The techniques and procedures described in this book have been used and refined in pastoral development projects for more than ten years. Except for RAPS (Resource Assessment for Pastoral Systems), they are mainly a synthesis of methods already in existence, but modified or adapted according to project and end-user requirements. The end-use-orientated pathway has resulted in techniques which are relevant, technically sound and easily understood and interpreted, that in consequence have been well received by local project staff, ranging from scientists and field officers to farmers. The integration of modelling with GIS and resource monitoring programmes is clearly relevant, particularly in addressing issues of environmental protection, resource management and sustainable livestock production.
In addition to its use within projects, RAPS modelling has proved a valuable tool during their identification and preparation. RAPS has facilitated an understanding of the pastoral systems operating within proposed project areas, whether they be extensive or intensive mixed systems. It has also been useful for identifying and quantifying critical components and estimating the potential impact of proposed development options.
Interventions in extensive grazing systems are far from purely technical; in traditional systems, any change, technical or administrative, will affect the community as a whole and have effects far beyond the livestock output and associated income that, simplistically, is often assumed to be a grazing project's main aim. Pastoralists obtain much more than livestock from their environment: their whole livelihood comes from its natural resources so the principal means whereby they meet their domestic needs is through natural resource management. The essence of pastoral development is also to ensure sustainable resource management and to link it with better living conditions; this is becoming increasingly understood by donor agencies. Natural resource management (NRM) is easy to understand as a general concept, but is complicated in practice. It is the product of three distinct systems: the natural system governed by the laws of nature; the user system which determines the use of these resources; and the overall geopolitical situation and the externalities which govern resource utilization. Computer-based modelling is well adapted to the needs of such an integrated approach and is the only way to handle the increasingly complex data needed for project design and monitoring.
The RAPS system has been well tested in the field on four continents and in a wide range of pastoral and agropastoral systems; its efficacy - and that of any modelling system - depends, of course, on the quality of the data used. Careful and adequate data collection is a prerequisite for any pastoral planning. The importance of adequate study of the social framework of pastoral society must also be stressed, and the following quote from Pratt (1997) is appropriate:
"Sweeping assertions about pastoralism arise partly as a by-product of advocacy; of hammering home the point that customary practice is better suited to arid rangelands than ranch-style management. However, they arise also from the dearth of in-depth studies of the workings of pastoral systems and ecological processes. Participatory rural appraisal (PRA) may suffice to get a process started, letting monitoring data guide subsequent inputs, but embarking on development in a state of relative ignorance is always a high-risk strategy."
The main points of the publication are summarized below, and the evolution and refinement of RAPS, which took place during the four case studies, is described thereafter.
This document presents an approach to integrated pastoral resource assessment that has been used for more than a decade in international development projects relating to land and pastoral resource management and environmental protection. The approach reflects the need for a more complete and realistic understanding of grassland-based livestock production systems, including their limitations and dynamics. The main objectives of integrated land, forage and livestock resource assessments are to quantify resource endowment, understand interrelationships between resource components, predict environmental impact, estimate livestock support capacity, and to evaluate development options. The procedures, including data collection, compilation, analysis and results presentation, were designed to accurately represent the pastoral systems. Integrated procedures based upon modelling ensure that assessments of development options relating to land, forage, grassland and livestock are realistically analysed from the viewpoint of the individual components and from a whole-system or holistic perspective.
Traditional methods of pastoral resource assessment fail to account for the complex and interrelated nature of land, forage and livestock. The simple "DM balance" does not account for seasonality, the qualitative differences between forages, differences in livestock feed requirements due to growth and physiological state, nor the effect of forage carryover loss. Modelling overcomes the limitations of traditional methods and significantly improves the understanding of complex pastoral systems. Fast, repetitive analyses spanning time, incorporating variability and more realistically representing the complex interactions within the system, become possible using modelling.
RAPS is a computer-based model designed specifically for use in pastoral development programmes. It aids managerial, policy and development decisions, primarily by analysing land and forage resources for their productivity and livestock support capacity. Development of RAPS began in the mid-1980s, and since then has undergone considerable enhancement and testing in a wide range of environments and countries. As a pastoral resource assessment tool, RAPS has been used in environments ranging from humid mountains to arid deserts. Sheep, goats, cattle, horses, camels and yaks are among the livestock types analysed. RAPS is suited to the assessment of individual pastoral properties, in local and regional development projects, and for land use planning.
The issues, components and methods described in the initial chapters should not be interpreted as limits to an assessment but as the basis of, or framework for, an assessment. The complete range and detail of information will rarely be available for a particular project, therefore the scope and detail of information is presented as a compendium of resource parameters, options and indicators to assist the analysis of grassland systems and the evaluation of development options.
A "universal" methodology or prescribed procedure for analysing grassland-based pastoral resources is not practicable because, while general similarities between grassland-based systems occur, each system has its own unique characteristics and resources. The characteristics of a system include environmental features, social and cultural factors, and markets and political systems.
Resource analyses should not be static, one-off procedures, but rather part of an ongoing programme. Initial databases and analyses provide important benchmarks and interpretations of pastoral systems, but should not be considered final or conclusive. As further data from sources such as grassland resource monitoring programmes become available, and as land use databases are updated from ongoing remote sensing programmes, the pastoral resources can be remodelled and the previous conclusions critically reviewed and refined. Databases representing a sequence of years facilitate the monitoring of trends in the status of forage resources and their livestock support capacity. This provides an early warning of any negative trends and facilitates a more timely and economic implementation of mitigation measures.
The four case studies presented in Chapter 6 emphasized the heterogeneity of pastoral environments and how the approaches used in resource assessment were modified in consequence. The differences in character between studies reflect the local environment, social and cultural factors, and prevailing markets and political systems. Each assessment, therefore, emphasized a distinct combination of resource components and identified a different mix of practicable development opportunities.
Despite the heterogeneity, all case studies involved the following activities:
The Bhutan case study was primarily concerned with the survey of mid- and high-altitude grazing lands. At the outset, a critical review of the Terms of Reference was required because the proposed scale of the task was too large for the time and resources available. Down-scaling and a focus on specific representative areas was essential, and ensured that the results were of sufficient quality to identify and address pastoral resource issues.
Analyses of livestock support capacity were carried out using a forerunner of the RAPS model, and used to identify periods of forage surpluses and deficits, and quantify the predicted impact of suitable development options. Most of the analyses of grazing management, especially comparisons between land units, used simple indices of stocking load. Specifically, standardized livestock units in the form of yak-equivalents were calculated.
The survey of the grassland resources was based upon the height-frequency technique developed by Scott in New Zealand during the 1960s. The benefits of this technique were clearly demonstrated during the survey. The graphical presentation of height-frequency profiles provided a practical and informative method of presentation of the botanical survey results, suited to non-statisticians. As an illustration, Figures 48 and 49 provide a seasonal comparison of vegetation for a single grazing area. The changes in profile between autumn and late winter are obvious, with a dramatic reduction in plants above the 5-cm stratum. Another use of graphical presentation was in the description of altitude and aspect distribution of seasonal grazing lands (Figure 52). The technique enabled quite complex information to be presented in a simple manner.
The study showed that practical avenues for forage and livestock development were constrained by environmental and socio-economic factors and that, in general, forage or grassland development could initially only be small scale. The impact of development was estimated using the modelling technique and indicated that, in the short to medium term, development could only effect a small percentage change in the total forage resource, and therefore any incremental forage had to be used for strategic or critical livestock feeding so as to maximize the benefit of development. Pastoral resource modelling of one of the survey areas (not described in the case study) showed that a previously promoted form of development, namely open enclaves within forests, was neither practicable nor viable.
The Ethiopian study contrasted with the Bhutanese one as it was a component of a major, ongoing land use planning project. As a consequence, the forage and livestock development components were based upon and integrated with comprehensive datasets that were already available. Extensive field surveys and land-user interviews were carried out to provide data of the detail required for the pastoral resource analyses. The presentation of results continued the graphical approach, with an in-house GIS system being used to present information on stocking loads, grazing pressures and vegetation types (Figures 38 - 40). In addition, the unit used in the land use planning procedures - the planning zone - was adopted as the basic unit for the pastoral resource analysis.
The patterns of livestock management within both of the provinces surveyed focused on the use of croplands and the grazing lands in their vicinity. In general, all of the seasonal grazing lands were within the study areas. The analyses focused upon identifying levels of livestock feeding relative to that considered optimal, quantifying the levels of under-feeding and identifying the seasonal patterns of forage balance (Figures 37 and 42). This facilitated estimation of the scale of development programmes required to avoid the regular forage deficits. Such development would not, however, address the extreme fluctuations in forage supply that occasionally occurred.
Causes of the deficits in forage supply were similar in both provinces: an irregular growing season, reductions in available grazing land, pasture of low forage quality, and the significance of low quality crop residues as feed. While forage deficits might occur at any time of year, the commonest period was from April through to October. The modelling procedures defined and contrasted the magnitudes of the deficits between Menagesha Province and Haikoch and Butajira; it indicated the more serious nature of forage deficits in the latter.
Practical development options to help alleviate forage deficits were identified by drawing upon the recent, localized development successes within the provinces. Options selected were generally simple, low cost and compatible with local knowledge and traditional practices. Among the range of techniques for improving forage supply, the most critical and relevant were: undersowing of crops with legumes, introduction of perennial legumes into hay lands, forage tree establishment, controlled-use fodder banks, over-sowing with competitive legumes, agroforestry, establishment of forage reserves, and livestock exclusion areas.
The potential impacts of a range of these options were assessed, using the model, by adjusting "current" forage and livestock databases to reflect yield increases on relevant land types, taking into account realistic levels of implementation relative to land types and the resources available. The analyses showed that a forage production increase of 8% was realistic for Menagesha; for Haikoch and Butajira, a 17% increase might be achieved. The analyses also showed that while all forage development options contributed to the production increase, the most significant was the undersowing of crops. This option was especially relevant in Haikoch and Butajira because there were more crops suitable for undersowing.
An important conclusion was that, while it was realistic to expect a significant reduction in the current forage deficits of both provinces, the scale of development that was practicable meant that it was unrealistic to expect a complete negation of all "normal" deficits.
The analysis of pastoral resources associated with forests in Sudan contrasted with the analyses described in the other case studies because it did not attempt to analyse a complete pastoral system. The Sudan analysis focused upon the pastoral resources associated with forests and how these resources were used during the northward and southward migrations of livestock between the "external" wet- and dry-season grazing lands. The analyses were confounded because the timing and extent of the movements between the northern and southern grazing lands were highly variable and the result of complex management decisions. Furthermore, the narrowing of livestock migration routes through the croplands near the forests created a greater reliance on the forage resources within the forest areas.
An early activity was to determine trends in the forage resources over recent decades, as indicated by changes in land use and livestock management. The information revealed that there had been a massive reduction in the once extensive grazing lands, although too many unknown factors prevented quantification of the actual impact on forage supply.
The second objective of the analysis was to determine the significance of the forest areas as a forage source, within the context of total livestock forage intake. To do this, the timing and extent of use of the forest areas, especially during both the northward and southward migrations, were investigated. The information obtained was overlain with estimates of seasonal patterns of forage intake, and the proportional contribution of the forest areas to annual livestock intake were estimated. The analyses showed that the relative contribution from the forests in the project area to annual forage intake were 26% and 31% (Figures 30 and 31).
A potent reason for using models in assessing pastoral resources is their ability to account for the complex seasonal patterns of forage supply and availability, including differences in characteristics and the patterns of seasonal supply of individual forage sources. The forage sources in the Sudan project illustrate such differences. For example, Figure 32 shows a relatively simple modal pattern of grass and forb production, while Figure 33 represents a more complex multi-modal pattern related to browse production and availability.
The use of convergent subjective indicators to identify the occurrence and pattern of forage resource limitations is a useful adjunct to objective data. In the case of Sudan, interviews with nomads defined patterns of forage shortages, supplementary feeding, crop residue use, poor livestock condition and livestock deaths (Table 11).
During the analyses it became obvious that opportunities for forage development and livestock production improvement associated with the project forests were inextricably linked to the multiple use of the forests, the relationship of the forest to other seasonal grazing lands, and the relationship to the adjacent cropping lands that supply residues. While identifying potential development, a range of options were assessed for their relevance and practicability based on socio-economic, technical and environmental factors. Potential development options were critically reviewed, particularly with regard to local evidence and potential scale of implementation. The resultant recommended development components were divided into priority and conditional activities. The priority components were relatively small scale and designed to encourage local responsibility and participation, and to verify some of the field techniques. The conditional components could only be initiated once priority components were completed and other constraints resolved.
The Gansu pastoral resource analysis represents the culmination of the modelling and analysis procedures that evolved with increased sophistication from the techniques used in the Bhutan case study. The land resource information of the Gansu project was also the most complex, with more than one thousand land units or forage sources described in the databases and simultaneously analysed. Pastoral system dynamics were assessed with a greater level of sophistication.
The RAPS pastoral resource modelling procedures were linked to a grassland resource monitoring programme and a GIS. The resource monitoring programme was used to help set up and calibrate the provisional land and forage databases and to provide a benchmark for monitoring long-term trends in the grassland resources within the project county. A GIS provided the primary land use data used in the compilation of the RAPS land and forage databases, and was used to spatially display the analysis results. That is, a two-way linkage was set up between the model and the GIS. Baseline data on the forage resources, livestock production and land type and use were used directly by both the RAPS model and the GIS (Figure 18).
An example of the increased level of sophistication of the analyses used in Gansu involved the estimation of the effects of intensity of livestock use centred upon livestock water sources within the extensive grazing lands. For each water source, three zones of differing intensities of livestock use were defined and forage production within each of those zones was adjusted according to variations in total annual production and levels of utilization. The availability of reliable data relating to the watering points was crucial to this component of the analyses. The net effect of incorporating zones of utilization on the interpretation of grassland-based forage yields is illustrated in Figure 20.
Inferred variability of forage production between and within years was used to assess the dynamics of the pastoral system and, with certain assumptions, to forecast long-term trends in livestock support capacity (Figures 23 and 24). The long-term forecasts were based upon an assumption of a neutral long-term trend in grassland productivity, and a trend of decreasing grassland productivity. The latter represented the impact of a decline in grassland productivity due to the widespread lowering of the watertable. The data on variability of livestock support capacity were used to create a graph representing estimates of livestock support capacity for deciles of years.
Estimates of the effects of implementing a number of forage development options were carried out using modified land and forage databases. Possible long-term impacts of the development options on the forage resources were forecast, incorporating the presence and absence of a long-term decline in grassland productivity (Figure 27).