Effective dimensionality and factors affecting crop-livestock integration in West African savannas: a combination of principal component analysis and Tobit approaches

Crop‐livestock integration (CLI) to improve natural resource management for increased productivity is evolving in Nigeria and West Africa. Processes in the evolution and factors that influence it still need to be well understood. This article proposes and tests a new framework for measuring the multiple dimensionality of CLI. The framework derives a CLI index using the principal components of its most common single measures; it develops geographic information systems (GIS)‐based village‐level ecological and market factors; and it estimates parameters of factors affecting CLI using the derived index as the dependent variable in a Tobit model. The framework is tested using empirical data from 634 farm households in 11 geo‐referenced villages in the Sudan savanna (SS) and northern Guinea savanna (NGS), Nigeria. Along a north–south (SS to NGS) axis, CLI initially increases, peaking around 11.2°N, and then declines. This latitude probably identifies the boundary below which disease challenge constrains traditional livestock production and CLI. This polynomial pattern of CLI is contrary to an expected linear increase along lines of perceived potentials for rain‐fed crop production. Household resources, GIS‐derived village‐level market factors, and institutional factors also significantly affect CLI. Ecological and institutional factors have most impact on the probability of adoption and use intensities of CLI. The incorporation of GIS‐derived market factors with household and institutional variables in an econometric model offers new opportunities for assessing patterns of evolution of CLI, comparing results across sites, and targeting recommendation domains objectively. A comparison with results from more common methods of running independent models for individual indicators of CLI shows that this new framework is an effective way of reducing the multiple dimensionality of CLI to gain quicker, well‐focused knowledge of the processes of agricultural intensification.