Article ID Journal Published Year Pages File Type
4511640 Field Crops Research 2007 9 Pages PDF
Abstract
Cultivar × location interaction for lucerne forage yield across northern Italy is large, repeatable and associated with summer drought-stress level and soil type. The objectives of this study were: (i) to investigate the genotypic factors associated with cultivar adaptive responses to drought-stress and soil factors; (ii) to identify adaptive traits exploitable for selection of widely adapted or specifically adapted material. Aerial dry matter (DM) over 12 harvests and shoot traits of 13 landraces and four varieties were evaluated in four artificial environments created by the factorial combination of drought-stress level (almost nil or high) and soil type (sandy-loam or silty-clay) (Exp. 1). Aerial and root DM over four or five harvests were evaluated in metal containers 55 cm × 12 cm × 75 cm deep for the factorial combinations of three varieties by two drought-stress levels by two soil types (Exp. 2), or six landraces by two drought-stress levels (Exp. 3). Cultivar × environment interaction was detected for forage yield, plant mortality after the second summer, leaflet size and stem weight. The environments of Exp. 1 reproduced well the variety adaptive responses across agricultural environments. The relationship of cultivar forage yield with shoot traits was environment-specific, i.e.: (i) strictly negative with plant mortality, in no-stress environments (where mortality and plant competition were severe); (ii) positive with stem number per plant and autumn-winter growth, in stress environments; (iii) positive with stem dry weight, in 'stress/sandy-loam soil'; and (iv) positive with leaflet size, in 'no-stress/sandy-loam soil'. Cultivars specifically adapted to no-stress or sandy-loam conditions showed consistently greater root DM across three soil layers than material with opposite adaptive response. Entry yields tended to inverse genetic correlation between two environments which represented contrasting geographical subregions and were characterized by different combinations of traits associated with higher yield, i.e.: (i) higher root biomass, lower plant mortality and larger leaflets, for 'no-stress/sandy-loam soil'; and (ii) more stems per plant and greater autumn-winter growth, for 'stress/silty-clay soil'. The difficulty of yield-based selection for wide adaptation may hardly be overcome by selection based on adaptive traits. Adaptive trait-based selection for specific adaptation may be envisaged especially for 'stress/silty-clay soil' conditions, for which the relevant traits are inexpensive to record, not correlated, and not subject to cultivar × environment interaction.
Related Topics
Life Sciences Agricultural and Biological Sciences Agronomy and Crop Science
Authors
,