Surf smelt accelerate usage of endogenous energy reserves under climate change

In the Salish Sea’s Puget Sound, elevated temperatures can be driven by human activities. Regionally, approximately one third of the coastline has been modified by humans [47]. Human activities and development alter surf smelt spawning grounds by reducing nearshore vegetation and the shading it provides, causing a significant increase in beach gravel tempera- ture and concurrent decreases in gravel humidity [7]. A survey of modified and natural beaches in the Puget Sound found that modified beaches had an average substrate temperature of 18.8˚C, which was 4.7˚C higher than the average temperature of natural beaches, and a tem- perature known to cause increased surf smelt embryo mortality [7]. The observed temperature increase caused by shoreline modification was represented by the high temperature treatment (18˚C) in this study. It is likely that contemporaneous shoreline modifications and subsequent temperature effects are realigning surf smelt energy budgets and, potentially, the timing of hatching [7, 10]. Ongoing shoreline development coupled with the large proportion of already developed beaches in the Salish Sea poses an immediate threat as surf smelt rely on a relatively small proportion of beaches for spawning each year, and disturbance to any of them could have population-level implications [8]. Additionally, climate models predict that Salish Sea seawater temperatures will increase by approximately 1.2˚C by the year 2040 [48]. This tem- perature increase was captured in this study as the 15˚C treatment and was found to reduce

yolk size by 22% compared to ambient seawater. Thus, findings here suggest that global cli- mate change will accelerate surf smelt endogenous energy usage in the northeast Pacific Ocean. We observed little direct effects of elevated pCO2 on early life history stages of surf smelt, which is in agreement with recent findings on another Salish Sea forage fish, Clupea pallasii [40]. Embryo heartrate, larval yolk size, and oil globule size were not directly affected by varia- tion in pCO2, despite the very high concentrations used in this study. Only surf smelt embryos were directly affected by elevated pCO2, with a significant positive effect on Y:E. A review of the effects of OA on marine fish metabolic rates found that 21% of studies reported a consis- tent reduction in metabolic rate in response to OA [49]. The positive effect of elevated pCO2 on surf smelt embryos and the absence of a pCO2 treatment effect on the surf smelt larval met- rics may reflect the unique spawning behavior and habitat use of surf smelt, and an overall robustness to OA sensitivity. In comparison to open-ocean species, marine fish that spawn in the intertidal or near-shore habitat possess broad tolerance to environmental variability due to the high frequency and magnitude of environmental change [7, 40, 50, 51]. With respect to habitat-dependent CO2 sensitivity, Baumann [52] called this the Ocean Variability Hypothesis. The surf smelt embryos used here were collected from a shallow embayment that experiences highly variable and ephemeral pH, ranging from � 6.5 to > 8.5, with the majority of observa-

tions between 7.15 to 8.45 [53]. Thus, the lack of a strong OA effect on surf smelt, even on early life histories, can be expected. While this study shows that the utilization of endogenous energy reserves in early life-his- tory stages of surf smelt are affected by elevated temperature and, to a lesser degree, pCO2, it remains unknown whether these effects will be detrimental to later life history stages, or in annual population abundance and recruitment. Despite this limitation, this study highlights the importance of including surf smelt in ecological and climate change research given the evi- dence provided here that surf smelt early life stages are affected by stressors associated with near-present climate change and under conditions associated with current nearshore urban development. Given their keystone role in marine ecosystems, reductions to forage fish fitness and population abundance have the potential to cause ecosystem-wide effects. For example, ecosystem models show that Puget Sound seabird populations are negatively affected by reduc- tion in forage fish biomass [5, 54] and piscivorous fish and marine mammals may share similar vulnerability to changes in forage fish abundance [4, 5]. Finally, surf smelt remain one of the PLOS ONE Surf smelt response to climate co-stressors PLOS ONE | https://doi.org/10.1371/journal.pone.0270491 June 27, 2022 10 / 14