This project focused on dead-wood-dependent (saproxylic) arthropods - an ecological community that is poorly known, yet critically important for maintaining healthy productive forests. The goals were to survey, quantify, and predict the spatial distribution of biodiversity across montane forests of the southern Appalachians. This region is well known global biodiversity hotspot for vertebrates such as salamanders, but is likely to contain a considerable amount of as-yet undiscovered invertebrate diversity. To achieve our goals we performed targeted fieldwork with replication, applied objective yet broadly connectable biodiversity metrics based on DNA sequence data (i.e., Phylogenetic Diversity and related measures) in order to circumvent challenges associated with the existence of many undescribed species / species complexes, and coupled this with identification of environmental correlates that could be used for geo-spatial projections of biodiversity hotspot and coldspot locations.
Results/Conclusions .
As suspected, current taxonomy underestimated the Phylogenetic Diversity contained within some groups considerably (e.g., lineages within a single named species can show DNA sequence divergences that are equivalent to genus-level divergences in other taxa), and a non-negligible proportion of samples were collected at larval or otherwise unidentifiable immature life stages. The major arthropod groups represented in our sampling were spiders and opiliones (class Arachnida), centipedes (class Chilopoda), millipedes (class Diplopoda) and insects such as termites and cockroaches (order Blattodea), ants (order Hymenoptera), and beetles (order Coleoptera). By far, spiders and beetles contributed the greatest Phylogenetic Diversity. As a group, the chosen suite of potential environmental correlates of biodiversity showed significant partitioning of variance over several contrasting spatial scales (i.e., among rotting logs within a site, across sites within a forest region, and between forest regions), and therefore provide a strong basis upon which to identify easy to measure surrogates. Development of models for mapping hotspots, and their potential applicability to forecasting distributional shifts under future climate change will be discussed.