5. Impact of Water Quality on Performance
The primary objective of a New Zealand survey of stock drinking water on sheep, cattle and deer farms (Belton et al. 1999) was to describe the water resources used, and to elicit the opinions of farmers on the quality of drinking water. About half of the farmers considered that the quality of stock drinking water had an effect on animal production, but most farmers were satisfied with the quality of stock water. Where there were perceived problems, half were due to either low flows of water (resulting in stagnant water) or high flows (resulting in dirty water), while a quarter were due to mineral contaminants and the rest were related to microbial contaminants (either algae or effluent). The survey concluded that farmers would support measures to protect the quality of water resources, although they did not consider microbial contamination a major concern.
In fact, as noted previously, few studies have actually investigated the influence of water quality on animal productivity. Most have focused on particular aspects such as microbiology or mineral content. The most comprehensive studies on the impact of water quality and productivity are the Canadian studies of Willms et al. (1996, 2000, 2002) who investigated quality in dams (measuring a number of different water components), and also carried out specific studies on the impact of faecal contamination and palatability. They were not able to identify any specific components as having particular influence except that faecal contamination influenced water palatability, and hence water consumption and productivity, as measured by weight gain.
5.1 The quality of water, and the impact on intake and productivity
The cow can detect organoleptic properties such as odour and taste, and if the water source is unpalatable, cows may avoid the water or not drink sufficient to meet production needs; causes of off-odour or off-taste are a result of physiochemical properties, or substances present in excess, or bacterial contamination (see review by Beede and Myers 2000). However, Embry et al. (1959) noted the palatability of water could be confused with toxicity from algae or associated with salts that decrease water intake at high concentrations. Response to these factors may be associated with physiological effects and not necessarily to preference or avoidance due to smell or taste.
The most common contaminant that may reduce the palatability of water is faeces. For example, Holechek (1980) reported a decrease in water consumption and weight gain in cattle drinking from a water source contaminated by faeces and urine. On the other hand, Crawford et al. (1996) found no difference in weight gain of cattle drinking water from a pond to which they had direct access compared with a clean source. Neither of these studies separated potential physiological impacts from reduced intake. However, more recently, Willms et al. (2000) conducted several studies to investigate the effects of cattle manure on acceptability of water to cattle. In one trial, cattle were offered uncontaminated water, or water contaminated with fresh manure (0.005 and 0.025% fresh manure). They found that cattle avoided the contaminated water when given a choice of clean water. They then went on to investigate the effect of contamination on the intake of water and the associated intake of feed in ad libitum fed cattle (Table 9). In these trials they compared the consumption of fresh water with contaminated water (fresh manure at rates of 0.005, 0.015, and 0.045% (Trial 1) or with 0.25, 0.50, or 0.75% (Trial 2)). There were no apparent effects on water or feed consumption in Trial 1 at the low rates, but there were small effects on water consumption, and some evidence of effects on feed consumption in Trial 2 at the higher rates of contamination. In a further study with penned cattle, animals were offered either clean water or water contaminated with either 0.25, 0.50 or 0.75% fresh manure (Table 10); the calculated ratios of the intake of water to feed were 4.9, 4.4, 3.7 and 4.2 litres per kg of feed, respectively (Willms et al. 2000).
Contamination of water with soil (soil taken from a dam to which cattle had direct access) at 5 and 15% (by weight) reduced water consumption and feed intake but had only a minor effect, if any, on the ratios of water to feed consumption (3.1 for fresh water, 3.0 and 2.9 litres per kg, respectively), although the water intake was reduced from 38 litres/head of clean water to 32 litres/head of the 15% contaminated water. These data provide some support for the proposition that cattle on dry feed will maintain a more or less constant water to dry feed intake ratio, although Utley et al. (1969) did record a decline in the ratio when water intake was restricted. However the balance of evidence indicates that any factor that reduces voluntary water consumption will also be expected to result in a reduced food intake and hence productivity.
Table 9: Effect of manure contamination on water intake (2 yearling steers per treatment over two, 5-day periods, from Willms et al. 2002, Table 7)
| Manure concentration (mg/gH2O or %) |
Water intake (litres) Period 1 |
Water intake (litres) Period 2 |
| 0 | 152 | 154 |
| 2.5 (0.25%) | 123 | 148 |
| 5.0 (0.50%) | 107 | 113 |
| 7.5 (0.75%) | 112 | 124 |
| Statistical analyses | ||
| SEM | 13 | 5 |
| Effect as probability | >0.10 | 0.003 |
The Willms team (Willms et al. 2002) claimed that the provision of clean water compared with highly contaminated water improved average daily weight gains by up to 23% (see Table 10 for a summary of some of the trials). This was based on several studies over several stock classes (eg. cows, yearlings, cows and calves), over at least 3 different sites/farms and different seasons. The improvements in water quality were brought about by techniques such as pumping dam water to troughs to avoid faecal contamination from cows entering the water and defecating or pugging the soil. However they did not directly assess the quality of the trough water, but in many cases the quality of the dam water was poor. For example, the water in the (stock-accessed) dams was often very high in sodium (e.g. around 800 ppm at one site/property and 1400 ppm at another, with 2600 ppm in the pumped water at that site) and very high in sulfates (around 1000, 6700 and 6600 ppm at the same sites as above respectively). They noted one situation in one season where the cattle gained virtually no weight and had diarrhoea, a classical symptom of salt overload.
Observations on the behaviour of cattle in the field support the notion that cattle having access to fresh water will consume more forage. For example, Willms et al. (2000) observed cattle from dawn to dusk over 5 to 10 days by and found that the animals spent considerably less time at a "drinking activity", more time grazing, and less time loafing, when they had access to fresh water than when drinking from a dam. They summarised that "One can only surmise, without becoming too anthropomorphic, that cattle drinking from an unpalatable water source tended to 'sip' and ingested less water per unit time. This is an hypothesis that obviously needs to be tested."
The Willms data show a strong avoidance of contaminated water, yet the water available to cattle directly from dams or from open water sources is often heavily contaminated by faeces, indicating that water consumption could well be reduced in such situations. They went on to attempt to put the potential for contamination into perspective, assuming an animal produces 25 kg of faeces per day with 10 defecations/day and defecates 25% each time it drinks from the dam or pond (drinks twice a day), each animal would then contaminate an equivalent volume of 5,000 litres per day with 0.25% fresh manure. Contamination will also likely be higher in shallow water. However, given no choice, cattle will drink contaminated water and intake may not be suppressed, except at concentrations of fresh manure beyond 0.25%.
Table 10: Effect of water source on growth of cattle (from Willms et al. 2002, Table 3)
| Water source | Daily live weight gain (kg/day) | |||
| Yearlings | Cow-calf pairs | |||
| Cow | Calf | |||
| Clean | 0.79 | 0.60 | 1.17 | |
| Pond water | ||||
| Trough | 0.66 | 0.48 | 1.14 | |
| Direct | 0.64 | 0.53 | 1.06 | |
| Statistical analyses | ||||
| Years (n) | 3-5 | 6 | 6 | |
| SEM | 0.09 | 0.28 | 0.06 | |
| Effect as probability | ||||
| Clean vs Trough | 0.076 | >0.10 | >0.10 | |
| Clean vs Direct | 0.045 | >0.10 | 0.056 | |
| Trough vs Direct | >0.10 | >0.10 | >0.10 | |
Although they examined a wide range of chemical and biological characteristics of the water, the researchers stated that criteria defining water quality that explained cattle performance were not likely to be associated with the chemical and biological parameters measured, but rather were defined by organic compounds that affected smell or taste. They proposed that the detrimental effect of pond water on weight gains of cattle appeared to be mediated through feed intake rather than by stress induced by pathogens, toxins or parasites. In the field studies, cattle that drank clean water spent longer time grazing and in penned studies they tended to have higher feed intakes, although the numbers of animals involved in the latter trials were few. In summarising their studies, the Willms group suggested a concentration of 1% faecal contamination as a threshold at which contaminated water would be completely rejected.
5.2 Mineral content and pH, water intake and productivity
The mineral content of water may impact on animal performance as previously noted. Weeth and co-workers have carried out a series of studies, mostly focussed on the impact of sulphate content of water on water intake (Weeth and Hunter 1971, Weeth and Capps 1972, Digesti and Weeth 1976). Other studies (Embry et al. 1959, Challis et al. 1987, Ward et al. 1992, Loneragan et al. 2001) have also revealed an impact of sulphate, or of dissolved salts. However the data from these studies are not conclusive as to the level of tolerance or of the impact of sulphate (see Tables 4 and 5).
The most interesting experiment is that of Challis et al. (1987) who compared the productivity of cows offered well-water containing 4000 to 5000 ppm of total dissolved solids (mostly as sulfates, but also chlorides and bicarbonates with some nitrates, in calcium, sodium and magnesium forms) with desalinated water (produced by reverse osmosis). The groups of cows that received desalinated water drank more water, consumed more concentrate and produced significantly more milk (nearly 7 kg more milk per cow per day) than the groups given raw well-water. However the actual cause and effect relationships cannot be deduced from this experiment, and given the apparent wide range of results from different investigators (as discussed previously), it may be that the impact of a specific component or the interaction of components may be the primary contributor to the outcome of this particular experiment.
A consideration of the contribution of water to the total intake of a particular element may be instructive, and thus the intake via the feed may also affect the likelihood of the mineral contamination of water having an effect on the animal. For example, sodium intakes from feed in New Zealand are often closer to the minimum required levels, and hence sodium from drinking water is likely to boost production by supplying a required dietary element rather than decrease it through toxicity. However on the contrary, the level of potassium in forage can be very high in New Zealand pastures, especially in the North Island. The potential significance of this is important given that dietary intakes of minerals may affect water intake by the animal, as seen in work with beef heifers comparing the water intake of animals offered diets containing high (4.2% of dietary dry matter) or low (0.6 or 1.7%) potassium levels; the higher potassium levels were associated with a significantly higher water intake (Omer and Roberts, 1967).
The situation may also be relevant for other minerals. For example, the actual contribution of iron or manganese from water to the total intake of a normal diet would generally be less than 2%, even at the proposed upper limit. However the availability of iron from the various sources may be different, and the conversion of the iron in the ferrous state to the ferric state on contact with air will actually reduce the potential intake as it is precipitated (see previous discussion).
The pH of water may impact on animal health. For example, Grant (1996) indicated that water with a pH of less than 5.5 may cause problems related to mild acidosis such as reduced milk yield, depressed milk fat percentage, low daily gains, more infectious and metabolic disease, and reduced fertility. Grant also stated that alkaline water of pH greater than 8.5 may result in problems related to mild alkalosis such as amino acid and B-vitamin deficiencies, and symptoms similar to mild acidosis. Such problems are unlikely to occur with normal sources of stock water in New Zealand.
5.3 Microbial contamination
Beede and Myers (2000) stated that contamination of a livestock water source by microorganisms typically is not of concern. However, under certain conditions, microbial populations can explode, creating problems for livestock.
Treated, as well as untreated, drinking water for animals is a potential reservoir and transmission route for Campylobacter (Savill et al. 2003), although it is not actually known whether ruminants are susceptible to infection via this route (Belton et al. 1999). However chlorine may be used to treat water for microbial contamination, and in this respect, OMAFRA (Factsheet 86-053) indicated that high chlorine levels of 50 to 100 ppm in stock water are not a problem.
5.4 Pathogen contamination
Although experimental work has found associations between pathogen (Giardia spp. or Cryptosporidium parvum.) infection and decreased animal performance in sucking animals (e.g. Harp et al. 1990, Olsen et al. 1995), field trials have found no such association in sucking and adult animals where pathogens were found to be present in the ruminant. In faecal testing of yearling cattle, Willms et al. (2000) found those that harboured cysts or oocysts of Giardia spp. or Cryptosporidium spp. (and also nematodes) had the same weight gains as those where no pathogens were detected. However the impact of a clinical loading of such pathogenic protozoa in adult animals is not known.
Contact for Enquiries
MAF Information Services
Pastoral House
25 The Terrace
PO Box 2526
Wellington, NEW ZEALAND
Fax: +64 4 894 0721
Contact this person
