Document: HAL-3-39-19

Matrix models for annual organisms: The missing link between phenology and demography.

CASWELL, H.*

Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA. ¹

Abstract:
Annual or sub-annual life cycles are common among insects and plants. The demographic analysis of annual organisms is difficult because it involves two time scales. Within a year, annuals exhibit richly complex life cycles, often tightly coupled to seasonal environmental changes. The literature is replete with detailed analyses of demography on this phenological scale. But population growth takes place on an inter-annual time scale, and the life cycle of many annuals collapses to a single stage, often a dormant or diapause stage, at some point in the year. Thus, many models of annual organisms forget the within-year processes and project, e.g., seeds in year t to seeds in year t+1. This makes it impossible to explore the link between within-year phenology and between-year population growth. Here I present a new class of matrix models based on periodic products of non-square matrices with heterogeneous projection intervals. I show how to derive them from a new kind of seasonal life-cycle graph. They explicitly link the two time scales, projecting the entire life cycle from year to year, without sacrificing phonological information. I will show how to calculate the population growth rate, the stable stage distribution and reproductive value of any stage at any point in the seasonal cycle, and the sensitivity and elasticity of population growth rate to perturbations of any stage at any point in the seasonal cycle. These models can be used, inter alia, to analyze the evolution of annual life histories and to evaluate conservation strategies. They permit incorporation of environmental stochasticity and density-dependence.

Keywords:

This abstract is being presented at: 10:30 AM in session:
PLANT DEMOGRAPHY