Modeling primary production: Developing the BZI model from the production-irradiance (P-I) curves

• Published in: Ecological modelling Vol. 458; p. 109707
• Main Authors: Sun, Detong, Doering, Peter H., Phlips, Edward J.
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 15.10.2021
• Subjects: BZI model; Light functions; Photic depth; Primary production; Subtropical; Subtropical; Photic depth; Light functions; BZI model; Primary production
• DOI: 10.1016/j.ecolmodel.2021.109707
• ISSN: 0304-3800

•The empirical linear BZI model was derived by vertically integrating a linearized production-irradiance (P-I) curve applying beer's law for light attenuation.•Slope of the linear BZI model was found to be dependent on the slope of the P-I curve.•The intercept often seen in the linear BZI regression was found to be caused by fitting a nonlinear P-I curve with a linear one.•A simple correction term was obtained for water depth shallower than the photic depth.•A nonlinear version of the BZI model was obtained by vertically integrating the nonlinear P-I curves.•The nonlinear BZI model is dependent on two parameters: the maximum production rate and the initial slope of the P-I curve. It can through origin naturally, i.e., zero production at zero light.•Both the linear and the nonlinear BZI models were successfully applied to three shallow estuarine systems in Florida. The nonlinear model yielded similar or slightly better results than the linear model. The linearity of the BZI (biomass, photic depth and irradiance) regression model for the estimate of depth-integrated primary productivity in the water column is evaluated. It is shown that the linear model can be derived from traditional production-irradiance (P-I) curves by assuming a linear production-irradiance relationship. A correction tem is obtained for water depth shallower than the photic depth. The analysis revealed that the slope of the linear BZI model is determined by the slope of the P-I curve. Following similar steps and assumptions, some of the well-known nonlinear light functions such as Steele's function, Smith's function and the Monod function were analytically integrated yielding nonlinear BZI models which go through the origin naturally allowing zero productivity at zero light. The nonlinear model integrated from Steele's function was successfully applied to three subtropical estuaries in Florida: the St. Lucie Estuary on the east coast, the Caloosahatchee River and Estuary and the Escambia Bay on the west coast. Despite being more faithful to theory (zero intercept), the non-linear version yielded only slightly better results than the linear model.