Comparing performance of five nutrient phytoplankton zooplankton (NPZ) models in coastal lagoons

Modeling nutrient cycle dynamics in the open ocean based on the well-known interactions among nitrogen, phytoplankton, and zooplankton (NPZ) is well-established. Difficulty arises in applying this methodology to multiple coastal systems because of differences among geography, water chemistry, microbial interactions, weather patterns, and sedimentary nutrient cycling. Current trends in ecological modeling are toward more complex modeling relationships and mathematical functions. Four published NPZ models and a new model are compared based on their varying number of equations, mathematical complexity, and required parameters. The new model adapts NPZ interactions to shallow estuary systems by adding a benthic consumption component as a nutrient producer consumer (NPC) system. Each model is calibrated and validated for two bays in the western Gulf of Mexico: San Antonio Bay, TX, USA, and Copano Bay, TX, USA. Daily riverine nutrient inputs are used as the model driver while historical measurements are used for calibration and validation. The five models are compared for their ability to simulate the observed bay response of primary production in both locations. The new model simulates primary production closer to measured observations than other models because of the inclusion of benthic consumption dynamics. However, differences of equation complexity between NPZ models had no relationship to overall goodness of fit in study area. Models produced nearly identical results regardless of different relationships and mathematical formulas. Although every location is unique, this study shows that adding mathematical complexity may only provide marginal gains in practice while decreasing portability when extending NPZ models for the coastal zone.