Effect of climate change in lizards of the genus Xenosaurus (Xenosauridae) based on projected changes in climatic suitability and climatic niche conservatism

Abstract Accelerated climate change represents a major threat to the health of the planet's biodiversity. Particularly, lizards of the genus Xenosaurus might be negatively affected by this phenomenon because several of its species have restricted distributions, low vagility, and preference for low temperatures. No study, however, has examined the climatic niche of the species of this genus and how their distribution might be influenced by different climate change scenarios. In this project, we used a maximum entropy approach to model the climatic niche of 10 species of the genus Xenosaurus under present and future suitable habitat, considering a climatic niche conservatism context. Therefore, we performed a similarity analysis of the climatic niche between each species of the genus Xenosaurus. Our results suggest that a substantial decrease in suitable habitat for all species will occur by 2070. Among the most affected species, X. tzacualtipantecus will not have suitable conditions according to its climatic niche requirements and X. phalaroanthereon will lose 85.75% of its current suitable area. On the other hand, we found low values of conservatism of the climatic niche among species. Given the limited capacity of dispersion and the habitat specificity of these lizards, it seems unlikely that fast changes would occur in the distribution of these species facing climate change. The low conservatism in climatic niche we found in Xenosaurus suggests that these species might have the capacity to adapt to the new environmental conditions originated by climate change.

can have serious potential consequences for most species (Albon et al., 2017), particularly those that strictly depend on specific conditions of habitat (Parmesan, 2006).
In this context, Wiens et al. (2010) and Liu et al. (2017) proposed that the threat of climate change can be analyzed by climatic niche conservatism. This concept refers to the retention of certain characteristics of the ancestral fundamental niche over time and space (Wiens et al., 2010). Accordingly, it has been suggested that if the climatic tolerance of a species is not extensive enough to face new environmental conditions, then those species with strong niche conservatism must migrate or become extinct (Jackson, Gergel, & Martin, 2015;Wiens et al., 2010). For example, some reports have shown that several lizard species tend to maintain similar thermal preferences despite inhabiting different types of environments (Andrews, 1998;Bogert, 1949; Díaz de la Vega-Pérez, Jiménez-Arcos, Manríquez-Morán, & Méndez-De la Cruz, 2013; Grigg & Buckley, 2013;Rocha & Vrcibradic, 1996;Schall, 1977). This represents strong evolutionary evidence of the presence of conservatism in optimal thermal preferences in these groups of vertebrates (Adolph, 1990;Menezes & Rocha, 2011;Pianka, 1970;Stevens, 1982;Winne & Keck, 2004).
Range-restricted species often are particularly vulnerable to extinction due to the climate change that is happening presently; therefore, endemic species could be strongly affected (Ballesteros-Barrera, Martínez-Meyer, & Gadsden, 2007;Malcolm, Liu, Neilson, Hansen, & Hannah, 2006). This condition could be even more problematic if these species tend to conserve their climatic niche (Wiens et al., 2010). Many studies have used species distribution modeling to predict habitat suitability of endemic species in the future (García, Ortega-Huerta, & Martínez-Meyer, 2013;Thuiller et al., 2006). However, evaluating whether these studied species conserve their niche has not been performed commonly. Therefore, besides evaluating the availability of areas with suitable climatic conditions for species in the future, it is also important to evaluate whether species show a tendency to conserve their climatic niche, which would help provide more information to better understand their capacity of response to climate change; these data could be obtained from reptiles, because as ectothermic organisms, they are good models to use to assess the impacts of climate change (Huey & Kingsolver, 1993;Valenzuela-Ceballos, Castañeda-Gaytán, Rioja-Paradela, Carrillo-Reyes, & Bastiaans, 2015).
Lizards of the genus Xenosaurus comprise a group of northern forms of Laurasian origin with a diversification in North America; therefore, they are considered as one of the oldest groups of lizards (Macey et al., 1999;Bhullar, 2011; Figure 1). Currently, distribution of the Xenosaurus species is associated with mountain chains that occur from northeastern Mexico to Alta Verapaz, Guatemala (Bhullar, 2011). Species of Xenosaurus occur in a broad altitudinal range of approximately 500-2,360 m, and they can be found in a wide variety of habitats, ranging from xerophytic tropical scrub to tropical montane cloud forest, and tropical rain forest (King & Thompson, 1968

| Selection of climatic variables
Climatic information was obtained from the 19 current climatic

| Climatic niche modeling (habitat suitability)
We used the Maximum Entropy modeling method (Phillips, Anderson, & Schapire, 2006) implemented in the software MaxEnt version 3.3.1 (Phillips, Dudik, & Schapire, 2004), to develop niche models of habitat suitability for each Xenosaurus species. MaxEnt is a maximum entropy-based program that estimates the probability distributions of a species in a geographic space using occurrence records and environmental data (Phillips et al., 2006). This modeling algorithm only requires presence data and has relatively good performance when compared to other presence-only methods (Elith et al., 2006). To establish the areas that maintain the environmental conditions of the species and that could be occupied by them (Burgman & Fox, 2003;Elith et al., 2011), MaxEnt generates habitat suitability maps scaling from zero (low suitability) to 1 (high suitability; Elith et al., 2011). Data of the seven bioclimatic variables used for the study were collected for the whole background using the software ArcMap version 10.3.
Successful calibration is important for datasets that suffer from sampling bias and for studies that require transfer models through space or time, for example, climate change response (Elith, Kearney, & Phillips, 2010;Moreno-Amat et al., 2015). Therefore, for the construction of the models, the dataset used in MaxEnt was randomly divided as follows: 75% of the data was used for the model construction or calibration, and the remaining 25% was used to evaluate the adjustment of the model (Cianfrani, Lay, Hirzel, & Loy, 2010).
Information obtained from each model was projected in order to identify potential areas under current climatic conditions and the scenario of the climate change RCP85-2070. We used projections to the year 2070 employing the environmental variables records available on the WorldClim database, which are calculated from future climate projections of General Circulation Models (Hijmans et al., 2005). The RCP 85 assumes that global greenhouse gas concentration trajectories will continue to rise throughout the XXI century and will stabilize in the year 2100 (Meinshausen et al., 2011). We evaluated MaxEnt model performance using a fifteenfold cross-validation of the area under the curve (AUC) of the receiver operating characteristic curve (ROC). Models with an AUC score of 0.5 indicated a model performing randomly, while models with an AUC score of 1 indicated a perfect model (Lobo, Jiménez-Valverde, & Real, 2007;Phillips, Dudik, & Schapire, 2004

| Similarity of climatic niche
Comparison of climatic niche among the 10 species of the genus was carried out using the analytical framework developed by Niche overlap between species in the environmental space was measured by the Schoener D metric (Schoener, 1970), and niche similarity tests were performed according to Warren, Glor, and Turelli (2008), using 100 randomizations in the null model. When the observed niche overlap value was significant (p < 0.05) based on this two-way test (similarity of species A vs. B and of species B vs. A), the climatic niches of both species were considered similar, indicating that one species predicts the climatic niche of the other better than would be expected by chance under a specific null model.

| Habitat suitability modeling
Distribution models showed an AUC value above 0.74 for the 10 species of Xenosaurus, thus suggesting that the obtained models had high quality and the bioclimatic variables that most contribute to the models' calibration were Bio7 = Temperature Annual Range, Bio12 = Annual precipitation, and Bio14 = Precipitation of driest month (Table 1).

Considering a scenario of climatic niche conservatism, all
Xenosaurus species showed a decrease in their areas of habitat  Figure 3).

| Similarity of climatic niche (conservatism)
The results of the analyses of climatic niche similarity are presented in  Figure 4). will decrease due to climatic change. Therefore, on the basis of the aforementioned evidence, it is reasonable to infer that the genus Xenosaurus will be exposed to a high vulnerability risk due to climatic change. In this sense, Lemos-Espinal et al. (2003a) pointed out that X. grandis might persist in croplands that provide the minimum requirements of microhabitat and native vegetation cover.

| D ISCUSS I ON
However, the areas predicted by its habitat suitability model in the future, include croplands and urban settlements, which reduce the TA B L E 1 Comparison of the top model runs for each species. Values of area under the receiver operating curve (AUC), maximum training sensitivity plus specificity of the models of habitat suitability of the 10 species of the genus Xenosaurus are represented. The bioclimatic variables that contributed the most to their construction for each species of Xenosaurus were: Bio1 (annual mean temperature), bio7 (annual range temperature), bio11 (annual mean temperature of the coldest quarter), bio12 (annual precipitation), bio 14 (precipitation of the driest month), bio17 (precipitation of the driest quarter), and bio18 (precipitation of the warmest quarter) possibility of occupying these sites due to the particular ecological characteristics of this species, such as high microhabitat specificity and low dispersion capacity, and therefore, the future survival of its populations will be at high risk (King & Thompson, 1968 (Lapola, Oyama, & Nobre, 2009). Even if environmental conditions were maintained in the future, changes in vegetation structure could affect these tropical species negatively (Turner, 1996).
According to the variables that contributed the most to the climatic niche models (see Table 1), it is evident that precipitation regime is very important for Xenosaurus lizards; also, our results indicate that a decline in precipitation conditions for this species in the area they inhabit is expected under all scenarios, causing more hostile environments for the lizards. For example, Lavergne, Mouquet, Thuiller, and Ronce (2010) noted that organisms might not be able to adapt to climate change if the rate of change is too rapid and the demography is not sufficiently dynamic. Therefore, one of the crucial questions in the debate on ecological effects of climate change is whether species will be able to adapt rapidly enough to keep up with the rapid pace of changing climate (Parmesan, 2006;Wiens et al., 2010).
Our results show that members of the genus Xenosaurus have a low similarity among their climatic niches, suggesting low niche conservatism and a tendency to niche shift. Such patterns have been found already in other animal species (Pyron & Burbrink, 2009;Rodrigues, Pacheco Coelho, & Diniz-Filho, 2016;Strubbe, Beauchard, & Matthysen, 2015). This also might explain the high environmental variability reported for most of the species of Xenosaurus (King & Thompson, 1968;Nieto-Montes de Oca, Campbell, & Flores-Villela, 2001   Sp1   (Ballinger et al., 1995;Lemos-Espinal & Smith, 2005). Therefore, our models could have biases in their accuracy, because these models only consider environmental variables at macro-scale levels and do not take into account micro-environmental conditions and interactions (Mateo, Felicísimo, & Muñoz, 2011). Therefore, we propose to do more fine-scale studies of ethology and thermal ecology (López-Alcaide, Nakamura, Macip-Ríos, & Martínez-Meyer, 2014) in order to establish greater precision in our understanding of the response of these species to increase in the ambient temperature.
Finally, in order to cushion the effects of global climate change, we propose to maintain the remnants of native vegetation, as well as rocky outcrops where Xenosaurus populations live, in order to maintain specific habitat conditions. In addition, the conservation status of these lizards must be carefully reviewed in order to fulfill the central obligations of species conservation, particularly for such a vulnerable group.

ACK N OWLED G M ENTS
The authors would like to thank Jonathan Campbell, Vicente

CO N FLI C T O F I NTE R E S T
None declared.