Delineation of site-specific management zones for pest control purposes: Exploring precision agriculture and species distribution modeling approaches

Abstract

Delineation of management zones is one key feature of precision agriculture (PA), an agricultural management strategy which aims to maximize productivity and sustainability of crops by making optimum use of inputs. Common means of achieving this include the delineation of site-specific management zones (SSMZ), which are subregions of crop fields where yield-limiting factors manifest homogeneously. For pest management, however, PAs standard approach to crop field zoning (e.g. mapping zoning factors, mapping responses, filtering zoning factors, field partitioning, validation of SSMZ) is difficult to implement. In this case, high dynamism of pest populations makes mapping zoning factors and response variables complex and resource-consuming endeavors. Moreover, partitioning crop fields for pest management purposes requires consideration of temporally dynamic zoning factors instead of stable ones. In this paper we propose the addition of an ecological layer to classic PA zoning approaches, in order to overcome obstacles to delineating SSMZ for pest management purposes. In particular, we explore the potential of precision sampling tools (unmanned aerial vehicles, Arduino data loggers) and species distribution modeling (SDM) as facilitators of such an ecological dimension. After contrasting two PA zoning methods against one novel SDM-based approach (described here) to partition a commercial orchard (9 ha, Citrus latifolia Tanaka) in terms of six virtual pests, results show several advantages in our SDM-based method and the occasional necessity of multi approach methodologies to assure adequate zone delimitation under particular infestation scenarios.