Analysis of Changes in Mean Milk Production in Turkiye Using Nonlinear Models and Assessment from the Perspective of Food Security

In this study, annual raw milk production data in Türkiye for the period 2001–2025 were analyzed using nonlinear models. The Gompertz, Gamma, Brody, Bertalanffy, Cubic, Cubic Piecewise, Logistic, Wilmink, and Wood models were compared in terms of their capacity to predict production changes. Model performance was evaluated using mean square error (MSE), adjusted coefficient of determination (R̅²), accuracy factor (AF), bias factor (BF), Durbin-Watson (DW), Akaike information criterion (AIC), corrected Akaike information criterion (AICc), and Bayesian information criterion (BIC). The results indicated that the Cubic model reflected changes in milk production most accurately, exhibiting the highest accuracy and lowest error. Cubic Piecewise and Logistic models ranked second. Gamma and Wood models showed low performance due to high error margins. The findings demonstrate the effectiveness of nonlinear modeling in Türkiye’s milk production and highlight the applicability of projections for food security strategies.