1 INTRODUCTION
Anthropogenic activities have led to rapid global changes that are creating unprecedented climatic conditions, which could compromise the viability of many species (Walther et al. 2002; Walther et al. 2009). Understanding how environmental conditions affect the survival of species’ populations and shape their distributions is fundamental for establishing conservation strategies (Bosch et al. 2018). However, estimating the relationships between species and their environment is a complex task, as it involves multiple factors (Gaston and Fuller 2009; Sexton et al. 2009; Wiens 2011). These factors can be summarized into three categories: 1) limiting or regulatory, which control the ecophysiological responses of the species, 2) disturbance factors, which describe the historical modifications in the occupation of habitats by both natural and anthropogenic processes, and 3) resource factors, which represent the supplies necessary for the survival of organisms (Guisan and Thuiller 2005). Some causal relationships between these factors and species’ attributes, such as abundance and distribution, have been demonstrated through physiological experiments in laboratories and wildlife demographic studies (Gause 1931; Hutchings and Myers 1994; Kordas et al. 2011). However, carrying out laboratory experiments or fieldwork to obtain demographic data has some drawbacks. First, both are generally costly and time-consuming. Secondly, implementing physiological experiments is not ethically feasible for many organisms (Fraser 1999). Finally, some environmental variables are more relevant at broad spatial scales (>200 km). For instance, limiting or regulating factors associated with climate are considered to be the most relevant in explaining the distribution of species across their entire range (Contreras-Díaz et al. 2022; Gillespie et al. 2008; Pearson and Dawson 2003).
An alternative approach to studying species-environment relationships in a spatial context is the use of correlative methods, known generically as ecological niche modeling (ENM) and species distribution modeling (SDM) (Bentlage et al. 2013; Pearson and Dawson 2003; Peterson and Soberón 2012). The inputs required by correlative methods are easier to acquire since they are generally available in online repositories. These inputs are sets of georeferenced localities where the presence of the species in question has been observed (some methods also require points that represent absence) and spatial surfaces that describe the environmental conditions of the area of interest (Randin et al. 2009). Thus, correlative methods are extensively used to model species-environment relationships and estimate their actual and potential distributions, both in terrestrial (Colwell and Rangel 2010; Eme et al. 2014) and marine ecosystems (Bosch et al. 2018; Bradie and Leung 2017). Although correlative methods have certain limitations (e.g., it can be difficult to discern whether their results are mere statistical associations or causal relationships; see Wiens et al. 2009), they allow us to recognize diverse ecological patterns (Ashcroft et al. 2011). As a result, these methods have become popular for analyzing the importance of environmental variables as limiting factors of distribution ranges and various ecological aspects across multiple taxa (Bosch et al. 2018; Bradie and Leung 2017; Lee-Yaw et al. 2016).
One of the most studied limiting factors is temperature, as it has a key relevance for many physiological and ecological processes (Bicego et al. 2007; Clarke 2003; Kordas et al. 2011). Thus, studying the thermal niche (TN) and realized thermal niche (TNR) of species has been a major focus of ecological research (Gvoždík 2018; Lee-Yaw et al. 2016; Magnuson et al. 1979; Stuart-Smith et al. 2017). According to Gvoždík (2018), TN refers to the range of body temperature that individuals of a species require for the population to experience positive growth, while TNR refers to the environmental range of temperatures that individuals of a species are exposed to (Collin et al. 2021; Stuart-Smith et al. 2017). Based on the concepts of TN and TNR, two hypotheses have been proposed to explain patterns of temperature tolerance and organismal response to global warming. The first hypothesis suggests that tropical and polar terrestrial species have lower TN than those from temperate climates (Pörtner 2001; Pörtner and Farrell 2008). The second hypothesis is the realized thermal limit asymmetry, which refers to significant differences in the variation of upper limits in a species or taxon compared to the variation of lower limits within the same taxon, with the variation being greater in the lower limits. This means that under climate change scenarios, high rates of global warming can severely impact the conservation of most species (Collin et al. 2021; Stuart-Smith et al. 2017). Ideally, evaluating these hypotheses requires direct estimation of the species’ TN in question through physiological population growth-mortality experiments (Dong and Somero 2009; Helmuth et al. 2006; Sánchez-Fernández et al. 2012; Stillman and Somero 2000). However, despite the high cost, time and ethical constrains, there are global initiatives to analyze this patterns (Bennett et al. 2018). Nonetheless, characterization of the TN for most taxa is still lacking.
Therefore, through the use of correlative methods, it is possible to characterize a portion of the TN called TNR, which represents the range of temperatures associated with the presence of a species (Descombes et al. 2015; Gaüzère et al. 2015; Quintero and Wiens 2013). Since correlative methods do not require experimentation and management of organisms, they are a more feasible option for studying species TNs. As a result, they have been commonly applied to many taxonomic groups (Huff et al. 2005; Kuo and Sanford 2009; Sánchez-Fernández et al. 2012; Sánchez-Fernández et al. 2016). Moreover, most TNs analysis have been focused on species from terrestrial environments (Bennett et al. 2018), leaving no information on many important marine groups, such as sea snakes.
Sea snakes are an ecologically diverse and marine-adapted lineage of elapids that includes two subfamilies, Hydrophiinae (true sea snakes) and Laticaudinae (sea kraits). The group comprises >60 species widely distributed in the tropical and subtropical regions of the Indian and Pacific Oceans (Sanders et al. 2013; Wallach et al. 2014). Sea snakes’ lineage has evolved with unique physiological adaptations to survive in marine environments, being their ability to thermoregulate a critical adaptation to adjust their metabolic rate and behavior in response to water changes in temperature (Heatwole et al. 2012). Recent research has also shown that temperature plays a critical role in the diving behavior and metabolism of sea snakes, as individuals exhibit different diving behaviors and metabolic rates in response to changes in water temperature (Dabruzzi et al. 2012; Udyawer et al. 2016). These findings highlight the complex ways in which temperature influences the behavior and physiology of sea snakes and underscore the importance of understanding the thermal biology of these reptiles for conservation and management efforts. Nevertheless, most sea snake research has focused on their toxicity, and their phylogenetic and evolutionary relationships (Udyawer et al. 2018). However, to date, there is no information on the factors that determine sea snakes’ distribution, nor the asymmetry of their realized thermal limits. In this study, we used correlative ecological niche modeling (ENM) to evaluate the relative importance of environmental variables in explaining the distribution of almost all sea snake species on a global scale. We also analyzed whether there is an asymmetry in the dispersion of their lower and upper realized thermal limits. The results of our research provide a better understanding of the biogeographical phenomena that determine the distribution patterns in this group and will allow us to infer their vulnerability to future ocean warming scenarios.