Introduction
At its core, genetic diversity is the foundation upon which biodiversity flourishes. It provides the raw material for adaptation to local environments and divergence among populations. As intraspecific genetic diversity is the necessary precursor to speciation, exploring global trends in genetic diversity may shed light on the mechanisms, or combinations of mechanisms, that drive patterns of species diversity. Similarly, elucidating the processes that generate genetic diversity helps create a common ground for evolutionary biology and community ecology around topics of diversity and patterns of speciation (Vellend & Geber 2005). Despite this importance, the general patterns of genetic diversity across species remains poorly understood at global scales (de Kort et al. 2021; Manel et al. 2020; Messmer et al. 2012; Miraldo et al. 2016). Efforts to understand such trends are vital for identifying the factors creating and maintaining biodiversity and for pinpointing high priority areas and taxa for conservation.
Here, we propose three distinct hypotheses for genetic diversity gradients, all of which are grounded in foundational community ecology and population genetics theory. The first is the Kinetic Energy Hypothesis, which posits that, like species richness, intraspecific genetic diversity should be greater at hotter temperatures due to faster evolutionary tempos (e.g., higher metabolic and mutation rates), particularly in mitochondrial DNA that is affected by oxidative damage from metabolic processes (Allen et al. 2002; Manel et al. 2020; Wright et al. 2011). While oxidative damage should not influence nuclear DNA mutation rates (Hoffman et al. 2004), genome-wide mutation rates are negatively correlated with generation times (Thomas et al. 2010), which are shorter in organisms with smaller body sizes (Martin & Palumbi 1993), and, by Bergmann’s rule, inversely related to temperature (Bergmann 1847; Fernández-Torres et al. 2018). Thus, nuclear genetic diversity may also be weakly correlated with temperature, albeit not to the degree of mitochondrial diversity (Gillooly et al. 2004). The second hypothesis, the Productivity-Richness Hypothesis (Evans et al. 2004), suggests that population size is often constrained by resource availability, such that regions of high primary productivity should support larger populations with higher intraspecific genetic variation, since large populations lose genetic diversity to genetic drift at a slower rate (Charlesworth 2009; Wright 1983). Finally, the Founder Effect Hypothesis proposes a negative relationship between latitude and genetic diversity, a lasting legacy from the last glacial maximum (LGM) (Hewitt 2000). As species expanded from equatorial to temperate and polar latitudes, a sequential series of founder and bottleneck events along the expansion front depleted standing genetic variation, leaving a latitudinal genetic footprint that is still apparent in many modern populations (Jenkins et al. 2018; Lessa et al. 2010; Mattingsdal et al. 2020). For marine species, this effect is particularly pronounced in the Northern hemisphere, as many contemporary high-latitude taxa in the Southern Ocean may have endured the LGM in local polar refugia (Allcock & Strugnell 2012; Fraser et al. 2012).
Much of our knowledge on intraspecific genetic diversity, including local and regional estimates in various taxa, has only been collected in recent decades. Recent studies compiled these data to construct broad-scale geographic patterns in order to better understand global distributions of genetic diversity (Li et al. 2021; Manel et al. 2020; Miraldo et al. 2016). For example, Martin & McKay (2004) revealed greater genetic divergence among vertebrate populations at lower latitudes, while Schmidt et al. (2022) found that environmental heterogeneity was an important predictor of genetic diversity in North American mammals. Large knowledge gaps still exist, however, as the strength and direction of latitudinal gradients in genetic diversity appear to vary across taxa and ecological systems (de Kort et al. 2021). In particular, it remains unclear how universal such patterns are and how influential underlying ecological drivers may be. This is particularly true of marine communities, as most meta-analyses of global genetic diversity to date have focused on either terrestrial or freshwater systems (but see Manel et al. 2020).
While the same evolutionary processes occur in all taxa, the strength of these forces (particularly selection, drift, and gene flow) differ substantially across terrestrial and marine realms (Steele et al. 2019; Strathmann 1990). Marine species tend to exhibit larger populations, higher gene flow, and wider species ranges (Steele et al. 2019). Alleles may be more easily transported throughout species ranges and latitudes in marine systems, muting the effects of the local environment and weakening the consequences of genetic drift. Such patterns have previously been documented, including evidence that strong dispersal helped maintain high diversity in range edge populations of Senegal seabream, Diplodus bellottii , and erased typical core-periphery patterns of genetic variation (Robalo et al. 2020). Moreover, global patterns of species richness tend to differ between land and sea. Marine taxa commonly display bimodal latitudinal gradients of species richness (Chaudhary 2016; Tittensor et al. 2010), peaking at mid-latitudes instead of along the equator as is more common in their terrestrial counterparts (Davies et al 2007; Rolland et al 2014). Given these unique differences, more work is needed to determine how the distinctive environmental conditions and life-history strategies in the ocean combine to shape macroecological patterns of genetic diversity.
In addition, most macrogenetic studies have primarily investigated patterns of mitochondrial genetic diversity, despite suggestions that such markers do not accurately reflect nuclear genetic diversity (Bazin et al. 2006; Leigh et al. 2021). For example, putatively neutral mitochondrial markers are linked without recombination to loci under strong selective constraints (Galtier et al. 2009). Mitochondrial diversity is therefore subject to selective sweeps and background selection as well as bottlenecks due to its small effective population size (Ne ), which is a quarter that of nuclear DNA (Ballard & Whitlock 2003; Birky et al. 1989). Mitochondrial diversity does not display a consistent relationship with population size, with strong variation across taxa that is not related to population size or life history characteristics (Bazin et al. 2006; James & Eyre-Walker 2020; Nabholz et al. 2009). The Resource Availability Hypothesis, therefore, may apply most strongly to nuclear diversity, and only weakly to mitochondrial diversity. However, the Founder Effect Hypothesis may apply more strongly to mitochondrial genetic diversity, as mitochondrial DNA should be more sensitive to bottleneck events and founder effects from the LGM due to its smaller Ne (Birky et al. 1989). With these caveats in mind, macro-scale patterns of mitochondrial genetic variation may not be generalizable across the genome. To gain a more complete understanding of global distributions of genetic diversity, neutral genetic variation should also be analyzed across the nuclear genome to understand forces acting across this much larger fraction of the genome.
To help close these knowledge gaps, we conducted a literature search to aggregate georeferenced data from population genetic studies in marine fish species and then used these data to evaluate the three hypotheses. We compiled environmental data on sea surface temperature (SST) and chlorophyll (a proxy for primary productivity) and assessed the generality of these hypotheses using both mitochondrial and nuclear (microsatellite) DNA. Specifically, we tested 1) the Kinetic Energy Hypothesis that temperature and genetic diversity will be positively related most strongly in mitochondrial DNA and more weakly in nuclear DNA, 2) the Resource Availability Hypothesis that primary productivity (e.g., chlorophyll) will be positively related with genetic diversity, particularly in nuclear DNA, and 3) the Founder Effect Hypothesis that both mitochondrial and nuclear genetic diversity will be negatively correlated with latitude and increase towards the equator, particularly in the Northern hemisphere and more strongly in mitochondrial diversity. To test among these three hypotheses, we fit generalized linear mixed effect models (GLMMs) and explored the extent to which each macroecological driver independently explained variation in mitochondrial or nuclear genetic diversity.