| Abstract |
- Accumulating human impacts on freshwater ecosystems have created a biodiversity crisis for freshwater fishes while conservation efficacy remains hampered by the Linnaean shortfall, the ignorance of many species that have yet to be discovered and described. The past discovery of most freshwater fish species was simply a matter of collecting in taxonomically blank or poorly explored places on the map and then describing whatever looked different. Increasingly, especially in well-sampled areas, undescribed species are likely to be morphologically cryptic and are only discovered after intensive research. Fortunately, biogeography has arranged biodiversity into patterns that, if revealed and understood, can be used to straighten the path to discovery.
The Cape Floristic Region (CFR) of the southwestern and southern coast of South Africa is famous for its 6,200 species of endemic plants but its coastal river systems contain only 19 primary freshwater fish species according to traditional (morphological) taxonomy. However, 16 of them are endemic and recent studies have revealed additional, cryptic species within them. Understanding how biogeographical processes have generated this endemic cryptic diversity could inform the search for species in other coastal regions around the world. This dissertation is focused on one primary freshwater fish, the Cape Kurper, Sandelia capensis (Cuvier 1829), a climbing gourami (Anabantidae) found in coastal river systems across the CFR. The results of previous studies suggest that it may also include undescribed cryptic species, and its ubiquity in the CFR makes it an excellent model species for studying the biogeographic processes that generate endemic species in coastal river systems.
After collecting a large number of specimens and tissue samples from across the geographic range of S. capensis, the first step was to obtain mitochondrial and nuclear DNA sequences and produce a molecular phylogeny and haplotype network. Based on these results I hypothesized that S. capensis sensu lato is actually comprised of three cryptic species: Sandelia sp. [West Coast], Sandelia sp. [Klein River], and Sandelia sp. [South Coast]. Using fossil calibrations I generated a dated phylogeny which provided divergence times for the three putative species and the various lineages within them. The divergence times and phylogeographic patterns were then compared to the history of geology, climate, and sea level to evaluate the relative significance of four biogeographic processes in driving the evolution and distribution of S. capensis lineages. The results indicate that the distribution of lineages is most consistent with a history of rare transdivide dispersal by river captures with a smaller role for dispersal via paleoriver confluences during low sea levels brought about by the Pleistocene glacial periods. Due to the narrow and steep continental shelf along the coast west of Cape Agulhas (the southern extreme of Africa) and north of the Cape of Good Hope, paleoriver dispersal was only likely to the east where the continental shelf is wide. However, S. capensis apparently did not use some paleoriver confluences, possibly because of intervening paleowaterfalls or fast currents. The divergence of the cryptic species, and the lineages within them, is mostly due to long isolation within coastal river systems. This isolation was maintained by the marine environment, the paucity of usable paleoriver confluences, and persistent drainage divides due to the relatively stable geology and dry climate which limit river capture events. There was little support for the influence of marine transgressions which could have isolated populations in tributaries by drowning the confluences between them.
The second phase of this research was an analysis of morphological variation based on linear morphometrics, meristics, and geometric morphometrics. The goal was to determine whether morphology supported the species hypotheses, and to discover diagnostic characters. While there was statistical and diagnostic support for all three species, the differences were very subtle. There was more variation within species than between them. The fact that 35–40 linear measurements, or 20 meristic counts, were needed in the discriminant functions indicated very strong morphological stasis. The relatively stable geomorphology and climate, the broad niche of S. capensis sensu lato, and the dearth of fish competitors and predators, point to niche conservatism as the most likely explanation for the cryptic nature of the three putative species.
Finally, in a revision of S. capensis sensu lato, I describe Sandelia sp. [Klein River] from the 980 km² Klein River basin, Sandelia sp. [West Coast] from the Langvlei, Verlorenvlei, Berg, and Diep river systems on the west coast, and Sandelia [South Coast] from drainages on the south coast other than the Klein River system. I also evaluate the conservation implications of this revision. Considering population losses already brought about by alien bass and trout, water extraction, and habitat degradation, I recommend prompt population assessments and strong protection, particularly for Sandelia sp. [Klein River] and Sandelia sp. [West Coast].
|
|---|
