Phylogeny and biogeography of Southern African limpets in the genus Siphonaria in the context of a global phylogeny

Doctoral Thesis


Permanent link to this Item
Journal Title
Link to Journal
Journal ISSN
Volume Title
Abundant, diverse and ecologically important, the 'false-limpet' genus Siphonaria has a near world-wide distribution. However, arising from uniformity of shell shape among species and its plasticity within species, frequent synonymisations and changes of names have created uncertainty about the number of species that exist and the names that should be applied. In this thesis I pursue three lines of genetic research to address this. First, I resolve the species that exist in two South African complexes: the Siphonaria carbo complex and the S. concinna/oculus complex. Second, I examine what is currently accepted as S. compressa from the only two South African localities where it has been recorded, to determine if the two populations are conspecific. Finally, I compare the genetics of a selection of species from around the world to place the South African species within, and contribute to, a world phylogeny for the genus and analyse geographic patterns of clades emerging from this phylogenetic reconstruction. With respect to the 'S. carbo' complex, Teske et al. (2007), using mitochondrial and nuclear sequence data on Siphonaria nigerrima, S. anneae, and S. tenuicostulata, failed to discern distinct monophyletic lineages, and concluded that these species, together with S. dayi, are different colour morphs of a single species, merged under the name S. nigerrima. This contrasts with a previous systematic revision by Chambers & McQuaid (1994), which considered these to be four separate species. An added complication is that the name S. nigerrima has been synonymised with S. carbo by various authors and in a range of data bases. Using phylogenetic analyses based on COI and 16S ribosomal RNA genes and incorporating additional GenBank sequences of east coast species, I confirm that S. nigerrima and S. dayi comprise a single species, to be merged under the name S. nigerrima because of its priority, and because examination of the type material of S. carbo indicates that it is unlikely that that species is present in South Africa. However, I also show that within the 'S. nigerrima complex' there exist four or possibly even five clades that warrant recognition at a specific level, two of which constitute undescribed and unnamed species (nov. sp. 1 and nov. sp. 2) while the other three can be accommodated under the names S. nigerrima sensu stricto, S. tenuicostulata and S. anneae. Using the same approach, I examined the distinction between S. oculus and S. concinna, which has become blurred in the literature, with only the latter being recognised by Teske et al. (2011). Both my molecular analyses and a consideration of shell features provide evidence that the two are distinguishable and valid species. Siphonaria compressa was first described by Allanson (1958) from specimens found in Langebaan Lagoon on the West Coast of South Africa, where it is associated with the seagrass Zostera capensis. In 2005 a population of what was considered the same species was recorded in Knysna Lagoon on the South Coast. On the basis of COI and 16S analyses coupled with shell morphometrics I found that the two populations have no shared haplotypes and are clearly diverged; IMa2 estimates indicate there has been no migration between these populations and the split between them occurred around 730 000 years ago. Evidence points towards recently expanding populations at both sites. Uni-, bi- and multivariate analyses of 13 morphometric variables revealed that the means for most variables are significantly different and clear differences in shell shape, size and weight exist between the two populations. I thus recognise two geographically delimited species, and in this thesis I distinguish the Knysna population as Siphonaria nov. sp. 3. The fact that the populations are considered separate species has important ramifications for conservation. The third aspect of my thesis was to integrate the southern African species that I now recognised into the most recent global phylogeny of the genus (Dayrat et al. 2014), which focussed strongly on the Indo-West Pacific species and omitted all South African species. I sequenced 120 individuals collected from 51 localities including 11 southern African species - three of which I consider to be undescribed species - and 42 other species from around the world, nine of which were not previously included in any published phylogeny, thus adding a total of 20 species to the global phylogeny. In total, incorporating GenBank sequences, I analysed 272 mtDNA sequences (16S + COI, 1202 nucleotide characters) within a Bayesian framework. Sacoglossans were used as the outgroup as per Bouchet et al. (2017). Similar to Dayrat et al. (2014), two major clades, A and B, were recovered. The South African species were not monophyletic but their positions were clear and distinct and did not influence the original division of the genus into two major clades. The inclusion of the South African species led to a rearrangement of the tree topology, especially within clade A, where all bar one of the South African species were recovered, with the the highly divergent S. serrata occurring in the second major clade B. Within clade A all bar two of the South African species were retrieved in one clade. The exceptions were S. compressa and Siphonaria nov. sp. 3 which fell elsewhere within clade A and were more closely related to Pacific Ocean species. The contents of clade B remained similar to the previously published phylogeny with the exception of a significant rearrangement of the basal species. Although many species remain unnamed, six previously unidentified molecular units in the global phylogeny are identified here and assigned species names. I also explored biogeographic patterns for all recognised species in the genus. There were strong latitudinal trends evident in species richness with high diversity in the lower latitudes, but with a dip at the equator, in contradiction to the prevailing paradigm of greatest diversity at the equator. Species with direct rather than planktonic development had both smaller geographic and smaller average latitudinal ranges. Overall, this thesis encompassed a view of the genus from the small-scale withinspecies population genetics of the highly endangered and range-restricted S. compressa and its sister taxon Siphonaria nov. sp. 3, through the broader geographic delimiting of the species contained within two southern African species complexes, to the placement of all the southern African species within a global context by updating the world phylogeny of the genus.