A Modelling Approach to Describe the Anthonomus eugenii (Coleoptera: Curculionidae) Life Cycle in Plant Protection: A Priori and a Posteriori Analysis

Abstract

Anthonomus eugenii Cano (Coleoptera: Curculionidae) is one of the most severe pests for sweet and hot varieties of pepper (Capsicum spp.; Solanaceae). The species is distributed widely, principally in Central America, but in 2013 it was detected for the first time in the Lazio region of Italy. Modelling plays a key role in reducing chemical treatments used on Capsicum spp., but reliable predictions of pest populations require adjusted tools, as well as intense knowledge of the insect's biology and its typical environment. The main goal of this work is to describe the life cycle of A. eugenii with a physiologically based model, which links the population dynamics with the environmental parameters. More specifically, this analysis focuses on the different response of the age-structured model in relation to the development rate function in input. Two methodologies to determine the best representative development rate function suitable for simulations are proposed; the first is "a priori analysis," whereas the second is the "a posteriori analysis." Simulations were compared with semi-field data, collected in a controlled experimental greenhouse where A. eugenii developed in varying temperature conditions. Results showed that the model used is adequate to describe A. eugenii population dynamics and highlighted how the a posteriori analysis can be essential to (i) analyze the simulation outputs, and (ii) determine the best representing development rate function, if the a priori analysis does not provide this information sufficiently clearly.

Modelling Drosophila suzukii Adult Male Populations: A Physiologically Based Approach with Validation

“…A high χ 2 value does not always imply a lower reliability of the simulation, because it also depends on the abundance of both experimental and simulated populations. Accordingly, even small differences among high values can provide a higher χ 2 [28]. This was the case in the year 2019 in the Montelibretti field, where the highest population density among the three years of survey was observed, and the corresponding R 2 value indicates that the simulation trend was well described.…”

“…With this precondition, let us introduce the physiologically based model applied to describe D. suzukii populations developing over time t and through the life stages x. The population density function, hence, is represented by a two-variables function N(t, x), solution of the following first-order partial differential equation [25,26,28]:…”

“…The comparison was conducted using the χ 2 function and the coefficient of determination R 2 , since this methodology was successfully applied to provide a clear idea of the level of overlap between field data and simulations with the Equation (1) [25,26,28,30,32].…”

“…More specifically, data of this type are called "life tables", and represent a powerful tool for knowing more about the biology of the insects [44]. The life tables published by Tochen et al [12] were used to estimate the best fit function parameters through the Minuit algorithm contained in the ROOT software [45], as described by Rossini et al [26,28].…”

“…Once the best fit parameters were available, the second step of the model application was the choice of the best representing expression for the life tables data through the "a priori analysis" proposed by Rossini et al [26,28,30]. The method provides for the choice of the expression which reported the best result from the χ 2 test and an R 2 value close to 1.…”