Developing a set of simulation-based indicators to assess harmfulness and contribution to biodiversity of weed communities in cropping systems

SPE EA EcolDurInternational audienceWeeds are both harmful for agricultural production and an essential component of biodiversity in agricultural landscapes. Therefore, new cropping systems aiming at both maximising weed-related biodiversity and minimising weed harmfulness are needed. New cropping systems are now increasingly designed with weed dynamics models but these usually only consider weed densities or crop yield losses. The present paper proposed a set of indicators for assessing the impact on crop production and biodiversity of weed communities simulated with a cropping system model. Five harmfulness indicators were developed to take account of the criteria most frequently listed by farmers via an internet survey: (1) crop yield loss, (2) harvest pollution by weed seeds, stems and leaves, (3) harvesting problems due to green weed biomass blocking the combine, and (4) field infestation represented by weed biomass averaged over cropping seasons. A fifth indicator was added, i.e. (5) the increase in crop disease (i.e. take-all disease of cereals) in the presence of weeds. The biodiversity indicators were chosen in collaboration with ecologists. Two indicators reflect the weed contribution to vegetal biodiversity: (1) species richness and (2) Pielou's index for species equitability. Three other indicators were developed to assess weeds as a trophic resource for other organisms in the agro-ecosystems: (3) the number of weed seeds present on soil surface in autumn and winter to feed field birds, (4) lipid-rich seeds on soil surface in summer to feed insects such as carabids, and (5) weed flowers in spring and summer to feed domestic bees. These indicators were tested in a series of contrasted cropping systems identified in farm surveys and simulated with FwaSYs. Analyses of variance showed that the cropping system and the crop sequence presented the highest impact on indicator values. Weather scenario and pedoclimate had little effect. Antagonisms and synergies between weed-related harmfulness and biodiversity were identified with Spearman correlations. Harmfulness indicators were all positively correlated, except for additional disease risk which was at best poorly correlated with other indicators. Most weed-related biodiversity indicators were also positively correlated, except species richness which was negatively correlated with species equitability, bird resource and insect resource. Weed harmfulness generally increased with increasing weed-related biodiversity. These correlations were though weak, and others were negative, showing that increased biodiversity could occur with decreased harmfulness (e.g. trophic resource for insects vs. yield loss or field infestation). Consequently, there are cropping systems that reconcile agricultural production and biodiversity. (C) 2014 Elsevier Ltd. All rights reserved