How to Evaluate Hypotheses When Data Availability is Poor

Selection criteria of appropriate study subjects may involve specific ecological traits of target groups such as niche breadth and dispersal ability. Suitable study subjects are also characterized by data availability, which is especially a problem in remote, poorly known regions. With respect to the Pantepui region, recent zoogeo-graphic studies have focused on the biogeography of amphibians and reptiles (i.e. herpetofauna), which are comparatively well known (McDiarmid and Donnelly 2005). Such herpetofaunal communities are perhaps better suited for biogeographic analyses than other terrestrial vertebrates since in general these animals are less mobile than birds and most mammals and may be closely tied to specific habitats (e.g. Hoogmoed 1979; McDiarmid and Donnelly 2005; MacCulloch et al. 2007). Sampling efforts largely varies among the tepuis and are hardly comparable between neighbouring tepuis. As a consequence, it may be more useful to compare herpetofaunal communities between different massifs rather than single tepuis (McDiarmid and Donnelly 2005). By doing so, the distinctness of local herpetofaunal

Herpetofaunal assembladges > 1500 m

Cerro Yavi Yutaje


Los Testigos -

Aprada -

Chimanta -

Auyan -

Cerro Guaiquinima Eastern Tepui Chain


Fig. 3 Comparison of herpetofaunal assemblages at massifs in the Pantepui region. Most herpeto-faunas show limited overlap only, even when compared to neighbouring massifs. Cluster analysis was performed based on a species list of amphibians and reptiles compiled from Donnelly and Myers (1991), Mägdefrau et al. (1991), McDiarmid and Donnelly (2005), MacCulloch et al. (2007), Means and Savage (2007), Schlüter and Rödder (2007), Myers and Donnelly (2008), Rödder and Jungfer (2008)

assemblages as a result of the high degree of local endemism becomes obvious (Fig. 3.). The wide lack of overlapping faunal elements makes broad scale assessments and evaluation of hypotheses difficult. Phylogeographic analyses would be the most suitable tool for an evaluation of the hypotheses formulated above; however, DNA samples, as required, are largely unavailable.

GIS-based Climate Envelope Distribution Models (CEDMs) may provide an easy-to-use-alternative. In recent times, there have been several examples using CEDMs for species or habitat climate envelopes under past, present and future climate scenarios (e.g. Hugall et al. 2002; Bonaccorso et al. 2006; Peterson and Nyari 2008). Such approaches rely on the assumption that climatic tolerances of species are the primary determinants of their current distributions and that specific climatic niches are conservative, at least within an evolutionary short time frames of some hundreds to thousands years (e.g. Wiens and Graham 2005; but see Pearman et al. 2007).

Peterson and Nyari (2008) and Carnaval and Moritz (2008) suggested that spatial modelling of species or habitats under current and palaeoclimate regimes can generate hypotheses about the potential existence and the extent of stable (refugial) regions. On the other hand spatial models may likewise provide information on past - or future - extents of current refuges.

Was this article helpful?

0 0

Post a comment