Spatiotemporal variation in the local population dynamics of the short-lived Opuntia macrorhiza (Cactaceae)

Background/Question/Methods

Spatial and temporal variation in demographic rates can have profound effects on long‐term persistence of populations. While effects of environmental stochasticity on population dynamics is well studied, it is not usually known how local dynamics can be generalized across populations. In particular, is it safe to assume that environmental variation has similar effects across populations within small spatial scales?  We use long-term demographic data from seven years of the short-lived cactus Opuntia macrorhiza (Cactaceae) at five different locations in Boulder County, Colorado, USA, to construct stochastic matrix models. This species is endangered in Iowa and salvage-restricted in Arizona. To predict future population response to increasing environmental variation we use stochastic elasticities of population growth rate to vital rates. To understand, the difference in local dynamics of different populations we use the recently developed stochastic life table response experiments (SLTRE) to quantify the relative contribution of vital rates (e.g., survival vs fecundity) to the difference in population growth rates across populations.        

Results/Conclusions

Our results showed that populations can vary dramatically within relatively small spatial scales.  All populations differed in their mean vital rates and also in the extent to which they were buffered against environmental variation.  Two of the populations had positive stochastic population growth rates log λ_s while the other three had log λ_s < 0. Stochastic elasticities were largest for transitions associated with smaller sized plants (plants 2-3 cladodes) suggesting that the fate of smaller plants is most important for future population growth rates. However, SLTRE showed that the observe differences in between the populations was mostly due to mean transition rates (survival, growth and regression) associated with medium sized plants (4-7 cladodes). Further, elasticities to the variance of reproductive rates were small but positive, indicating positive effects of increased variation in these vital rates. Our results suggest that relying exclusively on models parameterized by aggregating data from different years and sites can result in erroneous conclusions about population viability. Hence our results suggest that local dynamics should be considered in the management and conservation of species of concern.