Abstract
Animal welfare assessment in livestock systems requires integrating productive performance with internal system organization. Here, animal welfare is formulated as an emergent system-level property arising from the interaction between productive homogeneity and relative performance. Within an Information Theory framework, productive homogeneity is quantified using Shannon–Lad extropy, while performance is represented through a relative productivity factor defined with respect to a reference level. These components are combined into a Productivity–Extropy Welfare Index (PEWI), derived from explicit structural and functional properties without imposing ad hoc functional forms. The index was applied to five consecutive production cycles (2021–2025) of a livestock production unit using routine production records. Results show that extropy remained close to its theoretical maximum across cycles, indicating stable internal organization, whereas welfare variations were primarily driven by changes in relative productivity. Symmetric complementary Kullback–Leibler divergences were employed to characterize inter-cycle structural transitions, revealing reorganizations not detectable through homogeneity measures alone. Comparison with a conventional qualitative welfare index (CWI) showed consistent temporal trends, while the proposed index retained higher resolution under high-performance conditions. Overall, the PEWI provides a parsimonious and operational framework for welfare assessment in data-limited systems, situating animal welfare analysis within the context of non-equilibrium complex systems through the entropy–extropy symmetry.

This publication has 23 references indexed in Scilit: