Northern Prairie Wildlife Research Center
We identified substantial variation in structure of the insect community between the two prairie communities studied in Texas. There were major differences in abundance of taxa, e.g., the central prairie had a higher proportion of Hymenoptera and Orthoptera and a lower proportion of Homoptera and Coleoptera, and occurrence of unique species, e.g., higher proportion of unique species in the central prairie (Tables 1 and 2). These differences led to the discovery that there was a marked structure to the habitats on the coastal prairie (Fig. 1), whereas the habitats in the central prairie had a simple, hierarchical structure (Fig. 2). Nonetheless, clustering of habitats within both areas demonstrated similar contrasts among constituent habitat types. The major clusters in the coastal prairie, for example, contrasted woodland and mixed woodland/prairie sites (Sites 3, 4, 5, 7) with open sites, which were either riparian (Sites 1 and 2) or prairie (Sites 6 and 8). In the central prairie, sites were again clustered by the degree of herbaceous open vegetation (Sites 1, 2H, and 3H) or woodland (Sites 2W and 3W).
In spite of the taxonomic differences between the two areas and the difference in the hierarchical arrangements of habitats, diversity of insect communities was closely tied to species diversity and structure of the plant community in both areas, specifically to the structure resulting from the presence of woody vegetation. This relationship between vegetation structure and insect diversity has been found for other communities (Strong et al. 1984). We considered this assertion for Texas prairies by separately analyzing insects that were sampled from woody components of habitats in the central prairie from those sampled from herbaceous components. The results clearly substantiated that, as vertical woody structure was added to the plant community, a nearly independent component of insect fauna was added, thereby increasing overall species richness (Table 2). A similar conclusion was reached when comparing a single habitat, native prairie, between the two areas. Higher diversity was attained in the native prairie habitat (Site 1) in the central prairie than in the coastal prairie (Site 8) because prairies in the former area contained a heavy component of dicot vegetation (see description of sites in Methods).
The coastal prairie community contained 58% herbivores, 13% parasites, 19% detritivores, and 9.5% predators (Table 4), whereas the proportion of herbivores (65%) in the central prairie was higher with lower representation of the other trophic levels (10% predators, 17% parasites, and 9% detritivores; Table 5). In addition, a higher proportion of detritivores in woodlands led to among-habitat differences in trophic structure in the central prairie. The question as to whether proportional representation among guilds in communities is constant was addressed when Heatwole and Levins (1972) analyzed data on island re-colonization collected by Simberloff and Wilson (1969). Heatwole and Levins (1972) concluded that trophic structure was stable because apportionment of species among trophic levels was similar before and after defaunation. Moran and Southwood (1982) found that the proportional representation in guilds of insects on six species of forest trees in South Africa and Britain was constant in both countries. Studying early succession oldfields in the United States and Britain, Hendrix et al. (1988) also reported constancy in guild structure of arthropods and found greater constancy between years within a site than between countries. Our results illustrate significant alterations in both taxonomic and trophic structure between the two prairie ecosystems studied in Texas.
Species richness was greater in the coastal prairie (700 species) than the central prairie (501 species; Tables 1 and 2). We do not think that this difference reflects a difference between areas in the number of sites sampled because the number of sites was based on the number of different types of habitat; hence, we sampled the complete diversity of habitats, and the constituent assemblage of insect species at each study area. Similarity among sites, however, was higher among sites in the coastal prairie than in the central prairie. Differences in similarity indicated less regional habitat heterogeneity in coastal prairie caused by a smaller difference between wooded and herbaceous sites, and higher similarity among herbaceous sites than in the central prairie. Furthermore, lower among-habitat heterogeneity in the coastal prairie was reflected by a lower (27%) average number of unique species compared to the central prairie (49%; Tables 1 and 2). Finally, higher heterogeneity in the central prairie leads to significant shifts in taxonomic and trophic structure among study sites compared to the coastal prairie.
Several factors could account for such striking differences in the ecological structure between the coastal and central prairies. Aboveground net primary productivity (ANPP) on the coastal prairie is about 470 g/m² (Harcombe et al. 1993). ANPP for the central prairie can be computed from a regression equation relating ANPP to annual precipitation (APPT) and a scalar of soil waterholding capacity (WHC; Sala et al. 1988): ANPP = 32 + (0.45 * APPT) - (352 * WHC) + (0.95 * WHC * APPT). Using APPT of 817 mm and a low estimate of WHC = 0.20 (Sala et al. 1988), we estimate a productivity of 485 g/m² for the central prairie. Because of the similarity between study areas, ANPP does not appear to explain the taxonomic and trophic differences that were observed between the prairies. However, differences in extent and structure of native prairie may explain these differences. The extent of native prairie (Site 8) in the coastal prairie area was small (about 0.02 ha), and was surrounded by intensive agriculture and cattle grazing. In addition, this site was mowed annually for hay and, as a result, contained a low proportion of dicots. In contrast, native prairie (Site 1) in the central prairie was less isolated, less disturbed, and had a higher abundance of dicots, thereby creating more heterogeneity among sites. In addition, habitat sites in the central prairie were more isolated from each other than those in the coastal prairie. Differences in size and degree of isolation of constituent habitats likely led to observed differences in entomofauna between the coastal and central prairies.
Studies comparing within-ecosystem variation in species composition and community structure are lacking for North American grasslands. Species composition and structure of insect communities varied substantially in prairie communities located less than 300 km apart in the southern extent of the Great Plains. In both prairies, diversity of the entomofauna increased when vertical (woody) vegetation occurred as a component of the prairie community. Average taxonomic similarity was greater (32%) among sites in the coastal prairie than among sites in the central prairie (21%) because among-habitat differences and proportion of unique species per site were higher in the latter area, and trophic structure varied between areas and among habitats in the central prairie, which reflected the greater among-habitat variation in taxonomic structure in the central prairie. Such differences between sites in the larger prairie ecosystem underlie the fact that individualized conservation plans may be required when assessing ways to preserve the remaining fragments of this once widespread ecosystem (Kremen et al. 1993, Golden and Crist 1999).