4. The Factors Involved in Water Contamination
The review has outlined the issues of microbial contamination and contamination with high concentrations of particular minerals or other contaminants of water as the primary concern from the perspective of animal health. This section summarises the New Zealand work, along with some North American studies, on the most relevant aspects of contamination.
4.1 The impact of livestock and agricultural practices
Animals themselves can have a major impact on the quality of water in a catchment, principally through faecal contamination and through disturbance of watercourses and surrounding land, both of which contribute to downstream contamination. Such downstream effects have a significant impact on water quality and, potentially, on animal productivity. A number of studies in New Zealand have shown that increased farming intensity will, or is likely to, result in greater nutrient enrichment of waterways. Many of these are outlined below.
Microbes and nutrients from manure enter streams or watercourses by several pathways (Gary et al. 1983; Tiedemann et al. 1988; Larsen et al. 1994) including direct deposition from animals; via runoff or overland flow, where deposits of manure, along with organisms and nutrients in the manure, are transported to the stream; and via other pathways including sub-surface transport and filtration, and artificial drainage.
Bacterial densities in run-off from agricultural lands often exceed water quality standards (Baxter-Potter and Gilliland 1988). This is not surprising given concentrations of 100 million coliforms per litre in bovine faeces (Thelin and Gifford 1983). Several studies have shown that the extent of, or levels of, bacterial indicator organisms in streams in the US are proportional to cattle numbers, and inversely related to the area of pasture (Darling and Coltharp 1973, Gary et al. 1983). In New Zealand, Bagshaw (2002) quantified defecation by beef cattle on hill country and found that the number of defecations in the riparian zone was not influenced by the season, the presence of a trough, other resources next to the stream, the size of the field or pasture availability.
Although it has been suggested that the impact of defecation on aquatic ecosystems would be reduced by excluding stock from streams by fences and by providing alternative water in troughs, fencing would be economically prohibitive in much of the United States (Platts and Wagstaff 1984). While fencing may also be regarded as economically prohibitive in New Zealand (Bettjeman 1997), the reality is that it is now becoming necessary in order to meet supplier requirements, such as the Fonterra Clean Streams Accord now being introduced. In addition, regional and local authorities around New Zealand are also moving towards measures that will encourage or enforce waterway fencing for all pastoral farms.
The effect on water quality due to agricultural practices including stock having direct access to waterways has been illustrated in several studies in New Zealand as outlined below.
Measurements of suspended sediment, ammonia and faecal coliforms were taken upstream and downstream of a stock access point on a deer farm in Southland (Environment Southland 2000). Downstream levels were 19 to 35 times higher than upstream, exceeding the recommended guidelines for stock water (Parkyn et al. 2002), and ammonia concentration was at levels toxic to fish life. In a study where deer had access to a small stream near Te Anau, water had a turbidity measurement of 280 NTU (nephelometric turbidity units) with 32,000 faecal coliforms/litre; where deer were not present, the values were 0.4 and 710 coliforms/litre respectively (Rodway 2003). Similarly in hill country pasture in the Waikato, Quinn et al. (1998) showed that turbidity levels downstream of cattle wading in a stream increased dramatically from less than 10 to 50-250 NTU.
Sampling of stream water after a herd of around 250 cows had crossed twice showed a 50% increase in suspended sediment and a 400% increase in E. coli (Davies-Colley et al. 2001). This was associated with the disturbance of the streambed and banks, and defecation (the cows deposited about 37 kg of faeces into the stream).
Vant (1999) found that the nitrogen yield in eight large Waikato catchments was strongly correlated with the stocking density of dairy cows. In other work, Davies-Colley and Nagels (2002) studied the effects of dairying on water quality of eight lowland streams and rivers in New Zealand. The median conductivity, phosphorus and nitrogen concentrations and, to a lesser extent, faecal contamination (as indicated by E. coli concentrations), were all elevated in streams draining dairy pasture. The Waikato stream in the most intensively-farmed region had very high concentrations of nutrients and high faecal contamination. Dissolved reactive phosphorus (DRP) ranged 70-fold, total phosphorus ranged 30-fold, and total nitrogen 20-fold, across the sample sites. Faecal contamination (per E. coli) showed a broadly similar trend to that of nutrients and conductivity. Hamill and McBride (2003) examined trends in water quality variables at 29 river sites in Southland over a 6-year period in relation to the expansion of dairy farming in the area. An increase in the concentrations of DRP was associated with the increase in dairy farming. Although other water quality parameters deteriorated (oxidised N and dissolved oxygen), these trends also occurred in non-dairying catchments.
Similarly, overseas studies have shown that nutrient (nitrate and orthophosphate, Banasik et al. 1999) and faecal bacterial (Gary et al. 1983; Tiedemann et al. 1988) concentrations in streams and rivers increase with the intensification of agricultural land use.
4.2 Microbial contamination
A major New Zealand Microbiological Freshwater Project (McBride et al. 2000) covered surface waters (ponds, lakes, streams and rivers) but not groundwater. The main interest was in pathogenic organisms and so the survey did not include toxigenic organisms (such as blue-green algae) or chemical agents, which may also be of concern for livestock production. The presence of E. coli, the most commonly-used indicator of faecal contamination of fresh water, was detected in 99% of all samples. The detection rate of Campylobacter was 60% with the highest rates in the late summer-early autumn period with C. jejuni being the most frequently identified species (present in at least 48% of the positive samples). The largest proportion of high Campylobacter values occurred in the sheep/pastoral use catchments. The overall detection rate of Salmonella was low (10% of samples), with the highest incidence in August; it was associated with disease outbreaks in sheep and associated human cases. Giardia cysts and Cryptosporidium oocysts were detected infrequently (8% and 5% respectively) and at low concentrations in all catchment types. Catchments inhabited by birds (seagulls and waterfowl) were the most contaminated, across nearly all micro-organisms.
Contamination of the water is inevitable where animals can wander freely into the water source such as streams, ponds or dams. However faecal and urine contamination, as well as contamination via saliva or regurgitation, also occurs in troughs. Such contamination will add nutrients as well as bacteria, and may reduce palatability. For example, a small investigation of faecal contamination of trough water by grazing livestock was conducted by Belton et al. (1999) as part of the Microbiological Freshwater Project. Troughs are used for supplying drinking water to livestock onNew Zealand dairy farms and a significant proportion of drystock farms. The rationale was that livestock contamination of trough water by regurgitation of rumen contents, and direct faecal splashing is likely to significantly affect trough water quality as measured by faecal coliforms, total coliforms and E. coli. The study confirmed that faecal contamination of trough water supplies by grazing livestock is a normal occurrence on farms where water is reticulated via troughs. There was significant between-farm variation in the quality of source water as measured by total coliforms. On every property there was significant contamination of trough water with E. coli (total coliforms often exceeding 104 per litre and E. coli of more than 103 per litre) when cattle were grazing paddocks in which the sampled troughs were situated, indicating contamination of the trough waters by the grazing livestock.
A large US study of livestock drinking water (LeJeune et al. 2001) investigated 473 water troughs on 98 dairy farms. The authors concluded that troughs are a major source of exposure of cattle to enteric bacteria, including a number of food-borne pathogens, and the degree of bacterial contamination appeared to be associated with potentially controllable factors. The results of the study indicated that drinking water offered to cattle is often of poor microbiological quality with total coliforms and E. coli counts of around 105 and 104 cfu per litre respectively. E. coli O157 was isolated from 1.3% and Salmonella spp. from 0.8% of troughs. Interestingly, metal troughs had significantly lower coliform and E. coli counts (1.53 and 0.71 log10 cfu/g) compared with other construction materials (1.8 and 1.1, 2.0 and 1.1, and 2.6 and 1.4 log10 cfu/g, respectively, for concrete, plastic, and other materials); there were no significant differences in protozoal counts. The group also found that bacterial contamination was higher in troughs that were closest to the feed-trough. Proximity of the troughs to the feed-trough may have permitted a greater amount of feed to enter the trough, thus increasing the level of contamination, as well as providing a nutrient-rich substrate for bacterial growth at the bottom of the trough.
LeJeune et al. (2001) also noted the association between the quality of the water and the ecological parameters measured, suggesting that many of the same factors that influence the survival and proliferation of bacteria in natural aquatic ecosystems have parallels in water trough environments. Bacterial contaminants in troughs may arise from multiple sources (eg. cud or faecal material, and extraneous matter including dust or feed). In some instances, depending on the source, water may be heavily contaminated before it enters the trough. Overland and sub-surface flow of faeces into waterways are also likely to play a part in bacterial dissemination as E. coli can survive in bovine faeces for several weeks (Wang et al. 1996). Sediments within a trough may have much higher levels of microbial contamination (eg. 1000 times higher as reported by Ashbolt et al. 1993).
Two US surveys in Wisconsin have noted the role of drinking water in the dissemination of E. coli strains in dairy herds (Faith et al. 1996; Shere et al. 1998). E. coli survives for long periods in drinking water, particularly at lower temperatures (Rice and Johnson 1992, 2000; Wang and Doyle 1998). The E. coli issue is of particular concern due to the role of the O157 strain in many outbreaks of food poisoning, particularly in the US. This is much less of an issue in New Zealand with its grazing ruminants, as this strain thrives in the more acidic rumen and colon environments of the grain-fed animal (Russell et al. 2000). However the organism is widespread in the environment and pre-ruminants such as young calves are likely to be at risk.
4.3 Mineral and nutrient contamination
A US study investigating the variability of water composition (focussing mainly on the impact of mineral content on acceptability and animal performance) collected over 3600 water samples from livestock operations throughout the US (Socha et al. 2001). There was considerable variability in composition between and within regions. The average concentration in water did not exceed the upper desired level as shown in Table 7 for calcium, chlorine, copper, magnesium, sodium, sulfate, or zinc, but a significant proportion of samples did exceed the desired limits for calcium, sodium and sulfates, and for iron and manganese (Socha et al. 2003).
The authors claimed that the results indicated that water quality on a number of dairy farms could be limiting animal performance. Point of collection and source of water (surface, well, spring, etc.) were not recorded. Well-water samples are commonly collected as close to the well-head as possible. However, the researchers noted that it might be better to collect a sample close to the waterer to accurately assess the quality of water the cow is actually drinking. In this respect, Ensley (2000) found a difference in the concentration of iron and coliforms from the well-head to the drinking waterer. The iron content of water declines when water mixes with air, with oxidation from the soluble ferrous state to the less soluble ferric state resulting in precipitation of the iron, and less being available to livestock than what is measured at the well. The influence of the level and availability of iron on bacterial growth in troughs should also be considered, as iron is often the most-limiting nutrient for microbial growth (eg. as a case in point, this is the rationale behind the iron enrichment experiments in the Great Southern Ocean, Boyd et al. 2000)
Another US study (National Animal Health Monitoring System, Wagner et al. 2001) monitored water quality in feedlots with 1000 head or more capacity (a total of 263 feedlots from 10 states). The average nitrate, sulfate, and TDS content were 334, 205, and 800 mg/litre, respectively. No samples exceeded the recommended limit for nitrate. Approximately 23% of the samples had a sulfate concentration greater than 300 mg/litre, but less than 3% exceeded the recommended upper limit for TDS. The authors noted that a study at one site found water sulfate concentration greater than 1000 mg/litre and that this caused an increase in ruminal gas cap hydrogen sulfide concentration and a reduction in water intake and feedlot performance. In this respect, a very high concentration of sulfate in water (7200 mg/litre, Hamlen et al. 1993) has been associated with an increased incidence of polio-encephalomalacia (thiamine deficiency) in cattle. The normal source of thiamine in the ruminant is the digestion of ruminal microbes.
4.4 Other factors affecting water quality
Seasonal fluctuations were observed in the US trough-water study (LeJeune et al. 2001), with higher bacterial counts in summer. These paralleled the seasonal trend in total bacterial counts reported in a longitudinal study of troughs on a single farm (van der Veer 1992), and the issues around algal contamination in this season. Thus the greater exposure to bacterial (and algal) contamination combined with greater water demands by livestock in summer means that a greater effect on productivity would be expected in summer. While direct sunlight (via the strongly bactericidal ultra-violet component) would be expected to reduce the survival of E. coli in troughs as in other aquatic systems (Sinton et al. 1999, 2002), it may not be a very significant factor alongside the other factors influencing bacterial load, such as direct contamination. In this respect, LeJeune et al. (2001) also found lower E. coli densities in the troughs exposed to direct sunlight compared with shaded troughs. Interestingly these workers also found an inverse relationship between E. coli counts and protozoal counts. Competition between microbial species and predation, such as protozoal predation of bacteria, is also likely to be a very important factor influencing the bacterial population in the water supply (Mallory et al. 1983).
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
