From the herd’s perspective

Abstract

The farming systems approach has emerged as unique tool to deal with the enormous diversity of smallholder farming in the tropics. However, the focus was mostly placed on the cropping component of farming systems. In 1996, Seré and Steinfeld devised a global classification, which remains an important basis for classifying livestock farming systems. Although this classification system has been useful at global scale, its application at local level results in a few and rather broad categories of livestock farms. Such farms may not necessarily be similar regarding their size, resource base, enterprise patterns, household livelihoods and constraints. To address this limitation, we analyse the existing farming systems classification and propose a new approach that complements such farm level classifications with a classification of livestock management practices at herd level. We define the livestock management system as a typical combination of herd structure, feeding, and herd management, and their interlinkages. We applied an iterative process of reviewing, coding and classifying primary studies about livestock in Sub-Saharan Africa. The following main classification criteria were thus developed: main feed source, production goal, how animals are confined, and – if applicable – the mobility pattern. On this basis we identified 12 distinct livestock management systems.

For empirical application, we use data from 3 sites in ILRI’s IMPACTLite dataset, that fall into the category rainfed mixed farming systems in arid/semi-arid tropics and sub-tropics (MRA) according to Seré and Steinfeld. However, we identify distinct 3 livestock management systems, that are practices across these locations. These 3 systems describe the management practices of over 90% of animals in the datasets. By focussing on livestock management practices on herd level (rather than the household level). Our approach can support researchers and development practitioners in collecting high-quality data, developing better data structures, understanding livestock production systems, assessing change in livestock management and guide more targeted interventions.