Responses of adzuki bean as affected by row spacing, plant density and variety

Abstract

Agronomic responses of varieties of adzuki bean, Vigna angularis (Willd.) Ohwi and Ohashi, to row spacing and plant density were investigated at 3 locations, Kumbia and Warwick in Queensland, and Grafton in New South Wales. The treatments were: at Kumbia and Warwick row spacings of 17.8, 35.6 and 71.1 cm, 3 densities of 250 x 103, 500 x 103 and 750 x 103 plants/ha, and 4 and 3 varieties respectively; and at Grafton 4 densities of 200 x 103, 400 x 103, 550 x 103 and 700 x 103 plants/ha and 3 varieties at 30 cm row spacing. The varieties Bloodwood and Dainagon were common across locations. Grain yield was increased by 26% at Kumbia and 19% at Warwick for narrow rows versus wide rows, and by 31% at Kumbia, 22% at Warwick, and 19% at Grafton for high versus low plant density. These responses occurred over sites differing in environment and in mean yield. At Kumbia a significant interaction for yield occurred between plant density and row spacing and between variety and row spacing, whereas at the other locations yield responses to row spacing and to plant density were linear and additive. These responses differed for the large-seeded variety Dainagon with a trend for a yield decline at the highest plant density at Grafton. Phenologic responses to row spacing and to plant density were significant but very small at Warwick. Ground cover percentage at mid pod fill was reduced at both low plant densities and in wide rows, with smaller differences between the intermediate and high levels of each treatment. Variety, row spacing and plant density showed interactions for expression of ground cover at Kumbia and Warwick. Both seed weight and canopy height were less at low plant densities at Kumbia and at Warwick, but responses differed by variety at Grafton, whereas these traits were both less at wide row spacing at Kumbia but not at Warwick.Other traits with responses to plant density included; lodging percentage with a small increase at 250 x 103 ha at Kumbia, and shoot biomass which was reduced at low plant density and at wide spacing at Warwick. There were significant varietal differences in all traits except lodging percentage. Bloodwood yielded well at each site, equivalent with Erimo at Grafton and Warwick, but Dainagon was equivalent at Kumbia only and yielded significantly less at other sites. Dainagon was shorter and with larger seed than Bloodwood at each site, it reached maturity earlier at Kumbia but later at Warwick, and it had less ground cover than Bloodwood at Warwick but not quite significantly less at Kumbia. Bloodwood and Erimo were very similar over all traits except for Bloodwood having a taller canopy and later maturity at Warwick. Adzuki bean grain yield was strongly affected by both plant density and row spacing with increases from low to high plant densities and wide to narrow rows, in association with responses in ground cover. Shoot biomass, recorded at Warwick only, was less in wide rows but unresponsive to plant density. Harvest index increased with increased plant densities but was unresponsive to row spacing.