When are matrix models useful for management? An empirical test across plant populations

Demographic models are widely used to predict the consequences of management actions, such as sustainable harvest, possible restoration strategies of endangered species, and ways to control invasive species. These models, like all ecological models, simplify the complexity of natural systems and therefore are not expected to accurately reflect all aspects of population dynamics. However, if they are too simplistic, these models would not provide good guidance for management. The key question, then, is whether the models capture enough detail to support their common applications. Our working group will, first, quantify how models are interpreted in practice for management of plant populations. For example, do practitioners expect models to make exact predictions of population size, distinguish growing from declining populations, or simply tell them which management option is best? Next, we will test whether models provide useful guidance in these applications, in spite of their simplicity. We have just reached the point where models, developed for plant populations 5-20 years ago, can be compared to subsequent population dynamics. The combination of designing tests that reflect actual applications and comparing predictions with realized management outcomes will provide an important evaluation for this much used tool. Our project will also provide an example of potential tests of ecological models in other contexts, and a database of long-term studies that other researchers can use to address questions in basic and applied ecology.