M. J. Nicholson
International Livestock Centre for Africa, Addis Ababa, Ethiopia
Outlook on Agriculture Volume 14, No. 4, 1985
© 1985. Pergamon Press printed in Great Britain
Water availability and development
This article discusses the water needs of livestock in Africa and other hot regions; the physiological utilization of water both within and between species; and suggests some ways of optimizing water use and using water as a management tool.
Water is the most vital ingredient for life on earth. It is generally believed that life began in an aquatic environment and that the migration of species on to land depended on the ability to preserve an aqueous internal environment in which all biochemical reactions take place. As species diversified into drier and hotter regions, adaptations to withstand water loss became increasingly necessary. Livestock in hot environments must take advantage of two properties of water in order to assist in thermoregulation, namely a high thermal capacity and a high latent heat of evaporation.
Water is also essential for plants upon which the livelihood of livestock depends. In most African countries, drought is a recurring threat to plant and animal life and when combined with the increasing problems of rising human populations, overstocking, and improper land use, the result can lead to erosion, desertification, and eventually famine.
During the past three decades, much of the aid given to Africa has been directed towards the semiarid and arid zones where the provision of water for both humans and animals has formed the largest single item of expenditure .
Water is the main constituent of animals, normally comprising between 50 and 80 per cent of body weight. A reduction in this body water below a certain critical level is more life threatening than a shortage of any other component in the body, with the exception of oxygen. In man, a 7 per cent fall in body weight caused by water loss is described as severe dehydration, while water depletion leading to a 15 and 20 per cent loss of body weight causes coma and death respectively . Man, however, is not well adapted to dehydrating environments, in contrast to many animals which have efficient ways of conserving their body water. Many domestic animals in the drier regions of the world have adapted to water restriction by tolerating high levels of dehydration accompanied by a varying degree of water conservation efficiency. In comparison with desert adapted wild animals, even the camel is inefficient at conserving body water but the strategies adopted by wild animals lies outside the scope of this article.
To maintain the body-water pool within homeostatically acceptable limits, any water lost must be replaced. The proportion or quantity of water utilized in a given unit of time is known as the turnover and varies according to the species, size, and physiological status of the animal, and the environmental conditions. In general, animals adapted to dry environments have lower turnover rates than species in temperate zones, but in the case of many livestock a greater heat load gives rise to substantial water loss by evaporative heat dissipation so that turnover rates may in reality be much higher. Daily turnover can be expressed as millilitres per litre of body water (ml 1–1 d–1) or as mls per kg of body weight (ml kg–1 d–1). The disadvantage of the former is that total body water must be known and assumed to be constant. In practice, no accurate method of establishing body-water content in vivo has been found for large animals and validating estimates by the dilution method using total body dissection has revealed variable and substantial over-estimates [3, 4]. In contrast, body weight is an easily determined variable.
The range of turnover rates in four livestock species is shown in Table 1. Camels have the lowest turnover, zebu cattle and sheep have comparable rates, while goats have the highest turnover. This contrasts with R. E. McDowell  who said that goats native to tropical areas seem to have a lower turnover rate, 11 per cent lower than sheep at high temperatures. Large animals with a lower metabolic rate per unit weight and a low surface-to-volume ratio would be expected to have correspondingly low water turnover rates, and in this respect cattle appear to be the least efficient users of water.
Table 1. Turnover rates of four livestock species*.
Turnover rate (ml kg–1d–1)
King  reported a linear relationship between the log of body-water pool and the log of water turnover, the regression being:
Water turnover (1 d¹) = antilog [0.836 log body water (1) 0.619]
Water turnover is affected by many parameters and the relationships between them are complex. Since water requirements are related to water turnover and water loss, it is useful to discuss how water is lost and replaced before examining water requirements in detail and relating these to the factors that alter turnover rates.
Body water is usually lost by four routes: by evaporation, in the faeces, in the urine, and in the milk. In addition, major haemorrhaging, excessive salivation, and mucous or serous exudation can give rise to additional and serious water losses as a result of traumatic or pathological conditions, but these are not normal and will not be considered further.
Evaporative water loss rises in response to an increased heat load when the other heat dissipation mechanisms, such as re-radiation, convection, conduction, and vasodilation, are insufficient for thermoregulation. In hot areas, water loss and thermoregulation are inextricably linked in domestic animals. The efficiency of other features designed to lower heat absorption or increase heat dissipation will affect the extent of evaporative water loss and should be mentioned briefly. They include animal coats, coat colour, and appendages. Animal coats act differently in protection from heat loads: the short, smooth coats of Bos indicus (zebu cattle) and goats reflect light and facilitate convective loss; in contrast, the tight-packed wool of the Merino sheep re-radiates in the long wavelengths while the loose wool of the Awassi sheep absorbs more heat while convective heat loss is more effective . W. V. Macfarlane  calculated that sheared sheep doubled their water turnover and absorbed two or three times more heat than unshorn sheep. Coat colour may be equally important: zebu cattle with white coats reflect 40 per cent more radiation than black cattle  but black cattle absorb heat quicker in the early morning, reducing energy loss from thermogenesis. For this reason, V. A. Finch and D. Western  found that more white cattle died than black ones during a drought, but in general a positive correlation was found between potential evaporation and the incidence of white cattle .
Heat loss (or lowering of heat absorption) has also been associated with anatomical features such as the narrow dorsal profile of the camel, dewlaps, large ears, and skinfolds of Bos indicus or the long limbs of goats, sheep, camels, and pigs in hot regions which reduce absorption of reflected heat from the ground and presumably aid convection [12, 8, 13, 1]. Whilst the efficiency of some of these adaptations has been questioned  there is no doubt that the overall effect would be to reduce evaporative water loss, especially since re-radiation and convection will be enhanced by larger surface areas.
As a means of dissipating heat, evaporation is efficient: vaporization of 1 g of water releases 2.43 kJ but it is at the risk of accelerating dehydration, as up to 4 or 5 per cent of body weight per day can be lost by evaporation in hot conditions [14, 15].
Evaporation will therefore become an important source of water loss in hot areas especially when an animal is exposed to solar radiation and generating metabolic heat. Finch  found that a steer in the open gained 71 per cent of its total heat load from shortwave radiation and the balance from endogenous production. Of this absorbed heat, 21 per cent was dissipated by sweating and 5 per cent from respiratory evaporation, the balance being lost using methods not involving water loss. King  cited C. R. Taylor  (1972) and Finch  by saying that 80 per cent of the water loss in tropical ruminants may be from evaporative cooling. W. Bianca et al  compared steers in a climate room and found that evaporative water losses amounted to 30 per cent of water consumed at 15°C and 61 per cent at 40°C.
Evaporative losses can be cutaneous or respiratory and the contribution of each route varies according to species . Whereas in donkeys and camels sweating accounts for almost all evaporative heat loss, panting is the major method of evaporation in dogs and pigs. King  reviewed the data available on both methods and concluded that panting is more efficient for smaller animals. Working with Hereford steers and eland, C. R. Taylor and C. P. Lyman  demonstrated that between 22° and 40°C, 7080 per cent of evaporative water loss was cutaneous despite rapid panting. In contrast, 60 per cent of the evaporative water loss comes from panting in sheep and goats . 'Open-mouth breathing in livestock is an extreme form of panting often associated with excessive salivation (exacerbating water loss) and such animals are usually intolerant of heat .
Except in very hot environments, faecal water may often be the chief route of water loss. Species vary their ability to resorb colonic water and while camels can void faeces at 45 per cent dry matter (DM) and sheep at 50 per cent , cattle cannot dry their faeces under normal conditions to greater than about 30 per cent DM. Only Australian workers  quoted a figure of 40 per cent DM for B. taurus cattle under conditions of severe dehydration in a hot, unshaded environment. J. Quarterman et al  found that zebu cattle had significantly drier faeces than B. taurus while G. D. Phillips  found no differences. Using Herefords, Taylor and Lyman  showed that restricted watering only reduced faecal water from 78 to 75 per cent. Nevertheless, nearly all the data showed faecal DM of less than 30 per cent in cattle. In general, the hotter the environment, the more faecal water loss as a percentage of water consumption declines . Phillips  found that total faecal water loss amounted to 5057 per cent of water consumption in both zebu and Hereford steers. Bianca et al  found that at 15°C, Ayrshire steers lost 58 per cent of their water consumption as faecal water, whereas at 40°C, this was reduced to 20.5 per cent as water consumption and evaporative water loss climbed. W. Little et al  found that total faecal water loss of Friesian cross cows restricted to 60 per cent of their normal water intake, was 58 per cent of normal faecal water loss.
Since the moisture content of cattle faeces cannot be reduced substantially, a reduction in total faecal output would reduce water loss. This is associated with a reduction in dry matter intake that occurs in response to dehydration.
Urine is also a significant route for water loss and two methods are available to reduce it. The first lies in the ability to increase urine osmolarity up to a limit predetermined by kidney anatomy. Thus, while camels, goats, and sheep can excrete urine up to 27003000 mOsm kg–1, cattle cannot concentrate urine beyond 1400 mOsm kg–1 . The second method can be even more critical in water conservation, and that is to inhibit diuresis following large intakes of water. This is a common phenomenon in camels  and goats  but less well-known in cattle. B. D. Siebert and W. V. Macfarlane  recorded a rise in diuresis 24 hours after watering dehydrated B. taurus, while it took 26 hours for a similar effect in the camel which followed what appeared to be complete renal shutdown. However, Bianca et al  did not observe immediate diuresis in cattle following rehydration.
Milk production adds a considerable burden to the water economy of a mammal  and while high-yielding cows are rare in Africa, any lactating animal will have a greater water demand than a male or dry animal in a similar environment. For every kg of milk secreted, 85 per cent will be water, and in addition to this a considerable quantity of metabolic heat will be generated in producing it, necessitating yet more water for its dissipation. M. A. Barrett and P. L. Larkin  estimated that an extra 3 litres of water were required for each litre of milk produced.
The suggestion by King  that the milk of arid-adapted ruminants is not very different from that of other livestock because the young suckling animal needs water as much as nourishment is probably incorrect. Urine from calves at foot, even when their dams are water-restricted, is consistently hyposmotic to their plasma suggesting imperative water excretion (Nicholson, in preparation). The reason for the high water content of milk is undoubtedly osmotic, since hypertonic milk cannot be secreted by any ruminant. However, young animals, particularly calves and camels, are frequently restricted in their milk intake under pastoralist and subsistence management conditions and this may well lead to a water deficit where water is not freely on offer. The situation will be further exacerbated by diarrhoea.
The main method for replacing water lost is by drinking although at some times of the year, water on and in the food may be sufficient to meet most of the animal's demands. In many situations in Africa, water is not available ad libitum to livestock; where it is, the risk of dehydration is minimal and the deleterious effects of heat and exposure to the sun will be lessened. Assuming that access to a watering point is at best daily for a few minutes, the quantity of water that can be drunk is a critical parameter in determining whether all of the water lost can be made up and whether extra can be accommodated to serve as a store.
Of all the ruminant species, Bedouin goats appear to be the most adapted to large water intakes, imbibing an average of 30 per cent of their dehydrated weight  causing these and other workers to suggest that the rumen and the gastro-intestinal tract are serving as a water reservoir . Cattle, sheep, donkeys, and camels can all drink voluminously , although Macfarlane and Howard  found that dehydrated camels only drank 60 per cent of total water needed for rehydration at the first drink. In practice, however, it is often essential to recover completely the water lost at the first drink or livestock would become progressively more dehydrated .
Classen  found that Zebu cattle could drink a maximum of 45 kg of water on 3-day watering but at the risk of water intoxication and death, while more recent work has shown that up to 104 kg can be drunk within four minutes with no apparent signs of distress (Figure 1(a)(b); Nicholson, in preparation). King  cites S. Sandford  who noted that after a large drink, livestock often stagger about and lie down, and King suggests that the cause may be related to water intoxication. Similarly, weak cattle in pastoral situations will occasionally die an hour or two after drinking following three days without water .
Figures 1a, 1b. Cattle have been recorded as drinking 104 kg in three minutes. (a: left) Before (b: right) After
Water intoxication is an uncommon condition seen in calves ; the reasons for weakness and death following drinking are probably caused by malnourishment and the additional strain of the extra weight. Avoidance of serious sequelae following ingestion of large volumes of drinking water is important and it would appear that most livestock do not suffer at all. The two main threats would appear to be haemolysis caused by a sudden decrease in plasma osmolarity and, more rarely, nervous signs as the turgor pressure of cerebral cells is raised. Choshniak and Shkolnik  wondered whether erythrocytic stability is ever challenged in arid-adapted or water-restricted livestock, since ruminal water is retained in the rumen and plasma osmolarity does not usually fluctuate more than 3050 mOsm kg¹.
Moisture in the food is the second most important source of water for livestock, especially during the rainy season when 85 per cent or more of standing grass may be water. At this season, most African livestock can be independent of water (Figure 2). During the post-rains period, when maximum live weight gains are likely, the animals resort to ephemeral wet-season standing water (Figure 3) for their drinking. King  notes that an exception to this is the grazing of volcanic hills or areas where porous soils prevent standing water accumulating, in which case such areas can be exploited only during the rains or when they are within walking distance of water. As the grass dries out, many species may resort to browse in which the moisture content is seldom below 30 per cent . This in part explains why camels, smallstock, and wildlife often fare better in the dry season than cattle, although cattle may browse extensively when water is restricted.
Figure 2. Cattle and other livestock can be independent of water during the rains and can exploit outlying grazing.
Figure 3. Cattle using ephemeral wet season ponds (Rangeland, Sidamo, S. Ethiopia).
Water on plants may also be an important source of water intake both after rain or formed as dew during the night. G. D. Brown and J. J. Lynch  calculated that sheep obtained between 30 and 130 ml kg–1 d–1 from guttation and dew on improved grass. Even though many African livestock are not permitted to graze at night, taking them out at dawn would allow them to take advantage of this extra water source. It has also been observed that some plants elevate their moisture content hygroscopically, which would be of use to night-grazing animals .
Other less important sources of water replacement have been reviewed by King [1J and include metabolic water, in which water is formed by the oxidation of organic compounds. Although it is usually low in proportion to water intake of normally watered livestock, it was found to contribute 16 per cent of total water input in eland and 8 per cent in penned livestock. Water in inspired air was described by R. C. Weast et al  as contributing as much as 10 per cent of total water intake. However, this will depend upon the relative humidity, and in the semi-arid tropics outside of the rainy season it is unlikely to be a significant source of water intake, particularly when respiratory water loss is much greater, since expired air is usually warmer and more saturated than inspired air.
For growing animals, milk is an important source of water especially when they begin to eat solid food and rumen function starts. Hypotonic urine of calves can also be a useful source of water for their lactating mothers and the drinking of urine has been noticed in water-restricted cows (Nicholson, in preparation) but this phenomenon has not been observed in other livestock species. The urine will be of benefit only as long as it has a lower osmolarity than the drinker's urine but certain cows have been observed to drink their own calf's and other calves' urine for six or seven months after calving.
For those concerned with livestock development and the concomitant provision of water supplies, it is useful to know the relationship between the numerous factors which alter water-turnover rates and hence influence drinking requirements. It can be assumed that the total water requirement less that already obtained from other sources (in food, as metabolic water, etc.) constitutes that consumed in the form of drinking water.
While some wild ungulates can live almost independent of drinking water, all the common domesticated species must have access to drinking water at intervals. This frequency of watering is an important topic and will be discussed later. The quantity consumed will be influenced by four main factors, viz. animal, diet, environment, and management.
Various animal variables will influence drinking. The first is species: the camel's inherently low turnover rate means that expressed on a daily basis its drinking requirements will be low. However, its size coupled with its normally infrequent visits to water mean that total intake at drinking is likely to be large, up to 100 litres [5, 35]. Cattle and horses drink more than sheep and goats on a ml kg–1 basis because although their turnover rates are comparable, sheep and goats are likely to derive more of their water from food, particularly browse.
Within breeds differences can also be marked; several authors have observed that zebu cattle turnover water more slowly than B. taurus breeds under similar conditions [36, 37, 38]; this is due to a number of factors including drier faeces , anatomical differences reducing evaporative losses, diurnal hyperthermia, and lower heat production. Barrett and Larkin  reported that indigenous male cattle drink half as much as exotics under the same conditions. Other livestock species may exhibit similar interbreed differences, although even the desert-living Bedouin goat has a turnover of 140 ml kg–1 d–1  which is not low for a goat breed. It would seem reasonable to expect hair sheep and fat-ailed sheep to have lower water turnovers than wool breeds originating from temperate environments but comparative data on sheep breeds are scarce.
Since lactation exacerbates water loss, it causes similar increases in drinking requirements. Macfarlane and Howard  and McDowell  calculated that lactating livestock require up to 44 per cent more water than non-lactating animals. However, this is likely to be related to total milk yield.
Young animals require large quantities of water for growth, a requirement normally met by the water in milk. Young Merino lambs had water turnover rates twice that of adult sheep, as a result of higher metabolism, growth, and the increase in total body water [5, 7]. This higher water requirement will hold true for pregnant animals and those recovering from a dry season which gave rise to large weight losses. In both cases water intake must exceed water loss to allow for increasing body water content.
Climate has a marked influence on drinking. Cattle in a climate room study drunk nearly twice as much water at 40°C than at 15°C  and P. C. Fourie et al  found that a change from 17.2°C to 33.9°C increased water consumption in cattle by 67 per cent. Solar radiation is possibly more important than ambient temperature in affecting drinking requirements but published work on this topic is scarce. Rainfall can depress the quantity of water drunk at troughs by 100% as a result of cloud cover, surface water and water on the feed. In contrast, the combination of hot weather and feed intake increases water requirements. Not only does dry food increase water demand  but water requirements rise steeply with dry matter (DM) intake and rising temperature (Table 2). Such requirements must be modified during pregnancy or lactation; pregnancy will raise demand by a factor of 1.5, and lactation according to yield. As a guide, beef cattle and dry cows need about 40 1 per head per day but allowance should be made for wastage and evaporative losses.
Table 2. Water requirements for certain livestock.
Water intake (kg/kg DM d–1)
4.1 (Beef cattle);
4.7 (Beef cattle);
5.5 (Beef cattle)
Husbandry and management factors influence drinking requirements in several ways: the provision of shade has always been regarded as essential for optimum livestock performance although the benefits of natural shade remain largely unquantified [1, 7]. The utilization of shade will depend upon access to grazing and grazing behaviour. Where feed and bite size are limiting factors, grazing time will be extended and this will be compounded by night-kraaling. Where increased exposure to radiation during the hottest time of the day is inevitable, drinking requirements will rise.
No data are available on the effect of walking distances on drinking requirements: any increase in water consumed is likely to be a result of the added heat load from walking in the sun during the middle of the day, since heat production as a result of walking per se is not high and would probably be dissipated as re-radiation, provided walking speed was low.
Water frequency has a large effect on drinking requirements. Camels are well-known for their ability to conserve water and survive up to ten days or more between watering, resulting from their high tolerance of dehydration [14, 21, 40]. Sheep and goats can operate on three to five-day watering in many range-land areas while cattle are least efficient, requiring water at least every three or four days. Comparing performance of sheep on one-, two-, three- and four-day watering, G. C. Taneja  found that 169 sheep on three-day watering could be maintained on the water ration required for 100 daily watered sheep. Weight loss, however, was more pronounced but no details were given as to whether weight loss was fat, water or gastro-intestinal fill. R. G. Mares  reported that Somali sheep and the Galla goat need to drink only once a week. These frequencies represent the maximum for continuous production and even at these levels a decline in productivity can be expected. However, since the rumen capacity is likely to be the limiting factor to water intake, a depression of water intake is likely when watering frequency is restricted. The watering interval must not be so long as to allow the water lost by dehydration to exceed that which can be drunk in the time allowed for drinking. The ability of goats, camels, and sheep to replace 30 per cent of their dehydrated body weight [1, 8, 27] is comparable with the author's findings in Boran cattle under three-day watering where up to 28 per cent of dehydrated body weight has been drunk in three minutes (Figure 1 (a)(b) ). This compares favourably with the other species which required a longer period of time for full rehydration [1, 5, 43].
The quality of drinking water can seriously affect livestock productivity, and falls into three categories, total dissolved salts, toxic and contaminating substances, and disease-producing organisms . Different species vary in their tolerance to dissolved salts (Table 3) and this is partly related to the ability of the kidney to concentrate urine.
Table 3. Tolerance of salty drinking water by different livestock.
Recommended units of total dissolved salts in drinking water (%)
In practice, most species exhibit an adaptation to high salt levels in the water, and the levels given in Table 2 are frequently exceeded.
Specifically toxic substances may be natural or can result from pollution. A particular risk is found in the vicinity of cattle dips when arsenic, organophosphorus, or other pesticides could contaminate water supplies. Almost all the metals and common inorganic salts can be toxic above certain levels and data on these levels are widely available [45, 46, 47].
Water points are frequent sources of disease, particularly flukes and other helminths, protozoa, and bacteria both in the water and the surroundings. Disease may also be spread when carcasses are left near watering holes, when anthrax spores can easily be disseminated, or when sick animals or faecal material contaminate supplies. Schistosomiasis, rinderpest, foot and mouth disease, salmonellosis, and brucellosis can all be spread in this way.
The availability of water is often the first consideration in the development of livestock projects. Watering stock from naturally occurring supplies is the simplest form of exploitation, all other sources requiring a degree of technology which must be appropriate to local conditions . These authors suggested that past failures in livestock projects have often been a result of water development breakdowns caused by organizational and logistical shortcomings rather than lack of technology. They urged that when water supplies are planned, cost-effectiveness, capital and operating costs, reliability and ease of maintenance, availability of supplies needed for maintenance (labour, desiltation equipment, fuel, lubricants and spares) should be fully considered and provided for. The five factors which they regarded as likely to affect the success of water development projects are thorough surveys; long-term integrated planning; sound design; community participation; and, finally, monitoring and evaluation.
Edwards et al  stated that the exploitation of an available water resource must be linked to the ecology of the pasture and to its actual or potential carrying capacity in terms of human or livestock units. They further pointed out that overdevelopment of water is not only wasteful but is likely to upset the fragile ecology of drier zones by carrying the threat of increasing animal numbers and hence overgrazing. In contrast, full development of water supplies in the high rainfall areas is likely to increase overall productivity. The overall effect of the provision of more water has been recognised by some pastoralists who were quoted by Cossins  as saying 'Put in wells a long way from here otherwise people will come and the grass will be finished'.
Such an attitude is an exception to the rule, with most inhabitants of semi-arid and arid zones welcoming water development without being aware of the possible long-term effects on the environment.
Watering can be used as a management aid by varying the frequency or by controlling the use of watering holes or boreholes. If water is supplied by pumps, stock can be forced elsewhere when these pumps are turned off. In practice, this rarely happens in Africa and is more relevant to extensive cattle production in the arid areas of Australia. Similarly, the provision of seasonal ponds (Figure 3) may extend the utilization of wet season grazing areas but as they dry up, stock will have to move elsewhere. A disadvantage of 'permanent' artificial ponds is that people come to rely upon them, settling in the proximity, thus enhancing local erosion and overgrazing, while a severe drought can cause drying up of the water source.
Altering the watering frequency of livestock has a number of advantages in range management. The first and most easily quantifiable effect of two- and three-day watering in cattle is that water consumption falls by 8 per cent and 30 per cent respectively . H. J. Weeth and A. L. Lesperance  found that heifers on once a day watering drank 1117 per cent less than heifers watered ad libitum. Where provision of water is at the expense of human labour or fuel, this represents a considerable saving (Figure 4(a)(b)).
Figure 4a, 4b. Raising water in the Borana region of Ethiopia. Three day watering of cattle reduces water consumption by 30%, a factor to be considered when labour or fuel is used for raising water.
The second benefit of three-day watering in cattle is a significant reduction in dry matter intake of at least 8 per cent in comparison with daily watered cattle . Exotic cattle may lose 20–30% of their appetite if watered daily rather than ad libitum , which is clearly not advantageous, but where dry season feed is limited any method that contributes to its conservation is desirable. While such reduction will result in loss of body condition, in terms of controlling dry-season feed the effect is either to prolong the availability of the standing crop or to allow an increase in stock numbers per unit area. A further advantage of two- and three-day watering is access to more distant grazing. A daily watered animal grazing 7 km from water has to walk 14 km a day and could cover an area of 154 km2 on centrifugal grazing. If allowed to walk 21 km away from water, it would trek 42 km over the three-day period (14 km a day) giving access to an area of 1385 km2 or nine times the original area, while walking distances would not increase (Figure 5). In camels and smallstock this effect is amplified by the more infrequent waterings. Camels have an even larger grazing orbit making them especially suited to arid environments with low carrying capacity . Horses are intolerant of infrequent watering and need water at least every two days but usually daily; donkeys and mules can adapt themselves to three-day watering with no adverse effects (Nicholson, unpublished).
Figure 5. Diagram to illustrate the exploitation of grazing by cattle dependent upon watering frequency.
One of the major problems of fixed watering points is that of localized erosion caused by the density of animals. This gives rise to 'bicycle-spoke' erosion which is the convergence of numerous cattle tracks on to the central hub where water is provided. Three-day watering of cattle and smallstock reduces this erosion, so that in practice a given water point will serve three times the number of livestock for a similar degree of erosion risk produced by livestock on daily watering.
The division of herds according to age and lactation status is a strategy followed by several pastoral groups . The Borana of Kenya and Ethiopia divide their cattle herds into the home-based milking cows and followers which are normally watered more frequently than the wandering herds of predominantly dry cows and male animals. The latter group walk long distances and are less frequently watered but exploit more distant grazing areas (Figure 6). S. Sandford  gives a number of examples of how different pastoral groups manage their livestock in relation to the local water and feed resources. In the same report he suggested that increasing the density of water points from one to one hundred per 1500 km2 would more than treble milk supply by reducing energy required for walking. Such a highly simplified model should be approached with caution, since in practice the effect of walking on steer growth and weight loss was not found to be high  even though dry matter intake is depressed . However, it is accepted that there are trade-offs between lower animal productivity caused by the long distances walked to water and the exploitation of distant grazing areas, the control of dry matter intake, and the great savings in water. It has also been pointed out that an alternative to increasing numbers of water points is a mixed herding strategy adopted by many pastoralists both between and within species that allows a more even distribution of livestock within the grazing orbit.
Figure 6. Underground well systems in Borana region, S. Ethiopia. Watering points are often few and far between in rangelands
The issue of water development and its possible detrimental effects on the ecology is a difficult and controversial subject in which trade-offs exist between short and medium term output and long term degradation. For more detailed discussion of this subject reference should be made to Sandford .
Livestock vary in their ability to conserve water and tolerate dehydration and their water requirements vary according to so many factors that no hard and fast rules can be applied. It is clear that management factors can influence water needs, in particular the frequency at which stock are watered and that intermittent watering of stock can both save water and serve as a control to dry matter intake. The water needs of different livestock types and species can be exploited to ensure more even utilization of the grazing resource around water points. Whilst it is generally accepted that maximum productivity can be accomplished only by providing water ad libitum, a possible consequence of such a strategy in pastoral areas might be an increase in animal numbers and land degradation. Above all, it should be appreciated that the provision of an unlimited supply of water is likely to be detrimental to the ecology of the semi-arid and arid zones, which are not intrinsically areas of high primary or secondary productivity. Whilst the technology for water development is available, the history of this topic in Africa is an unfortunate one. In the final analysis it is the organization and management of the water resource that will determine the future success or otherwise of water development schemes.
At the International Livestock Centre for Africa, Addis Ababa, the author is working on the effects of watering frequency on cattle. Two trials have been established in the Rift Valley in Ethiopia and the Eastern Province in Kenya using 500 Boran cattle, both breeding herds and growing steers. Comparisons are made between one-, two-, and three-day watering both on the productivity of the animals and on the physiological aspects of chronic dehydration under rangeland conditions.
Results from the trials will shortly be published but after two years, no effect on calving rates and cow productivity has been observed as a result of three day watering despite substantial losses of condition by lactating cows in the early dry season. A depression of overall water consumption of 8 per cent by two day watered and 30% by three day watered cattle is in line with the work of French , who worked with stalled zebu cattle.
Other findings include a 1020 per cent depression of dry matter intake in the dry season, resulting in a small reduction in 20 month weights of three day watered steers in comparison with daily watered steers. However, at 24 months no differences were recorded as a consequence of compensatory growth following the rainy season. Lactating animals drink 2030 per cent more than dry stock and moderately severe dehydration measured by plasma osmolarity is seen only in lactating cows in the early stages of lactation which coincides with the early dry season. Cloudy weather and access to shade both depress water consumption in all groups by 30 per cent while during the rainy season, the cattle can be independent of drinking supplies. Water is conserved in three day watered cattle by a slight reduction in faecal moisture, a lower faecal output, an inhibition of diuresis following rehydration despite low plasma osmolarity, and active shade seeking. No increases in digestibility have been observed under three day watering, despite the reduction in dry matter intake.
The trial has been recently modified to include walking distances and night penning as variables to emulate conditions frequently found in extensive and pastoralist systems of animal production in Africa.
Colour photographs by N. Cossins, ILCA.