Modeling Phasic Development in Sorghum: New Hypotheses from Old Data.

Variation in sorghum (Sorghum bicolor) biomass and grain yields are often associated with large genetic differences in phenology. CERES-Sorghum, as implemented in the Cropping Systems Model (CSM) of the Decision Support System for Agrotechnology Transfer (DSSAT, version 4.5), simulates development over seven phases from germination to physiological maturity. For a given cultivar, durations of these phases are determined by six parameters for phase durations plus two additional parameters that determine photoperiod sensitivity. This could be interpreted as meaning that at least eight distinct genetic controls affect sorghum phenology. An alternative, simpler perspective is that the relative durations of the three phases from emergence to anthesis (i.e., juvenile, photoperiod sensitive, and pre-anthesis) likely are constant (in the absence of photoperiod effects), and that variation in these phases reflects differences in an intrinsic rate of development (D) that is manifested throughout pre-anthesis development. Thus, in the absence of photoperiod effects, genetic differences in time to anthesis would primarily be due to differences in D. We examined data from Alagarswamy et al. (1998; Field Crops Res. 55:1) and Caddel and Weibel (1971; Agron J. 63:799) to estimate the relative durations of the juvenile (P1), photoperiod sensitive (P2), pre-anthesis (PANTH) phases as P2 = 1.22 P1 (r² = 0.28*) and PANTH = 1.63 (P1 + P2) (r² = .82**).  These relations were applied to independent sets of sorghum cultivars and hybrids. Besides facilitating routine parameterization of CERES-Sorghum, this approach aligns well with the current understanding of the molecular control of flowering. Two fundamental questions that arise are what molecular mechanisms underlie genetic variation in D [which is essentially 1/(P1 + P2 + PANTH)] and do the cardinal temperatures assumed in estimating D differ among sorghum genotypes?