5.4.1 Cattle management
5.4.2 Cattle production
5.4.3 Cattle mortality and health
5.4.4 Cattle productivity and watering frequency
5.4.5 Cattle growth and implications for breed persistence
5.4.6 Small ruminants
5.4.7 Camels and donkeys
Typical patterns of cattle management, including separation of adults and immatures, nightly corralling of adults, milking cows with the calf at foot, allocation of milking quarters to human or calf consumption, intensive hand-rearing of calves, water restriction of adult cattle and calves in dry seasons, splitting cattle into home-based and satellite herds and differentiation of management roles according to sex and age among household members have been commonly observed to various degrees in similar systems throughout sub-Saharan Africa (Dahl and Hjort, 1976; McCabe, 1983; Wagenaar et al, 1986; de Leeuw and Wilson, 1987; Massey, 1987; Coppock et al, 1988; Grandin et al, 1991). Alberro (1986: p 37) commented that the Boran sustain an ecological balance in their system in par by not desiring to accumulate large numbers of cattle and by slaughtering unwanted calves in the dry season. This was speculated to be a management adaptation to a restricted resource base. It is notable that there is no support for such speculation based on observations reported here. Like other pastoralists, Borana herd owners seek to accumulate cattle for various social and economic reasons (see Section 4.5.4: Traditional marketing rationale).
The standard practice of watering cattle once every three to four days in dry periods in response to local constraints of forage, water and labour is somewhat unusual, but such practices are permitted by the relatively cool temperatures (Section 2.4.1.4: Climate, primary production and carrying capacity) which facilitate maintenance of cattle water budgets by reducing evaporative losses for thermo-regulation (King, 1983). In interviews with Borana elders, the hypothesis that restricted watering of cattle is a relatively recent innovation due to population pressure was rejected (D. L. Coppock, ILCA, unpublished data). According to the elders, the practice has apparently lasted for many generations. This suggests that labour and/or water limitations have been historically important in the region.
The proportion of females to males in Borana cattle herds found by Mulugeta Assefa (1990: p 19) of 74:26 is in agreement with previous surveys by AGROTEC/CRG/SEDES Associates (1974h) who found a ratio of 69:31. This is also similar to results for other pastoral and agropastoral groups throughout sub-Saharan Africa with an average of 65% females: (de Leeuw and Wilson, 1987). The high proportion of females is thought to help stabilise milk production by offsetting the longer calving intervals characteristic of these systems (Jahnke, 1982 cited by Mulugeta Assefa, 1990: p 20).
The average age at first calving of four years found by Nicholson and Cossins (1984) and 4.5 years by Mulugeta Assefa (1990: p 21) also agree with earlier surveys done on the Borana Plateau by AGROTEC/CRG/SEDES Associates (1974h) and Faulkner (cited in Donaldson, 1986). These figures are higher than the mean of 3.6 years for Bos indicus found in a number of traditional systems reviewed by Mukasa-Mugerwa (1989: p 59). Age at first calving is highly influenced by the nutritional environment (Mukasa-Mugerwa, 1989: p 59). Mulugeta Assefa (1990) also found cows among the low-producer class that reportedly had their first calf two months earlier than the high-yielding ones, which may underlie some fundamental differences among cow classes (see below). Nicholson and Cossins (1984) noted that age of first calving of improved Borana cattle on Kenya commercial ranches was only 2.4 years. This may be a particularly exceptional situation, as Pratt and Gywnne (1977: p 149) gave a range of 3 to 4.3 years for Borans managed on ranches and research stations in East Africa. Trail and Gregory (1981), in a study of 10 years of data from the Kenya Rift Valley, found 3.3 years to first calving for Borans. It may thus be concluded that age at first calving of Borana cattle on the plateau is at the upper end of the range for animals reared on traditional settings. The ability of some Borans to calve markedly earlier under improved management is presumably related to the secure forage base and upgrading of the genotype.
Estimates of weights for mature Boran cattle vary markedly. Nicholson (1983b) found an average size for male cattle at Sarite of 282 kg, less than the mean of 318 kg for mature males reported by Alberro (1986: p 33). The animals studied by Nicholson were probably younger than those considered by Alberro. Nicholson and Cossins (1984: p 19) reported that it takes five years to produce a maximum-sized male of 350 to 450 kg and this is in contrast to Church et al (1957) who estimated the time to be six to seven years. Alberro (1986) also noted a maximum size of about 500 kg for males and an average mature weight of 225 kg for females on the plateau. Genetic upgrading of animals under conditions reportedly produces mature males and females that weigh from 550 to 850 and 400 to 550 kg, respectively (Alberro, 1986: p 33).
Annual calving rates of 67% (Mulugeta Assefa, 1990: p 21) and 75% (Nicholson and Cossins, 1984: p 2) in average rainfall years exceed earlier estimates of 50 to 60 % by AGROTEC/CRG/SEDES Associates (1974h). Considering uncontrolled breeding and the variable environment, a rate of 67 to 75% is at the upper end of the scale for either traditional production systems (34 to 69%) or for research and commercial conditions (53 to 82%) in Africa (Mukasa-Mugerwa, 1989: p 643. Uncontrolled breeding is the norm in African traditional systems (Wagenaar et al, 1986; de Leeuw and Wilson, 1987; Mukasa-Mugerwa, 1989: p 59; Wilson, 1989) with the hope of having calving distributed through all seasons to provide milk year round. Higher calving rates are correlated with higher rainfall (Cossins and Upton, 1988a: p 122; Mukasa-Mugerwa, 1989: p 663 which underscores the fundamental role of the nutritional environment in regulating conception. Disease, genetics and management may also play key roles otherwise (Mukasa-Mugerwa, 1989: p 63). De Leeuw et al (1991) noted for Kenya's Maasailand (which, like the southern Ethiopian rangelands, is also under a bimodal rainfall regime) that with two seasons of average or above-average rainfall per year, the calving percentage would come to 75%. One average or above-average rainy season per year would yield an annual rate of 58% while two below-average rainy seasons in a year would yield 43%. This perspective is very relevant for the Borana Plateau, but it must also be qualified by stocking rate and competition for forage (see Section 7.2: A theory of local system dynamics).
Mulugeta Assefa (1990) found markedly lower calving rates for cows held by poor households (56%) versus those of middle-class or wealthy households (70-71 %). This could be related to the higher milking intensity he observed within poor households. The poor reportedly milked each cow more often, using more quarters and one month longer than wealthy families. Similar patterns of milking pressure were found among the Maasai (de Leeuw and Wilson, 1987: Grandin, 1988). More intensive milking has been noted to increase the period or lactation anoestrus in zebu cattle, the hormonal mechanism is reviewed by Mukasa-Mugerwa (1989: p 72). Evidence from other pastoral systems suggests milking effects on lactation anoestrus are minor. De Leeuw et al (1991) found that when milking is prolonged by one month the calving interval increased only by three days, as cows milked for four or five months calved 20 or 21 months later, respectively. For Mali, de Leeuw and Wilson (1987) estimated that every additional month of milking increased calving intervals by 10 days.
The generally higher level of milk deprivation of calves of poor families (see below) may also have a long-term effect on calving rates. Despite the ability of calves to compensate for the effects of early milk deprivation and reach maturity at a similar time across all wealth classes (see Section 7.3.3.4: The calf Prospects for growth acceleration), they may be impaired in terms of subsequent breeding success. Mulugeta Assefa (1990: p 23) cited Clutton-Brock et al (1987) who noted that the lifetime breeding success of different cohorts of red deer (Cervus elaphus) were strongly related to the differences in nutritional conditions of the years of their birth.
Calving intervals overall were estimated at 14 to 15 months (Nicholson and Cossins, 1984: p 3; Mulugeta Assefa, 1990: p 23) which agrees with AGROTEC/CRG/SEDES Associates (1974h). This estimate is lower than the 20-26 months reviewed by de Leeuw and Wilson (1987) and is at the lower end of the 12 to 26 month range for zebus proposed by Mukasa-Mugerwa (1989). Calving intervals on the Borana Plateau are strongly influenced by the environment, as over 90% of conceptions and births occur either in the long or short rains or shortly thereafter (Nicholson, 1983a; Mulugeta Assefa, 1990: p 24; Sovani, 1990). This effect of the environment has also been observed on ranches (Trail and Gregory,1981). The different calving rates found among wealth classes by Mulugeta Assefa (1990) also imply that calving intervals may vary at most from about 14 months for animals owned by the wealthy to around 20 months for those owned by the poor.
Assuming a succession of average rainfall years, Nicholson and Cossins (1984) reported that having reached four years of age a heifer will likely produce 6 to 6.5 calves in a reproductive lifetime of 8 to 8.5 years. Culling commonly starts around 12 years of age. A few animals in their survey were between 15 to 20 years old and reportedly still calving. Anecdotal observations of Alberro (1986: p 34) agreed with this assessment. This is a marked increase in performance over zebus in the Ethiopian highlands that reportedly yield only three to four calves over a reproductive life of 7.6 years Mukasa-Mugerwa et al, 1989). Data from Trail and Gregory (1981) suggest that improved Borans under conditions produced seven calves over a productive life of 7.7 years, only moderately higher than that observed on the Borana Plateau. Data from Trail and Gregory (1981) also indicated that the cows were culled around 11 years of age, which on average is only several years younger than that observed by Nicholson and Cossins (1984).
The results of Nicholson (1983a) for average calf birth weight (18 kg) and a low ADG for nursing calves (136 g/day) are similar to estimates for pastoral systems elsewhere (Wagenaar et al, 1986; de Leeuw et al, 1991; R. T. Wilson, ILCA, unpublished data). The estimates of median milk yield (843 kg) and lactation length (320 days) by Nicholson (1983a) were higher than those of Belete Dessalegn (1982; 488 kg over 249 days) or that of Mulugeta Assefa (1990; 436 kg over 237 days). These substantial differences reflect differences in research methods and the effects of the different sites and years, Nicholson (1983a) probably provides the most accurate baseline set of data. Nicholson and Cossins (1984) stated that milk production ranged from 680 to 1000 kg/lactation.
It is noteworthy that milk output of Borana cattle under highly seasonal range conditions appears: (1) similar to that of native came of the Ethiopian high lands under conditions; and (2) higher compared to native cattle of the Ethiopian highlands under smallholder conditions. Kiwuwa et al (1983: p 13) reported yields for Arsis and zebus as averaging 869 kg over a 287-day lactation. Mukasa-Mugerwa et al (1989) noted that zebus under traditional management yielded about 524 litres over a 239-day lactation, roughly 62% of that for Borans reported by Nicholson (1983a).
In sum, these contrasts give some credence to claims that the indigenous Boran is, even at minimum a moderately good and efficient milk producer (Pratt and Gwynne, 1977: p 147; Alberro, 1986). Pratt and Gwynne (1977: p 149) indicated that Borana cows under management in East Africa may yield up to 2641 kg/lactation, over three times the mean value for pastoral management found by Nicholson (1983a). Such production levels are presumably related to genetic effects as well as a secure forage base resulting from moderate stocking rates.
Mulugeta Assefa (1990) noted that there are three classes of milking cows in the Borana system that vary significantly with respect to several production aspects. Although the lowest producers were milked less intensively over the shortest length of time, compared to the highest producing cows, they had a slightly lower age at first calving and a shorter calving interval that could lead to a slightly higher lifetime productivity. The shorter calving interval could be due to a reduced period of lactation anoestrus; the high producing animals sacrifice more of their body stores to support milk production and the next conception is thus delayed (Mulugeta Assefa, 1990: p 24). This difference among cow classes is most likely to be genetic but it is doubtful that the Boran have a breeding regime that promotes one class of cow over another given that animals of various owners are herded together and breeding is uncontrolled. It is more likely that selection is fortuitous. If the high producing cows are more susceptible to death during droughts (Cossins and Upton, 1985; ILCA, 1986: p 25), a selective breeding regimen would be fruitless over the long teem, anyway.
In their study of different cattle breeds in a seasonal environment, Vercoe and Frisch (1983) argued that Bos taurus breeds and their crosses were more productive under favourable conditions than Bos indicus because the former have a higher basal metabolic rate. However, Bos indicus was postulated to be more likely to survive unfavourable enviromental conditions. This illustrates the important trade-off between productivity and survival and may apply to the different classes of milk cows in Borana.
The finding of Nicholson (1983a) that the shape of the lactation curve of Boran cows in the rangelands as being typically bimodal fluctuating markedly with season, is in sharp contrast to the classical lactation curve (Wood, 1967). This underscores the great influence of the nutritional environment under traditional production conditions. But with shifts in animal density, the bimodal lactation curve may vary in its occurrence from year to year (see Section 7.2: A theory of local system dynamics).
Nicholson (1983a) found that milk offtake for human use ranged from 30 to 40% of production and that the average daily Offtake was about 0.92 litre/cow. Work done by Holden et al (1991) and Mulugeta Assefa (1990) indicated similar average rates of Offtake overall, but again considerably more resolution was provided by differentiating households into wealth classes. The findings of Nicholson (1983a) are remarkably similar to those for Maasailand (40% Offtake with 0.94 litre/cow/day; de Leeuw et al, 1991), a transhumant system in Mali (20 to 35% Offtake with 0.7 kg/cow/day; Wagenaar et al, 1986) and other pastoral and agropastoral systems reviewed by de Leeuw and Wilson (1987; up to 1.1 kg/cow/day). Even smallholders may take 45% of the yield of local zebus (producing 2.2 litres/cow/day) in the Ethiopian highlands (Mukasa-Mugerwa et al, 1989).
The low growth rates of nursing calves on the Borana Plateau were hypothesised to be related to milk deprivation due to high Offtake (Nicholson, 1983a; Nicholson and Cossins, 1984; Cossins and Upton, 1988b). The same relation has been proposed by Pratt and Gwynne (1977: p 36), Wagenaar et al (1986), Wilson (1987), Preston (1989) and de Leeuw et al (1991; see Section 7.3.3.4: The calf: Prospects for growth acceleration). It was also found in the Borana Plateau that this competitive relation will vary markedly according to proximity to market, wealth strata and cow productivity class (Holder et al, 1991; see Section 4.3.5.3: Effects of distance to market, wealth and season on pastoral dairy marketing). It is thus a hypothesis that requires caution in generalising. Wagenaar et al (1986: p 50) noted that higher absolute offtakes of milk from Fulani cows were positively correlated with better calf performance, rather than the reverse. This has relevance to the work done by Mulugeta Assefa (1990) and Holden et al (1991) in which it was commonly stated by producers that they tried to avoid milking low-yielding cows so that their calves could have all the milk, milking only the high-yielding cows and sharing the milk with their calves. Even in this situation, calves of the high-yielding cows are likely to have more milk than calves of low-yielding ones. Thus, the relation between milk Offtake and calf performance is underpinned by cow productivity, which probably explains the result of Wagenaar et al (1986).
It has been further speculated that milk restriction for Borana calves could have a carry-over effect in causing permanent stunting, delaying time to puberty and therefore reducing lifetime productivity of females (Nicholson, 1983a; Nicholson and Cossins, 1984; Cossins and Upton, 1988b). This has also been postulated by Preston (1989). However, these claims have been tested and rejected to be invalid in the southern rangelands. Animal weight tends to become harmonised during the post-weaning period as a result of compensatory growth and seasonal effects (see Section 7.3.3.4: The calf: Prospects for growth acceleration).
In sum, considering the vagaries of the southern rangelands environment, there is no indication that the fundamental productivity of Borana cows is unusually poor. In fact, productivity appears to be similar to the lower end of the range for animals kept under African and research station conditions and even superior to that of indigenous breeds in the Ethiopian highlands. Major differences in productivity likely arise from genetic improvements, but institutionalising breeding to this end is very risky under existing conditions of drought vulnerability where survival capability, and not merely high productivity, is the ultimate measure of success. Although calves grow slowly, this may mean little in a system where high turnover of the inventory is not a production value, production costs are low, and increasing growth rates via more forage feeding and provision of extra water is expensive and very difficult (Section 7.3.3.4: The calf: Prospects for growth acceleration).
Excluding drought impacts, mortality rates for adult cattle were reported to be low after two years of age, with an annual rate of about 5% due mostly to disease and accidents (Cossins and Upton, 1988b). This low rate is in agreement with values for cows reported by de Leeuw and Wilson (1987) who concluded that adult cattle mortality is a minor determinant of overall herd productivity in pastoral systems. This relatively stable state of affairs on the Borana Plateau has been so probably more during the past 20 years or so, because of veterinary campaigns. Nicholson and Cossins (1984: p 5) cited Church et al (1957) as reporting large losses of stock on the plateau as a result of unchecked rinderpest outbreaks in the 1950s. Pratt and Gwynne (1977: p 38) contended that rinderpest control was still ineffective in southern Ethiopia in the early to mid-1970s; but this has changed since then (SORDU, 1988). Long-term destabilising effects of disease control on livestock populations have been mentioned by Pratt and Gwynne (1977: p 38) and Lamprey (1983; cited in Ellis and Swift, 1988). Wood (1989) noted the role of veterinary activities in the growth of cattle herds in western Zambia during the 1980s.
Nicholson and Cossins (1984: p 5) reported that mortality rates of calves, in contrast to those of mature cattle, varied markedly across seasons and years. Their estimates were 10% for average rainfall years and 23% in dry years. The sequence should also include the 90% losses in a drought (Donaldson, 1986; see Section 6.3.1.1: Livestock dispersal and herd composition). Mulugeta Assefa (1990) calculated calf death rates of 21,19 and 69% for average rainfall, dry and drought years, respectively. He also found that lowest mortality rates were rigisterd for animals held by middle-class households, which again justifies the segregation of data by wealth classes. Data from AGROTEC/CRG/SEDES Associates (1974h) and Church et al (1957) suggested that up to two years of age calf mortality was about 40%, presumably in average rainfall years. However, these claims were discounted by Nicholson and Cossins (1984: p 5) as inaccurate.
The overall average calf mortality rate of 18 to 25% for Borana calves up to one year of age during nondrought years is comparable to the 15 to 21 % rates found for some agropastoral systems in the review by de Leeuw and Wilson (1987), and that is markedly lower than that for transhumant systems with rates of 27 to 36% including peri-and postnatal losses (Wagenaar et al, 1986; de Leeuw and Wilson, 1987). De Leeuw et al (1991) noted that Maasai calves have particularly low rates of mortality (12% for calves up to seven months of age) and they attributed this to intensive hand rearing and the high cultural value placed on calves. Calf survival was mentioned as the top production priority of the Maasai, exceeding even milk supply for the family. The Boran also rear calves intensively but they have probably poorer access to veterinary inputs compared to Kenya's Maasai (Evangelou, 1984: pp 76-84; see Section 7.3.3.5: Calf mortality mitigation).
De Leeuw and Wilson (1987) concluded that high calf mortality is the most important factor causing the low output of traditionally managed herds. Likewise, Dahl and Hjort (1976) cited Williamson and Payne (1965) as recommending that calf mortality rates over 15% should attract serious attention for interventions. In contrast, Cossins and Upton (1988b: p 267) speculated that only very modest overall production gains would result from substantial reductions in calf mortality in the Borana system. This view was derived from modeling that employed a "steady state" approach which allowed no net herd growth and contrasted intervention options that shifted the proportion of various age and sex classes of cattle. An increase in the proportion of calves tended to increase overall herd mortality and decrease output by slightly reducing the proportion of milking cows. This approach was later critiqued by Upton (1989) concluding that the benefits of calf mortality mitigation were underestimated by Cossins and Upton (1988b). This is discussed further in Section 7.3.3.5: Calf mortality mitigation.
Nicholson and Cossins (1984: p 7) list seven major causes of mortality for calves and suspected pasteurellosis as the most important, followed by five other diseases and finally starvation. A more recent review of calf diseases is found in Mulugeta Assefa (1990: pp 31-34). Mulugeta Assefa (1990) more clearly segregated reports of calf deaths according to disease and starvation and noted that losses because of diseases were more common in average rainfall and dry years but that during drought nearly all deaths were due to starvation. Nutrition-related losses were most likely the result of competition for milk in poor households while losses from disease of calves belonging to wealthy households were most likely because of lower labour investment (Mulugeta Assefa, 1990).
The limited capability of the regional veterinary unit at SORDU to provide "farm-gate" services to immobile calves means that this large cohort, comprising 24.6% of the regional cattle herd according to Cossins and Upton (1987), remains particularly vulnerable to disease. Mulugeta Assefa (1990: p 33) cited SORDU's (1988) summary of animal health statistics from 1976-87 as estimating that although 25% of calf deaths on the plateau were caused by disease, only 2.6 and 6.1% of the calf population had been vaccinated or otherwise treated, respectively. The logistical difficulties in implementing disease control programmes for calves dictate that at least appropriate feeding management tactics must be devised to reduce calf mortality (see Section 7.3.3.5: Calf mortality mitigation).
The pre-eminent importance of milk production to the welfare of the Boran indicates that health problems specific to cows should receive more attention. Besides spreading several dangerous diseases on the Borana Plateau (but not East Coast Fever), ticks may cause swellings and lesions of soft tissues of cattle (Hill, 1982). The results of the survey on cow udders and teats damaged by ticks (Coppock, 1990b) have important implications for milk yields. A 13% damage rate for teats implies that a household with an average of eight cows may require one more cow simply to maintain current level of milk production. Assuming these data are representative of the entire study area, and if cows make up 45% of the total cattle population of 250000 head (Cossins and Upton, 1987: p 206; Assefa Eshete et al, 1987), it suggests that the number of cows would need to be increased by 14600. At current high stocking levels these additional cows would have to compete with or displace other cattle that are also valuable for economic security (Coppock, 1992b). Because a cow only loses one or two teats from tick damage (D. L. Coppock, lLCA, personal observation), it is impaired but remains productive; so that ticks are a burden on the system in terms of milk production efficiency. It is unclear, however, whether or not the undamaged portions of the udder increase production of milk to compensate for damaged portions. This should be a topic for future research.
One solution for tick control is implementation of appropriate chemical prophylaxis and range management practices such as prescribed burning (see Section 7.3.3.1: Mature cattle and Section 7.3.1.4: Site reclamation). Concern has been raised, however, if an unsustainable introduction of acaricides would not reduce the natural resistance of Boran cattle to tick-borne diseases (G. Smith, former TLDP consultant, personal communication). The literature is conflicting on the degree to which Boran cattle are immune to tick-borne diseases. On the basis of anecdotal observations, Alberro (1986: p 32) noted that Boran cattle were infested with ticks and yet they seemed to have a high degree of tolerance to disease due to early exposure as calves. He also remarked that the Boran herd owners do not feel it necessary to remove ticks because they believe their cattle possess such immunity. This is an understandable response to overwhelming tick problems since they consistently express concern about tick control in ILCA surveys (D. L. Coppock, ILCA, personal observation). In contrast, Pratt and Gwynne (1977: p 145) mentioned that Boran cattle are more susceptible to tick-borne diseases than the East African Zebu, but this remark may be biased because of the occurrence in Kenya of East Coast Fever which is deadly (Stobbs, 1966). Hill (1982) cited AGROTEC/CRG/SEDES Associates (1974h) as listing 21 different types of ticks on the Borana Plateau and that the local stock are resistant to tick-borne piroplasmosis, anaplasmosis, theileriosis and rickettsiosis except when nutritionally stressed. Despite these conflicting comments, still use of acaricides could be useful for adult animals that have been exposed to tick borne diseases as juveniles while the topic requires further investigation (see Section 7.3.3.1: Mature cattle).
One of the most important findings of the work of Nicholson (1987a) was that cattle possess the ability to compensate in the wet seasons for moderate amounts of live weight lost in dry periods as a result of water restriction. Water restriction acted primarily to overheat animals and thus depress their dry-matter intake, a fact which had been noted previously (King, 1983). The fundamental problem of water shortage in dry periods tends to discount the effectiveness of forage interventions specifically targeted to improve the nutrition of adult cattle in dry seasons. In contrast, because of their smaller size, improvements of forage and water resources for calves in dry seasons appear more feasible (Section 7.3.3.5: Calf mortality mitigation).
This ability to compensate for the ill effects of water restriction is significant for beef production and the recovery of cows to allow conception. However, the results obtained by Nicholson were probably positively influenced by conditions at Abernossa compared to the pastoral situation because of the high availability of forage at the and because the animals were apparently not corralled at night, which would allow them to forage when temperatures were cooler. Dry season standing crops may vary 10-fold between Abernossa and pastoral systems (D. L. Coppock, ILCA, personal observation), animals on the Borana Plateau are confined, for example nightly for about 11 hours (Belete Dessalegn, 1983). This suggests that compared to the pastoral situation, the animals may have lost weight more slowly in dry periods and while being able to recover more quickly in wet seasons. The biggest difference in animal response between the two systems may occur in dry periods. Assuming that forage quality is similar, cattle could probably have more opportunity to consume a higher quantity of forage during the first one or two days after watering compared to those grazing communal rangeland. Conversely, differences in animal response in the two systems may be less in rainy seasons, and this would depend more on differences in the quality and productivity of forage. Despite these points, however, it is clear that watering cattle once every three days is common during dry seasons on the Borana Plateau. Further study would be needed to assess the degree to which weight compensation by cattle is impaired under pastoral, as opposed to ranch, conditions.
Cattle watering once every four days was observed at some well groups on the Borana Plateau during the long dry seasons of 1988 and 1989, and this was reportedly due to a low volume of ground water in particular areas (D. L. Coppock, ILCA, personal observation). This variation with year may be related to the possibility that certain watersheds receive lighter precipitation during the long rains. In any case it supports comments by Alberro (1986: pp 34-35) that Borana cattle can be watered once every four days and are thus particularly tolerant to water deprivation, although as previously mentioned the relatively cool climate probably contribute to this through slower rates of dehydration (King,1983). The Boran recognise that cattle need to be adapted to thr regime of watering every three days. They thus refrain from daily watering of calves in dry seasons even when sufficient water is available (D. L. Coppock, ILCA, personal observation). The role of water restriction in calf growth and the effects of small-scale water development for use by calves and households is reviewed in Sections 7.3.3.4: The calf: Prospects for growth acceleration and 7.3.3.5: Calf mortality mitigation.
Nicholson (1987a) concluded that overall effects of water restriction were negligible and recommended restricted watering practices particularly for ers. Significant benefits such as water use efficiency (30%), extension of grazing area (ninefold), forage conservation and savings of labour and/or fuel were noted from restricted watering. In one striking example, Nicholson (1987a) noted a potential annual savings of 1.9 million tonnes of water as a result of increased water-use efficiency by 800000 cattle throughout the southern rangelands if they were all to go on a once every three days watering regime. Such advantages of water restriction probably outweigh most disadvantages to the Boran and are instrumental in helping a high density of both people and animals survive despite their being restricted to areas around the deep wells for a habitat (see Section 2.4.1.7: Water resources). In addition, implicit recognition that cattle can compensate for certain seasonal stresses such as water restriction is probably one reason why the Boran are apparently uninterested in supplementing the diet of cattle during dry seasons as long as the situation is not life threatening (see Section 4.4.4: Traditional marketing rationale).
Cossins and Upton (1988b: p 259) echoed the benefits of watering every three days and stated there were no apparent costs associated with the practice. While this may be true in terms of live weight loss and physical condition, the conclusion needs to be qualified with respect to milk yield, which is the key production parameter and of acute importance to the Boran in dry periods. Nicholson (1987a) noted that the reduced calf growth among the group watered every three days was indicative of a drop in milk production. The water turnover and weighing trials indicated that this reduction was about 13%. This is a considerable depravation for the calves, but the Boran obviously have the perspective that the interacting constraints of labour, water and forage outweigh this.
In a related eight-month experiment with four treatments not reported here, Nicholson (1987b) examined the interactive effects of corralling nightly and trekking to water once every three days on live weight and physical condition dynamics and activity budgets of growing and breeding cattle at Abemossa. The main effect of walking and corralling combined was to reduce grazing time by 40% when compared to control animals. Walking and corralling also depressed dry-matter intake but only 12% (compared to the controls), indicating that the restricted animals fed at faster rates when given the opportunity (Nicholson, 1987b). The overall effects of corralling and walking a total of 3040 km on animal response variables were interpreted to be small and overshadowed by the generally poor base level of nutrition, as reported earlier by Payne (1965).
Using calculations of maximum live-weight differences between walked and sedentary groups (20 kg) and the assumption that this live weight should contain 480 MJ GE (MAFF, 1984), Nicholson (1987b) calculated that the cost of walking was 0.6 KJ/kg/km, roughly one-third of the value reported by King (1983). Nicholson (1987b) speculated that this apparent energy saving may have been caused by walking adaptation, physical fitness and/or other energy-conserving mechanisms, but he also acknowledged that the true energy value of the lost live weight was unknown and involved complicated interpretation. The role of energy-sparing mechanisms in Bos indicus under stress has been dealt with by others (Ledger and Sayers, 1977; Western and Finch, 1986) but detailed and properly designed metabolic trials appear to be lacking. Nicholson (1987b) also commented that there was no evidence that walking caused dramatic declines in milk yield. This had been postulated by Sandford (1983a) as a likely energy trade-off with implications for improving the distribution of water points in a way that minimised travel. It is important to distinguish, however, whether the possible cause of such a trade-off is indeed energy allocation or merely foraging time and if productivity lost is because of walking. The latter is a simpler explanation as it is a more pervasive seasonal constraint for cattle in pastoral systems (Coppock et al,1988). Once again the direct transferability of results to a pastoral system may be questioned here because of the superior foraging conditions on ranches. Further details on this point are available in Nicholson (1987b).
The question of poor adaptability of cattle from higher elevations when taken to the lowlands was raised in the analysis of Sarite cattle by Nicholson (1983b) and this may be relevant to whether or not highland genotypes could have a sustained influence on diluting the Boran breed in the rangelands. Borana herd owners have been noted to obtain highland cattle in their attempts to re-stock after droughts (Negussie Tilahun, 1984 cited by Mulugeta Assefa, 1990: p 19), and highland zebus are regarded as inferior compared to the Boran (Alberro and Solomon Haile Mariam, 1982a, b). The ability of highland stock to eventually dilute the Boran is a component of the debates regarding genetic conservation of superior indigenous breeds in relation to livestock development in Ethiopia (Alberro, 1986; G. Smith, former TLDP consultant, personal communication).
Based on preliminary growth data, Nicholson (1983b) hypothesised that cattle procured from higher elevations on the plateau were poorer performers in the warmer and drier lowlands. He also speculated that these animals were probably genetically related more to highland stock than to the Boran. If the growth data are considered indicative, it may be further speculated that longterm climate patterns (including droughts that induce high levels of stress and cattle mortality; see Section 6.3.1.1: Livestock dispersal and herd composition) may serve to keep highland genotypes out of much of the lower elevations of the southern rangelands on a sustainable basis. Research to test this hypothesis should precede any large investment in an attempt to conserve the Borana genotype from alleged dilution by highland breeds.
To sum up, when the vagaries of the southern rangelands are considered, performance of Borana cattle appears quite satisfactory. Results for production parameters, however, are only indicative and must not be regarded as static. As often noted, the environment is the supreme regulator of production here and this interacts with a dynamic population density that competes with cattle for resources. Wagenaar et al (1986: p 50) noted that only a detailed time series of data, systematically collected over many consecutive years, could provide the most useful insights into production dynamics and system function. All that was presented in this chapter has been a group of discrete studies, often difficult to compare with each other and impossible to compare with work conducted 20 to 35 years earlier.
Although hard evidence is lacking, it appears that the current population densities of both humans and cattle on the Borana Plateau may be the highest on record (see Section 7.2: A theory of local system dynamics). This may set the stage for more regular density-dependent influences on cattle productivity. This depends on climatic patterns and whether such key inputs as veterinary campaigns and manpower can be maintained. If it is assumed that veterinary and labour inputs improve or remain constant, a long period of average rainfall will allow the cattle population to reach and maintain a density where it impacts the environment and consumes forage reserves otherwise saved for droughts.
Under these conditions it can be expected that milk production, weaning weights, cow condition and calving rates will fall and calving intervals will increase. Heavy wet-season grazing could even impair the ability of cattle to regain weight lost during the dry season as a result of restricted watering. Some herd owners reported that lower milk yields and a reduced condition of cattle are already occurring as a result of heavy grazing in 1990, less than six years after the 1983-84 drought (D. L. Coppock, ILCA, unpublished data). On the other hand, if there is regular drought, this will kill animals, de-couple interactions to more of a density-independent situation, help preserve the environment and create years of high productivity per cow in drought-recovery years as a result of low competition for forage. These scenarios of variable environmental check on production parameters are supported by the observation that Borana cows managed under conditions at Abernossa have reproduction cycles occurring throughout the year (S. Sovani, ILCA, personal communication). This presumably is related to the low stocking rates and high standing crops of forage on the ranch as well as the stock being upgraded genotypes.
General observations of small-stock management indicate that patterns of milking, herding, corralling and intensive rearing of kids and lambs practiced by the Gabra and Borana are similar to many reported for elsewhere in sub-Saharan Africa (Wilson, 1988). Two notable differences may be the lower likelihood of Boran or Gabra milking sheep compared to goats (Cossins and Upton, 1988b: p 265) and the apparent lack of any breeding control. De Leeuw et al (1991) noted that the Maasai use breeding aprons in an attempt to control the seasonality of conceptions in small ruminants. In his review, Wilson (1989) commented that uncontrolled breeding systems typical of indigenous management are usually more productive than those on research stations where restricted breeding occurs.
Small ruminants comprise about 5% of the livestock biomass held by the average, eight-cow Borana family (Cossins and Upton, 1987: p 213). This is considerably less than the 15 to 25% range reported for small ruminants in other pastoral settings (Wilson, 1988). The mean age at first parturition for sheep and goats at Beke Pond (17.4 months) is near the average (17 months) for both species reviewed by Wilson (1989), but the mean birth interval of 7.7 months appears lower than that given by Wilson (10 months). Litter sizes on the Borana Plateau of (1 for sheep and 1.1 for goats) also appear to be slightly lower than those in Wilson's (1989) review of 1.38 (goats) and 1.16 (sheep).
The greatest constraint to small ruminant production in the Beke Pond area is apparently disease, which is in agreement with the assessment of Wilson (1988: p 324) for sub-Saharan Africa in general. Contagious caprine pleuropneumonia (CCPP) is reportedly the most pervasive disease of small ruminants on the Borana Plateau (Sileshi Zewdie, SORDU veterinarian, personal communication). In the past there has been no vaccine for CCPP manufactured in Ethiopia but it is anticipated that production will commence in the early 1990s (Sileshi Zewdie, SORDU veterinarian, personal communication). The lack of veterinary service for small ruminants in the SORDU project area today means they will continue to be high-risk species to produce, although the management costs are much less compared to cattle in terms of herding, milking and watering. To compensate for the lack of veterinary drugs some pastoralists reportedly use human antibiotics from clinics to treat illnesses of small ruminants. That special health programmes for small ruminants are hard to sustain is apparent from remarks by Hill (1982: p 5) who noted that mobile clinics for small ruminants and camels were routinely sponsored by the Ethiopian Government prior to 1980.
Perhaps the most important information collected to date on small ruminants on the Borana Plateau is from aerial surveys which indicated that numbers of sheep and goats were less affected by the 1983-84 drought compared to cattle in general (see Sections 6.3.1.1: Livestock dispersal and herd composition and 6.3.1.3: Small ruminant productivity). Small ruminants probably contribute to household income diversification and stability. They may also offer a higher rate of economic return compared to cattle (Cossins and Upton, 1988b). Browsing goats in particular use vegetation in a fashion that is complementary to grazing cattle (Belete Dessalegn, 1985; Coppock et al, 1986a;
Woodward, 1988; see Section 3.3.5.1: Livestock food habits). Wilson (1988) noted a keen awareness regarding the value of small ruminants in rural African systems, and this is applicable here.
Wilson (1984: pp 17-18) noted that the upper limit of the camel's habitat range is around 550 mm of annual rainfall, and that Islam is an important factor in its distribution. Observations on the Borana Plateau confirmed these points as camels are most abundant in the 650-mm rainfall zone and appear to become very rare some 75 km north of Yabelo where annual rainfall approaches 750 mm. Disease in these more mesic locations may be an important limiting factor (see below). In addition, the camel's presence further north is also restricted by there being other ethnic groups. The pockets of Muslim Gabra to the north, south and south-west of the study area have been instrumental in introducing camels to the Boran.
Many of the Gabra moved to various pockets on the Borana Plateau from the Kenya border during the conflict with Somalia in the late 1970s. The concentration of Gabra in the upper semi-arid zone near Yabelo is most likely due to their free access to the large communal ponds constructed in the 1960s (see Section 1.4.5.2: Water development strategy). The Gabra are not allowed to have easy access to the traditional deep wells to the south that are the property of the Boran and therefore under their control. Ethnic conflicts in 1991 led to reported evacuation of Gabra from Borana-held territory (D.L. Coppock, Utah State University, unpublished correspondence). Wilson (1984: p 44) stated that Gabra camels of northern Kenya are typically small and compact in form. While this somewhat fits the description of the geleb, it does not describe the massive quorti. The quorti more resembles the Somali milk camel (Wilson, 1984: p 42). It is thus apparent that some diversification of camel holdings may have occurred among the Gabra.
Limited observations regarding camels indicate that their most common use by Borana and Gabra pastoralists is to transport grain, water, salt and other goods. They also have a critical role in milk production for the minority Gabra. The Borana, however, have insufficient numbers of camels to provide a meaningful supply of milk to their diets. A few Borana innovators have been observed using male camels to pull ploughs for cultivation.
Although limited in scope, data of Belete Dessalegn (1985) verifies the high milk yields and long lactation period of camels compared to cattle (1045 litres over 430 days), which is at the low end of the scale for camel milk production according to the review by Mukasa-Mugerwa (1981: p 61). Camels in East Africa may produce up to 2100 litres over an 18-month lactation (Dahl and Hjort (1976) cited by Mukasa-Mugerwa (1981: p 61)). Data from Coppock (1988) also illustrates that camels reportedly produced over 10 times more milk than cattle during the height of the 1983-84 drought (1 vs 0.095 litre/head/day; see Section 6.3.2.1: Livestock). Similar perspectives on the role of camels in improving stability of pastoral production systems have been reported by Coughenour et al (1985), Coppock et al (1986c) and McCabe (1987). The higher milk yields of camels are not only significant in terms of household milk consumption but also provided the Gabra with the chance to sell milk when prices were high so that they could buy grain (Coppock, 1988; see below).
Disease is reportedly the most important constraint of camel production in the Beke Pond region. Sileshi Zewdie (former SORDU veterinarian, personal communication) considered trypanosomiasis as the biggest threat to camels and he felt that infections are incurred by browsing along riverine location and during long-distance commercial treks out of the study area. Tick-borne diseases and skin infections are also prominent (Coppock, 1988). Today SORDU is unable to provide veterinary care for camels.
The Boran are very interested in acquiring camels but they often mention problems of cost and market access as major constraints in their attempts to procure them (see Section 7.3.3.2: Camels, donkeys and small ruminants). Other constraints may include low rates of reproduction (Wilson, 1984: p 136), the additional labour inputs required for their management and some peculiar aspects of biology such as inducible ovulation that requires management training (Wilson, 1984: p 84). Despite these obstacles, camels can play a more useful role on the Borana Plateau (Section 7.3.3.2: Camels, donkeys and small ruminants).
The very small population of donkeys and mules indicated from surveys is unexplained. This should receive more research attention in the future. It is possible that water shortage may be a factor in limiting the donkey population. Research in Kenya's arid zone, however, has demonstrated that donkeys perform adequately on a restricted-watering regime as well as on very poor quality forage (Coppock et al, 1986a, b; 1988). Another constraint on donkeys could be forage competition from numerically superior cattle (see Section 3.4.3: Use of native plants). If equines are restricted to graze in the vicinity of encampments, this may make them even more vulnerable to forage competition and hence to death during drought. The high loss of 60% of the equines during the 1983-84 drought attests to this situation (see Section 6.3.1.1: Livestock dispersal and herd composition). Donkeys could play a greater role in alleviating labour burdens on women (see Section 4.3.3: The labour of married women).
This chapter has reviewed some baseline features of livestock management and production in an attempt to provide a general perspective on the system. As will be shown in Chapter 7, however, the proper orientation is to view livestock production and management as potentially highly dynamic from year to year. Livestock production and management vary in terms of short-term cycles and long-term trends as defined by interactions of the cattle population with the environment. Integrated research and development perspectives are presented in Chapter 8: Synthesis and conclusions.
