|
|
SummaryMany long-term time series of animal population abundances show increased temporal variability with the length of the time series, which is associated with temporal autocorrelation and reddened spectral color. This might suggest that the temporally uncorrelated environmental fluctuations (white noise), used in most PVAs, are inadequate, and that reddened environmental noise needs to be explicitly incorporated in estimating extinction risks. However, population dynamics may become 'reddened' by various means: by inherited redness from variation in the environment (reddened environmental noise), or through trophic, density-dependent, spatially explicit, or age-structured dynamics. We used simulation models to explore the effects of various within-population processes on the spectral color of the simulated time series of abundances. Results show that reddened spectral color could be explained as the effect of a combination of measurement error and natural variability (in the form of white environmental noise). With both measurement error and natural variability, the spectral exponent is around +1, as in observed time series. Thus, it may not be necessary to invoke complex varieties of environmental noise to explain the observed spectral exponents; the typical value of +1 may be caused by white environmental noise affecting a population that is observed with measurement error.This study was presented at the 2003 Annual Meeting of the Society for Conservation Biology. This study is summarized in: |
|
| Software · Prices · Training · What's New · Forum | |
| Research · Support · Index · Contact Us · Home | |
| ©2003 by Applied Biomathematics | |
Date modified: 1-31-03