USGS Home
Contact USGS
Search USGS

Figure 2.  Average total number of ground beetles collected in pitfall traps from 9 April to 28 October 1996 during each biweekly sampling period from Anderson Prairie.

We collected a total of 468 ground beetles in plots burned in 1994, an average of 0.43 beetles per trap per day. Plots burned in 1995 yielded 718 ground beetles, an average catch of 0.59 beetles per trap per day. Plots burned in 1996 produced 586 ground beetles, an average of 0.52 beetles per trap per day (Table 1). Even though there were noticeable differences among the burn treatments, neither numbers of individual ground beetles (F=0.68, df=2, 8, p>0.05) nor beetles captured per trap per day (F=1.57, df=2, 8, p>0.05) were significantly different.

The overall species richness in the three burn treatments decreased significantly with increasing time since burning from 29 species found in plots burned in 1994, to 32 in plots burned in 1995, to 38 species in plots burned in 1996. By early August, all species collected from the burned 1994 and 1995 treatments had been detected, while new species continued to accumulate until the beginning of October in the burned 1996 treatment (Fig. 3). Species accumulation over time for the three burn treatments showed the burn 1996 treatment had significantly greater species over time than the other treatments (Friedman test S=17.42, df=2, p<0.05).

Figure 3.  Species accumulation curves from April to October 1996 for ground beetles collected from within each burn treatment in Anderson Prairie.

In contrast to increased species richness in more recently burned plots, species diversity (H') increased with time since burning, with the most diverse carabid fauna in plots burned in 1994, and the least diverse fauna in plots burned in 1996 (Table 1). This trend in species diversity is likely due to the more even distribution of beetles among species from plots burned in 1994 than 1995 and 1996, and is evidenced by the evenness (J) calculated for each treatment (Table 1), which also increased with time since burning. The beetle assemblages of the burned 1995 and 1996 plots were most similar (80.3%), with the burned 1994 and 1995 plots 70.7% similar. Least similar were beetle assemblages in the burned 1994 and 1996 plots, with a similarity of only 56.6%.

Trends shown by the short-term samples, i.e., three discrete one-week samples taken during the peak beetle activity period between late-June and early-August, were consistent with full-season biweekly sampling. The numbers of beetles collected showed the same increasing trend as the full year data (Fig. 4). However, the burned 1996 treatment did possess fewer species during the late-June and early-August activity peak than the other treatments (Fig. 5), a trend that differed from the biweekly full-season results. Shannon's diversity index (H') for the discrete peak activity period increased from 1.43 in the burned 1996 treatment to 1.74 in the burned 1994 treatment, a trend similar to the continuous full-season results (Table 1). Likewise, evenness (J) increased from 0.52 in the burned 1996 treatment to 0.57 in the burned 1994 treatment, similar to the full-season results (Table 1).

Figure 4.  Abundance of ground beetles collected in pitfall traps from treatments in Anderson Prairie burned in 1994, 1995, or 1996. Beetles were collected by either short-term sampling consisting of three one-week samples during the peak beetle activity period, or bi-weekly from April to October 1996.

Figure 5.  Species richness of ground beetles in Anderson Prairie collected by pitfall traps from treatments burned in 1994, 1995, or 1996. Beetles were collected by either short-term sampling consisting of three one-week samples during the peak beetle activity period, or bi-weekly from April to October 1996.