Effects of Location and Plant Density on Yield and Architectural Traits in Dry Beans1

Abstract

Plant architectural traits associated with yield have been suggested as indirect selection criteria for improving seed yield of dry beans (Phaseolus vulgaris L.) but critical data from selection experiments are lacking. Six dry bean experimental lines of growth habits I, II and III, bred for one or more plant architectural traits, were therefore compared at three locations in Colombia with six other cultivars which were conventionally developed experimental lines of similar growth habits. At each location plant densities of 5, 13, 22, and 30 plants/m2 were established. The objectives were to examine: a) whether lines bred for architectural traits outyielded cultivars and experimental lines bred conventionally, b) the effects of environment and plant density on architectural traits and seed yield, and c) the associations between architectural traits and yield. We did not find a genotype of any growth habit bred for specific architectural traits outyielding the commercial check cultivar. Indeterminate, prostrate type llI, and type II cultivars, in that order, were among the highest yielding regardless of environment and plant density. These also offer the best opportunity for the development of broadly adapted, high yielding cultivars of dry beans for monoculture. Locations, seasons within locations, and plant densities each affected yield and architectural traits. Significant interactions were observed between environment, growth habit, and plant density; and these factors affected yield and all architectural traits. Differences were greatest between determinate and indeterminate growth habits (type I vs. types II and III). For plant density the shape of the response curves for yield was parabolic whereas the shape for type I beans was asymptotic. Curvilinear (types I and II) and linear (type III) increases nodes/m2 were observed as plant density increased. Branch and node number/plant decreased linearly with increasing plant density for genotypes of all three growth habits. Linear reductions in node number on the main stem were observed in indeterminate types II and III with increasing plant density, but internode length did not change. In contrast, the number of nodes on the main stem of determinate type I remained unchanged but internode length increased linearly as plant density increased. Nodes/m2 and branches/plant were positively correlated with yield regardless of environment or plant density. In contrast, nodes/branch and internode length on the main stem were positively correlated with yield over the range of plant densities used, but showed variable correlation over environments. Therefore, development of bean plant ideotypes with enhanced expression of certain architectural traits can result in limited adaptation and reduced yield potential in some environments.