1. Environmental degradation, livestock and environmental management
2. Results of the present study
Functional vs ecological degradation
Diagnostic vs longitudinal evidence
Stocking rates and degradation
Density, integration and sustainability
The present Study is undertaken at a time when reservations about the conventional view of degradation in the SAZ are becoming commonplace, and both its linkages with management, and the evidence for its progression are being questioned (see, for example, Ahlcrona, 1988: Mortimore, 1989, a,b; Nelson, 1988; Olsson, 1985; Sandford, 1983). It is difficult to reconcile this perspective with the orthodox view of desertification as a man-made and irreversible process consuming large areas of productive land every year (UNEP, 1977: Tolba, 1986). In mixed farming areas, both the degradation of arable land under cycles of cultivation, and the degradation of rangeland under various levels of stocking, are issues. Relevant to both cropping and animal husbandry, as well as to the status of the environment in general, is the management of the woody vegetation.
Environmental status has traditionally been left to ecologists to define, even though it has long been recognised that low nutrient status in cultivated soils is primarily an aspect of their economic management, and may be remediable given the right incentives. Work on common access grazings in the Communal Areas of Zimbabwe has challenged conventional notions of carrying capacity and overstocking (Thiesen and Marastha, 1974; Sandford, 1982; Cousins et al, 1989; Scoones, 1989). Optimum stocking levels for commercial beef cattle may be lower than those of dairy herds whose functions include household subsistence, investment, breeding, manure and traction, and which are fed partly on residues and browse. What may appear as overstocking to the ecologist may be economically efficient to the stockowner. Alteration of the vegetation is not irreversible. The opportunity costs of alternative forms of management are more relevant to an understanding than a comparison between observed and potential vegetation; also, annual primary productivity may be higher under intensive grazing. It appears necessary to distinguish between ecological degradation (in the sense of the loss of primary potential productivity) and a functional, remediable degradation that reflects the economic rationale of a particular management system under certain constraints of capital, land or labour.
Reliance on diagnostic evidence (e.g., a substitution of annual grasses for perennials, of unpalatables for palatables, the appearance of bare ground, gully erosion, etc.) supported by intuitively convincing hypotheses linking management (or mismanagement) with degradation, has tended to obscure the scarcity of longitudinal data that would allow the rate and nature of degradation to be established. Proper examination of such data, increasingly available from the interpretation of air photos and earth satellite imagery, exposes many ambiguities and tends to emphasise the impact of rainfall fluctuations. Meanwhile the efficiency of some pastoral nomadic systems, in terms of energy conversion under conditions of fluctuating climatic stress, is becoming better understood (Western, 1982; Coughenour et al, 1985). Such studies would be appropriate in the SAZ also.
If the condition of the vegetation is not always a reliable guide to the quality of management, neither can stocking rates be used as a short cut to assessing degradational status. Overstocking (however defined) may occur at any point on the scale of farming intensity. If it truly occurs, then unless the livestock are fed from imported feed, there must be either cumulative ecological degradation, losses from sale or starvation, or both. It is a transitional, not a permanent condition. The persistence of livestock populations that are supposed to be much higher than local carrying capacities for decades, if not generations, is therefore of obvious significance.
Carrying capacity estimates tend to be related to the area of available land rather than to the total capacity of the managed ecosystem to feed livestock (natural grazing, browse, crop residues, weeds, fodder crops, field boundary plants, irrigation canal-sides, etc). Arable encroachment on grazing land has major implications for cattle management, even though the crop residues may support more LUs/ha on an annual basis than the natural grazings. A switch into small ruminants, however, may sidestep such problems, and there are mixed farming systems where comparatively high small ruminant stocking levels are maintained, although natural grazings have all but disappeared.
The following model is advanced linking human and animal population densities, farming intensity, crop-livestock integration and environmental management.
The first stage of the model is a low population density associated with farming enclaves and a predominance of grazing land. With increasing human population density, which is expressed in increasing availability of family labour, and given the economic conditions (uncertain market supply/prices of foodstuffs) that encourage a subsistence priority in the household economy, arable land expands at the expense of natural grazings. As the human population rises, and given the multipurpose value of livestock, so does the livestock population, subject to constraints imposed by household poverty, disease or starvation in drought. Diminishing natural grazings may favour small ruminants at the expense of cattle, or necessitate transhumance. The loss of natural woodland encourages the protection and eventually planting of browse (especially valuable for small ruminants) and other trees on farmland. Increasing frequency of cultivation (increasing labour inputs/ha) necessitates the use of animal manure and enhances this function of livestock, as well as favouring grain/legume crop mixtures. Crop residues increase in importance relative to natural grazings as sources of fodder. Leguminous trees, providing dry season browse as well as benefiting crop growth, increase in importance in the system. Trees and planted field boundaries (also sources of fodder) stabilise soil wash and reduce aeolian activity. The rising scarcity of land intensifies individual claims to access rights, and eventually raises the market price of land and the frequency of sale relative to other forms of transfer. Labour and capital investments are made in order to raise the productivity of land. Labour diversification out of agriculture, in response to alternative income-earning opportunities, need not cause the system to decline owing to the investment value of both the land and the livestock. Primary productivity of the system is low (constrained by the manure supply) but stable, and degradation is held in check.
This model provides a rationale for linking sustainable environmental management with high human and livestock densities, in contrast with much conventional wisdom that sees-rising tendencies as a certain road to environmental degradation. According to such a model, degradation is more likely to occur earlier in the sequence, if an increasing human population density is not associated with the introduction of intensive practices and crop-livestock integration.
The implication is that the link between the characteristics of mixed farming systems and environmental degradation, or sustainability, should be sought in the management of intensification, achieved through the integration of crops, livestock and (probably) trees.
Summary of chapters 2-4
Linking the systems typology to environmental management
Suggestions for further work
Chapter 2 reviewed seven available principles on which a typology of mixed farming systems in the SAZ may be based, and concluded that the most useful general principles (though not necessarily for all users) are the linked ones of crop-livestock integration and farming intensity.
Chapter 3 developed a regionalisation of the SAZ of sub-Saharan Africa in four orders of increasing scale. The first order subdivision is between W & N and E & S geographical regions. The second order subdivision follows LGP Zones by country, using data from the FAO's Land Inventory and Population Supporting Capacities project. The third order is according to agroclimate, employing moisture, modality, and monthly regimes. This subdivision exposes anomalies in the SAZ as defined by the LGP isolines of 75 and 180 growing days, and a functional redefinition is proposed. The fourth order subdivision develops a set of 83 Environmental Units based on a synthesis of mapped data from 8 available published sources (see Appendix 2).
Chapter 4 reviews the characteristics of mixed farming systems through a sample of published and unpublished literature, whose limitations for this purpose are noted. From 65 system characterisations reviewed, 43 case studies are systematically analysed on 32 variables (see Appendix 1), including scores for 8 integration variables, and for farming intensity as indicated by the cultivated percentage. The review provides a basis for a classification of systems, but the literature provides a very weak basis for estimating the territorial extent, livestock and human populations of the systems (Term of Reference 6: see Appendix 6).
It has not proved possible to identify direct and unambiguous linkages between system characteristics and trends in environmental degradation, or in other words, to link ecological sustainability to properties of system management on the basis of measured observations.
(1) The distribution of case studies (reviewed in Chapter 4) on the map of Environmental Units (Figure 6) leaves many EUs unrepresented by a system characterisation. A larger sample is needed. However the literature is unevenly distributed and many EUs will remain unrepresented even if a more thorough search is undertaken.(2) There is little reason to suppose that a system case study is always reliably representative of the EU in which it is situated. There is also little reason to expect that there is any general correspondence between system properties and EUs, since some of the criteria used for delimiting the EUs may have marginal significance for system management.
(3) No clear pattern of degradation risk or status emerges from the mapping of the EUs. This is partly because the sources are inadequate - the assessments of degradation risk are only available for areas north of Lat. 2°N, and elsewhere the broad categories of desertification risk provide an insufficiently detailed guide. More fundamentally, it is because actual degradation is linked to management as well as to environmental characteristics.
(4) Characterisations of mixed farming systems often ignore questions of sustainability, or deal with them in a superficial way. This arises from the differences in the professional skills required for the investigation of socio-economic, technical, and environmental variables, and from the relatively late arrival of sustainability on the research agenda of management-related studies.
(5) Unlike the EUs, the mixed farming systems identified in the present study do not comprise a spatially complete set, which, if it were available, would invite correlation with the map of EUs. Not only are many systems unrepresented in the literature, but of those that have been described, the territorial limits are rarely known.
Because it has not proved possible to link in a systematic way the organizational (management) aspects of systems directly to reliable indicators of environmental status, as set out in Term of Reference 4 (see Appendix 6), it has been necessary to proceed independently with the generation of Environmental Units and with the taxonomy of mixed farming systems.
However, in setting out a rationale for both of these operations, the present study provides a basis for further work.
Environmental classification
(1) A refinement of the EUs as defined and classified. Further subdivision is not considered useful since it would increase the number of units in the SAZ as a whole, and in some individual countries, to a level that would be complex. On the other hand, amalgamating the EUs into a smaller number would increase the internal variability of the units, and it is preferable, if a smaller number is required, to use divisions based on a smaller number of criteria, i.e. the third order (LOP) subdivisions or the second order (agroclimatic) subdivisions.(2) Further analysis of the FAO Land Inventory data with a view to (a) revising the system of 83 EUs derived from conventional published maps, and extending the scope of the accompanying inventory, and (b) linking the LGP sub-zones with livestock-related variables such as biomass production in natural pastures, and the availability of crop residues as fodder.
(3) Exploration of the GEMS system's capability for supplementing the FAO's LGP zonation and the system of EUs employed here. It may prove possible in future to substitute a computerised GIS-based regionalisation.
Systems typology
(4) An extension of the systems review to a larger number of cases, an intensification of selected cases from additional literature, and the filling of some gaps in the map of mixed farming systems. Given a larger and more complete set of case studies, systematic analysis of the patterns of similarity may be attempted.(5) Cross tabulation of selected system characteristics in order to explore in a preliminary way the existence of linkages between, say:
stocking rates (LUs/km2) and integration scores
cultivated percentages and human/livestock densities
access rights and market impact
livestock types and economic integration
system integration and environmental sustainability or degradation
investment value and effects of drought
(see the key to Appendix 1)This has not been attempted in the present study. It would be desirable to strengthen the review of the systems before doing so.
(6) Incorporation of livestock census data at the national level into the systems typology (and EUs), where available. The National Livestock Census of Nigeria, presently in progress, offers an opportunity.
The difficulty we have experienced in identifying clear patterns linking the systems typology with the environmental variables, notwithstanding the priority of the degradation-sustainability issue in the SAZ, underlines the need for both (a) more system characterisations and (b) a format to expose such linkages on a compatible basis.
