INTRODUCTION
Species is widely considered as the basic entity of biodiversity in nature, although different species concepts have been proposed and used (Zachos, 2016). The most popular biological species concept (BSC) uses reproductive isolation as the gold standard to designate species. In taxonomy, subspecies is used to denote recognizable infraspecific differentiation above populations. Conventionally, subspecies is defined as “a geographically defined aggregate of local populations which differ taxonomically from other subdivisions of the species” by E. Mayr (1940, 1963). In the context of BSC, subspecies can interbreed without a fitness penalty (Patten, 2015), although a test of this ability is usually not practically possible in the wild. The designation of subspecies has been criticized as arbitrary and rejected by some taxonomists (Wilson & Brown, 1953; Hawlitschek, Nagy, & Glaw, 2012; Phillimore & Owens, 2006; Torstrom, Pangle, & Swanson, 2014).) Others insist on the value of the subspecies rank (Durrant, 1955; Mayr, 1982; Phillimore & Owens, 2006), with the emphasis on unique geographic range or habitat, phylogenetically concordant phenotypic characters, and unique natural history (O’Brien & Mayr, 1991).
Given the taxa designated as subspecies, conventionally basing on morphological and geographical evidences, we would expect that the designation itself represents a level of genetic variation at somewhere between geographical populations and full species. It requires some examinations on the pattern of genetic variations, which may strengthen or weaken the subspecies designation. Many studies have tested whether defined subspecies are monophyletic on phylogenetic trees constructed using a handful of genetic markers (Moritz, 1994; Phillimore & Owens, 2006). However, it provides little assessment of the divergence level, because monophyly is also evidence of full species and the alternative is normally interpreted as a deficit of divergence. Instead, we propose that divergence among, given polymorphism within, subspecies should be assessed using population genetic analyses, since speciation proceeds at the population level.
The reduction in cost of sequencing allows us to quantify genetic divergence across the genome in large population samples. With such data in hand we can ask the following questions: What are the patterns of genomic variation among taxa a priori designated as subspecies? Do these patterns strengthen or weaken the subspecies designation? We start from investigating the taxa that can be reasonably designated as subspecies by conventional criteria. We perform such a study on the mangrove tree Avicennia marina . It is the most wide-ranging mangrove species, reaching the most marginal mangrove patches of the Indo-West Pacific region (Duke, 2006; Tomlinson, 2016). The taxonomy of Indo-West Pacific (IWP) Avicennia had been troublesome before Duke’s comprehensive revision (Duke, 1991). In that assessment, A. marina were divided into three varieties (Duke, 1991). After that division, “varieties” or “subspecies” were used to refer to the three groups by different authors (Duke, 2006; Duke, Benzie, Goodall, & Ballment, 1998; Maguire, Peakall, Saenger, & Maguire, 2002; Maguire, Saenger, Baverstock, & Henry, 2000).
The three subspecies show remarkable differences in morphological traits and geographical distribution (Duke, 1991, 2006). Details of these differences are described in the following section. A previous study used allozyme markers to determine genetic divergence among these subspecies, in which no fixed differentiation among subspecies was identified and the populations were genetically clustered into two groups (Duke et al., 1998). However, the reliability of that study is greatly compromised because they used very few genetic markers. We sought to obtain a comprehensive determination of genetic divergence among the three subspecies as well as reconstructing their evolutionary history through collecting single nucleotide polymorphisms across close to a hundred genomic loci.
The clarification of genetic divergence among subspecies may encourage us to treat these subspecies as different conservation units, particularly in projects such as transplanting and breeding. As the most widely distributed mangrove tree, this significance is valuable forA. marina . Mangrove species are all a conservation priority because these species are ecologically important in sheltering coastal regions from hurricanes, supporting the intertidal ecosystem, and sequestering carbon, but they are under great threat of global climate change in combination with more direct human disturbances (Gilman, Ellison, Duke, & Field, 2008; Guo et al., 2018a).