Austrofundulus Myers

Tomas Hrbek, Donald C. Taphorn & Jamie E. Thomerson, 2005, Molecular phylogeny of Austrofundulus Myers (Cyprinodontiformes: Rivulidae), with revision of the genus and the description of four new species., Zootaxa 825, pp. 1-39 : 2-4

[[ Genus Austrofundulus Myers ]]

Austrofundulus was last revised by Taphorn and Thomerson (1978). In that study Taphorn and Thomerson (1978) recognized only two species: Austrofundulus transilis and A limnaeus , and placed the other two then described species, A. myersi and A. stagnalis into synonymy with A. limnaeus . Austrofundulus transilis was at that time only known from the Río Apure basin of Venezuela, while A. limnaeus had a very wide and disjunctive distribution.

The type species, Austrofundulus transilis Myers 1932 , is known from the Venezuelan Llanos north of the Orinoco mainstream and from the lower Río Unare Basin (Thomerson et al. 1990). Taphorn and Thomerson (1978) recognized seven distinctive populations of A. limnaeus Schultz, 1949 : the Colombian population found on the coastal lowlands between Cartagena and Sincelejo previously described as A. myersi Dahl, 1958 ; a population from the Guajira Peninsula; three populations from the Lake Maracaibo basin and the adjacent coastal desert including a population from the southeastern Maracaibo basin bearing the name A. stagnalis ; a population from the coastal Caribbean drainages of Río Aroa and probably also Río Tocuyo in the vicinity of Tucacas, Falcón State, Venezuela; and a population from the upper Tacutu (Branco-Amazon) River drainage in the Rupununi Savannah of Guyana (Fig. 1). Based on the apparent availability of suitable habitat shown on topographic maps, we suspect that the Guyanese population likely extends further south to the savannahs around Boa Vista, Brazil, however, currently no specimens are known from this area.

Taphorn and Thomerson had initiated the 1978 study thinking that several of these populations of A. limnaeus , distinguishable by male color patterns, might prove to be valid species. However, because of considerable morphological variation among males of A. limnaeus within populations, few significant differences in meristic and morphometric characters, and great similarity among females of all Austrofundulus populations, Taphorn and Thomerson (1978) did not give formal taxonomic recognition to any of these populations. Subsequent popular articles (Thomerson and Taphorn 1992a; b) included updated distribution maps, photographs of fish from various populations to document differences in male color patterns, and comments on natural history and aquarium culture, but they also suggested no further taxonomic changes.

In recent years molecular analyses have been applied as an additional tool in identifying species, and in inferring phylogenetic relationships among species (e.g. Avise 1994; Avise 2000; Templeton 2001). Mitochondrial DNA (mtDNA) in particular has found a widespread use as a molecular marker. Although mtDNA is non-recombining and is maternally inherited, and thus it cannot reflect the tokogenetic relationships among individuals within a sexual species, it is useful for inferring phylogenetic relationships among species, as well as in identifying clusters of individuals that are significantly differentiated from other individuals. This naturally requires the caveat that the evolutionary history of the mtDNA is the same as the evolutionary history of the individuals bearing them. This caveat applies to any character, whether molecular or morphological. Allowing for these assumptions, we proceed to investigate phylogenetic relationships among populations of the annual killifish genus Austrofundulus Myers 1932 , known from Colombia, Venezuela and Guyana.

Molecular phylogenetic analysis allowed us to reassess former conclusions regarding these different populations of Austrofundulus . Assuming that the evolutionary history of mtDNA reflects the evolutionary history of the different populations of Austrofundulus , reconstructing phylogenetic relationships among mtDNA haplotypes originating from different areas should amount to reconstructing phylogenetic relationships among these areas. Phylogenetic analysis of mtDNA sequence data suggests that all A. limnaeus populations recognized by Taphorn and Thomerson (1978), except “ A. stagnalis ”, are monophyletic. Thus geographic areas are inhabited by clades of individuals (Figs. 1, 2, 3). Austrofundulus transilis is also monophyletic (Figs. 2, 3). However, A. transilis is sister to the A. limnaeus population from the Rupununi savannah, and together they are sister to A. limnaeus populations from the Río Aroa basin. Austrofundulus transilis is nested within A. limnaeus as defined Taphorn and Thomerson (1978) making A. limnaeus a paraphyletic entity. It is unlikely this paraphyly would result from a process such as incomplete lineage sorting. In this case we would not expect monophyly of sampled geographical areas, or male color pattern differences diagnostic for these same geographical areas. Furthermore, we would not expect to observe spatial and temporal concordance of phylogenetic relationships with the geological history of northern South America. Combined evidence suggests that these geographically restricted, monophyletic entities represent species.

Based on the combined phylogenetic, geographic-distributional, male color pattern, and hybridization (see discussion), we propose a revision of the genus Austrofundulus . We propose to remove A. myersi from synonymy with A. limnaeus , restrict A. limnaeus to populations occurring on the eastern side of Lake Maracaibo but retain within A. limnaeus “ A. stagnalis ” from the southeastern side of Lake Maracaibo, and to describe the populations from the Guajira peninsula, from the western side of Lake Maracaibo, from the Tucacas region, and from the Rupununi Savannah as new species.

Costa (1990) synonymized Austrofundulus Myers 1932 with Rachovia Myers 1927 , but in a later publication (Costa 1998) resurrected Austrofundulus as a separate genus without justification. We follow Taphorn and Thomerson (1978) in recognizing both genera. Although there are no universally accepted definitions as to what constitutes a genus, it is generally agreed that generic designations should encompass not only monophyletic units, but also a morphologically and ecologically distinct groups, thus conveying additional information above and beyond the species level. Since this analysis shows Austrofundulus species to be a monophyletic group, and Austrofundulus and Rachovia form morphologically distinct units (Taphorn and Thomerson 1978), the retention of Austrofundulus as a separate genus is justified.