Introduction
Biological invasions enable researchers to conduct studies at large spatial and temporal scales that capture substantial environmental variation, facilitating the ability to observe ecological and evolutionary processes over contemporary timescales (Sakai et al., 2001; Sax et al., 2007; Lawson et al., 2011; Hodgins et al., 2018). When non-native populations encounter new environments in terms of predators, competitors, or abiotic conditions, this variation can relax selection experienced in their native range and/or impose novel selective pressures (Sakai et al., 2001; Lee, 2002; Bock et al., 2015, 2021). Invasions have provided many recent examples of rapid adaptation (e.g., Colautti & Barrett, 2013; Stern & Lee, 2020; Battlay et al., 2023). Previous studies have demonstrated phenotypic shifts in non-native species driven by natural selection occurring over the course of tens to hundreds of generations (e.g., Huey et al., 2000; Shine, 2011; Colautti & Barrett, 2013; Bock et al., 2018). Yet not all invasions result in rapid adaptation, even when non-native populations experience novel environments, due to a variety of factors such as a lack of associated selective pressures, little relevant genetic variation, or possible genetic constraints (e.g., Alexander & Edwards, 2010; Bock et al., 2021, Baeckens et al., 2023, Bock et al., 2023).
In addition to examples of rapid adaptation, studies of biological invasions have revealed the importance of the invasion history for patterns of demographic, genetic and phenotypic variation (e.g., Kolbe et al., 2004; Cristescu, 2015). I
nvasion history includes the contribution to genetic ancestry through the location and number of source populations in the native range (Kolbe et al., 2004, 2007a), the spread of populations in the non-native range (e.g., leading to ‘expansion load’; Peischl et al., 2013), and the extent of admixture (i.e., intraspecific hybridization of previously isolated sources) within non-native populations (Rius & Darling, 2014; Bock et al., 2021), among other factors. As a result, genetic and phenotypic variation can increase or decrease in invasive populations during an invasion. For example, admixture will often increase variation (Kolbe et al., 2004, 2008; Rius & Darling, 2014), whereas random genetic drift will decrease variation through founder effects and population bottlenecks (Dlugosch & Parker, 2008; Ficetola et al., 2008; Zhu et al., 2017). The invasion history is expected to have a strong effect on patterns of genetic and phenotypic variation within and among non-native populations, and can strongly bias the interpretation of drivers of evolution during invasions, when not properly accounted for (Keller & Taylor, 2008; Colautti et al., 2009; Hodgins et al., 2018).
Therefore, understanding evolution in introduced populations will often require elucidating the combined effects of invasion history and adaptation to environmental conditions encountered in the novel range.
The biological invasion of the brown anole lizard (Anolis sagrei ) provides an excellent opportunity to test hypotheses for phenotypic evolution among non-native populations. Native to the northern Caribbean—Cuba, Bahamas, Cayman Brac and Little Cayman—the brown anole is a very successful invasive lizard that has been repeatedly introduced to Florida (USA) starting in the late 19th century (Garman, 1887) and expanding rapidly beginning in 1950s (Campbell, 1996). At least eight introduction events from genetically distinct native-range populations have resulted in non-native populations with greater genetic diversity than those observed in the native range (Kolbe et al., 2004, 2007a, 2008; Bock et al., 2021). Also, non-native Florida populations are more morphologically variable in body size and shape, number of toepad lamellae, and number of scales compared to native Cuban populations (Lee, 1992; Kolbe et al., 2007a). This increased phenotypic and genetic variation may be acted upon by natural selection, which could promote evolution and facilitate invasion success (e.g., Shine 2011, but see Jaspers et al., 2021). Introduced populations differ significantly from each other in genetic variation and certain morphological traits (Kolbe et al., 2007a, 2008; Bock et al., 2021). Thus, the invasion history may dominate patterns of phenotypic and genetic variation in these non-native populations. Indeed, genetic ancestry rather than local adaptation better explains variation among non-native brown anole populations in morphology and water loss traits (Kolbe et al., 2007a; Bock et al., 2021; Baeckens et al., 2023). Moreover, novel genetic interactions that result from hybridization (e.g., Bock et al., 2021) and high linkage disequilibrium in chromosomal segments of reduced recombination (i.e., Bock et al., 2023) may play a role in constraining adaptive responses in the invasive range.
Given the lack of previous evidence for local adaptation among non-native brown anole populations, we wanted to focus on a phenotypically and genetically complex trait that is subject of multiple selective pressures to increase the chance of detecting evidence of natural selection.
We therefore examined evolutionary change across the invasive range in the iconic Anolis signaling ornament: the dewlap. Dewlaps are extendable structures located on the throat of anoles that differ in size, shape, color, and patterning at both the inter- and intraspecific levels (Losos, 2009). This multifaceted signaling structure often plays an important role in territory establishment and defense (e.g., Fleishman, 1992), reproductive interactions (e.g., Crews, 1975), species recognition (e.g., Rand & Williams, 1970; Losos, 1985), and predator deterrence (e.g., Leal & Rodríguez-Robles, 1995; Leal & Rodríguez-Robles, 1997a,b). Considerable evidence supports that the dewlap is subject to a variety of selection pressures (Leal & Fleishman, 2002, 2004; Vanhooydonck et al., 2009; Ng et al., 2013a; Driessens et al., 2017). Previous research has shown strong, albeit contrasting, relationships between environmental conditions and dewlap design in native populations of Anolis lizards (Leal and Fleishman, 2002, 2004; Ng et al., 2013a; see Discussion for details). For brown anoles, native-range populations inhabiting mesic environments had primarily marginal dewlaps (i.e., red or orange covering most of the dewlap with yellow along the outer margin), showing high reflectance in red, whereas lizards occupying xeric environments had a higher proportion of solid dewlaps with higher ultraviolet (UV) reflectance (Driessens et al., 2017).
These results show that Anolis dewlap phenotypes are associated with variation in environmental conditions, likely resulting from differential selection on signal effectiveness in response to climatic conditions and physical habitat characteristics (Endler, 1992, 1993) as well as interactions with competitors, predators, and conspecifics (Cole, 2013).
In this study, we investigated multiple aspects of a complex trait in brown anole populations across a large portion of its non-native range in the southeastern United States. Our goal was to determine whether among-population variation in aspects of dewlap design can be best explained by environmental conditions (consistent with local adaptation), genetic ancestry, or a combination of both factors. Specifically, we tested whether variation in dewlap characteristics (i.e., color, pattern and size) were associated with 1) differences in prevailing environmental conditions (i.e., temperature, precipitation, and light conditions) and 2) genetic ancestry as measured by the relative contribution of genome-wide SNP variation inherited from different native-range source populations (Kolbe et al., 2004, 2007a, 2008, Bock et al., 2021). We complemented the latter ancestry-based tests with genome-wide trait association analyses, taking advantage of recent analytical developments that allow ancestry-specific associations to be detected along the genome (e.g., Skotte et al., 2019). Further, we relied on F ST outlier analyses and genotype-environment association analyses, which can identify loci involved in adaptation. The phenotypically diverse and putatively polygenic nature of complex traits, such as the Anolis dewlap, afford an excellent opportunity to assess whether changing environmental conditions during an invasion shape phenotypic variation in ways consistent with natural selection. The brown anole invasion is a particularly useful case because previous studies of morphological and physiological traits have failed to provide evidence of adaptation during the invasion (i.e., Kolbe et al., 2007a, Bock et al., 2021, Baeckens et al., 2023, Bock et al., 2023). A better understanding of the evolutionary dynamics during biological invasions is useful for predicting future impacts of invasive species, especially as global environments continue to change rapidly.