Nonparametric analysis of growth in replenishable resource stocks

We consider the problem of estimating growth rates and their determinants f or harvestable biomass using nonparametric regression techniques. Kernel re gression is used for the estimation of an unknown regression function and i ts derivatives with the appropriate degree of smoothing being determined vi a least-squares cross-validation. The application of nonparametric techniqu es reveals structure in the data which is not picked up by traditional para metric growth models.