Exposure to exogenous egg cortisol does not rescue juvenile Chinook salmon body size, condition, or survival from the effects of elevated water temperatures

Abstract Climate change is leading to altered temperature regimes which are impacting aquatic life, particularly for ectothermic fish. The impacts of environmental stress can be translated across generations through maternally derived glucocorticoids, leading to altered offspring phenotypes. Although these maternal stress effects are often considered negative, recent studies suggest this maternal stress signal may prepare offspring for a similarly stressful environment (environmental match). We applied the environmental match hypothesis to examine whether a prenatal stress signal can dampen the effects of elevated water temperatures on body size, condition, and survival during early development in Chinook salmon Oncorhynchus tshawytscha from Lake Ontario, Canada. We exposed fertilized eggs to prenatal exogenous egg cortisol (1,000 ng/ml cortisol or 0 ng/ml control) and then reared these dosed groups at temperatures indicative of current (+0°C) and future (+3°C) temperature conditions. Offspring reared in elevated temperatures were smaller and had a lower survival at the hatchling developmental stage. Overall, we found that our exogenous cortisol dose did not dampen effects of elevated rearing temperatures (environmental match) on body size or early survival. Instead, our eyed stage survival indicates that our prenatal cortisol dose may be detrimental, as cortisol‐dosed offspring raised in elevated temperatures had lower survival than cortisol‐dosed and control reared in current temperatures. Our results suggest that a maternal stress signal may not be able to ameliorate the effects of thermal stress during early development. However, we highlight the importance of interpreting the fitness impacts of maternal stress within an environmentally relevant context.


| INTRODUC TI ON
Climate change is altering habitats across the globe, introducing environmental stressors such as elevated mean temperatures (Stocker et al., 2013), increased frequency of extreme weather events (e.g., floods, droughts; Easterling et al., 2000;Fischer & Knutti, 2015), and novel competitor and predator interactions (e.g., via species range expansions toward the poles; Parmesan & Yohe, 2003). Climate change is occurring at an accelerated rate (Loarie et al., 2009) and is consequently affecting the capacity of organisms to respond adaptively (Palmer et al., 2017;Woodward et al., 2016). Organisms may respond to environmental stressors through phenotypic changes in their growth, morphology, reproduction, and survival (Barton, 2002), which can include rapid-acting mechanisms within and across generations such as phenotypically flexible responses, phenotypic plasticity, and contemporary evolution (Hendry, Farrugia, & Kinnison, 2008;Sih, Ferrari, & Harris, 2011). Climate change is expected to have strong direct and indirect effects on aquatic systems, with alterations in the hydrological cycle (e.g., changes in precipitation) leading to extremes in water availability (i.e., floods and droughts), increases in air temperature leading to increased water temperatures, and increased CO 2 leading to increasing water acidity (Bates, Kundzewicz, Wu, & Palutikof, 2008). Notably, increasing temperatures alone are expected to greatly impact ectothermic organisms such as invertebrates and fishes, and oftentimes are considered to be negative (Ficke, Myrick, & Hansen, 2007;Stoks, Geerts, & De Meester, 2014). Indeed, warming waters generate offspring phenotypic traits such as smaller body size (Kuehne, Olden, & Duda, 2012;Whitney, Hinch, Patterson, & a., 2014), faster growth (Beacham & Murray, 1990), and increased metabolism (Enders & Boisclair, 2016).
Maternal effects have long been recognized for the role in the finescale tuning of adaptive responses to larger environmental stressors (Räsänen & Kruuk, 2007). For example, maternal exposure to environmental stressors translates information to developing offspring about the relative quality of their future environment (Sheriff et al., 2017).
Maternally derived hormones have recently been acknowledged as a possible mechanism by which information about environmental quality can be translated to offspring via the mother (Meylan et al., 2012). In particular, maternally derived glucocorticoid (GC) hormones have been proposed as preparative mediators of offspring phenotype and performance in relation to the predicted quality of the offspring's future environment (Love, Chin, Wynne-Edwards, & Williams, 2005;. Glucocorticoids are strong candidates for this mediatory effect since they are involved in energy management and the stress response in vertebrates (Barton, 2002;Moore & Jessop, 2003;Romero, 2004). As such, environmental stressors can elevate maternal baseline GCs (Love, McGowan, & Sheriff, 2013;Wendelaar Bonga, 1997), and maternally derived GCs can be transferred to developing offspring in utero (Matthews, 2002) or via the lipid content of eggs in oviparous species (Love et al., 2009; providing a reliable signal of current and potentially future environmental quality for offspring . Recent studies suggest that when the maternal environment is indicative of the offspring environment (environmental match hypothesis; Sheriff et al., 2017;, maternal stress may elicit predictive adaptive responses (PARs) in offspring (Bateson, Gluckman, & Hanson, 2014), generating offspring phenotypes that may be better prepared for harsher environments (Bian et al., 2015;Gagliano & McCormick, 2009;Love et al., 2013;Love & Williams, 2008). From a climate change perspective, where the projected global mean surface temperatures are expected to increase 0.3-0.7°C by 2035 (Stocker et al., 2013) and where harsher events such as extreme water flow (i.e., flood and droughts) are expected to be more frequent (Woodward et al., 2016), females may be exposed to multiple environmental stressors during reproduction. Thus, maternally derived GCs may be involved in modulating the responses of ectothermic offspring to multiple environmental stressors due to climate change, by better-preparing offspring for managing harsher environmental conditions such as elevated temperatures during development .
Here, we apply the concept of environmental matching  to examine whether exposure to maternally derived GCs can generate phenotypes that better buffer the negative phenotypic effects of developing in elevated water temperatures using Chinook salmon Oncorhynchus tshawytscha ( Figure 1). Pacific salmon are an F I G U R E 1 Picture of Chinook salmon (Oncorhynchus tshawytscha) hatchlings. Photo Credit: T. Warriner important study species for these questions since they are ectothermic and are sensitive to changes in environmental temperatures (McCullough et al., 2009); are susceptible to additional stressors during migration (i.e., increase circulating plasma GCs) when they return to terminally spawn (Cook et al., 2014;McConnachie et al., 2012); and mothers and offspring overlap spatially in their spawning and developing environment, respectively, meaning that an honest signal about environmental quality passed from mother to eggs may be valuable for salmon offspring. Importantly, climate change has been implicated for declines in Chinook salmon populations across the west coast of North America, through direct and indirect effects of water temperature increases and droughts (Isaak, Wollrab, Horan, & Chandler, 2012;Mantua, Tohver, & Hamlet, 2010). Determining whether offspring are better prepared for warmer waters after having received a hormonal maternal signal about increased environmental stress is an important component for quantifying the adaptive capacity of Chinook salmon to climate change. To investigate how the effects of elevated temperatures and maternally derived GCs interact to affect Chinook offspring phenotype and performance, we exposed eggs to exogenous cortisol or a control solution immediately postfertilization and then split these eggs within females and raised the offspring in one of two temperature regimes: current (+0°C) and elevated (+3°C: as predicted in next century by current climate models, Vliet, Franssen, et al., 2013). In accordance with previous research (e.g., Daufresne, Lengfellner, & Sommer, 2009;Whitney et al., 2014), we predicted that individuals raised in elevated temperatures would have both lower survival and body size than those raised in current temperatures. However, since preparatory responses following exposure to a maternal stress signal may dampen the effects of an environmental stressor (i.e., environmental matching hypothesis; Sheriff et al., 2017), we predicted that prenatal exposure to exogenous cortisol would help to buffer these negative impacts, resulting in relative increases in survival and body size at emergence (fry stage) for fish raised under elevated water temperatures.

| Fish Origin
All work described here was approved and completed under University of Windsor Animal Use Project Proposals (AUPPs: # 14-25 & #15-15). Our study species was Chinook salmon from the Credit River (43°34′40.0″N 79°42′06.3″W), which drains into Lake Ontario, Canada. Chinook salmon were purposely introduced to Lake Ontario starting in 1967 and they continue to be stocked for recreational fishing purposes (OMNRF, 2015). Spawning in this Great Lake's population takes place in the early fall, where eggs incubate under gravel in these natal streams until hatching in February and emergence in March, with juveniles migrating out into the lake in summer (Johnson, 2014). Tributaries that flow into the Great Lakes are expected to increase in water temperature (Chu, 2015;Vliet, Franssen, et al., 2013), and therefore investigating this population's responses to these environmental changes may provide information relevant to the future status of Great Lakes fisheries under climate change. We collected eggs and milt from fifteen adult females and nine males on October 4, 2016, from the Credit River. We measured female body mass (mean ± SE, range: 7.89 ± 0.41, 5.5-11.7 kg), fork length (85.6 ± 0.89, 80.0-93.0 cm), and ovarian mass (ovarian mass = pre-post stripping mass: 0.98 kg, ±0.10, 0.55-2.00 kg). Eggs and milt were collected by applying pressure to the abdomen, and the gametes were transported on ice in coolers to the University of Windsor's Aquatic Facility.

| Egg cortisol exposure and incubation temperatures
At the aquatic facility, we filled individual containers with 90 g of eggs (~300 eggs) from each of the 15 females and added ~0.5 ml (4 drops) of pooled milt from the nine males. After gently swilling the egg-milt solution we added 60 ml of river water to activate the sperm (Lehnert, Helou, Pitcher, Heath, & Heath, 2018). After 2 min (when sperm should no longer be motile), we added river water The remaining eggs were transferred to 4-inch × 3-inch incubation cells (each tray divided into 16 cells using metal dividers) within a vertical egg-incubation stack that followed one of two temperature treatments. Eggs were incubated either at ambient water temperatures indicative of the "current" water temperature scenario or under the predicted "future climate change" scenario (i.e., elevated by 3°C based on predicted water temperature increases;

| Fertilization success, morphology, and survival
Offspring development and mortality were assessed every two days. We removed and stored dead eggs in Stockard's solution (5% formaldehyde, 4% glacial acetic acid, 6% glycerin, 85% water) to determine fertilization success and at which developmental stage death occurred. Due to the effects of water temperature on development in this ectothermic species, offspring raised in the two temperature regimes reached development stages at distinct calendar days ( Figure 2). We therefore equalized fish developmental stages based on Accumulated Thermal Units (ATUs), which is the sum of average daily temperatures, and is highly correlated to fish growth and development (Chezik, Lester, & Venturelli, 2014;Neuheimer & Taggart, 2007). Once offspring reached their emergence stage

| Egg cortisol assay
To verify the success of the cortisol manipulation, we measured cortisol concentrations in both unfertilized and 2-hr post-treatment eggs. Briefly, we sampled three eggs from each treatment (unfertilized, control-dosed, and cortisol-dosed) for each of the 15 females (N = 45). We weighed, and blended eggs in 1.2 ml assay buffer, and then extracted the cortisol according to protocol as detailed in Capelle et al. (2017). Samples were stored in -80°C freezer until use in assay. Egg samples were run in triplicate using ELISA Cayman Cortisol kits in 1:57 dilution and were read at 412 nm on a plate reader. Intra-and interassay (i.e., plate) coefficients of variation were 8.1% and 20.4%, respectively. F I G U R E 2 Recorded average daily temperature for each temperature regime in the rearing experimental design. Line and letter color represent rearing temperature regime: current-blue and elevated-red. Vertical lines represent sampling timepoints: solid-both current and elevated groups; dashed-elevated, dotted-current F I G U R E 3 Morphological measurements of emerged juvenile Chinook salmon taken from photographs analyzed using ImageJ. The nine morphological measurements were incorporated into a principal components analysis (PCA; see Table 1

| Fertilization success and survival
There was no interactive effect of temperature and cortisol ex-

| Early survival in warming waters
Warmer temperatures, but not early cortisol exposure, affected offspring survival with fish raised in elevated temperatures having lower survival at the eyed and hatchling stages. Elevated incubation temperatures above species' preferred temperature range result in lower survival in juvenile salmonids, regardless of whether the elevation follows a stable temperature increase (Tang, Bryant, & Brannon, 1987;Whitney et al., 2013) or an oscillating temperature regime (Geist et al., 2006;Taranger & Hansen, 1993). Since salmon are ectothermic, offspring are sensitive to temperature increases, especially during early development (Beacham & Murray, 1990;Tang et al., 1987) when offspring respond strongly to environmental cues (Monaghan, 2008). This associated decline in embryo survival has been attributed to temperature-dependent increases in yolk coagulation (McCollough, 1999) and increases in developmental deformities (Cingi, Keinänen, & Vuorinen, 2010;Fraser, Hansen, Fleming, & Fjelldal, 2015).
Contrary to our prediction, exposure to elevated egg cortisol  temperatures. However, studies should continue to explore the role of maternally derived cortisol, as the dose used in this study may not have matched the severity of environmental stressor (i.e., dose not representative of maternal environmental stress). Future studies could work to determine the exact cortisol dose that matches a given environmental stressor by taking a dose-response approach.
Additionally, it is possible that the prenatal cortisol dose used in this study may have downstream (indirect) impacts on survival at later developmental stages through changes in performance (e.g., morphology, physiology, and behavior), especially when under the effects of a stressful environment. These effects largely remain to be studied in fish within a "stressful" environmental context (although see Capelle, 2017).
Survival from egg to exogenously feeding fry represents an important bottleneck that limits offspring migration success to the lake or ocean (Groot & Margolis, 1991 (Kope et al., 2016).

| Temperature and prenatal cortisol effects on fry structural size and body condition
Morphology, and in particular, body size, impacts juvenile performance metrics such as foraging (Johnsson, 1993), or predator avoidance (Tucker, Hipfner, & Trudel, 2016) that ultimately contribute to variation in fitness. We found that fry raised in elevated temperatures were smaller than under current water conditions, with no apparent temperature effects on body condition. Previous studies indicate that elevated water temperatures generate smaller body size in fish larva and fry (Beacham & Murray, 1990;Murray & Mcphail, 1988;Whitney et al., 2014), presumably via the higher metabolic costs of living in warmer waters (Kingsolver & Huey, 2008;Sheridan & Bickford, 2011)  Previous work has shown that exposure to elevated prenatal cortisol often results in decreased body size in benign environments , but few studies have examined the interactive effects of prenatal cortisol and stressful environments on body size. Atlantic salmon fry whose mothers were exposed to cortisol close to spawning had smaller body length and body mass than control fry when incubated in (+2°C) elevated temperatures at fry stage (Eriksen et al., 2006). Likewise, Chinook salmon exposed to a similar dose of egg cortisol (1,000 ng/ml) as the current study and raised in lower quality water conditions were smaller (Capelle, 2017). However, under a lower dose of egg cortisol (300 ng/ml) and under the same low-quality water conditions fish were larger than higher dose fish suggesting a match between the degree of the prenatal signal and the relative quality of the postnatal environment (Capelle, 2017

| Early life in stressful environments
Wild fish populations are globally in decline (Dulvy, Jennings, Rogers, & Maxwell, 2006), and a number of salmonid species, including Chinook, are currently at population extinction risk, presumably due to the direct and indirect effects of climate change (Crozier, 2015).
Early development in juvenile salmon occurs in a dynamic environment; streams and rivers can change considerably in temperatures both daily and seasonally (Caissie, 2006). Due to this fluctuating tendency and higher surface area to volume, streams and rivers are more likely to be affected by climate change through temperature increases and extreme changes in flow (Hill, Hawkins, & Jin, 2014;. Since environmental variation during development plays a large role in generating variability in offspring phenotypes through developmental plasticity and flexibility, it is important to determine whether juvenile salmon have the capacity to respond to the rapid rate of climate change in their environments. In our study, we examined the role of a stress-induced maternal effect, egg cortisol, as a potential modulator of offspring phenotype and performance in response to elevated water temperatures induced by anthropogenic change. More broadly, we were also able to investigate predictions of the environmental matching framework to test whether maternal signals may modulate offspring responses adaptively in response to future stressful environments. Although we did not find support that elevated egg cortisol led to altered offspring phenotypes or survival during early freshwater stages, it is still possible that exposure to maternal stress modulates phenotypes or performance at later important developmental stages. Further work examining a range of cortisol doses on offspring phenotype and performance under matched environmental conditions is warranted.
Additionally, interpreting the fitness impacts of maternal stress within an environmental context continues to be highly important for determining how maternal effects may assist species' responses to rapid environmental changes.

ACK N OWLED G M ENTS
The authors would like to thank the OMNRF and the Pitcher laboratory for field assistance. K. Janisse, C. Harris, N. Sopinka, P.
Capelle, for their assistance with fish husbandry, and experimen-

CO N FLI C T O F I NTE R E S T
Authors have no conflicts of interest.

AUTH O R CO NTR I B UTI O N S
TRW, CAD, and OPL conceived ideas, and designed methodology; TEP provided rearing facilities; all authors contributed to data collection; TRW, CAD, and OPL ran analyses and led writing of the manuscript with input from all authors.