Organic weed suppression using diverse cover crop mixtures

Background/Question/Methods

Multi-species cover crop mixtures, or “cocktails,” are an increasingly popular ecological weed suppression tool on organic farms.  However, little is known about the effects of cover crop cocktail (CCC) species and/or functional diversity on weed suppression in organic grain crops.  We investigated CCC weed suppression in two stages of an organic corn/soybean/wheat rotation: after corn (Zea mays) and after wheat (Triticum aestivum).  Four replicate plots of 12 cover crop treatments ranging from zero to seven species were planted in late August (following wheat) and mid-October (following corn) of 2012.  We measured light attenuation below the cover crop canopy with a line quantum sensor and percent ground cover at multiple time points between fall planting and spring termination.  Paired plots were destructively sampled for biomass before winter dormancy and before termination.  We hypothesize that greater CCC species and functional diversity both enhance weed suppression by increasing cover crop competition with weeds at critical weed developmental stages.  We assess the roles of species and functionaldiversity (the latter using these categories: grasses, legumes, brassicas, and mixtures of the preceding types) using regression on final weed biomass.  We also assess the time-dependence of these dynamics using partial least squares (PLS) path modeling.

Results/Conclusions

Preliminary results from fall data indicate that rapid canopy closure and CCC functional diversity may be important determinants of CCC weed suppression, but not species diversity.  Simple univariate regressions show that light interception at first non-destructive sampling, 39-44 days after planting (DAP), explains a substantial portion of the variation in final fall weed biomass, 75-79 DAP (R² = 0.46; p < .001).  CCC species diversity is not a strong predictor of light interception at 39-44 DAP (R² = 0.079, p = .065) or weed biomass at 75-79 DAP (R² = 0.035; p = .23).  CCC functional diversity, however, is a strong predictor of light interception at 39-44 DAP (R² = 0.72; p < .001).  Further, CCC functional diversity explains a substantial portion of the variation in weed biomass at 75-79 DAP (R² = 0.38; p < 0.001), primarily due to poor legume weed suppression.  PLS path modeling of the early-planted cover crops indicates that Early Cover Crop Canopy (a latent variable reflecting multiple measures of the cover crop canopy at 39-44 days) is more negatively correlated with final fall weed biomass (R = -0.42) than Late Cover Crop Canopy (prior to fall destructive sampling, R = -0.29).