michi tobler wrote:So, I think we should be skeptical if species are delineated simply based on morphological data
This is certainly true, every species diagnosis is based only on a random sample out of the entire population/species (BTW this applies also to phylogenetic analyses based on morphological as well as molecular characters). It is probably an overemphasis on the 'diagnosability' component of the Phylogenetic Species Concept, which is used as a justification for establishing species based on smallest phenotypic variation. Although the diagnoses of the Australoheros
species described by Ottoni & Costa are very long (due to the many, many species names
), some of the species are distinguished only by single or a few characters, mostly counts (e.g. anal or dorsal soft rays, pterygiophores, rib pairs etc.) known to cover broader ranges in other (also congeneric) species. It needs only to discover slightly more variation within a population of a certain river, and the delimitation of several species will collapse. However, if several independent (and of course suitable) characters are used to delimit species, there is an increasing probability that they prove to be valid. In some instances it is certainly useful to have also large sample sizes to asses if morphometric characters or counts are statistically significant, but in others, such as some species from the Great African Lakes, which differ only in male nuptial coloration, it would be nonsense to erase half the population just to have a large list of examined material.
Of course, every population which is physically isolated from others, does in fact start to go its own evolutionary way. However, as evolution is a continuous process rather than a series of steps we have at least theoretically an infinite number of intermediate states between populations, species etc., so that it is, from an evolutionary point of view, impossible to say were a species 'begins'. All taxonomic categories are in fact artificial and may be not absolutely necessary for understanding evolutionary processes. Nevertheless, the evolution is slow enough to recognize something like species. Also I think we need species for manyfold practical purposes, including such concerning conservation, zoogeography, ecology, ethology etc.. This practical need requires in my view that a species must be something tangible, observable, hence it is useless to have species which are not phenotyoically distinct. From this point of view it would perhaps be the best to adopt something like a 'prudent specieces concept', I think there will never be a technical guidance allowing us to say with certainty "This is a species and this is not".
Mark Smith wrote:On the other hand, Nandopsis tetracanthus, thus far known, has 7 distinct morphological variants, and possibly more once more of Cuba is explored. All 7 variants of N. tetracanthusreside within a 100 mile radius of Havana. What might be in the middle and eastern portions of the island? At least one morphological variant will likely be seen as a distinct species, in my opinion.
Well, Mark, these variants of Nandopsis
are not as distinct as the drawings imply. None of them is well diagnosed. Heros nigricans
, which was regarded as the most distinct form, is based on a single specimen, which may well be somewhat malformed for it shows also some squamation anomalies. Furthermore, the jaw morphology is not very different from other N. tetracanthus
, Eigenmann's drawing is inaccurate in this respect. Compare the photo and x-ray of the holotype: http://research.calacademy.org/redirect ... /index.asp
Even Eigenmann (1903) himself expressed some reservation about the validity of the subspecies (in his 1910 work, he raised all to species level without giving reasons for doing so), and Hubbs (1920) gave further arguments for synonymizing all of Eigenmann's taxa. Here are links to Eigenmann's and Hubbs' works:http://fisherybulletin.nmfs.noaa.gov/22-1/eigenmann.pdfhttp://deepblue.lib.umich.edu/handle/2027.42/56529
I won't deny that N. tetracanthus
as currently understood may comprise several species, but these do not necessary correspond to the named subspecies, most of which do not even reflect geographic variation. Furthermore, it is not clear what exactly would be the true N. tetracanthus
, as the original description was based only on a drawing and there is no exact type locality known. Likewise, the identity and status of the other two synonyms (Chromis fusco-maculata
Guichenot and Acara adspersa
Günther), both of which have precedence over Eigenmann's taxa, has to be resolved before applying the available names to taxa if there should be more than one (disregarding the clearly distinct N. ramsdeni
). Hence, it would be premature to recognize any of Eigenmann's taxa as valid, and be it only at the subspecies level.