Design and implementation of a parallel fish model for South Florida

A parallel, spatially explicit landscape fish population model (ALFISH) is presented hereby to model the impacts of different water management strategies in the South Florida region on the fresh water fish population, which in turn provides the information on the food resource available to wading birds. Adopting a static domain partitioning scheme and using message-passing, the parallel ALFISH model mimics the basic behaviors of fresh water fish based on the interaction of four components - landscape, hydrology, lower trophic biomass, and fish, over a time span up to several decades. The parallel ALFISH model delivers accurate results in simulations. Compared to the average simulation time of the sequential model, which is about 35 hours, the parallel model yields substantial speed improvement. On a symmetric multiprocessor (SMP), the execution time of the parallel ALFISH model on 13 processors is less than 4 hours - a speedup factor of near 9.