Gap Analysis of Five Orders in Great Smoky Mountain National Park: A Quantification of Inventory Gaps

Global biodiversity is currently in a state of crisis with human alterations to the environment exacerbating extinctions so that extinction rates now exceed 1,000 times normal background rates (Lees & Pimm, 2015, Nicholas & Langdon, 2007). To better understand and protect global biodiversity, the All Taxa Biodiversity Inventory (ATBI) project was founded to determine the identity, distribution, and function of every species present within a specific study location (Sharkey, 2001). The most famous ATBI was established in 1998 at Great Smoky Mountain National Park and has since identified over 20,000 species with almost 1,000 species new to science (Discover Life in America, 2014; Nichols & Langdon, 2007; Parker & Bernard, 2006). To help the GSMNP ATBI project use its resources more efficiently, I conducted a taxonomic gap analysis for five orders to identify whether more species may potentially exist within the Park’s boundaries that have yet to be added to species inventories. If inventory gaps were present, I then estimated the total species richness to determine which order had the largest taxonomic gap and should thus be the focus of future sampling efforts. The Park had previously identified five orders that it believed may contain taxonomic gaps: crustaceans, diptera, hemiptera, hymenoptera, and acari. Given species presence locations for these orders, I generated species accumulation curves to determine if taxonomic gaps were present in the Park’s inventories. The species richness modeling program EstimateS was then used to quantify total species richness for each of the focal orders within the Park (Colwell, 2013). Given the estimated total species richness and the number of species previously found by the Park for each order, I was able to quantify the taxonomic gaps in each order’s inventory in terms of the total number of species and the percent of the order identified. To determine where future sampling efforts should be focused to identify the remaining species, I used the species distribution program, Maxent to locate areas of high species richness for each order within the Park (Philips, Dudik, &Shapire, 2010). I compiled 15 environmental predictor layers at a 30m resolution which were uploaded into Maxent along with the presence points of all species that were present at 15 or more locations. Habitat suitability produced by the model was then thresholded and stacked for all species within each order to identify areas of high species overlap. For cases in which the data were spatially structured, bias files were constructed to remove the direct and indirect influences of spatial bias on the models. From the species accumulation curves, it was determined that crustaceans, diptera, hemiptera, and acari all likely had species yet to be found within the Park, while the hymenoptera accumulation curve approached an asymptote at around 550 species. After conducting the species richness analysis, hemiptera was found have the largest gap with about 334 species potentially yet to be identified. However, acari has the largest gap in terms of the percent of the order identified by the Park (45.74%). This Park can now decide whether to view taxonomic gaps in terms of the potential number of species or percent identified. After order-level species distributions were calculated, hemiptera and hymenoptera had the highest richness in the northern and central parts of the Park, diptera and acari were found primarily along streams, and crustaceans had the highest richness in the western parts of the Park. Distance to streams, soil type, vegetation, and slope all play critical roles in defining habitat suitability for these focal orders and all five focal orders can be sampled in the areas in close proximity to Park streams. This analysis identified that past sampling efforts have primarily occurred along trails and roads within the Park so future sampling should be tailored to focus away from these anthropogenic features to avoid under sampling. One of the benefits of this analysis is that its accuracy improves as more sampling is done. With more sample presence locations, both the species richness and species distribution models better reflect reality. Combining species richness and species distribution modeling to structure species inventory efforts will result in more efficient and effective use of resources which will allow the ATBI to better conserve and protect biodiversity within the Park