Canadian Journal of Fisheries and Aquatic Sciences
Thermal dependence of size-at-hatch in the lake whitefish (Coregonus clupeaformis)
Journal: Canadian Journal of Fisheries and Aquatic Sciences
Manuscript ID cjfas-2018-0097.R2
Manuscript Type: Article
Date Submitted by the 12-Dec-2018 Author:
Complete List of Authors: Mitz, Charles; McMaster University, Medical Physics Thome, Christopher; Northern Ontario School of Medicine Cybulski, Mary Ellen; Northern Ontario School of Medicine Somers, Christopher; University of Regina, Biology Manzon, Richard;Draft University of Regina, Biology Wilson, Joanna; McMaster University Boreham, Douglas; McMaster University, Medical Physics; Northern Ontario School of Medicine, Medical Sciences
MODELS < General, PHYSIOLOGY < General, TEMPERATURE EFFECTS < Keyword: General, GROWTH < General, FRESHWATER FISHES < General
Is the invited manuscript for consideration in a Special Not applicable (regular submission) Issue? :
https://mc06.manuscriptcentral.com/cjfas-pubs Page 1 of 43 Canadian Journal of Fisheries and Aquatic Sciences
1 Thermal dependence of size-at-hatch in the lake whitefish
2 (Coregonus clupeaformis)
3 Charles Mitz1*, Christopher Thome1,5, Mary Ellen Cybulski1, Christopher M. Somers3,
4 Richard G. Manzon3, Joanna Y. Wilson4 and Douglas R. Boreham1,2,5
5 Abstract
6 Lake whitefish embryos incubated at low temperatures have a longer incubation period and hatch 7 at a significantly greater size than those incubated at warmer temperatures. We examined hatch 8 timing and morphological characteristics for whitefish embryos reared under different constant 9 and varying temperatures to determine whether the thermal dependence of hatching size reflects 10 differences in their development stage. OurDraft results show that lake whitefish embryos hatch at 11 different temperature-dependent developmental stages, and this is the dominant factor affecting 12 size-at-hatch. The term “heterograde” is proposed for the thermal dependence of hatching stage 13 to differentiate it from hatching that occurs at a fixed developmental stage. A method to quantify 14 this effect is given using a ratio that describes the difference in relative development at hatching 15 between different viable constant incubation temperatures. Heterograde hatching is proposed as 16 possible mechanism to synchronize the timing of hatch to the break-up of winter ice cover 17 despite variability in the date of spawning and in the onset of spring break-up.
1Department of Medical Physics and Applied Radiation Sciences, McMaster University 2Bruce Power, Tiverton, ON, Canada 3Department of Biology, University of Regina 4Department of Biology, McMaster University 5Northern Ontario School of Medicine *corresponding author
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 2 of 43
18 Introduction
19 The size of newly hatched lake whitefish (Coregonus clupeaformis) embryos is strongly
20 temperature-dependent. Embryos incubated at low temperatures hatch at a larger size than those
21 incubated at warmer temperatures (Hart 1931; Price 1940; Colby and Brooke 1973; Brooke
22 1975; Griffiths 1979; Luczynski 1984; Luczynski & Kirklewska 1984; Eme et al. 2015, Mueller
23 et al. 2015; Lee et al. 2016, and Lim et al. 2017). The question remains whether size-at-hatch is
24 largely driven by differences in the thermal scaling of growth and development, or in
25 development stage at the time of hatching.
26 Differences in the thermal scaling of growth and development have been proposed as a
27 mechanism governing adult size in ectothermsDraft (Sibly and Atkinson 1994; van der Have and
28 Dejong 1996, and Forster and Hirst 2012), and the mechanism underlying the temperature-size-
29 rule in general (Zuo et al. 2012). The larger size-at-hatch produced by lower incubation
30 temperature may result from differences in the thermal scaling of growth and development or
31 from temperature-dependent differences in growth efficiency (Perrin 1995; Atkinson and Silby
32 1997; Angilletta and Dunham 2003). Alternatively, the greater size-at-hatch may simply reflect
33 lake whitefish embryos hatching at a more advanced stage of development at lower temperatures
34 (Luczynski and Kolman 1987; Jordaan 2002; Jordaan et al. 2006).
35 Hatching is a life stage transition that may occur over a range of development stages (Spicer and
36 Burggen 2003; Touchon et al. 2015) as an active response to an environmental signal/cue (e.g.
37 Martin et al. 2011; Warkentin 2011), or passively (Schulte et al. 2011) as a function of
38 temperature. This latter possibility need not violate the rule of equiproportional development
39 which states that the duration of development stages occupies constant proportion of the
https://mc06.manuscriptcentral.com/cjfas-pubs Page 3 of 43 Canadian Journal of Fisheries and Aquatic Sciences
40 embryonic development across temperatures (Peterson 2001; Jordaan et al. 2006). The focus of
41 this study is to assess whether developmental stage at hatch differs in whitefish reared under
42 different temperatures and assess whether differences in the developmental stage at hatch can
43 account for the thermal dependence of size-at-hatch.
44 We propose the term heterograde for hatching that occurs at different temperature-dependent
45 development stages, while the term homograde would be used for hatching that occurs at a fixed
46 development stage, independent of incubation temperature. The use of specific descriptive terms
47 is useful to differentiate the thermal dependence of stage-at-hatch from other forms of
48 heterochrony such as hypermorphosis (Fink 1982) that act on an evolutionary time-scale, and 49 from phenomena in which hatching mayDraft be triggered by an active response to an environmental 50 cue.
51 Both heterograde hatching and differences in the thermal scaling of growth and development are
52 able to account for a thermal dependence hatching size, but the two hypotheses yield different
53 predictions for hatch timing. The predicted hatch timing may be clarified by describing
54 embryonic growth in terms of relative development (RD) which, for a constant temperature, is
55 typically calculated as the cumulative incubation time divided by the time to median hatch.
56 Heterograde hatching implies an instantaneous change in RD accompanying a change in
57 incubation temperature (Figure 1). Thus, for any pool of embryos (e.g. a clutch or group of
58 embryos spawned at the same time and place), an increase in temperature at a sufficiently
59 advanced stage of development could raise RD beyond 100% and trigger immediate hatching.
60 Conversely, a downward shift in incubation temperature would lead to an instantaneous decrease
61 in RD and a delay in the hatching of the remaining embryos, widening the hatch window. Under
62 the alternate hypothesis (i.e. differences between the thermal scaling of growth and
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 4 of 43
63 development), RD is continuous and monotonic (i.e. it must continually increase at a greater or
64 lesser rate but can never reverse itself and decline).
65 In this paper we use embryo morphology, hatch timing under constant and varying (fluctuating
66 and asymmetric) thermal regimes, and apparent changes in the thermal scaling of hatch timing
67 following temperature shifts to differentiate between heterograde hatching and alternate size-at-
68 hatch hypotheses. We include fin morphology, a recognized means to developmentally stage
69 larval fish, to provide further evidence that development stage at hatch differs. Analysis of yolk
70 conversion efficiency is used to account for temperature-dependent differences in growth
71 efficiency.
72 Methods and Materials Draft
73 Egg collection and Incubation Procedures
74 Whitefish eggs were obtained from spawning Lake Huron fish gillnetted on November 21, 2013
75 in 3 to 6 m of ~7°C water south of Whitefish Island (44°42’37.74”N, 81°18’38.94”W). Eggs
76 were stripped from multiple ripe females (mF) and fertilized with pooled milt from ripe males
77 followed by disinfection using Ovadine® (5 ml L-1, Syndel Laboratories Ltd.). In 2014,
78 additional eggs were obtained from the Ontario Ministry of Natural Resources and Forestry
79 (MNRF) from a single female (sF) from Lake Simcoe fertilized by pooled milt from different
80 males.
81 The fertilized eggs were transported in an ice bath to McMaster University where they were
82 reared in refrigerated incubators using apparatus and methods described in Mitz et al. (2014).
83 Incubation temperatures were held at nominal temperatures of 2, 5 and 8°C with an additional
https://mc06.manuscriptcentral.com/cjfas-pubs Page 5 of 43 Canadian Journal of Fisheries and Aquatic Sciences
84 0.5°C (nominal) temperature point obtained by incubating dishes on an unsaturated bed of
85 crushed ice. Photoperiod was ambient with natural lighting through the laboratory windows and
86 was consistent across all treatments. Data loggers (Schlumberger Mini-Diver and Onset Hobo)
87 were placed at multiple locations within the incubator units and set to record water temperature
88 every 5 minutes.
89 The embryos were reared in multiple 24-well plates (one embryo per well) filled with
90 dechlorinated municipal tap water and numbered (to ensure that measurements were made
91 without knowing which treatment group was being measured). Generally, two plates were used
92 for each constant incubation temperature with four plates used for the 0.5 oC constant 93 temperature incubation to offset elevatedDraft mortality previously observed at this temperature. 94 Embryos were also reared in multiple un-numbered 24-well plates (approximately 60) to provide
95 a pool of embryos for measurements of unpreserved dry weight and standard length at different
96 stages of development, and for experiments incorporating a temperature shift at different discrete
97 stages of development. Embryos were moved to a clean multi-well plate during weekly water
98 changes to minimize fungal contamination. Viable hatches were recorded, and dead embryos
99 removed daily. Embryo mortalities were recognized by the opacity or disintegration of the
100 embryo.
101 Morphological indicators of growth and development
102 Hatched larvae were euthanized using a solution of approximately 0.5% MS-222 (ethyl 3-
103 aminobenzoate methanesulfonate, Sigma Aldrich) prior to being imaged using an Axio Zoom
104 V16 microscope (Carl Zeiss) equipped with a Canon EOS Rebel T1i camera. Dimensional
105 measurements (Figure 2) were made to the nearest 1 μm using Zeiss Axiovison (Rel 4.9.1)
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 6 of 43
106 digital image processing software. Following imaging, the embryos were placed intact on
107 aluminum foil tares and dried overnight to an asymptotic constant weight at a temperature of
108 70oC in a VWR Brand Model 1500EM drying oven. Dried samples were kept on a bed of
109 anhydrous silica gel prior to weighing (Mettler-Toledo XA105DU, ±0.01 mg). Separate pools
110 (10 embryos per pool) of yolked and deyolked embryos were used to determine an
111 experimentally-derived (see Results) relationship between standard length (SL) and somatic dry
112 weight (weight of deyolked embryo) at six different times during development (covering the
113 range of approximately 30% to 100% RD). This relationship was used to estimate the dry weight
114 of individual larvae based on the measured length. The yolk dry weight was calculated by
115 subtracting the measured somatic dry weight from the total embryo dry weight (weight of
116 embryo including intact yolk). All measurementsDraft were taken using unpreserved embryos to
117 avoid confounding effects resulting from chemical preservation (Smith and Walker 2003; Melo
118 et al. 2010; Sreetharan et al. 2015).
119 Since existing staging tables (Price 1940; Colby and Brooke 1973; Sreetharan et al. 2015) lack
120 readily identifiable morphological stage indicators late in development, we used dorsal fin
121 indentation as an indication of development stage near hatch. Dorsal fin indentation increases
122 progressively in lake whitefish with complete separation of the dorsal and adipose fins occurring
123 three to five weeks post-hatch under natural thermal regimes in the Bay of Quinte, Canada (Hart
124 1931). Fin definition (i.e. indentation) has been successfully used as a means of continuous
125 staging in post-hatching zebrafish (Parichy et al. 2009). Our calculated fin indentation ratio and
126 additional morphological measurements are shown on Figure 2. The area of the yolk sac and
127 yolk oil globule were calculated from the measured parameters assuming an ellipsoidal shape.
https://mc06.manuscriptcentral.com/cjfas-pubs Page 7 of 43 Canadian Journal of Fisheries and Aquatic Sciences
128 Hatch Timing
129 Embryos from the two populations (Huron and Simcoe) were incubated under a variety of
130 constant and asymmetrically varying thermal regimes. Asymmetrically variable regimes included
131 progressively increasing or decreasing regimes where initial development took place at either 2
132 or 8oC for approximately 1/3rd of expected RD, followed by transfer to progressively higher
133 (→5→8°C) or lower (→5→2°C) temperatures for the remaining thirds of expected RD.
134 Asymmetrically varying regimes also included initial incubation at 2oC for different lengths of
135 time (81, 93, 106, 121, 138, and 151 days) before transfer to 5 or 8oC for their remaining
136 development. The experimental thermal regimes are summarized on Table 1. Development time 137 was determined using the median hatchDraft time-to-hatch for a given batch of eggs. The duration of 138 hatching was taken as the difference between 10% and 90% hatching to minimize the
139 confounding effect of outliers and differences in sample size.
140 Q10 following a shift in incubation temperature
141 A quantitative estimate of the effect of heterograde hatching can be derived using the Q10
o 142 measure which describes changes in growth rate, R, across a 10 C change in temperature:
10 R (T2 T1 ) 143 2 [1] Q10 R1
144 where R1 and R2 are the inverse of the time from the temperature shift to the timing of mean
145 hatch. A one-time shift from a lower base temperature (T1 ) to higher incubation temperature
146 occurs at a specific point in RD, which we denote f 0 . We may quantify heterograde hatching
T 147 using the term f het , which we define as the difference in RD for a post-shift temperature, 2 ,
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 8 of 43
148 compared to T1 (e.g. fhet 0.1 means that median hatch th at T2 occurs at an RD equivalent to
149 90% of t at T ). For T , R is proportional to (1 f )1 or f 1 but for T , R is proportional h 1 1 1 0 0 2 2
1 150 to f 0 f het . If we substitute these values into Equation 1 we see that the post-shift Q10
151 remains unchanged where hatching is homograde (f het 0 ) but if fhet 0 the post-shift Q10
152 increases with f 0 according to:
10 T T f0 2 1 153 Post-shiftQ10 = Pre-shiftQ10 [2] T1T2 f0 fhet
154 Thus, heterograde hatching is predicted to result in an apparent post-shift value for Q10 (from a
155 low base temperature to a higher post-shiftDraft temperature) that steadily increases as the time of the
156 temperature shift approaches the time of hatch. Alternate size-at-hatch hypotheses predict no
157 significant change.
158 The value for f het may be related to the apparent thermal scaling (i.e. Q10) of development and
159 the thermal scaling for hatching under constant temperatures according to:
T2 T1 development T1T2 Q ) 10 f 1 10 160 het hatching [3] Q10
161 The value for fhet over a 10 degree change in temperature is then:
Development Q10 162 fhet 1 Hatching [4] Q10
https://mc06.manuscriptcentral.com/cjfas-pubs Page 9 of 43 Canadian Journal of Fisheries and Aquatic Sciences
163
164 We determined post-shift Q10 values at six points in development (81, 93, 106, 121, 138, 151
165 dpf) beginning at the eyed stage (Stage 9 after Sreetharan et al. 2015) and continuing to pre-
166 hatching for shifts from a base temperature of 2oC to 5 and 8oC respectively (Figure 3).
167 Thermal Dependence of Yolk Conversion Efficiency (YCE)
168 If hatching is heterograde, then conventional determinations of YCE in newly hatched embryos
169 (e.g. Rombough 1988) may reflect differences in development stage rather than actual
170 differences in yolk conversion efficiency owing to the increasing proportion of energy devoted to 171 somatic maintenance as body mass Draft increases. We corrected for this potential effect by 172 normalizing YCE (see supplementary material) using the relationship between somatic dry
173 weight (WS ) and yolk dry weight (Wy ) according to an empirically-derived hyperbolic function:
Wy 174 WS [5] (1 aWy )
175 where, is interpreted as a limiting efficiency1 of yolk conversion (0.87) as the maintenance
176 allocation approaches zero, and a is experimentally determined and related to YCE. If we know
177 WS and Wy then we can substitute their values into Equation 5 to solve for a and use this value
178 to normalize YCE to the point of theoretical yolk depletion. We emphasize that Equation 3
179 cannot be used at the point of actual yolk depletion since embryos begin to lose dry mass before
180 yolk consumption is complete (Blaxter and Hempel 1963; Meyer et al. 2012). We have also
181 assumed that the limiting efficiency of yolk conversion, , is unaffected by temperature and
1 Ignoring any differences in the specific energy content of the yolk and somatic tissue.
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 10 of 43
182 therefore a constant. If is temperature-dependent, then both a and will vary, and
183 normalized YCE will not be a function of a alone. This possibility does not meaningfully affect
184 the normalization of YCE over the range of incubation temperatures and values for WS
185 considered in this paper but normalization over large differences in temperature and WS should
186 be treated with caution.
187 Statistical Analysis
188 Statistical analysis was conducted using Sigmaplot V11.0. Normalized YCE differences
189 between temperature groups were analyzed using a one-way analysis of variance (ANOVA),
190 while differences in dry weight vs SL for different temperatures were linearized using an
191 exponential transformation followed byDraft a one-way ANOVA. The time to 10, 50, and 90% hatch
192 and the time to median hatch were determined after pooling the results of replicate dishes (two
193 per experimental thermal regime). Linear and non-linear regressions were solved by means of the
194 least-squares method. Statistical differences between related curves were assessed by using a
195 Johnson-Neyman ANCOVA with significance among pairwise slopes and elevations tested
196 following a Bonferroni adjustment. Differences were considered statistically significant at P≤
197 0.05.
198 Results
199 Morphological indicators of growth and development
200 Constant temperature incubation
https://mc06.manuscriptcentral.com/cjfas-pubs Page 11 of 43 Canadian Journal of Fisheries and Aquatic Sciences
201 Viable hatches were recorded at all temperatures; however, survival was poor at 0.5oC (9% for
202 mF Lake Huron and 0% for sF Lake Simcoe) and 8oC constant temperature treatments (9% for
203 the mF Lake Huron and 7% for sF Lake Simcoe). Higher survival ranging from 18 to 58% was
204 observed at other constant temperatures. We did not exclude data from groups where survival
205 was poor and included the size-at hatch and YCE for all treatments for completeness; poor
206 survival came from incubations at temperature extremes. Survival data is included in Table 2.
207 Larval standard length (SL) and dry weight at hatch were affected by incubation temperature. At
208 2oC, the hatchlings were significantly longer and heavier (P<0.01) than those reared at 5 and 8oC,
209 with the 2oC hatchlings more than 50% heavier than those hatching at 8oC. This difference was 210 similar for both the sF-Simcoe and mF-HuronDraft populations, although standard deviations were 211 greater with the mF-Huron trials. An inverse correlation was found between yolk area and
212 hatching temperature (P<0.001), with the 8oC sF trial having a larger yolk area (5.5±0.174 mm2)
213 than the 5o and 2oC sF trials (3.5±0.160 and 1.5±0.088 mm2 respectively). Oil globule area
214 followed a similar relationship (1.20±0.061, 0.90±0.031 and 0.66±0.023 mm2 for the 8, 5, and 2
215 oC trials).
216 Size-at-hatch (length and weight) varied depending on the relative timing of hatch (i.e. rank) for
217 the 5oC sF trial with late hatching larvae both longer (P=0.10) and heavier (somatic dry weight,
218 P=0.12) than the early-hatching larvae. Larval height (H) increased significantly with hatching
219 rank for the 5oC sF trial (P=0.006). The 2oC sF embryos exhibited a significant change in SL
220 with hatching rank (P=0.007), but no significant change in height was observed (P=0.29). The
221 8oC sF trial showed no significant difference in size between early and late hatching larvae for
222 either length, dry weight, or height.
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 12 of 43
223 Yolk reserves varied with hatching rank as indicated in the analysis of yolk area for both the 5oC
224 (P=<0.001) and 2oC (P=<0.001) sF trials (Figure 4B). Oil globule area also varied with hatching
225 rank for the 5 and 2oC constant temperature incubations. No relationship was observed with
226 hatching rank for either yolk or oil globule area for the 8oC trial for which hatching occurred
227 over a short time interval. Total embryo dry weight at hatching varied with incubation
228 temperature for the sF trials with mean weights of 1.822±0.015, 1.930±0.024, and 2.066±0.017
229 obtained for 2, 5, and 8oC incubations respectively.
230 Yolk reserves at hatch displayed an inverse relationship to temperature at the time of hatching,
231 with the larger embryos hatching with smaller yolks and smaller lipid globules. No significant 232 relationship existed between yolk area Draftand mean incubation temperature. A summary of hatch 233 timing, SL, yolk and oil globule areas, fin indentation ratio (F) and YCE is provided for different
234 tested regimes on Table 2.
235 Dorsal fins were more deeply indented (i.e. fin indentation ratio F decreased) for embryos
236 hatching at low temperatures. F values ranged from 0.79±0.05 for the 8oC sF trial to 0.66±0.06
237 and 0.57±0.05 for the 5 and 2oC sF incubations respectively (Figure 4C). Fin indentation also
238 increased with hatching rank within the 5oC sF trial (P=<0.001) but no significant difference
239 between early and late-hatching embryos was observed for the 2 and 8oC sF trials. The
240 relationship between the different morphological and dry weight measures and time for different
241 constant incubation temperatures is shown on Figure 4 A to C. The same relationships for
242 embryos reared under progressively increasing and decreasing temperatures is shown on Figure 4
243 D to F.
244 Variable temperature incubation
https://mc06.manuscriptcentral.com/cjfas-pubs Page 13 of 43 Canadian Journal of Fisheries and Aquatic Sciences
245 Embryos reared in an asymmetrically decreasing thermal regime hatched at 2oC with a longer
246 body (13.35±0.06 vs. 12.54±0.05 mm, P<0.001 for sF embryos from Lake Simcoe) and
247 significantly greater fin indentation ratio (0.65±0.081 vs. 0.79±0.041, P<0.001) than those reared
248 in an asymmetrically increasing thermal regime (Figure 4 D and F). Morphological indicators of
249 stage such as fin indentation and body size under asymmetrically increasing or decreasing
250 thermal regimes were similar to, but not identical to, those observed at their constant temperature
251 equivalents. Fin indentation was slightly but significantly greater for the increasing thermal
252 regime compared to the 8oC constant regime, and less for the decreasing regime than that
253 observed for larvae incubated at 2oC for their full development. Larvae hatching under an
254 increasing thermal regime were significantly (P<0.001) longer (mean SL 12.54±0.05) compared
255 to the 8oC constant temperature larvaeDraft (mean SL 12.12±0.10), perhaps reflecting a
256 proportionately greater fraction of development spent at a more energetically efficient
257 temperature. Similar relationships were apparent for the mF Lake Huron embryos.
258 Dorsal fins were more deeply indented (i.e. fin indentation ratio F decreased) for embryos
259 hatching at low temperatures. F values ranged from 0.79±0.05 for the 8oC sF trial to 0.66±0.06
260 and 0.57±0.05 for the 5 and 2oC sF incubations respectively (Figure 4C). Fin indentation also
261 increased with hatching rank within the 5oC sF trial (P=<0.001) but no significant difference
262 between early and late-hatching embryos was observed for the 2 and 8oC sF trials. The
263 relationship between the different morphological and dry weight measures and time for different
264 constant incubation temperatures is shown on Figure 4 A to C. The same relationships for
265 embryos reared under progressively increasing and decreasing temperatures is shown on Figure 4
266 D to F.
267 Hatch timing
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 14 of 43
268 The time to median hatch (th ) increased as an exponential function of constant temperature (T)
0.152T 2 0.163T 2 269 for both Lake Huron th 231e (R = 0.999, P < 0.001) and Lake Simcoe th 251e (R =
270 0.999, P < 0.001) embryos. Hatching duration was negatively associated with temperature
271 ranging from 21 to 22 days at 2oC to 7 to 13 days at 8oC (Table 2).
272 The time to median hatch differed markedly between the asymmetrically increasing thermal
273 regimes (102 and 104 days) and the asymmetrically decreasing regimes for which hatching
274 occurred nearly 70 days later (168 and 174) despite similar mean temperatures. The time to
275 median hatch for the asymmetrically variable incubation regimes deviated from expectations
276 based on the mean incubation temperature (e.g. Thome et al. 2016). Expressed in terms of 277 expected RD (if embryos hatched at Draft the same stage of development), the median hatch for 278 increasing thermal regimes occurred at 81 and 86% RD and at 119 and 117% RD for the
279 decreasing thermal regimes for mF Lake Huron and sF Lake Simcoe respectively. These
280 relationships persisted when the temperatures are corrected (Mitz et al. 2017) for the effect of
281 Jensen's inequality (Huey and Berrigan 2001; Martin and Huey 2008; Ragland and Kingsolver
282 2008). The duration of hatching in the asymmetrically decreasing thermal regimes ranged from
283 21 to 28 days, while embryos reared under asymmetrically increasing regimes hatched in only 2
284 to 3 days.
285 Yolk Conversion Efficiency (YCE)
286 The dry weight of 124 eggs from Lake Huron females (mF) averaged 2.763±0.234 mg. The
287 variance in initial egg weight was much lower for a 90 egg sample from the single female from
288 Lake Simcoe (sF) which had a mean dry weight of 2.735±0.081 mg. The dry weight of the sF
https://mc06.manuscriptcentral.com/cjfas-pubs Page 15 of 43 Canadian Journal of Fisheries and Aquatic Sciences
289 eggs ex chorion was determined by subtracting the mean chorion dry weight (0.316±0.00424
290 mg) from the total egg dry weight.
291 Somatic dry weight increased exponentially with increasing SL (Figure 5). No significant
292 difference existed between the 2 and 5oC trials during late development when hatching was
293 possible; however, embryos reared at 8oC were significantly lighter for a given SL (P<0.001). No
294 significant difference in the relationship between SL and dry weight was observed for the
295 different variable thermal regimes (data not shown).
296 Somatic dry weight increased linearly with time for the 2 and 5oC treatments. High mortality and
297 short development at 8oC prevented an estimation of pre-hatching dry weight increase. The 298 relationship between somatic dry weightDraft and time was found to be approximately linear for both 299 the 5 and 2oC trials over the portion of development for which reliable dry weight measurements
300 can be practically obtained (i.e. a condition for YCE normalization as described in
301 supplementary material).
302 Normalized YCE was significantly different (Figure 6) for embryos reared under different
303 constant temperatures (P <0.001). The yolk conversion ratio (dry weight basis) decreased with
304 increasing incubation temperatures with the YCE for the 8oC trials significantly lower than that
305 for embryos reared at 5 and 2oC (P<0.001). YCE was highest for the 2oC embryos but the
306 difference between normalized YCE for the 2 and 5 oC trials was non-significant (P=0.169). It is
307 noteworthy that the nominal YCE values show the reverse relationship with the 8oC embryos
308 having the highest nominal YCE although the differences are non-significant.
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 16 of 43
309 Calculated Q10 following a one-time increase in Incubation Temperature
310 Q10 values calculated post shift increased with decreasing interval between shift and hatching as
311 shown in Figure 7. The increase in the nominal post-shift Q10 to triple-digit values is sufficient
312 to reject the hypothesis of an unchanging Q10 that would result if hatching were homograde. The
313 relationship between post-shift Q10 values and the shift timing appeared to follow a hyperbolic
314 relationship approximating that predicted under the heterograde hypothesis (R2=0.36,
315 P=0.04).The reduction in incubation time (relative to the homograde expectation) following an
316 increase in incubation temperature may be illustrated by plotting the percentage of calculated RD
occurring pre- and post-shift. This allows f to be determined from the intercept of pre- and 317 het 318 post-shift RD (Figure 8). normalized forDraft a 10oC temperature difference using Equation 3.
319 Discussion
320 We observed that lake whitefish embryos incubated at low temperatures hatch at larger size but
321 with lower yolk reserves than those reared at higher temperatures. Morphological indicators of
322 development stage (i.e. fin indentation) suggest that the larger embryos hatched at a more
323 advanced development stage than the smaller embryos reared in warmer temperatures. Embryos
324 reared under an asymmetrically increasing regime hatched earlier and at smaller body size than
325 expected based on mean incubation temperature, while embryos reared under an asymmetrically
326 decreasing regime hatched later and at a larger body size than predicted by mean incubation
327 temperature. Both these observations are consistent with a thermal dependence of hatching stage.
328 The calculation of post-shift Q10 values following an upward shift in temperature conformed to
329 the heterograde prediction with values exceeding 100 compared to a base value of approximately
330 5 that would be expected if hatching occurred at an identical stage of development. Observed
https://mc06.manuscriptcentral.com/cjfas-pubs Page 17 of 43 Canadian Journal of Fisheries and Aquatic Sciences
331 differences in yolk conversion efficiency are sufficient to account for only some 20 to 25% of the
332 observed size difference at hatching between embryos reared under different constant
333 temperatures. The observed thermal dependence of size-at-hatch cannot be accounted for by
334 differences in growth efficiency alone. Our results therefore support the hypothesis that the
335 larger size of low temperature hatchlings is the result of continued development within the
336 chorion to a more advanced stage (i.e. heterograde hatching).
337 The magnitude of the heterograde effect may be quantified using a measured ratio, fhet , for the
338 difference in hatching stage between the highest and lowest viable constant incubation
339 temperatures for a species. We have taken this range to be between approximately 8 and 0oC for 340 the lake whitefish based on this study andDraft work by others (Price, 1940; Brooke, 1975; Griffiths, o 341 1979). fhet may be normalized to a 10 C temperature difference and related to the thermal
342 scaling (i.e. Q10) of development and the thermal scaling for hatching under constant
343 temperatures:
T2 T1 development T1T2 Q ) 10 f 1 10 344 het hatching [6] Q10
345 The value for fhet over a 10 degree change in temperature is then:
Development Q10 346 fhet 1 Hatching [7] Q10
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 18 of 43
T1 T2 347 Since we know the value of fhet across a specific temperature range, Equation 6 may be used
Development 28 348 to solve for the thermal scaling of development (Q10 ). Our measured values fhet ≈ 0.24
25 Development 349 and fhet ≈ 0.20 (Figure 8), correspond to Q10 of 3.2 and 2.5 respectively.
350 While we lack reliable independent measurements of the thermal scaling of pre-hatching
351 development rates, such scaling may be approximated using the temperature dependence of heart
352 rate measured over different development stages. Eme et al. (2015) found that the thermal
353 scaling of heart rate in later stages of development (intermittent fin flutter or Colby and Brooke
354 Stage 19) followed a relatively consistent Q10 of approximately 3.The thermal scaling of median 355 hatch is tightly constrained to a narrowDraft range of values (Q10 values between 4.6 and 5.2) 356 depending on the population and experimental conditions. The difference between a Q10=5
357 (hatching) and Q10=3 (development) is sufficient to account for the differences in size-at-hatch
358 that we observed (Figure 7).
359 We found no evidence that the thermal scaling of absolute growth was significantly different
360 than the scaling of development based on measures such as the degree of fin indentation. The
361 observed decrease in the efficiency of growth with increasing temperature is consistent with the
362 findings of Mueller et al., (2015), and is inconsistent with the predictions derived from
363 differential scaling between growth and development (Zuo et al. 2012). However, a decrease in
364 growth efficiency with increasing temperature is consistent with predictions based on the von
365 Bertalanffy-Perrin growth efficiency model (Angilletta and Dunham 2003).
366 Although the biological mechanism underlying the observed thermal dependence of hatching
367 stage remains unknown, our results indicate that the dependence is not an active response to an
https://mc06.manuscriptcentral.com/cjfas-pubs Page 19 of 43 Canadian Journal of Fisheries and Aquatic Sciences
368 environmental signal/cue (e.g. Martin et al. 2011; Warkentin 2011), Embryos differing in their
369 thermal history but experiencing identical constant temperatures in late-stage development,
370 hatch at different relative development. An unchanging temperature can provide no active signal,
371 so the development stage at hatch must vary as a passive function of temperature (e.g. Schulte et
372 al. 2011).
373 The existence of heterograde hatching does not preclude the simultaneous existence of actively
374 plastic hatching mechanisms, or differences in thermal scaling between growth and development.
375 Indeed, the timing of hatching and size-at-hatch under real-world conditions is likely to reflect
376 the interaction of multiple factors which may exert a greater or lesser influence under different 377 environmental conditions. Czerkies etDraft al. (2001) noted that hypoxia triggered precocious 378 hatching in Coregonus lavaretus, while similar hatching stimulation has been recorded in
379 response to fungal infection (Wedekind 2002) and predation (Wedekind and Müller 2005).
380 Lake whitefish spawn over shallow cobble shoals in the fall (October and November) during a
381 period of several weeks before the formation of winter ice cover (Hart 1931; Scott and Crossman
382 1973; Anras 1999). The natural thermal regimes experienced by incubating whitefish embryos
383 are therefore asymmetrically variable, dropping from temperatures of between 5 and 8oC to near
384 zero where the temperature remains until spring break-up, the timing of which may vary
385 significantly with latitude and from year to year (Duguay et al. 2006). Following break-up, water
386 temperatures rise rapidly due to warming and the mixing of deep isothermal waters with those of
387 the shallow shoals where incubation takes place. Under these conditions, both early and late
388 spawned embryos may reach a stage of development that would permit viable hatching under the
389 higher temperatures that rapidly follow the breakup of ice cover (Weyhenmeyer et al. 2004).
390 This is consistent with reported observations of thermally-induced mass hatching (Luczynski
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 20 of 43
391 1984; Dostatni et al. 1999; Hooper 2006) in response to a small increase in temperature late in
392 development. Heterograde hatching provides a mathematical basis for this phenomenon, which
393 superficially resembles environmentally-cued hatching. However, in this case the temperature
394 increase may trigger rapid hatching through the instantaneous increase in RD necessitated by the
395 thermal dependence of hatching stage.
396 An implication of heterograde hatching is that the effect of variable incubation temperatures on
397 hatch timing will depend not only on the amplitude of the temperature variation and mean
398 temperature, but on whether variation is symmetric (i.e. fluctuating around an unchanging mean
399 temperature) or asymmetric (temperatures that fluctuate around an increasing or decreasing mean 400 temperature). Transient temperature peaksDraft will also affect hatch timing differently depending on 401 whether they are experienced early or late in development. Early in development, a transient
402 temperature peak would have a relatively minor effect on hatch timing (consistent with
403 observations by Lee et al. 2016). Once hatching competence is reached, a similar transient
404 increase in incubation temperature may trigger mass hatching if it brings the thermally dependent
405 RD to 100%. Therefore, hatching may be synchronized for both early and late spawned embryos,
406 although the embryos of early spawners will be more developed with a larger body size and
407 commensurately less yolk.
408 Heterograde hatching has been reported for only a few species (Jordaan et al. 2006; Steenfeldt et
409 al. 2010; Luczynski and Kolman 1987). We speculate that the hatching of other fall spawning
410 Coregonids (e.g. Griffiths 1980) may also prove to be heterograde. Luczynski and Kirklewska
411 (1984) noted mass hatching was triggered by an increase in temperature when applied to late
412 stage C. albula embryos, and it is possible that other fish species having an apparently high
413 developmental Q10 or exhibiting mass hatching with an upward temperature shift may share a
https://mc06.manuscriptcentral.com/cjfas-pubs Page 21 of 43 Canadian Journal of Fisheries and Aquatic Sciences
414 thermal dependency in hatching stage. Moreover, recent experimental data suggests that
415 heterograde hatching might exist in at least some insects. Salis et al. (2016) compared the time
416 for egg development in the winter moth (Operophera brumata) following a shift from a base 10
417 oC incubation temperature to either warmer (15 oC) or cooler (5 oC) temperatures at different
418 stages during development. While Salis et al. (2016) speculated that the effect of temperature on
419 development rate depends on development stage, their results are also consistent with anfhet of
420 about 0.2.
421 For the lake whitefish, development stage at hatching varies according to a consistent and
422 predictable relationship to temperature. We also show that this relationship is the dominant factor 423 accounting for the thermal dependenceDraft of size-at-hatch observed in numerous studies. As 424 heterograde hatching affects not only the duration of incubation, but the size and yolk reserves of
425 the hatching embryos, it should be recognized as a potential confounding factor in the
426 determination of yolk conversion efficiency.
427 Acknowledgements
428 The authors wish to thank Lisa Stoa, Emily Hulley and Shayen Sreetharan for assistance with the
429 experimental work and morphometric analysis. The Ontario Ministry of Natural Resources and
430 Forestry is gratefully acknowledged for their assistance in donating whitefish embryos from
431 Lake Simcoe. We are grateful to Bruce Power for its support of this work, both financially and
432 technically, and for funding through a Natural Sciences and Engineering Research Council of
433 Canada (NSERC) Collaborative Research and Development Grant.
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 22 of 43
434 References
435 Angilletta, M.J. and Dunham, A.E. 2003. The temperature-size rule in Ectotherms: Simple
436 evolutionary explanations may not be general; The American Naturalist, Vol. 162, No. 3 2003,
437 332-342. doi:10.1086/377187
438 Anras, M.L.B. 1999. Movement and habitat use by lake whitefish during spawning in a boreal
439 lake: integrating acoustic telemetry and geographic information systems; Transactions of the
440 American Fisheries Society 128: 939-952. doi: 10.1577/1548-
441 8659(1999)128<0939:mahubl>2.0.co;2
442 Atkinson, D. and Sibly, R.M. 1997. WhyDraft are organisms usually bigger in colder environments? 443 Making sense of a life history puzzle; Trends in Ecology & Evolution 12:235–239. doi:
444 10.1016/s0169-5347(97)01058-6
445 Blaxter, J.H.S. and Hempel, G. 1963. The influence of egg size on herring larvae (Clupea
446 harengus L.); ICES Journal of Marine Science 28.2 (1963): 211-240. doi:
447 10.1093/icesjms/28.2.211
448 Brooke, L.T. 1975. Effect of different constant incubation temperatures on egg survival and
449 embryonic development in lake whitefish (Coregonus clupeaformis). Transactions of the
450 American Fisheries Society. 104 (3), 555-559. doi: 10.1577/1548-
451 8659(1975)104<555:eodcit>2.0.co;2
452 Colby, P. J. and Brooke, L.T. 1973. Effects of temperature on embryonic development of Lake
453 Herring (Coregonus artedii); Journal of the Fisheries Research Board of Canada 30.6: 799-810.
454 doi: 10.1139/f73-135
https://mc06.manuscriptcentral.com/cjfas-pubs Page 23 of 43 Canadian Journal of Fisheries and Aquatic Sciences
455 Czerkies, P., Brzuzan, P., Kordalski K. & Luczynski, M. 2001. Critical partial pressures of
456 oxygen causing precocious hatching in Coregonus lavaretus and C. albula embryos; Aquaculture
457 196 (2001). 151–158. doi: 10.1016/s0044-8486(00)00545-7
458 Dostatni, D., Mamcarz, A.J. and Poczyczyński, P. 1999. Feeding of early- and late-hatched
459 whitefish in illuminated cages; Archives of Polish Fisheries. Vol. 7, Fasc. 2 (1999) 281-295.
460 Duguay, C.R., Prowse, T.D., Barrie, R., Bonsal, R., Brown, R. D., Lacroix, M.P., and Ménard, P.
461 2006. Recent trends in Canadian lake ice cover; Hydrological Processes. 20, 781–801.
462 10.1002/hyp.6131
463 Eme, J., Mueller, C.A., Manzon, R.G., Somers,Draft C.M., Boreham, D.R., and Wilson, J.Y. 2015. 464 Critical windows in embryonic development: Shifting incubation temperatures alter heart rate
465 and oxygen consumption of Lake Whitefish (Coregonus clupeaformis) embryos and hatchlings.
466 Comparative Biochemistry and Physiology Part A; Molecular & Integrative Physiology. 179, 71-
467 80. 10.1016/j.cbpa.2014.09.005
468 Fink, W.L., 1982. The conceptual relationship between ontogeny and phylogeny; Paleobiology
469 Vol. 8, No. 3 (Summer, 1982), pp. 254-264. 10.1017/s0094837300006977
470 Forster, J., and Hirst, A.G. 2012. The temperature-size rule emerges from ontogenetic
471 differences between growth and development rates; Functional Ecology Volume 26, Issue 2,
472 483–492. 10.1111/j.1365-2435.2011.01958.x
473 Griffiths, J.S. 1979. Potential effects of unstable thermal discharges on incubation of Lake
474 Whitefish eggs; Ontario Hydro Research Division Report No. 79-521-K.
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 24 of 43
475 Griffiths, J.S. 1980. Potential effects of unstable thermal discharges on incubation of Round
476 Whitefish eggs; Ontario Hydro Research Division Report No. 80-140-K.
477 Hart, J.L. 1931. The spawning and early life history of the whitefish, Coregonus clupeaformis
478 (Mitchill), in the Bay of Quinte, Ontario; Contributions to Canadian Biology and Fisheries 6 (1),
479 165-214. 10.1139/f31-007
480 Hooper, G.W., 2006: Techniques used for the intensive culture of Lake Whitefish (Coregonus
481 clupeaformis) in Ontario, Canada; Special Publication Aquaculture Association of Canada (11):
482 169-172. http://www.aquacultureassociation.ca/assets/Uploads/AAC-Spec-Pub-No-11-
483 Freshwater.pdf, accessed March 27, 2017 Draft 484 Huey, R.B., and Berrigan, D. 2001. Temperature, demography, and ectotherm fitness; The
485 American Naturalist 158.2: 204-210. 10.2307/3079165
486 Jordaan, A., 2002. The effect of temperature on the development, growth and survival of Atlantic
487 Cod (Gadus morhua) during early life-histories; Electronic Theses and Dissertations. Paper 141.
488 Jordaan, A., Hayhurst, S.E. and Kling, L J. 2006. The influence of temperature on the stage at
489 hatch of laboratory reared Gadus morhua and implications for comparisons of length and
490 morphology; Journal of Fish Biology 68.1 (2006): 7-24. 10.1111/j.0022-1112.2006.00857.x
491 Kingsolver, J. G., Izem, R. and Ragland, G.J. 2004. Plasticity of size and growth in fluctuating
492 thermal environments: Comparing reaction norms and performance curves; Integrative and
493 Comparative Biology 44.6 (2004): 450-460. https://doi.org/10.1093/icb/44.6.450
https://mc06.manuscriptcentral.com/cjfas-pubs Page 25 of 43 Canadian Journal of Fisheries and Aquatic Sciences
494 Lee, A.H., Eme, J., Mueller, C.A., Manzon, R.G., Somers, C.M., Boreham D.R., and Wilson,
495 J.Y. 2016. The effects of increased constant incubation temperature and cumulative acute heat
496 shock exposures on morphology and survival of Lake Whitefish (Coregonus clupeaformis)
497 embryos; Journal of Thermal Biology 57 (2016) 11–20. 10.1016/j.jtherbio.2016.01.010
498 Lim, M.Y, Manzon, R.G., Somers, C.M., Boreham D.R., and Wilson, J.Y. 2017. The effects of
499 fluctuating temperature regimes on the embryonic development of lake whitefish (Coregonus
500 clupeaformis); Comparative biochemistry and physiology. Part A, Molecular & integrative
501 physiology 214. 10.1016/j.cbpa.2017.08.010
502 Luczyński, M. 1984. Improvement in the efficiency of stocking lakes with larvae of Coregonus 503 albula L. by delaying hatching; AquacultureDraft 41: 99 111. 10.1016/0044-8486(84)90087-5
504 Luczyński, M., and Kirklewska, A. 1984. Dependence of Coregonus albula embryogenesis rate
505 on the incubation temperature; Aquaculture, Vol. 42, no. 1, pp. 43-55. 10.1016/0044-
506 8486(84)90312-0
507 Luczyński, M., and Kolman, R. 1987. Hatching of Coregonus albula and C. Lavaretus embryos
508 at different stages of development; Environmental Biology of Fishes 19.4 (1987): 309-315.
509 10.1007/bf00003231
510 Martin, K., Bailey, K., Moravek, C., and Carlson, K. 2011. Taking the plunge: California
511 grunion embryos emerge rapidly with environmentally cued hatching; Integrative and
512 Comparative Biology 51.1 (2011): 26-37. 10.1093/icb/icr037
513 Martin, T.L., and Huey, R.B. 2008. Why “suboptimal” is optimal: Jensen’s Inequality and
514 ectotherm thermal preferences; The American Naturalist 171: E102-118. 10.1086/527502
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 26 of 43
515 Melo, M.T., Saturnino, C., Santos, J.N.S., Vasconcellos, R.M., Cruz-Filho, A G., and Araujo,
516 F.G. 2010. Correction of the weight and length for juveniles Atherinella brasilliensis
517 (actinopterygii: atherinopsidae) after fixation in formalin and preservation in ethanol; Zoologia.
518 27: 892-896. 10.1590/s1984-46702010000600009
519 Meyer, S., Caldarone, E., Chícharo, M., Clemmesen, C., Faria, A., Faulk, C., Folkvord, A., Holt,
520 G., Høie, H., Kanstinger, P., Malzahn, A., Moran, D., Petereit C., Støttrup J., and Peck, M. 2012.
521 On the edge of death: Rates of decline and lower thresholds of biochemical condition in food-
522 deprived fish larvae and juveniles; Journal of Marine Systems 93 (2012): 11-24.
523 10.1016/j.jmarsys.2011.09.010
524 Mitz, C., Thome, C., Cybulski M.E., LaframboiseDraft L., Somers, C.M., Manzon, R.G., Wilson, 525 J.Y., and Boreham D.R., 2014. A self-contained, controlled hatchery system for rearing lake
526 whitefish embryos for experimental aquaculture; North American Journal of Aquaculture. 76 (3),
527 179-184. 10.1080/15222055.2014.893472
528 Mitz, C., Thome, C., Thompson, J., Manzon, R.G., Wilson J.Y., and Boreham, D.R. 2017. A
529 method to transform a variable thermal regime to a physiologically equivalent effective
530 temperature. Journal of Thermal Biology, 65, 21-25 10.1016/j.jtherbio.2016.12.011
531 Mueller, C.A., Eme, J., Manzon, R.G., Somers C.M., Boreham, D.R. and Wilson J.Y., 2015.
532 Embryonic critical windows: Changes in incubation temperature alter survival, hatchling
533 phenotype and cost of development in lake whitefish (Coregonus clupeaformis); Journal of
534 Comparative Physiology. B, Biochemical, Systemic, and Environmental Physiology. 185 (3),
535 315-331. 10.1007/s00360-015-0886-8
https://mc06.manuscriptcentral.com/cjfas-pubs Page 27 of 43 Canadian Journal of Fisheries and Aquatic Sciences
536 Parichy, D.M., Elizondo, M.R., Mills, M.G., Gordon, T.N., and Engeszer R.E. 2009. Normal
537 table of post-embryonic Zebrafish development: Staging by externally visible anatomy of the
538 living fish. Developmental Dynamics; an Official Publication of the American Association of
539 Anatomists, 238(12), 2975–3015. 10.1002/dvdy.22113
540 Perrin, N., 1995. About Berrigan and Charnov’s life history puzzle; Oikos 73:137–139.
541 10.2307/3545737
542 Peterson, W.T., 2001. Patterns in stage duration and development among marine and freshwater
543 calanoid and cyclopoid copepods: a review of rules, physiological constraints, and evolutionary
544 significance. Hydrobiologia 453, 91–105. 10.1023/A:1013111832700 Draft 545 Price, J.W., 1940. Time-temperature relations in the incubation of the whitefish, Coregonus
546 clupeaformis (Mitchill); Journal of General Physiology. 23 (4), 449-468. 10.1085/jgp.23.4.449
547 Ragland, G.J., and Kingsolver, J.G. 2008. The effect of fluctuating temperatures on ectotherm
548 life-history traits: comparisons among geographic populations of Wyeomyia smithii;
549 Evolutionary Ecology Research 10: 29-44.
550 Rombough, P.J, 1988. Growth, aerobic metabolism, and dissolved oxygen requirements of
551 embryos and alevins of steelhead, Salmo gairdneri. Canadian Journal of Zoology 66:651–660.
552 10.1139/z88-097
553 Salis, L., Lof, M., Asch, M.V., and Visser, M.E. 2016. Modeling winter moth Operophtera
554 brumata egg phenology: nonlinear effects of temperature and developmental stage on
555 developmental rate. Oikos, 125(12), 1772-1781. doi:10.1111/oik.03257
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 28 of 43
556 Scott, W.B. and Crossman, E.J. 1973. Freshwater fishes of Canada; Fish Research Board of
557 Canada Bulletin 184. 10.2307/1442619
558 Schulte, P.M., Healy, T.M., and Fangue, N.A. 2011. Thermal Performance Curves, Phenotypic
559 Plasticity, and the time scales of temperature exposure; Integrative and Comparative Biology
560 (51) No. 5, 691-702. 10.1093/icb/icr097
561 Sibly R. M., and Atkinson, D. 1994 How rearing temperature affects optimal adult size in
562 ectotherms. Functional Ecology 8, 486–493. 10.2307/2390073
563 Smith, B.B., and Walker, K.F. 2003. Shrinkage of 0+ carp (Cyprinus carpio L.) after 564 preservation in ethanol; Marine and FreshwaterDraft Research. 54: 113-116. 10.1071/MF02089 565 Spicer, J.I., and Burggren, W.W. 2003. Development of physiological regulatory systems:
566 Altering the timing of crucial events; Zoology 106.2 (2003): 91-99. 10.1078/0944-2006-00103
567 Sreetharan, S., Thome, C., Mitz, C. Eme, J., Mueller, C.A., Hulley, E.N., Manzon, R.G., Somers,
568 C.M., Boreham, D.R., and Wilson, J.Y. 2015. Embryonic development of lake whitefish
569 (Coregonus clupeaformis): a staging series, analysis of growth and impacts of fixation; Journal
570 of Fish Biology 87.3 (2015): 539-558. 10.1111/jfb.12725
571 Steenfeldt, S., Lund, I., and Höglund, E. 2010. Is batch variability in hatching time related to size
572 heterogeneity and cannibalism in pikeperch (Sander lucioperca)?; Aquaculture Research
573 42(5):727 – 732. 10.1111/j.1365-2109.2010.02681.x
574 Thome, C., Mitz, C., Somers, C.M., Manzon, R.G., Wilson, J.Y., and Boreham, D.R. 2016.
575 Incubation of Lake Whitefish (Coregonus clupeaformis) embryos in cooling water discharge and
https://mc06.manuscriptcentral.com/cjfas-pubs Page 29 of 43 Canadian Journal of Fisheries and Aquatic Sciences
576 the impacts of fluctuating thermal regimes on development: Canadian Journal of Fisheries and
577 Aquatic Sciences, ttp://dx.doi.org/10.1139/cjfas-2015-0286.
578 Touchon, J.C., McCoy, M.W., Landberg, T., Vonesh J.R., and Warkentin, K.M. 2015. Putting
579 u/g in a new light: Plasticity in life history switch points reflects fine-scale adaptive responses;
580 Ecology 96.8 (2015): 2192-2202. 10.1890/14-1301.1
581 van der Have, T.M., and de Jong, G. 1996. Adult size in ectotherms: temperature effects on
582 growth and differentiation. Journal of Theoretical Biology. 138: 329–340.
583 10.1006/jtbi.1996.0224
584 Warkentin, K.M., 2011. EnvironmentallyDraft cued hatching across taxa: Embryos respond to risk 585 and opportunity; Integrative and Comparative Biology 51.1 (2011): 14-25. 10.1093/icb/icr017
586 Wedekind, C., 2002 Induced hatching to avoid infectious egg disease in whitefish; Current
587 Biology, Vol. 12, 69–71. 10.1016/s0960-9822(01)00627-3
588 Wedekind, C., and Müller, R. 2005. Risk-induced early hatching in Salmonids. Ecology
589 86:2525–2529. 10.1890/04-1738
590 Weyhenmeyer, G.A., Meili, M., and Livingstone, D.M. 2004. Nonlinear temperature response of
591 lake ice breakup; Geophysical Research Letters 31.7 10.1029/2004gl019530
592 Zuo, W., Moses, M.E., West, G.B., Hou, C., and Brown, J.H. 2012. A general model for effects
593 of temperature on ectotherm ontogenetic growth and development: Proceedings of the Royal
594 Society B (2012) 279, 1840–1846. 10.1098/rspb.2011.2000
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 30 of 43
595 Figure Captions
596 Figure 1: Effect of heterograde hatching on hatch timing under an asymmetrically variable 597 thermal regime. Under homograde hatching, hatching occurs at developmental stage S3 598 independent of the incubation temperature. However, under heterograde hatching, hatching will 599 occur at developmental stages that are dependent on the incubation temperature. In this 600 illustration, the development stage is shown for an embryo developing at 2oC to Development 601 Stage S1 (70% relative development under both the heterograde and homograde assumptions) 602 when it is shifted to 5oC (broken line with arrows). Time is not shown but may be visualized as 603 the Z-axis. As shown, the post-shift time to hatch will be overestimated under a homograde 604 assumption (70% relative development under the homograde assumption vs. 80% relative o 605 development under the heterograde model). A shift from 2 to 5 C occurring at S2 brings the 606 embryos to 100% relative development under the heterograde model and is predicted to trigger 607 mass hatching.
608 Figure 2: Morphological measurements included standard length (SL), body height 609 measured immediately post-vent (H), yolk sac length (Yl), yolk sac height (Yh) oil globule 610 length (Ol) and oil globule height (Oh). The area and volume of the yolk sac and yolk oil 611 globule were calculated from the measured parameters assuming an ellipsoidal yolk shape. 612 Fin indentation ratio F = dm/((da +dp)/2Draft) where dm is the dorsal minimum, dp is the posterior 613 maximum, and da is the anterior maximum. 614 615 Figure 3: Temperature shifts used to determine implicit post-shift thermal scaling (Q10) 616 for time-to-hatch. If hatching is a valid developmental stage then temperature shifts from a 617 low base temperature to higher temperatures will have little effect on the apparent post-shift 618 value for Q10. If hatching is heterograde then shifts in incubation temperature will accentuate 619 the apparent thermal scaling and result in progressively higher calculated values for Q10 as 620 development proceeds since the upward shift occurs closer to the hatching stage. 621 622 Figure 4: Graphical summary of measurements A – C: Summary of measurements of SL, 623 yolk area, and fin indentation ratio for sF Lake Simcoe embryos hatching at different constant 624 incubation temperatures ( – 2oC, - 5oC, - 8oC ). D – F: measurements of SL, yolk area, 625 and fin indentation ratio for sF Lake Simcoe embryos hatching under progressively increasing 626 () and progressively decreasing () thermal regimes. mF data for Lake Huron trials is not 627 shown.
628 Figure 5: Relationship between unpreserved dry weight (DW) and standard length (SL) 629 for constant temperature incubation ( – 2oC, M- 5oC, Ú - 8oC) for Lake Simcoe sF 630 embryos. Regression lines are given for the 2 and 5 oC (DW = 0.0492exp0.244SL, R2=0.97, 631 P<0.001), and 8oC trials (DW = 0.0268exp0.279SL, R2=0.947, P<0.001). 632 633 Figure 6: Observed conversion of yolk dry weight to somatic dry weight for embryos 634 hatching at different constant incubation temperatures ( – 2oC, M- 5oC, Ú - 8oC). The -1 635 dashed lines represent different regression lines for the relationship Ws=εWy((1+aWy) ) for 2 636 and 8oC. A representative image of a hatchling for each incubation temperature is included.
https://mc06.manuscriptcentral.com/cjfas-pubs Page 31 of 43 Canadian Journal of Fisheries and Aquatic Sciences
638 Figure 7: Nominal values for Q10 following a discrete temperature shift. Observed 639 relationship between post-shift values for Q10 and the timing of the temperature shift. The 640 dashed line shows the predicted hyperbolic relationship between post shift Q10 and the timing 641 of the shift (: 2→8oC, : 2→ 5oC). 642 643 Figure 8: Relationship between calculated relative development (RD) and the timing of 25 o o 644 the temperature shift (: 2→8 C, : 2→5 C). f het is illustrated graphically. The solid 645 line shows a linear regression for the 2→5oC data. RD is calculated as the cumulative 646 incubation time divided by the time to median hatch.
Draft
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 32 of 43
1 Table 1. Experimental Thermal Regimes mF Lake Huron 2013-2014 and 2014-2015
o Regime Mean T ( C) Details Constant (0) 0.5±0.2 Fertilization → hatch Constant (2) 2.5±0.87 Fertilization → hatch Constant (5) 5.2±0.33 Fertilization → hatch Constant (8) 8.1±0.28 Fertilization → hatch Asymmetric (8→5→2°C) 3.4 17 d → 33d → hatch Asymmetric (2→5→8°C) 4.2 45d → 33d → hatch sF Lake Simcoe 2014-2015 Constant (2) 2.0±0.15 Fertilization → hatch Constant (5) 5.0±0.08 Fertilization → hatch Constant (8) 7.9±0.17 Fertilization → hatch Asymmetric (8→5→2°C) 3.3 17 d → 33d → hatch Asymmetric (2→5→8°C) 4.1 45d → 33d → hatch Asymmetric (2→5°C at different development times) varies (81, 93, 106, 121, 138, 151d) → hatch Asymmetric (2→8°C at different development times) varies (81, 93, 106, 121, 138, 151d) → hatch 2 Temperatures are given as mean ± Standard Deviation.Draft No standard deviations are given for the asymmetrically 3 variable regimes although they experienced the same thermal variance as the constant temperature regimes for the 4 time spent at the respective constant temperatures. 5
https://mc06.manuscriptcentral.com/cjfas-pubs Page 33 of 43 Canadian Journal of Fisheries and Aquatic Sciences
6 Table 2. Summary of time-to-hatch and morphological measurements for lake whitefish 7 hatchlings reared under different constant and variable thermal regimes. Error represents 8 standard error of the mean. Live Yolk conversion efficiency Regime Yolk area Oil globule Fin indentation hatch h h -h SL (mm) (oC) 50 10 90 (mm2) area (mm2) ratio, F (nominal) (normalized) (%) Lake Huron Embryos (mF) 0.5 9 204 199-207 13.99 ±0.07(9) 1.88±0.08(9) 0.55±0.09(9) 0.52±0.02(9) … … 2 58 158 141-163 13.50±0.13(27) 2.71±0.14(27) 0.88±0.04(27) 0.61±0.01(27) … … 5 35 109 100-119 13.08±0.12(13) 8.15±0.55(11) 1.12±0.04(12) 0.68±0.02(12) … … 8 8 67 64-71 11.74±0.17 (6) … … … … … 2→5→8 20 104 101-104 12.66±0.05(32) 8.79±0.11(30) 2.00±0.03(30) 0.73±0.07(30) … … 8→5→2 42 154 154-174 13.41±0.11(33) 4.15±0.39(32) 1.24±0.08(32) 0.66±0.02(32) … … Lake Simcoe Embryos (sF) 2 46 181 168-188 13.97±0.06(44) 3.04±0.17(44) 1.33±0.05(44) 0.57±0.01(44) 70.5±0.52(36) 69.2±0.49(36) 5 38 112 96-134 13.24±0.08(33) 6.99±0.32(33) 1.90±0.06(33) 0.66±0.01(33) 70.8±0.96(30) 67.5±0.85(30) 8 12 68 62-75 12.12±0.10(12) 11.02±0.35(12) 2.53±0.12(12) 0.79±0.02(10) 71.5±1.10(12) 65.8±0.38(12)* 2→5→8 18 104 101-104 12.54±0.05(25) 10.33±0.19(27) 2.25±0.03(27) 0.73±0.01(23) … … 8→5→2 37 170 153-179 13.35±0.06(33) Draft4.45±0.27(27) 1.52±0.08(27) 0.68±0.01(20) … … * – conservative value derived using the standard length (SL) – dry weight relationship for the 2 and 5 degree groups which tended to have a higher condition factor. YCE would be lower for the 8-degree group if we applied a condition factor correction. hx refers to fractional hatching with median hatching occurring at h50. F is fin indentation ratio defined as the dorsal minimum divided by the average dorsal maxima. The number of embryos (n) varies within treatment groups as not all measurements could be obtained for each individual (e.g. damaged fins or ruptured yolk). Percentage hatch data for the 2→5→8 and 8→5→2 treatments has been corrected for initial pre-shift mortality. 9
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 34 of 43
1.6
1.4 Hatching (homograde) S3 - 1.2 Hatching (heterograde) 1S2 -
S1 - 0.8
0.6
0.4 Development Stage (S) Stage Development
0.2
0 0 1 2 3 4 5 6 7 8 9 Incubation Temperature (oC) 1 2 Figure 1: Effect of heterograde hatchingDraft on hatch timing under an asymmetrically 3 variable thermal regime. Under homograde hatching, hatching occurs at developmental 4 stage S3 independent of the incubation temperature. However, under heterograde hatching, 5 hatching will occur at developmental stages that are dependent on the incubation 6 temperature. In this illustration, the development stage is shown for an embryo developing o 7 at 2 C to Development Stage S1 (70% relative development under both the heterograde and 8 homograde assumptions) when it is shifted to 5oC (broken line with arrows). Time is not 9 shown but may be visualized as the Z-axis. As shown, the post-shift time to hatch will be 10 overestimated under a homograde assumption (70% relative development under the 11 homograde assumption vs. 80% relative development under the heterograde model). A shift o 12 from 2 to 5 C occurring at S2 brings the embryos to 100% relative development under the 13 heterograde model and is predicted to trigger mass hatching.
https://mc06.manuscriptcentral.com/cjfas-pubs Page 35 of 43 Canadian Journal of Fisheries and Aquatic Sciences
14 15 16 17 18 SL 19 da dm dp
H
20 21 Ol Oh Yl Yh 22 23 24 25 Figure 2: Morphological measurements included standard length (SL), body height 26 measured immediately post-vent (H), yolk sac length (Yl), yolk sac height (Yh) oil globule 27 length (Ol) and oil globule height (Oh). The area and volume of the yolk sac and yolk oil 28 globule were calculated from the measured parameters assuming an ellipsoidal yolk shape. 29 Fin indentation ratio F = dm/((da +dp)/2Draft) where dm is the dorsal minimum, dp is the posterior 30 maximum, and da is the anterior maximum. 31
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 36 of 43
32 33 34 8oC shift 35 hatch Shift in incubation Post-shift Q 36 10 temperature to 5 or 8oC 5oC shift hatch calculated for 37 a different times 2→5 & 2→8 38 o fertilization 2 C base hatch 39 40 41 42 Figure 3: Temperature shifts used to determine implicit post-shift thermal scaling
43 (Q10) for time-to-hatch. If hatching is a valid developmental stage then temperature shifts 44 from a low base temperature to higher temperatures will have little effect on the apparent 45 post-shift value for Q10. If hatching is heterograde then shifts in incubation temperature will 46 accentuate the apparent thermal scaling and result in progressively higher calculated values 47 for Q10 as development proceeds since the upward shift occurs closer to the hatching stage. 48 Draft
https://mc06.manuscriptcentral.com/cjfas-pubs Page 37 of 43 Canadian Journal of Fisheries and Aquatic Sciences
2 oC
5 oC
8 oC
49 Draft 50 51 Figure 4: Observed conversion of yolk dry weight to somatic dry weight for embryos 52 hatching at different constant incubation temperatures (⚫ – 2oC, - 5oC, - 8oC ). -1 53 Dashed lines represent different regressions for the relationship Ws=εWy((1+aWy) ) for 2 54 and 8oC. A representative image of a hatchling for each incubation temperature is 55 included. 56
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 38 of 43
A E
B F
C G Draft
D H
57 58 Figure 5: Graphical summary of measurements A – D: Summary of measurements of 59 SL, H, yolk area, and fin indentation ratio for sF Lake Simcoe embryos hatching at different 60 constant incubation temperatures (⚫ – 2oC, - 5oC, - 8oC ). E – H: measurements of 61 SL, H, yolk area, and fin indentation ratio for sF Lake Simcoe embryos hatching under 62 progressively increasing () and progressively decreasing (⚫) thermal regimes. mF data 63 for Lake Huron trials is not shown.
64
https://mc06.manuscriptcentral.com/cjfas-pubs Page 39 of 43 Canadian Journal of Fisheries and Aquatic Sciences
1.6
1.4
1.2
1
0.8
0.6
Larva dry weight (mg) dry weight Larva 0.4
0.2 Range of SL at hatching 0 4 6 8 10 12 14 16 SL (mm) 65 66 Figure 6: Relationship between unpreservedDraft dry weight (DW) and standard length 67 (SL) for constant temperature incubation ((⚫ – 2oC, - 5oC, - 8oC ) for Lake Simcoe 68 sF embryos. Regression lines are given for the 2 and 5 oC (DW = 0.0492exp0.244SL, 69 R2=0.97, P<0.001), and 8oC trials (DW = 0.0268exp0.279SL, R2=0.947, P<0.001). 70
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 40 of 43
1000
100 10 shift shift Q - 10 Post
1 0 50 100 150 200 250
71 Time of temperature shift (dpf)
72 Figure 7: Nominal values for Q10 following a discrete temperature shift. Observed 73 relationship between post-shift valuesDraft for Q10 and the timing of the temperature shift. The 74 dashed line shows the predicted hyperbolic relationship between post shift Q10 and the timing 75 of the shift (: 2→8oC, : 2→ 5oC).
https://mc06.manuscriptcentral.com/cjfas-pubs Page 41 of 43 Canadian Journal of Fisheries and Aquatic Sciences
100% C) o 80%
60%
40%
20% shift Incubation (%RD at 5 5 or 8 at (%RDshift Incubation
-
Post 0% 0% 20% 40% 60% 80% 100% Pre-shift incubation (%RD at 2oC) 76 77 Figure 8: Relationship between calculatedDraft relative development (RD) and the timing of →52 o o 78 the temperature shift (: 2→8 C, : 2→5 C). f het is illustrated graphically. The solid 79 line shows a linear regression for the 2→5oC data. RD is calculated as the cumulative 80 incubation time divided by the time to median hatch.
https://mc06.manuscriptcentral.com/cjfas-pubs Canadian Journal of Fisheries and Aquatic Sciences Page 42 of 43
Basis for normalization of YCE across development stages
We can treat ontogenetic growth as a simplified energetic case where energy is partitioned
dW dR between growth, S and maintenance, S according to: dt dt dW dW dR y 1 S [1] dt dt dt and dR f (W ) [2] dt s
where WS is somatic (yolk-free) dry weight, Wy is the dry weight of yolk consumed at any point in time, and is a limiting efficiency ofDraft yolk conversion as the maintenance allocation approaches zero.
Substituting [2] into [1] gives: dW dW y 1( S ) f (W ) [3] dt dt s
dW If S is a constant over a meaningful portion of embryonic development (Sreetharan et al. dt
2015 and this study), then over that portion of development at any point in time:
dWy is a function of WS , and WS may be related to Wy in the form of a function where dt dW S is equal to the instantaneous yolk conversion ratio (YCE). YCE may then be normalized dWy
to a consistent value for Wy .
https://mc06.manuscriptcentral.com/cjfas-pubs Page 43 of 43 Canadian Journal of Fisheries and Aquatic Sciences
2.5
2
1.5 (mg) S W 1
0.5
0 0 0.5 Draft1 1.5 2 2.5
Wy (mg)
Figure A-1: Observed relationship between somatic dry weight (WS) and the dry weight of yolk consumed (Wy) for lake whitefish embryos incubated at a constant 5oC (heavy solid line). The light dashed line represents 100% conversion of yolk dry weight to somatic dry -1 weight. Regression of the pooled data (+) using Ws=εWy((1+aWy) ) yielded a near-perfect curve fit (R2=0.999). Regression of the calculated points for individual embryos (), resulted in a fit similar to that obtained for the pooled embryos (R2=0.983, P<0.0001). Individual points closely matched the predicted value (RMSE 0.041 ± 0.012 mg) with no significant trend in the magnitude of error with development time (P=0.543).
https://mc06.manuscriptcentral.com/cjfas-pubs