The aim of this study was to explore the effects of different density treatments on potato spatial distribution and yield in spring and fall. Plant density influenced yield and composition, horizontal, and vertical distribution distances between potato tubers, and spatial distribution position of tuber weights. The results indicated that: (1) Spring potato yield had a convex quadratic curve relationship with density, and the highest value was observed at 15.75 × 10(4) tubers per hectare. However, the yield of fall potatoes showed a linear relationship with plant density, and the highest value was observed at 18 × 10(4) tubers per hectare; (2) Density had a greater influence on the tuber weight of spring potatoes and fruit number of single fall potatoes; (3) The number of potato tubers in the longitudinal concentration exhibited a negative linear relationship with density, whereas the average vertical distribution distance of tubers exhibited a positive incremental hyperbolic relationship. For spring and fall potato tubers, the maximum distances were 8.4152 and 6.3316 cm, and the minimum distances 8.7666 and 6.9366 cm, respectively; and (4) Based on the artificial neural network model of the spatial distribution of tuber weight, density mainly affected the number and spatial distribution of tubers over 80 g. Tubers over 80 g were mainly distributed longitudinally (6-10 cm) and transversely (12-20 cm) within the high density treatment, and the transverse distribution scope and number of tubers over 80 g were reduced significantly. Spring potato tubers over 80 g grown at the lowest density were mainly distributed between 12 and 20 cm, whereas those at the highest density were primarily distributed between 10 and 15 cm.
Keywords: artificial neural network model; growing season; plant density; potato; spatial distribution; tuber yield.