The lack of ND5 haplotype sharing between Manta alfredi and M.
birostris and the reciprocal monophyly in RAG1 support the
hypothesis that the speciation is complete. Our initial tests using
relatively slowly evolving gene CO1 failed to distinguish the two
species, but the employment of relatively fast-evolving gene ND5
(Billington, 2003) provided sufficient resolution. Our findings support
other claims that there can be difficulties with the use of a single
marker (CO1) to detect recently diverged species pairs
(Hickerson et al., 2006; Moritz and Cicero, 2004). Non-monophyly
in ND5 can be interpreted under two scenarios: IM (m > 0, t > 0)
and ILS (m = 0, t > 0). Although both scenarios are not necessarily
mutually exclusive, it appears that the IM scenario explains our
data better. First, if there was no gene flow, mtDNA is expected
to achieve reciprocal monophyly faster than nDNA. Our data
showed the opposite pattern. Second, the geographic distribution
of paraphyletic M. alfredi haplotypes (i.e. MA4 and MA5) is spatially