Pante et al. (2014) performed a literature review of marine population connectivity in order to illustrate the biased estimates of connectivity which can result from the failure to recognize an evolutionary-relevant unit, such as a species.
When exploring the connectivity of a set of populations, it may be necessary to revise and reassess taxonomic status. This is particularly true in the marine environment, which is vastly under-sampled as compared to terrestrial habitats.
Poor species delimitation doesn’t just affect an individual connectivity study, but can affect meta-analyses and reviews investigating evolutionary and ecological trends. It can also affect studies of speciation, phylogenetic studies, invasion biology and biodiversity inventories.
The authors review relevant examples of over- and under-estimation of connectivity due to poor species delimitation. They also provide a primer on delimiting a species and treating them as scientific hypotheses.
But, it’s important to note that the results from careful connectivity studies can provide evidence about divergence between different lineages. However, in order to carefully explore connectivity, we need to keep in mind:
(1) the state of knowledge on the biology of the studied organisms, (2) the state of taxonomic treatments of the studied organisms, (3) the spatial and temporal scales of sampling, (4) the characters used to infer connectivity patterns and (5) how to synthetize information in multimarker studies
In other words, we need to take into account life history and ecological traits. If the above knowledge is limited or nonexistent, the authors propose incorporating this uncertainty into the sampling design. It could also be possible to include closely related taxa for groups in which the phylogeny is poorly understood, for example, deep-sea organisms.
The authors also stress the importance of including life-history traits and their spatio-temporal variability into the design of sampling effort, such as clonal reproduction.
Finally, they articulate the use of multiple and diverse markers, while pointing out the importance of moving away from the sole use of mitochondrial genes.
Pante E, Puillandre N, Viricel A, Arnaud-Haond S, Aurelle D, et al. (accepted) Species are hypotheses: avoid connectivity assessments based on pillars of sand. DOI: 10.1111/mec.13048