LTER Cross-site: Collaborative Research - Assessing the Geographic and Temporal Consistency of Life History and Demographic Patterns: A Long-term, Multi-site Comparison

Duke University, Durham NC

Investigators

Abstract

0087096 Morris Ecologists increasingly understand the importance of spatial variability in natural systems, a recognition that led to the creation of the LTER network. However, we still know surprisingly little about spatial variation in the demography of individual species, creating an important gap in knowledge for both applied and basic ecology. The paucity of full demographic studies that span multiple populations across substantial geographic scales creates two general problems. First, population biologists are forced to extrapolate from the population dynamics and life history patterns found in single, well-studied populations to those occurring in distant parts of a species range. Such extrapolation is commonplace in conservation biology, comparative life history analysis, and climate change predictions, and is currently done with little understanding of how much spatial variation actually exists in the demography of individual species. Second is the question of what sets the range limits of species. While the physiological effects of climate are often well-known, the lack of broad-scale demographic studies makes it less clear how population dynamics vary near range limits, and thus the extent to which poor demographic performance versus other proposed factors, such as metapopulation dynamics, determine the distributional limits of species. As a result, we have little ground for building mechanistic predictions about how species ranges will shift as climate changes. To address these problems, a demographic study of two widespread tundra plant species, moss campion and alpine bisort, will be conducted as part of this LTER cross-site research project. These species will be studied in 16 populations across two habitats and four geographic sites. These sites span most of the latitudinal range of the species in North America. The study is designed to parameterize stochastic matrix models for each population of each species. While multi-site demographic studies are unusual, ones that measure and properly analyze temporal stochasticity in vital rates are even rarer. In addition to statistical analysis of the means and variance in individual vital rates, the PIs will base much of the analysis on outputs of these stochastic models, including stochastic growth rates, stochastic elasticity values and extinction risk predictions. Furthermore, the work will utilize LTER climate data to tie spatial and temporal variability in vital rates to specific environmental factors. The results will be used to test for the strength and pattern of life history variation across geographic sites and habitats, and to determine how best to make predictions about demographic variability with limited data. Finally, specific tests will be conducted for demographic factors leading to both northern and southern range boundaries. Overall, this study will be the first multi-site, multi-year demographic analysis to carefully include temporal stochasticity and to explicitly test how and in what ways life history patterns vary across species' ranges.