Plant population studies on peanut (Arachis hypogaea L.) in subtropical Australia. 2. Water-limited conditions

Abstract

The effects of plant population density on total dry matter (TDM) production, and on pod and kernel yields, of 2 peanut (Arachis hypogaea L.) cultivars (Virginia and Spanish) were investigated under a range of contrasting soil water availability regimes. Protracted crop water deficits were applied to each plant population density treatment in 3 experiments: (i) from planting until the early pod-filling phase (DSWF, dry start, wet finish); (ii) during the pod-filling to maturity phase (WSDF, wet start, dry finish); (iii) from flowering to maturity (TS, terminal stress). Crop water deficits of varying timing and severity were shown to modify substantially the effect of plant population on yield response compared with that observed under well-watered conditions. In most cases, TDM was maximised at the lowest density (40000 Plants/ha). In the WSDF and DSWF experiments, significant cultivar x plant population interactions for pod yield were found. The Spanish cultivar, McCubbin, showed strong pod yield response to S30000 plants/ha, while the Virginia cultivar, Early Bunch, did not respond to increases in plant population above 40000 plants/ha. These differing responses were probably associated with cultivar differences in branching pattern. Under extreme water stress situations where crops were forced to rely solely on soil water reserves (TS), pod yields were highest at the lowest plant population density and declined rapidly as plant population increased. The Gardner and Gardner (1983) model provided a useful framework to characterise the plant population-pod yield response under reduced water availability. The assumptions that both partitioning of dry matter to pods and the hypothetical minimum plant size capable of producing pods were crop constants, irrespective of crop water deficits experienced, were shown to be invalid. These constants may, however, be linearly related to water availability. Relationships relating these parameters to an index of crop or soil water status may improve the predictive capability of the model under water-limited conditions.