| Abstract: | Research was conducted to formulate a temperature-dependent population level model for johnsongrass (Sorghum halepense (L.) Pers.) phenology. A nonlinear poikilotherm rate equation was used to describe development rate as a function of temperature and a temperature-independent Weibull function was used to distribute development times for the population. Seed and rhizome johnsongrass germination and emergence data and rhizome johnsongrass flowering data were collected under constant temperature conditions to parameterize both the poikilotherm equation and the Weibull function. Germination rate was higher for rhizomes than seed up to 32 C, but lower at higher temperatures. Shoot elongation rate was higher for seedling johnsongrass than for rhizome johnsongrass at all temperatures. No high temperature inhibition was detected for seedling emergence up to 40 C. Time to first germination or emergence event was longer for seedling than rhizome johnsongrass, but the progression of development for the population was faster. Shoot elongation differences between planting depths of seed or rhizome identified problems with the model relating to plant death. Coupling the poikilotherm rate with the Weibull function resulted in a population level temperature-dependent model suitable for characterizing plant development. The model was tested and validated against independent data sets under conditions of variable temperatures. Accurate predictions were obtained for seed and rhizome johnsongrass germination, seed and rhizome johnsongrass emergence from shallow depths (2.5 cm for seed, 5 cm for rhizomes), and rhizome johnsongrass flowering from plants emerging in the spring. The model performed poorly for seed and rhizome johnsongrass emergence from deeper depths (5 cm for seed, 10 cm for rhizomes) and for rhizome johnsongrass flowering from plants emerging in the summer. Poor performance of emergence could be related to plant death. Adjustments to the high-temperature parameters of the poikilotherm rate equation improved model performance in the summer without affecting spring predictions. |
|---|