Use of field observations to characterise genotypic flowering responses to photoperiod and temperature: a soyabean exemplar

Thirty-nine accessions of soyabean Glycine max (L.) Merrill] and 1 of wild annual soyabean (Glycine soja L.) were sown at two sites in Taiwan in 1989 and 1990 and on six occasions during 1990 at one site in Queensland, Australia. On two of the occasions in Australia additional treatments extended natural daylengths by 0.5 h and 2 h. The number of days from sowing for the first flower to appear on 50% of the plants in each treatment was recorded (f), and from these values the rate of progress towards flowering (1/f) was related to temperature and photoperiod. In photoperiod-insensitive accessions it was confirmed that the rate is linearly related to temperature at least up to about 29°C. In photoperiod-sensitive genotypes this is also the case in shorter daylengths but when the critical photoperiod (P _c) is exceeded flowering is delayed. This delay increases with photoperiod until a ceiling photoperiod (P _ce) is reached. Between P _c and P _ce, 1/f is linearly related to both temperature (positive) and photoperiod (negative), but in photoperiods longer than P _ce there is no further response to either factor. The resulting triple-intersecting-plane response surface can be defined by six genetically-determined coefficients, the values of which are environment-independent but predict time to flower in any environment, and thus quantify the genotype x environment interaction. By this means the field data were used to characterise the photothermal responses of all 40 accessions. The outcome of this characterisation in conjunction with an analysis of the world-wide range of photothermal environments in which soyabean crops are grown lead to the following conclusions: (1) photoperiod-insensitivity is essential in soyabean crops in temperate latitudes, but such genotypes flower too rapidly for satisfactory yields in the tropics; (2) photoperiod-sensitivity appears to be essential to delay flowering sufficiently to allow adequate biomass accumulation in the warm climates of the tropics; (3) contrary to a widely held view, some degree of photoperiod-sensitivity is also needed in the tropics if crop-duration homeostasis is required where there is variation in sowing dates (this is achieved through a photoperiod-controlled delay in flowering which counteracts the seasonal increase in temperature that is correlated with increase in day-length); and (4) a greater degree of photoperiod-sensitivity is necessary to provide maturity-date homeostasis for variable sowing dates — a valuable attribute in regions of uncertain rainfall. Since the triple-intersecting-plane response model used here also applies to other species, the use of field data to characterise the photothermal responses of other crops is discussed briefly.