Global variation in native and non-native richness of weeds in maize is shaped by the length of cereal cultivation and spillover of crop weeds

Abstract

Weeds are a major cause of production losses in arable lands, but the relative contributions of native and non-native weeds vary considerably among countries. Understanding the factors that drive these differences could help improve biosecurity and weed management in arable lands by targeting measures to the unique agricultural contexts of different countries. However, global-scale analyses aimed at understanding the variations in weed richness are rare. We examined the factors associated with native and non-native weed richness in maize (Zea mays) fields worldwide. Using species records from 40 countries, we used random forest analysis to assess the importance of nine explanatory variables, selected a priori, on the richness of native and non-native weed species in maize cultivation. We found that maize harvested area, arable land area and the number of scientific publications on weeds from the region were positively correlated with both native and non-native weed richness. While the length of maize cultivation in a country was weakly positively correlated with both native and non-native weed richness, the length of cereal cultivation was negatively correlated with non-native weed richness. The length of cereal cultivation also had the strongest negative correlation with the percentage of non-native weed species. Our findings suggest that most native weeds in maize fields probably result from ‘spillover’ from other cereal crops cultivated prior to maize cultivation in each country. Consequently, the percentage of non-native weeds is relatively low in countries with a short history of maize cultivation but a long history of other cereal cultivation. In countries with a long history of cereal cultivation, native weeds of other cereal crops (e.g., sorghum and wheat) significantly shape weed assemblages of maize fields. These native weeds, originally associated with other cereal crops, have become introduced worldwide as contaminants of maize cultivation and now occur as non-native weeds in countries with shorter cereal cultivation histories. This macroecological insight offers new opportunities for targeted weed management and potentially could inform strategies to mitigate the impacts of native and non-native weeds in different agricultural contexts. In countries with long cereal cultivation histories, understanding the drivers of spillover from other crops could help identify specific high-risk weed species that could facilitate targeted control measures to improve productivity. Conversely, in countries with shorter cereal cultivation histories, emphasis should be placed on preventing non-native weeds from being introduced, particularly from seed contaminants.