The effect of demographic correlations on the stochastic population dynamics of perennial plants

Abstract

Understanding the influence of environmental variability on population dynamics is a fundamental goal of ecology. Theory suggests that, for populations in variable environments, temporal correlations between demographic vital rates (e.g., growth, survival, reproduction) can increase (if positive) or decrease (if negative) the variability of year-to-year population growth. Because this variability generally decreases long-term population viability, vital rate correlations may importantly affect population dynamics in stochastic environments. Despite long-standing theoretical interest, it is unclear whether vital rate correlations are common in nature, whether their directions are predominantly negative or positive, and whether they are of sufficient magnitude to warrant broad consideration in studies of stochastic population dynamics. We used long-term demographic data for three perennial plant species, hierarchical Bayesian parameterization of population projection models, and stochastic simulations to address the following questions: (1) What are the sign, magnitude, and uncertainty of temporal correlations between vital rates? (2) How do specific pairwise correlations affect the year-toyear variability of population growth? (3) Does the net effect of all vital rate correlations increase or decrease year-to-year variability? (4) What is the net effect of vital rate correlations on the long-term stochastic population growth rate (λ s )? We found only four moderate to strong correlations, both positive and negative in sign, across all species and vital rate pairs; otherwise, correlations were generally weak in magnitude and variable in sign. The net effect of vital rate correlations ranged from a slight decrease to an increase in the year-to-year variability of population growth, with average changes in variance ranging from −1% to +22%. However, vital rate correlations caused virtually no change in the estimates of λ s (mean effects ranging from −0.01% to +0.17%). Therefore, the proportional changes in the variance of population growth caused by demographic correlations were too small on an absolute scale to importantly affect population growth and viability. We conclude that, in our three focal populations and perhaps more generally, vital rate correlations have little effect on stochastic population dynamics. This may be good news for population ecologists, because estimating vital rate correlations and incorporating them into population models can be data intensive and technically challenging.

Multiple mechanisms confer stability to isolated populations of a rare endemic plant

“…While we saw statistical evidence for both patterns, we did not find any strong effect of these mechanisms on population growth rates. Some other studies have found potentially significant effects of demographic rate buffering on extinction risk (Garcia 2003, McDonald et al 2016), while others have not (Compagnoni et al 2016). In part, our negative result may be due to the very limited data we have with which to evaluate these effects; with only four annual transitions of data, characterizing the pattern of elasticity vs. variation of vital rates, or the covariance of different rates, is difficult.…”

“…), while others have not (Compagnoni et al. ). In part, our negative result may be due to the very limited data we have with which to evaluate these effects; with only four annual transitions of data, characterizing the pattern of elasticity vs. variation of vital rates, or the covariance of different rates, is difficult.…”

Reproductive losses due to climate change‐induced earlier flowering are not the primary threat to plant population viability in a perennial herb

“…Methods for vital rate estimation and population modelling closely followed our previous study (Compagnoni et al, ). We used the long‐term demographic data to fit statistical models for six individual‐level processes ("vital rates") that collectively determined population dynamics: (a) survival, (b) growth, (c) flowering (yes or no), (d) fertility of flowering plants (total # flowering stalks), (e) flower stalk abortion rate (flower viability), and (f) seedling recruitment (detailed methods are available in Appendix ).…”

“…Methods for vital rate estimation and population modelling closely followed our previous study (Compagnoni et al, 2016). We used the long-term demographic data to fit statistical models for six in- (detailed methods are available in Appendix S1).…”

Savanna canopy trees under fire: long‐term persistence and transient dynamics from a stage‐based matrix population model

“…For plants, such models can be particularly powerful tools for assessing population viability or recovery in highly variable environments (cf. Tuljapurkar 1990, Caswell 2000, Caswell and Kaye 2001, Tuljapurkar et al 2003, Koons et al 2005, Boyce 2006, Buckley et al 2010, Ellis and Crone 2013, Davison et al 2013, Compagnoni et al 2016, McDonald et al 2016, Needham et al 2018 and references therein).…”