A Simulation Approach to Assessing Sampling Strategies for Insect Pests: An Example with the Balsam Gall Midge

R Drew Carleton,Peter J Silk,Stephen B Heard,Frederic Amblard

doi:10.1371/journal.pone.0082618

Abstract

Estimation of pest density is a basic requirement for integrated pest management in agriculture and forestry, and efficiency in density estimation is a common goal. Sequential sampling techniques promise efficient sampling, but their application can involve cumbersome mathematics and/or intensive warm-up sampling when pests have complex within- or between-site distributions. We provide tools for assessing the efficiency of sequential sampling and of alternative, simpler sampling plans, using computer simulation with “pre-sampling” data. We illustrate our approach using data for balsam gall midge (Paradiplosis tumifex) attack in Christmas tree farms. Paradiplosis tumifex proved recalcitrant to sequential sampling techniques. Midge distributions could not be fit by a common negative binomial distribution across sites. Local parameterization, using warm-up samples to estimate the clumping parameter k for each site, performed poorly: k estimates were unreliable even for samples of n∼100 trees. These methods were further confounded by significant within-site spatial autocorrelation. Much simpler sampling schemes, involving random or belt-transect sampling to preset sample sizes, were effective and efficient for P. tumifex. Sampling via belt transects (through the longest dimension of a stand) was the most efficient, with sample means converging on true mean density for sample sizes of n∼25–40 trees. Pre-sampling and simulation techniques provide a simple method for assessing sampling strategies for estimating insect infestation. We suspect that many pests will resemble P. tumifex in challenging the assumptions of sequential sampling methods. Our software will allow practitioners to optimize sampling strategies before they are brought to real-world applications, while potentially avoiding the need for the cumbersome calculations required for sequential sampling methods.

Highlights

Insects and other pests are responsible for enormous financial and production losses in agriculture and forestry
Perhaps the most basic requirement for any pest management program is the availability of a sampling method for assessing the level of infestation
Study system: Balsam gall midge in Christmas tree stands In eastern Canada, the sale of Christmas tree and wreath products from Abies balsamea (L.) Mill. is a multimillion dollar industry, with trees shipped to markets throughout the western hemisphere [23]

Summary

Introduction

Insects and other pests are responsible for enormous financial and production losses in agriculture and forestry. Estimating insect densities in the field is far from a simple task, and it involves decisions about when to sample during host or insect phenology (e.g., [1]), what to sample (quadrat, whole plant, appropriate organ, or representative module; e.g., [2]), and which and how many plants, or other sampling units, to sample from the large number available at a site This last decision in particular has spawned an enormous literature [3], with thousands of idiosyncratic recommendations for different systems but with a simple underlying truth: in general, more accurate estimation is achieved by including more samples and selecting them in more sophisticated ways; but doing so requires more time, money and labour. Achieving the most accurate estimates from the smallest investment of effort can involve ingenuity in field technique (e.g., [4]), but great returns can come from the development of statistical methods for handling sampling data and for evaluating the efficiency of alternative sampling designs (e.g., [5,6,7])

Objectives

Methods

Results

Conclusion