Morphological Trait-based Prediction of Konjac (Amorphophallus muelleri Blume) Corm Yield Across Mother Corm Sizes Under Shaded and Potassium-fertilized Conditions

Konjac (Amorphophallus muelleri Blume) is a tuberous crop grown for its glucomannan-rich corms, but yield determinants under semi-intensive production remain poorly understood. This study aimed to identify size-specific morphological traits that could serve as indicators of corm yield of A. muelleri and to determine the growth stage at which these traits showed the strongest relationships with final yield.

A factorial field experiment was conducted using two mother corm size classes (small and large) and four potassium (K) fertilization rates (0, 75, 150 and 225 kg K/ha) under 70% shade. Non-destructive and destructive measurements of root and shoot traits were taken at 12 and 18 weeks after planting (WAP), and simple linear regression analyses were conducted separately for each corm-size class to examine the relationships between selected vegetative traits and final corm yield.

Adjusted coefficients of determination for the strongest fitted relationships ranged from 0.42 to 0.71. In plants derived from small mother corms, Root Fresh Weight (RFW) at 18 WAP showed the strongest association with yield, represented by the regression equation: Yield (t/ha) = 0.25 × RFW (g/plant) + 38.49 (adjusted R² = 0.43). In plants derived from large mother corms, Petiole Fresh Weight (PFW) showed the strongest association with yield. The relationship at 12 WAP explained more variation in yield than that at 18 WAP and was represented by the regression equation: Yield (t/ha) = 0.08 × PFW (g/plant) + 33.98 (adjusted R² = 0.71).

The contrasting trait-yield relationships suggest different physiological limitations between corm sizes, with yield in small-corm plants being more closely linked to belowground establishment and resource acquisition, whereas in large-corm plants it is more influenced by canopy development and assimilate supply.

Final corm yield of A. muelleri was most strongly associated with RFW in plants derived from small mother corms and with PFW in plants derived from large mother corms. The strongest relationship was observed between PFW at 12 WAP and final yield in large-corm plants. These equations should be interpreted as exploratory, condition-specific trait–yield relationships rather than validated prediction models. Because they were derived from a single-location, single-season experiment conducted under 70% shade and a fixed management system, validation across locations, seasons, shade levels, and management practices is required before any predictive application.