
==== Front
Integr Comp Biol
Integr Comp Biol
icb
Integrative and Comparative Biology
1540-7063
1557-7023
Oxford University Press

38839599
10.1093/icb/icae057
icae057
Symposium
AcademicSubjects/SCI00960
Water Availability and Temperature as Modifiers of Evaporative Water Loss in Tropical Frogs
https://orcid.org/0000-0002-5474-596X
Juarez Bryan H Department of Biology, Stanford University, Stanford, CA 94305, USA
Earth System Science Department, Stanford University, Stanford, CA 94305, USA

Quintanilla-Salinas Isaac Mathematics Department, California State University Channel Islands, Camarillo, CA 93012, USA

Lacey Madison P Department of Biology, Stanford University, Stanford, CA 94305, USA

https://orcid.org/0000-0002-2706-4077
O'Connell Lauren A Department of Biology, Stanford University, Stanford, CA 94305, USA

E-mail: bryanhjuarez@gmail.com
8 2024
05 6 2024
05 6 2024
64 2 354365
05 4 2024
22 5 2024
22 5 2024
28 6 2024
© The Author(s) 2024. Published by Oxford University Press on behalf of International Council for the Exploration of the Sea.
2024
https://creativecommons.org/licenses/by/4.0/ This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.

Synopsis

Water plays a notable role in the ecology of most terrestrial organisms due to the risks associated with water loss. Specifically, water loss in terrestrial animals happens through evaporation across respiratory tissues or the epidermis. Amphibians are ideal systems for studying how abiotic factors impact water loss since their bodies often respond quickly to environmental changes. While the effect of temperature on water loss is well known across many taxa, we are still learning how temperature in combination with humidity or water availability affects water loss. Here, we tested how standing water sources (availability) and temperature (26 and 36°C) together affect water loss in anuran amphibians using a Bayesian framework. We also present a conceptual model for considering how water availability and temperature may interact, resulting in body mass changes. After accounting for phylogenetic and time autocorrelation, we determined how different variables (water loss and uptake rates, temperature, and body size) affect body mass in three species of tropical frogs (Rhinella marina, Phyllobates terribilis, and Xenopus tropicalis). We found that all variables impacted body mass changes, with greater similarities between P. terribilis and X. tropicalis, but temperature only showed a notable effect in P. terribilis. Furthermore, we describe how the behavior of P. terribilis might affect its water budget. This study shows how organisms might manage water budgets across different environments and is important for developing models of evaporative water loss and species distributions.

National Science Foundation 10.13039/100000001 2109850 New York Stem Cell Foundation 10.13039/100003194
==== Body
pmcIntroduction

Hydro- and thermoregulation are critical elements of physiological processes in all organisms (Potts 1994; Chown et al. 2011; Lombardini and Rossi 2019; Rozen-Rechels et al. 2019). Water is necessary for the transport and maintenance of solute concentrations found in the body, which may be altered by temperature and water loss (Takei and Hwang 2016). The importance of temperature in organisms ranges from controlling the rates of chemical reactions to the rates of organism-level traits, such as metabolic rates (Robinson et al. 1983; Hochachka and Somero 2002; Clarke and Fraser 2004; Angilletta 2009). While water and temperature are distinct aspects of the abiotic environment, their effects may be correlated (Riddell et al. 2019), and organisms often have to respond to changes in water and temperature in tandem. Terrestrial organisms that are challenged by relatively hotter and drier conditions must hydro- and thermoregulate or risk experiencing lethal levels of dehydration or heat stress (Kearney et al. 2013; Dupoué et al. 2017; O'Sullivan et al. 2017; MacMillan 2019). Mortality caused by dehydration or heat stress can have important ecological effects, including changes to population sizes or species distributions (Kearney et al. 2018; Camacho et al. 2023). For example, species distribution models that account for the effects of water loss on the voluntary thermal limits of lizards outperform those that do not (Camacho et al. 2023). However, relative to temperature, we have a limited understanding of the role of water loss in acclimation to hydrothermal environments (Weaver et al. 2023), shifting geographical distributions (Kearney et al. 2018), and vulnerability to climate change (Pintor et al. 2016; Riddell et al. 2023). Thus, learning how organisms are impacted by water loss and temperature will help us support conservation efforts for at-risk species.

Measurements of evaporative water loss (EWL) rates are commonly used to understand how organisms in terrestrial environments respond to hydrothermal challenges. Most EWL happens through fluid exchange along moist surfaces in the respiratory, sensory, and integumentary systems like the lungs, eyes, and skin (Mautz 1980; Senzano and Andrade 2018; Pirtle et al. 2019). EWL is determined by various factors, including water vapor pressure, temperature (Riddell et al. 2019), and traits such as body size, skin resistance, or metabolic rate (Eynan and Dmi'el 1993; Oufiero and Van Sant 2018; Senzano and Andrade 2018; Howard et al. 2020). EWL rates are important for understanding how organisms manage their water budgets, or water influx and efflux, across different environments (Kearney et al. 2013; Gouveia et al. 2019; Pirtle et al. 2019). Quantifying water loss is key for learning how organisms respond to changes in mass and about the resulting correlations between mass and traits such as body temperature, metabolism, and movement ability (Losos 1990; Marsh 1994; Iriarte-Díaz 2002).

There are two major challenges to learning how water loss may impact organismal traits across different environments. The first challenge is determining whether water and temperature have non-additive effects on EWL and body mass (Kearney et al. 2013; Rozen-Rechels et al. 2019; Weaver et al. 2023). For instance, water may reduce EWL by covering the body and limiting body surface area exposed to air. Furthermore, some organisms might offset EWL by absorbing water from the environment (Dainton 1954; Withers 1993; Yoder et al. 2007). A second challenge is determining how organisms behaviorally modify how they experience their abiotic environment to alter dehydration rates (Pough et al. 1983; Pirtle et al. 2019; Dezetter et al. 2023). For example, many animals burrow to avoid dehydration or use evaporative cooling to reduce their body temperature (Hall and Root 1930; Duellman and Trueb 1994). Learning how water and temperature contribute to dehydration in the context of water sources and behavior is necessary to understand how organisms perform hydro- and thermoregulation.

Anuran amphibians are an ideal system for determining relationships between water loss, water absorption, temperature, and body size. Frogs have a rich 200+ year history of research describing how they manage their water budgets (Jørgensen 1997). While endotherms exhibit body temperatures above ambient temperatures, ectotherms maintain temperatures at or near ambient levels, relative to endotherms (Withers 1992; Muñoz-Garcia et al. 2014). This makes frogs useful in quantifying the effect of temperature on water loss. Furthermore, anurans exhibit a variety of integumentary, circulatory, and behavioral adaptations for delaying or slowing dehydration (Toledo and Jared 1993; Lemenager et al. 2022). Some examples include lipid secretions, integumentary cocoons, increased skin texturing, or water conservation postures (Jørgensen 1997). Notably, frogs use vascularization along the pelvic skin (the pelvic patch) to absorb water instead of drinking (Parsons and Mobin 1991; Suzuki et al. 2007). Frogs may also manage their water budget through behavior by climbing, jumping, or burrowing (Duellman and Trueb 1994) to reach new microhabitats that provide water or shelter from drying conditions (Brusch et al. 2019; Gastón and Akmentins 2023). Additionally, traits used in managing water budgets (Toledo and Jared 1993), including behavioral and phenotypic traits, are often linked to microhabitat use. For example, terrestrial frogs exhibit some of the lowest rates of dehydration among frogs while aquatic frogs exhibit some of the highest (Young et al. 2005; Tracy et al. 2014; Cruz-Piedrahita et al. 2018). Learning how frogs manage their water budgets should help us develop better predictions of how ectothermic organisms respond to extreme environments and climate change.

While we generally understand water budgets in temperate frogs, we have a much narrower understanding of water budgets in tropical frogs (Scheffers et al. 2013; Reider et al. 2021; Womack et al. 2022; Bovo et al. 2023). This is a critical knowledge gap because most of the world’s frogs (>7000 species) are found in the tropics (Jenkins et al. 2013). Globally, 50% of frogs are classified as threatened, endangered, or critically endangered, and the world may face catastrophic losses of amphibian biodiversity in the tropics without intervention (Stuart et al. 2004; González-Del-Pliego et al. 2019). In this study, we chose to learn about water loss in three species: cane toads (Rhinella marina), golden poison frogs (Phyllobates terribilis), and Western clawed frogs (Xenopus tropicalis). We chose these species by selecting primarily terrestrial or aquatic frogs with large expected differences in EWL. Rhinella marina is a large (∼85–225 mm; AmphibiaWeb 2023) primarily terrestrial toad native to Central and South America. Rhinella marina is also a prevalent invasive species around the world and has high skin resistance to dehydration (Kosmala et al. 2020). Phyllobates terribilis is a small (∼46 mm; AmphibiaWeb 2021) terrestrial frog native to Colombia. To our knowledge, dehydration rates have not been measured in P. terribilis but we expect it to experience rapid dehydration due to its small size (Tracy et al. 2010). Xenopus tropicalis is a small (∼28–40 mm; AmphibiaWeb 2024) aquatic frog native to Western and Central Africa with low skin resistance to dehydration (Mokhatla et al. 2019).

The goal of this study is to determine how frogs respond to the dual effects of water availability and temperature. We use R. marina, P. terribilis, and X. tropicalis to test the hypothesis that water can buffer against EWL by limiting body surface area exposed to air. Alternatively, water sources may not affect EWL if the animals do not use water sources, as this would not alter the body surface area exposed to air. We demonstrate in Fig. 1 how water availability may buffer or overcome EWL due to temperature, resulting in mass loss or mass gain. Whether mass is lost or gained in each environment depends on the relative rates of water uptake and evaporation. Additionally, we hypothesize that larger bodies are more resistant to EWL since larger organisms have a relatively lower surface area-to-volume ratio. Lower ratios limit the surface area over which evaporation might happen, relative to volume (Gouveia et al. 2019; Castro et al. 2021). We also expect that the extent to which hydrothermal environments affect body mass varies across species. We predict that R. marina is the most resistant to EWL, followed by P. terribilis, and then X. tropicalis based on our expectation that terrestrial frogs should experience less EWL than aquatic frogs. Overall, this study advances our understanding of ecology in ectotherms by determining how water budgets are managed in different environments and is important for developing our models of EWL and species distributions.

Fig. 1 A conceptual model of how body mass may change due to interactions between water availability and temperature. These examples are not exhaustive. We assume mass gain is only possible in water and relatively higher temperatures always result in greater mass loss. Outcomes (O) 1–5 show treatment effects (solid blue lines) for water at 26°C. For simplicity, we only show two 36°C (dashed red) lines near the lower slope = 0 limit for each potential outcome, corresponding to the lack of an interaction with temperature for O1 and O4. 36°C treatments may also yield results as shown in O1–5, representing an interaction (if statistically distinct from the 26°C effect). O1 and O4 show no temperature interaction and only a small effect of water availability, resulting in mass gain or mass loss. O2 shows no marginal effect of water. O3 and O5 show stronger possible temperature interactions (relative to 36°C). Relative to O5, O3 shows more body mass is lost without water, and only a small body mass increase when in water.

Alt text: Body mass change is on the y-axis and water availability is on the x axis. The y-axis shows only the baseline of no change at zero. The x-axis shows two water availability treatments: in water or no water. Outcome 1 shows body mass loss in both treatments but less mass loss in water. The 36°C line is similar. Outcome 2 shows similar body mass loss in both water treatments. Outcome 3 shows mass loss without water, but mass gain in water. Outcomes 4 and 5 show mass loss when not in water, but different degrees of mass gain in water.

Methods

Specimen collection and housing

Wild R. marina (N = 32) were obtained from Oahu, Hawaii, USA (HI Permit No. EX-23-04). We caught these toads by hand and placed them in containers with moist soil and a water source prior to and during shipping to Stanford, CA. Phyllobates terribilis (N = 31) were purchased from Indoor Ecosystems (Whitehouse, Ohio, USA). Xenopus tropicalis (N = 32) were purchased from Xenopus 1, Corp. (Dexter, Michigan, USA). All animal procedures were conducted in compliance with Stanford University’s research ethics review board (APLAC 34069).

Each animal was placed in species-specific housing. These tanks were approximately 101.0 cm L x 54.6 cm W x 45.5 cm H for the R. marina and 76.2 cm L × 45.7 cm W × 45.7 cm H for the other species. Rhinella marina and P. terribilis were housed socially in plastic tanks in photoperiod-, temperature-, and humidity-controlled rooms with a 12 h light-dark cycle (night from 3 pm to 3 am) at an average of 26°C and 100% humidity. The animal tanks contained moist soil, sphagnum moss, and water sources for each species (e.g., water in petri dishes or glass bowls). We also included shelter for each species consisting of Philodendron plants and/or coconut endocarp. Rhinella marina and P. terribilis were misted daily and fed every other day. We provided each toad three crickets and each poison frog roughly 60 flies on each feeding day. We provided approximately 60 flies per frog to make up for the fact that a large portion of flies escape or are not eaten, and therefore the real number of flies eaten is likely far less. Xenopus tropicalis were housed socially in aquaria at 28°C and a 12 h (night from 9 pm to 9 am) photoperiod. The aquaria included shelters made of PVC. Xenopus tropicalis tanks received a daily 20% water change, and each frog was fed approximately five aquatic frog pellets every other day. Animals were kept in these standard housing conditions for at least one week before data collection.

Data collection

We exposed each frog to a starvation period of at least 3 days prior to acclimation and application of treatments and did not feed any animals after the first day of experimentation. This procedure let us standardize for potential metabolic effects on body mass, independent of body size differences between species. We assigned frogs randomly to experimental treatments and exposed them to an acclimation period of 24 h prior to applying each treatment. Following the starvation period, we transferred each frog to individual plastic terraria for 24 h (at 26°C) to acclimate them to their experimental containers. We used larger containers (29.8 cm L × 19.7 cm W × 20.3 cm H) for R. marina and smaller containers (23.5 cm L × 15.2 cm W × 17.8 cm H) for the other species. To prevent stress and mortality during this period, we provided the frogs with shelter and a water source (R. marina and P. terribilis) or enough water to cover their bodies and the shelter (X. tropicalis). We removed the shelter and water sources and dried the individual tanks with a paper towel, as needed, immediately prior to applying each water and temperature treatment. We measured four frogs per day (one per treatment) for 8 days (4 frogs/day × 8 days = 32 frogs).

We exposed each species to four treatments, including all combinations of water presence/absence and incubation at 26 or 36°C, resulting in an average of N = 8 samples per treatment (32 frogs/4 treatments) for each species. Following Shibata et al. (2014), we filled each container with 1 cm of water for the water presence treatment and provided a 1 cm L × 1 cm W × 0.5 cm H moist sponge to frogs in the water absence treatment to minimize potential discomfort but limit potential water absorption. We used a Fisherbrand™ Isotemp™ BOD Refrigerated Incubator to control experimental temperatures. We weighed each frog prior to each treatment and approximately every 20 min for an hour, since preliminary experiments indicated high mortality for P. terribilis for longer periods. We gently patted each frog with a paper towel to remove excess water prior to weighing. We euthanized each frog by administering intracoelomic injections of 1% MS-222 followed by decapitation at the conclusion of the experiment after 1 h, or after individuals: (1) lost > 20% body mass, or (2) did not exhibit a righting response. We recorded body size (snout-vent length) for each frog immediately after administering MS-222 but before decapitation.

Data analysis

All analyses were done in R 4.3.2 (R Core Team 2024). We implemented some corrections to account for urination, defecation, and missing values. In total, only six R. marina and two P. terribilis urinated and/or defecated (0.5–3% body mass) throughout the experiment. Since mass changes associated with urination and defecation are not due to evaporation, we corrected the raw body masses by adding the weight of the urine and stool to prevent overestimation of the EWL. Furthermore, we removed N = 9 individuals from the study due to lack of righting response or >20% mass loss. Removing the N = 9 individuals resulted in ten total missing values for body mass at various times. We estimated these missing values by predicting body mass at each corresponding time point using a linear regression of body mass and time with all available data. Lastly, we could not measure the mass of an additional N = 8 frogs for one time point per frog since the measurement period overlapped with endpoints for other frogs. In these cases, we estimated the missing values as the average mass of the previous and next measurements. For the whole study, we had N = 18 missing mass values of a possible 380 measurements (4.74%).

We analyzed the data in a Bayesian framework by fitting a phylogenetic longitudinal generalized linear mixed-effects model using a Hamiltonian Monte Carlo algorithm implemented using Stan 2.26.1 (Stan Development Team 2024), which we accessed using the R library cmdstanr (Gabry et al. 2024). We used four Markov Chains with each chain having 30,000 burn-in and 50,000 samples. We verified appropriate model fit using standard diagnostics for multi-level longitudinal models (Fitzmaurice et al. 2004). We regressed body mass (natural log grams) onto time (minutes), water availability, temperature, and snout-vent length (natural log mm). We included interactions between water and temperature to test our hypothesis that EWL depends on water availability and a second interaction between water and time to model separately the time effects associated with water loss and water uptake. We also included random effects for individuals, housing group, date, and species. To account for phylogenetic and time autocorrelation, we used a composite covariance matrix obtained from the error variance and correlation matrix. We obtained the phylogenetic correlation matrix after estimating mean branch lengths from 1000 trees drawn from the pseudo-posterior distribution of Jetz and Pyron (2018). Since we did not measure body mass at exact 20 min intervals, we modeled temporal autocorrelation using a Gaussian function that incorporates the time difference between measurements (Diggle et al. 2013).

We modeled all parameters of interest and the corresponding priors using normal or inverse gamma distributions, where appropriate (Gelman et al. 1995). We designed priors with mean effects estimated from preliminary experiments on P. terribilis or prior literature (Mokhatla et al. 2019; Kosmala et al. 2020) and enough variance for the effects to include zero on the raw scale. We allowed water uptake rates to be up to double the dehydration rates based on prior evidence that water uptake can happen very quickly (Jørgensen 1997). The prior for the water-temperature interaction assumed that on average, half the surface area of X. tropicalis and one-quarter of the surface area of P. terribilis and R. marina were covered with water in those treatments (based on our observations). The priors for body size were set at the midpoints of the minimum and maximum body sizes for each species. The variance priors were modeled with an inverse-gamma distribution with error variance modeled with a narrow range and the random effects variances modeled with a wide range. The correlation parameter was modeled as uniform distribution. We list and describe all prior distributions in the Supplementary Materials (Table S1).

Results

Generally, our raw data show that most instances of water loss occur in the treatments without water (Fig. 2). We also plotted individual stepwise changes in mass through time in the Supplementary Materials (Fig. S1). Comparison of the posterior and prior distributions of each model effect shows how our data updated our prior expectations, generally resulting in better estimates of mean effects with higher precision (Fig. 3). Summary plots for evaluating model fit are found in the Supplementary Materials.

Fig. 2 Raw body mass data were collected over 1 h for each water availability and temperature treatment. Each panel is a different species. Red hues correspond to 36°C, and blue hues correspond to 26°C. Circles show water presence and triangles show water absence treatments. We assumed even sampling at 20 min intervals for plotting purposes, but we analyzed the real times at which measurements were taken (see the “Methods” section).

Alt text: The figure is composed of 3 panels, each with their own y-axis, representing body mass in grams for Rhinella marina, Phyllobates terribilis, and Xenopus tropicalis. The mass for each ranges from about 100–400, 1–4, and 5–25 grams. The x-axis for each panel is the same and has time ranging from 0–60 minutes. Each panel has mass measurements for 0, 20, 40, and 60 minutes connected by lines for each of the 31 or 32 individuals used for each species. Each of the individual frogs corresponds to one of the 4 treatments of water and temperature. The R. marina lines are mostly flat showing very little mass change regardless of treatment. The P. terribilis lines for the frogs in water seem to be mostly flat whereas the lines for frogs not in water seem to mostly decrease, with different individuals showing different rates of decrease. The X. tropicalis lines all seem to decrease at similar rates.

Fig. 3 Prior and posterior distributions of model effects. Posterior distributions (left) are shown in blue, and prior distributions (right) are shown in yellow. Snout-vent length is a measure of body size in amphibians. All prior distributions shown here were normally distributed. Please see text for how we chose to specify the mean and variance for each prior.

Alt text: 18 graphs corresponding to the posterior and prior distribution for each model effect. The y-axis for each corresponds to different values for the regression coefficient of each effect. The x-axes are centered at 0 with the left and right halves corresponding to the densities of the posterior and prior, respectively. Overall, the range for the priors are wider than the range of the posteriors. The posteriors tend to occur at the edges of the priors, with a narrower variance. Two exceptions include the R. marina water-time interaction and the R. marina time effect, where the posteriors are quite different from the priors.

We found larger body sizes and water availability reduce EWL and these effects, like the EWL rate, differ between species (Table 1). For instance, we found larger bodies prevented dehydration at an average rate of 14.83% body mass per 2.72 mm increase in snout-vent length (=0.1383 log g/log mm) after 30 min in Xenopus. We found water availability reduces net EWL by allowing for water uptake. This effect was similar for X. tropicalis and P. terribilis at an average rate of 0.23% body mass per minute (=0.0023 log g/min), whereas R. marina did not absorb any water. We also found X. tropicalis and P. terribilis showed similar EWL rates of 0.2% body mass per minute (=0.0019 log g/min), whereas R. marina lost less mass over time at a rate of 0.02% body mass per minute (=0.0002 log g/min). Each of these effects had 95% credible intervals that excluded 0. Furthermore, we found frog ID, housing, and experimental date did not explain much variation in body masses, especially compared to interspecific differences. We found interspecific differences accounted for 20, 40, and 65 times more variance than housing, date, and frog ID, respectively.

Table 1 Summary of posterior distributions of factors affecting dehydration

Model term	Mean	Q2.5	Q97.5	
Intercept (ln g)	0.0887	−0.3665	0.5457	
Time (ln g/min, not in water)	−0.0019	−0.0028	−0.0010	
Water (ln g, not in water)	0.0373	−0.3286	0.4024	
Temp. (ln g/°C, not in water)	−0.0092	−0.0208	0.0023	
SVL (ln g/ln mm)	0.1383	0.0633	0.2137	
R. marina	3.1515	0.5209	5.6464	
P. terribilis	−1.9741	−2.4284	−1.5218	
R. marina-Temp.	−0.0057	−0.0186	0.0070	
P. terribilis-Temp.	−0.0092	−0.0272	0.0086	
Water-Temp. (ln g/°C, in water)	0.0028	−0.0130	0.0186	
P. terribilis-Water	0.0139	−0.3864	0.4134	
R. marina-Water	−0.2219	−0.6034	0.1616	
P. terribilis-Water-Temp.	0.0051	−0.0098	0.0201	
R. marina-Water-Temp.	0.0074	−0.0063	0.0211	
P. terribilis-Time	−0.0002	−0.0014	0.0011	
R. marina-Time	0.0017	0.0005	0.0030	
Water-Time (ln g/min, in water)	0.0023	0.0011	0.0035	
P. terribilis-Water-Time	0.0001	−0.0016	0.0018	
R. marina-Water-Time	−0.0023	−0.0040	−0.0005	
Frog ID variance	0.0787	0.0535	0.1146	
Housing variance	0.2555	0.1163	0.5413	
Date variance	0.1280	0.0688	0.2340	
Species variance	5.1479	0.9350	19.6249	
Error variance	0.0174	0.0135	0.0224	
We obtained these results using a phylogenetic time series generalized linear mixed model implemented in a Bayesian framework. Mean is the mean of the effect. Q2.5 and Q97.5 are the quantiles corresponding to the two-tailed 95% credible interval. Rows corresponding to main effects in bold exclude 0 from the 95% credible interval. SVL is snout-vent length. R. is Rhinella. P. is Phyllobates. Housing is a grouping factor indicating the experimental housing. Date is the date of measurement. We provide units for each effect on the baseline of X. tropicalis. ln is natural log. Temp. is temperature.

While temperature seemed unimportant at first, we found higher temperatures increased EWL in X. tropicalis and P. terribilis. We estimated the posterior marginal effects of temperature for each species across water treatments and plotted them (Table 2; Fig. 4). We found P. terribilis experienced greater water loss at higher temperatures (when not in water) at a rate of 1.02 g/°C (0.0185 log g/°C). The 95% credible interval for the previous result excluded 0. Interestingly, the marginal effect of temperature for P. terribilis in water is visually similar to that of P. terribilis not in water (with similar quantiles and mean). While the 95% credible intervals corresponding to P. terribilis in water and X. tropicalis in both water treatments included 0, we also found very high probabilities (91.14–97.32%) that higher temperatures contribute to greater dehydration in X. tropicalis and P. terribilis, regardless of water treatment.

Fig. 4 Posterior distributions of the marginal effect of temperature by species and water treatment. These effects correspond to body mass. The height of each distribution corresponds to the probability density. The vertical line denotes an effect of 0.

Alt text: Y-axis ticks correspond to each water treatment (presence and absence) for each of Xenopus tropicalis, Rhinella marina, and Phyllobates terribilis. The x axis shows the posterior distribution of each effect. All distributions show roughly the same effect between 0 and -0.05 (on the natural log scale). The two effects of greatest magnitude correspond to both water treatments in P. terribilis, these distributions are almost entirely overlapping. The treatments for X. tropicalis are the next greatest in magnitudes with qualities similar to P. terribilis. The distributions for R. marina are nearly centered at 0.

Table 2 Summary of the posterior distributions of the marginal effects of temperature across species and water treatments

Model term (ln g/°C)	Mean	Q2.5	Q97.5	
R. marina in water	−0.0047	−0.0252	0.0153	
R. marina not in water	−0.0064	−0.0235	0.0104	
P. terribilis in water	−0.0191	−0.0393	0.0010	
P. terribilis not in water	−0.0185	−0.0365	−0.0004	
X. tropicalis in water	−0.0150	−0.0304	0.0002	
X. tropicalis not in water	−0.0092	−0.0208	0.0023	
Mean is the mean of the effect. Q2.5 and Q97.5 are the quantiles for the two-tailed 95% credible interval. Rows corresponding to effects in bold exclude 0 from the 95% credible interval. ln is the natural log.

Discussion

To our knowledge, this is one of the first studies to examine the dual effects of water sources and temperature on EWL. Overall, we found body masses change in predictable ways due to EWL, water uptake, size, and temperature. While we did not find an interaction between water and temperature, we did find interspecific differences in how temperature affected EWL. Our results support our hypotheses that water sources dampen rates of EWL and that the precise mechanisms affecting EWL vary by species. Contrary to our expectation, rates of water uptake and EWL were most similar between the aquatic X. tropicalis and terrestrial P. terribilis, rather than between both species with terrestrial lifestyles. The latter result supports assertions that species-specific traits may control water budgets (e.g., Pirtle et al. 2019), but further research is needed to determine how body size and ecology are related to water uptake.

Compared to prior research, we found differences in the rates of water loss and uptake in R. marina, but these may be explained by differences in body size, temperature, or experimental design. We found the R. marina experienced lower rates of EWL than X. tropicalis and P. terribilis at ∼33 mg/min, corresponding to a median starting mass of 165 g and mean experimental temperature of 31°C. This rate is almost four times previous estimates of 8.8 mg/min at 26°C (Kosmala et al. 2020) for frogs weighing 110 g. Since larger bodies and hotter temperatures are associated with higher rates of evaporation (Mautz 1980), much of the difference can be explained by body mass and temperature alone. Kosmala et al. (2020) also report a water uptake rate of 0.5 g/min, whereas R. marina in this study did not absorb any appreciable amount of water. This major difference is likely due to Kosmala et al. (2020) measuring water uptake in toads that had lost 30% body mass, whereas we allowed our toads to hydroregulate at will prior to experimentation. Based on our findings for R. marina, it is evident that spending 1 h in drying conditions (with or without water) had little impact on this species’ body mass (Fig. 2). Since R. marina did not absorb much water even after spending 1 h at 36°C, this suggests these conditions were either not stressful to this species or did not last long enough to provoke the frogs to absorb water. This result shows how understanding responses of frogs to specific environmental conditions might benefit from further research on hormones like arginine vasotocin, which is linked to dehydration and rehydration in frogs (Morel and Jard 1963; Cartledge et al. 2008; Uchiyama et al. 2014).

We also found differences in our estimated rates of EWL and those in other studies for X. tropicalis. We found X. tropicalis (average body mass = ∼13.5 g) dehydrated at a rate of 12% body mass/h, which is higher than previous findings of 1.6% body mass/h (Mokhatla et al. 2019) in Xenopus laevis (average body mass = ∼57 g). Since larger frogs dehydrate at a slower rate, body mass alone explains much of this difference, but the reported rate of 1.6% mass/h seems to be an average across temperatures ranging from 0 to 35°C, complicating direct comparisons. Furthermore, our estimates of water uptake (∼0.0322 g/min) were similar to previous findings in anesthetized X. laevis who were roughly twice as large as our animals and whose rate of water uptake was 0.012 g/min (Ireland 1973), given animals with a mean body mass of 27.5 g. The latter study did not report an experimental temperature. Comparing rates of water transfer depends on standardizing a variety of variables and while we explain how body size, temperature, and experimental design might account for our observed differences, humidity might be an additional confounding factor that should be investigated in future research.

While P. terribilis and R. marina share a terrestrial lifestyle, P. terribilis seems to manage its water budget more similarly to the aquatic X. tropicalis. In other words, while shared ecology is sometimes informative of EWL (Wygoda 1984), it may not always be a reliable indicator of EWL. Additionally, unlike the other two species, there is little literature on how P. terribilis manages its water budget. While P. terribilis and X. tropicalis had similar water loss and uptake rates, P. terribilis showed greater dehydration at higher temperatures when not in water. Since they had similar water loss and uptake rates, this means the two species differ in some other factor related to rates of water transfer. We suspect P. terribilis has a higher skin resistance relative to X. tropicalis, making water transfer rates similar despite temperature having a greater impact on P. terribilis. To our knowledge, however, no studies have measured skin resistance in P. terribilis. We also showed how the marginal effects of temperature on P. terribilis were quite similar across water treatments. Anecdotally, we believe this is due to ∼25–50% of individuals refusing to remain in the water during the experiment and choosing to climb up to the corner of the tank instead. There might be two reasons why we observed this behavior. First, climbing might be a stress response where the frogs chose to seek shelter at the expense of water loss. Second, it is possible the animal was not stressed but chose to use the corner as a refuge to reduce the surface area directly exposed to the air, resulting in less EWL. Either way, this climbing behavior and the roles of stress hormones in hydro- and thermoregulation of poison frogs, or other frogs that can climb, are important areas of future study.

This study shows the importance of water sources when considering how EWL occurs in nature. Humidity (van Dyk et al. 2019) and hydration states (Senzano and Andrade 2018) are undoubtedly important factors that can interact with temperature to affect EWL. While we did not observe an interaction between water availability and temperature, this might be due to the short time and temperature range we used. For comparison, Senzano and Andrade (2018) ran their experiment for 2 h and over a 20°C range. Future work might show that water availability indeed interacts with temperature, since our prediction was founded on the theoretical premise that EWL rates depend on the animal’s surface area (Gouveia et al. 2019). In nature, water may change the exposed body surface area on frogs while they are in water or floating on it, after they rub against wet substrate (e.g., wet plants), or as it is raining. Two of the latter examples involve behaviors whose role in EWL is understudied (Tracy et al. 2014). As we might have observed with P. terribilis, organisms may seek hydric refugia (Kearney et al. 2013) and the types of available refugia depend on the environment. Previous studies have found that R. marina depends on water availability and not heat tolerance for moving across dry environments, although high heat tolerance enabled Rhinella granulosa to retain high jumping performance even at high temperatures (Prates et al. 2013; Brusch et al. 2019). Thus, we may generally expect that behaviors (like water conservation postures or jumping), together with the environment, determine how we should interpret the importance of dehydration (Davis and DeNardo 2010; Pintor et al. 2016; Pirtle et al. 2019; Dezetter et al. 2023).

Mitigating future amphibian declines depends on our ability to predict how different groups are affected by changing hydrothermal environments (Roznik et al. 2018; Greenberg and Palen 2021). Some have considered how acclimation to different environments might allow organisms to survive challenging conditions (Weaver et al. 2023). In this study, we exposed frogs to sharp environmental changes with no acclimation period. Evaluating how organisms respond to various environmental conditions, with and without acclimation, should allow us to determine which species might survive gradual versus extreme changes in climate, reflecting the spatial heterogeneity of climate change (Kaufmann et al. 2017). Furthermore, if climate change proceeds slowly in some regions, we may expect organisms to adapt. This implies an eco-evolutionary framework for thinking about global climate impacts on ectotherm diversity is needed. In the context of geographical ranges, water and temperature have unique effects on behavior and species distributions (Kearney et al. 2018; Delgado-Suazo and Burrowes 2022; Camacho et al. 2023). However, potential trade-offs between traits like water balance and jumping performance are also important for predicting how organisms will move throughout the landscape in response to climate change (Moore and Gatten 1989; Titon and Gomes 2015; Mitchell and Bergmann 2016). Since metabolic rates increase EWL by increasing respiratory rates (Preest et al. 1992; Tomlinson and Phillips 2012), considering abiotic effects on movement, behavior, and physiology seems crucial. In this context, we hope this study will be used to improve our models of EWL and ecogeographical gradients (Gouveia et al. 2019; Rubalcaba et al. 2019), and species distributions (Riddell et al. 2023). In summary, the future of amphibian conservation depends on understanding the complex relationships among many abiotic and biotic variables (Kearney et al. 2013; Schulte 2015; Gouveia et al. 2019; Rubalcaba et al. 2019). Future studies should seek to unravel how trade-offs or plastic responses of biotic variables, such as those associated with water balance or movement, are distributed both geographically and phylogenetically (Garland et al. 2022; Telemeco et al. 2022).

Author contributions

B.H.J.: Conceptualization, Methodology, Software, Validation, Formal Analysis, Investigation, Resources, Data Curation, Writing—Original Draft, Writing—Review and Editing, Visualization, Project Administration, Funding Acquisition; I.Q.-S.: Methodology, Software, Validation, Formal Analysis, Resources, Data Curation, Writing—Review and Editing, Visualization; M.P.L.: Investigation, Resources, Writing—Review and Editing; L.A.O.: Conceptualization, Methodology, Resources, Writing—Review and Editing, Supervision, Project Administration, Funding Acquisition.

Supplementary Material

icae057_Supplemental_File

Acknowledgement

We thank D. Shaykevich for providing valuable assistance in collecting and housing cane toads, S. Weaver for providing a bibliography of relevant studies that significantly improved this study, and E. Martinez for caring for our Western clawed frogs. We also thank the O'Connell lab and three anonymous reviewers for providing valuable comments on previous versions of this manuscript.

Funding

This work was supported by a National Science Foundation Postdoctoral Research Fellowship in Biology Program award [2109850 to B.H.J.]. B.H.J. and L.A.O. are supported by the New York Stem Cell Foundation. L.A.O. is a New York Stem Cell—Robertson Investigator.

Conflict of interest

The authors listed above declare no conflicts of interest.

Data availability

The data presented in this article are available in the Dryad Digital Repository (doi.org/10.5061/dryad.mw6m90650).

Notes

From the symposium “Feel the flow: how water movement shapes organisms and ecosystems” presented at the annual meeting of the Society for Integrative and Comparative Biology, January 2–January 6, 2024.
==== Refs
References

AmphibiaWeb . 2021. Phyllobates terribilis: Golden Poison Frog. https://amphibiaweb.org/species/1707  (Accessed May 9, 2024), University of California: Berkeley, CA, USA.
AmphibiaWeb . 2023. Rhinella marina: Marine Toad. https://amphibiaweb.org/species/229  (Accessed May 9, 2024), University of California: Berkeley, CA, USA.
AmphibiaWeb . 2024. Xenopus tropicalis: Tropical Clawed Frogs. https://amphibiaweb.org/species/5246  (Accessed May 9, 2024), University of California: Berkeley, CA, USA.
Angilletta  MJ . 2009. Thermal adaptation: a theoretical and empirical synthesis. OUP: Oxford.
Bovo  RP, Simon  MN, Provete  DB, Lyra  M, Navas  CA, Andrade  DV. 2023. Beyond Janzen’s hypothesis: how amphibians that climb tropical mountains respond to climate variation. Integr Org Biol. 5 :obad009.37151602
Brusch  GA, Christian  K, Brown  GP, Shine  R, Denardo  DF. 2019. Cane toads (Rhinella marina) rely on water access, not drought tolerance, to invade xeric Australian environments. Oecologia. 189 :307–16.30535812
Camacho  A, Brunes  TO, Rodrigues  MT. 2023. Dehydration alters behavioral thermoregulation and the geography of climatic vulnerability in two Amazonian lizards. PLoS One. 18 :e0286502.37910524
Cartledge  VA, Withers  PC, Bradshaw  SD. 2008. Water balance and arginine vasotocin in the cocooning frog Cyclorana platycephala (Hylidae). Physiol Biochem Zool. 81 :43–53.18040971
Castro  KMSA, Amado  TF, Olalla-Tárraga  MÁ, Gouveia  SF, Navas  CA, Martinez  PA. 2021. Water constraints drive allometric patterns in the body shape of tree frogs. Sci Rep. 11 :1218.33441858
Chown  SL, Sørensen  JG, Terblanche  JS. 2011. Water loss in insects: an environmental change perspective. J Insect Physiol. 57 :1070–84.21640726
Clarke  A, Fraser  KPP. 2004. Why does metabolism scale with temperature?. Funct Ecol. 18 :243–51.
Cruz-Piedrahita  C, Navas  CA, Crawford  AJ. 2018. Life on the edge: a comparative study of ecophysiological adaptations of frogs to tropical semiarid environments. Physiol Biochem Zool. 91 :740–56.29211619
Dainton  BH . 1954. The activity of slugs: I. The induction of activity by changing temperatures. J Exp Biol. 31 :165–87.
Davis  JR, Denardo  DF. 2010. Seasonal patterns of body condition, hydration state, and activity of gila monsters (Heloderma suspectum) at a Sonoran Desert site. J Herpeto. 44 :83–93.
Delgado-Suazo  P, Burrowes  PA. 2022. Response to thermal and hydric regimes point to differential inter- and intraspecific vulnerability of tropical amphibians to climate warming. J Therm Biol. 103 :103148.35027199
Dezetter  M, Le Galliard  J-F, Lourdais  O. 2023. Behavioural hydroregulation protects against acute effects of drought in a dry-skinned ectotherm. Oecologia. 201 :355–67.36564481
Diggle  P, Heagerty  P, Liang  KY, Zeger  S. 2013. Analysis of longitudinal data. Oxford: Oxford University Press.
Duellman  WE, Trueb  L. 1994. Biology of amphibians. JHU Press: Baltimore, USA.
Dupoué  A, Rutschmann  A, Le Galliard  JF, Miles  DB, Clobert  J, Denardo  DF, Brusch  GA, Meylan  S. 2017. Water availability and environmental temperature correlate with geographic variation in water balance in common lizards. Oecologia. 185 :561–71.29018996
Eynan  M, Dmi'el  R. 1993. Skin resistance to water loss in agamid lizards. Oecologia. 95 :290–4.28312954
Fitzmaurice  GM, Laird  NM, Ware  JH. 2004. Applied longitudinal analysis. Hoboken, USA: John Wiley & Sons.
Gabry  J, Češnovar  R, Johnson  A. 2024. cmdstanr: r Interface to ‘CmdStan’. https://mc-stan.org/cmdstanr/, https://discourse.mc-stan.org.
Garland  T, Downs  CJ, Ives  AR. 2022. Trade-offs (and constraints) in organismal biology. Physiol Biochem Zool. 95 :82–112.34905443
Gastón  MS, Akmentins  MS. 2023. Differential effect of dehydration on the voluntary activity of a diurnal toad. Zoology. 159 :126105.37536073
Gelman  A, Carlin  JB, Stern  HS, Rubin  DB. 1995. Bayesian data analysis. Chapman and Hall/CRC: New York, USA.
González-Del-Pliego  P, Freckleton  RP, Edwards  DP, Koo  MS, Scheffers  BR, Pyron  RA, Jetz  W. 2019. Phylogenetic and trait-based prediction of extinction risk for data-deficient amphibians. Curr Biol. 29 :1557–1563.e3.31063716
Gouveia  SF, Bovo  RP, Rubalcaba  JG, Da Silva  FR, Maciel  NM, Andrade  DV, Martinez  PA. 2019. Biophysical modeling of water economy can explain geographic gradient of body size in anurans. Am Nat. 193 :51–8.30624109
Greenberg  DA, Palen  WJ. 2021. Hydrothermal physiology and climate vulnerability in amphibians. Proc Biol Sci. 288 :20202273.33593188
Hall  FG, Root  RW. 1930. The influence of humidity on the body temperature of certain poikilotherms. Biol Bull. 58 :52–8.
Hochachka  PW, Somero  GN. 2002. Biochemical adaptation: mechanism and process in physiological evolution. New York, USA: Oxford University Press.
Howard  JM, Griffis  H-B, Westendorf  R, Williams  JB. 2020. The influence of size and abiotic factors on cutaneous water loss. Adv Physiol Educ. 44 :387–93.32628526
Ireland  MP . 1973. Studies on the adaptation of Xenopus laevis to hyperosmotic media. Comp Biochem Physiol A Physiol. 46 :469–76.
Iriarte-Díaz  José . 2002. Differential scaling of locomotor performance in small and large terrestrial mammals. J Exp Biol. 205 :2897–908.12177154
Jenkins  CN, Pimm  SL, Joppa  LN. 2013. Global patterns of terrestrial vertebrate diversity and conservation. Proc Natl Acad Sci USA. 110 :E2602–10.23803854
Jetz  W, Pyron  RA. 2018. The interplay of past diversification and evolutionary isolation with present imperilment across the amphibian tree of life. Nat Ecol Evol. 2 :850–8.29581588
Jørgensen  CB . 1997. 200 years of amphibian water economy: from Robert Townson to the present. Biol Rev. 72 :153–237.9155243
Kaufmann  RK, Mann  ML, Gopal  S, Liederman  JA, Howe  PD, Pretis  F, Tang  X, Gilmore  M. 2017. Spatial heterogeneity of climate change as an experiential basis for skepticism. Proc Natl Acad Sci USA. 114 :67–71.27994143
Kearney  MR, Munns  SL, Moore  D, Malishev  M, Bull  CM. 2018. Field tests of a general ectotherm niche model show how water can limit lizard activity and distribution. Ecol Monogr. 88 :672–93.
Kearney  MR, Simpson  SJ, Raubenheimer  D, Kooijman  SALM. 2013. Balancing heat, water and nutrients under environmental change: a thermodynamic niche framework. Funct Ecol. 27 :950–66.
Kosmala  GK, Brown  GP, Shine  R, Christian  K. 2020. Skin resistance to water gain and loss has changed in cane toads (Rhinella marina) during their Australian invasion. Ecol Evol. 10 :13071–9.33304517
Lemenager  LA, Tracy  CR, Christian  KA, Tracy  CR. 2022. Physiological control of water exchange in anurans. Ecol Evol. 12 :e8597.35169455
Lombardini  L, Rossi  L. 2019. Ecophysiology of plants in dry environments. In: D'Odorico  P, Porporato  A, Wilkinson Runyan  C, editors. Dryland ecohydrology. Springer International Publishing: Dordrecht, The Netherlands. p. 71–100.
Losos  JB . 1990. The evolution of form and function: morphology and locomotor performance in West Indian Anolis lizards. Evolution. 44 :1189–203.28563896
Macmillan  HA . 2019. Dissecting cause from consequence: a systematic approach to thermal limits. J Exp Biol. 222 : jeb191593.30796157
Marsh  RL . 1994. Jumping ability of anuran amphibians. Adv Vet Sci Comp Med. 38B :51–111.7810380
Mautz  WJ . 1980. Factors influencing evaporative water loss in lizards. Comp Biochem Physiol A Physiol. 67 :429–37.
Mitchell  A, Bergmann  PJ. 2016. Thermal and moisture habitat preferences do not maximize jumping performance in frogs. Funct Ecol. 30 :733–42.
Mokhatla  M, Measey  J, Smit  B. 2019. The role of ambient temperature and body mass on body temperature, standard metabolic rate and evaporative water loss in southern African anurans of different habitat specialisation. PeerJ. 7 :e7885.31660269
Moore  FR, Gatten  RE. 1989. Locomotor performance of hydrated, dehydrated, and osmotically stressed anuran amphibians. Herpetologica. 45 :101–10.
Morel  F, Jard  S. 1963. Inhibition of frog (Rana esculenta) antidiuretic action of vasotocin by some analogues. Am J Physiol-Legacy Content. 204 :227–32.
Muñoz‐Garcia  A, Ben‐Hamo  M, Korine  C, Pinshow  B, Williams  JB. 2014. A new thermoregulatory index for heterothermy. Methods Ecol Evol. 5 :141–5.
O'sullivan  JDB, Macmillan  HA, Overgaard  J. 2017. Heat stress is associated with disruption of ion balance in the migratory locust, Locusta migratoria. J Therm Biol. 68 :177–85.28797478
Oufiero  CE, Van Sant  MJ. 2018. Variation and repeatability of cutaneous water loss and skin resistance in relation to temperature and diel variation in the lizard Sceloporus consobrinus. J Comp Physiol B. 188 :671–81.29619510
Parsons  RH, Mobin  F. 1991. Water flow across the pectoral and ventral pelvic patch in Rana catesbeiana. Physiol Zool. 64 :812–22.
Pintor  AFV, Schwarzkopf  L, Krockenberger  AK. 2016. Hydroregulation in a tropical dry-skinned ectotherm. Oecologia. 182 :925–31.27384338
Pirtle  EI, Tracy  CR, Kearney  MR. 2019. Hydroregulation: a neglected behavioral response of lizards to climate change?In: Bels  V, Russeull  A, editors. Behavior of Lizards. 1st ed. CRC Press. p.343–74.
Potts  M . 1994. Desiccation tolerance of prokaryotes. Microbiol Rev. 58 :755–805.7854254
Pough  FH, Taigen  TL, Stewart  MM, Brussard  PF. 1983. Behavioral modification of evaporative water loss by a Puerto Rican frog. Ecology. 64 :244–52.
Prates  I, Angilleta  MJ, Wilson  RS, Niehaus  AC, Navas  CA. 2013. Dehydration hardly slows hopping toads (Rhinella granulosa) from xeric and mesic environments. Physiol Biochem Zool. 86 :451–7.23799839
Preest  MR, Brust  DG, Wygoda  ML. 1992. Cutaneous water loss and the effects of temperature and hydration state on aerobic metabolism of canyon treefrogs, Hyla arenicolor. Herpetologica. 48 :210–9.
R Core Team.  2024. R: A language and environment for statistical computing. R Foundation for Statistical Computing. Vienna, Austria. https://www.R-project.org/
Reider  KE, Larson  DJ, Barnes  BM, Donnelly  MA. 2021. Thermal adaptations to extreme freeze–thaw cycles in the high tropical Andes. Biotropica. 53 :296–306.
Riddell  EA, Burger  IJ, Tyner-Swanson  TL, Biggerstaff  J, Muñoz  MM, Levy  O, Porter  CK. 2023. Parameterizing mechanistic niche models in biophysical ecology: a review of empirical approaches. J Exp Biol. 226 :jeb245543.37955347
Riddell  EA, Mutanen  M, Ghalambor  CK. 2023. Hydric effects on thermal tolerances influence climate vulnerability in a high-latitude beetle. Global Change Biol. 29 :5184–98.
Riddell  EA, Roback  EY, Wells  CE, Zamudio  KR, Sears  MW. 2019. Thermal cues drive plasticity of desiccation resistance in montane salamanders with implications for climate change. Nat Commun. 10 :4091.31501425
Robinson  WR, Peters  RH, Zimmermann  J. 1983. The effects of body size and temperature on metabolic rate of organisms. Can J Zool. 61 :281–8.
Rozen‐Rechels  D, Dupoué  A, Lourdais  O, Chamaillé‐Jammes  S, Meylan  S, Clobert  J, Le Galliard  J‐F. 2019. When water interacts with temperature: ecological and evolutionary implications of thermo-hydroregulation in terrestrial ectotherms. Ecol Evol. 9 :10029–43.31534711
Roznik  EA, Rodriguez-Barbosa  CA, Johnson  SA. 2018. Hydric balance and locomotor performance of native and invasive frogs. Front Ecol Evol. 6:1-8 .
Rubalcaba  JG, Gouveia  SF, Olalla‐Tárraga  MA. 2019. A mechanistic model to scale up biophysical processes into geographical size gradients in ectotherms. Global Ecol Biogeogr. 28 :793–803.
Scheffers  BR, Brunner  RM, Ramirez  SD, Shoo  LP, Diesmos  A, Williams  SE. 2013. Thermal buffering of microhabitats is a critical factor mediating warming vulnerability of frogs in the Philippine biodiversity hotspot. Biotropica. 45 :628–35.
Schulte  PM . 2015. The effects of temperature on aerobic metabolism: towards a mechanistic understanding of the responses of ectotherms to a changing environment. J Exp Biol. 218 :1856–66.26085663
Senzano  LM, Andrade  DV. 2018. Temperature and dehydration effects on metabolism, water uptake and the partitioning between respiratory and cutaneous evaporative water loss in a terrestrial toad. J Exp Biol. 221 :jeb188482.30385484
Shibata  Y, Sano  T, Tsuchiya  N, Okada  R, Mochida  H, Tanaka  S, Suzuki  M. 2014. Gene expression and localization of two types of AQP5 in Xenopus tropicalis under hydration and dehydration. Am J Physiol-Regul Integr Comp Physiol. 307 :R44–56.24717674
Stan Development Team . 2024. Stan modeling language user’s guide and reference manual, 2.26.1. https://mc-stan.org.
Stuart  SN, Chanson  JS, Cox  NA, Young  BE, Rodrigues  ASL, Fischman  DL, Waller  RW. 2004. Status and trends of amphibian declines and extinctions worldwide. Science. 306 :1783–6.15486254
Suzuki  M, Hasegawa  T, Ogushi  Y, Tanaka  S. 2007. Amphibian aquaporins and adaptation to terrestrial environments: a review. Comp Biochem Physiol A: Mol Integr Physiol. 148 :72–81.17270476
Takei  Y, Hwang  P-P. 2016. 6–homeostatic responses to osmotic stress. In: Schreck  CB, Tort  L, Farrell  AP, Brauner  CJ, editors. Fish physiology. Academic Press: Amsterdam, The Netherlands. p. 207–49.
Telemeco  RS, Gangloff  EJ, Cordero  GA, Rodgers  EM, Aubret  F. 2022. From performance curves to performance surfaces: interactive effects of temperature and oxygen availability on aerobic and anaerobic performance in the common wall lizard. Funct Ecol. 36 :2544–57.
Titon  B, Gomes  FR. 2015. Relation between water balance and climatic variables associated with the geographical distribution of anurans. PLoS One. 10 :e0140761.26469787
Toledo  RC, Jared  C. 1993. Cutaneous adaptations to water balance in amphibians. Comp Biochem Physiol A Physiol. 105 :593–608.
Tomlinson  S, Phillips  RD. 2012. Metabolic rate, evaporative water loss and field activity in response to temperature in an ichneumonid wasp. J Zool. 287 :81–90.
Tracy  CR, Christian  KA, Tracy  CR. 2010. Not just small, wet, and cold: effects of body size and skin resistance on thermoregulation and arboreality of frogs. Ecology. 91 :1477–84.20503879
Tracy  CR, Tixier  T, Le Nöene  C, Christian  KA. 2014. Field hydration state varies among tropical frog species with different habitat use. Physiol Biochem Zool. 87 :197–202.24642537
Uchiyama  M, Maejima  S, Wong  MKS, Preyavichyapugdee  N, Wanichanon  C, Hyodo  S, Takei  Y, Matuda  K. 2014. Changes in plasma angiotensin II, aldosterone, arginine vasotocin, corticosterone, and electrolyte concentrations during acclimation to dry condition and seawater in the crab-eating frog. Gen Comp Endocrinol. 195 :40–6.24184110
Van Dyk  M, Noakes  MJ, Mckechnie  AE. 2019. Interactions between humidity and evaporative heat dissipation in a passerine bird. J Comp Physiol B. 189 :299–308.30805750
Weaver  SJ, Mcintyre  T, Van Rossum  T, Telemeco  RS, Taylor  EN. 2023. Hydration and evaporative water loss of lizards change in response to temperature and humidity acclimation. J Exp Biol. 226 :jeb246459.37767755
Withers  P . 1993. Cutaneous water acquisition by the thorny devil (Moloch horridus: agamidae). J Herpeto. 27 :265–70.
Withers  PC . 1992. Comparative animal physiology. Philadelphia, USA: Saunders College Publishing.
Womack  MC, Steigerwald  E, Blackburn  DC, Cannatella  DC, Catenazzi  A, Che  J, Koo  MS, McGuire  JA, Ron  SR, Spencer  CL  et al.  2022. State of the Amphibia 2020: a review of five years of amphibian research and existing resources. Ichthyol Herpetol. 110 :638–61.
Wygoda  ML . 1984. Low cutaneous evaporative water loss in arboreal frogs. Physiol Zool. 57 :329–37.
Yoder  JA, Buchan  DR, Ferrari  NF, Tank  JL. 2007. Dehydration tolerance of the Rocky Mountain wood tick, Dermacentor andersoni stiles (Acari: ixodidae), matches preference for a dry environment. Int J Acarol. 33 :173–80.
Young  JE, Christian  KA, Donnellan  S, Tracy  CR, Parry  D. 2005. Comparative analysis of cutaneous evaporative water loss in frogs demonstrates correlation with ecological habits. Physiol Biochem Zool. 78 :847–56.16052451
