Incidence and measurement of autumn seed softening within Medicago polymorpha L

Abstract

Softening of hard seeds during autumn rather than summer is a desirable characteristic for reducing seed losses in annual legumes from false breaks of season in Mediterranean environments. The incidence of this characteristic in Medicago polymorpha L. was determined in 34 lines grown at Toodyay, Western Australia, in 1993 and collected in December. Patterns of seed softening during summer and autumn 1993-94 were studied in pods placed on the soil surface in the field at Merredin, and in laboratory and glasshouse simulations at Perth. Summer seed softening was simulated in the laboratory by the removal of seeds from pods by hand, subjecting them to a diurnally fluctuating temperature of 60/15¦C for 16 weeks, and testing their permeability. Residual hard seeds were subjected to 4 gradual diurnal temperature fluctuations of 35/10¦C and re-tested for permeability as a measure of autumn seed softening. In the glasshouse study, pods were placed on the surface of soil in boxes, and emerged seedlings were counted after watering in March and June. Field softening in the first year ranged from 6.8 to 69.6%, but exceeded 40% in only 6 of the 34 lines. Proportions of total soft seeds present in the field in June that had softened after 1 March in the 34 lines were normally distributed, and ranged from 2.5 to 78.7%. The laboratory simulation markedly underestimated both autumn and total seed softening in 13 of the lines but effectively predicted field softening behaviour in the other 21 lines. The glasshouse technique overestimated the proportions of seeds softening in autumn in most lines and underestimated total softening in 12 of the 34 lines. A technique involving the use of a rain-out shelter is proposed for routine determination of the incidence of autumn seed softening in medic evaluation programs.