The Limits to Knowledge in Conservation Genetics

Abstract

Science is a “way of knowing” (Moore, 1984) that distinguishes itself by developing theories capable of prediction; however, in studies at the interface of evolution and the environment, this task can become formidable. The future direction of evolutionary change is intrinsically unpredictable, because the unit of study (the population) cannot be isolated from changes in its environment. In contrast, the physical sciences and most of Biology have been able to achieve the goal of prediction by studying systems that generally can be understood in terms of their internal properties. Thus, a particular type of cell or organism studied today is expected to be much the same when studied by subsequent generations of biologists. However, this expectation is lost when we are asked to consider populations and communities over even moderate periods of time: changes on the time scale of tens of years are commonplace, often dramatic, and often caused by unpredictable events external to the study unit. Short-term changes are primarily numerical, but evolutionary changes, both adaptive (e.g., Reznick et al., 1997) and random (e.g., due to a population bottleneck), can accumulate rapidly. The stochastic models of ecology and population genetics include unpredictable environmental effects, allowing us to make probabilistic predictions that can be quite precise when we consider averages over large numbers of populations, large numbers of genes, or long periods of time.