Field studies were conducted in three site-years to measure no-till Glycine max yield loss in relation to Apocynum cannabinum vegetative shoot density. Apocynum cannabinum densities of 28 to 40 shoots m−2 reduced predicted G. max yield 58 to 75% and 62 to 94% with the rectangular hyperbolic and linear regression models, respectively. Differences between locations were attributed to rainfall and temperatures, with delayed G. max canopy closure and higher yield loss where soil moisture remained high and temperatures were relatively cool. Application of these predictive G. max yield loss equations to field populations of A. cannabinum showed that between 19 and 36% and 20 and 29% G. max yield loss could be expected from within A. cannabinum patches for the rectangular hyperbolic and linear regression models, respectively. The rectangular hyperbolic regression model appeared to describe the relation between G. max yield loss and A. cannabinum density accurately; however, the model appeared to be dominated by the initial linear phase. This may indicate a lack of high levels of intraspecific competition among A. cannabinum shoots. The results of this study indicate that there is a strong linear relation between G. max yield loss and A. cannabinum shoot density. We conclude that the biological basis for the use of the rectangular hyperbolic model for creeping perennial weeds is questionable.
Nomenclature: Apocynum cannabinum L. APCCA, hemp dogbane; Glycine max (L.) Merr., soybean.