Northern Prairie Wildlife Research Center
Within the coastal prairie, 700 species of insects, i.e., gamma diversity, representing 13 orders and 126 families were taken in sweep samples (Table 1). Nearly 65% of families were included in Diptera, Coleoptera, and Hymenoptera, with 19% of families represented in Hemiptera and Homoptera, 6% and 4% represented in Orthoptera and Lepidoptera, respectively, and 1 to 2 families in each of the remaining orders. An average of 27.4% (range = 20 to 32% among the sites) of the species were unique to a site, with no difference in proportion of unique species between areas dominated by woody ( = 27.8%) or prairie ( = 26.7%) vegetation.
|Table 1. Number of families, species (above), and total number of individuals (below) for each order, as well as proportion of unique species, species diversity (Shannon-Wiener H'), and equitability (J) in eight study sites in coastal prairie. Number of species for each study site cannot be added to generate total number of species in an order because species could be common to several sites. Sites 1, 2, 4, and 5 = woody vegetation with bordering prairie/pasture; Site 3 = wooded bottomland; Sites 6 and 7 = grazed/abandoned pasture with some woody vegetation; and Site 8 = coastal prairie.|
Grand Total (%)
|Number of Sweeps||600||800||600||1200||1200||2400||1200||1200|
|% Unique Species||29.4||32.4||30.5||19.9||26.6||24.1||25.7||30.4||56.2|
|* Neuroptera, Mecoptera,
Psocoptera, Odonata, Tricoptera, and Thysanoptera.
200 sweeps also taken in woody vegetation at all sites except 6 and 8.
There was no significant difference within the coastal prairie in the overall distribution of the number of species among sites by any insect order (G = 39.35, df = 48, P > 0.05; Table 1). However, sites dominated by woody vegetation (Sites 1 to 5) contained more species than those dominated by pasture or prairie (Sites 6 to 8; Table 1); for example, Site 2 contained the most species with Homoptera, Diptera, and Coleoptera being the most abundant orders. Grazed pastures (Sites 6 and 7) had the highest number of individuals. Nearly half of the individuals collected at Site 6 were Diptera, while at Site 7, which contained some woodland, Orthoptera, Homoptera, and Coleoptera were abundant. Despite a high sampling intensity (1200 sweeps), coastal prairie contained a much lower total number of species and number of individuals than the other areas. This reduction occurred because of a decrease in Homoptera, Coleoptera, Diptera, and a smaller decrease in other orders. The reduction in these orders was primarily because of fewer phytophagous forms (see Trophic Structure below).
Alpha diversity within the coastal prairie was higher in wooded sites (average H' = 1.64) than those dominated by pasture or prairie (average H' = 1.32; F = 3.65, df = 1,6, P < 0.10; Table 1). For wooded areas, Site 2 had the highest diversity, corresponding with the highest number of species. Homoptera, Hymenoptera, Coleoptera, and Diptera contained the highest diversity for wooded areas, which corresponded with their abundance. Sites dominated by pasture or prairie contained the lowest diversity, with species concentrated in Homoptera, Hymenoptera, and Coleoptera. As with the number of species, diversity at Site 7 was higher because of the presence of the riparian woodland. Diversity at prairie/pasture sites was low because richness, not equitability, was low (Table 1). Equitability was similar among all areas although equitability was higher in Sites 2 and 4 because individuals were more evenly distributed among taxa. In contrast, an abundant species of Coleoptera (Chrysomelidae) in Site 5, an abundant species of Diptera (Anthomyiidae) in Site 6, and abundant species of Diptera (Bibionidae), Orthoptera (Tettraigidae), and Hemiptera (Miridae) in Site 8 lowered equitability.
Within the central prairie, 488 or 97% of the 501 species of insects collected, i.e., gamma diversity, were (in decreasing order of species abundance) Hymenoptera, Diptera, Coleoptera, Homoptera, Hemiptera, Orthoptera, and Lepidoptera, while the remaining 3% were distributed among six other orders (Table 2). Total number of species in herbaceous habitat was nearly identical at each study site: 202, 202, and 198 in Sites 1, 2H, and 3H, respectively. Woody vegetation in Sites 2W and 3W contained 64 and 71 species, respectively. Because species taken from woody vegetation were different from those in the herbaceous component (see section on Similarity Analyses below), total number of species collected in Sites 2 and 3, 249 and 242, respectively, was greater than for Site 1. The proportion of unique species averaged 49% (range = 42-57%) for sites in the central prairie. Site 3 contained the largest proportion of unique species, attributable primarily to a higher proportion of Diptera (Table 2).
|Table 2. Number of families, species (above), and total number of individuals (below) for each order, as well as proportion of unique species, species diversity (Shannon-Wiener H'), and equitability (J) in each study site in central prairie (Site 1 = upland prairie; Site 2 = semi-open woodland; and Site 3 = woody vegetation with herbaceous understory along creek). Number of species in each habitat cannot be added to generate total number of species in an order because species could be common to several sites. H = herbaceous, W = woody; and 2Total and 3Total = combined samples from herbaceous and woody vegetation.|
|Study Sites||Grand Total (%)|
|Number of Sweeps||500||750||200||950||600||200||800|
|% Unique Species||42.1||47.8||57.0||68.3|
|* Collembola, Neuroptera, Odonata, Plecoptera, Psocoptera, and Thysanoptera.|
There were significant differences among sites within the central prairie in the distribution of species across orders (G = 42.49, df = 10, P < 0.005; Tables 2 and 3). When sites were subdivided by habitat, one maximally nonsignificant subset included all sites except Site 1, which indicated that Site 1 was significantly different from Sites 2H, 2W, 3H, and 3W (top contrast; Table 3). Site 1 differed by having a greater abundance of Orthoptera (ratio of species abundance relative to other sites = 1.23; Table 3) in contrast to a greater abundance of Coleoptera and Hymenoptera on the remaining sites (ratio of species abundance = 0.27 and 0.20, respectively; Table 3). Two other nonsignificant subsets also were found. Sites 1 and 2H were excluded from a subset including Sites 2W, 3H, and 3W (middle contrast; Table 3). The former sites contained higher numbers of species of Homoptera and Hemiptera (ratio = 1.75 and 1.83, respectively), but much higher numbers of species of Orthoptera (species ratio = 3.46; Table 3). Finally, Sites 2H, 3W, and 3H formed a nonsignificant subset excluding Sites 1 and 2W (bottom contrast; Table 3). The latter sites contained a higher proportion of Orthoptera (ratio = 1.32; Table 3). Overall, these comparisons indicated a basic difference between Site 1 and the remaining sites.
|Table 3. Maximally nonsignificant subsets of sites identified by log-likelihood ratio tests for the central prairie. Sites in the left column indicate those not included in subset; there was significant heterogeneity at P < 0.05 if these sites were included. H = herbaceous, W = woody; and 2Total and 3Total = combined samples from herbaceous and woody vegetation.|
|Order||Number of species in sites||Ratio|
|1||2H + 2W + 3H + 3W|
|Order||1 + 2H||2W + 3H + 3W||Ratio|
|Order||1 + 2W||2H + 3H + 3W||Ratio|
Total number of individuals within the central prairie collected in herbaceous habitats was nearly identical; 1,123, 1,128, and 1,186 individuals were in Sites 1, 2H, and 3H, respectively (Table 2). Distribution of individuals among habitats was similar to the distribution of species. Site 1 contained more individuals of Homoptera and Orthoptera, Site 2H contained the most Hemiptera, Coleoptera, and Hymenoptera, and Site 3H had the highest number of Diptera. With the exception of one dipteran species in Site 3H and one coleopteran species in Site 2H, no single species was numerically dominant. Subtracting these two species resulted in a total of 929 individuals for Site 2H and 764 individuals for Site 3H, which indicated a greater productivity in Site 2H. Woody vegetation in Sites 2W and 3W contributed more individuals to these areas, which brought the total number of individuals for all three sites to 1123, 1249, and 1335, respectively. Because individuals in woody vegetation represented species largely distinct from those in herbaceous vegetation (see below), the increased total number is meaningful taxonomically and is not simply an increase resulting from additional sweep samples.
Alpha diversity in the herbaceous vegetation within the central prairie was approximately the same for the three sites, although Site 3H was somewhat less diverse (and less equitable) because of higher abundance of a single species of Diptera (Table 2). Without this species, average diversity of the herbaceous component of Site 3H was 1.89, indistinguishable from that for Site 1 (1.93). Site 1 had the greatest diversity of Homoptera (although indistinguishable from Site 2H), Diptera, Hemiptera, and Orthoptera, which reflected the herbaceous nature of this site, and Site 3H had the highest diversity of Coleoptera and Hymenoptera. Addition of samples from woody vegetation increased overall species diversity of both Sites 2 and 3. Equitability of insects in herbaceous vegetation at Site 3H was lower because of the presence of one Diptera species. When that species was removed, overall distribution of individuals among species was approximately the same in all sites, with J = 0.84, 0.77, and 0.82 for Sites 1, 2H, and 3H, respectively. With samples from woody vegetation added, equitabilities of Sites 2 and 3 were lower than for Site 1.
There were significant shifts in abundance of species (both total numbers and proportions) within orders between the two prairie areas (G = 13.35, df = 5, P < 0.05; Tables 1 and 2; i.e., coastal versus central prairie, with Homoptera through Hymenoptera in the test because these orders contained the majority of taxa). There was a greater proportion of Hymenoptera and Orthoptera and a lower proportion of Homoptera and Coleoptera in the central prairie (compare Tables 1 and 2).
Average pairwise similarity was 32% among sites in the coastal prairie, which showed a marked structure of the habitats (Fig. 1). The closest pairings were between Sites 4 and 5 (woodland with bordering pasture; S = 45%) and Sites 1 and 2 (riparian woodland; S = 42%), but the major trend was a contrast of open, more exposed habitats (Sites 6 and 8; pasture/prairie) with the remaining sites containing some canopy (S = 23%). Sites 3, 4, 5, and 7 (pasture with riparian woodland) formed one subgroup within the canopy sites, while Sites 1 and 2 (riparian woodland) formed the remaining group.
|Figure 1. Dendogram displaying results of cluster analysis of the similarity matrix among sites based upon sweep samples in coastal prairie. Numbers at tips of clusters represent study sites that were joined at the level indicated by the horizontal lines. (See Methods for study site numbers).|
Pairwise similarity among sites in the central prairie was lower, averaging 20.9% (Fig. 2). There was a simple, hierarchical structure, unlike the more structured association among sites in the coastal prairie. This result indicated that there were no distinctly different subsets of habitats within this area. Nevertheless, the herbaceous components of Sites 1 and 2 formed a single, mutually close, faunal pairing with approximately 45% of faunal elements overlapping (Fig. 2). Samples from the herbaceous vegetation along the creek in Site 3 then joined this cluster at a slightly lower level of similarity (S = 30%), forming a cluster of all three herbaceous or understory communities. Similarity among grass and herbaceous components (S = 35%) of the study sites was greater than among woody components (S = 12%) or between herbaceous and woody (S = 14%) components. Hence, all herbaceous samples were mutually closest and separate from the woody samples within each site.
|Figure 2. Dendogram displaying results of cluster analysis of the similarity matrix among sites based upon sweep samples in central prairie. Numbers at tips of clusters represent study sites that were joined at the level indicated by the horizontal lines. H = herbaceous and W = woody. (See Methods for study site numbers).|
There was no significant difference in the distribution of species by trophic category among sites within the coastal prairie (G = 16.45, df = 21, P > 0.05). Averaged over all sites, herbivores constituted 58.6% ± 1.36% (mean ± SE), predators 9.5% ± 0.89%, parasites 12.9% ± 0.99%, and detritivores 19.0% ± 1.09% of the total fauna (Table 4).
|Table 4. Number of species (above) and percent of total fauna (below) in each trophic level for all study sites in coastal prairie. Number of species in each trophic level for all study sites in coastal prairie. Number of species in each trophic type may not sum to the total because some species were found in several habitat types.|
|Trophic type||Study Sites||Total|
There was a significant difference in the distribution of species among trophic levels for sites within the central prairie (G = 47.37, df = 12, P < 0.005; Table 5). There were two nonsignificant subsets of trophic levels over all sites (Table 6). The first included herbivores, predators, and parasites, and excluded detritivores. This reflected a higher proportion of detritivores in wooded areas (Sites 2W and 3W; average ratio = 4.25 versus average ratio of detritivores of 12.6 in the sites containing herbaceous vegetation (Sites 1, 2H, and 3H; Table 6). The contrast of herbivores versus the remaining trophic categories also represented significant heterogeneity. This contrast was similar to the first contrast in reflecting higher proportions of herbivores in nonwooded sites (Sites 1, 2H, and 3H, average ratio of 2.1) than wooded sites (average ratio of 0.99; Table 6).
|Table 5. Number of species (above) and percent of total fauna (below) in each trophic level for all study sites in central prairie. Number of species in herbaceous (H) and woody (W) vegetation does not necessarily sum to the total for an area as some species were found in both habitat types. 2Total and 3Total = combined samples from herbaceous and woody vegetation.|
|Trophic Level||Study Sites||Total|
|Table 6. Maximally nonsignificant subsets for trophic categories among sites identified by log-likelihood ratio tests for central prairie. Detritivores (D) were excluded from herbivores (H), predators (P), and parasites (Pa), in upper contrast and herbivores were excluded in the lower contrast.|
|Trophic Category||Study Sites|
|H, P, Pa||191||194||53||183||61|
|P, Pa, D||57||60||32||88||39|
As in the coastal prairie, herbivores constituting 64.3% ± 5.1% of the insect fauna also were the dominant trophic component in the central prairie, (Table 5), but contained few species in great abundance. Parasites accounted for 16.7% ± 1.8%, predators for 10.0% ± 2.0%, and detritivores for 9.1% ± 1.6% of the fauna. There were proportionately more herbivores in prairie (Site 1) and herbaceous vegetation (Sites 2H and 3H) than in woody vegetation, which contained more detritivores. Woodland (Site 3) had a greater number and proportion of parasites, and twice as many predators and detritivores as Sites 1 or 2.
The distribution of total number of species in each trophic category was significantly different between coastal and central prairies (G = 24.66, df = 3, P < 0.005; Tables 4 and 5). The number of species classified as herbivores, predators, and parasites was significantly different from those classified as detritivores (upper comparison; Table 7); both herbivores and parasites were separated from predators and detritivores (lower comparison; Table 7). As indicated by the ratio of species, this resulted from a much higher abundance of predators and detritivores in the coastal compared to the central prairie.
|Table 7. Maximally nonsignificant subsets of trophic categories between central and coastal prairie. Detritivores (D) were excluded from remaining categories in upper set, reflecting a greater proportion of this trophic level in coastal prairie. Both predators (P) and detritivores were separated from herbivores (H) and parasites (Pa) in the second comparison reflecting a greater proportion of the former in coastal prairie.|
|Trophic Category||Central Prairie||Coastal Prairie|
|H, P, Pa||474||598|