Conservation of aquatic insect species across a protected area network: null model reveals shortfalls of biogeographical knowledge

The effective conservation of species requires some understanding of where populations occur in a landscape. Gaps in this knowledge base (the “Wallacean Shortfall” of some authors) may coincide with hotspots of diversity for different plant and animal species, requiring the cooperation of a number of different federal, state, local and non-governmental agencies for effective conservation. In this example, the distribution and abundance of benthic macroinvertebrates are widely used as metrics for water quality monitoring, but far less is known about these organisms qua species (taxonomic orders EPT—Ephemeroptera, Plecoptera and Trichoptera). In this study, we inventoried a network of individual US National Park units for species in these orders. These parks are located in geological, ecological and historical places of interest across the states of Alabama, Georgia, Kentucky, North Carolina, South Carolina, Tennessee and Virginia. We sampled these parks in a multi-year intensive inventory in order to determine the composition of the aquatic insect fauna in each park. Since there are no comprehensive accounts of the geographic ranges of these species, we compiled published accounts of species occurrences in these and adjacent states (Arkansas, Florida, Louisiana, Mississippi, West Virginia) to construct a potential species pool for each state. This pool comprised our best estimate of the EPT species that might potentially occur in each state. We used these source pools to test null hypotheses on whether parks disproportionately under- or over-protect species in different categories of risk of imperilment. We find that parks have fewer rare (G1) species than expected from a null model, and parks over-protect some of the most common (G5) species in the network. This pattern would be expected if the actual landscape distributions of the most imperiled (G1) species are small and/or disjunct and tend to occur outside of the national parks in the region. Interactions between park shape (and size) and individual species geographic ranges are likely to influence the precision of estimates of the potential species pool within a protected area. More research is needed on the distribution of imperiled species across the entire geographic range of species, and the traditional practice of compilation and reporting of occurrence records by state is not sufficient for informed conservation practice. State natural heritage programs and biodiversity conservation database efforts (e.g. NatureServe) implicitly recognize the importance of species ranges, but our analysis demonstrates the need to assess these patterns at a finer spatial grain in order for these state lists to serve as meaningful expectations of the composition of species assemblages. Our analysis considers only a tiny fraction of the protected lands in the region, and an enormous additional area of protected lands exists where many of these rare species occur. More precise and accurate reporting of EPT species occurrences in this region will allow resource managers to target the conservation of particular species within single parks, or across protected area networks.