Published online by Cambridge University Press: 19 September 2008
Both temperature (T) and water potential (ψ) have consistent and quantifiable effects on the rate and extent of seed germination (radicle emergence). Germination at suboptimal T can be characterized on the basis of thermal time, or the T in excess of a base (Tb) multiplied by the time to a given percentage germination (tg). Similarly, germination at reduced ψ can be characterized on a hydrotime basis, or the ψ in excess of a base (ψb) multiplied by tg. Within a seed population, the variation in thermal times to germination for a specific percentage (g) is based upon the normal distribution of ψb values among seeds (ψb(g)). Germination responses across a range of suboptimal T and ψ might be accounted for by a general hydrothermal time model incorporating both T and ψ components. We tested this hypothesis for tomato (Lycopersicon esculentum Mill.) seeds of two genotypes differing in germination rates and tolerance of suboptimal T and ψ. For combinations of T (10−25°C) and ψ (0 to −0.9 MPa), a general hydrothermal time model accounted for approximately 75% of the variation in times to germination within the seed populations of both genotypes, and over 96% of the variation in median germination rates. However, ψb(g) distributions were sensitive to both the T and ψ of imbibition, resulting in a poor fit of the model to specific time course data. Analysis of germination timing separately for low and high ψ ranges within a given T resulted in specific models accounting for 88−99% of the variation in individual germination times and >99% of the variation in madian germination rates. Thus, for a given T and ψ range, the hydrotime model closely matched tomato seed germination time courses. Accumulated hydrothermal time accounted well for germination rates at ψ> −0.5 MPa across suboptimal T if ψb(g) was allowed to vary with T. Germination did not show a consistent response to T at ψ < −0.5 MPa, and estimated Tb values varied over different T ranges. Generalization of the hydrothermal time model across the entire range of suboptimal T and ψ was limited by physiological adjustments of the seeds to their current environment. The hydrothermal time model detected and quantified these adjustment processes that would otherwise not be evident from inspection of germination time courses. Temperature and water potential influence the time to germination via physiological mechanisms that reciprocally interact.