2 Department of Sciences, Museums Victoria, Melbourne, VIC, Australia
Abstract
Differences in the geographic scale and depth of phylogeographic structure across co-distributed taxa can reveal how microevolutionary processes such as population isolation and persistence drive diversification. In turn, environmental heterogeneity, species’ traits and historical biogeographic barriers may influence the potential for isolation and persistence. Using extensive SNP data and a combination of population genetic summary statistics and landscape genomic analyses, we explore predictors of the scale and depth of phylogeographic structure in co-distributed lizard taxa from the topographically and climatically complex monsoonal tropics (AMT) of Australia. We first resolve intraspecific lineages and then test whether genetic divergence across space within lineages is related to isolation by distance, resistance and/or environment, and whether these factors differ across genera or between rock-related versus habitat generalist taxa. We then test whether microevolutionary processes within lineages explain differences in the geographic scale and depth of intraspecific phylogeographic lineages. Results indicate that landscape predictors of phylogeographic structure differ between taxa. Within lineages, there was prevalent isolation by distance, but the strength of isolation by distance is independent of the taxonomic family, habitat specialization and climate. Isolation by environment is the strongest predictor of landscape-scale genetic divergence for all taxa, with both temperature and precipitation acting as limiting factors. The strength of isolation by distance does not predict the geographic scale of phylogeographic structure. However, more localized lineages had higher mean individual heterozygosity and less negative Tajima’s D. This implies that finer-scale phylogeographic structuring within species is associated with larger and more stable populations and, hence, persistence.
Keywords
Population structure, gene flow, isolation by distance, geographic barriers, environmental limitation
Introduction
The immense variety of life forms that we observed today is a result of the speciation process- and all the factors that promote speciation initiation. Speciation is an extended process where the potential for persistence of population isolates is central to whether intraspecific lineages are ephemeral, or will eventually transition to full species under the protracted speciation model (Etienne et al., 2012; Rosenblum et al., 2012). Population isolation and high persistence of isolates manifests as a strong phylogeographic structure across the range of species (or parapatric species complexes). It follows that the phylogeographic structure of taxa depends on how a combination of species traits (e.g. specialization to spatially patchy habitats), environmental heterogeneity and biotic interactions (Harvey, Aleixo, Ribas & Brumfield, 2017; Riginos, Buckley, Blomberg, & Treml, 2014) influence isolation and/or persistence of populations (Funk et al., 2005; Oliveira et al., 2018; Pease et al., 2009). Understanding how genetic differentiation is shaped by environmental heterogeneity and dispersal limitation can help predict potential for adaptive divergence (Wang & Bradburd, 2014), in addition to informing how geographical and climatic barriers influence phylogeographic structure (Li, Huang, Sukumaran, & Knowles, 2018).
A broad spatial view of the population dynamics associated with species formation provides a bridge of two largely isolated research approaches – macro and microevolution (Harvey, Singhal & Rabosky, 2019). While macroevolution is broadly concerned with rates of speciation and extinction among groups, microevolutionary processes include intrinsic dispersal limitation, estimated using isolation by distance (Singhal et al., 2018; Strien, Holderegger, & Van Heck, 2015; Wright, 1943), extrinsic dispersal limitation related to the presence, currently or in the past, of geographic and physiological dispersal barriers (Myers et al., 2019; Pelletier & Carstens, 2018; Yannic, Hagen, Leugger, Karger & Pelissier, 2020), and population persistence through past climates (Bell et al., 2010; Ortego, Gugger, Riordan, & Sork, 2014; Vasconcellos et al., 2019).
Here we apply landscape genetics methods to a tropical reptile fauna in which most species have phylogeographic structuring, but with varying spatial and temporal scales. We explore whether dispersal within lineages is limited by geographic distance (isolation by distance (IBD); Wright, 1943), environmental heterogeneity (isolation by environment (IBE); Wang & Bradburd, 2014), or geographic barriers (isolation by resistance (IBR); Cushman, McKelvey, Hayden, & Schwartz, 2006; Spear, Balkenhol, Fortin, McRae, & Scribner, 2010), and how these relationships vary across rock-restricted vs. habitat-generalist species, or across genera, representing broader trait variation. We also test whether phylogeographic structuring is more geographically localized in taxa with lower dispersal (stronger IBD), larger local population size (higher average heterozygosity) or more stable population size assessed using departure from mutation-drift equilibrium (i.e. lower absolute Tajima’s D). We apply this comparative approach to the lizard fauna of the Australian Monsoonal Tropics (AMT).
The AMT is a diverse and biologically rich region characterized by disjunct sandstone plateaus or areas of relatively high topographic relief separated by regions of flat savanna woodland (Figure 1). Aridity increases from north to south (Figure 1A). The AMT has strongly contrasting rainfall across wet and dry seasons, set against consistently warm temperatures (Figure 1B), and is the most fire-prone region in Australia (Bowman et al., 2010). Previous studies identified three biogeographically distinct areas within the AMT — the Top End, Kimberley and Cape York Peninsula (Figure 1A; Bowman et al., 2010; Woinarski, Mackey, Nix, & Traill, 2007), with the latter being the most biogeographically distinct (Fujita, McGuire, Donnellan, & Moritz, 2010; Lee & Edwards, 2008; Smith, Harmon, Shoo, & Melville, 2011). The Top End is more mesic than the Kimberley (Figure. 1A). Large river plains separate major plateaus in both regions (Figure 1A) and, along with climatic and edaphic barriers (e.g. the Gulf plains and Ord arid intrusion), represent a potential driver of allopatric divergence within terrestrial species (Catullo, Lanfear, Doughty, & Keogh, 2014; Eldridge, Potter, & Cooper, 2011).
Recent multilocus phylogeographic analyses of the AMT lizard fauna have consistently revealed strong phylogeographic structure within taxonomically recognized species. In some cases, cryptic species complexes have subsequently been taxonomically revised (Afonso Silva et al., 2017; Doughty et al., 2018; Melville, Date, Horner, & Doughty, 2019; Oliver et al., 2019; 2020), while assessments of species boundaries for other groups are still in progress (Catullo et al., 2014; Fujita et al., 2010; Laver, Doughty, & Oliver, 2018; Laver, Nielsen, Rosauer, & Oliver, 2017; Melville, Ritchie, Chapple, Glor, & Li, 2011; Potter et al., 2018). In some taxa (notably rock-specialist geckos; Laver et al., 2018; Moritz et al., 2018; Oliver et al., 2020) this phylogeographic structure can occur at very fine spatial scales, whereas in other taxa (e.g. more generalist Carlia skinks, Afonso Silva et al., 2017); Diporiphora dragons, Smith et al., 2011); andHeteronotia geckos, Moritz et al., 2016) the phylogeographic units typically span broader geographic ranges (Figure 1C). Phylogeographic lineages (whether or not now reclassified as separate species) are also much older in Gehyra than are those in the other genera (Figure 1C). There are also regional differences relating to climate and topography. Short-range phylogeographic lineages are concentrated in the more mesic regions of the Top End and west Kimberley, including on islands (Rosauer et al., 2016, 2018) and isolated karst limestones (Oliver et al., 2017). For Carlia, there were more variable demographic responses to past climate change in the drier Kimberley than the more mesic Top End (Potter et al., 2018).
This variation in the spatial and temporal scale of phylogeographic structure provides the opportunity to apply landscape genetics within lineages to test how intrinsic dispersal limits and ecological specialization interact with environmental heterogeneity in shaping the scale of phylogeographic structure within species. Using newly obtained SNP data and landscape-scale sampling with phylogeographic lineages, we 1) tested whether dispersal limitation within lineages (IBD) varies among taxa (families), ecologies (habitat requirement) or environmental condition; 2) explored what environmental features best explain genetic divergence within lineages (Isolation by distance/resistance/environment); and 3) tested whether microevolutionary processes, reflected in within-lineage genetic parameters, explain differences among taxa in the geographic scale and depth of phylogeographic lineages.
Material and Methods