Acute warming tolerance (CT_max) in zebrafish (Danio rerio) appears unaffected by changes in water salinity

Introduction

With climate warming, average water temperatures are increasing, and in many places, temperature extremes are increasing in frequency, duration and maximum temperatures (Seneviratne et al., 2014). In some aquatic habitats, heat waves may push populations to or above their acute warming tolerance limits. Such situations may already be occurring in some coral reefs (Genin et al., 2020), and freshwater systems (Morgan et al., 2019; Wegner et al., 2008). As the increasingly severe heat waves will increase the impacts of extreme temperatures, there is increased research interest in understanding the mechanisms behind warming tolerance. Portions of this publication were previously published as part of a preprint (Åsheim et al., 2022).

The most commonly used measure of acute warming limits in aquatic ectotherms is the critical thermal maximum test (CT_max), where the water temperature is gradually increased until the animal displays a loss of balance or locomotor function (Becker & Genoway, 1979; Lutterschmidt & Hutchison, 1997; Morgan, Finnøen & Jutfelt, 2018; Morgan et al., 2020; Desforges et al., 2023). The temperature at loss of equilibrium (LOE) is commonly used as the endpoint for actively swimming fish. However, the mechanisms causing the LOE are partly unknown (Andreassen et al., 2022; Ern, Andreassen & Jutfelt, 2023). With a lack of understanding of the underlying physiology, it is also difficult to predict how temperature may interact with other factors (Ern, Andreassen & Jutfelt, 2023; Jutfelt et al., 2024).

One such other factor is water salinity. Salinity stress can occur in aquatic environments through processes such as migration between water bodies, evaporation, ice melt, flooding and precipitation (Kültz, 2015). As climate change affects weather patterns, many fish populations such as those residing in estuaries, rivers and lakes will likely experience changing water salinities (reviewed by Gillanders et al., 2011; Mosley, 2015; Robins et al., 2016; van Vliet et al., 2023).

Examining how water salinity affects warming tolerance can be useful for predicting interactions of factors during heat waves in nature. If water salinity modulates warming tolerance limits, then differences in precipitation or evaporation may affect the warming tolerance of some populations. Additionally, modulation of warming tolerance by other factors may give information on which physiological mechanisms are controlling these traits. There have been a few indications in the recent literature suggesting that water salinity can modulate warming tolerance in both freswhater and seawater fishes, but there has been no consistent direction of effects with salinity manipulation (Fig. 1). Some studies report a reduction in warming tolerance with increased salinity (Morgan et al., 2019; Shaughnessy & McCormick, 2018), some show increased warming tolerance with increasing salinity (King & Sardella, 2017; Metzger, Healy & Schulte, 2016), some show a peak in warming tolerance around some optimal salinity (Haney & Walsh, 2003; Sardella, Sanmarti & Kültz, 2008; Loeb & Wasteneys, 1912), whilst others report no effect (Davis et al., 2019; Hines et al., 2019) (Table S1). Several different mechanisms linking osmotic stress and acute warming tolerance have been suggested to be responsible where CT_max has been affected, such as oxygen availability (via changes in blood haematocrit and lactate) (Shaughnessy & McCormick, 2018), expression of heat shock proteins (Metzger, Healy & Schulte, 2016), and changes in gill permeability causing an osmoregulatory compromise (Haney & Walsh, 2003; Wood & Eom, 2021). However, none of these suggested mechanisms have so far been confirmed.

A field study by Morgan et al. (2019) suggested a reduction in CT_max with higher salinities. In that study, CT_max of wild zebrafish (Danio rerio) correlated negatively with water salinity, with fish from higher salinity streams (measured as conductivity) showing lower warming tolerance. It was, however, not clear if that effect was due to water salinity or some other confounding factor of the sampled river systems. This curious observation that stream salinity appeared to affect warming tolerance, together with the extensive amount of research tools available for the species, suggested that zebrafish could make a good study system for research on the effects of salinity on warming tolerance.

To test the findings by Morgan et al. (2019), to explore the zebrafish as a model species for interactions between salinity and warming tolerance, and to build on the existing body of knowledge on thermal tolerance, we designed an experiment testing CT_max of zebrafish after exposure to different salinity treatments. We were also interested in knowing whether a potential effect is immediate (e.g., due to acutely disturbed plasma ion concentrations), and if such an effect is modulated by longer exposures and acclimation to the water salinity levels. Therefore, we tested fish at two different time points: after two hours and after ten days of salinity exposure.

Materials & Methods

Results

CT_max was not significantly affected by water salinity or treatment time (Table 1, Fig. 2). The mean CT_max for all individuals was 40.8 ± 0.03 °C (Mean ± SE, N = 263). Three fish were showing signs of inactivity after two days in the 5 ppt treatment (lying still on the bottom of the tank), resulting in us lowering the salinity of the highest treatment to 3 ppt for the remainder of the experiment (as described in ‘Methods’). In the highest salinity treatment (3–5 ppt), five (of 71) individuals died during the experiment. These deaths happened on days three (n = 1), seven (n = 1) and ten (n = 3). Furthermore, six individuals in the highest salinity treatment were removed from the CT_max test before reaching CT_max because they displayed abnormal behaviour during heating (see discussion). Four individuals jumped out of the CT_max test tank and were excluded from the analysis (three from the 0 ppt treatment and one from the 1 ppt treatment. Total number of fish included in the analysis was 263.

Discussion

In this experiment, we aimed to test a potential connection between water salinity and warming tolerance as reported in wild zebrafish by Morgan et al. (2019), and to better understand interacting factors on acute warming tolerance. We found no effect on CT_max from a 2-hour exposure to salinities ranging from 0-5 ppt, or a 10-day exposure to salinities ranging from 0-3 ppt (Table 1, Fig. 2).

The highest salinities used (3 and 5 ppt) were 6 and 10 times higher than normal holding salinities for zebrafish (0.5 ppt), and 50 and 70 times higher than what has been observed in the wild for this species (0.06 ppt). While both salinities are comparably high, these salinities are still hypoosmotic to the fish’s internal ion concentrations (approximately 9 ppt), and as such may not require large changes in the osmoregulatory activity of the fish, but only a modulation of their ion- and water intake (Kültz, 2015). Here, we chose relatively low salinities as there is little information on what salinities zebrafish can tolerate long-term. Our observation of three fish lying immobile in one of the 5 ppt tanks after two days indicates that this salinity is too high for long-term survival, but this may also be due to the acute transfer to this salinity. It is possible that the fish could have endured those salinities if more gradually exposed to them. Further studies on interactions between thermal- and osmotic stress in zebrafish would benefit from more research on the ranges of survivable salinities for these fish in the long term, as well as blood measurements to indicate the degree of osmotic stress experienced. Nonetheless, our results show that high, but still hypoosmotic water salinities do not affect acute warming tolerance in zebrafish.

During CT_max testing of fish from the highest salinity treatment (at ten days), an unexpected behaviour occurred in which fish would sometimes perform a sudden burst in swimming, followed by a cramped seizure-like spasm lasting a few seconds that caused loss of equilibrium. This happened at temperatures down to 10 °C lower than the normal CT_max and was first mistaken as their final LOE at CT_max, hence the premature removal of six fish before their true CT_max. This happened to several more individuals and we found that they always recovered after these cramping events and would continue swimming until reaching the normal LOE reaction at similar temperatures to other individuals. The total number of individuals this happened to was not recorded, but it only occurred in the highest salinity treatment after two weeks of exposure and has not been observed in other experiments. These seizure-like events could indicate that the highest salinity treatment had some negative effect on the tested fish and that there could be some connection between acute warming tolerance and water salinity. As these observations were anecdotal, follow-up studies are needed to confirm and quantify this effect.

Acute warming tolerance was previously reported to be negatively correlated with habitat water salinity in a survey of wild zebrafish populations (Morgan et al., 2019; Sundin et al., 2019). That result does not match the observed robustness of CT_max to even larger salinity differences in the current study. We believe the present result is more reliable due to the higher sample size and broader range of salinities. The discrepancy may be due to an additional factor(s), associated with stream salinity that confounded the correlation observed in Morgan et al. (2019), or other factors differing between the wild- and laboratory setting (Morgan et al., 2022). In any case, these results show that variation in salinity beyond what was measured in natural streams by Morgan et al. (2019) does not seem to influence warming tolerance in zebrafish. However, more extreme variations in salinity may occur in freshwater environments due to various events such as heatwaves, droughts, and saltwater intrusion, all of which are increasing in frequency with climate change (reviewed by Mosley, 2015; van Vliet et al., 2023). The excact magnitude of salinity variation during such events varies with a range of environmental factors, but two studies have reported median increases around 21% (Jones & Van Vliet, 2018) and 24% (Graham, Bierkens & Van Vliet, 2024). In our study, we used far greater increases in water salinity and observed no effect on acute warming tolerance, indicating that acute warming tolerance might not be affected by salinity increases during droughts and heatwaves in zebrafish at large. However, our results are likely only applicable for warming challenges that can be resolved in the short term (hours); whether changes in salinity may affect thermal tolerance- and/or performance over longer durations (days) is an open question.

To get a better overview of the diverse relationship between salinity and CT_max in previous studies, we compiled a list of published papers where water salinity was manipulated experimentally, and where CT_max was tested as the outcome variable (Fig. 1. Table S1). Compared to these previous studies, our study adds to the list of studies showing no effect, together with Davis et al. (2019) and Hines et al. (2019). A notable difference between the studies showing an effect and those not showing one is the salinity tolerance of the species used. This study and the study by Hines et al. (2019) used relatively low water salinities (up to 5 ppt and 12 ppt), while studies reporting an effect (Haney & Walsh, 2003; King & Sardella, 2017; Metzger, Healy & Schulte, 2016; Sardella, Sanmarti & Kültz, 2008; Shaughnessy & McCormick, 2018) used far higher ranges (up to 25 ppt, and up to 60 ppt in one study). The studies showing an effect all used highly euryhaline species and acclimation times of at least one week, indicating that the effects on warming tolerance may be connected to salinity acclimation itself or through some other general stress response. Still, the study by Hines et al. (2019) found no effect despite using salinities up to 30 ppt and acclimation times up to 600 days. Overall, looking at the published literature on the subject, it looks like the association between salinity and warming tolerance is highly complex and species- and context-dependent.

Conclusions

We explored whether the acute warming tolerance of zebrafish is affected by water salinity. We found no difference in CT_max between fish tested at water salinities ranging from 0 to 5 ppt in the short term (2 h) or from 0 to 3 ppt in the long term (10 days). These results indicate that there is little interaction between acute warming tolerance and environmental salinity at these salinity ranges in zebrafish, and that variation in salinity levels encountered by wild zebrafish might not impact their acute warming tolerance.

Supplemental Information