Discrete phenotypes are not underpinned by genome-wide genetic differentiation in the squat lobster Munida gregaria (Crustacea: Decapoda: Munididae): a multi-marker study covering the Patagonian shelf

Chen Wang, Shobhit Agrawal, Jürgen Laudien, Vreni Häussermann, Christoph Held

DNA barcoding has demonstrated that many discrete phenotypes are in fact genetically distinct (pseudo)cryptic species. Genetically identical, isogenic individuals, however, can also express similarly different phenotypes in response to a trigger condition, e.g. in the environment. This alternative explanation to cryptic speciation often remains untested because it requires considerable effort to reject the hypothesis that the observed underlying genetic homogeneity of the different phenotypes may be trivially caused by too slowly evolving molecular markers. The widespread squat lobster Munida gregaria comprises two discrete ecotypes, gregaria s. str. and subrugosa, which were long regarded as different species due to marked differences in morphological, ecological and behavioral traits. We studied the morphometry and genetics of M. gregaria s. l. and tested (1) whether the phenotypic differences remain stable after continental-scale sampling and inclusion of different life stages, (2) and whether each phenotype is underpinned by a specific genotype. A total number of 219 gregaria s. str. and subrugosa individuals from 25 stations encompassing almost entire range in South America were included in morphological and genetic analyses using nine unlinked hypervariable microsatellites and new COI sequences. Results from the PCA and using discriminant functions demonstrated that the morphology of the two forms remains discrete. The mitochondrial data showed a shallow, star-like haplotype network and complete overlap of genetic distances within and among ecotypes. Coalescent-based species delimitation methods, PTP and GMYC, coherently suggested that haplotypes of both ecotypes forms a single species. Although all microsatellite markers possess sufficient genetic variation, AMOVA, PCoA and Bayesian clustering approaches revealed no genetic clusters corresponding to ecotypes or geographic units across the entire South-American distribution. No evidence of isolation-by-distance could be detected for this species in South America. Despite their pronounced bimodal morphologies and different lifestyles, the gregaria s. str. and subrugosa ecotypes form a single, dimorphic species M. gregaria s. l.. Based on adequate geographic coverage and multiple independent polymorphic loci, there is no indication that each phenotype may have a unique genetic basis, leaving phenotypic plasticity or localized genomic islands of speciation as possible explanations.