The Great Smoky Mountains National Park All Taxa Biodiversity Inventory: A Search for Species in Our Own Backyard 2007 Southeastern Naturalist Special Issue 1:215–228 1Department of Ecology and Evolutionary Biology, 569 Dabney Hall, University of Tennessee, Knoxville, TN 37996. 2Department of Zoology, 381 David Clark Labs, North Carolina State University, Raleigh, NC 27695. 3USGS Biological Resources Discipline, 1314 Cherokee Orchard Road, Gatlinburg, TN, 37738. *Corresponding author - jlessar@utk.edu. Rarity and Diversity in Forest Ant Assemblages of Great Smoky Mountains National Park Jean-Philippe Lessard1,2,*, Robert R. Dunn2, Charles R. Parker3, and Nathan J. Sanders1 Abstract - We report on a systematic survey of the ant fauna occurring in hardwood forests in the Great Smoky Mountains National Park. At 22-mixed hardwood sites, we collected leaf-litter ant species using Winkler samplers. At eight of those sites, we also collected ants using pitfall and Malaise traps. In total, we collected 53 ant species. As shown in other studies, ant species richness tended to decline with increasing elevation. Leaf-litter ant assemblages were also highly nested. Several common species were both locally abundant and had broad distributions, while many other species were rarely detected. Winkler samplers, pitfall traps, and Malaise traps yielded samples that differed in composition, but not richness, from one another. Taken together, our work begins to illuminate the factors that govern the diversity, distribution, abundance, and perhaps rarity of ants of forested ecosystems in the Great Smoky Mountains National Park. Introduction Robert Whittaker's (1952, 1956) classic work showed that the elevational gradient in the Great Smoky Mountains National Park (GSMNP) strongly infl uences plant and insect communities. Since Whittaker’s work, few studies (see Stiles and Coyle 2001, Van Pelt 1963, Watson et al. 1994) have explored elevational gradients in diversity in the southern Appalachians. As for most taxa, ant diversity often varies systematically along elevational gradients ( Brühl et al. 1998; Fisher 1996, 1998; Olson 1994; Sanders 2002; Sanders et al. 2003), but no studies to date have explicitly examined elevational gradients in ant diversity in the southern Appalachians. At least two early investigators focused on the ant fauna of the southern Appalachians. First, Cole (1940) performed one of the earliest ant surveys of any National Park and one of the first systematic surveys of any taxon in the GSMNP (though he focused only on the Tennessee side of GSMNP). Cole (1940) thoroughly inventoried the ant fauna of the GSMNP and provided notes on the distribution and autecology of the species he observed in his forays. Second, Van Pelt (1963) studied the ant communities of the southern Blue Ridge Mountains with a particular interest in the variation in regional 215 216 Southeastern Naturalist Special Issue 1 patterns of ant diversity and community composition at low and high-elevation sites. These early studies by Cole and Van Pelt provide contemporary ecologists with a unique knowledge of the history of ant assemblages in the southern Appalachian Mountains. Because of the ecological importance and near ubiquity of ants in most terrestrial ecosystems (Hölldobler and Wilson 1990), understanding the causes and consequences of ant diversity is critical to preserving both ecosystem functions and services that ants provide (Folgarait 1998). As part of the All Taxa Biodiversity Inventory (ATBI; [Sharkey 2001]), we have sought to understand the biotic and abiotic controls on ant diversity. Our goals in this paper are to: (i) estimate litter ant species richness at 22 forest sites in the GSMNP; (ii) document the major spatial patterns in the diversity, distribution, and abundance of ants in the GSMNP; (iii) assess whether ant assemblages are nested, where nestedness is a measure of the extent to which species-poor assemblages are subsets of species-rich assemblages; and (iv) examine how different sampling techniques yield different components of ant assemblages. To accomplish the first three goals, we used data from leaf-litter ant assemblages collected at 22 sites during 2004 and 2005. To accomplish the fourth goal, we used a combination of different sampling techniques (pitfall traps, Malaise traps, and Winkler extractors) to inventory ant workers and alates at eight sites in the GSMNP. Methods Sampling We collected ants at 22 forested sites using Winkler extractors in 2004 and 2005. At eight of those 22 sites, we also collected ants using pitfall traps and Malaise traps as part of C. Parker’s “ATBI Pilot Study.” The 22 sites were chosen to cover nearly the entire range of elevation in the GSMNP (260–2021 m). Our sampling design ensured that about 80% of the elevational range in the GSMNP was sampled. Sites were chosen on both the NC and TN side of the GSMNP, and from the southern to the northern boundaries of the park (Fig. 1). All sites were located in mixed-hardwood forests, which is the main forest type found throughout the park (White 1983), and all sites were located away from roads, heavily visited trails, or other human disturbances. Leaf-litter ant sampling (Winkler Extraction). At each of the 22 sites described above, we randomly placed a 50-m x 50-m plot. Within the corners of this plot, we placed a 10-m x 10-m sub-plot, and within the corners of each 10-m x 10-m sub-plot, we sampled ants in four 1-m2 plots. Thus, at each site, there were 16 1-m2 plots. We collected the leaf litter inside each 1-m2 plot and sifted it through a coarse mesh screen of 1-cm grid size to remove the largest fragments and concentrate the fine litter (see Longino and Colwell 1997, Longino et al. 2002). The litter fragments that did not fit through the mesh, twigs, and sticks in each 1-m2 plot were inspected for colonies. The concentrated fine litter from each sample was then suspended 2007 J.-P. Lessard, R.R. Dunn, C.R. Parker, and N.J. Sanders 217 in mini-Winkler sacks for two days in the lab. All worker ants that were extracted from the 1-m2 plots were identified and enumerated and are stored in Sander's ant collection at the University of Tennessee. Pitfall trapping. We used pitfall traps to sample ants at eight forested sites (Table 1) within the GSMNP every two weeks from October 2000 through April 2003. At each of the eight sites, ten pitfall traps were placed at least 3 m apart along an approximately 30-m transect. Traps were 6-cm diameter plastic cups buried fl ush with the soil surface and partially filled with propylene glycol. Pitfall traps effectively capture ground-foraging ants (Bestelmeyer et al. 2000), while not always capturing those that are exclusively litter-dwelling species. Malaise traps. Two Malaise traps (1.6 m × 1.8 m × 1.0 m) were placed on the ground 75–100 m from one another at each of the eight sites at which pitfall traps were placed (Longino and Colwell 1997, Longino et al. 2002). The contents of the alcohol-filled traps were collected every two weeks from January 1999 through January 2002. Analyses For each site and sampling technique, the observed number of species is simply the tally of species collected at the site. We examined how species richness (the total number of ant species occurring at a site) varied with elevation. To estimate species richness, the number of species that would be collected if sampling were to go to completion, we used the Chao2 estimator (Chao 1987, Colwell and Coddington 1994) as: SChao2 = SObs + Q1 2 / 2Q2, where SObs is the number of species that occurred in the sample, Q1 is the number of species that occur in only one sample (uniques), and Q2 is Figure 1. Map of the Great Smokies National Park (GSMNP) showing the 22 sites (white dots) where leaf-litter ants were sampled. The shadings represent the repartition and proportion of the main habitat-cover types in the GSMNP. 218 Southeastern Naturalist Special Issue 1 Table 1. Minimum and maximum elevation and number of sites occupied for forest ant species sampled in Great Smoky Mountains National Park. The elevation records are based on 8 sites for which data for the three different sampling techniques (malaise, pitfall, and Winkler sacks) were available. The number of occurrences are shown for each of the sampling techniques. Elevation (m) Species Min Max Pitfall Winkler Malaise Amblyopone pallipes (Haldeman) 594 1530 5 5 6 Aphaenogaster fulva Roger 594 1530 3 1 0 A. lamellidens Mayr 1530 1530 1 0 0 A. rudis Enzmann 594 1673 6 7 6 A. sp. 594 1380 0 0 2 Brachymyrmex depilis Emery 594 896 2 0 0 Camponotus americanus Mayr 594 594 0 0 1 C. chromaiodes Bolton 594 1033 4 0 4 C. mississippiensis Smith 594 594 0 0 1 C. nearcticus Emery 594 1380 1 0 3 C. pennsylvanicus (De Geer) 594 1530 3 1 5 C. snellingi Bolton 594 594 0 0 1 C. subbarbatus Emery 594 594 0 0 1 Crematogaster ashmeadi Mayr 594 594 0 0 1 C. pilosa Emery 594 1673 1 0 1 C. sp. 594 594 0 0 1 Cryptopone gilva (Roger) 594 896 2 0 0 Formica subsericea Say 594 994 0 0 2 Lasius alienus (Foerster) 594 1673 3 5 4 L. latipes (Walsh) 594 1530 0 0 2 L. nearcticus Wheeler 594 594 1 0 0 Monomorium minimum (Buckley) 594 594 0 0 1 Myrmecina americana Emery 594 1530 6 5 5 Myrmica latifrons Cole 594 896 0 0 2 M. pinetorum Wheeler 594 896 2 0 2 M. punctiventris Roger 594 1673 2 2 2 M. sp. 594 594 0 0 1 Paratrechina sp. 594 594 0 0 1 P. sp1. 594 594 0 0 1 Ponera pennsylvanica Buckley 594 1673 3 2 4 Prenolepis imparis Emery 594 1530 4 1 2 Proceratium croceum (Roger) 594 594 0 0 1 P. pergandei (Emery) 594 594 0 0 1 P. sp. 594 594 0 0 1 P. sp.1 594 594 0 0 1 Pyramica ohioensis (Kennedy & Schramm) 594 594 0 1 0 P. ornata (Mayr) 594 594 1 0 0 P. rostrata (Emery) 594 594 0 1 0 P. sp. 594 594 0 0 1 Solenopsis molesta (Say) 594 1033 2 0 1 Stenamma brevicorne (Mayr) 594 1530 5 3 0 S. diecki Emery 594 1673 7 6 1 S. impar Forel 1033 1530 1 1 0 S. meridionale Smith 594 1673 6 5 0 S. schmittii Wheeler 594 1673 5 3 0 S. sp. 594 594 0 0 1 S. sp.1 994 1530 0 0 3 S. sp.2 594 1673 0 0 7 Tapinoma sessile (Say) 594 1530 0 0 2 Temnothorax curvispinosus Mayr 594 1033 0 0 3 T. longispinosus Roger 594 1380 2 1 4 T. sp. 594 1033 0 0 2 T. sp.1 1380 1380 0 0 1 2007 J.-P. Lessard, R.R. Dunn, C.R. Parker, and N.J. Sanders 219 the number of species that occur in two samples (Colwell and Coddington 1994). The Chao2 index uses data on the rare species collected in the samples (Q1 and Q2) to estimate the number of additional species that are present at the site, but were not recorded in the samples. As in other biodiversity studies of this kind, (Colwell et al. 2004, Kaspari et al. 2000, Longino et al. 2002, Ratchford et al. 2005), we treated each sampling unit (a 1-m2 quadrat for the Winkler sampling or a pitfall trap for the pitfall trapping) as a sample. Because the Chao2 estimator is sensitive to sample size (Colwell and Coddington 1994), we used Colwell’s EstimateS (Colwell 2004) to construct 50 randomized accumulation curves for each site to calculate the standard deviation of the estimated species richness. Across all sites, the asymptotic richness estimator was very similar to the observed total species richness (r2 = 0.64, p < 0.001), and sampled diversity reached a plateau at all but two sites. Therefore, for ease of interpretation, we report only the observed richness for each site and sampling technique. Elevational gradient. To examine whether leaf-litter ant diversity varied systematically with elevation, we plotted observed species richness from the Winkler samples from 22 sites against elevation. We assessed the relationship using linear regression. Nestedness and rarity. We conducted two analyses to understand the distribution of rare ant species in the GSMNP hardwood forests: a nestedness analysis and a simple correlation between the diversity of rare ants and overall diversity. Nestedness analyses provide a measure of the extent to which the species found at species-poor sites are exclusive of or a subset of those found at species-rich sites. If sites are nested, those species at speciespoor sites are a subset of those at species-rich sites. If sites are not nested, then species at species-poor sites are not necessarily found in species-rich sites. We performed an analysis of community nestedness using a presence/ absence matrix comprising 22 sites and 38 species detected in the leaf-litter ant sampling. Nestedness can be assessed using the “nestedness temperature calculator” (NTC) implemented by Atmar and Patterson (1995). The NTC provides a T value between 0 and 100 describing the degree of nestedness of a given set of communities. A T value close to 0 is highly nested (where species at less diverse sites are strict subsets of those at more diverse sites), whereas a T value near 100 describes a random assemblage. The NTC further allows testing for statistical significance by generating 50 random matrices based on the original data set. The mean T value produced in the process is then compared to the observed T value and used to calculate a confidence interval. As an additional test of the distribution of rare ant species in the GSMNP forests, we looked at the correlation between the diversity of rare species (defined as those species found at no more than 4 sites) and overall ant diversity. In many regions, the diversity of rare species and overall diversity do not co-vary (Jetz et al. 2004), and as a consequence, conservation of diversity per se will not necessarily conserve those species most at risk. Abundance and distribution. We examined the shape of the abundance220 Southeastern Naturalist Special Issue 1 distribution curve by ordering species’ frequencies of occurrence (the number of times they were detected) in litter samples, from the most rare to the most common species. This allowed us to illustrate the relative proportion of rare and common species in our litter samples. Then, to test whether abundant ants also tend to be widespread within the GSMNP forests, we regressed the number of occurrences (the number of 1-m2 plots in which a species was detected) against the number of sites (out of 22) at which it was detected. If many species were uniques and/or singletons, it would be an indication that sampling at the scale of the park was relatively incomplete even though sampling in individual sites seemed relatively complete (see above). Comparing sampling techniques. To compare the three sampling techniques (Winkler samples, pitfall traps, and Malaise traps), we first asked whether the number of species collected by one technique was correlated with the number of species collected by the other two techniques at the eight sites at which each of the three sampling techniques were employed. Then, to assess the similarity in composition of the assemblages sampled with each technique, we used the Jaccard’s similarity index. The comparison was limited to the 8 sites where all three sampling techniques had been used. Results Leaf-litter ant assemblages. In total, we detected 38 leaf-litter ant species at the 22 sites. The number of species per m2 ranged from 0–10, and the number of species per site varied from 2–22. In 20 of the 22 sites, the estimators reached an asymptote, indicating that further sampling would have added no new species. Leaf-litter ant species richness declined significantly with increasing elevation (Fig. 2, r2 = 0.63, p < 0.001). Nestedness and rarity. The 22 litter-ant assemblages sampled were nested (T =18.37°). The NTC randomization process generated 50 matrices that had an average T value of 62.2 ± 4.11°. The original matrix had a significantly lower temperature than the mean T for the simulated matrices (p = 2.64e-26), indicating that species-poor assemblages were composed of a subset of species-rich assemblages. The core species of most assemblages were Aphaenogaster rudis, Myrmecina americana, Stenamma diecki, S. meridionale, Lasius alienus, Amblyopone pallipes and Ponera pennsylvanica (Fig. 3). The diversity of rare species was well correlated with overall diversity, such that the most diverse sites had the most rare species (Pearson r = 0.87, p = 0.001 ). Abundance and distribution. The species abundance distribution was approximately log normal, with a few common species and a tail of rare species (Fig. 4a). The abundance of individual species (the number of 1-m2 plots a species was detected in) increased with the number of sites they occupied (Fig. 4b, r2 = 0.74, p < 0.0001). One “species,” A. rudis, represented a large percentage of all occurrences. Aphaenogaster rudis is likely a group of species rather than a single species (Umphrey 1996), but distinguishing species 2007 J.-P. Lessard, R.R. Dunn, C.R. Parker, and N.J. Sanders 221 within the group is possible only by examining their karyotypes. Regardless, the A. rudis species group appears to be tremendously successful in the southern Appalachia, a success that warrants further examination. Pitfall and Malaise samples. Our inventory of the eight ATBI sites yielded a total of 15,340 ant individuals and 53 species, all sampling techniques (Winkler + pitfall + Malaise traps) combined (Table 1). Malaise traps at the same eight sites captured 30 species and over 10,000 alates ; pitfall traps captured 25 species and 2796 workers; and litter samples yielded 17 species and 2242 workers. Due to the limited availability of taxonomic work on ant alates, we identified them to morphospecies and tried to be as conservative as possible in splitting species to avoid infl ating the count of the number of species. We expect this conservatism to cause only a slight, if any, increase in the number of species recorded, and thus no significant effect on the relative species richness yielded by the different sampling techniques. Comparing sampling techniques. To assess whether different sampling techniques provide equivalent estimates of the number of species sampled at our sites and the overall pattern of diversity, we correlated the different measures of species richness with one another. Measures of observed species richness were positively correlated to each other (Table Figure 2. Elevational gradient in leaf-litter ant species richness for 22 sites surveyed during summer 2004–2005 in the Great Smokies National Park. 222 Southeastern Naturalist Special Issue 1 2). The myrmecofauna collected with pitfall traps and Winkler extractors were similar to one another, but the fauna detected in the Malaise traps differed markedly from the faunas collected in Winkler extractors and pitfall traps (Table 3). Discussion To our knowledge, our work is the first systematic sampling of the ant fauna in the GSMNP. As with other taxa in the GSMNP (e.g., Whittaker Figure 3. Presence/absence matrix of forest ant species illustrating the nestedness pattern of the 22 leaf-litter ant assemblages. Species are sorted from top to bottom by the number of sites they occcupy. Table 2. Correlation coefficients showing relationship between richness detected by three sampling techniques. First sample Second sample Correlation coefficient P Pitfall traps Winkler samples 0.835 0.010 Pitfall traps Malaise traps 0.878 0.004 Winkler samples Malaise traps 0.847 0.008 Table 3. The similarity in species composition among assemblages collected with three sampling techniques. Jaccard’s values near 1 indicate more similarity, and values near 0 indicate less similarity. First sample Second sample Jaccard similarity index Pitfall traps Winkler samples 0.703 Pitfall traps Malaise traps 0.480 Winkler samples Malaise traps 0.392 2007 J.-P. Lessard, R.R. Dunn, C.R. Parker, and N.J. Sanders 223 Figure 4. a) Species abundance distribution in 22 sites (black lines) and 288 onem2 plots of leaf-litter samples, and b) correlation between the relative abundance of leaf-litter ant species detected and the number of sites at which they occur. 224 Southeastern Naturalist Special Issue 1 1952), elevation strongly infl uenced leaf-litter ant diversity. Ant species richness decreased monotonically with elevation, a common pattern for many insect elevational diversity gradients (Rahbek 2005). Our results support the observations made by Van Pelt (1963), who found a greater number of ant “forms” and “nests” at lower elevations in the Blue Ridge mountains than at higher elevations. Similarly, Wang et al. (2001) found that ant species richness decreased with elevation in an oak forest of the central Appalachians. A regional faunistic survey of the ants of Georgia (Ipser et al. 2004) also found that ant species richness generally declined with elevation. Of course, the ants do not really respond to elevation per se. Instead, they respond to some biotic (e.g., productivity) or abiotic (temperature, geometric constraints) variable that covaries with elevation. The next step in our work in the GSMNP is to understand the biotic and abiotic factors that shape both spatial and temporal gradients in diversity (Dunn et al. 2007). Leaf-litter ant assemblages in the GSMNP are highly nested. Speciespoor assemblages (generally at the highest elevations) are made up of a subset of those species that occur at the most species-rich assemblages (generally at the lowest elevations). As a consequence, the diversity of rare species tracked the overall pattern of diversity. Ellison et al. (2002) calculated a T value of 15.1 for the bog ant assemblages in New England, comparable to the value we observed here. The ant fauna of the New England bogs is characterized by a few “bog specialists;” here we find a few low-elevation specialists. In our study, ant species collected at the high-elevation sites were widely distributed across the elevational gradient surveyed, whereas the distributions of low-elevation species were often restricted to a few low-elevation sites. Aphaenogaster rudis, M. americana, P. pennsylvanica, and S. diecki were widely distributed along the elevational range covered by our study. The most frequently collected ant species, A. rudis, was abundant within plots, common across multiple plots within sites and found at most sites in the GSMNP, while others species were represented by just a few individuals in one or a few plots (e.g., Cryptopone gilva, Pyramica ornata, Proceratium pergandei). The causes of rarity and abundance in ants remain poorly explored. Some authors (e.g., Davidson et al. 2003) have suggested that the most abundant ants (in tropical forest canopies) tend to be homopteratenders that can monopolize large territories and pools of sugar resources. In contrast to ants in tropical systems, A. rudis, a behaviorally subordinate ant (Fellers 1987, Smallwood 1982), is not known to rely extensively on homoptera exudates, and does not maintain exclusive foraging territories (Lessard, pers. observ.; Smallwood 1982). Instead, Aphaenogaster rudis, like other species in the genus Aphaenogaster (ants in this genus are referred to as the gypsy ants; T.G. McGlynn, California State University, Dominguez HillsCarsn, California, pers. comm.), migrates from nest to nest frequently and feeds on a wide variety of food resources. Interestingly, all of these behaviors are shared by Aphaenogaster araneoides Emery in Costa Rica, 2007 J.-P. Lessard, R.R. Dunn, C.R. Parker, and N.J. Sanders 225 where A. araneoides is extremely abundant in an even more diverse ant community (McGlynn 2006, McGlynn et al. 2004). The rare ants, those species found at only a few sites, represented a mix of different life histories. Rarity may be real or only apparent. Apparent rarity means that the low abundance of certain species in a set of samples is strictly a sampling artifact. In the current study, the combination of sampling techniques and the high correlation between observed and predicted richness suggest that the species that we describe as being rare actually are rare. Furthermore, species that were rare in space (litter samples) were also rare in time (two years of pitfall trapping), suggesting that they have both low local abundance and low frequency of occurrence at the meso-scale. Most of the rare species may be at the edge of their climate envelope within the GSMNP and hence found predominately at the lowest, warmest, and most diverse sites within the park. Perhaps not surprisingly, different sampling techniques detected different ant species (Longino et al. 2002, Martelli et al. 2004). The composition of assemblages collected solely by litter extraction did not differ dramatically from the assemblages detected by pitfall traps. However, it is worth noting that there were two Pyramica spp. that were detected in the leaf-litter samples that were not collected in the pitfall traps. Similarly the reproductive ant fauna collected in the Malaise traps was different from either the pitfall or leaf-litter samples. Taken together, these results suggest that documenting ant diversity in the GSMNP will require multiple sampling techniques, as is the case in other systems (Delabie et al. 2000, Longino et al. 2002). However, it is worth noting that most (68%) of the identifiable ant species collected here were collected by litter extraction and that the overall patterns of diversity detected by the different methods were highly concordant such that all sampling methods showed similar declines in diversity with increasing elevation. Conclusions The ant species that occur at these high-elevation sites generally have broad elevational ranges, whereas many of the species that occur at low elevations are found almost exclusively at low-elevation sites. Similarly, many of the species that occur at high elevations, with broad elevational ranges, also occur at high latitudes (e.g., Lessard and Buddle 2005) and have broad latitudinal ranges. Ant assemblages in GSMNP are highly nested, and most species-poor sites in GSMNP occur at high elevations. This suggests that these broad-ranged species are able to tolerate climatic extremes that are frequent at high elevations and latitudes. Unlike examples of elevational gradients in the southwestern US (e.g., Fleishman et al. 2000), we find few high-elevation endemics and a number of relatively rare species at low elevation. More extensive sampling, especially of other habitat types in GSMNP, could further illuminate the causes and consequences of ant diversity. 226 Southeastern Naturalist Special Issue 1 Acknowledgments Thanks to Jaime Ratchford, Melissa Geraghty, Raynelle Rino, Kerri Crawford, Donny Mai, and Kristin Lane for help in the field and Matt Fitzpatrick for help with Figure 1. We are grateful to Noa Davidai, James Trager, and two reviewers who provided comments on the manuscript. This research was funded by grants to N.J. Sanders and R.R. Dunn by Discover Life in America. Literature Cited Atmar, W., and B.D. Patterson. 1995. The Nestedness Temperature Calculator: A visual basic program, including 294 presence-absence matrices. AICS Research Inc., University Park, NM. Bestelmeyer, B.T., D. Agosti, L. Alonso, C. Roberto, F. Brandao, W.L. Brown, J.H.C. Delabie, and R. Sylvestre. 2000. Field techniques for the study of ground-dwelling ants. Pp. 122–154, In D. Agosti, J.D. Majer, L.E. Alonso, and T.R. Schultz (Eds.). Ants: Standard Methods for Measuring and Monitoring Biodiversity. Smithsonian Institution Press, Washington, DC.. Brühl, C.A., M. Mohamed, and K.E. Linsenmair. 1999. Altitudinal distribution of leaf-litter ants along a transect in a primary forest on Mount Kinabalu, Sabah, Malaysia. Journal of Tropical Ecology 15:265–267. Chao, A. 1987. Estimating the population size for capture-recapture data with unequal catchability. Biometrics 43:783–791. Cole, A.C., Jr. 1940. A guide to the ants of the Great Smoky Mountains National Park, Tennessee. American Midland Naturalist 24:1–88. Colwell, R.K. 2004. EstimateS: Statistical estimation of species richness and shared species from samples. Version 7.0b1 Beta. http:// purl.oclc.org/estimates. Colwell, R.K., and J.A. Coddington. 1994. Estimating terrestrial biodiversity through extrapolation. Philosophical Transactions of the Royal Society of London, Series B 345:101–118. Colwell, R.K., C.X. Mao, and J. Chang. 2004. Interpolating, extrapolating, and comparing incidence-based species accumulation curves. Ecology 85:2717–2727. Davidson, D.W., S.C. Cook, R.R. Snelling, and T.H. Chua. 2003. Explaining the abundance of ants in lowland tropical rainforest canopies. Science 300:969– 972. Delabie, J.H.C., B.L. Fisher, J.D. Majer, and I.W. Wright. 2000. Ants: Standard Methods for Measuring and Monitoring Biodiversity. Pp. 145–154, In D. Agosti, J.D. Majer, L.E. Alonso, and T.R. Schultz (Eds.). Ants: Standard Methods. Smithsonian Institution Press, Washington, DC. Dunn, R.R., C.R. Parker, M. Gerhaghty, and N.J. Sanders. 2007. Reproductive phenologies in a diverse temperate ant fauna. Ecological Entomology 32: 135–142. Ellison, A.M., E.J. Farnsworth, and N.J. Gotelli. 2002. Ant diversity in pitcher-plant bogs of Massachusetts. Northeastern Naturalist 9:267–284. Fellers, J.H. 1987. Interference and exploitation in a guild of woodland ants. Ecology 68:1466–1478. Fisher, B.L. 1996. Ant diversity patterns along an elevational gradient in the Reserve Naturelle Integrale d’Andringitra, Madagascar. Fieldiana Zoology 85:93–108. Fisher, B.L. 1998. Ant diversity patterns along an elevational gradient in the Reserve Special d’Ajanaharibe Sud and on the western Masoala Peninsula, Madagascar. Fieldiana Zoology 90:39–67. 2007 J.-P. Lessard, R.R. Dunn, C.R. Parker, and N.J. Sanders 227 Fleishman, E., J.P. Fay, and D.D. Murphy. 2000. Upsides and downsides: Contrasting topographic gradients in species richness and associated scenarios for climate change. Journal of Biogeography 27:1209–1219. Folgarait, P. J. 1998. Ant biodiversity and its relationship to ecosystem functioning: A review. Biodiversity and Conservation 7:1221–1244. Hölldobler, B., and E.O. Wilson. 1990. The Ants. Belknap, Cambridge, MA. Ipser, R.M., M.A. Brinkman, W.A. Gardner, and H.B. Peeler. 2004. A Survey of ground-dwelling ants (Hymenoptera: Formicidae) in Georgia. Florida Entomologist 87:253–260. Jetz, W., C. Rahbek, and R.K. Colwell. 2004. The coincidence of rarity and richness and the potential signature of history in centres of endemism. Ecology Letters 7:1180–1191. Kaspari, M., S. O’Donnell, and J.R. Kercher. 2000. Energy, density, and constraints to species richness: Ant assemblages along a productivity gradient. The American Naturalist 155:280–293. Lessard, J.P., and C.M. Buddle. 2005. The effects of urbanization on ant assemblages (Hymenoptera: Formicidae) associated with the Molson Nature Reserve, Quebec. Canadian Entomologist 137:215–225. Longino, J.T., and R.K. Colwell. 1997. Biodiversity assessment using structured inventory: Capturing the ant fauna of a tropical rain forest. Ecological Applications 7:1263–1277. Longino, J.T., J. Coddington, and R.K. Colwell. 2002. The ant fauna of a tropical rain forest: Estimating species richness three different ways. Ecology 83:689–702. Martelli, M.G., M.M. Ward, and A.M. Fraser. 2004. Ant diversity sampling on the southern Cumberland Plateau: A comparison of litter sifting and pitfall trapping. Southeastern Naturalist 3:113–126. McGlynn, T.P. 2006. Ants on the move: Resource limitation of a litter-nesting ant community in Costa Rica. Biotropica 38:419–427. McGlynn, T.P., R.A. Carr, J.H. Carson, and J. Buma. 2004. Frequent nest relocation in the ant Aphaenogaster araneoides: Resources, competition, and natural enemies. Oikos 106:611–621. Olson, D.M. 1994. The distribution of leaf-litter lnvertebrates along a neotropical altitudinal gradient. Journal of Tropical Ecology 10:129–150. Rahbek, C. 2005. The role of spatial scale and the perception of large-scale speciesrichness patterns. Ecology Letters 8:224–239. Ratchford, J.S., S.E. Whittman, E.S. Jules, A.M. Ellison, N.J. Gotelli, and N.J. Sanders. 2005. The effects of fire, local environment, and time on ant assemblages in fens and forests. Diversity and Distributions 11:487–497. Sanders, N.J. 2002. Elevational gradients in ant distributions: Area, species richness, and Rapoport’s rule. Ecography 25:25–32. Sanders, N.J., J. Moss, and D. Wagner. 2003. Patterns of ant species richness along elevational gradients in an arid ecosystem. Global Ecology and Biogeography 12:93–102. Sharkey, M. J. 2001. The all taxa biodiversity inventory of the Great Smoky Mountains National Park. Florida Entomologist 84:556–564. Smallwood, J. 1982. The effect of shade and competition on emigration rate in the ant Aphaenogaster rudis. Ecology 63:124–134. Stiles, G.J., and F.A. Coyle. 2001. Habitat distribution and life history of species in the spider genera Theridion, Rugathodes, and Wamba in the Great Smoky Mountains National Park (Araneae, Theridiidae). Journal of Arachnology 29:396–412. Umphrey, G.J. 1996. Morphometric discrimination among siblings in the Fulva228 Southeastern Naturalist Special Issue 1 Rudis-Texana complex of the ant genus Aphaenogaster (Hymenoptera: Formicidae). Canadian Journal of Zoology 74: 528–559. Van Pelt, A. 1963. High-altitude ants of the Southern Blue Ridge. American Midland Naturalist 69:205–223. Wang, C.L., J.S. Strazanac, and L. Butler. 2001. Association between ants (Hymenoptera: Formicidae) and habitat characteristics in oak-dominated mixed forests. Environmental Entomology 30:842–848. Watson, J.K., P.L. Lambdin, and K. Langdon. 1994. Diversity of scale insects (Homoptera, Coccoidea) in the Great Smoky Mountains National Park. Annals of the Entomological Society of America 87:225–230. White, P.S. 1983. Eastern Asian-eastern North American fl oristic relations: The plant community level. Annals of the Missouri Botanical Garden 70:734–747. Whittaker, R.H. 1952. A study of the summer foliage insect communities in the Great Smoky Mountains. Ecological Monographs 22:1–44. Whittaker, R.H. 1956. Vegetation of the Great Smoky Mountains. Ecological Monographs 26:1–69.