A real-time decision-making tool based on dynamic thresholds for Phthorimaea absoluta management in greenhouse tomato

Abstract

The onset of digital information and computing power brings the opportunity to optimize the use of agricultural inputs and favors the transit to an information-based agriculture. The tomato leafminer (TLM), Phthorimaea absoluta, is a major pest of greenhouse tomatoes, and its control relies heavily on chemical pesticides. This study presents a decision-making system for TLM that dynamically adjusts decision thresholds throughout the crop cycle. Through a controlled experiment, we obtained pest damage curves for three crop phenological stages, which were used to derive a set of non-linear functions that estimates a decision threshold expressed as the maximum tolerable number of larvae/plant for any given crop age. We show that produce losses due to TLM are far from constant across the crop cycle, and that such variation can be modeled as a function of the effect of TLM on yield, cumulative control costs, potential yield, and produce prices. The decision threshold function was linked to a previously validated sampling plan, which was adjusted to work in serial bouts in frequencies that vary according to sampling results and pest growth rates. The decision-making system is presented as a web-based app that delivers a management recommendation, the timing for the next sampling bout, and an indication of how close the population is from the threshold. This study presents a step forward towards a modern application of the economic injury level concept by adapting pest management decision criteria to the changing ecological and economic context of each production system.