Does deep-soil habitat influence the maintenance of ancient asexual lineages in acariform mites?

The acariform mites are a hyperdiverse chelicerate lineage (estimated 350,000 species) that includes an array of detritivores, fungivores, phytophages, predators and animal parasites. The Acariformes are among the most ancient groups of chelicerate arthropods. Putative acariform asexuals are known from the late Devonian (410 Mya), with members of Alicorhagiidae (Endeostigmata) bearing a close morphological resemblance to modern mites of this family (which are mostly asexual). Acariform mites, with bdelloid rotifers and darwinulid ostracods, have been used as model lineages in the study of the evolution and loss of sexual reproduction. Certain acariform taxa have been regarded as “scandalous”, a term coined by Maynard Smith in 1986 to describe apparently ancient lineages that reproduce exclusively by thelytokous parthenogenesis, contrary to the prevailing theory that suggested such lineages should be short-lived.

We use acariform mites, in particular Endeostigmata (a lineage regarded by specialists as a basal paraphyletic assemblage at the acariform root) as model organisms to test hypotheses about origin and maintenance of asexuality. We first infer an acariform phylogenetic tree (76 ingroup taxa, 5 nuclear genes, 11 kb aligned) and then test if the distribution of ancient asexual lineages (as defined by the paleontological record and phylogeny) is correlated with habitat. There are two basic habitats for endeostigmatan mites: deep soil and upper soil with related environments (eg., litter, moss). Deep soil harbors specialist asexual species (15.6%), while upper soil is inhabited by both sexual and asexual species (28.6% and 15.6%, respectively). There are also generalist species living in both habitats (33.3% of sexuals and 6.8% asexuals). Deep soil is a stable, impoverished environment that can support only very low population densities. Hence, asexuality may be more favorable there because there is no need to invest resources in producing males and maintain genetic diversity beside the most adapted genotype. In contrast, sexuality may be advantageous in more unpredictable and resource rich habitat, such as upper soil.

We test if phylogenetic and environmental distribution of asexual lineages can be explained by two hypotheses: (i) rates of diversification (speciation minus extinction) of asexual lineages are higher in deep soil, and (ii) slower at upper soil (unpredictable environment, competition with sexual species). As a consequence, a mix of basal and recent asexual lineages is expected in deep soil because of asymmetric diversification and migration between deep and upper soil.  

We test this model in the program GeoSSE, which can estimate habitat-dependent rates of speciation, extinction, and migration in a phylogenic context (using branching pattern and timing). Another parameter, the rate of between-habitat speciation, is available for species from both deep and upper soil. All these rates are estimated given the topology, and then the best model is compared with a model constrained for rates of speciation, extinction and migration to be lower in deep soil. As such, GeoSSE results provide an additional test for ‘ancient’ asexuality (i.e., if speciose asexual-only clades are indeed ancient, but not a result of asymmetrical diversification).

            Our study provides a unique window into how environment may affect evolution and long-term maintenance of asexuality in the oldest asexual metazoan lineage.