Nansen Christian, Meikle William G, Campbell James, Phillips Thomas W, Subramanyam Bhadriraju
Department of Entomology, AgriLife Research, Texas A&M University, 1102 E FM 1294 Lubbock, TX 79403-6603, USA.
J Econ Entomol. 2008 Dec;101(6):1719-28. doi: 10.1603/0022-0493-101.6.1719.
Traps for monitoring of flying insect pests constitute a critical part of integrated pest management strategies. However, interpretation of trap captures is hampered by 1) factors associated with the performance of traps (i.e., lure, trap design, placement); 2) an often poorly defined relationship between trap captures and population density; and 3) interpretation approaches being highly specific to a certain insect species, trapping method, or trapping environment. The main purpose of this study was to identify a trap capture interpretation approach with little sensitivity to characteristics specific to a given data set, which would allow easier comparison of trapping data sets and make it easier to standardize sampling plans across insect pests and trapping environments. Based on fits of trapping data sets to standard distributions (normal, Poisson, and negative binomial), evaluations of the index of aggregation, k, and linear regression coefficients from Taylor's power law, we concluded that these characteristics varied considerably among data sets, which means that enumerative sampling plans may not be appropriate. Across 13 trapping data sets of six insect species, we showed a consistent nonlinear relationship between average trap captures and number of traps with zero captures and that the k can be stabilized by converting trapping data into binomial data. A trap interpretation approach based on number of zero captures is both easy to use, was found to be species-independent, and means that it may be possible to establish meaningful and reliable action thresholds based on trap captures of flying insects. Although developed using trapping data from food facilities, this approach may have application to trapping data from other environments as well.
用于监测飞行害虫的诱捕器是综合害虫管理策略的关键组成部分。然而，诱捕器捕获量的解读受到以下因素的阻碍：1）与诱捕器性能相关的因素（即诱饵、诱捕器设计、放置位置）；2）诱捕器捕获量与种群密度之间通常定义不明确的关系；3）解读方法高度特定于某一昆虫物种、诱捕方法或诱捕环境。本研究的主要目的是确定一种对给定数据集的特定特征不太敏感的诱捕器捕获量解读方法，这将便于比较诱捕数据集，并使跨害虫和诱捕环境的抽样计划更容易标准化。基于诱捕数据集与标准分布（正态分布、泊松分布和负二项分布）的拟合、聚集度指数k的评估以及泰勒幂法则的线性回归系数，我们得出结论，这些特征在数据集之间差异很大，这意味着枚举抽样计划可能不合适。在六种昆虫的13个诱捕数据集上，我们发现平均诱捕器捕获量与零捕获诱捕器数量之间存在一致的非线性关系，并且通过将诱捕数据转换为二项数据可以使k稳定下来。基于零捕获数量的诱捕器解读方法既易于使用，被发现与物种无关，这意味着有可能根据飞行昆虫的诱捕器捕获量建立有意义且可靠的行动阈值。尽管该方法是利用食品设施的诱捕数据开发的，但这种方法也可能适用于其他环境的诱捕数据。
