Eunectes, Wagler, 1830

Tarkhnishvili, David, Hille, Axel, Waller, Thomas, Todua, Mariam, Murtskhvaladze, Marine & Böhme, Wolfgang, 2022, Morphological trends and genetic divergence in anacondas, genus Eunectes Wagler, 1830 (Serpentes: Boidae), Amphibia-Reptilia 43 (4), pp. 379-393 : 388-390

Morphological vs genetic differentiation in Eunectes

The monophyletic genus Eunectes comprises two clades, large-bodied and small-bodied anacondas. According to different authors, the time of this separation varies between five ( Tonini et al., 2016) and 20 ( Wright et al., 2015) Mya, with a median value 10.8 Mya ( Zheng et al., 2016; Kumar et al., 2017). The first clade has a single species, the Green Anaconda ( E. murinus ), with a continuous distribution range covering a considerable part of South America; the second clade includes three taxa spatially separated from each other by distributional gaps. According to our estimate based on the analysis of mitochondrial Cyt-b gene, the diversification of small-bodied anacondas started ca. 2.4 Mya (fig. 5), i.e., it can be connected with the range fragmentation in the early Pleistocene. The Green Anaconda can easily be distinguished from its small-bodied congeners by higher numbers of head scales and lateral head stripes. These characters show an apparent hiatus between these lineages (table 1, fig. 2). The discrimination of the two clades is supported both by morphological and genetic evidence, and posterior probabilities for monophyly of the small-bodied anacondas vary between 0.67 (ODC) to 0.99-1.00 (Cyt-b and BNDF). However, the analysis of RAG-1 and NT3 genes did not show reciprocal monophyly between the two clades of Eunectes . This suggests either gene introgression between big-bodied and small-bodied anacondas in the geological past, or low rates of evolution of these two genes, which might hamper identification of the lineages ( Brito and Edwards, 2009; Freitas et al., 2020).

Neither of the analyzed datasets showed reciprocal monophyly or full discrimination between three nominal species of the smallbodied anacondas. Moreover, the initial split between the nominal species of small-bodied anacondas, judging from the mt-DNA sequence divergence, happened even later than the basal split between the current lineages of E. murinus (fig. 5a). Simultaneously, there are some diagnostic characters for the studied species. E.g. the number of dorsal blotches do not overlap between the studied E. notaeus and E. beniensis ; RAPD profiles fully separate E. notaeus and E. deschauenseei ; there is a diagnostic mitochondrial haplogroup of E. beniensis (but not in two other small-bodied species).

The incongruence of unlinked gene-based trees and paraphyly of individual species suggest either incomplete lineage sorting in sense Avise (1999), or gene introgression. Concerning small-bodied species, the former scenario looks reasonable, assuming their relatively recent divergence. Gene introgression can continue millions of years after the initial split of the lineages ( Kronforst, 2008), which could also cause limited gene flow between E. murinus with coexisting species of the small-bodied anacondas. In any event, neither of the small-bodied anaconda species reached the stage of genealogical concordance (Avise, 1999, 2000), although they show signs of both phenotypic and genotypic distinctness.

The question remains as to the consequences of divergence events in small anacondas. The analysis of morphology indicates a higher similarity between E. notaeus and E. deschauenseei than between these two and E. beniensis . This is in contrast with RAPD analysis which places Beni anaconda in the same cluster as E. notaeus . However, the mitochondrial topology is consistent with the morphological analysis and supports a more distant position of E. beniensis from two other small- bodied anaconda species than that of E. notaeus and E. deschauenseei . We should acknowledge here the limitation of RAPD approach, which helped to distinguish between the small – and large-bodied anacondas and confirmed probability of incomplete lineage sorting between the small-bodied species; however, the method can hardly be used for building a topology with a high resolution between the studied taxa. In future, genomic data can shed light on the evolution of small-bodied Eunectes , and provide a clear reply on the questions on the relative time of divergence, present or past gene flow between the individual species.

Evolution of Eunectes in biogeographic context

All anaconda species are dependent on seasonally flooded wetland areas ( Pizzatto et al., 2007). However, they are found in different landscapes and climates. The central part of the range of Green Anacondas coincides with the distribution of tropical rainforests, whereas the ranges of all three small-bodied anacondas lay in tropical moist and dry forests and seasonally flooded savannah and treeless grassland, which emerged in lowland Amazonia in the Middle Pliocene ( Burnham and Graham, 1999; Cheng et al., 2013).

Considering increasing long-lasting fluctuations of earth climate since the Middle Miocene ( Miller and Fairbanks, 1983), one can hypothesize that fragmentation of tropical rainforest paralleled by the increase of dry forest and grassland areas could have triggered the initial split between the big-bodied, more rainforestadapted, and the small-bodied species, which are adapted to more open biomes. The inferred time of the divergence within Eunectes different from that accepted here does not affect this suggestion. Both the highest ( Wright et al., 2015) and the lowest (Tonini et al., 2015) estimated divergence times are within the period between the middle and late Miocene, within the period of continuous decline of the land temperature (and hence continuous forest fragmentation).

The further split within big-bodied anacondas occurred, probably due to climate and landscape changes in early Pliocene. In small-bodied anaconda lineages, the divergence of the main evolutionary lineages occurred later, probably during the Pleistocene glacial-interglacial cycles. Kershaw et al. (2013) showed that the dispersal of E. notaeus is limited by precipitation seasonality and presence of sandy cambisol soil; the two other small-bodied anaconda species also are absent from the areas with high precipitation seasonality. Currently, large grassland areas in South America, more or less coinciding with the ranges of the small-bodied anacondas, are separated by rainforest massifs. During the last glacial period, the climate of the continent was drier ( Cheng et al., 2013; Häggi et al., 2017) and, probably, grasslands were less fragmented (see also Wüster et al., 2005). Rainforest repeatedly expanded during interglacials, separating optimal habitats of the small-bodied anacondas, and shrinked / fragmented again during the glacial waves. The split between E. notaeus and E. deschauenseei , which started ca. 1 Mya, coincides in time and space with the split between two lineages of the Neotropical rattlesnake Crotalus durissus which occurs in seasonally dry and open habitats, and can be attributed to “Transamazonian Vicariance” as defined by Wüster et al. (2005). However, earlier inferred time of the split between the diagnostic lineages of E. beniensis and E. notaeus + E. deschauenseei draws us to conclude that the split between the eastern and western fragments of South American grasslands and light forests happened earlier than the Transamazonian split.

Taxonomic suggestion

The studied gene phylogenies, including Cyt-b based phylogeny, do not support reciprocal monophyly of allopatric species of small-bodied anacondas. This suggests that the small-bodied anacondas did not reach the stage of genealogical concordance in the sense of Avise (1990), and their differences are probably maintained due to long-lasting geographic isolation rather than pre- or postzygotic isolation mechanisms.

Therefore, our study reveals (1) reciprocal monophyly of E. murinus and the smallbodied anacondas within the genus Eunectes ; (2) incomplete lineage sorting between the small-bodied anaconda species; (3) probably, closer relationships between E. notaeus and E. deschauenseei than between each of them and E. beniensis ; and (4) a distinct phenotypic position of the Peruvian E. murinus population. This separation, however, is inconsistent with the traditional distinction of the two E. murinus subspecies E. m. murinus and E. m. gigas ( Stimson, 1969) and supports the synonymy of both taxa (see Dirksen and Böhme, 1998b).

In conclusion, we suggest that the genus Eunectes has four phenotypically distinct species, which did not reach the stage of complete lineage sorting. The genetic differences between small-bodied anacondas are not fixed. However, they are highly significant, and currently their ranges are separated by less suitable tropical forest landscapes and limited dispersal capabilities due to the relative autonomy of the different wetland and riparian systems (cf. McCartney-Melstad, 2012). If Mayr’s (1969) polytypic species concept is considered, the three small anacondas might be attributed to the same polytypic species complex. A new taxonomic committee of SEH does not suggest a consolidated view on species definition ( Speybroeck et al., 2020); however they agree that (1) all species are evolutionary lineages, (2) some gene flow between them is possible, if they remain distinct geographically. We conclude that E. notaeus , E. deschauenseei , and E. beniensis are evolutionary species in the sense of Wiley (1978). De Queiroz (2007) defined species as the lineages/ groups of individuals characterized by a unique evolutionary pathway. The divergence of the smallbodied anacondas developed at least several glacial cycles before present; this means that they might have come into contact repeatedly during the glacial climatic oscillations, which, however, did not cause secondary panmixis, and thus phylogenetic independence of these evolutionary lineages appears to be sustainable.