Warming reinforces non-consumptive predator effects on prey growth,
physiology and body stoichiometry
1
2
3
LIZANNE JANSSENS 1, MARIE VAN DIEVEL1 AND ROBBY STOKS 2
4
5
6
Laboratory of Aquatic Ecology, Evolution and Conservation, University of Leuven, Leuven,
7
Belgium
8
9
1
Joint first author
10
2
Corresponding author
11
12
Running title: Warming, non-consumptive predator effects and elemental stoichiometry
13
1
14
Abstract. While non-consumptive effects of predators may strongly affect prey populations,
15
little is known how future warming will modulate these effects. Such information would be
16
especially relevant with regard to prey physiology and resulting changes in prey
17
stoichiometry. We investigated in Enallagma cyathigerum damselfly larvae the effects of a 4
18
°C warming (20 °C vs. 24 °C) and predation risk on growth rate, physiology and body
19
stoichiometry, for the first time including all key mechanisms suggested by the general stress
20
paradigm (GSP) on how stressors shape changes in body stoichiometry. Growth rate and
21
energy storage were higher at 24 °C. Based on thermodynamic principles and the growth rate
22
hypothesis, we could demonstrate predictable reductions in body C:P under warming and link
23
these to the increase in P-rich RNA; the associated warming-induced increase in C:N may be
24
explained by the increased synthesis of N-rich proteins. Yet, under predation risk, growth rate
25
instead decreased with warming and the warming-induced decreases in C:N and C:P
26
disappeared. As predicted by the GSP, larvae increased body C:N and C:P at 24 °C under
27
predation risk. Notably, we did not detect the assumed GSP-mechanisms driving these
28
changes: despite an increased metabolic rate there was neither an increase of C-rich
29
biomolecules (instead fat and sugars contents decreased under predation risk), nor a decrease
30
of N-rich proteins. We hypothesize that the higher C:N and N:P under predation risk are
31
caused by a higher investment in morphological defense. This may also explain the stronger
32
predator-induced increase in C:N under warming. The expected higher C:P under predation
33
risk was only present under warming and matched the observed growth reduction and
34
associated reduction in P-rich RNA. Our integrated mechanistic approach unraveled novel
35
pathways of how warming and predation risk shape body stoichiometry. Key findings that (i)
36
warming effects on elemental stoichiometry were predictable and only present in the absence
37
of predation risk, and that (ii) warming reinforced the predator-induced effects on C:N:P, are
2
38
pivotal in understanding how non-consumptive predator effects under global warming will
39
shape prey populations.
40
41
42
43
Key words: climate change; elemental stoichiometry; non-consumptive predator effects;
life history; metabolic ecology; nutrient cycling; predator-induced plasticity
INTRODUCTION
Understanding the interplay between temperature and biotic interactions is essential for
44
fully anticipating how populations, communities and ecosystem functions will respond to
45
global warming (Traill et al. 2010, Angert et al. 2013, Blois et al. 2013). While mild warming
46
may positively affect fitness in many ectotherms (Deutsch et al. 2008), these positive effects
47
of warming on prey populations may no longer hold in the presence of predators (De Block et
48
al. 2013, Barton and Ives 2014). The ability of populations to survive locally under global
49
warming will therefore not only depend on their ability to deal with the temperature increase
50
itself, for example through physiological adjustments (Chown and Gaston 2008), but also on
51
their ability to deal with temperature-induced changes in the interactions with predators
52
(Traill et al. 2010). While there is a growing appreciation for the importance of non-
53
consumptive effects imposed by predators, which can often outweigh the importance of direct
54
feeding (Preisser et al. 2005), the way non-consumptive predator effects change under
55
warming is largely unknown (but see e.g., Miller et al. 2014). This is especially relevant for
56
physiological stress effects imposed by predators as these can shape prey stoichiometry and
57
thereby scale up to ecosystem functions linked to elemental cycling (Hawlena and Schmitz
58
2010a, Hawlena et al. 2012). Studying how warming and predator-induced stress jointly shape
59
prey stoichiometry may therefore provide crucially important mechanistic insights for
60
forecasting future consequences of global warming (Schmitz 2013).
61
According to metabolic ecology (Sibly et al. 2012), mild warming likely will affect prey
62
stoichiometry through its positive effects on growth rate. Based on thermodynamic principles,
3
63
reaction rates inevitably increase with absolute temperatures because the kinetic energy of a
64
system increases with absolute temperature. Therefore, animals reared at higher non-lethal
65
temperatures typically have higher growth rates (Nilsson-Örtman et al. 2012). Further, the
66
growth rate hypothesis (Elser et al. 1996) asserts that faster growth rates will require the
67
allocation of resources to P-rich ribosomal RNA to increase the synthesis of N-rich proteins,
68
resulting in decreased body ratios of C:P and N:P (Sterner and Elser 2002). Taken together,
69
we therefore expect reductions in C:P and N:P under mild warming.
70
Central to our understanding of how predator stress shapes prey stoichiometry is the
71
general stress paradigm (GSP, Hawlena and Schmitz 2010a). According to the GSP, prey
72
under predation risk will increase their metabolic rate, thereby mobilize energy for predator
73
escape and divert energy toward costly defense mechanisms away from production of new
74
tissues. This is associated with an increased gluconeogenesis (creation of C-rich biomolecules
75
from N-rich proteins) and the excretion of excess nutrients, mostly N and P, resulting in
76
increased body ratios of C:N and C:P (Hawlena and Schmitz 2010a). While a predator-
77
induced increase in C:N has been demonstrated in Melanoplus femurrubrum grasshoppers
78
(Hawlena and Schmitz 2010b), the very few follow-up studies in other taxa showed deviating
79
patterns and highlighted additional behavioral and morphological responses to predation risk
80
that may overwhelm or counteract the stoichiometric changes predicted by the GSP (Costello
81
and Michel 2013, Dalton and Flecker 2014).
82
To gain insight in how mild warming and predation risk affected body nutrient condition
83
we investigated the combined impact of a mild (4 °C) temperature increase and predation risk
84
on prey physiology with special attention for changes in body stoichiometry. We studied this
85
in Enallagma damselfly larvae, among the most extensively studied invertebrates with regard
86
to physiological responses to predators (e.g., McPeek et al. 2001, Stoks et al. 2005b, 2006,
87
Slos and Stoks 2008, Culler et al. 2014, Janssens and Stoks 2014). We predicted decreased
4
88
C:P and N:P under a mild temperature increase (growth rate hypothesis, Elser et al. 1996,
89
Sterner and Elser 2002, Sibly et al. 2012), and as temperature-induced growth increases have
90
been shown to be less pronounced under predation risk in the damselfly Enallagma vesperum
91
(Culler et al. 2014), we expected less strong effects of warming on body stoichiometry under
92
predation risk. We predicted increased C:N and C:P under predation risk (GSP, Hawlena and
93
Schmitz 2010a), and as the effects of predation risk predicted by the GSP are driven by
94
increases in metabolic rate, we expected stronger effects of predation risk on prey growth,
95
physiology and stoichiometry under mild warming (see also Laurila et al. 2008, Kuehne et al.
96
2012, Culler et al. 2014, but see Touchon and Warkentin 2011). To advance our mechanistic
97
understanding, we quantified all variables associated with the assumed key mechanisms put
98
forward by these theories: individual growth rates, RNA:DNA ratios (as a proxy for protein
99
synthesis, Karasov and Martinez del Rio 2007), the activity of the electron transport system
100
(ETS, as a proxy for metabolic rate, De Coen and Janssen 2003), and energy storage
101
molecules (fat, sugars and proteins). This multifaceted approach provided unique information
102
how mild warming and predation risk shape body stoichiometry.
103
METHODS
104
Collecting and housing
105
Mated females (n = 20) of the damselfly Enallagma cyathigerum were collected in “Het
106
Stappersven” (Kalmthout, Belgium), a fishless lake with Anax dragonfly larvae as top
107
predators, and transferred to the laboratory for egg laying. After hatching, larvae were placed
108
individually in 200 ml cups in temperature-controlled water baths set at 20 °C and 24 °C
109
(photoperiod 14:10 L:D). Importantly, 20 °C is the mean summer water temperature in
110
shallow lakes occupied by the study species in Flanders, yet water temperatures of 24 °C are
111
frequently observed during summer. The 4 °C temperature difference corresponds with the
112
predicted surface temperature increase by 2100 under IPCC scenario RCP8.5 (IPCC 2013).
5
113
Damselfly larvae were fed ad libitum with Artemia nauplii five days a week (mean daily dose
114
± 1 SE: 673 ± 53 nauplii, n = 10 daily doses).
115
Experimental setup
116
To test for the single and combined effects of rearing temperature and predation risk on
117
growth rate and associated physiological variables, we set up a full factorial experiment with
118
two temperature treatments (20 °C and 24 °C) crossed with two predation risk treatments
119
(absence and presence). While the temperature treatment started when the eggs hatched, the
120
predation risk treatments were imposed during a 7-day exposure period starting when larvae
121
molted into their final instar. Larvae entered the exposure period after 144 ± 1 (mean ± 1 SE)
122
days at the 20 °C rearing temperature and after 135 ± 1 days at the 24 °C rearing temperature.
123
During the exposure period, larvae were placed individually in glass vials (100 ml) at their
124
respective rearing temperature. Sets of four vials were placed together in a larger outer
125
container (750 ml). To avoid any bias due to larvae being associated with a specific container
126
(set of conspecific larvae and predator), we randomly re-distributed vials among containers of
127
the same temperature-by-predation risk combination on a daily basis (see McPeek 2004).
128
Throughout the exposure period, larvae were daily fed ad libitum with Artemia nauplii (mean
129
daily dose ± 1 SE: 1224 ± 108 nauplii, n = 10 daily doses). The number of larvae tested at
130
each treatment combination was 45 (total of 180 larvae).
131
Predation risk was manipulated using visual and chemical predator cues, reflecting the
132
cocktail of predator cues that damselfly larvae encounter in nature. To ensure visual predator
133
cues, a large Anax dragonfly larva, important predators of damselfly larvae (Stoks et al.
134
2005a), was placed in the outer container of the treatment with predator cues. Additionally,
135
larvae could see the conspecific larvae in the other vials in the container (damselfly larvae are
136
cannibalistic, De Block and Stoks 2004). When no complete sets of four larvae could be made
137
for the treatments with predation risk at a given temperature, we added vials with ‘dummy
6
138
larvae’, these were final instar conspecific larvae not included in the experiment. To avoid
139
visual predator cues in the treatment without predation risk, the walls of these vials were
140
made non-transparent using tape (this did not affect light levels in the vials). For the chemical
141
predator cues, we homogenized one E. cyathigerum larva in 20 ml of water from an aquarium
142
filled with 300 ml aged tap water in which a large Anax dragonfly larva had eaten one E.
143
cyathigerum larva. We daily added 1 ml of this predator medium to each vial of the treatments
144
with predator cues, to the other vials we daily added 1 ml of aged tap water.
145
Response variables
146
To quantify growth rate, we weighed each larva to the nearest 0.01 mg at the start and at
147
the end of the 7-day exposure period. Growth rate was calculated as [ln(final mass) – ln(initial
148
mass)] / 7 days (McPeek et al. 2001) for all 180 larvae. After determining final mass, the
149
larvae were directly frozen on dry ice and stored at -80 °C for physiological analyses. Given
150
that not all physiological variables could be measured on the same larva, we worked with
151
three randomly chosen sets of larvae.
152
A first set of larvae (20 per treatment combination, total of 80 larvae) was used to quantify
153
RNA:DNA ratios based on the protocol by Vrede et al. (2002). Larvae were homogenized and
154
diluted 15 times in extraction buffer (50 mM EDTA, 0.05 % SDS in 50 mM Tris). In a first
155
step, we measured the total amount of RNA and DNA by filling a 96 well black microtiter
156
plate with 100 µl sample and 2 µl ethidium bromide (100 µg/ml). After an incubation on ice
157
for 15 minutes, fluorescence was measured using a spectrophotometer (Infinite M2000,
158
TECAN) at an excitation/emission wavelength of 535:595 nm. In a second step, we measured
159
the total amount of DNA by filling a 96 well black microtiter plate with 100 µl of the sample
160
and 1 µl of RNase (20 mg/ml) (which break downs the RNA). After an incubation of 60
161
minutes at room temperature, 2 µl ethidium bromide (100 µg/ml) was added. After an
162
incubation on ice for 15 minutes, fluorescence was measured. Subtracting the amount of DNA
7
163
from the total amount of RNA and DNA, resulted in the RNA concentration in the samples.
164
RNA and DNA concentrations were measured in triplicate and the means per larva were used
165
for the statistical analyses.
166
A second set of larvae (10 per treatment combination, total of 40 larvae) was used to
167
quantify electron transport system (ETS) activity based on the protocol of De Coen and
168
Janssen (2003). Larvae were homogenized and diluted 15 times in a homogenization buffer
169
(100 mM Tris-HCl, pH 8.5, 15 % polyvinyl pyrrolidone, 153 µM MgSO4 and 0.2 % Triton X-
170
100). A 96 well microtiter plate was filled with 150 µl buffered substrate solution (0.13 M
171
Tris HCl, 0.3 % Triton X-100, 1.7 mM NADH, 250 µM NADPH, pH 8.5) and 50 µl of the
172
supernatant. We started the reaction by adding 100 µl INT (8 mM p-iodonitrotetrazolium) and
173
followed the increase in absorbance at 490 nm (Infinite M2000, TECAN) and 20 °C during 5
174
minutes with readings every 30 seconds. Using the formula of Lambert-Beer we calculated
175
the concentration of formazan (extinction coefficient 15,900 M-1cm-1) and afterwards
176
converted this to cellular oxygen consumption based on the theoretical stoichiometric
177
relationship that for each 2 µmol of formazan formed, 1 µmol of O2 was consumed in the ETS
178
system. Measurements were done in duplicate. The means were used for statistical analyses
179
and expressed as nmol O2/min.
180
A third set of larvae (15 per treatment combination, total of 60 larvae) was used to quantify
181
energy reserves and C:N:P ratios. Larvae were homogenized and diluted 5 times in milli-Q
182
water. Afterwards, 35 µl of the supernatant was further diluted 3 times with milli-Q water and
183
used to measure energy reserves. The rest of the homogenate was used to quantify C:N:P
184
ratios. Fat content was measured following Janssens and Stoks (2014). We mixed 8 µl of the
185
body supernatant and 56 µl sulfuric acid (100 %) in a glass tube. The tubes were heated for 20
186
minutes at 150 °C. Afterwards, we added 64 µl of milli-Q water. We filled a 384 well
187
microtiter plate with 30 µl of the sample and measured absorbance at 340 nm (Infinite
8
188
M2000, TECAN). Fat concentrations were calculated using a standard curve of glyceryl
189
tripalmitate. Measurements were done in triplicate and the means per larva were used for
190
statistical analyses. For total sugar content (glucose + glycogen), we used the protocol
191
described in Stoks et al. (2006) based on the glucose kit of Sigma Aldrich USA. In a first step,
192
all glycogen was transformed to glucose. Therefore, we mixed 50 µl milli-Q water, 20 µl
193
body supernatant and 10 µl amyloglucosidase (1 U/10 µl) (Sigma A7420) in a 96 well
194
microtiter plate. After 30 minutes of incubation at 37 °C, all glycogen is transformed to
195
glucose. We measured the glucose levels by adding 160 µl of glucose assay reagent (Sigma
196
G3293) to each well. After another incubation period of 20 minutes at 30 °C we measured
197
absorbance at 340 nm (Infinite M2000, TECAN). We calculated sugar concentration based on
198
a standard curve of known concentrations of glucose and their absorbance. Measurements
199
were done in duplicate and the means per larva were used for statistical analyses. The results
200
for glucose and glycogen concentrations were very similar, therefore we only report the total
201
sugar content in the results and discussion sections. Protein content in the body homogenates
202
was measured using the Bradford method (Bradford 1976). Measurements were done in
203
triplicate and the means per larva were used for statistical analyses. Fat content, total sugar
204
content and protein content were expressed as µg per mg dry mass.
205
For the quantification of C:N:P ratios, we divided the homogenate in two parts: ¼ of the
206
sample was used for C and N analyses and ¾ of the sample for P analyses. For the
207
quantification of the C and N content, samples were placed in tin cups and dried for 24 h (60
208
°C), where after C and N were quantified using an elemental analyzer (Carlo Erba 1108). For
209
P analysis we mixed the sample with 1000 µl HNO3 in a glass tube (based on Van
210
Moorleghem et al. 2013). The tubes were heated for 15 minutes at 150 °C. Afterwards, the
211
digests were diluted to 10 ml with milli-Q water and analyzed using inductively coupled
212
plasma mass spectrometry (Agilent 7700x ICP-MS). The measurements (n = 8) of the
9
213
Spectrapure Standard SPS-SW2 Batch 128 showed 1.41 % deviation. C:N, C:P and N:P were
214
expressed as molar ratios. All measured physiological variables were within the range
215
observed for aquatic invertebrates, including damselfly larvae (Appendix A, Table A1).
216
Statistical analyses
217
We used two-way ANOVAs to test for the effects of temperature and predation risk on the
218
different response variables. We compared means using linear contrasts which were corrected
219
for multiple testing using the false discovery rate procedure as outlined in Benjamini and
220
Hochberg (1995). Given the interest for temperature as well as for predation risk, their two-
221
way interactions will be described from the perspective of each factor separately. All tests
222
were done in STATISTICA v12 (StatSoft, Tulsa, OK, U.S.A.). Values of p < 0.05 were
223
considered significant. For all variables the assumptions of ANOVA (normal distribution and
224
homogeneity of variances) were met without the need for transformations.
225
RESULTS
226
Life history, RNA:DNA and ETS
227
Rearing temperature and predation risk interacted for growth rate, the RNA:DNA ratio and
228
ETS activity, but their interaction patterns differed (Table 1, Fig. 1). All three variables were
229
affected by temperature, yet these effects strongly depended upon predation risk. While
230
growth rate increased at 24 °C in the absence of predation risk, it decreased at 24 °C when
231
predation risk was present. The RNA:DNA ratio increased at 24 °C but only in larvae not
232
exposed to predation risk. ETS activity increased at 24 °C, especially in larvae exposed to
233
predation risk.
234
All three variables were affected by predation risk and these effects were stronger at 24°C.
235
Growth rate was lower under predation risk and this predator-induced growth reduction was
236
ca. 2× stronger at 24 °C (44 %) than at 20 °C (21 %). Exposure to predator cues resulted in a
237
reduction of the RNA:DNA ratio and this predator-induced reduction was 2× stronger at 24
10
238
°C (21 %) than at 20 °C (11 %). Exposure to predator cues resulted in an increased ETS
239
activity, and this increase was ca. 2× stronger at 24 °C (31 %) than at 20 °C (17 %).
240
Energy storage
241
All three storage molecules had higher levels at 24 °C than at 20 °C in the absence of
242
predation risk (Table 1, Fig. 2). Yet, in the presence of predator cues this was only true for
243
protein content. Instead, fat content and total sugar content did not increase with temperature
244
under predation risk (Temperature × Predation risk, Table 1). Exposure to predator cues
245
resulted in lower fat and sugar contents and these reductions were stronger at 24 °C
246
(reductions in fat content: 34 % at 24 °C versus 17 % at 20 °C; reductions in sugar content: 39
247
% at 24 °C versus 24 % at 20 °C, Fig. 2A-B). Neither predation risk, nor the interaction
248
between rearing temperature and predation risk affected the protein content (Table 1).
249
250
Body stoichiometry
Both the C:N and the C:P ratios were lower at 24 °C than at 20 °C, yet only in the absence
251
of predation risk (Temperature × Predation risk, Table 1, Fig. 3A-B). In the presence of
252
predation risk, the temperature had no effect on these ratios. Animals exposed to predator
253
cues had a ca. 2× stronger increase of C:N ratios at 24 °C (14 %) than at 20 °C (6 %), while
254
the predator-induced increase in C:P ratio (ca. 16 %) was only present at 24 °C (Fig. 3A-B).
255
The N:P ratio was higher in larvae exposed to predation risk (Table 1, Fig. 3C). Neither the
256
rearing temperature, nor the interaction between rearing temperature and predation risk
257
affected the N:P ratio.
258
DISCUSSION
259
Both temperature and predation risk strongly affected growth rate, RNA:DNA ratio, ETS
260
activity, energy storage and the C:N:P signature of the damselfly larvae. Mild warming in the
261
absence of predation risk induced the predicted reduction in C:P (Elser et al. 1996, Sterner et
262
al. 2002, Sibly et al. 2012). Opposite to our predictions, warming did not decrease N:P,
11
263
instead, warming reduced C:N. The effects of warming on most end points were as predicted
264
less pronounced in the presence of predation risk. Predation risk induced the expected
265
increases in C:N and C:P at 24°C (Hawlena and Schmitz 2010a). In addition, also N:P
266
increased under predation risk. Finally, the impact of predation risk was as expected stronger
267
under mild warming (Laurila et al. 2008, Kuehne et al. 2012, Culler et al. 2014). We will
268
integrate the patterns of all variables measured with special focus on the explanation of
269
response patterns in body stoichiometry for these four scenarios.
270
Temperature effects in the absence of predation risk
271
In accordance with thermodynamic principles, larval growth rate and energy storage were
272
higher at 24 °C than at 20 °C. This fits the observation in other damselflies (e.g., Stoks et al.
273
2012, Culler et al. 2014) and insects (Karl and Fischer 2008) that food intake and growth
274
efficiency increase in this temperature range. This can in turn explain the increases in growth
275
rate and in energy storage. Following the metabolic theory of ecology (Gillooly et al. 2001),
276
metabolic rate (as measured by ETS activity) was also higher in larvae reared at the higher
277
temperature.
278
Integrated changes in growth rate, RNA:DNA and protein content can explain the observed
279
stoichiometric changes under mild warming in the absence of predation risk. In line with the
280
growth rate hypothesis (Elser et al. 1996), the increased growth rate at 24 °C was associated
281
with a higher RNA:DNA ratio and lower C:P. A lower C:P is expected because of an
282
increased allocation of resources to P-rich ribosomal RNA to increase the synthesis of
283
proteins (Karasov and Martinez del Rio 2008), as here indicated by the increased RNA:DNA
284
ratio. This is in line with other studies showing higher RNA:DNA ratios in faster growing
285
animals (e.g., Weider et al. 2004, Elser et al. 2006). While often not explicitly addressed by
286
the growth rate hypothesis (but see e.g., Watts et al. 2006, Chen et al. 2010, Reef et al. 2010)
287
we also observed a lower C:N in the faster growing animals at 24 °C. This can be explained
12
288
because rapid growth requires increased allocation to ribosomes (hence also ribosomal
289
proteins), RNA and protein synthetic products (Watts et al. 2006). This might explain the here
290
observed pattern in damselflies as we did indeed found higher protein levels in the faster
291
growing larvae at 24 °C. Higher growth rates have been shown to be inversely correlated with
292
C:N in several other taxa (e.g., clovers: Chen et al. 2010, fruit flies: Watts et al. 2006; trees:
293
Reef et al. 2010). Note that the observed increases in C-rich fat and sugar storage molecules
294
under warming were overruled by the increases in N and P. Furthermore, the increases in N
295
and P balanced out, resulting in no change in the N:P ratio under mild warming.
296
The very few studies that experimentally tested for an effect of temperature on body
297
stoichiometry did not quantify the underlying physiological mechanisms. Liess et al. (2013)
298
documented similar decreases in C:N and C:P when Rana temporaria tadpoles were reared
299
from the egg stage at 23 °C compared to 18 °C. In contrast with our study, tadpole growth rate
300
was, however, lower at the higher temperature. The higher N content at the higher
301
temperature was therefore assumed to be the result of an increased protein synthesis in
302
relation to thermal tolerance; the higher P content at the higher temperature was hypothesized
303
to be the result of morphological changes (Liess et al. 2013). Schmitz (2013) instead
304
documented a higher C:N in Melanoplus femurrubrum grasshoppers kept in outdoor
305
mesocosms warmed 2.5-3 °C above ambient. This was assumed to be driven by a shift of
306
nutrient demands from N-rich to C-rich food and an excretion of excess nitrogen in response
307
to increased metabolism under thermal stress. Compared to our study, the opposite change in
308
C:N at higher temperature in the study of Schmitz (2013) may be because the high
309
temperature was considered as stressful and the grasshoppers could change food preference.
310
Temperature effects in the presence of predation risk
311
When mild warming was combined with predation risk, the damselfly larvae showed the
312
highest metabolic rate (as measured by ETS activity) which likely increased energetic costs
13
313
(Lemoine and Burkepile 2012) to such extent that the positive effect of temperature on growth
314
rate changed into a negative effect and that the temperature-induced increases in RNA:DNA
315
ratio and in fat and total sugar contents disappeared. As an important consequence, while the
316
growth rate had the highest values at the high temperature in the control larvae, it reached the
317
lowest values when the high temperature was combined with predation risk. In line with this,
318
the temperature-induced growth increase in the damselfly Enallagma vesperum was less
319
pronounced under predation risk, which was also attributed to a higher metabolic rate (Culler
320
et al. 2014). The disappearance of the positive temperature effects on growth rate and
321
RNA:DNA ratios combined with an increased morphological defense (see below) can explain
322
why the decrease in C:P under mild warming in the absence of predation risk disappeared
323
when mild warming was combined with predation risk.
324
Effects of predation risk in the absence of warming
325
As predicted by theory (Abrams and Rowe 1996) and widely documented (Benard 2004),
326
larvae decreased growth rate under predation risk. This can be explained by a combination of
327
a lower food intake and a lower efficiency to convert food into body mass (e.g., McPeek et al.
328
2001, Stoks et al. 2005c, Trussell et al. 2006, Miller et al. 2014). According to the general
329
stress paradigm (GSP, Hawlena and Schmitz 2010a), when exposed to predation risk, prey
330
increase metabolic rate, thereby mobilize energy for predator escape and divert more energy
331
toward the upregulation of costly defense mechanisms and less toward production of new
332
tissues (hence growth). This may explain the observed increase in ETS activity (as a measure
333
of metabolic activity), and reductions in energy storage (fat and total sugars) (see also Stoks et
334
al.2005b, 2006, Thaler et al. 2012) and in RNA:DNA ratio under predation risk.
335
The observed predator-induced increase in body C:N is as expected by the GSP (Hawlena
336
and Schmitz 2010a, Costello and Michel 2013) and documented before in M. femurrubrum
337
grasshoppers (Hawlena and Schmitz 2010b, Hawlena et al. 2012). To meet the energetic
14
338
demands of the predator-induced stress response, prey are expected to increase their
339
production of C-rich glucose through the breakdown of the N-rich proteins (gluconeogenesis)
340
and release the excess nutrients, mostly N and P (Hawlena and Schmitz 2010a). Yet, despite
341
indications of a general stress response in our study species when exposed to predator cues
342
(see above, Slos and Stoks 2008), we did not detect an increased investment in C-rich
343
biomolecules (instead the fat and sugar contents were lower in larvae exposed to predation
344
risk), neither a decreased protein content. Note that the only studies that explicitly tested for
345
changes in N excretion under predation risk reported, in contrast with the GSP prediction,
346
reduced N excretion in guppies (Dalton and Flecker 2014) and in caterpillars (Thaler et al.
347
2012), indicating that also other mechanisms are shaping predator-induced patterns in these
348
biomolecules. One recently invoked mechanism by Dalton and Flecker (2014) that may have
349
contributed to the here observed patterns in these biomolecules, is the physiology of food
350
restriction associated with the lowered food intake under predation risk. Animals faced with
351
food restriction prefer to mobilize glycogen and lipid stores for energy production and spare
352
the resource stores (i.e. proteins) most needed for future physiological activities (Wang et al.
353
2006). In support of this mechanism, larvae of the study species decrease food intake under
354
predation risk (Stoks et al. 2005a, Janssens and Stoks 2014), and food-restricted damselfly
355
larvae show reduced levels of fat and sugars (Stoks et al. 2006). This food restriction
356
mechanism may overrule the GSP response and has been assumed to contribute to the
357
predator-induced reduction in C:N in guppies (Dalton and Flecker 2014).
358
Although the pattern of predator-induced increase in body C:N in our study species is as
359
predicted by the GSP, this pattern cannot be fully explained by the GSP (more particularly by
360
the changes in biomolecules). We hypothesize that the higher C:N under predation risk is
361
caused by a higher investment in the exoskeleton, which mainly consists of chitin, a
362
polysaccharide with a high C:N (5.1) compared to other biomolecules (Sterner and Elser
15
363
2002). Such a morphological response is assumed to be a widespread structural defense
364
strategy to reduce predator attack efficiency (Rabus et al. 2013) and has been observed in
365
aquatic insects (e.g., mayfly larvae, Dahl and Peckarsky 2002). Future work on predator-
366
induced effects on body stoichiometry would therefore benefit from explicitly measuring
367
cuticle thickness and chitin content. Notably, Costello and Michel (2013) attributed the
368
absence of a change in C:N under predation risk in Hyla versicolor tadpoles to strong
369
morphological defenses (i.e. increased size of the protein-rich tail muscles). Taken together,
370
this suggests a major role in morphological defense mechanisms for shaping C:N under
371
predation risk and thereby overwhelming or counteracting some key GSP-mechanisms.
372
Together with Costello and Michel (2013), we show that also body N:P increased under
373
predation risk. They hypothesized this was the result of an increase in protein content linked
374
to more N-rich muscle tissue. In our study, the protein content, however, was not influenced
375
by predation risk. Moreover, also the predator-induced decrease in RNA:DNA would, if
376
anything, work to reduce the N:P content. The above invoked higher investment in the
377
exoskeleton (and higher chitin content) in the presence of predation risk may also explain the
378
higher N:P. Indeed, chitin contains nitrogen but no phosphorus (Sterner and Elser 2002).
379
Therefore, a thicker exoskeleton in the presence of predation risk would also result in a higher
380
N:P ratio.
381
382
Effects of predation risk under mild warming
As expected the predator-induced reductions in growth rate and the associated RNA:DNA
383
ratio and the reductions in fat and total sugar contents were stronger under mild warming.
384
This can be explained because these predator-induced changes are generated by predator-
385
induced increases in metabolic rate (GSP, Hawlena and Schmitz 2010a) while metabolic rates
386
were the highest under the combination of predation risk and mild warming. In accordance
387
with our findings, stronger predator-induced growth reductions under mild warming have
16
388
been observed in Chinook salmon (Oncorhynchus tshawytscha) (Kuehne et al. 2012) and E.
389
vesperum damselfly larvae (Culler et al. 2014), and in both cases attributed to increased
390
metabolic demands at the higher temperature.
391
In line with the GSP we also found an increased C:P under predation risk, but only at 24
392
°C. The only other study on this also demonstrated a higher C:P in tadpoles exposed to
393
predation risk (Costello and Michel 2013). They assumed this could be explained by the GSP
394
as the result of an increased gluconeogenesis (i.e. production of C-rich glucose) and the
395
excretion of P. In our study, however, this explanation is unlikely given the lower sugar
396
content under predation risk. Instead, in the condition with increased C:P (predation risk at 24
397
°C), the growth reduction and the associated lowered RNA:DNA ratio were the strongest,
398
which can explain the decreased P content, hence the higher C:P under predation risk at 24°C.
399
The predator-induced increase in C:N which was stronger under warming, could not be
400
explained by the GSP as we did not observe increases in fat and total sugar content nor a
401
decrease in protein content in this condition. Instead, the above suggested increase in chitin
402
content might also explain the stronger predator-induced increase in C:N at 24 °C as
403
morphological defenses against predators have been shown to be more pronounced at higher
404
temperatures (e.g., Laurila et al. 2008).
405
406
Synthesis and conclusions
By combining a mild 4 °C temperature increase (matching IPCC [2013] warming scenario
407
RCP 8.5) and predation risk, we obtained three novel insights directly relevant for
408
understanding the impact of global warming on prey organisms. Firstly, the effects of
409
warming did strongly depend upon predation risk. In accordance with thermodynamic
410
principles (Sibly et al. 2012) larvae reared at the higher temperature performed better (had
411
higher growth rates and energy reserves). Yet, when larvae were exposed to predation risk,
412
the positive effect of the higher temperature on the growth rate reversed into a negative effect,
17
413
and the increase in energy reserves disappeared. This indicates that laboratory studies should
414
be used with caution when making predictions under global warming, as in nature prey
415
organisms typically experience predation risk. Secondly, our results indicate that the negative
416
impact of predation risk on performance (growth rate and energy storage) is expected to
417
increase under global warming in this study system. Predation risk is increasingly appreciated
418
as an important driver of population dynamics (Preisser et al. 2005) and even community
419
composition (Peacor et al. 2012). While recent studies stressed the importance of increased
420
consumptive predation under global warming (e.g., De Block et al. 2013), we here highlight
421
another pathway through which warming may indirectly affect ecological communities: by
422
changing non-consumptive effects of predators (see also Miller et al. 2014). This mechanism
423
may contribute to the observed stronger top-down effects by predators on their prey at higher
424
temperatures (e.g., O’ Gorman et al. 2012, Kratina et al. 2012).
425
Thirdly, how temperature (Schmitz 2013) and predation risk (Hawlena and Schmitz 2010a,
426
b) shape body stoichiometry may profoundly affect the functioning of food webs, particularly
427
elemental cycling, and is therefore key to understand the impact of global warming in natural
428
systems. For example, soil samples that received carcasses of grasshoppers with a 4 % higher
429
C:N body content due to exposure to predators, showed a threefold decrease in the
430
mineralization of plant litter (Hawlena et al. 2012). By integrating metabolic ecology and the
431
growth rate hypothesis, we could demonstrate predictable changes in body stoichiometry in
432
the absence of predation risk. Given the generality of the underlying mechanisms this likely
433
provides a useful predictive framework for many other ectotherms. Notably, we extended the
434
very few studies that experimentally studied the separate effects of temperature and predation
435
risk on body stoichiometry by also focusing on their combined effects and by quantifying the
436
full set of assumed driving key biomolecules (i.e. RNA:DNA, fat, sugars and proteins).
437
Building on the inspiring and innovative studies of Hawlena and Schmitz (2010a, b) and
18
438
Schmitz (2013), our integrated approach provided novel insights about the interplay of
439
warming and non-consumptive effects on prey stoichiometry and the underlying mechanisms.
440
Current study together with the very few other ones that looked at the effects of warming
441
(Liess et al. 2013) and predation risk (Costello and Michel 2013, Dalton and Flecker 2014)
442
highlighted the interplay of the physiological GSP-responses with behavioral changes
443
affecting food restriction and morphological defenses. Further validating this predictive
444
mechanistic framework is an important challenge linking stress ecology and ecosystem
445
functioning and will be pivotal in understanding how non-consumptive predator effects under
446
global warming will shape prey populations.
ACKNOWLEDGMENTS
447
448
We thank Lin Op de Beeck and Ria Van Houdt who assisted during the experiment. Sara
449
Debecker, Lin Op de Beeck, Nedim Tüzün and two anonymous reviewers provided valuable
450
feedback on the manuscript. LJ is a postdoctoral fellow of FWO-Flanders and benefited a
451
PDM fellowship of the KULeuven. Financial support came from the Belspo project SPEEDY,
452
KULeuven Excellence Center Financing PF/2010/07 and FWO research grant G.0943.15.
453
LITERATURE CITED
454
Abrams, P.A., and L. Rowe. 1996. The effects of predation on the age and seize of maturity of
455
456
457
458
459
460
461
prey. Evolution 50: 1052-1061.
Angert, A.L., S.L. LaDeau, and R.S. Ostfeld. 2013. Climate change and species interactions:
ways forward. Annals of the New York Academy of Sciences 1297:1-7.
Barton, B.T., and A.R. Ives. 2014. Direct and indirect effects of warming on aphids, their
predators and ant mutualists. Ecology 95:1479-1484.
Benard, M.F. 2004 Predator-induced phenotypic plasticity in organisms with complex life
histories. Annual Review of Ecology, Evolution and Systematics 35: 651-673.
19
462
Benjamini, Y., and Y. Hochberg. 1995. Controlling the false discovery rate: a practical and
463
powerful approach to multiple testing. Journal of the Royal Statistical Society, Series B 57:
464
289-300.
465
466
467
Blois, J.L., P.L. Zarnetske, M.C. Fitzpatrick, and S. Finnegan. 2013. Climate change and the
past, present and future of biotic interactions. Science 341: 499-5040.
Bradford, M.M. 1976. A rapid and sensitive method for the quantification of microgram
468
quantities of protein, utilizing the principle of protein-dye landing. Analytical
469
Biochemistry 72: 248-254.
470
Chen, M.M., H.B. Yin, P. ‘O Connor, Y.S. Wang, and Y.G. Zhu. 2010. C:N:P stoichiometry
471
and specific growth rate of clover colonized by arburuscal mycorrhizal fungi. Plant Soil
472
326: 21-29.
473
474
475
476
477
478
479
480
Chown, S.L., and K.J. Gaston. 2008. Macrophysiology for a changing world. Proceedings of
the Royal Society B Biological Sciences 275: 1469-1478.
Costello, D.M., and M.J. Michel. 2013. Predator-induced defenses in tadpoles confound body
stoichiometry predictions of the general stress paradigm. Ecology 94: 2229-2236.
Culler, L.E., M.A. McPeek, and M.P. Ayres. 2014. Predation risk shapes thermal physiology
of a predaceous damselfly. Oecologia 176: 653-660.
Dahl, J., and B. Peckarsky. 2002. Induced morphological defenses in the wild: predator
effects on a mayfly, Drunella coloradensis. Ecology 83: 1620-1634.
481
Dalton, C.M., and A.S. Flecker. 2014. Metabolic stoichiometry and the ecology of fear in
482
Trinidadian guppies: consequences for life histories and stream ecosystems. Oecologia
483
176: 691-701.
484
De Block, M., K. Pauwels, M. Van den Broeck, L. de Meester, and R. Stoks. 2013. Local
485
genetic adaptation generates latitude-specific effects of warming on predator-prey
486
interactions. Global Change Biology 19: 689-696.
20
487
488
De Block, M., and R. Stoks. 2004. Cannibalism-mediated life history plasticity to combined
time and food stress. Oikos 106: 587-597.
489
De Coen, W.M., and C.R. Janssen. 2003. The missing biomarker link: relationships between
490
effects on the cellular energy allocation biomarker of toxicant-stressed Daphnia magna
491
and corresponding population characteristics. Environmental Toxicology and Chemistry
492
22: 1632-1641.
493
Deutsch, C.A., J.J. Tewksbury, R.B. Huey, K.S. Sheldon, C.K. Ghalambor, D.C. Haak, and
494
P.R. Martin. 2008. Impacts of climate warming on terrestrial ectotherms across latitude.
495
Proceedings of the National Academy of Sciences USA 105: 6668-6672.
496
497
Elser, J.J., D.R. Dobberfuhl, N.A. MacKay, and J.H. Schampel. 1996. Organism size, life
history and N:P stoichiometry. Bioscience 46: 674-684.
498
Elser, J.J., T. Watts, B. Bitler, and T.A. Markow. 2006. Ontogenetic coupling of growth rate
499
with RNA and P contents in five species of Drosophila. Functional Ecology 20: 846-856.
500
Gillooly, J.F., J.H. Brown, G.B. West, V.M. Savage, and E.L. Charnov. 2001. Effect of size
501
502
503
and temperature on metabolic rate. Science 293: 2248-2251.
Hawlena, D., and O.J. Schmitz. 2010a. Physiological stress as a fundamental mechanism
linking predation to ecosystem functioning. American Naturalist 176: 537-556.
504
Hawlena, D., and O.J. Schmitz. 2010b. Herbivore physiological response to predation risk and
505
implications for ecosystem nutrient dynamics. Proceedings of the National Academy of
506
Sciences USA 107: 15503-15507.
507
508
509
510
Hawlena, D., M.S. Strickland, M.A. Bradford, O.J. Schmitz. 2012. Fear of predation slows
plant-litter decomposition. Science 336: 1434-1438.
Intergovernmental Panel on Climate Change. 2013. Climate change 2013: the physical
science basis. Cambridge University Press, Cambridge, UK and New York, USA.
21
511
512
513
Janssens, L., and R. Stoks. 2014. Reinforcing effects of non-pathogenic bacteria and predation
risk: from physiology to life history. Oecologia 176: 323-332.
Karasov, W.H., and C. Martinez del Rio. 2008. Ecological stoichiometry. In: Physiological
514
ecology: how animals process energy, nutrients and toxins. Princeton University Press,
515
Princeton, USA.
516
517
Karl, I., and K. Fischer. 2008. Why get big in the cold? Towards a solution to a life-history
puzzle. Oecologia 155: 215-225.
518
Kratina, P., H.S. Greig, P.L. Thompson, T.S.A. Carvalho-Pereisa, and J.B. Shurin. 2012
519
Warming modifies trophic cascades and eutrophication in experimental freshwater
520
communities. Ecology 93: 1421-1430.
521
Kuehne, L.M., J.D. Olden, and J.J. Duda. 2012. Costs of living for juvenile Chinook salmon
522
(Oncorhynchus tshawytscha) in an increasing warming and invading world. Canadian
523
Journal of Fisheries and Aquatic Sciences 69: 1621-1630.
524
525
Laurila, A., B. Lindgren, and A.T. Laugen. 2008. Defenses along a latitudinal gradient in
Rana temporaria. Ecology 89: 1399-1413.
526
Lemoine, N.P., and D.E. Burkepile. 2012. Temperature-induced mismatches between
527
consumption and metabolism reduce consumer fitness. Ecology 93: 2483–2489.
528
Liess, A., O. Rowe, J. Gou, G. Thomsson, and M.I. Lind. 2013. Hot tadpoles from cold
529
environments need more nutrients – life history and stoichiometry reflects latitudinal
530
adaptation. Journal of Animal Ecology 82: 1316-1325.
531
532
533
McPeek, M.A. 2004. The growth/predation risk trade-off: so what is the mechanism?
American Naturalist 163: E88-E111.
McPeek, M.A., M. Grace, and J.M.L. Richardson. 2001. Physiological and behavioral
534
responses to predators shape the growth/predation risk trade-off in damselflies. Ecology
535
82: 1535-1545.
22
536
Miller, L.P., C.M. Matassa, and G.C. Trussell. 2014. Climate change enhances the negative
537
effects of predation risk on an intermediate consumer. Global Change Biology 20: 3834-
538
3844.
539
Nilsson-Örtman, V., R. Stoks, M. De Block, F. Johansson. 2012. Generalists and specialists
540
along a latitudinal transect: patterns of thermal adaptation in six species of damselflies.
541
Ecology 93:1340-1352.
542
O’Gorman, E.J., D.E., Pichler, G. Adams, et al. 2012. Impacts of warming on the structure
543
and functioning of aquatic communities: individual- to ecosystem-level responses.
544
Advances in Ecological Research 47: 81-176.
545
Peacor, S.D., K.L. Pangle, L. Schiesari, and E.E. Werner. 2012. Scaling-up anti-predator
546
phenotypic responses of prey: impacts over multiple generations in a complex aquatic
547
community. Proceedings of the Royal Society B-Biological Sciences 279: 122-128.
548
549
550
Preisser, E.L., D.I. Bolnick, and M.F. Benard. 2005. Scared to death? The effects of
intimidation and consumption in predator-prey interactions. Ecology 86: 501-509.
Rabus, M., T. Sollradl, H. Clausen-Schaumann, and C. Laforsch. 2013. Uncovering
551
ultrastructural defences in Daphnia magna - an interdisciplinary approach to assess the
552
predator-induced fortification of the carapace. PLoS One 8: e67856.
553
Reef, R., M.C. Ball, I.C. Feller, and C.E. Lovelock. 2010. Relationships among RNA:DNA
554
ratio, growth and elemental stoichiometry in mangrove tress. Functional Ecology 24: 1064-
555
1072.
556
557
558
559
Schmitz, O.J. 2013. Global climate change and the evolutionary ecology of ecosystem
functioning. Annals of the New York Academy of Sciences 1297: 61-72.
Sibly, R.M., J.H. Brown, and A. Kodric-Brown. 2012. Metabolic ecology: a scaling approach.
Wiley & Blackwell, Chichester, UK.
23
560
561
562
563
564
565
566
567
Slos, S., and R. Stoks. 2008. Predation risk induces stress proteins and reduces antioxidant
defense. Functional Ecology 22: 637-642.
Sterner, R.W. and J.J. Elser. 2002. Ecological stoichiometry: the biology of elements from
molecules to the biosphere. Princeton University Press, Princeton.
Stoks, R., M. De Block, and M.A. McPeek. 2005b. Alternative growth and energy storage
responses to mortality threats in damselflies. Ecology Letters 8: 1307-1316.
Stoks, R., M. De Block, and M.A. McPeek. 2006. Physiological costs of compensatory
growth in a damselfly. Ecology 87: 1566-1574.
568
Stoks, R., M. De Block, F. Van de Meutter, and F. Johansson. 2005c. Predation cost of rapid
569
growth: behavioural coupling and physiological decoupling. Journal of Animal Ecology
570
74: 708-715.
571
Stoks, R., J.L. Nystrom, M.L. May, and M.A. McPeek. 2005a. Parallel evolution in ecological
572
and reproductive traits to produce cryptic damselfly species across the Holarctic. Evolution
573
59: 1976-1988.
574
Stoks, R., I. Swillen, and M. De Block. 2012. Behaviour and physiology shape the growth
575
accelerations associated with predation risk, high temperatures and southern latitudes in
576
Ischnura damselfly larvae. Journal of Animal Ecology 81: 1034-1040.
577
Thaler, J.S., S.H. McArt, and I. Kaplan. 2012. Compensatory mechanisms for ameliorating
578
the fundamental trade-off between predator avoidance and foraging. Proceedings of the
579
National Academy of Sciences USA 109: 12075-12080.
580
Touchon, J.C., and K.M. Warkentin. 2011. Thermally contingent plasticity: temperature alters
581
expression of predator-induced color and morphology in a Neotropical treefrog tadpole.
582
Journal of Animal Ecology 80: 79-88.
24
583
Traill, L.W., M.L.M. Lim, N.S. Sodhi, and C.J.A. Bradshaw. 2010. Mechanisms driving
584
change: altered species interactions and ecosystem function through global warming.
585
Journal of Animal Ecology 79: 937-947.
586
587
588
Trussell, G.C., P.J. Ewanchuk, and C.M. Matassa. 2006. The fear of being eaten reduces
energy transfer in a simple food chain. Ecology 87: 2979-984.
Van Moorleghem, C., N. De Schutter, E. Smolders, and R. Merckx. 2013. Bioavailability of
589
organic phosphorus to Pseudokirchneriella subcapitata as affected by phosphorus
590
starvation: an isotope dilution study. Water Research 47: 3047-3056.
591
Vrede, T., J. Persson, and G. Aronsen. 2002. The influence of food quality (P:C ratio) on
592
RNA:DNA ratio and somatic growth of Daphnia. Limnology and Oceanography 47: 487-
593
494.
594
Wang, T., C.C.Y. Hung, and D.J. Randall. 2006. The comparative physiology of food
595
deprivation: from feast to famine. Annual Review of Physiology 68: 223-251.
596
Watts, T., H.A. Woods, S. Hargand, J.J. Elser, and T.A. Markow. 2006. Biological
597
stoichiometry of growth in Drosophila melanogaster. Journal of Insect Physiology 52:
598
187-193.
599
Weider, L.J., K.L. Glenn, M. Kyle, and J.J. Elser. 2004. Associations among ribosomal
600
(r)DNA intergenic spacer length, growth rate and C:N:P stoichiometry in the genus
601
Daphnia. Limnology and Oceanography 49: 1417-1423.
602
603
604
605
SUPPLEMENTAL MATERIAL
Appendix A: Comparison of the values of physiological variables with literature data
606
25
607
TABLE 1. Results of ANOVAs testing for the effects of rearing temperature and predation risk
608
on life history and physiology in the damselfly Enallagma cyathigerum.
Temperature (T)
Predation risk (P)
TxP
df
F
p
df
F
p
df
Growth
rate
1, 176
2.24
0.14
1, 176
76.99
< 0.001
1,
176
15.02 < 0.001
RNA:DNA
1, 76
10.65
0.0017
1, 76
37.31
< 0.001
1, 76
5.45
0.022
ETS
1, 36
59.76
< 0.001
1, 36
57.6
< 0.001
1, 36
7.50
0.0097
Energy
storage
Fat
1, 56
16.87
< 0.001
1, 56
32.08
< 0.001
1, 56
6.62
0.013
Sugars
1, 56
7.92
0.0068
1, 56
120.87
< 0.001
1, 56
12.98 < 0.001
Protein
1, 56
8.40
0.0054
1, 56
0.12
0.73
1, 56
0.39
0.54
Body
stoichiometry
C:N
1, 56
15.24
< 0.001
1, 56
56.34
< 0.001
1, 56
5.91
0.018
C:P
1, 56
3.11
0.062
1, 56
27.66
< 0.001
1, 56
4.59
0.034
N:P
1, 56
0.79
0.52
1, 56
5.15
0.029
1, 56
0.11
0.74
609
26
F
p
FIGURE LEGENDS
610
611
FIG. 1. Mean (A) growth rate, (B) RNA:DNA ratios and (C) oxygen consumption (ETS
612
activity) of E. cyathigerum larvae as a function of rearing temperature and predation risk
613
exposure. Given are least-squares means + 1 SE. Means having different letters were
614
significantly different (based on linear contrasts corrected for multiple testing using the false
615
discovery rate procedure).
616
617
FIG. 2. Mean levels of energy storage molecules; (A) fat content, (B) sugar content and (C)
618
protein content of E. cyathigerum larvae as a function of rearing temperature and predation
619
risk exposure. Given are least-squares means + 1 SE. Means having different letters were
620
significantly different (based on linear contrasts corrected for multiple testing using the false
621
discovery rate procedure).
622
623
FIG. 3. Mean (A) C:N ratios, (B) C:P ratios and (C) N:P ratios of E. cyathigerum larvae as a
624
function of rearing temperature and predation risk exposure. Given are least-squares means +
625
1 SE. Means having different letters were significantly different (based on linear contrasts
626
corrected for multiple testing using the false discovery rate procedure).
27
627
Fig. 1
Growth rate (/d)
A
Predation risk absent
Predation risk present
0.016
b
a
0.012
c
d
0.008
0.004
B
7
b
RNA:DNA ratio
a
5
c
c
3
1
ETS (nmol O2/min)
C
0.30
c
0.25
b
b
0.20
a
0.15
0.10
20 °C
628
24 °C
Temperature
28
629
Fig. 2
Predation risk absent
Predation risk present
A
Fat content (µg/mg dry mass)
30
b
25
a
20
c
c
15
10
Sugar content (µg/mg dry mass)
B
16
12
b
a
c
c
8
4
0
Protein content (µg/mg dry mass)
C
800
b
b
700
a
600
a
500
20 °C
630
24 °C
Temperature
631
29
632
Fig. 3
Predation risk absent
Predation risk present
A
5.2
c
c
4.8
C:N ratio
a
4.4
b
4.0
B
110
a
C:P ratio
105
100
a
a
95
b
90
85
C
26
b
b
N:P ratio
24
22
20
a
a
18
16
20 °C
633
24 °C
Temperature
30