|
certainty in our estimates of natural mortality and growth for this population, so simulations were conducted for three different growth rates and two rates of natural mortality. Regardless of natural mortality, fast growth increased the benefit of the size limits on total catch (Figure 1). If natural mortality was low (10%), an 18-inch length limit was predicted to triple angler catch rates over those found with no length limit if growth is medium or fast (Figure 1). The 14-inch minimum size limit and slot limit provided intermediate angler catches. However, at high natural mortality (25%), the model indicated that length limits would slightly increase but not greatly change total angler catch (Figure 1). Thus, implementing a size limit is predicted to either increase or not greatly change angler catch rates for the range of growth and mortality simulated. Even with uncertainty in our data estimates, the model provides useful information to managers that when combined with social and economic factor can form the basis of sound management decisions.
One of the most powerful uses of models is to assess where additional field data should be collected or improved. A population model allows biologists to incorporate all available data into the simulations. Areas where data are missing will be recognized and can be prioritized for collection in future sampling efforts. Alternately, modeling across a range of parameters often identifies specific parameters that greatly influence model output. This is often termed Asensitivity analyses,@, which simply describes the procedure of modifying each input parameter by a certain level or percentage (e.g., percent change in growth, recruitment, mortality) to assess which factor has the most influence on total catch, yield, or population size. Model parameters that greatly influence the simulated population can then be targeted for improved estimates from field data. For example, the magnitude and variation in fish recruitment across years may greatly influence population responses to a size limit, and this can be assessed by modeling. Agency resources could then be allocated to attain more precise estimates of these influential parameters.
Johnson (1995) noted that variation and bias in field data is common, as are biases in simulation models. Thus, differences between model outputs and the population response could result from error in the field data, error in the model, or both (Johnson 1995). Students using population models should realize the limitations of both the model and the input data! Population models will continue to serve an important role in fisheries management, but they will not make decisions for us. Models are just one tool in the overall assessment of fish populations. However, when modeling is conducted as an exploratory venture to guide future collection of field data, it can be a powerful tool for assessing fish populations.
Johnson, B. L. 1995. Applying computer simulation models as learning tools in fishery management. North American Journal of Fisheries Management 15:736-747.
œMike Allen is an Assistant Professor at the University of Florida's Department of Fisheries and Aquatic Sciences.
Contact him at msal@mail.ifas.ufl.edu or visit his website at http://grove.ufl.edu/~msal.
|
|
