Zandt H S
Zoologisches Institut, Rheinsprung 9, CH-4051, Basel, Switzerland.
Oecologia. 1994 Apr;97(3):399-406. doi: 10.1007/BF00317331.
Three sampling techniques commonly used to estimate the population size of caterpillars and sawfly larvae in trees (branch samples, frass production, water basins), were compared with respect to sampling error and economic costs. At the level of tree populations (e.g. forests), on an arbitrary date, the mean caterpillar intensity per tree (expressed in numbers of larvae or their biomass per 100 shoots) was predicted from the mean frass production per tree (expressed in mg frass per m forest floor per day). At the level of the single tree, the frass production on an arbitrary date was related to the population intensity, but, due to the large sampling error, did not provide an accurate prediction. Summing the frass produced over the whole season reduced this error and predicted the seasonal abundance of larvae in single trees, estimated as their maximum intensity or their density (numbers of larvae or their biomass per m forest floor). The maximum population intensity was not related to the population density. The sampling techniques suffer from large errors unrelated to larval abundance. The main sources of error (i.e. weather or predation of the larvae) usually cause an underestimation of population size. Labour, the main cause of high costs, was low in the basin technique and high in the frass production technique. Possible ways of reducing errors and applications of the three techniques are discussed.
比较了常用于估计树上毛虫和叶蜂幼虫种群数量的三种抽样技术（枝条样本、粪便产量、水盆法）在抽样误差和经济成本方面的差异。在树木种群层面（如森林），在任意日期，根据每棵树的平均粪便产量（以每天每平方米林地的粪便毫克数表示）预测每棵树的平均毛虫强度（以每100个嫩枝上的幼虫数量或其生物量表示）。在单棵树层面，任意日期的粪便产量与种群强度相关，但由于抽样误差较大，无法提供准确预测。将整个季节产生的粪便量相加可减少这种误差，并预测单棵树中幼虫的季节性丰度，以其最大强度或密度（每平方米林地的幼虫数量或其生物量）来估计。最大种群强度与种群密度无关。这些抽样技术存在与幼虫丰度无关的较大误差。主要误差来源（如天气或幼虫被捕食）通常会导致种群数量估计偏低。劳动力是成本高昂的主要原因，在水盆法中劳动力成本较低，而在粪便产量法中劳动力成本较高。讨论了减少误差的可能方法以及这三种技术的应用。
