Natural enemy abundance in a crop plot depends on its prey presence and also influenced by habitats close to field. Landscape changes are also important factors driving pest and natural enemy population abundance in a specific crop field. Examining these kinds of effects on insect pests or biocontrol agents, as well as analysis of their functional food webs, would be asset to make a fruitful pest management programme at local scales. Therefore, this study was undertaken to evaluate the impact of surrounding landscape on the abundance insect predators in rice field. This study revealed a dependency on rice bund margin width, with spider populations increasing with increased bund widths. Conversely, population abundance did not rely on the number of weed species observed on earthen ridge around the rice field. In general, relative abundances of predator populations differed significantly across the three landscapes tested. Among the four predators of rice insect pest, the green mirid bug showed highest number irrespective of landscape. Comparatively, higher predator diversity (Shanon diversity) was observed in landscape I followed by landscape III and landscape II. All landscapes showed different diversity indices indicating heterozygosity existed in each study site. These landscape diversity influences the predator’s abundances at a local scale. Variogram derived from this study also indicated the landscape heterozygosity existed in studied locations which can also explain the predator’s abundances in rice field at locale scale. These findings suggest that predators of rice insect pests are landscape specific. Therefore, characterization of each local landscape in Bangladesh rice production landscapes are necessary before planning and implementation of integrated pest management. Geospatial analysis of local landscape would be more effective to analyze such unique characteristics. As a step in this direction, preliminary variography analyses using the RED spectral band of December 2016 LANDSAT 8 imagery propose an initial learning suite of methods for describing useful local characteristics affecting rice pest predators.

中文翻译：

---

周围景观影响稻田中昆虫捕食者的数量

作物田间天敌的丰度取决于其猎物的存在，并且还受田野附近栖息地的影响。景观变化也是驱动特定作物田中害虫和天敌种群数量增加的重要因素。检查这些对害虫或生物防治剂的影响，以及对其功能性食物网的分析，对于在地方范围内制定富有成效的害虫管理计划将是有益的。因此，本研究旨在评估周围景观对稻田中大量昆虫捕食者的影响。这项研究揭示了对水稻外缘边缘宽度的依赖性，蜘蛛种群随着外缘宽度的增加而增加。相反，人口丰度并不取决于在稻田周围土上观察到的杂草种类的数量。一般来说，在所测试的三个景观中，捕食者种群的相对丰度差异显着。在水稻害虫的四种捕食者中，无论景观如何，绿色的虫虫数量最多。相比之下，在景观I中观察到更高的捕食者多样性（Shanon多样性），随后是景观III和景观II。所有景观均显示不同的多样性指数，表明每个研究地点均存在杂合性。这些景观多样性在局部范围内影响捕食者的丰度。这项研究得出的方差图也表明了研究地点的景观杂合性，这也可以解释区域规模下稻田中捕食者的丰度。这些发现表明，水稻害虫的天敌是特定于景观的。因此，在计划和实施病虫害综合治理之前，有必要对孟加拉国稻米生产景观中的每个局部景观进行特征描述。对局部景观进行地理空间分析将更有效地分析此类独特特征。作为朝这个方向迈出的一步，使用2016年12月LANDSAT 8影像的RED光谱带进行的初步变异学分析提出了一套初始学习方法，用于描述影响水稻害虫天敌的有用局部特征。