Molecular evidence for sequential colonization and taxon cycling in freshwater decapod shrimps on a Caribbean island.

Taxon cycling, i.e. sequential phases of expansions and contractions in species' distributions associated with ecological or morphological shifts, are postulated to characterize dynamic biogeographic histories in various island faunas. The Caribbean freshwater shrimp assemblage is mostly widespread and sympatric throughout the region, although one species (Atyidae: Atya lanipes) is geographically restricted and ecologically and morphologically differentiated from other Atya species. Using patterns of nucleotide variation at the COI mtDNA gene in five species of freshwater shrimp (A. lanipes, A. scabra, A. innocuous; Xiphocarididae: Xiphocaris elongata; Palaemonidae: Macrobrachium faustinum) from Puerto Rico, we expected to detect a signature of sequential colonization in these shrimp, consistent with the concept of taxon cycling, and expected that A. lanipes would be at a different taxon stage (i.e. an early stage species) to all other species. We also examined patterns of genetic population structure in each species expected with poor, intermediate and well-developed abilities for among-river dispersal. Population expansions were detected in all species, although the relative timing of the expansions varied among them. Assuming that population expansions followed colonization of Puerto Rico by freshwater shrimp, results bear the hallmarks of sequential colonization and taxon cycling in this fauna. A. lanipes had a star phylogeny, low mean pairwise nucleotide differences and recent (Holocene) estimates for an in situ population expansion in Puerto Rico, and it was inferred as an early stage species in the taxon cycle undergoing a secondary phase of expansion. All other species were inferred as late stage species undergoing regional population expansions, as their mean pairwise nucleotide differences were relatively high and phylogenetic patterns were more complex than A. lanipes. High rates of gene flow without isolation by distance among rivers were detected in all species, although results should be treated cautiously as some populations are unlikely to be in mutation-drift equilibrium. Nested clade analysis produced inconsistent results among species that all have high rates of gene flow and expanding populations.