According to assessments made for USAID's Greater Horn of Africa Initiative, over $4 billion have been spent by donors in the Greater Horn of Africa between 1985 and 1992. In the late 1980s, an estimated 71 million people, or 46 percent of the region's population, were chronically food insecure. Food insecurity reduces people's quality of life and fosters the social, political, environmental and economic instability associated with recurring crises. The limitations of the natural environment in East Africa place certain constraints on improving food security.
The chances of drought occurring in parts of the Greater Horn have increased from a probability of one in six years to one in three years for areas affected. Repeated occurrences of drought and high variability in precipitation have reduced the ability of many smallholders to maintain their assets or to respond when conditions are good. Inter-annual variability of rainfall has been increasing in the crescent from Kenya to Sudan, including parts of Ethiopia and Tanzania. Other natural disasters, such as pest infestations and periodic flooding, destroy are-specific production levels. Analysis of these factors supports arguments for a more effective spatially coherent early warning system ion the region, especially as it affects livestock throughout the pastoral and mixed farming region of East Africa. We must think crisis prevention and early warning.
Such programs as USAID's Famine Early Warning Systems (REWS) and the Crisis Mitigation project within the Greater Horn of Africa program have been designed to focus attention on problems of monitoring emerging famine situations and gaining a better understanding of the causes of famine, addressing:
"...the broader causes of disaster by placing a strategic focus on sustainable development while responding to the existing and impending crises in the region...new ways of thinking, new ways of acting and new institutions should be adopted and supported by all partners in the region."The "relief-to-development continuum" approach of the Greater Horn of Africa program has been promoted as the critical framework for this region, considering simultaneous integration of short-term emergency responses and long-term development assistance. Development of multi-scale early warning systems are a primary objective within this regional program.
Crisis prevention involves the ability to foresee and the means to prevent, prepare for and mitigate or resolve crisis and conflict. Effective prevention requires monitoring and analytical capacity at the regional, national and local levels, as well as the ability and desire to respond to warning signs of all kinds. Of the current set of information generated by donor-based monitoring programs, such as the FEWS program, the rainfall and NDVI data offers information on locations of "initiating conditions" while the on-ground monitoring programs of markets, human condition and animal herd situations reflect, mainly, a "post-effect" appraisal system. However, many of the problems besetting livestock (e.g., weight loss, body condition loss) have already occurred before the human eye can detect the response, no matter the level of personal experience. Other human indicators such as upper arm diameter of children under five years or cereal/meat consumption ratios, are further down the food chain within the pastoral ecosystem and offer even more delayed post-effect monitoring of emerging crisis.
The proposed monitoring and analysis system, based on NIRS fecal profiling technology and spatially referenced modeling environments, can add a new dimension to the existing monitoring programs in East Africa. We interviewed several community-based drought preparedness coordinators at the Crisis Mitigation Workshop in Kenya, July 7-11, 1997. They felt that the integration of these proposed tools into their current program would provide an additional 6-8 weeks lead time on the current early warning systems in East Africa and improve the deliberation process of their recommendations to government decision makers (e.g., Drought Preparedness Project-Isiolo District Kenya, Drought and Livestock Project in Marsabit District-Kenya, Turkana District Drought Monitoring Group and Early Warning System for East Africa - Dr. Agastiva - GAO).
These new facilitating technologies would bridge the current "standoff" monitoring systems (e.g., NDVI, meteorological data) and on-ground post-effect surveys of human/animal conditions, vegetation cover, food consumption habits, and movement and market patterns associated with pastoralists in East Africa. The ability to predict supply and decline in milk production, allows more flexibility to design timely destocking strategies that allow the pastoralists to maintain their assets through crisis and allows for greater ecosystem integrity to respond more rapidly after droughts have run their cycle.
USAID awarded the Texas A&M University System (TAMUS) an assessment grant to develop an early warning livestock system proposal as part of the Global Livestock CRSP program. The TAMUS assessment proposal seeks to establish the necessary methodologies, analytical tools, organizations and infrastructures to develop an early warning system for livestock nutrition and health as an integral part of existing early warning systems for drought and famine in East Africa, particularly FEWS, NOAH-USGS,FAO-GIEWS, ASARECA'a in-country programs, IGAD and international donor-sponsored regional drought preparedness programs.
There are a series of technologies, developed by TAMUS, that provide critical support to the effort; including the advanced GIS-based Spatial Characterization Tool (SCT), the NIRS fecal profiling technology for nutritional assessment of free-ranging livestock, the NUTBAL nutritional management system, the APEX cropping systems model and the PHYGROW multi-species growth/hydrology grazingland systems model. Each of these analytical tools are unique to the assessment team (AT) assembled at Texas A&M University. Dr. Corbett is the co-developer of SCT. Dr. Jerry Stuth is the coordinator for the Ranching Systems Group that developed the fecal NIRS profiling system, the NUTBAL nutritional decision support system and the PHYGROW grazinglands modeling environment. Dr. Jimmy Williams, along with Dr. Paul Dyke, are developers of the APEX farming systems model and will be leading the effort to integrate ENSO infromation in the project.
When the analytical capacity of this research group is blended with the robust network of professionals and organizations in East Africa, which was developed during the assessment phase of the program, the formula for successful development and implementation of a regional livestock early warning system emerges. Our challenge is to package these technologies in a manner that is adaptable to East African conditions and organize a critical mass of personnel and institutions (in-country and internationally) taht provides high-quality information on trends of well-being of livestock and projections to emerging crisis. These provisions and projections provide added value to the monitoring and assessment systems in East Africa, improving the overall response time of those programs.
The thrust of the program would be to provide information that reaches all levels of decision making, from community-based programs to national policy makers and the international monitoring programs. After conducting three planning meetings in East Africa, between January and July of 1997, our assessment team of 17 professionals from 5 East African countries and key key staff from a series of GO and international donor programs agreed to the following objectives for this GL-CRSP:
NIRS/NUTBAL-Nutritional Management System. Although NIRS technology has been used extensively in ariculture to directly detect nutritional quality of hand-plucked forage species this research program utilizes a new technique developed by the Ranching Systems Group at Texas A&M University. It predicts dietary concentration of protein and digestible organic matter from the spectral charcteristics of feces which, in turn, reflect the end products of the digestive process, i.e., the technology allows prediction of diets selected by the animal under their unique forage conditions.
Lyons and Stuth (1992) first proved that identification of key organic chemical bonds in the feces, through NIRS, could successfully predict dietary constituents which formed the precursors to the formation of those bonds in the feces. This concept resluted in the develpment of NIRS equations for scanning feces and predicting the diet crude protein and digestible organic matter of free-ranging livestock (Lyons and Stuth 1992, Lyons et al. 1993, with the NUTBAL nutritional balance analyzer decision support system, a new capability has emerged to assess the nutritional status and weight change for free-ranging animals (Tuth and Lyons 1995). A complete overview of the system can be found at the URL http://cnrit.tamu.edu/ganlab on the worldwide web. Currently the Grazingland Animal Nutrition Lab at TAMU serves over 1400 ranches in 42 states in the USA and has been in service for three years.
NIRS fecal profiling equations are derived from a reference data set of known diet values of animals in relation to the spectral reflectance of feces on animals grazing the forage for which diet quality is known (esophageal fistula or stall fed). Statistical techniques (modified stepwise multiple regression or partial least squares analysis) allow exploration of relationships of reflectance/absorption values of a diverse set of chemical bonds in the feces and wet chemistry values of consumed forage. Dr. Stuth has assembled a reference set of over 2000 samples from diverse geographical locations, plant types and environmental regimes throughout the USA, Africa (Ethiopia, Nigeria, Niger), Australia and Argentina.