The effect of spatio-temporal heterogeneity on the dynamics of plant populations

Background/Question/Methods

Spatio-temporal variation in demographic rates is still an area of considerable theoretical and empirical interest in population ecology. Variation in demographic rates affects fitness, and it therefore has implications for population dynamics and evolution. In general, demographic rate variation is expected to decrease long term population growth, but this need not always apply. For example, such variation might have no effect if demographic rates covary negatively among each other. We used long term plant demographic data from three perennial plant species to estimate the variation and covariation of vital rates across space and time. To estimate demographic rate covariation and to quantify parameter uncertainty, we fit models in a hierarchical Bayesian framework. We used parameter estimates to build stochastic Integral Projection Models (IPMs) for each species, which we used to address three questions: 1) What is the effect of interannual and spatial variation in demographic rates on long term population growth? 2) Is there evidence for temporal and / or spatial co-variance across vital rates? 3) Does covariance among demographic rates change the effect of uncorrelated demographic variation?

Results/Conclusions

First, we found that both spatial and temporal variation in demographic rates have a negative effect on the population growth rate of one of our three plant populations. The effect of spatial variation was an order of magnitude smaller than that of temporal variation. Second, we found evidence for negative covariation among demographic rates. Third, we found that negative vital rate covariance lessens the negative effect of temporal variability on long term population growth. Our results support theoretical expectations that variable demographic rates tend to decrease fitness. Negative covariance among demographic rates is rare, and this is the first study, to our knowledge, to consider the effects of both temporal and spatial covariance. Overall, our results suggest that negative covariance among demographic rates might be selected for, at least in some populations.