6.2.1 Effects of drought in the lower semi-arid zone
6.2.2 Effects of drought in the upper semi-arid zone
6.2.3 Assessment of effects of the 1990-91 drought
Negussie Tilahun (1984) provided a baseline description of Borana households for average rainfall years during 1981-82 (see Section 4.3.1: General household structure and economy in average rainfall years). This work was followed by another survey of 20 households conducted for 21 months from March 1984 to November 1985 coincident with the 1983-84 drought (Donaldson, 1986: p 28).
The families in the drought survey of 1984-85 were distributed among five encampments in three madda (Did Hara, Medecho and Melbana). During November 1984 to November 1985 data were collected weekly for 15 families; observations of family structure and food intake were recorded, and sources and amounts of cash income and expenditure elicited using questionnaires. For the diet study, one enumerator in each encampment recorded total food consumption for members of four families for 24 in/week (Donaldson, 1986: p 52). Amounts of each food consumed were estimated by calibrating appropriate household containers. Gross energy (GE) intake was quantified by multiplying dry-matter intake of each food item by its calorific value from nutritional tables (Donaldson, 1986: p 53). Daily requirements for GE were based on 10.6 MJ GE for an African Adult Male Equivalent (AAME), where a male >16 years of age = 1 AMME; a female >16 = 0.8 AAME; and all youths <16 = 0.6 AAME (FAO, 1973, cited in Cossins and Upton, 1987: p 120). Cossins and Upton (1988a) and Coppock and Mulugeta Mamo (1985) reported on human mortality during the drought based on samples of five and 60 encampments, respectively.
An important aspect of resource management is the use of resident (warra) and satellite (forra) cattle herds (Helland, 1980b). The former are dominated by milk cows, calves and yearlings, whereas the latter consist of dry cows, bulls, steers and others that do not produce milk but that are hardy and able to travel (see Section 5.3.1: General aspects of cattle management). The warra/forra system serves to reduce grazing pressure near wells and encampments by sending away animals that are not contributing to immediate subsistence needs (Helland, 1980b). One objective of the drought study was to examine the dynamics of warra/forra allocation. The age and sex structure and size of warra and forra herds were recorded at five encampments for eight months between November 1984 and June 1985 and research noted, why cattle were shifted from one herd to another and tallied all births, deaths, slaughters, sales and purchases. Age and sex structure and size of warra and forra herds were obtained for November 1983 (in the first dry year) using recall from the same herd owners (Donaldson, 1986: p 29). Description of drought dynamics of cattle dispersal and composition of cattle herds were based either on an average of 621 head (forra; four sample dates) and 509 head (warra; eight sample dates). Pre-drought data were based on several thousand head enumerated in a water-point survey during 1982 (Donaldson, 1986: p 29).
Detailed data on cattle allocation dynamics were supplemented with information from five aerial surveys conducted during March 1983, March and September 1984 and March and June 1985 (Milligan, 1983; Cossins and Upton, 1985: pp 138-142; Assefa Eshete et al, 1987). A follow-up, post-drought survey was conducted during July 1986 (Assefa Eshete et al, 1987). General aspects of the systematic reconnaissance flight (SRF) methodology were reviewed in Section 4.2.6: Grain cultivation. Results were combined from 1983-85 for surveys that had subtle differences in methodology. The main difference in methods between Milligan (1983) and subsequent surveys was that Milligan (1983) collated data according to madda and particular range units, whereas Assefa Eshete et al (1987) collated data in 1984-85 according to six regions distinguished by spatial and ecological criteria. Regions included Did Hara (2400 km2) and Sarite (3125 km2) to the north, Web (3200 km2) and Gayu (3050 km2) to the east, Medecho (1700 km2) in the central region and Dilo (2000 km2) to the west. These are depicted by Assefa Eshete et al (1987: map 1). Data collection of Milligan (1983) was expanded by Assefa Eshete et al (1987: p 4) to include visual estimates of per cent canopy cover for grasses, woody plants, crop cultivation, potentially arable landscapes and hardpan soil surfaces. This chapter will emphasise the effects of drought on population densities of cattle, small ruminants, camels, equines and people as derived from these surveys.
Milk cows are the critical group of animals to pastoralists, and as will be shown, they may be particularly vulnerable to drought (Donaldson, 1986). The subsequent recovery of the cow population after drought in several different regions was analysed by Cossins and Upton (1988a: p 124) using a standard-herd growth model (Upton, 1986b), with parameters for production features characteristic of average rainfall years in 1981-82 (Cossins and Upton, 1987: p 207). Actual recovery of the regional livestock population was assessed by Solomon Desta (nd) in a comprehensive census of animals using over 3000 water points throughout the SORDU sub-project area during five weeks in February and March 1987.
Milk offtake from cattle during the height of the drought was measured one day per week for five months from November 1984 to March 1985 for each lactating cow in the warra herds at the five olla (Donaldson, 1986: p 44). Offtake included milk from morning and evening milkings. Milk intake of calves was not measured. Heart-girth measurements were taken monthly on warra cattle of all age and sex classes during the same period and converted to live weight by the equation:
LW = e-8.88 (HG+1)2.87
where LW is live weight (kilograms) and HG is heart girth (centimetres) [from Nicholson (1983) cited in Donaldson, 1986: p 43]. Estimates of calving percentage from November 1984 to November 1985 were based on records for 176 cows of prime reproductive age at the five encampments (Donaldson, 1986: p 42).
Fewer studies were conducted on other livestock species. Milk offtake was monitored for 129 sheep and 197 goats one day per week from November 1984 to March 1985 at the five encampments (Donaldson, 1986: pp 53-55). A population of 788 goats and sheep were monitored at the five encampments for incidence of mortality, sales and slaughter from November 1983 through March 1985 (Donaldson, 1986: pp 127-128). Complete daily records of milk offtake were compiled for lactations for three camels over 14 months during 1983-84 (Belete Dessalegn, 1985). The camel data are reported in Section 5.3.7.2: Camels.
The effects of drought on the household economy of 96 sympatric Borana and Gabra families (48 each) in the Beke Pond area were evaluated using questionnaires in two separate surveys conducted in 1987 and 1989 (Coppock, 1988; Webb et al, 1992). The main objectives were to contrast how cattle owners fared in comparison to camel owners, and how families poor in livestock fared compared to those wealthy in livestock. Families were stratified into wealthy, intermediate and poor categories according to family size and reported livestock holdings before the drought (i.e. AAME/TLU ratio; where 1 TLU = 250 kg live weight (Jahnke, 1982); see Section 4.3.5.3: Effects of distance to market, wealth and season on dairy marketing). Livestock losses to mortality, sale and slaughter, and gains from births and trade were recorded for each family to quantify loss of pre-drought assets due to drought and post-drought recovery and were aggregated for each wealth strata. Comparative persistence of milk production in cattle, camels and small ruminants at the height of the drought was estimated from verbal recall estimates of the median number of reported lactating stock per household, lengths of lactation and daily milk yields calibrated using traditional containers. Human births, deaths and incidences of migration were recorded, as was reliance on famine relief, traditional social networks, other sources of food, employment, non-livestock assets and trade. Changes in reported livestock holdings and family size during drought were analysed using a factorial ANOVA blocked by ethnic group and stratified by wealth. Changes in livestock holdings were analysed on a percentage basis using SAS (1987).
The 1990-91 drought was again driven by several consecutive rainy seasons with deficient rainfall and had detrimental effects on the production system. Extensive field research by ILCA, however, had been terminated by 1990. Observations on the effects of this drought have therefore, been contributed as personal communications from SORDU and CARE-Ethiopia staff. Interviews of 30 Borana leaders in late 1989 (D. L. Coppock, ILCA; unpublished data; Coppock, 1992b) revealed important insights regarding the nature of cattle population regulation in the system, and these are reported here. Interview methods are described in Section 4.2.3: Livestock marketing.
