1: \begin{abstract}

2: Systems of reaction-diffusion equations are commonly used in biological models of food chains.  The populations and their complicated interactions present numerous challenges in theory and in numerical approximation.  In particular, self-diffusion is a nonlinear term that models overcrowding of a particular species.  The nonlinearity complicates attempts to construct efficient and accurate numerical approximations of the underlying systems of equations.  In this paper, a new nonlinear splitting algorithm is designed for a partial differential equation that incorporates self-diffusion. We present a general model that incorporates self-diffusion and develop a numerical approximation. The numerical analysis of the approximation provides criteria for stability and convergence. Numerical examples are used to illustrate the theoretical results.

3: \end{abstract}

4: