Plant functional traits may serve as a useful tool to predict ecosystem services from cover crops (non-harvested crops) in agroecosystems, and trait-based ecology provides a generalizable framework for linking agroecosystem diversity and function. However, limited knowledge of the ranges and drivers of trait variation, and associated influences on ecosystem services, currently constrains the application of a trait-based approach to cover cropping. This knowledge gap is especially pronounced for functionally diverse multi-species mixtures, where complex biotic interactions (e.g., competition, facilitation) likely impact these dynamics. Our field experiment addressed two questions: How do species interactions influence functional trait expression in cover crop mixtures? And, can cover crop functional traits predict the delivery of ecosystem services? We planted warm season cover crops in a randomized, complete block design with four replicates in southeast Michigan, with species selection guided by contrasts in plant functional traits. Treatments included cowpea, sorghum-sudangrass (sudex), and buckwheat monocultures; cowpea-sudex and cowpea-sudex-buckwheat mixtures; and a weedy fallow control. Using standardized protocols, we measured plant height, specific leaf area (SLA), leaf N concentration, root:shoot ratio, and aboveground biomass and C:N. Ecosystem services measured included weed suppression, nitrogen (N) assimilation, and potential NO3- leaching during the cover crop season.
Results/Conclusions:
Supporting our hypotheses, traits related to resource acquisition, including plant height, SLA, and leaf N, were in some cases significantly higher in mixture than in monoculture, indicating that complementary interactions likely occurred between species in mixture. As expected, interspecific trait variation drove several ecosystem service outcomes (e.g., aboveground biomass and C:N were positively related to aboveground N assimilation (R2 = 0.864 and 0.217, and p = 0.0001 and 0.007, respectively), and root:shoot corresponded with greater total (above- and below-ground) N assimilation (R2 = 0.578, p = 0.001)). These relationships provide evidence that traits can predict ecosystem services, and highlight the importance of species selection when targeting specific services from cover crops. We also found that aboveground N assimilation was positively related to sudex SLA (R2 = 0.312, p = 0.059), but the cause of the intraspecific trait variation remains poorly understood, indicating the need for further efforts to better parameterize such variation in order to more accurately predict and manage ecosystem services. Continued refinement of this ecological approach to cover crop selection could improve efforts to strategically manage plant diversity for specific ecosystem services, which may, in turn, increase agricultural sustainability.