Hydrothermal Germination Models: Assessment of the Wet‐Thermal Approximation of Potential Field Response

Abstract

Hydrothermal (HT) models can be used to characterize seed germination response to field‐variable conditions of temperature (T) and water potential (Ψ). Hydrothermal response data are relatively difficult to generate, which limits their utility for large‐scale comparisons of inter‐ and intraspecies germination response. Previous studies have hypothesized that HT germination response can be estimated using a simple model for thermal‐time accumulation above a fixed threshold of environmental Ψ (wet‐thermal [WT] model). The purpose of this study was to test this hypothesis by explicitly comparing HT and WT germination rates and estimated cumulative germination response of 13 rangeland grass seedlots under simulated conditions of field‐variable T and Ψ. We used a 44‐yr weather record to parameterize a seedbed‐microclimate model for estimation of hourly T and Ψ at seeding depth for a sandy loam soil type at the Orchard field test site in southwestern Ada County, Idaho. Hydrothermal and WT germination responses for each hour of the simulation were estimated for 13 range grass seedlots. We found that overestimation of germination rate by thermal accumulation at low Ψ contributed relatively little to WT model error rates, and that >95% of the variability in predicted germination response could be explained by WT germination response above a Ψ threshold of −0.3 to −0.5 MPa. Given the pulse‐like nature of favorable germination conditions in the field, this modeling approach may have immediate and wide potential application, as there are a relatively large number of thermal‐germination datasets currently available for rangeland plant species.