Masamu Institute for Research, Training, and Advanced Studies

AbstractThe purpose of my talk is to give a progress report on a Masamu research program, the elephant demography project.  The elephant project has morphed over time, and during the past couple of years turned its focus to the topic of demographic stochasticity.  Demographic stochasticity is a property of population growth that results from dealing with integer values; that is, fractional individuals do not exist in nature, even though population models generally allow fractional individuals.  To investigate the importance of this distinction in natural populations, we measured population change (growth or decrement) using temporal censusing of populations over long periods of time.  Population growth can be measured (1) as year-to-year changes in the number of individuals.  Using some fairly simple matrix algebra, one can also project year-to-year changes in population size (2) based on probabilities of reproduction and survival (and numbers of offspring for breeders), and on the initial population size in the first year.  These two approaches do not yield the same subsequent population sizes, because the latter “controls” for demographic stochasticity and can be used to estimate the resulting demographic variation.  To learn more, we applied both approaches to 10 populations of birds and mammals (including elephants!) that had extensive long-term data, and asked whether demographic stochasticity was important to annual changes in population sizes. What did we learn?  … come to the talk!