Analysis of a transition matrix model for Gaura neomexicana ssp. coloradensis (Onagraceae) reveals spatial and temporal demographic variability

The fates of individuals in nine fixed plots in three areas were recorded for a population of Gaura neomexicana ssp. coloradensis on F. E. Warren Air Force Base in Cheyenne, Wyoming, U.S.A., from September 1992 to September 1994. The data were used to construct a stage-based transition matrix model for the population across two transitions. The specific goals were to determine if the population size is increasing, decreasing, or stable; if there is significant spatial and/or temporal variation in demographic parameters; which demographic transitions have the most impact on population growth; and the probability of extinction. An additional objective was to compare Gaura with other plant taxa that have been similarly modeled. The population is increasing, although some plots experienced a decline, as evidenced by the asymptotic growth rates (lambda) calculated for the individual plots and for the population as well as the actual numbers of plants in the plots. The population stage structure varied across the two transitions and was different from that predicted by the matrix model. Log-linear analysis revealed significant temporal and fine-scale spatial variation in demographic transition frequencies for Gaura. Elasticity analysis indicated that growth from one stage to another had the largest effect on population growth rate and that the most important single (nonsummed) transitions were from large rosette to flowering and recruitment into the small rosette stage. The population as a whole will persist, but there is a chance for local extinction under certain conditions. The mean growth (G), survival (L), and reproduction (F) elasticity values placed Gaura in the range of other semelparous herbs.