Simulación de paisajes agrícolas para estudiar las dinámicas poblacionales de plagas y fauna auxiliar

Abstract

[ES] El hongo Ampelomyces quisqualis es un parásito del oidio del tomate Oidium neolycopersici. En este estudio, hemos demostrado que su efecto protector sobre el tomate puede no requerir el mantenimiento de un alto nivel de humedad relativa. Este resultado se observó para dos niveles de temperatura: 21°C y 28°C. Esto contrasta con la literatura publicada sobre el uso de este micoparásito contra el oídio en otras especies de plantas y con algunos resultados del presente estudio que muestran la necesidad de una humedad relativa del aire del 100 % para la germinación in vitro de esporas de A. quisqualis. La temperatura también jugó un papel importante en el desarrollo de A. quisqualis y O. neolycopersici. In vitro, ambos germinaron mejor en un rango de temperatura de 15-25°C. Su germinación se redujo significativamente a 30°C y se suprimió totalmente a 35°C. En plantas de tomate, A. quisqualis tuvo un efecto protector significativo contra O. neolycopersici incluso a altas temperaturas (28°C). El presente estudio también demostró que el tratamiento de plantas con A. quisqualis no redujo la agresividad del inóculo secundario de O. neolycopersici.

[EN] Crop pests remain a constant threat to global food security despite of the intensive measures of crop protection. Loss of habitat heterogeneity because of landscape simplification and excessive use of external fossil and agrochemical inputs has partially eliminated the ecological functions provided by communities of potentially beneficial organisms. So, it is crucial to understand on-farm management and composition of landscapes surrounding agricultural fields, which strongly impact biological control. Up to now, most studies in the literature are limited to particular fields and often they do not consider the spatial and temporal heterogeneity of landscape at a larger scale. In this context, we propose to develop a landscape model that allows us to study ecological characteristics linked to landscape composition and organization at a large spatial scale. This generative stochastic landscape model allows simulating large numbers of landscapes based on various predefined configurations of parameters controlling habitat quality. These simulations are used as an input to run a spatially explicit population dynamics model. We use the combined model to study how the variations in landscape structure lead to fluctuations in pest and auxiliary population densities depending on large-scale habitat heterogeneity, which ultimately affects biological control service. The particular pest-predator system that has motivated this theoretical study is composed of the codling moth pest and its predators in apple orchards, but our model is generic enough to be adaptable to and interpretable for other insect species. Based on an extensive simulation study, we found that spatial heterogeneity and landscape structure have a strong effect on pest-predator population dynamics. Integration of such concepts and knowledge into agricultural land-use management can pave the way to modern, productive and environmentally friendly crop production systems.