Impact of daily variable temperatures in life-history traits of tropical anurans

Anuran embryos and tadpoles are daily exposed to wide thermal variations in their ponds, with maximum temperatures at midday. The aim of this research was to study the impact of three daily variable thermal environments (with maximum experimental temperatures between 10:00 and 16:00 hours), on the survival, developmental time and body size of metamorphs of four tropical anuran species from lowland habitats in Colombia. A total of 50 embryos (Gosner stage ten) to metamorphosis (Gosner stage 46) of Rhinella humboldti, Hypsiboas crepitans and Engystomops pustulosus were exposed to each one of the three daily variable temperature treatments: high temperature (mean = 27.5 °C; maximum temperature = 34 ± 1 °C; range = 23-35 °C), medium temperature (25.5 °C; 29 ± 1 °C; 23-30 °C), and low temperature (24 °C; 24 ± 1 °C; 23-25 °C). For the other species, Espadarana prosoblepon, 40 embryos to metamorphosis were exposed to each one of the following thermal treatments: high temperature (mean = 22 °C; maximum temperature = 25 ± 1 °C; range = 18-26 °C), medium temperature (20.5 °C; 22 ± 1 °C; 18-23 °C), and low temperature (19 °C; 19 ± 1 °C; 18-20 °C). For all species, the thermal variable environment with the highest temperature showed the greatest accumulated survival, reduced significantly the developmental time from embryos to metamorphs, and the snout-vent-length of metamorphs. Therefore, under field conditions where ponds are exposed to thermally variable environments, the highest temperatures may promote a decrease in the period of time to metamorphosis, and a positive increase for the anuran survival; nevertheless, extreme temperatures were also found in the microhabitat of the species studied, higher than their upper thermal limits reported, which suggest a vulnerable situation for them and other tropical anurans from similar habitats. Rev. Biol. Trop. 65 (1): 55-63. Epub 2017 March 01.

The effect of constant temperatures on life-history traits in amphibians has been long known, but less so under variable temperature environments.Studies incorporating thermal variations may represent more natural field conditions, since temperature fluctuations are typical of most of the shallow and ephemeral pools where tadpoles live in.Despite the remarkable plasticity of amphibians through the embryonic and larval periods (Álvarez, & Nicieza, 2002;Moore, 1939), environmental temperature is well documented as one of the most predominant abiotic factor to influence the anuran development and growth (Atkinson, 1996;Bradford, 1990;Kern, Cramp, & Franklin, 2015;Smith-Gill, & Berven, 1979).Growth is the trajectory of increase in the somatic mass owned by the uptake, transformation and allocation of materials, while development is the trajectory of differentiation from a fertilized egg or some later stage to its adulthood, which is regulated by gene-by-environment interactions (Zuo, Moses, West, Hou, & Brown, 2011).Temperature has differential effects on larval growth and development, with differentiation decreasing at low temperatures but growth keeps on progressing until temperatures eventually become too low (Smith-Gill, & Berven, 1979).As a consequence, anuran growth tends to get prolonged larval periods at low temperatures and higher stage-specific growth than those growing in warmer conditions (Morrison, & Hero, 2003).The differential effect of temperature on development and growth may be explained by the enzymatic kinetic reactions (Ultsch, Bradford, & Freda, 1999), which is faster for the differentiation rate than for the growth rate (Smith-Gill, & Berven, 1979).Additionally, because temperature exerts a greater effect on the differentiation than on the growth rate (Gomez-Mestre et al., 2010), and the growth rate functionally depends upon the differentiation rate whereas the contrary does not (Berven, & Gill, 1983).Temperature can also interact with an individual's genome to affect the developmental time, body shape, body size, and mass at metamorphosis (Arrighi et al., 2013), although the genetic contribution tends to be masked by some differences resulting from the variation of the environmental conditions (Morrison, & Hero, 2003).
Daily temperature fluctuations on amphibians have also shown to induce a variation in traits associated with fitness, such as an increase in the developmental rate, tadpole body length and jumping performance in Limnodynastes peronii (Niehaus, Wilson, & Franklin, 2006), a positive increase in individual performance and development in Triturus alpestris (Meřaková, & Gvoždík, 2009), and to cause embryos to hatch at a younger developmental stage and longer snout-vent-length in Bombina orientalis (Kaplan, & Phillips, 2006).However, there is a lack of studies on the possible effects of thermally variable environments on life-history traits in amphibians, particularly in tropical anurans, which are important to define the thermal reaction norms for different species and to quantify the anuran vulnerability to global warming under more real variable conditions in nature (Paaijmans et al., 2013).Therefore, we examined the effect of three daily variable temperatures on the expression of several lifehistory traits in four tropical anuran species from Colombia, Rhinella humboldti, Hypsiboas crepitans, Engystomops pustulosus and Espadarana prosoblepon, that inhabit water bodies susceptible to large daily variations in the thermal environment.We studied these species as they are typical of the tropical dry forest, which is one of the most threatened ecosystems in the study region, the Tolima department, in Colombia and the Neotropics (Pizano, Cabrera, & Garcia, 2014;Portillo-Quintero, & Sánchez-Azofeifa, 2010).Specifically, we evaluated the larval survival, and the larval development time and body length at the end of metamorphosis, due to the strong influence that have on adult fitness (Altwegg, & Reyer, 2003).We were particularly interested in knowing whether the high temperatures, in the daily variable thermal environments, would have a greater effect on the life-history traits assessed, than the medium and low experimental temperatures.
Microhabitat temperatures for embryos and tadpoles of R. humboldti, H. crepitans and E. pustulosus were registered on the water surface of ponds (Fig. 1A) and underwater (about 10 cm depth) (Fig. 1B), and for E. prosoblepon embryos and tadpoles on the leaves of ferns (air temperature) (Fig. 1C) and underwater of streams (about 10 cm depth) (Fig. 1D), respectively.These data were collected with several MicroLite 16 L data loggers (Contoocook, NH, USA), programmed to obtain temperatures for every hour from September to December 2012, during the breeding season.This information was used to make the thermal profiles of the study species.
For this study, we collected several egg masses (of the different species) in early mornings of variable days during the rainy season, between September 2012 and March 2013.The egg masses were taken to the experimental site, the Herpetology laboratory at the University of Tolima, Ibagué, at an environmental temperature around 23 °C for R. humboldti, H. crepitans and E. pustulosus, and in an acclimatized room for E. prosoblepon (19 ± 1 °C).The aquatic embryos were placed directly in the aquariums (33 x 28 x 13 cm), whereas arboreal embryos of E. prosoblepon were initially located on wet towels in small petri dishes (51 mm) until they reached the stage 25, then the tadpoles were transferred to the aquariums.All the aquariums contained 8 L of previously declorinated tap water, 1 cm base layer of soil and sand, five small stones (150 g approximately) and four macrophytes (Pistia stratiotes), so as to simulate some natural conditions for these species.A fine screen nylon cloth (0.5 mm mesh) was also laced at the bottom of the aquariums to facilitate the subsequent checking of embryos and tadpoles.The tadpoles were fed with 0.2 g of fish food once every day and water quality was maintained by regular manual removal (using a small net) of uneaten food and a partial (50 %) (every eight days) and total renewal (every 24 days) of water.A photoperiod of approximately 12 hour light/12 hour darkness was set throughout the experiment.
The survival, development and growth from embryos (Gosner stage ten) to metamorphosis (Gosner stage 46) were the life-history traits evaluated in this study.The survival was analyzed on the accumulated number of animals that reached the metamorphosis in the three thermally variable environments, using a generalized linear model (GLM) with a binomial probability distribution and logit link function (logistic regression).We tested the effects of the species, thermal treatments, aquariums and interactions.The developmental time, from embryos to metamorphs, was recorded in days at the three experimental temperatures, while the growth was assessed by digital measurements of the snout-ventlength for each metamorph, from photos processed with ImageJ software (http://rsbweb.nih.gov/ij/) after checking for scaling artifacts.The developmental time and snout-vent-length were independently analyzed with linear models (LM) testing the effects of the species, thermal treatments, aquariums and their interactions.We did not show the statistical results neither for the aquariums, since they never had a significant effect on all the life-history traits evaluated, nor for the species, due to the remarkable differences in the development and growth, which were statistically significant in all cases.We used the Tukey's post-hoc test to compare the effects of the thermal treatments.All statistical analyses were conducted in SPSS Statistic 19.0 (SPSS Inc., Chicago, Illinois).
The protocols for this research were approved by the Bioethics Committee of the Tolima University (permit number: 2.3-193, May 14, 2012).

RESULTS
The accumulated survival of metamorphs increased significantly at the highest experimental variable temperature, followed by the medium and then the lowest temperatures (GLM: Wald X 2 = 45.9,P = 0.000).This trend was similar in all four species (Fig. 2).Pairwise comparisons among the three thermal treatments showed significant differences in all cases (high-medium, P = 0.00; high-low, P = 0.00; medium-low, P = 0.001).
High daily temperature variations significantly decreased the developmental time from embryos to metamorphs (LM: F = 26.85,P = 0.00), which was observed in all species (Fig. 3).This difference was statistically significant between the high and low thermal treatments (Tukey: P = 0.000), high and medium (Tukey: P = 0.003), and medium and low treatments (Tukey: P = 0.000).Finally, high daily temperature variation also significantly decreased the snout-vent-length of metamorphs (LM: F = 9.34, P = 0.001) (Fig. 4), although the Tukey's post-hoc tests only indicated a significant difference between the high and low temperatures (P = 0.004).

DISCUSSION
Environmental temperature determines, at a large extent, the larval body size and length of larval period, which in turn are directly related to survival (Alford, 1999;Sibly, & Atkinson, 1994).Then, the greatest survival that we found at the highest temperatures in the daily variable thermal environments, could be explained by the faster-growing tadpoles that metamorphosed early and consequently experienced lower cumulative mortalities.On the contrary, when exposed to lower experimental temperatures, metamorphs were larger, and tadpoles showed a slower growth, which is a typical response in amphibian larvae (Berven, & Gill, 1983;Smith-Gill, & Berven, 1979).Overall, these results were similar to those found in larval anurans raised in constant temperatures (Atkinson, 1996;Laugen, Laurila, & Merila, 2003), nevertheless, the effects of a constant environment does not always equal those of a fluctuating environment at the same mean temperature, according to the Jensen's inequality (Ruel, & Ayres, 1999).Ecologically, high temperature variations may promote a decrease in time to metamorphose and an increase in the anuran survival, when ephemeral ponds are overheated during the pond drying, as is the case for the temporal breeding study species R. humboldti, E. pustulosus and H. crepitans, and less for E. prosoblepon.
We found out that daily differences among the three thermally variable environments were enough to observe changes in the development and growth of all species.However, we are not confident if these responses were the consequence of the daily mean temperatures, the maximum thermal treatments, from 10:00 to 16:00 hours, the daily temperature ranges, or from a combined effect of all these regimens.We did not experimentally separated these three thermal conditions, as they co-occur in nature for the study species, which are more remarkable for those from open habitats, and we wanted to study these simultaneous daily thermal variations on some life-history traits in anurans, as opposed to other experiments that test the effects of daily temperature fluctuations, at a same experimental mean, with respect to constant temperatures (Arrighi et al., 2013;Kaplan, & Phillips, 2006;Kern et al., 2015;Meřaková, & Gvoždík, 2009;Niehaus et al., 2006).
Global warming has increased not only the mean and variance (Bozinovic et al., 2011) of the environmental temperatures, but also the frequency of extremely high temperatures (Easterling et al., 2000).Amphibians, on the other hand, are highly susceptible to temperature variations in their environments (Deutsch et al., 2008), and especially tropical ectotherms have relatively narrow thermal tolerances and the optimal temperatures are only a few degrees below the upper limits (Bozinovic et al., 2011).So, tropical amphibians may be particularly vulnerable to climate warning (Tewksbury, Huey, & Deutsch, 2008).This assertion can be confirmed with our field data, as we found extreme environmental temperatures higher than 42 °C in the habitats of R. humboldti, H. crepitans and E. pustulosus, which are 100 % lethal for their embryos (R. humboldti = 38.7 °C, H. crepitans = 36.9°C, E. pustulosus = 37.4 °C, E. prosoblepon = 27.0 °C) and tadpoles (R. humboldti = 40.7 °C, H. crepitans = 40.0°C, E. pustulosus = 40.6 °C, E. prosoblepon = 32.98)(Turriago, Parra, & Bernal, 2015).Since other anurans are exposed to environmental temperatures very similar to the study species, particularly those from lowlands and ephemeral ponds, these results suggest a vulnerable situation for many tropical anurans.
In conclusion, we found out that differences in the daily variable thermal environment did have a significant effect on the anuran life-history traits studied, with the highest experimental temperatures decreasing the accumulated mortality, the time to reach the metamorphosis and the snout-vent-length of metamorphs.
Thus, these results demonstrate that, within the thermal limits, high temperatures in thermally variable environments would be favorable for anuran survival, especially for those species that breed in temporary ponds.However, our field data also recorded temperatures higher than the upper thermal tolerances reported for the species studied, showing a vulnerable situation for them and other tropical anurans from similar environments.

ACKNOWLEDGMENTS
This research was supported by Fondo de Investigaciones de la Universidad del Tolima (Project number 370112).Collection permit was obtained from Corporación Autónoma Regional del Tolima, CORTOLIMA, resolution number 2046, June 13, 2012.We thank Liliana Marcela Henao for her valuable contribution to the lab experiments and the anonymous reviewers for their constructive comments which helped to improve the manuscript.

SUPPORTING INFORMATION
The data on the accumulated survival, the developmental time and growth of the four species exposed to the three daily thermal variable environments are shown in table 1.

Fig. 1 .
Fig. 1.Profiles of daily environmental temperatures (black lines) and three thermal treatments (curved lines) experienced by embryos and tadpoles of R. humboldti, H. crepitans, E. pustulosus (A, B), and E. prosoblepon (C, D).The squares indicate the mean temperatures and the lines the minimum and maximum value.

TABLE 1
Survival, developmental time and growth (snout-vent-length of metamorphs) of four tropical anuran species, exposed from embryos to metamorphs to three daily variable thermal environments