Weed resistance to different herbicide modes of action is driven by agricultural intensification

Abstract

Context or problem: The continued increase in numbers of herbicide-resistant weed species in field crops constrains sustainable agricultural practices worldwide. Countries differ markedly in numbers of herbicide-resistant weed species in field crops yet the extent this reflects global variation in agricultural intensification is not known. Objective or research question: To what extent does the global variation in the number of herbicide resistant weeds reflect differences in the magnitude of direct measures of agricultural intensification such as agrochemical inputs, indirect measures such as per capita GDP, or non-agronomic factors such as research intensity? Methods: Best-subset regression analysis quantified whether national scale estimates of agricultural intensification such as per capita GDP, cropland area, as well as inputs of N-fertilizer and herbicide explained variation in numbers of herbicide-resistant weed species in field crops worldwide. The number of publications addressing herbicides was used as a proxy for sampling effort, while the time since the first record of resistance estimated the window of opportunity for weed species to become resistant. Analyses were undertaken across all herbicides as well as separately for four herbicide modes of action (ACCase, ALS, PSII and EPSPS inhibitors). Results: Over 70% of the global variation in numbers of herbicide-resistant weed species was explained by national scale estimates of herbicide inputs, cropland area, per capita GDP as well as measures of research effort and the time since the first resistant weed was recorded. The explanatory ability of models for individual herbicide modes of action ranged from 40% (ACCase inhibitors) to 68% (PSII inhibitors), with per capita GDP and time since first record the most consistent explanatory variables. Conclusions: Agricultural intensification, as captured by herbicide inputs and per capita GDP, was associated with increased numbers of herbicide-resistant weed species worldwide but the limited herbicide expertise in many countries means the scale of the problem is underestimated. The number of resistant weed species depended on how long resistance had been observed suggesting for many countries the problem will increase in the future, especially as different modes of action become more widely used. Implications or significance: Many countries that have only recently recorded herbicide resistance in weeds are already on a trajectory for future increases but have limited capability to address this problem. Implementation of integrated weed management strategies to reduce the risk of herbicide-resistant weeds evolving should therefore be implemented proactively before the problem gets worse rather than reactively as has occurred in other countries.