Monitoring for cereal aphids in winter wheat

Abstract

A simple monitoring method for cereal aphids is described which diminishes observation time and can be used by farmers themselves. By repeated observations the farmers can follow the population upsurge of the aphids. A description of the yield decrease -aphid density relation is given which may help to decide whether control measures against cereal aphids are needed.Additional keywords: Metopolophium dirhodum, Sitobion avenae, Rhopalosiphum padi, population dynamics.The yearly rapid increase of cereal aphids requires continuous monitoring. However, continuous monitoring of aphids is a very time consuming procedure and therefore methods which reduce sampling and monitoring activities are urgently needed. Population models which simulate the population curve of the aphids from immigration until the flattening of the curve may help in this. Detailed population models of Sitobion avenae (Rabbinge et al., 1979, Carteret al., in prep.) explain the upsurge of the population and the period of flattening but are still inadequate in explaining completely the collapse of the aphid population. Based on a sensitivity analysis with these explanatory models, simple formulas were developed which are used to produce short term predictions (Rijsdijk et al., in prep.).With these short term predictions of the population density of the aphids it is possible to determine at what time measures are needed to prevent the aphids from reaching density levels at which they may cause economic damage. Definition of the damage threshold (costs of treatment = costs of potential yield loss) requires also knowledge of the level of yield loss and the way it is generated (Vereijken, 1979).It appears that the relation between aphid numbers per ear in the peak and yield loss has a correlation coefficient of 0.75 (Vereijken, 1979). The maximum number of aphids per culm (main shoot and tillers) has a higher correlation coefficient and a slightly lower regression coefficient (r = 0.83; y = 26.1 x + 8.02). Figure 1 based on an analysis of 20 experiments (PAGV, CABO, IPO*; ten Hag, de Vos and Ubels, pers. comm.) during the years [1974][1975][1976][1977][1978][1979] shows that Metopolophium dirhodum and S.avenae may cause considerable yield loss. Situations in which M. dirhodum

Simulation of damage in winter wheat caused by the grain aphid Sitobion avenae. 3. Calculation of damage at various attainable yield levels

“…To test this hypothesis and to compare the accuracy of these simulation-based regression models to the accuracy of empirical regression models for S. avenae damage by Vereijken (1979), Rabbinge and Mantel (1981) and Entwistle and Dixon (1987), the predictions of each model are compared to damage measured in field experiments. An overview of the various models is given in Table 3.…”

“…An overview of the various models is given in Table 3. For the evaluation of the models, data sets from 21 field experiments carried out in the Netherlands are available, 19 of which have been used previously by Rabbinge and Mantel (1981). The two additional data sets have been described by Rossing and Van de Wiel (1990).…”

“…Although care is taken to select independent data sets for model evaluation, no information is available on the data used for calibration of the models 2 and 3. As the performance of these models is evaluated with 21 data sets, 8 of which have been used in their development (Rabbinge and Mantel, 1981), some overestimation of the accuracy of these models is likely.…”

Simulation and systems management in crop protection

“…Generating the populations' distribution; and 5. Calculating the resulting parasitation rate.Choice of a distribution (1) Field data on S. avenae, when compared with the results predicted by the Poisson, Negative Binomial, and Neyman A distributions, have been shown in most cases to be best described by the Negative Binomial Distribution (Rabbinge & Mantel, 1981). The model, therefore, assumes that this is the appropriate form.…”

“…This has been shown to describe the mean-variance relation for a wide range of species (Taylor et al, 1978) Alternatively, k can be estimated using Equation 53, by substituting a descriptive relation between m and P 0 (e.g. those found by Rabbinge & Mantel, 1981;Nachman, 1981). The simplest method, which will be used here, relies on a regression of the maximum likelihood estimates of k on the mean density.…”