Density responses and spatial distribution of cotton yield and yield components in jujube (Zizyphus jujube)/cotton (Gossypium hirsutum) agroforestry

Trees are the dominant species in agroforestry systems, profoundly affecting the performance of under story crops. Proximity to trees is a key factor in crop performance, but rather little information is available on the spatial distribution of yield and yield components of crop species under the influence of trees in agroforestry systems. Also, little information is available on how crop density may be exploited to optimize the yield in such systems. Here we studied the performance of cotton in jujube/cotton agroforestry. Field experiments were conducted in 2012 and 2013 in Hetian, Xinjiang, China. Cotton was grown at a row distance of 60 cm in three densities, 13.5, 18.0 and 22.5 plants m(-2) in six m wide paths between tree lines in a jujube plantation. Plant density affected both cotton aboveground dry matter and yield significantly. The highest yield was attained at the intermediate density of 18.0 plants m-2 (20.0 plants m(-2) corresponding in sole cotton), lower than the optimal density in sole cotton (25.0 plants m(-2)). Yield at the lower density was constrained by the low number of bolls per m(2) as a direct consequence of the low density, whereas at the high plant density yield was constrained by a lower allocation of assimilates to cotton seed and lint, as a consequence of intraspecific and interspecific competitions. There were strong gradients in yield and yield components in relation to the distance from the tree rows. Leaf area and total dry matter of cotton in rows close to the tree lines were reduced, especially in the rows next to the trees. Moreover, biomass allocation to cotton fruits was reduced in these rows. Competitive influences from the trees on cotton performance extended two rows deep in a six-year old jujube stand, and even three rows deep in a seven-year old stand. Shading effects on cotton yield were compensated by increasing plant density as a result of greater boll numbers per unit ground area. Data from this study help guide the design of optimal plant density of cotton in jujube plantations and give insight in the spatial distribution and dynamics of competitive effects in agroforestry systems in general. (C) 2016 Elsevier B.V. All rights reserved.