摘要
The soil insect community was studied in grey desert soil district in September 2004. 90 soil samples and 100 pitfalls were collected from 10 treatments, i.e., abandonment （Aband.）, CK, N, NP, NK, PK, NPK, MNPK （fertilizer N：organic N = 3：7）, 1.5MNPK, and SNPK. 4915 soil insects （128 unknown）, as individuals belonging to 9 orders and 33 families, were obtained by pitfall traps and modified Tullgren methods. The results showed that, based on the number of individuals and groups, the macro fauna in total reached their peaks in abandonment, whereas meso and micro fauna in N and PK, respectively. Of the 10 treatments, the most dominant of soil insect composition was in MNPK and most evenness was N. The result by Kruskal-Wallis test indicated that the distribution of the arable soil insect was significantly impacted by different fertilizer treatments （X0.05（9） = 23.38, P 〈 0.005）, and soil insect group of the abandonment was significantly different from that of other fertilizer treatments. The soil insect community was divided into five groups by non-metric- MDS analysis： （1） NPK, MNPK, 1.5MNPK, CK, （2） NP and PK, （3） NK and N, （4） SNPK, and （5） abandonment, which indicated that distribution of soil insect was related to the character of the fertilizer. In the principal component analysis, two factors explained 98.51% of the total variation among the 10 treatments, and the factor one explained that N and SNPK positively affected soil insect community, whereas factor two explained that 1.5MNPK positively affected soil insect community, which showed that the diversified fertilizer did not evenly affect the soil insect community.
The soil insect community was studied in grey desert soil district in September 2004. 90 soil samples and 100 pitfalls were collected from 10 treatments, i.e., abandonment （Aband.）, CK, N, NP, NK, PK, NPK, MNPK （fertilizer N：organic N = 3：7）, 1.5MNPK, and SNPK. 4915 soil insects （128 unknown）, as individuals belonging to 9 orders and 33 families, were obtained by pitfall traps and modified Tullgren methods. The results showed that, based on the number of individuals and groups, the macro fauna in total reached their peaks in abandonment, whereas meso and micro fauna in N and PK, respectively. Of the 10 treatments, the most dominant of soil insect composition was in MNPK and most evenness was N. The result by Kruskal-Wallis test indicated that the distribution of the arable soil insect was significantly impacted by different fertilizer treatments （X0.05（9） = 23.38, P 〈 0.005）, and soil insect group of the abandonment was significantly different from that of other fertilizer treatments. The soil insect community was divided into five groups by non-metric- MDS analysis： （1） NPK, MNPK, 1.5MNPK, CK, （2） NP and PK, （3） NK and N, （4） SNPK, and （5） abandonment, which indicated that distribution of soil insect was related to the character of the fertilizer. In the principal component analysis, two factors explained 98.51% of the total variation among the 10 treatments, and the factor one explained that N and SNPK positively affected soil insect community, whereas factor two explained that 1.5MNPK positively affected soil insect community, which showed that the diversified fertilizer did not evenly affect the soil insect community.
