthorsson O, Blindheim J, Jonsson J, Krysov A, Malmberg S, and others (1998) Distribution, migration and abundance of norwegian spring spawning herring in relation to the temperature and zooplankton biomass in the norwegian sea as recorded by coordinated surveys in spring and summer 1996. Sarsia 83:117-127
Moellmann C, Kornilovs G, Fetter M, Koester FW, Hinrichsen H- (2003) The marine copepod, Pseudocalanus elongatus, as a mediator between climate variability and fisheries in the Central Baltic Sea. Fish Oceanogr 12:360-368
440441442
443444
445446447
448449450
451452453
454455456
457458
459460461
462463
464465
466467468469470
471472473
Munk P, Larsson PO, Danielssen DS, Moksness E (1999) Variability in frontal zone formation and distribution of gadoid fish larvae at the shelf break in the northeastern North Sea. Marine Ecology-Progress Series 177:221-233
Nielsen TG, Lokkegaard B, Richardson K, Pedersen FB, Hansen L (1993) Structure of Plankton Communities in the Dogger Bank Area (North-Sea) during a Stratified Situation. Marine Ecology-Progress Series 95:115-131
Olsen EM, Heino M, Lilly GR, Morgan MJ, Brattey J, Ernande B, Dieckmann U (2004) Maturation trends indicative of rapid evolution preceded the collapse of northern cod. Nature 428:932-935
Pitois SG, Shaw M, Fox CJ, Frid CLJ (2009) A new fine-mesh zooplankton time series from the dove sampling station (north sea). J Plankton Res 31:337-343
Planque B, Fromentin JM (1996) Calanus and environment in the eastern North Atlantic .1. Spatial and temporal patterns of C-finmarchicus and C-helgolandicus. Marine Ecology-Progress Series 134:101-109
Planque B, Hays GC, Ibanez F, Gamble JC (1997) Large scale spatial variation in the seasonal abundance of Calanus finmarchicus. Deep-Sea Res I 44:315-326
Planque and taylor (1998) Long-term changes in zooplankton and the climate of the North Atlantic. ICES J Mar Sci 55:644-654
Raab K, Nagelkerke LAJ, Boerée C, Rijnsdorp AD, Temming A, Dickey-Collas M (2012) Dietary overlap between the potential competitors herring, sprat and anchovy in the North Sea. Marine Ecology Progress Series: In press DOI:10.3354/meps09919
Stephens DW, Krebs JR (1986). Foraging theory. Princeton University Press (1986), pp. 13-24
Uye S (1982) Length-weight relationships of important zooplankton from the inland sea Japan. Journal of the Oceanographical Society of Japan 38:149-158
van Deurs M, Christensen A, Frisk C, Mosegaard H (2010) Overwintering strategy of sandeel ecotypes from an energy/predation trade-off perspective. Marine Ecology-Progress Series 416:201-U215
van Deurs M, van Hal R, Tomczak MT, Jonasdottir SH, Dolmer P (2009) Recruitment of lesser sandeel Ammodytes marinus in relation to density dependence and zooplankton composition. Marine Ecology-Progress Series 381:249-258
van Deurs M, Behrens JW, Warnar T, Steffensen JF (2011) Primary versus secondary drivers of foraging activity in sandeel schools (ammodytes tobianus). Mar Biol 158:1781-1789
Wanless S, Harris MP, Redman P, Speakman JR (2005) Low energy values of fish as a probable cause of a major seabird breeding failure in the North Sea. Marine Ecology-Progress Series 294:1-8
474475476
477478479
480481
482483
484485486
487488
489490
491492493
494495
496497
498499500
501502503
504505
506507
Winslade P (1974) Behavioral-Studies on Lesser Sandeel Ammodytes-Marinus (Raitt) .3. Effect of Temperature on Activity and Environmental-Control of Annual Cycle of Activity. J Fish Biol 6:587-599
Yamaguchi A, Ikeda T (2000) Vertical distribution, life cycle, and developmental characteristics of the mesopelagic calanoid copepod Gaidius variabilis (Aetideidae) in the Oyashio region, western North Pacific Ocean. Mar Biol 137:99-109
508509510
511512513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
Number of fishDate Number of included in the
Sample ID dd/mm/yy Latitude Longitude fish per sample diet analysis
1 18/04/2006 54.633 1.183 19 32 21/04/2006 54.584 1.264 22 43 22/04/2006 55.030 1.337 24 54 23/04/2006 54.617 1.200 24 35 23/04/2006 55.317 1.648 21 36 28/05/2006 54.533 1.283 22 37 28/05/2006 54.327 0.872 24 38 29/05/2006 54.327 0.872 23 39 29/05/2006 54.327 0.872 26 210 30/05/2006 55.317 1.648 24 411 31/05/2006 54.584 1.264 19 212 01/06/2006 54.584 1.264 29 213 03/06/2006 54.584 1.264 29 714 04/06/2006 54.145 2.044 22 315 06/06/2006 54.145 2.044 22 416 10/06/2006 54.145 2.044 21 417 10/06/2006 53.907 1.926 24 418 14/06/2006 53.907 1.926 24 419 14/06/2006 54.195 1.529 29 320 16/06/2006 54.277 0.735 35 221 14/04/2007 55.350 1.300 34 422 14/06/2007 55.400 1.300 29 723 22/05/2009 54.200 2.000 66 724 24/05/2009 53.500 1.000 68 425 30/05/2009 54.100 0.900 24 1526 07/04/2010 54.567 1.217 25 2227 12/04/2010 55.317 0.883 26 1028 18/04/2010 55.347 1.512 17 1329 22/05/2010 55.100 1.350 35 1530 07/06/2010 55.067 1.083 16 16
Total 823 181
Table 1. Overview of sandeel samples.
529
530
531
532
2006 2007 2009 2010Small copepods in stomachs (group-1) 0.011 no data 0.009 0.011
(± 0.0011) (± 0.0009) (± 0.0009)Large copepods in stomachs (group-2) 0.022 0.023 no data no data (± 0.0017) (± 0.0029)
Table 2. Geometric mean relative consumption C (g d-1 g-1 fish) by year and copepod size category
(± standard error).
533
534
535
536
Figure 1. Study area. Sandeel habitats in relation to the inflow (arrows) and southerly boarder of
the Calanus finmarchicus protrusion into the North Sea (solid line). The square window defines the
Dogger area. Sandeel habitats were re-produced from (Christensen et al. 2009) and the southerly
boarder of C. finmarchicus was adopted from Planque and Fromentin (1996). Stars represent the
positions of LIFECO copepod samples (n = 62).
537
538
539
540
541
542
543
544
Figure 2. Mean prosome length of copepods found in the sandeel stomachs calculated separately
for each sample. Samples are sorted according to the mean prosome length. Solid whiskers
represent standard errors and dashed whiskers represent standard deviations. Sample identifiers
correspond to sample ID in Table 1. The horizontal dashed line represents the threshold used to
divide samples into group-1 and group-2.
545
546
547
548
549
550
551
552
Figure 3. The effect of copepod size (prosome length) on consumption rate C [g (g fish) -1 d-1].
Graph displays the geometric mean food consumption (+/- s.e.) for small copepods (group-1) and
large copepods (group-2) and for small (black) and large (white) sandeels. n = 173, 293, 183, and
174 for group-1/small sandeels, group-1/large sandeels, group-2/small sandeels, and group-2/large
sandeels respectively. Mean size of small sandeels: 10.65 cm +/- 1.26 (s.d.) and 3.3 g; and large
sandeels: 15.14 cm +/- 1.28 (s.d.) and 10.3 g.
553
554
555
556
557
558
559
560
Figure 4. The effect of copepod size on maximum allowed handling time MAH [s prey-1]. Graph
displays the geometric mean MAH assuming a daily effective foraging period τ of 13 hours (± 25%)
for each of three scenarios of daily ration and copepod size. A: Daily ration as observed in group-1
(see Fig. 3) and prosome length = 0.85 mm, B: Daily ration as observed in group-2 (see Fig. 3) and
prosome length = 0.85 mm, and C: Daily ration as observed in group-2 and prosome length = 1.75
mm. Black symbols represent small sandeels (< 12 cm) and white symbols represent large sandeels
(> 12 cm). Horizontal dashed line defines the upper limit of realistic handling times reported for
zooplanktivorous fish of similar size (e.g. Hjelm and Persson 2001).
561
562
563
564
565
566
567
568
569
Figure 5. Co-occurrence of small and large copepods in the vicinity of North Sea sandeel areas.
Graphs show the concentration of Calanus versus the concentration of small copepods (A: numbers
m-2; B: g m-2) and each point represents a given sampling station. Small copepods include Acartia,
Temora, Centropages, Pseudocalanus, Oithona, Microcalanus, Oncaea, and Microsetella. p-values
were derived using the spearman correlation test.
570
571
572
573
574
575
576
577
Figure 6. Trends in total length at age for Lesser sandeel in the Dogger area in relation to the
regime shift in the late 1980s, total stock biomass and fishing mortality. Upper graph displays
length at age for age 1 (grey) and age 2 (black) together with trend lines (dashed lines) fitted to the
period before and after 1987. Lower graph depicts total stock biomass (bars) and fishing mortality
(lines) back to 1983 as provided by the ICES working group on Lesser sandeel in the North Sea.
578
579
580
581
582
583
584
585
586
587
588
589