摘要：The spatial distribution of crop and non-crop habitats over segmented agricultural landscapes could be used as a means to reduce insect pest populations. Seven land cover categories such as wheat, rapeseed, vegetable, water, built-up, paved road, and unsurfaced road were extracted from GeoEye satellite images dating from late May to late June of 2010. Three diversity metrics and three evenness metrics were estimated from the abovementioned land cover categories for quantifying the effect of landscape composition on nymphal and adult Laodelphax striatellus Fallén. The degree of correlation between the proportion of crop cover and adjacent spatial scales ( r : 0.651–0.983) was higher than the correlation between the proportion of crop cover and nonadjacent spatial scales ( r : −0.255–0.896). While the degree of correlation between diversity indices and abundance of L. striatellus decreased gradually when the spatial scales varied from large (>100 m radius buffer) to small (<100 m). Our study suggests that when using natural biological pest control and ecological engineering practices in the rural-urban fringes, the crop field’s width should be less than 200 m and increasing vegetation diversity within such a scale will be helpful to regulate the insect pests under a certain density.
关键词：paddy rice; small brown planthopper; satellite remote sensing; landscape metrics; fragmented agricultural landscapes
