It is generally recognized that estimates of geometric growth rate (lambda) represent the most robust integration of the combined, integrated interaction of population demography on the growth, persistance and response to disturbance of populations. Comparisons among populations often invoke the life-table response experiment (LTRE), an analytical approach which decomposes contributions to lambda in a linear fashion, allowing analysis in a manner analogous to standard linear models ANOVA. However, prior applications have only considered deterministic growth projections, and have ignored stochastic variation through space or time. Our work extends the standard LTRE approach to include measures of stochastic growth rate.

Assessment of impact on populations requires robust identification of which aspects of the population`s demography have been most affected by the perturbation, scaled by the proportional contribution of that particular component to population growth. Since most populations experience some level of stochastic variation, it is critical to be able to differentiate perturbation impacts from stochastic fluctuations (i.e., to parse signal from noise). Most assessment modalities do not take this into account in any form. Our work proposes the first description of such a formalism.

As with much basic research, we do the research, and write the papers, and hope that people involved in application (in this instance, individuals and agencies involved with development and implementation of monitoring programs) read our work, and apply our basic results. However, as with a lot of basic research, there is often a high degree of technical disconnect between the "basic research" and "implementation", owing principally to the technical abilities of many agencies to understand, let alone implement, some of the basic results.

Basic science is rarely directly involved in any "corporate" sense. Our work is no different. No demonstrable benefits to date, unless you consider contributing to our basic conceptual understanding of "how things work" as a benefit -- as beneficial as a direct "implementation" (without basic research, there is nothing to implement, after all).

None - we do this on our own time. Much of this work involves `ideas`, and `thinking` - with some computer time.

Hal Caswell (Woods Hole Oceanographic Institute)