A strong comprehensive framework to identify the underlying evolutionary processes operating on contemporary species dynamics is emerging (Harrison and Grace 2007; Ricklefs 2007). Many recent theoretical and empirical studies have highlighted the role of historical biogeography in explaining current geographic patterns of biodiversity (Harrison and Grace 2007; Ricklefs 2007; Wiens 2007; Leibold et al. 2010). Tropical forests with their stunning diversity have especially attracted the attention of many researchers (Schneider et al. 1999; Moritz et al. 2000; Haffer 2008; Vences et al. 2009; Hoorn et al. 2010). In this regard, much needs to be done with respect to understanding species dynamics and biogeography of the Western Ghats (WG) of tropical South Asia, which has been identified as a global biodiversity hotspot because it harbors high levels of biodiversity and endemicity (Myers et al. 2000). WG of peninsular India (PI) have also experienced complex geological history (Briggs 2003). Here, we present a brief review of the geological and climatic history of the WG and then propose three biogeographic scenarios for the wet evergreen species of the WG with explicit predictions.
The WG is a chain of mountains running along the west coast of PI for over 1600 km (8° to 21°N) with one major low-elevation break called the Palghat Gap (PG) (Fig. 1) (Subramanyam and Nayar 1974; Ali and Ripley 1987). In addition to high endemicity, WG exhibits substantial heterogeneity in vegetation and landscape types, as well as a distinct north–south gradient in seasonality and rainfall (Pascal 1988). Based on plant species composition, WG has been divided into four zones, namely (1) northern WG (NWG), (2) central WG (CWG), (3) Nilgiri or Blue Mountains, and (4) southern WG (SWG) (Fig. 1) (Subramanyam and Nayar 1974).
Interestingly, species richness and endemicity are not distributed uniformly across the WG. Studies on plants from across the WG have revealed that the southern parts of the WG are extremely diverse and have high endemicity compared with the central and NWG (Pascal 1988; Gimaret-Carpentier et al. 2003; Pascal et al. 2004; Davidar et al. 2007). Additionally, birds, amphibians, and fishes also exhibit similar diversity and endemicity patterns, that is, the SWG being more diverse with a higher proportion of endemics than the central and NWG (Dahanukar et al. 2004; Aravind and Gururaja 2010). Most of these studies have identified contemporary ecological factors like seasonality, productivity, climate, and short dry period length as the processes governing the aforementioned patterns. However, speciation and biogeographic processes are poorly understood for the WG biota.
As the WG is embedded in PI, its geological and climatic history is tied to that of PI. PI was dominated by warm tropical climate during the Late Cretaceous period (Singh et al. 2006; Samant and Mohabey 2009). Toward the end of the Cretaceous (around 70–65 mya), PI experienced extensive volcanic activity that resulted in the formation of the Deccan traps (Singh et al. 2006; Samant and Mohabey 2009). The Deccan traps are one of the largest continental flood basalts in the world and are confined to the northern parts of PI. This volcanic activity that spanned between 0.5 and 5 million years (my) has been speculated to have triggered mass extinctions in PI and elsewhere (Samant and Mohabey 2009). Recent pollen analyses from western and north-western parts of PI indicate the reestablishment of wet evergreen forests in the Early Paleocene after the extensive volcanic activity (Prasad et al. 2009). Thus, SWG may have served as refugia for wet evergreen plant species that were once widely distributed during the prevolcanic period. Another study on diversity and distribution of wet evergreen plants also discussed the plausibility of southern parts of the WG being a refuge for the wet evergreen plant species (Gimaret-Carpentier et al. 2003). Today, wet evergreen forests are one of the most dominant forests types in the WG and are confined predominantly to the western escarpment of this mountain range (Pascal 1988).
Biogeographical scenarios for WG
Given that SWG might have served as refugia for wet evergreen forest species during periods of volcanism, the advent of suitable conditions in the post-volcanic period could lead these species to have dispersed from SWG to central and NWG. This scenario (S1) can be tested in a historical biogeographical framework that uses molecular phylogenies in conjunction with molecular dating and event-based biogeographical analysis. If this scenario was true, then in the phylogeny of a group of WG endemics, the CWG and NWG species would be nested within a clade constituting SWG species. Additionally, the ancestral areas of the deeper nodes in the phylogeny should be in SWG and dispersals from SWG into CWG and NWG should fall in the post-volcanic period. Recent molecular phylogenetic studies on the amphibians and centipedes of the WG reported ancient lineages in the SWG (Biju and Bossuyt 2003; Bossuyt et al. 2004; Roelants et al. 2007; Joshi and Karanth 2011a), but these studies did not speculate or test the SWG refugia hypothesis, which is being presented here for the first time.
Alternately, physiographic or ecological barriers might have played an important role in the current distribution of species. As mentioned earlier, the WG is interrupted at 11°N by a 30-km wide gap called the PG. Distribution data of plants, birds, dragonflies, arboreal frogs (Philautus), and fishes suggest that this gap might have served as a potential geographical barrier (Fraser 1936; Subramanyam and Nayar 1974; Ali and Ripley 1987; Dahanukar et al. 2004; Biju et al. 2005). Few molecule-based biogeographic studies have assessed the role of the PG in shaping current distribution of flora and fauna. Microsatellite data have shown strong population structuring across the gap in the case of Asian Elephants (Vidya et al. 2005). For plants, like Eurya nitida and Gaultheria fragrantissima, PG emerged as an important barrier structuring populations within the species across the gap (Bahulikar et al. 2004; Apte et al. 2006). It also appears to be an important barrier for the high-elevation endemic bird, the White-bellied Shortwing (Brachypteryx major) (Robin et al. 2010). Thus, these studies suggest that PG might have served as a barrier to gene flow in species distributed on either side of the gap. However, all these studies have focused on intraspecific variation, and the role of the PG in shaping interspecific patterns of distribution needs to be evaluated. Interestingly, a study on a species of a group of caecilians, traditionally considered to have limited dispersal ability, showed that the gap did not function as a barrier for the species (Gower et al. 2007). Nevertheless, results from a molecular work which included multiple taxonomic groups from India and Sri Lanka pointed toward the role of PG in shaping species assemblages (Bossuyt et al. 2004). Thus, if PG is indeed a major biogeographical barrier, then the distributions of either sister species or clades in a phylogeny would not overlap and would fall on either side of the gap (Scenario 2, S2).
In addition to the PG, the four biogeographic subdivisions of the WG might have also restricted species movements and they serve as ecological barriers. These subdivisions are spatially separated and differ in their climatic envelope and floral composition (Subramanyam and Nayar 1974) (also see Fig. 1). Thus, species in each subdivision might be more closely related to each other than to species from other subdivisions. This scenario (S3) would be reflected in the phylogenetic tree by species from each subdivision forming distinct clades. To date, there is no published molecular work that addresses this ecological barrier scenario for WG.
Three explicit biogeographical scenarios are presented to explain current distribution of flora and fauna in the WG. These hypotheses can be used to generate specific predictions which can then be tested in a phylogenetic framework using biogeographic methods. These scenarios need not be mutually exclusive as they might have operated simultaneously at same/different spatial and temporal scales. To test the predictions of the above-mentioned hypotheses, the centipede genus Digitipes (Scolopendridae: Otostigmini) of the WG was selected. A recent molecular phylogenetic study on scolopendrid centipedes established the monophyly of Indian Digitipes species (Joshi and Karanth 2011a) and another study on species delimitation proposed a comprehensive phylogenetic hypothesis along with revised distribution maps (Joshi and Karanth 2011b). Current distribution patterns of Digitipes species in the WG suggest that SWG is more diverse and has more endemic species than CWG and NWG. Specifically, the SWG have five species of which three are endemics, Nilgiri hill range has three species with no endemics and the CWG has three with one endemic species, whereas the NWG has one nonendemic species (Joshi and Karanth 2011b). Furthermore, most of the species are found at lower (>500 mean sea level [msl]) and midelevation forests (500–1500 msl), except for one in the high-elevation areas of the WG (<1500 msl) (Joshi and Karanth 2011b). Additionally, diversification in the Digitipes lineage started during the Late Cretaceous when the peninsular Indian plate was on its northward journey, and experienced long periods of isolation (Joshi and Karanth 2011a). Thus, Digitipes of WG is ideally suited to test the biogeographical scenarios presented above as they represent an ancient endemic lineage (originating in the pre-volcanism period) and also exhibits a trend of decreasing species richness and endemicity with increasing latitude.
To this end, a recently published Digitipes phylogeny was used to estimate the divergence dates via a Bayesian approach. Additionally, historical biogeographic analysis was performed on the Digitipes phylogeny to reconstruct ancestral areas and test biogeographic hypotheses.