ABSTRACT
Testing acclimation plasticity informs our understanding of organismal physiology and applies to conservation management amidst our rapidly changing climate. Although there is a wealth of research on the plasticity of thermal and hydric physiology in response to temperature acclimation, there is a comparative gap for research on acclimation to different hydric regimes, as well as the interaction between water and temperature. We sought to fill this gap by acclimating western fence lizards (Sceloporus occidentalis) to experimental climate conditions (crossed design of hot or cool, dry or humid) for 8 days, and measuring cutaneous evaporative water loss (CEWL), plasma osmolality, hematocrit and body mass before and after acclimation. CEWL changed plastically in response to the different climates, with lizards acclimated to hot humid conditions experiencing the greatest increase in CEWL. Change in CEWL among individuals was negatively related to treatment vapor pressure deficit and positively related to treatment water vapor pressure. Plasma osmolality, hematocrit and body mass all showed greater changes in response to temperature than to humidity or vapor pressure deficit. CEWL and plasma osmolality were positively related across treatment groups before acclimation and within treatment groups after acclimation, but the two variables showed different responses to acclimation, suggesting that they are interrelated but governed by different mechanisms. This study is among few that assess more than one metric of hydric physiology and that test the interactive effects of temperature and humidity. Such measurements will be essential for predictive models of activity and survival for animals under climate change.
INTRODUCTION
Water availability and temperature are two interrelated drivers of physiological adaptations and plastic responses in terrestrial animals (Bodensteiner et al., 2020; Cox and Cox, 2015; Domínguez-Guerrero et al., 2021; Hochachka and Somero, 2002; Rozen-Rechels et al., 2019). In extreme heat, both endotherms and ectotherms use evaporative cooling to thermoregulate (DeNardo et al., 2004; Fuller et al., 2021; Pirtle et al., 2019). However, osmotic homeostasis in terrestrial organisms is challenged by water loss to the environment, which can lead to facultative hyperthermia in birds (Gerson et al., 2019) and reduced activity (Davis and DeNardo, 2009) and low selected temperatures in lizards (Pintor et al., 2016; Sannolo and Carretero, 2019). Water availability is a primary limiting factor of the geographic distribution of wild animals (Dunkin et al., 2013; Kearney et al., 2018), yet the amount of research into hydric physiology lags far behind the wealth of information on thermal physiology (Rozen-Rechels et al., 2019; Taylor et al., 2020). Rapid global change is testing the physiological limits of wild animals, highlighting the importance of understanding the extent to which animals can plastically respond to dramatically altered thermal and hydric regimes (Gunderson and Stillman, 2015; Huey et al., 2012; Kearney et al., 2018; Riddell et al., 2018; Seebacher et al., 2015; Somero, 2010; Stillman, 2003; Urban et al., 2014).
There is a wealth of evidence supporting the plasticity of thermal and hydric physiology in response to temperature acclimation, especially for ectotherms that depend on the environment to regulate body heat. Lizards acclimated to higher temperatures may have thermal preferences that are lower (Ryan and Gunderson, 2021; Wilhoft and Anderson, 1960) or higher than lizards acclimated to cooler temperatures (Ryan and Gunderson, 2021). In lizards, the critical thermal maximum and/or minimum can be impacted by acclimation temperature (Llewelyn et al., 2018), egg incubation temperature (Abayarathna et al., 2019; Llewelyn et al., 2018) or acute heat shock (Deery et al., 2021). There is also evidence of acclimation at the molecular level: grasshoppers raised in hot conditions experienced lipid phase change at higher temperatures (Gibbs et al., 1991). For hydric physiology, acclimation to high temperatures leads to decreased evaporative water loss in birds (Muñoz-Garcia et al., 2008; but see McKechnie and Wolf, 2004), lizards (Vicenzi et al., 2021) and salamanders (Riddell et al., 2019). Animals from Drosophila to lizards that are incubated at higher temperatures, then raised in common temperatures after hatching, show higher desiccation rates as adults than individuals incubated at lower temperatures (Llewelyn et al., 2018; Parkash et al., 2014). Finally, on a short time scale of only 2 h, warm-acclimated lizards had lower evaporative water loss than cool-acclimated lizards (Oufiero and Van Sant, 2018), whereas a similar study found the opposite effect for toads (Senzano and Andrade, 2018).
There is comparatively less research into how thermal and hydric physiology respond to hydric acclimation. This could be due to the comparable difficulty of modulating and measuring hydric factors: whereas environmental temperature can be changed and measured directly, environmental water could refer to water vapor, free-standing drinking water, or water from the diet and metabolism, each posing a unique set of methodological challenges. Most studies modulate drinking water or relative humidity, or harness environmental differences in aridity. For example, acclimation to humid conditions resulted in higher evaporative water loss than acclimation to dry conditions for birds (Clement et al., 2012; Muñoz-Garcia et al., 2008), lizards (Kattan and Lillywhite, 1989; Kobayashi et al., 1983; Weaver et al., 2022) and snakes (Moen et al., 2005). Finally, while thermal ecology and physiology metrics have become extraordinarily streamlined (Taylor et al., 2020), hydric physiology remains understudied. The most prevalent metric of hydric physiology, evaporative water loss, is usually measured with hygrometry or estimated based on mass loss (Le Galliard et al., 2021). Both estimates largely preclude our understanding of the contribution of different tissues (e.g. skin, eyes, respiration, cloaca). Hydration state metrics, including body mass, plasma osmolality, hematocrit, total body water and percentage wet mass, also vary in what exactly they measure and how much they align. Thus, our aim was to assess the effects of temperature and humidity acclimation on several metrics of hydric physiology and to focus specifically on the skin as a major organ of evaporative water loss.
The skin is a primary site of osmoregulation (Pirtle et al., 2019) and cutaneous evaporative water loss can be much greater than respiratory water loss (Cohen, 1975; McKechnie and Wolf, 2004; Tieleman and Williams, 2002). When the environment is dry, water moves down its vapor pressure gradient, from organisms, across their skin and into the environment (Campbell and Norman, 1998; Monteith and Campbell, 1980; Spotila and Berman, 1976). Thus, within a species at acute time scales, evaporative water loss tends to be higher in low relative humidity (Cooper and Withers, 2008; Pintor et al., 2016; Shoemaker and Nagy, 1977; Warburg, 1965; Willis et al., 2011) and at high temperatures (Dmi'el, 2001; Dmi'el and Tel-Tzur, 1985; Riddell et al., 2019; Toolson and Hadley, 1979; Vicenzi et al., 2021). We quantify this desiccation stressor as vapor pressure deficit, the difference between saturation vapor pressure (how much water vapor the air could hold) versus actual vapor pressure (how much water vapor is in the air at that time) at a given temperature. Thus, vapor pressure deficit incorporates both temperature and water vapor to assess the biophysical force driving the rate of evaporation. There is also evidence that skin becomes more resistant to water loss when vapor pressure deficit increases (Cooper and Withers, 2014; Eto et al., 2017; Riddell and Sears, 2015; Withers and Cooper, 2014), indicating that some organisms physiologically counteract the biophysical effect of vapor pressure deficit to help maintain hydration. The goal of our study was to assess whether the skin permeability of lizards changes to physiologically compensate for desiccation stress.
We investigated the acclimation of cutaneous evaporative water loss (CEWL) and related changes in hydration for a lizard in response to different combinations of temperature and relative humidity. This experiment is among the first to test plasticity in response to both temperature and humidity, and to incorporate several metrics of hydric physiology. We used a factorial design with high and low temperatures and humidities, each resulting in a unique water vapor pressure and vapor pressure deficit, to assess the acclimation capacity of skin resistance to water loss, measured as CEWL. CEWL was measured at common conditions before and after exposure to different climate treatments, with any differences therefore representing physiological acclimation. We hypothesized that CEWL would change in response to experimental climate conditions, and lizards exposed to the highest vapor pressure deficit would have the lowest CEWL at the end of the experiment to conserve water in response to desiccation stress. To assess this, we acclimated western fence lizards (Sceloporus occidentalis) to four different climate treatments and examined the plasticity of CEWL and related changes in hydration state, measured as plasma osmolality, hematocrit and body mass. If skin permeability can acclimate to experimental climate conditions, then we expect to find a negative relationship between treatment vapor pressure deficit and CEWL, whereby the lizards exposed to the greatest vapor pressure deficit will decrease skin permeability to water loss and have the lowest CEWL at the end of the experiment.
MATERIALS AND METHODS
Baseline measurements
We used hand-held lassos to capture 138 adult male Sceloporus occidentalis Baird and Girard 1852 throughout the campus of California Polytechnic State University in San Luis Obispo, CA, USA on five dates in June–August of 2021. In the lab, we recorded snout–vent length (SVL; ±0.5 mm) and mass (±0.1 g), drew blood from the postorbital sinus of the right eye using self-sealing heparinized micro-hematocrit capillary tubes (Clay Adams, Becton Dickinson, Sparks, MD, USA) and measured CEWL (g m−2 h−1) using a handheld AquaFlux evaporimeter (model AF200; BioX Systems, London, UK). The AquaFlux measures CEWL as transepidermal water loss of a 3 mm diameter area within a closed chamber sealed against the skin, thus excluding the influence of immediate external physical factors such as ambient vapor pressure deficit. This medical-grade device is calibrated before each measurement session and has extremely high repeatability (Imhof et al., 2014). It measures the instantaneous movement of water across the skin, enabling CEWL measurements that specifically answer questions related to physiological acclimation of skin permeability. Additional explanations and use of the AquaFlux evaporimeter are available (Elkeeb et al., 2010; Imhof et al., 2009; Lourdais et al., 2017; Weaver et al., 2022). The AquaFlux also measures ambient temperature and relative humidity at the time of measurement, which we used to assess covariates of CEWL. Blood samples were centrifuged in a micro-hematocrit centrifuge (model IEC MB; Damon IEC Division, Thermo Fisher Scientific, Waltham, MA, USA) for 2 min, then hematocrit (±1%) was recorded. Hematocrit is the percentage of red blood cells in blood; the remaining percentage is mostly plasma, which is water based, and its relative quantity might be useful to infer hydration state. Plasma osmolality (±3 mmol kg−1) was measured in 1–4 technical replicates using a vapor pressure osmometer (VAPRO, model 5600; Wescor, ELITech, Logan, UT, USA). Plasma osmolality refers to the concentration of solutes in the blood; dehydrated animals should have high plasma osmolality. CEWL was measured in 4–6 technical replicates in the same region of the mid-dorsum while lizards were held by their forelimbs to minimize conductive heat gain from the researcher's hands. Immediately after measuring CEWL, we recorded lizard internal body temperature (±1°C) by inserting a thermocouple into the cloaca (Traceable Pocket-size K-type Thermocouple, model 14-649-81; Thermo Fisher Scientific). All procedures were approved by Cal Poly IACUC (protocol #2103) and permitted by the California Department of Fish and Wildlife (SC-192880002).
Experimental acclimation
We ranked lizards by mass to systematically assign individuals to one of four climate treatments: cool humid, hot humid, cool dry or hot dry. Ordering lizards by mass then alternating their assignment ensured equivalent average starting mass among treatment groups (mean±s.d.: cool humid: 11.6±1.4 g; hot humid: 11.6±1.7 g; cool dry: 11.8±1.6 g; hot dry: 12.0±1.7 g). We ran the 8-day acclimation experiment in five separate trials, with a total of 134 lizards completing acclimation (two lizards died during acclimation and two lizards were removed from treatment owing to extreme rates of mass loss). We chose an 8-day acclimation because we previously determined this to be the limit of how long these lizards could maintain good health while fasting, and other studies have elicited acclimatory changes in lizards within the same period (Kattan and Lillywhite, 1989; Weaver et al., 2022). While 8 days is unlikely to encapsulate a full shedding cycle, several lizards shed at some point during or soon after the acclimation study. Lizards were kept in individual transparent plastic containers (25×15×13 cm) in an environmental chamber (Low Temperature Illuminated Incubator, model 818; Thermo Fisher Scientific) maintained at ∼24°C for the cool treatment, which we estimate to be the low end of ambient temperatures when S. occidentalis are active, and ∼35°C for the hot treatment, which is the thermal preference for S. occidentalis (McGinnis, 1966; Wilhoft and Anderson, 1960), with a 12 h light:12 h dark photoperiod (Table 1). Every lizard had a sponge in their container, which was kept wet for lizards in humid treatments and dry for lizards in the dry treatments. Acclimation temperature and relative humidity inside six of the containers for each trial group were recorded every 30 min (HOBO External Temperature/RH Sensor Data Loggers, model MX2302A; Onset Computer, Bourne, MA, USA). Each probe sensed both temperature and relative humidity, and values were consistent across probes when recording the same conditions, but we also rotated probes across treatments to account for any variation. See Table 1 for the resulting temperature, relative humidity, water vapor pressure and vapor pressure deficit values. Compared with 24 h weather recorded on the Cal Poly campus (Irrigation Training and Research Center, 2021) during the active season of S. occidentalis, only the hot humid treatment conditions are extremely different from what these lizards might experience in the wild (Fig. 1). We refer to our treatments as ‘dry’ and ‘humid’ for their relative values; our dry treatments were only ∼18% and 34% relative humidity (Table 1), while truly dry conditions would be closer to 5–10% relative humidity. However, we were still able to achieve vapor pressure deficits for dry treatments that were at the high end for what S. occidentalis might experience in the wild (Fig. 1) and lizards were exposed to those conditions continuously, compared with their ability to seek refuge and avoid similarly dry conditions in the wild. Treatment conditions varied slightly among trials (Table S1), demonstrating the importance of recording actual experimental conditions in replicates despite using identical chambers and methods. Therefore, we included trial as a random factor in our acclimation models to account for this variation. Containers had 14 holes, each 4 mm in diameter, along the top for air flow and a wire partition to prevent direct access to the sponges. Containers were kept tilted so any pooled water from sponges was behind the wire partition and could not be accessed by the lizards.
In addition to the baseline measurements taken after capture on day 0 of acclimation, we again recorded mass on days 4–8, drew blood on days 4 and 8, and measured CEWL on day 8. All CEWL measurements for all treatment groups, before and after acclimation, were taken at common lab conditions: 26.7±0.8°C (mean±s.d.); 47±7% relative humidity; 1.9±0.3 kPa vapor pressure deficit. Because measurements were taken under common conditions and the evaporimeter measures instantaneous water flux across the skin, CEWL assesses skin permeability to water loss. Any changes in CEWL measured for a given lizard are therefore due to physiological acclimation. Even so, ambient temperature and humidity in the lab varied slightly on a given measurement day, so we also examined whether lizard body temperature, ambient temperature and ambient vapor pressure deficit at the time of measurement impacted post-acclimation CEWL. Lizards were fasted during acclimation, so we gave them ad libitum access to crickets and water prior to toe-clipping (to prevent recapture) and release at their sites of capture.
Statistical analysis
We only measured CEWL before and after acclimation, so we calculated ΔCEWL as the difference between CEWL measured after the experiment versus before for each lizard. The LMMs for ΔCEWL included trial as a random effect. Plasma osmolality, hematocrit and body mass were measured three or more times throughout the experiment, so the LMMs included the interaction of time in acclimation (numeric days) and nested random effects of trial and individual lizard ID. We also used two-sided t-tests to determine whether ΔCEWL values were different from zero. Owing to the visual nonlinearity in plasma osmolality values throughout the experiment, for each treatment group, we ran an ANOVA followed by Tukey's honest significant difference test to quantify pairwise differences in plasma osmolality among experiment days within treatment groups.
We quantified the effect sizes of acclimation by creating LMMs for each of CEWL, plasma osmolality, hematocrit and body mass, with time in acclimation (numeric days), treatment group (categorical cool humid, hot humid, cool dry, hot dry) and their interaction as predictors. The LMM for CEWL included trial as a random effect, and the other LMMs included nested random effects of trial and individual lizard ID. We used these models to estimate the marginal means of each linear trend. We also created LMMs for the day 8 measurements of each response variable after the experiment, with categorical treatment as the predictor and trial as a random effect. These models were used to estimate the marginal means for each treatment group at the end of acclimation.
Body temperature is typically a covariate of CEWL, so we created a LMM of the day 8 CEWL measurements with body temperature at measurement, categorical treatment and their interaction as predictors, with trial as a random effect; this will determine the impact of body temperature at the time of measurement versus acclimation effects on CEWL. We ran the same model with ambient temperature, water vapor pressure and vapor pressure deficit at the time of measurement.
We created an additional LMM for ΔCEWL with each of numerical water vapor pressure and vapor pressure deficit across trials and treatments as the predictor to assess which could explain the observed changes in CEWL across individual lizards. We tested both linear and polynomial effects and compared models using AIC values.
We tested whether different metrics of hydric physiology were interrelated by running a simple linear regression on baseline CEWL and plasma osmolality across lizards upon capture. To determine whether a relationship persisted post-acclimation, we also created LMMs of day 8 CEWL and plasma osmolality for each treatment group, with trial as a random effect. None of the models we present meaningfully deviate from assumptions of linearity, normality and equal error variance.
Models used a cleaned dataset consisting of averaged technical replicates for plasma osmolality and CEWL. Any outliers from technical replicate groups that were identified based on boxplot distributions were removed. For plasma osmolality, all values except one extreme outlier were retained. Most CEWL technical replicate groups (per lizard per date) had one outlier omitted, and the maximum number of measurements omitted per replicate group was two. All data analyses for plasma osmolality and CEWL used the average of the remaining replicates after outlier exclusion. We also removed one CEWL outlier at the measurement level for a lizard that was actively shedding and had CEWL far above the normal measurement range, because ecdysis is associated with exceptionally high evaporative water loss (Bogert and Cowles, 1947; Claussen, 1967; Cohen, 1975). We removed 10 plasma osmolality measurements because of a technical error with the osmometer.
All statistics and figures were done in R v.4.2.2 (https://www.r-project.org/) using tidyverse workflow (Wickham, 2022). See Table S2 for a list of R functions and packages used. ANOVAs were run with type 3 partial sum of squares and Kenward–Roger degrees of freedom. Data and code are archived on Zenodo (https://doi.org/10.5281/zenodo.10018753).
RESULTS
The best model to explain Δ CEWL during acclimation was the humidity treatment model (Table 2). Lizards in Humid treatments had the greatest increases in CEWL following acclimation: on average, CEWL increased 1.92 g m−2 h−1 day−1 for lizards in the hot humid treatment, versus 1.18 for cool humid lizards, 0.46 for hot dry lizards and 0.45 for cool dry lizards (Fig. 2). ΔCEWL was greater than zero for each treatment group (t-tests: cool humid: estimate=9.51, t(32)=8.75, P<0.0001; hot humid: estimate=15.35, t(32)=8.73, P<0.0001; cool dry: estimate=3.59, t(32)=2.75, P=0.01; hot dry: estimate=3.70, t(33)=2.51, P=0.02). ΔCEWL had a negative linear relationship with treatment vapor pressure deficit (Fig. 3A). The linear model of ΔCEWL and vapor pressure deficit satisfied assumptions of linearity, normality and equal error variance, with no high leverage or influential points. ΔCEWL also had a positive linear relationship with treatment water vapor pressure (Fig. 3B). The water vapor pressure model also satisfied assumptions. Polynomial models did not better explain either relationship (ΔAIC<2 compared with linear relationship of the same variable).
Ambient temperature at the time of measurement had an effect on post-acclimation CEWL measurements (LMM: SS=310, F1,59=9.24, P=0.003; treatment effect: SS=57, F3,122=0.57, P=0.6), but the effect of ambient temperature did not differ among treatment groups (interaction effect: SS=65, F3,122=0.65, P=0.6). We measured no effect of body temperature (LMM body temperature effect: sum of squares (SS)=1, F1,115=0.03, P=0.9; treatment group effect: SS=15, F3,123=0.13, P=0.9; interaction effect: SS=28, F3,123=0.25, P=0.9), water vapor pressure (LMM ambient water vapor pressure effect: SS=141, F1,9=4.08, P=0.07; treatment effect: SS=175, F3,122=1.68, P=0.2; interaction effect: SS=130, F3,122=1.25, P=0.3), or vapor pressure deficit (LMM ambient vapor pressure deficit effect: SS=115, F1,32=3.36, P=0.08; treatment effect: SS=113, F3,122=1.1, P=0.4; interaction effect: SS=153, F3,121=1.49, P=0.2) on CEWL at the time of measurement.
Change in plasma osmolality, hematocrit and body mass throughout acclimation were best explained by the model including time in acclimation and acclimation temperature (Table 2). Although plasma osmolality is visually nonlinear (Fig. 4A), pairwise comparisons between measurement days for each treatment group showed no differences (all P>0.05; Table S3). On average, plasma osmolality increased 1.59 mmol kg−1 day−1 for lizards in the hot dry treatment, 0.48 for hot humid, 0.13 for cool dry, and decreased 0.65 for lizards in the cool humid treatment (Fig. 4A). Hematocrit decreased in all lizards throughout acclimation, with lizards in hot treatments decreasing slightly faster than lizards in cool treatments: on average, hematocrit decreased 1.55% day−1 for lizards in the hot humid treatment, versus 1.36 for hot dry, 1.17 for cool humid, and 1.14 for lizards in the cool dry treatment (Fig. 4B). All lizards also decreased in body mass throughout acclimation: on average, body mass decreased 0.31 g day−1 for lizards in the hot dry treatment group, 0.24 for hot humid, 0.16 for cool dry, and 0.12 for cool humid (Fig. 4C).
At the completion of the experiment, CEWL was highest for lizards in the hot humid treatment, followed by lizards exposed to cool humid conditions; lizards exposed to dry conditions had the lowest final CEWL, with no difference between the two dry treatment groups (Fig. 5A). Plasma osmolality did not differ among treatment groups at the completion of the experiment (Fig. 5B). Post-acclimation hematocrit and body mass were impacted only by temperature treatment: both hematocrit and body mass were higher for lizards exposed to cool conditions than for lizards exposed to hot conditions (Fig. 5C,D).
Upon capture of lizards, plasma osmolality explained a significant amount of the variation in baseline CEWL, with increases in plasma osmolality accompanied by small increases in CEWL (Fig. 6). The relationship between CEWL and plasma osmolality remained post-acclimation only for lizards in the cool dry treatment (LMM: estimate=0.07, s.e.=0.03, F1,28=4.88, P=0.04; other treatment groups P>0.05).
DISCUSSION
Our results indicate that exposure to certain climatic conditions can affect lizard CEWL, with skin resistance to water loss demonstrating an acclimatory response based on both water vapor pressure and vapor pressure deficit. We measured physiological changes in response to humidity and temperature within just 8 days. CEWL for humid-acclimated lizards increased to levels 2–4 times that of dry-acclimated lizards (Fig. 2), a result consistent with previous humidity acclimation studies (Clement et al., 2012; Kattan and Lillywhite, 1989; Kobayashi et al., 1983; Moen et al., 2005; Muñoz-Garcia et al., 2008; Weaver et al., 2022). ΔCEWL was explained best by the humidity treatment to which lizards were exposed (Table 2), suggesting that humidity exerts an influence on skin permeability to water loss. It is possible that binary humidity treatment explained ΔCEWL better than the numeric vapor pressure deficit model because humidity affects the skin in ways not measured through physical vapor pressure deficit. Lizards in the hot humid treatment had the largest increase in CEWL, but were not exposed to the lowest vapor pressure deficit (Fig. 2). While higher CEWL for hot-acclimated lizards could be related to additive effects of heat and humidity, CEWL was measured at ∼27°C, which was much cooler than the hot acclimation temperature. Thus, neither evaporative cooling (Case, 1972; Fuller et al., 2021) nor skin lipid organization (Champagne et al., 2016; Gibbs, 2002; Rourke and Gibbs, 1999; Williams et al., 2012), could explain the especially high ΔCEWL for lizards in the hot humid group. Alternatively, lizards in the hot humid treatment may have exhibited the greatest acclimation response (Fig. 2) because they were exposed to extreme conditions that differ from what they experience in the wild (Fig. 1).
The large increases in CEWL we measured are surprising, assuming that it would be beneficial to conserve water in all conditions. High humidity could be a cue that water is replaceable, and if skin resistance to water loss is energetically costly, then reducing skin resistance might be beneficial. Studies have yet to quantify the energetics of skin permeability or to test how humidity might signal skin permeability changes. We predicted that CEWL would decrease for the lizards acclimated to dry conditions, but it slightly increased (Fig. 2). It is possible that the dry conditions in this experiment did not increase vapor pressure deficit enough for CEWL to decrease; the vapor pressure deficits we exposed lizards to were in the upper half of values S. occidentalis might experience in the wild, but they might experience vapor pressure deficits 1.5–2 times higher than our treatment conditions (Fig. 1). A previous study successfully elicited a decrease in CEWL for lizards exposed to dry conditions (Weaver et al., 2022) but these authors measured post-acclimation CEWL at the same temperature as the acclimation temperature. The measurement temperature in this study was higher than the cool and lower than the hot acclimation temperatures, and that temperature difference could influence CEWL (Dmi'el, 2001; Riddell et al., 2019; Vicenzi et al., 2021). Although we did find a relationship between post-acclimation CEWL and ambient temperature at the time of measurement, the effect of ambient temperature was the same across treatment groups. CEWL values at the end of the experiment reflect the same pattern as the acclimation response (Fig. 2, Fig. 5A) and these results support our hypothesis that CEWL is plastic in response to climate acclimation.
The evaporimeter we used measures instantaneous skin permeability and not total water lost over an extended period, so our results provide strong evidence that lizard skin physiologically acclimated to our climate treatments. The negative relationship between vapor pressure deficit and ΔCEWL (Fig. 3A) indicates that desiccation stress leads to lower skin permeability to water loss. However, we also measured a strong positive relationship between water vapor pressure and Δ CEWL (Fig. 3B), which suggests that the amount of water in the air could also affect changes in skin permeability. Our experiment was not designed to tease apart whether vapor pressure deficit or ambient vapor pressure is the biophysical cue that lizards use to adjust skin permeability, but these relationships provide a foundation for further investigation. The acclimation changes we measured would be beneficial to lizards because lower water loss rates when there is limited water in the environment (i.e. low water vapor pressure) or when desiccation stress is high (i.e. high vapor pressure deficit) would help conserve water, thereby reducing lethal desiccation risk. Acclimation changes in CEWL are mostly attributed to the concentration and composition of lipids in the skin; for example, a higher concentration of ceramides versus cerebrosides is associated with lower CEWL rates (Cox et al., 2008; Haugen et al., 2003a). Lipids also play a role in CEWL differences among populations and species, with lower water loss rates linked to higher overall lipid concentration, higher percentage of ceramides, longer hydrocarbon chains and more saturated fatty acids (Haugen et al., 2003b; Lillywhite, 2006; Muñoz-Garcia and Williams, 2005; Williams et al., 2012). Changes in blood flow have been implicated for more acute changes in CEWL (Marder and Raber, 1989) and these changes could theoretically persist during measurement if acclimation involved mechanisms like angiogenesis or changes in blood volume.
For changes in plasma osmolality, hematocrit and body mass, temperature was a more influential variable than humidity or vapor pressure deficit (Table 2), likely because temperature has a direct effect on metabolism (Angilletta, 2009) and therefore overall water loss. The lizards in our experiment had no drinking water to replenish what was used, so they became dehydrated and plasma osmolality increased, which is similar to other results showing increased plasma osmolality in response to water restriction (Dupoué et al., 2017a). Change in plasma osmolality across treatment groups was not directly related to the vapor pressure deficit lizards experienced (Fig. 4A), but it is possible that within temperature treatments, respiratory evaporative water loss was minimized in humid conditions (Lovegrove et al., 2014; Warburg, 1965). Notably, plasma osmolality was not statistically different between any of the treatment groups at the end of the acclimation experiment (Fig. 5B). On average, we measured 1–13 mmol kg−1 change in plasma osmolality over the course of acclimation, while we found S. occidentalis to have a natural plasma osmolality range of ∼100 mmol kg−1 among individuals. Considering our relatively extreme experimental climates (Fig. 1) and the naturally high variation in plasma osmolality, we hypothesize that dehydration effects on plasma osmolality might take longer than 8–10 days to observe.
Because lizards were fasted, hematocrit and body mass declined in all individuals (Fig. 4B,C). Blood draws were also likely too frequent and too large of volume for red blood cells to be replenished, especially considering the lack of natural sunlight which contributes to erythropoiesis via vitamin D (Smith and Tangpricha, 2015). Lizards acclimated to hot conditions decreased body mass by 1.5–2.5 times as much as lizards acclimated to dry conditions (Fig. 4C). At the completion of the experiment, body mass was significantly lower for hot-acclimated lizards than cool-acclimated lizards (Fig. 5D), likely due to the elevated metabolism and thus the increased total and respiratory water loss, that they experienced. This could also be related to fat and muscle catabolism to maintain the relatively consistent plasma osmolality values we measured (Fig. 5B; Brusch et al., 2018). The evaporimeter we used measures CEWL as instantaneous skin permeability to water loss, but cumulative water lost might be better estimated with a decline in body mass. Even though skin permeability to water loss can change (Fig. 2), lizards likely still lost more water when experiencing high vapor pressure deficits (Fig. 4C). Other studies should initiate tests to determine the relationship between instantaneous skin permeability to water loss versus cumulative water/mass loss, especially because each is referred to as evaporative water loss (Le Galliard et al., 2021; Žagar et al., 2022).
We hypothesized that CEWL and plasma osmolality would be related, given that they are each a way to measure organismal water balance. We had predicted a negative relationship between their baseline values such that lower hydration (high plasma osmolality) would be associated with water conservation (low CEWL), as indicated in toads (Anderson and Andrade, 2017). However, CEWL had a significant positive relationship with plasma osmolality (Fig. 6), which could be due to lizards with higher CEWL becoming dehydrated, thus having higher plasma osmolality as a result of their higher water loss rates. The relationship detected between CEWL and plasma osmolality in this study is small (Fig. 6), but it applies to well-hydrated lizards at the time of capture, where a causal, positive relationship between CEWL and osmolality is reasonable. This relationship was not detected previously (Weaver et al., 2022), but our study is among very few that measured both hydration state and evaporative water loss rates (see also Dupoué et al., 2017b; Kattan and Lillywhite, 1989).
Assuming a relationship between CEWL and plasma osmolality, we also expected their changes during acclimation to mirror each other, with the group that decreased CEWL the most having the largest increase in plasma osmolality. However, we did not find evidence to support this (Figs 2 and 4A). It is possible that the two hydric physiology factors are ruled by different mechanisms: water and food consumption could primarily affect plasma osmolality whereas ambient climate could primarily affect CEWL (Kattan and Lillywhite, 1989). Hydration is likely impacted by metabolism (Angilletta, 2009) and while there is evidence that respiratory evaporative water loss is also tied to metabolism (Welch and Tracy, 1977; Wolf and Walsberg, 1996), CEWL might not be. The same positive trend between CEWL and plasma osmolality held for values after acclimation, but the relationship was only significant for one treatment group. There could also be a time lag effect that complicates measurement of the relationship. In our study, plasma osmolality was maintained despite dehydrating conditions (Fig. 5B), which could be enabled by fat or muscle catabolism (Brusch et al., 2018), thereby enabling acclimation effects of CEWL unconstrained by short-term dehydration. Organismal water could also be moderated through water resorption from feces and behavioral modifications such as closing the eyes and mouth (Karasov, 1983; Pirtle et al., 2019; Waldschmidt and Porter, 1987). Future studies should incorporate measures of both hydration and water loss to further assess their relationship and what type of water availability (i.e. humidity versus water in the diet) each responds to.
Conclusion
Acclimation to different humidity and temperature regimes led to significant changes in CEWL and related effects on hydration in a lizard in only 8 days. Our data support our hypothesis that skin permeability to water loss is plastic, and this plasticity could enable resilience in the face of climate change. However, we did not measure the rapidity or reversibility of these acclimation changes, or whether plasticity differs among S. occidentalis populations based on habitat, and these studies will also be essential to inform conservation management. Temperature-dependent effects of metabolism appear to have the greatest impact on hydration, measured as plasma osmolality, hematocrit and body mass. Finally, baseline variability in CEWL was significantly explained by plasma osmolality, but the two variables responded differently to experimental climate acclimation. Future studies should incorporate both variables to determine how their interplay might affect physiological responses to environmental change. The mismatch we observed between CEWL and plasma osmolality suggests that they are controlled at least in part by separate mechanisms. The organismal-level results presented here have molecular, ecological and evolutionary bases, which should continue to be investigated to further our understanding of and ability to respond to the climate crisis.
Acknowledgements
We thank and honor the Northern Chumash tribe yak tityu tityu yak tiłhini for being the original stewards of the land that continues to support an abundant S. occidentalis population where we captured lizards for this study. We thank R. Bedard, S. Messina, E. Hutchinson, R. Johnson, M. Temple, S. Swan, P. Maier, G. Moosman, E. O'Brien, S. Han, S. Escalona, K. Choe, A. Sutcliffe, G. Post, and M. V. Xiong for their time catching lizards. We thank A. Hamrick, K. Doctor, and K. Dunham for logistical support, and we especially thank K. Bodwin for statistical guidance, and T. Bean and E. Riddell for feedback on manuscript drafts.
Footnotes
Author contributions
Conceptualization: S.J.W., E.N.T.; Methodology: S.J.W., R.S.T., E.N.T.; Formal analysis: S.J.W., R.S.T.; Investigation: S.J.W., T.M., T.v.R., E.N.T.; Resources: S.J.W., E.N.T.; Data curation: S.J.W.; Writing - original draft: S.J.W.; Writing - review & editing: S.J.W., T.M., T.v.R., R.S.T., E.N.T.; Visualization: S.J.W., R.S.T.; Supervision: E.N.T.; Project administration: S.J.W., E.N.T.; Funding acquisition: E.N.T.
Funding
This work was supported by an award from the Bureau of Land Management and a California Polytechnic State University Aryan Roest Memorial Scholarship to S.J.W.; and by Cal Poly Bailey College of Science and Mathematics William and Linda Frost Fund undergraduate summer research awards to T.M. and T.v.R. The U.S. Bureau of Land Management provided funding for the evaporimeter, and all other supplies and equipment were provided by the Cal Poly Biological Sciences Department and the William and Linda Frost Fund in the Cal Poly Bailey College of Science and Mathematics.
Data availability
All data and code for this study are available in Zenodo at https://doi.org/10.5281/zenodo.10018753.
References
Competing interests
The authors declare no competing or financial interests.