Varanus (Euprepiosaurus)

Ziegler, Thomas, Schmitz, Andreas, Koch, André & Böhme, Wolfgang, 2007, A review of the subgenus Euprepiosaurus of Varanus (Squamata: Varanidae): morphological and molecular phylogeny, distribution and zoogeography, with an identification key for the members of the V. indicus and the V. prasinus species groups, Zootaxa 1472 (1), pp. 1-28 : 19-21

Zoogeography of Euprepiosaurus

Members of Euprepiosaurus have an Australo-Papuan distribution mainly restricted to the Sahul continental shelf east of Lydekker’s line ( Figs. 1–3 ). This applies particularly to the members of the V. prasinus species group. However, several species of the V. indicus species group ( V. caerulivirens , V. cerambonensis , V. indicus , V. melinus , V. yuwonoi , V. zugorum , and Varanus sp. n.) have crossed this zoogeographic boundary of strictly Australian fauna in the past. Only two taxa succeeded in proceeding further west across Weber’s line to the Sula Islands ( V. melinus ) as well as to the Talaud Islands north of Sulawesi and Timor ( V. indicus ), respec- tively. As some of the above mentioned members of the V. indicus species group are endemic to the area between Wallace’s and Lydekker’s lines ( Figs. 1–2 ), better known as Wallacea, the importance of this biogeographic region for speciation processes becomes obvious. Until recently it remained unclear whether Euprepiosaurus had reached Sulawesi ( Iskandar & Tjan 1996), although a few voucher specimens exist ( Boulenger 1885, Böhme et al. 1994, Koch & Böhme 2005). Recent field studies on Sulawesi, however, have revealed that these locality data are probably incorrect (A. Koch pers. observ.) or due to single introduced specimens. The occurrence of V. salvator occupying a similar ecological niche on Sulawesi and in Southeast Asia west of Wallace’s line, may have restrained or inhibited the colonization of islands further west by Euprepiosaurus . Moreover, fossil evidence so far does not support the hypothetical replacement of a Sulawesian member of Euprepiosaurus by the larger V. salvator arriving at a later date on the island (Molnar 2004). Interestingly, fossil bones similar to those of V. indicus were found in New Caledonian deposits probably dating only about 1,750 years back ( Balouet 1991), thus suggesting a possible distribution of Euprepiosaurus further southeast into the Pacific region in the very recent geological past. Obviously, the adaptive radiation of Euprepiosaurus in the Australo-Papuan region correlates with the absence of large placental predators east of Wallace’s line ( Hooijer 1975, Diamond 1987, Sweet & Pianka in press). Sweet & Pianka (in press), however, speculate that predation rather than resource competition between varanids and placental carnivores was responsible for the observed pattern.

There exists a striking difference in the extension of the respective ranges of single species within Euprepiosaurus . While V. indicus inhabits the second largest area of all varanid species, several other species, particularly those of the V. prasinus species group like V. beccarii or V. macraei , are locally restricted island endemics. However, the widespread populations of V. indicus on several small Pacific island groups such as the Carolines and Marshall Islands probably refer to anthropogenic transportation ( Dryden 1965, Uchida 1967, Rodda et al. 1991). Similar to several Mediterranean island lizards ( Mertens 1934), a replacement may have occurred of plesiomorphic and stenoecious forms of Euprepiosaurus (e.g., V. finschi ) on mainland New Guinea to marginal islands (e.g., the Aru and Kai Islands) by more advanced and euryoecious species such as V. indicus (Philipp 1999, Philipp et al. accepted). This replacement hypothesis for the New Guinean area, as already proposed by Jacobs (2002) for the species pair V. kordensis vs. V. prasinus , requires confirmation by further ecological and faunistic studies. Due to distinct habitat preferences, V. indicus lives in broad sympatry in New Guinea with the closely related V.jobiensis and V. doreanus (Philipp 1999) . Possibly a fourth congener of the V. indicus species group, V. cerambonensis , has reached the western part of New Guinea. Accordingly, members of both the V. indicus and the V. prasinus species groups inhabit largely overlapping areas within the whole range of Euprepiosaurus resulting in multiple sympatric occurrences. Different ecological adaptations (semi-aquatic to terrestrial versus arboreal habits) between these distinct monophyletic lineages have influenced their distribution patterns.

The origin of Euprepiosaurus may be best located in the northern parts of its range (the Moluccas) where the highest diversity and endemicity is found. The recent sister taxon of Euprepiosaurus , V. olivaceus from the Philippines in the northwest of the Indo-Australian Archipelago ( Ast 2001, see also Fuller et al. 1998), adds support to this proposition. The findings of How & Kitchener (1997) with respect to the importance of the northern and southern Moluccas for speciation in snakes adds further weight to this hypothesis. They even suggested that these islands should be seen as “unique biogeographic subregions with (…) a relatively high degree of endemism and areas for incipient speciation”. In addition to the Moluccas in general, particularly Halmahera remained geographically isolated throughout the Pleistocene ( Voris 2000). New Guinea and the Solomon Islands (unpubl. data) are identified as less important centres of diversity and endemism rates in Euprepiosaurus .

Due to the paucity of the fossil record (Molnar 2004), it is currently not feasible to satisfactorily date the split-up of the two monophyletic lineages of Euprepiosaurus , i.e. the V. indicus and the V. prasinus species groups. Likewise, it is impossible to date single or multiple migration events (e.g., between New Guinea and North Australia), and to determine the time since the divergence of several taxa within the V. indicus and the V. prasinus species groups. Fitch et al. (2006), however, calculated the splitting event occurring during a fourth period of increased speciation rate for Australo-Papuan monitor lizards. At the same time, they hypothesized an arboreal ancestor of Euprepiosaurus undergoing an ecological radiation as a progenitor to the semi-aquatic to semi-arboreal V. indicus species group and the arboreal V. prasinus species group respectively. These assumptions correspond to our own interpretation of morphological character states and habitat adaptations for Euprepiosaurus (see above). In addition, Schulte et al. (2003) using molecular dating techniques re-analyzed the data set of Ast (2001) estimating the split-up between the Philippine V. olivaceus and Euprepiosaurus to have occurred about 112 Million years ago. Hence, they postulated the invasion of the ancestors of Australo-Papuan monitor lizards including those of Euprepiosaurus into the Australian-New Guinean region as early as the Late Mesozoic.

Our molecular studies revealed V. indicus as a derived member of Euprepiosaurus turns out to be paraphyletic. It seems reasonable to suggest that the ancestor of V. indicus evolved somewhere around New Guinea and subsequently colonized areas to the north ( Moluccas) and to the south ( Solomon Islands). The importance of New Guinea as a major source area for the herpetofauna of the Southwest Pacific region was summarized by Allison (1996). Thereafter, these newly established island populations were subjected to independent developments and allopatric speciation, possibly closely related to the complex geological history of the Indo-Australian region ( Hall 2001, 2002).

Molnar & Pianka (2004) suggested that members of Euprepiosaurus have dispersed from New Guinea to northern Australia in the recent past. This recent southward migration would explain why the north Australian populations of V. indicus , V. doreanus and V. finschi are conspecific with those found in New Guinea ( Ziegler et al. 2001). However, the ancestor of V. keithhornei must have reached the Australian continent prior to this latter colonization event because this lineage independently evolved into unique morphological ( Tab. 6, 7) and genetical features thus justifying the species status. The same might have applied to the remaining members of the V. prasinus species group inhabiting several satellite islands around New Guinea on the Sahul continental shelf ( Fig. 3 ).

Within the last years, several studies about the phylogeography of freshwater fishes ( McGuigan et al. 2000), pythons ( Rawlings & Donnellan 2003, Rawlings et al. 2004), rodents ( Godthelp 2001), and king brown snakes ( Kuch et al. 2005) revealed a complex pattern of faunal exchanges between New Guinea and Australia during repeated land bridge connections in times of glacially lowered sea levels of up to 120 meters ( Fairbanks 1989). Voris (2000) clearly illustrated that New Guinea apparently still remained largely connected with Australia when sea levels were as little as 10 meters below present-day levels. He assumed this situation for more than half the period since the last glacial maximum about 17,000 years ago. Due to their strict arboreality ( Greene 1986, 2004), V. prasinus and its close relatives are poor dispersers who were probably not able to cross distances of shallow waters and small islands between New Guinea and adjacent areas when global sea levels decreased in the past. Successful migration across the Torres Strait land bridge would have required a continuous forested connection between New Guinea and continental Australia. Growth of rainforest, however, was substantially inhibited by drier climate and cooler temperatures during periods of continental glaciations ( Heaney 1991, Hope 1996, Voris 2000). Another major factor might have been the high salinity of the emerged land. In contrast, coastal areas without tropical forest and dense vegetation in times of emerged land in nowadays Torres Strait separating Australia and New Guinea were no barrier for the migration of semiaquatic V. indicus and its predominantly terrestrial allies. Even the narrow but deep water-gap between New Guinea and New Britain still persisting when sea levels fell about 120 meters ( Hope 1996, Hall 2002), could not stop at least two members of the V. indicus species group, viz. V.indicus and V.finschi , to colonize islands northeast of New Guinea. Advantageous sea currents may have supported these dispersal events.

We think the isolation period of the allopatric Euprepiosaurus populations was not sufficient for speciation processes. A replacement hypothesis seems to us more reasonable. In fact, the evolution and adaptive radiation of the V. indicus and V. prasinus species groups may have been greatly influenced by both replacement and geographic isolation. Nevertheless, speciation processes in members of the subgenus Euprepiosaurus still remain highly complex and little understood.

Figures

FIGURE 1. Distribution of the members of the V.indicus species group in relation to significant biogeographic discontinuities within the Indo-Australian Archipelago (Wa — Wallace’s line, We — Weber’s line, Ly — Lydekker’s line): Varanus indicus (1), V. doreanus (2), V.jobiensis (3), V.finschi (4), V. melinus (5), V. cerambonensis (6), V. juxtindicus (7), V. yuwonoi, V. caerulivirens, V. zugorum, and Varanus sp. n. (8). Note, the Pacific island populations of V. indicus are not included in the map for a better resolution and due to their possible anthropogenic origin; dotted lines indicate presumed distributions.

FIGURE 2. Detailed distribution maps of the members of the V. indicus species group: A and B — Australo-Papuan region, C — Moluccas, D — Solomon Islands; question marks and dotted lines indicate presumed distributions.

FIGURE 3. Distribution of the members of the V. prasinus species group in relation to significant biogeographic discontinuities within the Indo-Australian Archipelago (We — Weber’s line, Ly — Lydekker’s line): V. prasinus (1), V. beccarii (2), V. kordensis (3), V. bogerti (4), V. keithhornei (5), V. telenesetes (6), V.macraei (7), V.boehmei (8), V.reisingeri (9).