Abstract:

In this paper, ground-dwelling beetle communities (GDBCs) from 32 sampling sites with five main vegetation types, two well-conserved ones and three degradated ones, were investigated in Yeyahu Wetland of Beijing by pitfall traps from April to December, 2007, and the influence of wetland degradation and vegetation variation on the composition of GDBCs was studied at the family level. A total of 42 families were collected in the whole sampling period. Among these, Carabidae and Staphylinidae were dominant, while Anthicidae, Eumolpidae and Lathridiidae were subdominant. In the five types of vegetation investigated, the activity density, family level richness and the Shannon-Wiener diversity index (H') of GDBCs were not significantly different between the well-conserved Phragmites communis and Cyperus glomeratus ones. However, the beetle activity density in these two well-conserved vegetations, and the family richness in the Cyperus glomeratus one were significantly higher than those of the three degradated ones. It was showed by canonical correspondence analysis (CCA) to the relationship between the composition of GDBCs and nine environmental factors, that there was an obviously correspondance between the distribution of the 32 sampling sites in CCA ordination diagram and the vegetation types, and soil water content, plant coverage, plant biomass, and litter coverage were the major factors affecting on the composition and spatial distribution of GDBCs in the wetland. According to the correlation analysis, we also found that the correlation among beetle activity density and the soil water content, plant biomass and coverage was all significantly positive (P<0.05 or P<0.01), and it was between beetle family richness and plant biomass (P<0.05); while that between the beetle diversity index (H') and plant coverage was significantly negative (P<0.01). Moreover, in a nonlinear regression analysis, the change of soil water content could explain about 57% of the variation in beetle activity density. A synthetic environmental variable, WBC (Water-Biomass-Coverage), which reflects the situation of soil water content, plant biomass and coverage, was obtained by principal component analysis (PCA) of the nine environmental factors. On the basis of relationship between beetle activity density and WBC, the five vegetation types could be divided into three extremely different groups. Our results showed that maintaining appropriate wetland landscapes is vital for beetle protection.

Key words: vegetation type, wetland, canonical correspondence analysis, principal component analysis