Data from: Corridors increase ant diversity over time
Data files
Jul 13, 2026 version files 96.76 KB
-
Corridor_AntCommunityData_2015to2019.csv
87.62 KB
-
FunctionalGroups.csv
4.58 KB
-
README.md
4.56 KB
Abstract
Globally, habitat loss and fragmentation pose significant threats to biodiversity. Habitat corridors are often used in fragmented landscapes to reconnect isolated patches. However, understanding how corridors function can be challenging, as their effects may be confounded with simultaneous changes to habitat area, edge amount, and connectivity among patches. Here, we used a long-term, landscape-scale experiment to investigate two potential mechanisms by which corridors may influence biodiversity - via increasing habitat connectivity or through increasing the amount of edge. With pitfall trap sampling conducted over 5 years (9 to 15 years since site establishment), we assessed the relative effects of connectivity and edge effects by measuring ant biodiversity in four ways: (1) taxonomic diversity, (2) functional diversity, (3) community composition, and (4) site-to-site variance in community composition (i.e., beta diversity). We found evidence of both connectivity and edge effects on ant diversity and community composition. Taxonomic diversity increased linearly in all patch types over time yet accumulated more quickly in connected than isolated patches. The faster accumulation of taxonomic diversity in connected patches was associated with greater site-to-site variance in community composition. Although we found no differences in mean taxonomic diversity within patches (alpha diversity), the faster accumulation of taxonomic diversity and greater beta diversity among patches with corridors suggests that ant diversity in connected patches may eventually surpass that of isolated patch types. Finally, while we found no effect of edge amount for how fast taxonomic diversity accumulated over time, functional group diversity accumulated more quickly in winged patches that have more edge habitat relative to interior habitat than rectangular patches. Taken together, these results add to a body of literature that shows the potential for habitat corridors to have positive effects on biodiversity by increasing connectivity and that these effects may continue to accrue over time.
Dataset DOI: 10.5061/dryad.z08kprrwn
Description of the data and file structure
Ant community data were collected from the Savannah River Site Corridor Experiment between 2015 and 2019. Experimental landscapes within the Corridor Experiment were established at the Savannah River Site (SRS) in South Carolina in the winter of 1999-2000 and 2006. The Savannah River Site is a National Environmental Research Park, a United States Department of Energy site that is managed by the United States Department of Agriculture Forest Service. Each experimental landscape (N = 7) consists of a center square patch (1 ha) surrounded on each side by four peripheral patches that are each equal in area (~1.4 ha, Fig. 1) and located 150 m away. For each experimental landscape, one peripheral patch is connected to the center patch by a corridor (150×25 m; hereafter "connected"). The three remaining peripheral patches are isolated and are one of two types. They either have 75×25 m projections equal in area to the corridor on opposite sides (hereafter “winged”) or are rectangular in shape. Comparisons among these treatments allow us to disentangle the effects of connectivity and edge amount that typically confound the interpretation of habitat corridor effects. Thus, to isolate effects of connectivity, we compare winged and connected patches, and to isolate effects of edge amount we compare the winged patches with the rectangular patches.
To characterize the ant community, in each year we deployed pitfall traps in the experimental landscapes between July and August, the time of peak ant activity in our study system. Once per year, we placed 22 pitfall traps within a permanent grid established in each of the landscape patches (Fig. 1; 66 traps per experimental landscape × 7 experimental landscapes = 462 traps/year). Pitfall traps were evenly dispersed across each patch, 25m away from the nearest neighboring pitfall trap with traps located along the edge 12.5m away from the matrix tree line. We randomly chose one of the duplicated isolated peripheral patch types at the start of sampling. Each pitfall trap consisted of a 50-mL centrifuge tube filled with 25 mL of a 50/50 solution of propylene glycol and 70% ethanol. We added a couple of drops of clear, unscented dish soap to each gallon of trapping solution to help break surface tension when arthropods fell into the trap. To account for digging-in effects, we dug traps at least 48 hours prior to opening (Greenslade 1973). To dig in traps, we removed a core of soil the same size as the 50 mL centrifuge tube. We opened traps for 48 hours when the chance of rain was low. After collecting traps at the end of sampling, we replaced the trapping solution with 95% ethanol and identified all ants to species or to species complexes with keys and species descriptions in (MacGown 2003).
Files and variables
File: FunctionalGroups.csv
Variables
- Species: Genus and species epithet
- Subfamily: Subfamily assigned to each ant species.
- FunctionalGroup_Andersen: Functional grouping assigned to each species according to Andersen 1997. Functional groupings describe a species association with environmental attributes and their competitive abilities. These functional groupings include: dominant Dolichoderinae, subordinate Camponotini, hot climate specialists, cold climate specialists, tropical climate specialists, opportunists, generalized Myrmicinae, and cryptic species.
- Functional_Group_DelToroetal: Functional grouping assigned to each species according to Del Toro et al. 2015. Del Toro et al.’s classification scheme separates ant species into four functional roles in ecosystems. These functional groupings include soil movers, decomposers, invertebrate community regulators, and seed dispersers
- Native_YorN: Describes whether a species is native to the southeastern United States (Y = native, N = non-native).
File: Corridor_AntCommunityData_2015to2019.csv
Variables
- Year: Sampling year
- Block: Identity assigned to each experimental landscape (n=7).
- Patch: Letter assigned to each patch within a landscape.
- Type: Experimental treatment assigned to each patch within each experimental landscape. Experimental treatments include (Connected, Rectangular, and Winged)
- Species: Species of ant identified
- ProportionOfTrapsWithSpp: Proportion of traps containing the species summarized at the patch level for each year.