Thermal ecological physiology of native and invasive frog species: do invaders perform better?

Biological invasions are an important threat to biodiversity. We studied the effect of thermal acclimation on thermal tolerance and locomotor performance in the invasive Xenopus laevis and the Chilean native Calyptocephalella gayi. We show that X. laevis is a better performer than C. gayi. Furthermore, thermal tolerance did not differ between the species.


Introduction
Biological invasions are recognized as an important biotic component of global change (Richardson et al., 2000;Ricciardi, 2007;Lockwood et al., 2013). Invasive species alter the composition, structure and functioning of ecosystems, resulting in loss of biodiversity and displacement of native species (Talley et al., 2001;Rochlin et al., 2013). An emergent concern is that other components of global change, such as climate warming, might enhance the capacity of alien species to invade new areas (Dukes and Mooney, 1999;Stachowicz et al., 2002;Hellmann et al., 2008;Rahel and Olden, 2008;Walther et al., 2009;Robinet and Roques, 2010). Lejeusne et al. (2014) proposed that both components 'constitute a deadly-duo threatening species abundance, distributions and biotic interactions'.
At present, however, the organismal attributes that make invading species ecologically successful, in comparison to a native, remain as a poorly answered question (Rejmánek and Richardson, 1996;Snyder and Evans, 2006;Devin and Beisel, 2007). The successful establishment and spread of invasive species in a recipient environment would be facilitated by the ability of an invasive species to maintain high physiological performance over a wide range of environmental conditions (i.e. generalist behaviour; Marvier et al., 2004;Snyder and Evans, 2006;Angilletta, 2009;Knop and Neusser, 2012). Nevertheless, among native species a higher performance is typically constrained to a narrow range of conditions (i.e. specialist). As temperature has profound effects on organismal functions, this biotic environmental factor is gaining attention as a major driver of invasion success.
In this context, Kelley (2014) proposed, tested and provided evidence supporting multiple hypotheses to untangle the role that thermal physiology plays in species invasion. Indeed, this author hypothesized that the ability to maintain physiological function across an extensive range of temperature tolerances (i.e. eurythermality) might explain the success of invasive species over native ones ('greater eurythermal hypothesis'; see Lockwood and Somero, 2011;Zerebecky and Sorte, 2011;Tepolt and Somero, 2014). In addition, it has been proposed that acclimation to higher temperatures is associated with broader thermal tolerance in invasive species (Braby and Somero, 2006;Chown et al., 2007;Slabber et al., 2007;Hoffmann and Todgham, 2010;Tepolt and Somero, 2014).
The Chilean frog Calyptocephalella gayi is an endemic aquatic species inhabiting central Chile (Donoso- Barros and Cei, 1962;Vélez, 2014). This species is the only representative of the genus Calyptocephalella and is sometimes referred to as a living fossil (Pyron and Wiens, 2011). Calyptocephalella gayi is currently restricted to deep ponds and small water reservoirs in central Chile (Veloso and Navarro, 1988;Rabanal and Nuñez, 2008;Muzzopappa and Nicoli, 2010;Veloso et al., 2010). This species is classified as vulnerable by the IUCN (Veloso et al., 2010), and current evidence indicates that populations of C. gayi are declining in its native range (Glade, 1983;Díaz-Páez and Ortiz, 2003). Moreover, the Chilean frog has been declared as a protected species by the Chilean government, who have prohibited its capture (Glade, 1983). The extensive alteration of Chile's temperate and water regime that have occurred over the last decade plays a key role in creating adverse abiotic conditions for amphibian fauna (Diaz et al., 2007;Gutiérrez et al., 2012). More importantly, this species is also threatened by the introduction of the aquatic African clawed frog, Xenopus laevis (Lobos and Measey, 2002;Lobos and Jaksic, 2005). Xenopus laevis is probably one of the invasive amphibian species with the greatest worldwide distribution (Lowe et al., 2000;Tinsley et al., 2015). This generalist predator was introduced into Chile in the early 1980s, occurring in a very wide range of habitats, including those occupied by C. gayi (Glade, 1983;Veloso and Navarro, 1988). In fact, since its invasion in Chile, this species has colonized an area of~21 200 km 2 (Lobos and Jaksic, 2005). Currently, there is scarce knowledge on the factors and processes underlying its distributional pattern.
It is surprising to see that, despite the substantial negative impact of the African clawed frog on the Chilean frog populations, the basic organismal traits of what makes X. laevis successful in comparison to C. gayi are practically unknown. This is particularly relevant in times when climate is changing more rapidly than an amphibian can adjust (Pörtner and Farrell, 2008;Bozinovic and Pörtner, 2015).
Locomotor performance is an ecologically relevant parameter because it affects the ability of an organism successfully to reproduce, forage, escape predators or disperse into novel habitats (Garland and Losos, 1994;Higham, 2007;Husak and Fox, 2008;Seebacher and Franklin, 2011). Extreme environmental temperatures can affect muscle contractile properties and enzyme activities and, consequently, locomotor performance and survival (Bennett, 1990;Navas, 1996;Johnston and Temple, 2002;Herrel and Bonneaud, 2012;Careau et al., 2014). A rich history of research on the ability of ectothermic vertebrates to make adjustments that alter their locomotor performance after exposure to different thermal regimes has revealed that acclimation at low and moderate temperatures tends to improve locomotor performance. At higher temperatures, in contrast, locomotor performance is decreased (see review and examples in Navas et al., 2008 and references therein). These studies also have reported that acclimatory responses of locomotor performance take several weeks to occur (Wilson and Franklin, 1999;Wilson et al., 2000). The success of invasive species may also be influenced by the thermal dependence of locomotor performance. In particular, the capacity for invasion and range expansion of most ectothermic vertebrates relies upon active dispersal, which in turn can be influenced by thermal effects on physiological mechanisms that support locomotor performance (Seebacher and Franklin, 2011;McCann et al., 2014;Winwood-Smith et al., 2015).
Among the locomotor performance traits that can be studied in reptiles and amphibians, the righting response (i.e. the time it takes an individual to return to a prone position after being placed upside down) has received particular attention because it is rather easy to quantify in experimental conditions, shows a strong thermal dependence, can be related directly to muscle physiology and is related to fitness (John-Alder et al., 1988;Lutterschmidt and Hutchison, 1997;Steyermark and Spotila, 2001;Freedberg et al., 2004;Elnitsky and Claussen, 2006;Delmas et al., 2007). Moreover, the temperature at which animals are incapable of a coordinated locomotor response and their righting response is lost has been used as an indicator of tolerance to extreme temperatures in anuran amphibians (John-Alder et al., 1988;Lutterschmidt and Hutchison, 1997;Herrel and Bonneaud, 2012).
In this study, we tested the effect of thermal acclimation on the comparative thermal tolerance and locomotor performance in the invasive X. laevis and the native C. gayi, in central Chile. We use these data to address the following two questions. (i) Does the African clawed frog perform better than the Chilean frog? (ii) Are African clawed frogs more eurythermal than Chilean frogs? If the hypothesis of a greater eurythermality of invasive species is supported (Zerebecki and Sorte, 2011;Barahona-Segovia et al., 2016) then we predict that African clawed frogs will exhibit higher and broader performance than the native Chilean frogs, i.e. the invasive species will have a broader physiological tolerance than the geographically overlapping native species (Kelley, 2014).

Experimental animals
Ten adult individuals of C. gayi and X. laevis were purchased from an animal breeding store in Central Chile. While held captive, animals were maintained in a natural thermal regime. Animals were transported to our laboratory and placed in individual plastic cages of 45 cm (length) × 30 cm (width) × 20 cm (height) filled with water and fed with beef liver. All plastic cages were maintained inside a climatic chamber (Pitec). The Universidad Católica animal care committee approved all experimental procedures.

Thermal acclimation and measurements of the righting response
Before experimental procedures, individuals of both species were randomly divided into two groups (10 specimens for each species; five subjected to treatment and five to control conditions). The first group was maintained at 20 ± 1°C (warm acclimation) in a climatic chamber (Pitec), while the second group was maintained at 10 ± 1°C (cold acclimation). These temperatures were chosen because they represent average environmental temperatures experienced by X. laevis and C. gayi during summer and winter, respectively (meteorological database obtained from the Chilean government agency Dirección General de Aguas). All groups were acclimated for 6 weeks and kept with a photoperiod of 12 h light-12 h dark. This period is in accordance with previous studies performed in X. laevis and other frog species (Wilson and Franklin, 1999;Wilson et al., 2000).
After acclimation, we measured the righting response, i.e. the time it takes an individual to return to a prone position after being placed upside down, in four groups at eight different temperatures during eight consecutive days. We used the righting response as a proxy of locomotor performance, because it is rather easy to quantify in experimental conditions, shows a strong thermal dependence, is related to fitness and, more importantly, is an extremely safe method that does not compromise the animal's integrity. The experimental protocol was as follows. At 10.00 h, animals were exposed to the experimental temperature. At 18.00 h, animals were placed in a temperature-controlled room at 20 ± 1°C, and righting response measurements were performed immediately (<2 min). The procedure outlined above was repeated for eight consecutive days, and the sequence for experimental temperatures was as follows: day 1, 20 ± 1°C; day 2, 5 ± 1°C; day 3, 15 ± 1°C; day 4, 25 ± 1°C; day 5, 10 ± 1°C; day 6, 30 ± 1°C; day 7, 35 ± 1°C; and day 8, 1 ± 1°C (Fig. 1). Body mass (m b ) was measured prior to all experimental proceedings with an electronic balance (Sartorius PT-600 Portable Digital Precision Balance; precision, ±0.01 mg).

Data analysis
To address our questions, data on maximal righting performance (μ MAX ), optimal temperature (T O ), which represents the temperature at μ MAX , lower (CT min ) and upper critical thermal limits (CT max ), thermal breadth (T br ) and the area under the performance curve (AUC) were computed by adjusting the righting response data to a second-order polynomial function using the free software CurveExert professional 2.0. The performance curve breadth can be determined when performance goes to zero at either extreme. Also, the AUC is an indicator of the total integrated individual's performance and can be estimated by integrating the polynomial function.
The effect of thermal acclimation and species on all variables curve was tested by two-way analysis of covariance using structural body size as a covariate, with temperature and species as fixed factors. Structural body size was estimated by a combination of multiple morphometric measurements (body mass, head-cloaca length and body width) using a principal component analysis, and we used the first component as our measure of structural body size. The μ MAX data were normalized with respect to the maximal value in the database. Normality and homoscedasticity of data were tested using the Kolgomorov-Smirnov and Cochrane C test, respectively. When differences were significant (P < 0.05), we used the a posteriori Tukey test for multiple comparisons. Data analysis was performed using Statistica 6.1 (2006), and results are shown as means ± SEM.

Results
Summary statistics for the thermal physiological traits obtained from the thermal performance curves are presented in Table 1 Righting response, optimal temperature and area under the thermal tolerance curve Maximal performance or μ MAX was affected by species (F (1,15) = 25.23, P < 0.01) and acclimation temperatures (F (1,15) = 5.02, P < 0.05), but not by the interaction between species and acclimation temperatures (F (1,15) = 2.00, P = 0.18). Xenopus laevis had higher μ MAX than C. gayi regardless of the acclimation temperature. The effect of temperature, however, differed between species. Indeed, cold-acclimated C. gayi had higher μ MAX compared with warm-acclimated animals, whereas X. laevis maintained μ MAX independent of thermal treatment. In contrast, T O did not differ between species (F (1,15) = 1.96, P = 0.18) or temperature (F (1,15) = 0.00, P = 0.98), nor by the interaction between species and temperature (F (1,15) = 1.07, P = 0.32).
Regarding AUC, there was a significant effect of species (F (1,15) = 31.90, P < 0.01) and temperature (F (1,15) = 8.79, P < 0.01), but not an effect of the interaction between species and temperature (F (1,15) = 0.18, P = 0.68). Independently of the acclimation temperature, the invasive X. laevis had the greater AUC in comparison to the native C. gayi. The area enclosed by the thermal tolerance curve in both species was greater in cold-acclimated than in warm-acclimated frogs.

Discussion
This study addressed an important question: does the alien invasive species, X. laevis, perform better than the co-habiting native species, C. gayi? (Daehler, 2003;Jumbam et al., 2008;Van Kleunen et al., 2010;Anacleto et al., 2014;Barahona-Segovia et al., 2016). Moreover, there is increasing evidence that climate warming might also influence the dynamics of biological invasion (Dukes and Mooney, 1999;Stachowicz et al., 2002;Hellmann et al., 2008;Walther et al., 2009;Robinet and Roques, 2010). We evaluated the effect of thermal acclimation on thermal tolerance and locomotor performance in the invasive X. laevis and the native living fossil C. gayi, two aquatic amphibian species inhabiting the same habitats in central Chile.
Within this context, our key findings were higher values of μ max and AUC in the invasive species in comparison to the native one. In contrast, X. laevis showed lower values of CT min in comparison to C. gayi. The values of CT max , T O and T br showed no inter-specific differences. Moreover, we found that both X. laevis and C. gayi have the ability to acclimate their locomotor performance and lower thermal tolerance limit at low temperatures.

Is the African clawed frog a better performer than the Chilean frog?
It is widely recognized that invasive species not only have higher values for traits associated with performance than Figure 1: Performance curves of righting response in both the native Calyptocephalella gayi (circles) and the invasive Xenpus laevis (triangles) acclimated for 6 weeks at 10°C (cold acclimation; a) and 20°C (warm acclimation; b). There were 10 specimens of each species; five subjected to treatment and five to control conditions. Values are shown as means ± SEM.
Recent theoretical work by Palaima (2007) suggests that the AUC may act as an indicator of the total integrated individual's performance. In addition to the maximal value of performance and the breadth of the curve, comparisons of AUC among species also provide very valuable information to distinguish a generalist from a specialist species (Gilchrist, 1995;Gvoždík and Van Damme, 2008). Theoretically, a thermal generalist can persist in a wide range of thermal conditions and, therefore, may have a broader tolerance range than a specialist. However, a generalist will have also a much lower maximal performance than a specialist (Gilchrist, 1995(Gilchrist, , 1996Angilletta et al., 2003). Palaima (2007) also states that a thermal generalist would have higher values of AUC compared with the specialist.
Recent evidence suggests that invasive species are more likely to be generalists than specialists, and thus more successful than natives in adapting or acclimating to new habitats (Moyle and Marchetti, 2006;Evangelista et al., 2008). Our results showed that: (i) X. laevis exhibited a higher AUC in comparison to the native C. gayi; (ii) X. laevis exhibited higher values of maximal locomotor performance in comparison the native C. gayi; and (iii) the alien and native frogs did not display differences in tolerance breadth. Overall, our findings did not provide strong evidence that the alien X. laevis exhibits a 'thermally generalist' strategy, but demonstrated that X. laevis always performed better than C. gayi over the entire range of temperatures evaluated in this study.
Time use and activity of frogs and other ectotherms depends on abiotic factors, such as temperature. An acute decrease in temperature results in lower body temperatures, reduced contraction velocity of muscles and, ultimately, lower locomotor performance (Christian and Tracy, 1981;Else and Bennett, 1987;Navas et al., 1999;Herrel and Bonneaud, 2012;Sun et al., 2014). In the present study, μ MAX and AUC of both X. laevis and C. gayi acclimated to 10°C were higher than those of animals acclimated at 20°C. Our findings are in agreement with previous studies showing that African clawed frogs inhabiting Europe have the ability to compensate for the effects of lower temperatures by enhancing their swimming performance (Wilson et al., 2000). Moreover, our results are in accordance with the 'colder is better' hypothesis, which predicts that organisms acclimated to low temperatures could achieve greater locomotor performance than others held in warm conditions (Deere and Chown, 2006;Frazier et al., 2008;Esterhuizen et al., 2014).
Are the African clawed frogs more eurythermal than the Chilean frogs?
Previous evidence suggested that both CT min and CT max are highly flexible traits that respond to changes in temperature (Li et al., 2009;Chanthy et al., 2012;Ruiz-Aravena et al., 2014;Davies et al., 2015). As we mentioned above, there are higher inter-specific differences and variability in CT min , in comparison to CT max and T br , among X. laevis and C. gayi.
The CT min of cold-acclimated X. laevis was generally below the minimal temperature encountered by these species in nature during the coldest month (June 2015, −2.0°C) in Central Chile. In contrast, the CT min of both cold-and warm-acclimated C. gayi were generally above the minimal temperature encountered in nature. The CT max of both species was generally above the maximal temperature encountered in nature during the hottest month (April 2015, 37.8°C). Lobos et al. (2013) made a bioclimatic niche model considering areas susceptible to invasion. From the habitat projection in northern and southern South America, they predict high habitat suitability for X. laevis in Chile. Our results on cold tolerance are in accordance with the projected expansion of X. laevis in Chile, where it is expected that X. laevis will have a higher probability to survive, invade and establish in colder places in Chile (Lobos et al., 2013). In this context, the physiological analyses used in our study could provide new and valuable information to be considered in species invasions using ecological niche modelling, and thus improve our understanding of areas susceptible to invasion by X. laevis. This is particularly important under a global warming scenario, because the temperature is expected to increase as much as 2-4°C in central and southern Chile by 2050 (Fuenzalida et al., 2007;Cabréet al., 2010). Hence, our findings suggest that both the invasive X. laevis and the native C. gayi could be resilient to climate warming expectations in Chile.
In the present study, we found that X. laevis and the native C. gayi were able to change the lower, but not the higher, thermal limit after 6 weeks of acclimation. This finding provides support for the idea that physiological adjustments to cold and heat stress are at least uncoupled in ectotherms (Klok and Chown, 2003;Jumbam et al., 2008;Lalouette et al., 2012;Hoffmann et al., 2013;Bozinovic et al., 2014;Davies et al., 2015). For example, Araújo et al. (2013) show a direct and positive relationship between lower thermal limits and environmental temperature in amphibians.
Nevertheless, recently Bozinovic et al. (2016) reported that flies acclimated to environments with changing thermal variance reduce the scope of thermal tolerance. Furthermore, they observed that the heat tolerance of flies seems to be more conserved with lower variation among acclimation treatments. Nevertheless, this is not exact in flies acclimated to a scenario of changing mean with changing variance, where flies exhibited significantly lower CT max values. This could be explained by the fact that the physiological mechanisms underlying cold and warm tolerance could be different. Ectotherms would tolerate high temperatures by increasing the expression of genes related to protein folding as well as chaperones and proteasome proteins (Gehring and Wehner, 1995;Healy et al., 2010;Zhang et al., 2015), whereas the expression of cryoprotectants and antifreeze proteins is associated with cold tolerance (Constanzo and Lee, 2013;Long et al., 2013;Kawarasaki et al., 2014;Zhang et al., 2015).
Invasive species are often assumed to establish and spread in a new habitat as a consequence of their ability to make physiological adjustments that maintain performance across an extensive range of environmental conditions. As pointed out before, the greater eurythermal hypothesis predicts that invasive species have a broader physiological tolerance than native species occupying the same habitat (Zerebecki and Sorte, 2011;Kelley, 2014). Today, many studies support this hypothesis in both terrestrial and aquatic ectotherms (Zerebecki and Sorte, 2011;Yu et al., 2012;Bates et al., 2013;Lejeusne et al., 2014;Tepolt and Somero, 2014). In fact, a meta-analysis performed by Kelley (2014) revealed that a higher temperature tolerance was positively related to the thermal width for invasive but not native species. Our results show that although there are differences in the lower thermal critical limit, the invasive and native frogs did not differ in their thermal tolerance, rejecting the greater eurythermal hypothesis in this case. These results, however, must be viewed with caution, because locomotion is a wholeorganism functional trait whose expression requires elaborate biochemical and metabolic integration. Hence, the establishment and spread of invasive species into new habitats can be explained by adjustments of the physiological systems supporting locomotion (Lefrançois et al., 2005;Cano and Nicieza, 2006;Anacleto et al., 2014). For example, Seebacher and Franklin (2011) documented that mitochondrial ATP production capacity could constrain the advance of cane toads to cooler southern areas of Australia. Finally, we emphasize the importance of future studies that consider other physiological traits to understand the mechanisms that underlies the thermal differences in performance between the invasive X. laevis and the native C. gayi, two frog species inhabiting central Chile.