The reconciliation of crop productivity and biodiversity maintenance is one of the most important challenges of global agriculture. Weed management is recognized as an integral component of ecological intensification in agriculture because weeds can generate severe yield losses but also form the base of trophic networks in agriculture. Research in weed science has thus often pitted two different schools of thought against one another. The dramatic decline in biodiversity of agricultural landscapes has been considered either as a signal of efficient weed management or as an erosion of natural capital upon which sustainable crop production is based (e.g. support for pollination). Scientists from INRA and the Scuola Superiore Sant’Anna (Pisa, Italy) sought to bring these two seemingly opposite visions together.
Weed-crop competition has historically been studied in experiments that considered only one weed species at a time. Thus, little is known about the competitive effect weeds have in complex communities (multiple weed species), which is the most common scenario in agricultural fields. Nevertheless, major yield losses may arise from the dominance of small number of competitive species. Additionally, higher trait diversity (characteristics) within the weed community could induce complementarity in the use of resources (light, water, nitrogen, etc.) and alleviate weed-crop competition. The authors thus quantified the effect of naturally occurring weed communities and the diversity of their composition on winter wheat yield, across 54 zones (36 unweeded, 18 weeded) and 3 years.
Not all weed communities generate significant crop yield losses
The researchers established that out of the six weed communities identified, only four generated significant yield losses in unweeded zones, ranging from 19 to 56%. The number of ears per plant and the number of grains per ear (yield components) were systematically affected when there was the detection of grain yield losses.
Higher weed diversity mitigates yield losses
Weed diversity was characterized in terms of taxonomy (i.e. evenness of observed species) and function (i.e. diversity of species traits), based on weed biomass data. Evenness is maximal when total weed community biomass is equally divided among species. Results show that when community evenness was high, weed biomass was low and weed-crop competition was alleviated. In this case, yield losses are reduced because all weed species produce low biomass. Authors also reveal a positive effect of weed functional diversity on winter wheat yield, even though no relationship with weed biomass was observed. This could hint that a higher diversity of traits within the weed community limits competition with the crop and hence, yield losses.
These relationships do not imply that high crop productivity is necessarily associated with high weed diversity but rather that, in the presence of weeds, high evenness limits the probability of dominant and competitive species susceptible of generating important yield losses. Moreover, authors insist that the generality of these results needs to be confirmed in other production situations (i.e. climate, crops, weed species pool, resource availability, etc.)
New implications for weed management
Current weeding practices target competitive and dominant species, but are not harmless towards subordinate species. Authors argue that weeding operations should exclusively target competitive and dominant species. However, current weed control practices do not allow targeting a specific species in a complex community. Therefore, future studies need to identify if weed diversity could rather be indirectly promoted by diversifying weed management tools.