农业景观格局对玉米害虫-捕食性天敌定性食物网结构的影响

农业过度集约化带来的农业景观均质化已成为农田生物多样性降低的主要原因之一。为研究农业景观格局对害虫-捕食性天敌定性食物网结构的影响,本文选择辽宁省昌图县为研究区,选取8个典型田块为样区,在分析定性食物网结构的基础上,采用回归分析和最优模型确定食物网参数与景观指数之间的关系。结果表明:1)互作丰度(IR)与各景观指数无显著相关性。2)连接密度(LD)与蔓延度指数(CONTAG, x1)、聚集度指数(AI, x2)呈显著正相关(P0.05),最优模型为:LD=-64.621+0.780x1+0.739x2。农业景观中非耕作斑块越聚集,玉米害虫-捕食性天敌定性食物网结构越复杂。3)连接性(C)与CONTAG(x1)、香农多样性指数(SHDI, x3)呈显著正相关(P0.05),而与香农均匀度指数(SHEI,x4)呈显著负相关(P0.05),最优模型为:C=-178.500+1.831x1-106.808x4。景观类型越多样,且同类斑块连接度越好,害虫与捕食性天敌的相互作用越频繁,也越有利于复杂食物网结构的维持。4)普遍性(G)与景观形状指数(LSI, x5)、斑块结合指数(COHESION, X7)、AI(x2)呈显著正相关(P0.05),而与斑块密度(PD,x6)呈显著负相关(P0.05),最优模型为:G=-2 994.798+26.891x2+27.090x5-0.491x6+2.851x7。非耕作斑块破碎化程度越低,天敌的搜寻行为和聚集行为越强,越有利于食物网结构的稳定。5)易损性(V)与SHEI(x4)呈显著正相关,而与CONTAG(x1)呈显著负相关(P0.05),最优模型为:V=8.411+5.351x4。斑块类型在景观中分布越均匀,害虫多样性越高,群落结构也越复杂。总体而言,农业景观异质性越强越有利于玉米害虫-捕食性天敌定性食物网的构建和抗干扰性的增强。而利用田间数据构建食物网矩阵的方法可以成为研究如何增强农业景观异质性的有力工具。

Effects of agricultural landscape pattern on qualitative food web structure of corn pest-predatory natural enemies

The homogenization of agricultural landscape caused by excessive agricultural intensification has been one of the main reasons for the reduction of farmland biodiversity. Research on the impact of the agricultural landscape pattern on the qualitative food web structure of pest-predatory natural enemies was conducted in Changtu County, Liaoning Province, where eight typical fields were sampled. Regression analysis and optimal model determined the relationship between food web parameters and landscape indexes. The results showed an insignificant correlation between food web interaction richness (IR) and landscape indexes. However, a significant positive multiple correlations between food web linkage density (LD) and contagion index (CONTAG, x₁) and aggregation index (AI, x₂) were observed. The corresponding optimal model was: LD=-64.621+0.780x₁+0.739x₂. The complexity of the qualitative food web structure of the corn pest-predatory natural enemies was dependent on the degree of concentration of the non-cultivated patches in the agricultural landscape. Furthermore, food web connectance (C) was significantly positively correlated with CONTAG, (x₁) and Shannon diversity index (SHDI, x₃), but was significantly negatively correlated with Shannon evenness index (SHEI, x₄). The corresponding optimal model was: C=-178.500+1.831x₁-106.808x₄; the more diverse the landscape types, the better the connectivity of the same patches; the more frequent the interaction between pests and predatory natural enemies, the more beneficial it is to maintain the complex food web structure. Food web generality (G) was significantly positively correlated with landscape shape index (LSI, x₅), cohesion index (COHESION, x₇), and AI (x₂); however, it was significantly negatively correlated with patch density (PD, x₆). The corresponding optimal model was: G=-2 994.798+26.891x₂+27.090x₅-0.491x₆+2.851x₇; the lower the degree of non-cultivated patch fragmentation, the stronger the search and aggregation behavior of natural enemies, which is beneficial and increases the stability of the food web structure. Food web vulnerability (V) was significantly positively correlated with the SHEI (x₄), but was significantly negatively correlated with CONTAG (x₁). The corresponding optimal model was: V=8.411+5.351x₄; the more evenly distributed patch types in the landscape, the higher the pest diversity and the increased complexity of the community structure. In general, the construction of the qualitative food web of corn pest-predatory natural enemies and the enhancement of anti-interference largely depends on the strength of the heterogeneity of the agricultural landscape. The use of field data in the construction of a food web matrix is a method that can be a powerful resource for studying ways to enhance the heterogeneity of agricultural landscapes.