View all publications

Pressures on our water environment

We tend to under-estimate how much water we use directly and indirectly simply to get through an average day. The total amount harnessed each year for electricity production, crop and livestock production, industrial production, and normal household activities has been estimated at almost 102,000 million cubic metres (Table 7.4.). Broken down to a scale we can deal with this comes to some 82,000 litres per person per day. When the huge amount used for hydroelectricity generation is subtracted, the total still exceeds 1,500 litres per person per day. Much of this is used by agriculture and industry in providing the products and services which are an integral part of our economy and lifestyle, but about 160 litres are used by each of us personally in the home and at the workplace. In addition to this we use large amounts of unharnessed water, such as the rain that falls on our gardens, crops and pastures, and the natural waterways that nurture the fish we eat and provide us with recreational opportunities.

With such a variety of water uses, it is not surprising that people have had an impact on New Zealand's water flows and water quality. Some of the heaviest impacts, however, have not come from water use, but from land use. By changing the vegetation cover we have altered the amount of rainfall that runs off the land into streams and rivers. The greatest single source of pressure was probably the removal of our hill and riparian (riverbank) forests for pastoral agriculture. This caused an increase in the scale of erosion, sedimentation and flooding. Most of the deforestation occurred before 1920 (see Chapter 8), but the continuing use of deforested hill and riparian areas for livestock production is a persistent source of pressure on streams and rivers, causing pollution from animal waste, sediment and fertilisers.

Table 7.4: Yearly use of harnessed water in New Zealand.
Water Use Quantity (Mm3/y)
Electricity generation 100,000
Irrigation 1,100
Livestock consumption 350
Industry consumption 260
Household consumption 210
  • Bathroom (assuming one 5-minute shower a day)
  • Lavatory (assuming five 11-litre flushes a day)
  • Washing machine (varies from 70-240 litres per cycle)
  • Kitchen (7 litres per half-sink, 24-45 litres per dishwasher cycle)
  • Outdoors (summer hosing for 30 minutes per day)
Total yearly use 101,920

Sources: McConchie (1992); Mosley (1993)

New Zealand's water has also come under heavy pressure from urban populations and industry. Many are situated on estuaries and in former wetlands where drainage and urban development have destroyed or modified the aquatic ecosystems. The increasing urban demand for water has reduced the levels of some rivers and aquifers, and the demand for electricity has led to river flows being disrupted by dams. In some areas, the increasing consumption of water by households and industry has led to water shortages and costly pipeline and reservoir extensions. In some areas, too, the increasing production of human and chemical waste has led to water pollution from sewage, leachates, and stormwater run-off.

Several surveys in recent years have asked local authorities to identify and rank the pressures on water quality in their areas. One survey covered local authorities, the Department of Conservation, and Fish and Game Councils (C.M. Smith, 1993), while another concentrated on regional councils (Hoare and Rowe, 1992). The most serious pressures identified in both surveys were:

  • sedimentation and nutrient enrichment (eutrophication) of surface waters by agricultural run-off and urban stormwater;
  • point source pollution in some lower reaches of streams and rivers; and
  • nitrate contamination of groundwater.

In a similar survey by the Ministry of Agriculture and Fisheries, regional officials ranked agriculture as the greatest source of pressure on water quality, closely followed by urban sewage (see Table 7.5). Below these, the officials ranked urban stormwater, industrial wastes and agricultural processing as important sources of pressure, and, further below these, mining and forestry. The main agricultural impacts were identified as sedimentation, nutrient and microbiological contamination, altered physical characteristics of surface waters, and nitrate contamination of groundwater (Sinner, 1992).

In summary, the various pressures on water can be grouped as: agricultural pressures (including vegetation clearance, land drainage and channelling, draw-off for irrigation and stock watering, and run-off and waste discharges from farms and agricultural processing facilities); urban pressures (including sewage and industrial waste, stormwater run-off, draw-off for household and industrial uses; and urban expansion into wetlands and estuaries); dams (including hydro-electric and water supply dams); and forestry and mining. In addition to these pressures, concern has also been expressed about the impacts on natural waterways from introduced pests and weeds and from the potential impacts of climate change. A brief discussion of each of these pressures follows.

Table 7.5: Impacts on water quality as ranked by regional officials.

Sources of impacts on water quality

Source Average rank1
Agriculture 4.9
Human sewage 4.8
Urban storm water 3.9
Industry 3.8
Agricultural processing 3.7
Mining 2.6
Forestry 2.6
Agricultural impacts on water quality
Type of Impact Average rank1
Sedimentation 6.4
Nutrient contamination 6.2
Alteration of physical characteristics 5.6
Faecal contamination-surface water 5.4
Nitrate contamination-ground water 4.6
Pesticide contamination-surface water 2.8
Faecal contamination-ground water 2.8
Pesticide contamination-ground water 1.6

Source: Sinner (1992)

1Ranked on a scale from 0 = no damage to 10 = severe damage

Agricultural pressures

Agriculture outranks other sources of pressure on water largely because of the scale of pastoral farming. Pasture grass and farm animals dominate more than half of New Zealand's land surface, and affect nearly all catchments. The livestock populations excrete about 40 times more organic waste than New Zealand's human population. The treeless pastures are a constant source of run-off, washing some of this waste, as well as sediment and fertiliser residues, into waterways. The animals and irrigated plants consume three times more water from rivers and aquifers than all the country's households and industrial sites combined. Pastoral products are also a major source of organic waste discharges from processing industries such as meatworks and dairy factories.

In recent decades some of these pressures have improved noticeably as the agricultural sector has become more aware of the problems and as waste water treatment technologies have improved. Economic conditions and soil conservation initiatives have also brought trees and scrub back to some of the steeper pastures. The extent of these trends has not been measured, however, and on current evidence, pastoral agriculture remains a significant source of pressure on New Zealand's water.

Other forms of agriculture also consume water and generate pollutants (e.g. horticulture, cropping, pig and chicken factory farms). These may have significant impacts in a particular locality, but their national impact is small compared to pastoral agriculture. Pasture covers 14 million hectares of land, whereas crops and horticulture cover barely 400,000 hectares (see Chapter 8). The estimated daily BOD5 loading from the organic waste of 9.3 million cattle, 48.8 million sheep, 0.3 million farm goats, and 1.2 million farm deer comes to over 10,000 tonnes, compared to a daily BOD5 loading of less than 5 tonnes from the nation's 430,000 pigs, and about 380 tonnes from the 54 million chickens, most of which is removed as dry waste.

This calculation assumes an average daily BOD5 load for individual animals of 980 grams for pasture-fed cattle, 32 grams for sheep and goats, 10-12 grams for pigs, and 7 grams for chickens (Vanderholm, 1984). The comparable human load is about 70 grams, although the discharge of industrial waste into domestic sewers increases the daily BOD5per person to 100 grams (Hauber, 1995). BOD5 varies considerably with size and diet, but these averages provide a basis for comparing the waste load of different species.

Agricultural pressures on water stocks and flows

The three main pressures which farming exerts on water flows are pastoral land use (which determines the dominant vegetation cover in most catchments and hence the amount and quality of rainwater that runs off into streams and rivers), irrigation (which takes water from rivers and aquifers and thereby affects their volumes and flows), and land drainage (which continues to exert major effects on the extent and depth of wetlands, and the volume of groundwater resources in most lowland floodplains).

Deforestation pressures on water flows

Before people arrived here, New Zealand was a forested country. The forests acted as natural sponges for much of the rainfall, with tree leaves and mosses intercepting the rain as it fell, and tree roots sucking the water out of the soil. Indigenous forests, plantation pine forests and, to a lesser extent, the native tussock grasses, all have a similar effect on the water cycle, absorbing a sizeable proportion of rainfall and allowing it to evaporate and transpire back into the atmosphere. Exotic pasture grasses also intercept and absorb water, but not to the same extent as forests and tussock. The result is greater water run-off in pastoral catchments than in forested or tussock-dominated ones.

Vegetation change began with Māori settlement. Large areas of forest were burnt off causing considerable sedimentation and flood flows in some areas, though not in others (McGlone, 1989; Trustrum and Page, 1991). Dry naturally erodible areas, such as the South Island high country, were heavily affected, but wetter areas, such as the East Coast of the North Island were relatively unscathed because of the the quick regrowth of deep-rooted bracken fern and small shrubs which held the soil. Following European settlement, however, large areas of forest, tussock and scrub were replaced with short, shallow-rooted, pasture grasses (see Chapter 8). This wholesale vegetation change increased the amount of water running off the land, causing higher river flows, more frequent floods, increased erosion rates and increased water sedimentation. The removal of streambank vegetation contributed to the loss of water quality and destruction of aquatic habitat (see Box 7.4).

Evidence that vegetation change significantly affects river flows has been obtained from several catchment studies in various parts of New Zealand, including Berwick Forest in Otago (Smith, 1987), Glendhu Forest in Otago (Fahey and Watson, 1991), Moutere near Nelson (Duncan, 1994), Tarawera in the Bay of Plenty (Dons, 1986; Pang, 1993) and Purukohukohu near Rotorua (Dons, 1987).

In the Purukohukohu study several small catchments with different vegetation covers were compared (Dons, 1987). One catchment was predominantly covered in pasture, one was covered in pine forest, and one was covered in indigenous forest. In the forested catchments 70-75 percent of rainfall was absorbed by vegetation compared to only 55 percent in the pasture catchment (see Figure 7.7). Run-off from the forested catchments was only half to two-thirds that of the pasture catchment. As a result, the pasture catchment had higher average stream flows and flood flows.

Figure 7.7: Patterns of water run-off from pasture, native forest and pine forest catchments at Purukohukohu, central North Island

The majority of water run-off from pasture, native forest and pine forest is from evaporation. On pasture, the proportion of water run-off by both quick and delayed run-off is greater than for native and pine forest. The proportion lost as ground water is similar for all.

Source: Dons (1987)

Figure 7.8: Tarawera river flows before and after catchment afforestation in the 1960s

Tararua river flows have declined since 1963 when they peaked with a 3 year average of around 40 cumecs. Since then, flows have decreased, and in 1991 the 3 year average was around 23 cumecs.

Source: Pang (1993)

More dramatic results were obtained from a 30-year study of a small catchment at Moutere, Nelson, where pine forest replaced pasture (Duncan, 1994). Average stream flows, flood flows, and low flows were reduced by between 50 percent and 80 percent. Extreme floods were reduced to such an extent that even a 50-year flood would now produce only half as much flow as before.

A study of vegetation changes in the much larger catchment surrounding the Tarawera River came up with similar results (Pang, 1993). In this 90,000 hectare catchment, some 25,000 hectares (28 percent of the catchment) were converted from a light cover of regenerating forest (40 percent) and scrub (60 percent) to pine forest. This caused a drop of about 13 percent (4.5 cumecs) in the average yearly flows of the Tarawera River (Pang, 1993). Declining rainfall caused a further flow reduction of 6.4 cumecs, making a total reduction of about 11 cumecs (see Figure 7.8). The impact of afforestation at Tarawera was similar to that found in Glendhu forest where 67 percent afforestation caused a flow reduction of 25 percent, and where complete afforestation of the catchment could reduce flows by 45 percent (Fahey, 1994).

These experiments corroborate the evidence from sediment studies suggesting that the deforestation of New Zealand has had a considerable impact on river flows (McSavaney and Whitehouse, 1989; McGlone, 1989; Trustrum and Page, 1991). They also suggest that an increase of tree cover on pasture land will reduce peak flood flows and average stream flows, but may cause problems in some catchments for downstream water uses and for maintaining water levels in surviving wetlands.

However, it should be noted, that the effects of vegetation change on river flows do not apply uniformly to all catchments. Pine forest was planted over 21 percent of the 26,000 hectare Esk catchment in Hawke's Bay but had no affect on flow rates (Black, 1991). The Esk river is in one of New Zealand's drier areas, so there was little rain for the newly established forest to intercept. It appears that most of the Esk's water enters as run-off and seepage from limestone rocks high up in the catchment headwaters, where rainfall is also highest. This has far more effect on flow rates than the comparatively small amount of rainfall that runs off pasture and forest lower down in the catchment. The obvious conclusion to be drawn from this is that the distribution of rainfall within a catchment is as important as the distribution of vegetation cover. In dry areas, therefore, afforestation of the lower part of a catchment is unlikely to have any great impact on downstream water uses.

Box 7.4: Banks in crisis

Before humans arrived, New Zealand's stream banks, or riparian zones, were also biodiversity banks. Most lowland streams were bordered by dense native forest teeming with at least 50 species of vascular plants and many species of mosses and fungi. The riparian zones provided key habitat for half our native land birds, many of which are now extinct or threatened (e.g. the blue duck, brown teal and even the takahe). The stream banks were also home to native frogs, skinks and lizards, slugs, snails, flatworms, earthworms and nematodes, and many insect species, including half the world's Taenarthus carabid beetles and many species of Tanyderidae crane flies (Collier, 1995). The overhanging vegetation and falling leaves and branches provided shade, food and habitat for native fish such as the extinct grayling, the banded kokopu, the shortjawed kokopu and the koaro. Mudfish dwelt in the holes around tree roots and in the moist earth of the stream banks and wetland edges. The shade also kept algae at low levels and maintained cool water temperatures, even through summer. The tree roots limited bank erosion, keeping the water clear of sediment (Gilliam et al., 1992).

Today, most stream banks have been cleared of their vegetation to create pasture. Where hundreds of species once dwelt, there are now just a few species of exotic grass accompanied by some sheep or cattle and the occasional thistle, sedge or lone clump of cabbage trees. Loss of shade has increased light levels in many lowland streams, driving the native fish away and leading to the proliferation of algae and introduced weeds. The situation has been made worse by the increased amount of run-off containing nutrients from animal waste and fertilisers and sediment from eroding hillsides and collapsing streambanks (Collier, 1995).

Awareness of these problems has led to riparian protection measures in some parts of the country, notably the streams draining into Lake Taupo which became riparian protection zones in the 1970s. In most cases, however, riparian protection has been undertaken by landowners only where significant bank erosion has demanded it. Where collapsing banks have caused serious pasture loss introduced willow and poplar trees have been planted. Because of their fast growth rate, these species can quickly restore bank stability but have done little to restore biodiversity. Some species (especially the crack willow, Salix fragilis) invade channels and destroy habitat for aquatic life (Collier, 1995).

Protecting the natural character of rivers and their margins is among the matters of national importance set down in the Resource Management Act and, as a consequence, a number of district and regional councils are now actively promoting riparian retirement planting and conservation. Taranaki Regional Council, for example, has adopted a riparian management strategy whose key elements are education, advocacy and free technical advice to interested landowners. Between 1993 and 1996, the council prepared 45 personalised riparian management plans for individual landowners (Taranaki Regional Council, 1996). To assist councils and private landowners, the Department of Conservation commissioned NIWA to develop guidelines for the management of riparian zones (Collier, et al., 1995). These have been published as a two volume set and have become an unexpected 'best seller' for DoC. By April 1996, copies had been sent to more than 600 locations (Green, 1996). The riparian guidelines outline techniques for retaining or restoring bank stability and biodiversity through: protecting remnant vegetation; planting trees and shrubs; maintaining a vegetated ground cover; protecting wetlands; controlling livestock, and planting and maintaining native riparian plants. Otago Regional Council has also recently published a report called Riparian Management which is a set of local guidelines for different types of land activity in Otago.

Although no native plants can match the vigour and protection of willows, several show promise for bank stabilisation where erosion is less severe, or where willows have already stabilised a bank. Some also provide food for birds. These plants include makomako ( Aristotelia serrata), kotukutuku (Fuschia excorticata), kawakawa (Macropiper excelsum), mahoe ( Melictus ramiflorus), karo (Pittosporum crassifolium), pate (Schefflera digitata), kowhai ( Sophora spp.) and pohutukawa (Metrosideros excelsa ). The popularity of the guidelines suggests a growing interest in stream and river bank conservation. Although it appears to involve only a fraction of the country's 80,000 or so farms at present, it is a much needed step toward improving stream quality and bringing biodiversity back to the farm.

Irrigation pressures

Every year, pastoral farmers, orchardists and market gardeners remove (or abstract) approximately 1,500 million cubic metres of water from surface waters and groundwater (Table 7.4). This is an estimate, as no national statistics exist on water abstraction. Much of the removed water returns to rivers and groundwater, but a large amount is absorbed by plants, soil and animals. Such a large draw-off can potentially reduce the water flows available for urban water supplies, recreational activities (e.g. swimming, fishing, boating), habitat (e.g. for fish and insects), and ecological processes (such as dilution and the prevention of eutrophication).

The bulk of the abstracted water is used to irrigate pasture, or is consumed directly by farm animals. Although horticulture is a heavy user of irrigation water, its overall share is relatively small because of the smaller land area involved. Horticulture covers 95,000 hectares, not all of which is irrigated, while the total irrigated land area is more than 300,000 hectares. About 50 large schemes, with a potential coverage of about 160,000 hectares, were established with government assistance between 1910 and 1987. The government share of these schemes was sold in 1989 and 1990, and all the irrigation schemes are now privately owned. Most of the irrigated land is pasture land in the South Island, particularly Central Otago and Canterbury. The much smaller area irrigated for horticulture, is mostly in drier parts of the North Island.

Agriculture's demand for water is highest during summer months when river flows and groundwater levels are at their lowest. This has led to intensive competition for water in some rivers and aquifers, particularly in cropping, market gardening, and horticultural areas such as Canterbury, South Auckland, the Waimea Plains, and Hawke's Bay (Armstrong, 1993). Large-scale irrigation can reduce river flows to the point where fish, wetlands, and recreational activities are affected. In Canterbury, for example, where up to 30 cumecs (m3/s) of water are drawn from the Rangitata River, the flow halves during the irrigation season from a yearly average of 98 cumecs to about 50 cumecs (Mosley, 1993).

The effects of irrigation were even greater under the old system of notified water rights. During the 1985 drought, when the combined flow of all streams and rivers in South Canterbury fell from an average of 180 cumecs to just 80, irrigators were entitled to draw off 51 cumecs, leaving a downstream flow of barely 30 cumecs (Scarf, 1988). That same year, Canterbury's Rakaia River became one of the first to be protected under a Water Conservation Order, partly because of the potential impacts of irrigation (see Table 7.13). Under the new system of water management plans, irrigation draw-offs from these rivers have been reduced.

Quite apart from its impact on river flows, irrigation also has the potential to affect groundwater quality in some areas through the leaching of nutrients and contaminants. The effect is only detectable in areas where shallow aquifers underlie thin soils and where rainfall is low enough to allow irrigation to play a dominant role, such as the Levels Plain and Temuka area in South Canterbury (Smith, 1994).

After decades of continual expansion, large-scale irrigation development halted when Government assistance came to an end in the mid-1980s. The prolonged spate of El Niño-associated dry seasons between 1989 and 1995 led to revived interest among some farming groups in developing new irrigation schemes, particularly in Canterbury. To date, the only recent schemes have been small ones, mostly on dairy farms. The small schemes draw off less water than the large formerly subsidised schemes, but, in some areas, their combined impact may have the potential to be significant.

Land drainage pressures

Beginning last century, land drainage has had a marked effect on New Zealand's wetlands, and also on some rivers, shallow lakes and groundwater. The main reason for draining large areas of low-lying land has been to increase the area of pasture, but flood control schemes and urban development have also played significant roles. Most of the changes occurred between 1920 and 1980, but in some areas, the drainage work is continuing, and even increasing, partly in association with the increase in dairying farming.

The drainage schemes have removed most of the nation's wetlands and have altered the natural character of some rivers and shallow lakes. Rivers were straightened and stopbanked, lake levels fell, and, some groundwater levels were lowered. Today, New Zealand is criss-crossed by several thousand kilometres of channels and ditches which quickly divert unwanted water into straightened rivers and eutrophying lakes. The drainage schemes have also caused greater scouring of rivers and riverbeds because of the increased volume of water flowing through them. The result in many rivers is a gradual rising (or aggradation) of the riverbed, causing peak flood flows to also rise. If the drainage schemes had caused the wholesale disappearance of lakes they would have been questioned much sooner, but communities have allowed their wetlands to disappear quietly with scarcely a question raised, despite a National Wetlands Policy approved by Cabinet in 1986. Some regional councils (e.g. Bay of Plenty) now restrict drainage in vulnerable catchments. While the rate of drainage of wetlands for conversion to pasture has declined, new land development methods such as 'humping and hollowing' are still having impacts on adjoining wetlands.

Agricultural pressures on water quality

Because it extends over half New Zealand's land area, agriculture dominates the middle and lower catchments of most streams and rivers. In the course of each year, significant parts of these catchments are defecated on by millions of farm animals, sprayed with fertilisers and pesticides, and rained on. As a result, tonnes of faecal matter, nutrients (i.e. nitrogen and phosphorus), and sediment are washed into surface waters, while nutrients and other contaminants leach into groundwaters. Additional organic waste is discharged into surface waters from facilities that process agricultural products and animal carcasses.

Where these pollutants are discharged from specific sites (e.g. meatworks, dairysheds, piggeries), they are referred to as point source pollutants. Where they wash into streams as run-off from land surfaces (e.g. paddocks, roads, forests) they are referred to as non-point source pollutants. Surveys in the 1970s showed that agricultural run-off accounted for 58 percent of the biochemical oxygen demand (BOD5) in freshwater, 45 percent of total phosphorus (TP), and 88 percent of total nitrogen (TN) (McColl and Hughes, 1981; McColl, 1982). A decade later, agricultural non-point sources were still the main cause of water pollution, accounting for 75 percent of the total nitrogen loading to surface waters (see Table 7.6 and Figure 7.9).

Table 7.6: Estimated yearly nitrogen loadings to New Zealand surface waters.
Non-point sources 122,000 tonnes
Agriculture 100,000 tonnes
Native forest 15,000 tonnes
Exotic forest 7,000 tonnes
Point sources 10,200 tonnes
Agriculture 7,000 tonnes
Sewage/urban 2,400 tonnes
Pulp and paper 800 tonnes

Source: Cooper, 1992

Figure 7.9: Estimated yearly nitrogen loadings to New Zealand surface waters

The largest nitrogen discharges come from agriculture (over 100 thousand tonnes) with a much greater proportion coming from non-point sources rather than point sources. Less than 20000 tonnes is from non-point native forest sources and around 10000 tonnes from non-point exotic forest sources. Urban sewage and pulp and paper point sources are both below 5000 tonnes.

Source: Cooper, 1992

Figure 7.10: The major point sources of organic pollution (BOD5) in surface water

Approximate percentage of all point source organic pollution (BOD5) in the main sources:

  • Meat processing factories: 25% before discharge; 31% discharged.
  • Farm dairies (cow sheds): 15% before discharge; 27% discharged.
  • Dairy factories: 17% before discharge; 13% discharged.
  • Pulp and paper mills: 6% before discharge; 13% discharged.
  • Piggeries: 5% before discharge; 9% discharged.
  • Reticulated sewage plants: 33% before discharge; 9% discharged.

Source: Hickey and Rutherford (1986)

Point source discharges into surface water

Ten years ago, meatworks and dairysheds were particularly important point sources of organic waste pollution (see Figure 7.10). Today, meatworks have declined in number, but dairysheds have increased, with milk production rising by 23 percent through the decade 1986-1996. Production from dairy factories and from pulp and paper mills has also increased. It is difficult to estimate the significance of these changes because the treatment and disposal of waste from all the major point sources has improved since 1986 (see Box 7.5).

Other point sources associated with agriculture include stockyards, holding pens, rendering plants, wool scourers, tanneries, canneries, and fertiliser plants. On some farms they also include rubbish dumps, sheep dips, and storage sites for pesticides, fertilisers and fuels.

Non-point source run-off into surface water

Most non-point source pollution is caused by rainwater washing organic matter, sediment and nutrients from land surfaces into streams, rivers and lakes. Non-point source pollution also occurs when nutrients or other contaminants are leached through the soil into groundwater. Apart from urban stormwater run-off, most of the significant non-point sources are associated with agriculture simply because agriculture dominates the land surrounding most streams. A decade ago, about 30 percent of New Zealand's pastoral land (more than 4 million hectares) was estimated to be adversely affecting water quality through run-off (Wilcock, 1986).

The organic matter in non-point source pollution comes mostly from farm animals. Faecal contamination of surface waters is most severe in areas of high cattle density (e.g. dairy farms). In 1995, the faeces from New Zealand's 9.3 million cattle, 49.1 million sheep and farm goats, and 1.2 million deer produced a waste load equivalent to a human population of 153 million people, up from 142 million just two years earlier. The vast majority of this came from cattle whose total BOD5 loading was about 9,000 tonnes per day in 1995, compared to about 1,600 tonnes from the more numerous sheep and goat population. Much of this waste entered the soil and was re-absorbed by plants, some was leached into groundwater, and a significant amount was washed into streams, rivers and lakes.

The sediment in non-point source pollution comes mostly from deforested slopes that have been converted to pasture. Rivers which drain pastoral catchments tend to have sediment loads 2 to 5 times greater than those flowing from forested catchments. Where the bedrock is particularly erodible, the sediment loads may be even greater. The impacts of deforestation are particularly great during the removal of vegetation, when sediment loads in streams and rivers may increase 100-fold. Sediment loads generally decline as vegetation is re-established, but this may take 5-10 years.

The quality of the pasture growth also affects sedimentation rates. Dense pasture growth has relatively low sedimentation rates (but may have high concentrations of unabsorbed fertilisers and animal waste), while sparser grass cover has high sedimentation rates (but lower concentrations of fertiliser and animal waste). Urban land development and forest harvesting operations can also cause significant sedimentation of surface waters, but affected areas are small compared to the area of eroding farmland.

The nutrients in non-point source run-off come from animal waste, the soil itself (which contains nitrate from the nitrogen-fixing clover) and fertilisers which have been applied to the soil. On land of a given slope and soil type, the rate of nutrient run-off is affected by the density of vegetation cover and the density of animal populations. In the experimental catchments at Purukohukohu, where no nitrogenous fertiliser had been applied, pasture lost about 15 times more phosphorus (P) and about 3 and 10 times more nitrogen (N) than the native and pine forest catchments respectively (see Table 7.7). In general, nitrogen and phosphorus losses were similar from the two forested catchments.

Box 7.5: Tackling the point sources

In most New Zealand rivers, pollution from point sources has declined noticeably. This is partly because the total number of point sources has fallen, and partly because waste treatment processes have improved. For example, 200 or so dairy factories operated in 1970. Today they have been replaced by just 30 very large dairy plants whose waste disposal systems are more sophisticated (Barnett et al., 1994). One river that illustrates these trends is the 161-kilometre Waimakariri River which traverses Canterbury from the mountains to the sea. For more than a century the lower reaches and tributaries of the river have been used for point source waste disposal. Thirty years ago one tributary, the Otukaikino Creek, received wastes from two freezing works, a wool scour, and a soap factory. Another tributary, the Kaiapoi River, received point source discharges from two fellmongeries, a flour mill, a woollen mill, a freezing works, a milk powder factory and many septic tanks from Rangiora and Kaiapoi townships. Many small milking sheds also discharged waste into the river's tributaries. Since that time, most of the large discharges have ceased and the few that continue have greatly improved the quality of their effluent. From 1956 to 1994, the amount of organic matter in the river (BOD5) fell by an estimated 80 percent (Canterbury Regional Council, 1995) Many other regions can report similar trends, though the gains are sometimes masked by increasing pollution from non-point sources (e.g. dairy pastures).

The difficulties in controlling river pollution are highlighted by the Bay of Plenty's Tarawera River, sometimes nicknamed 'the black drain'. Despite a large improvement in point source discharges of toxic substances, the river's organic matter pollution has worsenedand not just because of farm run-off. The Tarawera is not a typical example of New Zealand's rivers. Besides receiving discharges from surrounding dairy farms, and stormwater and sewage from the town of Edgecumbe, the lower part of the river also receives effluent from the pulp and paper industry. The effluent comes from the Tasman and Caxton mills, their nearby geothermal power plant, and the mill town of Kawerau. The mills discharge more than 160 million litresof industrial waste every day, including organic matter, tannins, lignins, resin acids and organochlorines (e.g. dioxins). The geothermal plant adds sulphates and heavy metals to the mix. The water below the discharge points is significantly discoloured and gives off an odour typical of chemical pulp and paper mill discharges. Compared to its upper reaches, the lower part of the river has low concentrations of dissolved oxygen (often falling below the minimum guideline for aquatic life of 5 mg/m3), elevated temperature, more chemical and microbial contaminants, little or no submerged vegetation, a restricted range of fish and invertebrates, and a highly mobile pumice river bed seething with oxygen-sapping micro-organisms (Bay of Plenty Regional Council, 1995).

For four decades, the Tasman mill was subject to special legislation, the Tasman Pulp and Paper Enabling Act 1954, which gave it immunity from water and soil legislation, the Health Act, and any other laws prohibiting water pollution or nuisance effects from industrial waste. Prior to 1971, all the mill's discharges were untreated. Trout vanished from the lower Tarawera within months of the mill's opening. Through the next two decades other fish kills occurred from time to time. Tannins in the sewage discoloured the river, giving it the appearance of strong tea. Discharge limits were imposed in the 1960s, but were frequently exceeded. During the 1970s, the mill began to improve its discharge practices, principally by installing oxidation ponds. Between 1974 and 1981, the number of days on which excessive BOD5 levels were recorded fell from 44 percent to 13 percent. However, the river quality was still below its permitted "D" classification and was unsuitable for livestock, irrigation, horticulture, domestic consumption and recreation.

In the past decade, millions of dollars have been spent at the Tasman mill to reduce the discolouration and organochlorine discharges (Ogilvie, 1995b). The organochlorines come from the mill's pulp bleaching process which, unlike many other mills, is still chlorine-based rather than oxygen-based. The geothermal plant has also reduced its toxic discharges. As a result, water clarity has improved, and laboratory analyses of water samples can find no evidence of toxicity (Bay of Plenty Regional Council, 1995). However, set against these improvements is the fact that organic matter discharges from both the Caxton and Tasman mills have increased, particularly since Caxton opened a new treatment plant in 1992. Dissolved oxygen is now lower than it was in 1985. As a result, the river's ability to support balanced invertebrate populations has fallen in recent years, with the sensitive and important mayfly and caddisfly species declining both downstream and upstream of the discharges (Bay of Plenty Regional Council, 1995).

It will be some time before the Tarawera loses its 'black drain' tag, but progress has been made and more is expected now that the Tasman mill's special legislation has expired and all discharges into the river are under the jurisdiction of the Bay of Plenty Regional Council. The council has developed a management plan for the whole catchment which targets not only point source discharges but also dairy farm run-off and land drainage. These non-point sources have reduced dissolved oxygen in some of the region's streams and rivers to concentrations as much as five times lower than those in the lower Tarawera. So, even in the Tarawera catchment, where the battle against point source discharges is far from over, attention is shifting to the far larger problem of controlling non-point pollution sources.

Groundwater contamination

Although groundwater is vulnerable to contamination from many sources, including chemicals leaching from urban and industrial sites (e.g. landfills), most groundwater runs beneath agricultural land where it is at risk of contamination from nitrate-nitrogen, faecal matter, and pesticides. Nitrate-nitrogen is the most widespread of these contaminants. It enters soil from animal wastes (predominantly urine), nitrogen-fixing legumes (e.g. clover), and fertilisers, and seeps into groundwater more easily than other nutrients. Each year an estimated 30-70 kg per hectare can enter groundwater from dairy pastures (Burden, 1982). In some situations this may reach 100 kilograms of nitrate-N per hectare per year (Selvarajah, 1994). Most of this groundwater ends up in rivers, lakes, or estuaries.

Changes in Agricultural Pressures on our Water Environment

Relatively depressed economic conditions during the 1980s and early 1990s may have temporarily reduced some of the agricultural pressures on water while increasing others. Several studies have reported that farmers responded to the combined effects of economic restructuring, removal of subsidies, rising interest rates and declining sheep and beef export returns by cutting their expenditure on farm maintenance and development (Wilkinson, 1994; Wilson, 1994; Smith and Saunders, 1996).

Reduced maintenance meant less fertiliser to run off pasture and leach into water, but also a deterioration in hill pasture cover and a reduction in erosion control works. Between 1985 and 1988, phosphate fertiliser use fell by about 45 percent, but has since picked up. Of greater significance is the increased use of nitrogen fertiliser (see Figure 7.11). In recent years, this has soared to record levels as dairy farming has expanded into new areas, particularly in the South Island, and intensified in established areas, mostly in the North Island.

The growing cattle numbers themselves also increase the pressures on waterways from animal waste and stream bank damage. Because a cow's excreta has a BOD5 loading about 30 times greater than that of a sheep, the nation's total livestock BOD5 has actually risen in recent years, despite the fall in sheep numbers. The expansion of dairying is part of a broader change in the composition of our livestock populations. This change is reducing pressure in steep catchments but increasing it in lower lying areas.

Table 7.7: Nutrient losses in run-off to streams in the Purukohukohu experimental catchments.
  Nutrient losses in run-off from: (kilograms/hectare/year)
Nutrient indicators Pasture Pine forest Native forest
Total (Kjeldahl) nitrogen1 10.76 0.76 0.83
Nitrate nitrogen (NO3-N) 1.19 0.55 2.84
Total nitrogen2 11.95 1.31 3.67
Dissolved reactive phosphorus 0.37 0.04 0.02
Total phosphorus 1.67 0.1 0.12

Source: Cooper and Thomsen (1988)

1 Total Kjeldahl nitrogen (TKN): organic nitrogen and ammonia

2 Total nitrogen (TN): Total Kjeldahl nitrogen + nitrate nitrogen + nitrite nitrogen

As sheep numbers decline in steeper pasture lands native vegetation is regenerating in some areas and pine forests are being planted in others. This will have the long-term effect of reducing erosion rates, run-off, sedimentation, and flood flows. About 70,000 hectares of exotic forest are now being planted each year, 80 percent of it on marginally productive pasture land (e.g. hill country). If planting continues at this rate into the next century, the area of unforested agricultural land will decrease by almost 4 percent, from around 14 million hectares to about 13.5 million. At present, we have no estimate of the area of regenerating native vegetation.

In contrast to non-point source pollution, point source discharges from farms appear to have improved considerably in the past two decades, as have point source discharges from most industries. A big improvement occurred in the 1970s, as farmers began using either two-pond treatment systems or land irrigation to dispose of their waste water. Irrigation returns waste nutrients, such as nitrogen, phosphorus and potassium, to land rather than water. However, too much irrigation can cause nutrients to wash off into streams or leach into shallow groundwater so some local authorities have imposed limits on waste water irrigation.

In a two-pond system, the effluent is discharged into streams or rivers after passing through an anaerobic pond and then an aerobic, or facultative, pond. The system is designed to remove organic matter (BOD5) from waste water but it does not remove nutrients or reduce their impact on aquatic life (Hickey et al., 1989; Rutherford et al., 1992; Bolan, 1996).

Furthermore, even though it removes 95 percent of the BOD5 in dairy waste, the sheer volume of waste means that pond discharges of BOD5 often still exceed a stream's assimilative capacity (Barnett et al., 1994).

From 1972 to 1982 the percentage of farmers disposing untreated dairyshed waste into open drains declined from 50 percent to under 30 percent and the percentage using pond treatment or spray irrigation rose from 13 percent to 36 percent (McColl, 1982). More recent national figures are unavailable, but data from some regional councils show that the improvements continued into the mid-1990s.

In Taranaki, for example, only 250 (12 percent) of the 2,100 or so dairysheds had pond treatment systems in 1975, and only 39 of these were licensed (see Figure 7.12). However, since the early 1980s, the council has required on-farm treatment of dairyshed wastes and has issued licenses for those which meet the required discharge standards. By 1996, all of Taranaki's 2,593 dairysheds were licensed, with 60 percent disposing their waste to treatment ponds and 40 percent disposing it to land through irrigation (Taranaki Regional Council, 1996). About half the 6,000 or so dairy farms in the neighbouring Waikato region use treatment ponds, with most of the remainder using irrigation systems.

Figure 7.11: Phosphate and nitrogen fertilisers sold by major fertiliser works

The total amount of phosphate and nitrogen fertilisers sold dropped between 1985 and 1991, but has since increase to approximately 2 million tonnes in 1995. Since 1990 the proportion of super-phosphatic fertiliser sold has dropped in favour of other phosphatic and nitrogenous fertilisers (such as reactive rock, phosphate and urea) and in 1995 was over one quarter of the total fertiliser sales.

Source: Ministry of Agriculture (1996)

Urban Pressures

Although agriculture is New Zealand's dominant land use, most of us are town and city dwellers. In fact, 85 percent live in urban areas of 1,000 or more people. This concentration of about 3 million people and their pets, vehicles, homes, gardens, workplaces, schools, swimming pools, sports fields, and shopping centres into some 730,000 hectares, places heavy pressures on local waterways and aquifers. The populations tend to be concentrated in areas which are prone to water shortage (e.g. Auckland, Nelson, Hawke's Bay, Canterbury) and they also tend to be located on estuaries, harbours and river mouths most of which have been heavily modified by development, recreation and waste disposal. All of our large coastal cities, for example, are situated on estuaries (i.e. Invercargill, Dunedin, Christchurch, Nelson, Wellington, Lower Hutt, Napier, Tauranga, Auckland and Whangarei).

The main pressures which urban populations exert on water are: consumption for household, garden, and industrial use; pollution from sewage and stormwater (and also, in some cases, from leachate and run-off draining from landfills and other contaminated sites); and degradation of aquatic ecosystems through drainage, channelling, land reclamation, infilling, construction, roads, causeways, and shoreline developments around harbours, estuaries, wetlands and river mouths.

Consumption pressures

All urban centres have public water supplies. In areas prone to low rainfall, consumption pressures on those supplies can sometimes be very intense (e.g. Auckland, Napier, Hastings, Christchurch). Total urban and industrial water consumption is estimated at 470 Mm3 (or 470 billion litres) each year (see Table 7.4). Surface water provides 60 percent of this, and groundwater supplies 40 percent. Daily water use in most main urban areas in the late 1980s ranged from around 180 litres per person to over 900 litres per person (see Table 7.8).

Figure 7.12: Improving dairyshed discharges in Taranaki

In 1975 there were around 2100 dairy sheds in Taranaki with around 10% of them having ponds or other licensed treatment systems. In 1989 there were around 2700 dairy sheds with over 80% of them having treatment systems. By 1994 all dairy sheds had treatment systems.

Source: Taranaki Regional Council (1996)

Water use peaks in summer and is mostly for domestic purposes, though commercial and industrial uses are also significant, as are reticulation losses from taps, pipes, tanks and reservoirs (see Figure 7.13). In drier areas, such as Renwick (Marlborough), maximum daily demand can reach 2,000 litres per person during the summer. The increased seasonal demand for water puts pressure on surface water resources, which are usually at their lowest levels in summer. It may also stretch the ability of the supply system itself to meet demand. In many places, restrictions have to be placed on the use of water in gardens during the summer.

Figure 7.13: Main uses of public water supplies in the Auckland urban area
Main uses of public water supplies in the Auckland urban area:
  • Domestic use 51%
  • Industrial use 18%
  • Commercial and recreational use 16%
  • Unaccounted (for example, fire fighting, leakage) 15%

Source: Watercare Services (Auckland) (1994)

Until recently, demand for water was increasing steadily throughout New Zealand. From 1970 to 1990 the amount used by Wellingtonians and Aucklanders increased by 25 percent and 32 percent respectively (see Figure 7.14). However, since the Auckland water crisis of 1993, and the adoption of water conservation strategies, Auckland's water use has reverted to early 1980s levels. The average Aucklander now uses around 300 litres per day, 21 percent down on the 1988 figure of 380 litres per day.

Figure 7.14: Water use in the Auckland and Wellington regions, 1973 to 1995

Auckland region's water consumption has been around double the consumption in the Wellington region for the period 1973 to 1995. Consumption in both areas peaked in the late 1980s, and in 1995 was around 110 million cubic metres annually in Auckland and 55 million cubic metres annually in Wellington.

Source: Watercare Services (Auckland); Wellington Regional Council

Urban pressures on water quality

The most severe urban pressures on water quality in streams, rivers and coastal waters are from the point source discharge of sewage and the non-point source discharge of stormwater run-off. According to the water quality experts surveyed by Sinner (1992), the impacts on surface water from these pressures rank second and third in severity after agriculture. Most sewage is now piped through treatment installations to remove or reduce pollutants before discharge. Treatment systems vary, however, and some are also vulnerable to flooding and stormwater infiltration. Stormwater quality is determined by the road debris and other contaminants it picks up when flowing from land to water. Pollution from motor vehicles through oil leakage and dust contamination has a significant impact on urban stormwater quality, particularly in Auckland which has large volumes of road traffic passing by estuarine waters (Auckland Regional Council, 1995).

Table 7.8: Daily water use in some cities and districts.
Locality Average daily water use (litres per person)
North Island Auckland 380
Taupo 610
Gisborne 180
Opotiki 500
Kapiti Coast 670
Wellington 550
South Island Nelson 190
Renwick 760
Greymouth (metered) 400
Greymouth (unmetered) 600
Christchurch 450
Dunedin 500
Invercargill 360

Mosley (1993); Watercare Services (Auckland); Wellington City Council

The quality of urban drinking water is a special case. Where possible, public water supplies are drawn from rivers, lakes, built reservoirs or deep aquifers that are not exposed to agricultural and urban waste. To further reduce the possibility of contamination, supplies are hygienically treated to remove micro-organisms, potentially harmful chemicals and sediment. Variations in drinking water quality are more often caused by variations in treatment method than by pressures from the surrounding environment.

Sewage point source pressures

Households and workplaces discharge a vast amount of human excreta, detergents and other substances into the nation's sewers. All towns with populations of 5,000 or more have a reticulated sewerage system which pipes everybody's waste water to a common discharge point. The vast majority of this sewage ends up in rivers or the sea, with only a small percentage being disposed of on land.

Human waste has a BOD5 loading of about 74 grams per person per day (Vanderholme, 1984; Hauber, 1995). This means that in 1950, when the New Zealand population numbered 1.9 million people, the bacteria working to decompose their waste would have removed 140 tonnes of dissolved oxygen from the water each day. By 1996, with a population of 3.6 million, this daily BOD5 loading from human wastes has nearly doubled to about 266 tonnes. Fortunately, these days most of this decomposition occurs at sewage treatment plants rather than in natural waterways. In 1950, New Zealand had only about five sewage treatment plants. Today we have more than 220, and about 80 percent of our households are connected to them. This means the total BOD5 loading to waterways from human waste has probably declined in 45 years to less than 50 tonnes per day, though precise figures are not available.

The sewage pollutants of greatest concern are rotting organic matter (measured as BOD5), disease-causing micro-organisms (measured by counting faecal coliform bacteria), excess nutrients (particularly nitrogen and phosphorus) and suspended solids. These can be harmful to aquatic ecosystems and to humans who accidentally eat contaminated shellfish, or, in badly polluted areas, swallow the water. Sometimes, too, ammonia from urine may reach concentrations that are toxic to fish (Hoare and Rowe, 1992).

Sewage treatment systems can be classified as primary, secondary or tertiary according to their ability to remove progressively smaller pollutants. Primary treatment removes suspended solids by filtering or milliscreening them or allowing them to precipitate to the bottom of settling ponds. Secondary treatment reduces the organic matter and nutrients in primary-treated effluent by allowing micro-organisms to consume and decompose them in oxidation ponds or wetlands. Tertiary treatment uses various methods, including chemical treatment and fine filtering, to reduce the nutrients and micro-organisms in secondary treated effluent.

In general, New Zealand's sewage treatment systems appear to be very good at reducing the organic matter, suspended solids and faecal coliform bacteria in sewage, but are less successful at removing the nutrients. A 1992 survey of 17 sewage plants revealed that BOD5 was reduced by a factor of 91 percent on average (though the actual decreases ranged from 30 percent to 99 percent, depending on the system), suspended solids by 86 percent (ranging from 53 percent to 99 percent), and faecal coliforms by 98 percent (ranging from 89 percent to 100 percent) (Hauber, 1995). However, the average reduction in nitrogen was only 31 percent (ranging from an actual increase of 12 percent to a maximum decrease of 77 percent), and that of phosphorus was just 24 percent (ranging from a 9 percent increase to a 57 percent decrease). As a result, sewage is a significant source of nutrient inputs to many harbours. In Whangarei, for example, which has relatively few dairy farms in the harbour catchment, sewage is the main nitrogen source (see Figure 7.15).

Some of the older sewerage systems are prone to developing cracks in pipes and joins, often caused by tree roots or earth tremors. This can lead to sewage escaping or to stormwater infiltrating the sewerage system, causing it to flood. Local authorities are steadily repairing and upgrading these older systems, but the cost of doing so makes the process a gradual one.

National data on our sewerage systems are only now beginning to be compiled after a gap of over a decade (Woodward-Clyde, 1996). The last national survey was 15 years ago (Ministry of Works and Development, 1981). Up until then local authorities had been surveyed every five years for information on their sewerage systems. The surveys came to end when the Ministry of Works and Development was dissolved in the government restructuring of the mid-1980s.

The 1976 and 1981 surveys showed that just over 60 percent of the population were connected to sewerage treatment plants. Around 17 percent of the population had their sewage discharged untreated, mostly into the sea, and around 20 percent were not connected to a sewerage system at all, but relied on septic tanks (Ferrier and Marks, 1982). In the intervening decade, the percentage connected to treatment plants is believed to have risen to about 80 percent, while those discharging untreated sewage are just a few percent. Some 15-20 percent of people probably still use septic tanks.

Septic tanks are used in small towns, rural communities and beach settlements. They are an efficient form of sewage treatment but when they malfunction contaminants can leach into nearby waterways (Higgins, 1991). This can cause problems in enclosed groundwaters and shallow ponds and lakes, but is not a problem in aquifers with a large throughput of groundwater. For example, it has been estimated that septic tank discharge into the gravels west of Christchurch would be rapidly diluted one million-fold by the time it had flowed 5 metres below the water table and 20 metres down-gradient (Sinton, 1982).

To get some information on the changes that were occurring in sewage treatment, the magazine Terra Nova in 1991 commissioned a survey of local authorities serving populations greater than 20,000 (Shields, 1991). Of the 46 councils approached, 42 (90 percent) replied. In total, they gave information on 74 sewerage systems which process the wastes of more than 2.5 million people (three quarters of the total population). The systems varied from 'state-of-the-art' facilities (e.g. Whangarei and Rotorua) to very basic disposal of raw sewage through ocean outfalls (e.g. Wellington and Napier).

Figure 7.15: Sources of nitrogen inputs into the Whangarei Harbour

Main sources of nitrogen inputs into the Whangarei Harbour:

  • Sewage 67.3%
  • Rural run-off and point sources 20.1%
  • Urban storm water 12%
  • Air 0.6%

Source: Williamson (1991)

Seven (9 percent) of the sewerage systems provided tertiary treatment. Half provided secondary treatment and 30 (41 percent) provided primary treatment. A number of authorities indicated that they were upgrading their systems. The authorities were also asked about the final destination of the treated effluent. Of the 65 sewerage systems for which this information was given, 18 discharged into the sea, 27 discharged into rivers or, in one case, a lake, 18 discharged into estuaries or harbours, and 10 discharged to land (see Figure 7.16). Land disposal was used by four of the seven tertiary treatment systems but by very few of the primary and secondary treatment systems (Shields, 1991).

Although this survey covered sewerage systems that serve three quarters of the population, it missed the majority of the nation's sewerage systems. Two thirds of our treatment plants serve communities with populations of less than 20,000. Two recent surveys have attempted to get a more comprehensive view of sewage treatment. Greenpeace New Zealand (1996) received 40 replies to its survey of 75 local authorities (a 53 percent response rate). A total of 69 treatment plants were described but, because the response rate was low and the survey was not restricted to communities above a certain size, interpretation of the results is difficult.

Figure 7.16: Sewage treatment and disposal in communities of more than 20,000

For primary treatment sewage systems, the majority discharge into freshwater followed by discharge into the sea or unspecified destinations. A small number discharge onto land and a few into harbours.

For secondary treatment sewage systems, the majority discharge into freshwater followed by discharge into the sea or harbours. A small number discharge onto land.

Tertiary treatment sewage systems have the smallest total number and mostly discharge onto land followed by discharge into harbours and freshwater.

A much more successful result was obtained from a survey of 74 local authorities which was commissioned by the New Zealand Wastewater Association with funding from the Ministry for the Environment's Sustainable Management Fund (Woodward-Clyde, 1996). Responses were obtained from 57 of these (77 percent), half of whom also replied to a second, more detailed, questionnaire, providing information on 242 of the nation's 258 sewerage systems. The database is still being developed and, at the time this report was being prepared, only limited analyses had been conducted on it. The unpublished results showed that many small communities discharge their sewage to rivers and lakes (64 percent compared to only 37 percent of the larger communities in the Terra Nova sample) and to land (28 percent compared to 14 percent), while very few discharge to sea (5 percent compared 25 percent) and harbours and estuaries (3 percent compared to 25 percent). This appears to reflect the New Zealand settlement pattern in which most of the larger communities are located near the coast and most of the smaller ones are located near rivers or lakes.

Apart from household and community sewage, a considerable amount of wastewater is discharged by factories and power plants. Some of these discharge into the community sewerage system while others discharge independently into rivers or coastal waters.

Industrial discharges into the sewerage system account for an estimated 9 percent of sewage wastewater, though it can be as high as 25 percent in some cities and nil in others (Hauber, 1995). Most large companies discharge their wastes separately under a permit from the local council.

Among the largest industrial dischargers are the Tasman Pulp and Paper Mill, which sends 160 million litres of treated wastewater into the Tarawera River each day (see Box 7.5), and the Kinleith Pulp and Paper Mill which discharges over 110 million litres per day into a tributary of the Waikato River. By comparison, the combined daily discharges from Wellington and Lower Hutt come to around 100 million litres, those from Christchurch total 140 million litres, and those from the Mangere sewerage plant, which processes the wastes of 70 percent of the Auckland population, come to nearly 230 million litres.

Most community and industrial discharges end up in the sea. In 1976 and 1981 60 percent of the population had their sewage discharged into estuaries, harbours or the ocean. The more recent surveys do not give population estimates, but it is likely that the figure has not changed markedly. A 1982 survey identified at least 52 coastal discharges, a third of them industrial (Abbott and Leggat, 1982), while a review three years later listed 32 'major' outfalls (i.e. discharging more than 1,000 litres per second) of which only seven were industrial (Williams, 1985).

No recent statistics on coastal outfalls have been compiled. Outfall pipes are a visible, and often controversial, means of disposing of wastewater, particularly where they discharge on the shoreline, near swimming or shellfish gathering areas, or in semi-enclosed estuarine waters (Abbott and Leggat, 1982; Smith et al., 1985). In estuaries and harbours with restricted water circulation, outfalls have caused pollution or deoxygenation problems (e.g. Avon-Heathcote, Manukau, Waimea Inlet and Nelson Haven, upper Whangarei Harbour and the Otamatea River estuary). The Moa Point and Pencarrow outfalls at Wellington Harbour have also caused problems and rendered the harbour's shellfish inedible.

Most of these problems are now being addressed through better effluent treatment or new disposal systems. For carefully-sited, long, deep, outfalls, however, most research indicates that the sea's dispersal powers quickly dilute any contaminants (Smith et al., 1985). The impacts are greatest in the immediate vicinity of the outfall and become undetectable within half a kilometre.

Incidents of severe river or coastal pollution from sewage have declined markedly in recent decades as treatment systems in most areas have improved, and more stringent environmental requirements have caused local authorities to look towards land disposal systems. An informal survey of regional councils in 1992 found that most water management officials considered sewerage systems and other point source discharges within their regions to be under adequate control (Hoare and Rowe, 1992). Although point source pollution of the lower reaches of some rivers was considered a problem, this was more often attributable to accidents or uncontrolled discharges rather than inadequate treatment systems.

Most regional water managers considered non-point source pollution from grazing animals as the number one cause of water quality problems, though urban stormwater was considered a main cause in some urban locations (Hoare and Rowe, 1992). Malfunctioning septic tanks can also cause non-point source pollution in some areas near enclosed waters, such as lakes (Higgins, 1991).

Urban stormwater

Because urban areas are largely paved, rainwater cannot be absorbed by soil and vegetation. To reduce the risk of flooding during storms, most towns and cities have stormwater systems which channel the rainwater into gutters and drainage pipes, eventually discharging it through outfall pipes into streams, lakes and coastal waters. This can be a significant source of pollution. In fact, urban stormwater is often similar in quality to secondary treated sewage (Jessen and Hawley, 1981; Melville, 1991; Williamson, 1991; Hoare and Rowe, 1992; Auckland Regional Council, 1995).

Pollution comes from the substances that are washed off the street and adjacent surfaces, and also from the accidental mixing of stormwater and sewage. Many sewerage systems are prone to stormwater invasion either through faulty pipework, illegal connections, or undetected past connections. Stormwater in a sewer can quadruple the volume of effluent, placing a heavy burden on treatment and disposal facilities, and reducing the effectiveness of treatment. If sewerage systems are flooded, untreated effluent can escape. In Auckland, stormwater can amount to 75 percent of the total quantity of sewage effluent and is a major environmental engineering problem for local authorities.

Contaminants that are washed off streets, construction and industrial sites, and other surfaces, include sediment, organic matter, nutrients, disease-causing organisms, and toxic substances ranging from oil products and contaminated dust from vehicle exhausts, to industrial chemicals (Williamson, 1991; Auckland Regional Council, 1995). Any of these may threaten water quality and aquatic life, or endanger the health of swimmers, other recreational water users and seafood gatherers and eaters. Stormwater is also a major source of marine debris, such as floating plastic, which is both unsightly and hazardous to marine mammals and birds (Island Care New Zealand Trust, 1995).

Stormwater that runs off construction sites can carry very high levels of sediment, particularly where the vegetation and topsoil are stripped beforehand. Sediment yields from small catchments that are undergoing construction throughout 100 percent of their area may be as high as in the most intensely eroding country of the East Cape (Williamson, 1991). A study of the Pauatahanui inlet near Wellington, for example, found that almost a tonne of sediment per hectare per month ran into the estuary from a catchment undergoing urban development (Curry, 1982). This was fifteen times greater than the sediment loss from a neighbouring non-urbanised catchment. When a flood struck, the sediment loss was seventy times greater in the urbanised catchment.

Pressures from dams

Dams alter the flows of rivers and streams, and create barriers to fish movement. Large dams have flooded valleys, raised lake levels, and reduced the flows of some major rivers to residual trickles. Even small dams can be insurmountable obstacles to our native fish, many of which need to migrate to estuaries or the ocean in order to breed. New Zealand has thousands of dams, most of which are small water-supply dams on farms. However, more than 400 dams have storage capacities greater than 18,500 m3 (18.5 million litres). They range in height from 1.8 metres to the imposing 118 metres of Benmore dam in the Waitaki headwaters. Some of these large dams were built to store water for irrigation, others for power generation, and others for domestic and industrial supply or floodwater control (Freestone, 1992).

The largest dams are for power generation. In fact, 98 percent percent of the water which is harnessed for human use in New Zealand is used to generate electricity (see Table 7.4). The controlled release of large torrents of water from dams provides the raw energy to spin the electricity-generating turbines at some 80 power stations around the country. These stations have power generating capacities ranging from 1,000 Megawatts (MW) at Clyde, to less than 1 MW at small stations operated by local electricity supply authorities.

The impact of a dam on river flows varies with the size of both the dam and the river. Most hydroelectric dams in New Zealand are 'run of the river' schemes, with enough storage for only a few hours or days of generation. This means they do not modify the seasonal flow patterns but may cause large day-night fluctuations in response to varying power demands. For example, daily flows in the Clutha River vary from 200 to 600 cumecs as the Roxburgh power station responds to changing demand for power. Such disturbance of the flow regime can affect river channel stability and reduce fish habitat and spawning areas.

Large dams have dramatically altered the flows of some of New Zealand's rivers and lakes. They include Otago's Clutha, Waitaki and Waipori rivers, and the North Island's Waikato River. Other rivers have been affected by diversions of water, such as Southland's Waiau River whose flow was halved when water was diverted to the Manapouri power station, and the Whanganui River in the central North Island. In the South Island's Waitaki River catchment, the diversion of flows into a system of canals and power stations reduced the flows in the Tekapo, Pukaki and Ohau rivers to 'residual' levels (see Figure 7.17).

Figure 7.17: Effects of the Upper Waitaki power scheme on river flows in the Mackenzie Basin.

Power stations are located at Tekapo (Tekapo A) and Lake Pukaki (Tekapo B), between Lake Ohau and Lake Benmore (Ohau A, B and C) and at Benmore.

There are dams at the south ends of Lake Tekapo and Lake Pukaki (Pikaki High Dam), between Ohau A and B, and at Benmore.

Canals run between Tekapo A and Tekapo B, from the Pukaki High Dam to Lake Ohau, and from the Ohau dam through Ohau B and C to Lake Benmore.

Residual flows run from Tekapo and Pukaki to Lake Benmore.

Source: Cooper and Thomsen (1988)

Some natural lakes have had their levels artificially altered for electricity generation. They include Lakes Tekapo, Pukaki and Ohau in the Waitaki headwaters, Lake Hawea in the Clutha headwaters (which does not generate power directly but feeds into Lake Dunstan) and Lake Manapouri. The level of Lake Taupo is also controlled to help manage generation from the Waikato River. In some cases, natural features have been lost or degraded by flooding from the construction of artificial lakes (e.g. the Aratiatia Rapids and Orakeikorako Geothermal Field on the Waikato River, and the Cromwell Gorge on the Clutha River).

Pressures from forestry

Forestry can exert pressure on water flows and water quality, though the area affected is relatively small and the negative effects are generally short-term. Tree felling and road or track construction can degrade stream and river quality by increasing the amount of sediment and nitrate-nitrogen (NO3-N) that washes into water (O'Loughlin, 1994). Other nutrient concentrations in the water may also increase, such as potassium (K), magnesium (Mg), calcium (Ca), ammoniacal nitrogen (NH4-N) and total phosphorus (TP) (Rowe and Fahey, 1991). These concentrations are reduced if narrow riparian (riverbank) strips of forest are retained, and they decline rapidly in the second year after replanting.

Although the scale of the hydrological impact depends on the area harvested, logging in moderate-to-high rainfall areas can cause a 60-80 percent increase in water run-off for three to five years after clearfelling and can raise flood peaks by 50 percent (Fahey, 1994). At present, the overall impact of forestry on water is likely to be positive. While about 20,000 hectares are harvested and replanted each year, a further 70,000 hectares of unforested land, much of it pasture, are converted to plantation forests. As discussed earlier, stream flows and sedimentation generally return to pre-harvesting levels within six to eight years of replanting, and pasture run-off generally falls by 30 percent to 50 percent within five to ten years of afforestation.

A more indirect impact of forestry is the effect of roading and vehicles on sedimentation and run-off into streams, and the impact of timber treatment chemicals at some contaminated sites where toxic substances have leached into streams or groundwater (see Chapter 8 for discussion of contaminated sites).

Pressures from mining

Mining can affect water in several ways. It can alter the flow pattern or volume of a stream by diverting or abstracting water, it can reduce the water table because of the need to remove groundwater from the mine shafts, it can change the amount of water, silt and sediment entering streams, and it can also increase their concentrations of processing chemicals (e.g. cyanide) and of heavy metals from ground rock. However, compared to the large impacts of gold and coal mining last century, the impacts of modern mining are better controlled and are confined to fewer locations (Glasby, 1991; Barker and Hurley, 1993).

For example, in the 1970s, the open-cast coal mines at Huntly discharged huge but unmeasured loads of silt to Lake Waahi and the Waikato River. These days each mine has a properly designed and maintained treatment system and the treated waste water is used for things such as spray irrigation of rehabilitated areas (Hoare and Rowe, 1992; Barker and Hurley, 1993). Every day, more than 6,000 cubic metres of waste water are discharged from the Huntly East underground mine into a nearby stream and wetland. Before being discharged, the water is held in settling ponds to allow silt and clay to settle out so that the water quality is within acceptable limits.

Gold mining, by its nature, tends to occur in areas where the rocks have a naturally high metal content (e.g. the Coromandel peninsula). When the rocks are exposed by other natural processes or land disturbance to oxygen and water, the metals within them can react chemically. The dissolved metals can then enter nearby streams and groundwater, causing them to become more acidic and to have elevated concentrations of metals. Mineral processing can speed up this process by exposing large mounds of finely ground waste rock to air and water. The mining industry is well aware of this risk and has developed methods of dealing with it.

The standard process is to pump the finely ground rock (or tailings) as a slurry to a tailings dam where it settles and compacts. Contaminated water from the slurry collects in the tailings dam and in the drainage system beneath it and is then pumped to an on-site treatment plant where most of the heavy metals and other contaminants are removed. The clean water is recycled through the mine and any excess is discharged to nearby rivers. The stream water quality is virtually unchanged by the mining discharges.

However, failures are not impossible. The large tailings dam at the Golden Cross mine in the Coromandel area is built on land which, while stable on the surface, is moving beneath the surface. The dam's location was approved before the passing of the Resource Management Act. Investigations into the possible relocation of the dam are currently being undertaken. Although attempts are being made to stabilise the dam, it is not known how successful these will be in the long term. For example, there is a possibility that the drains beneath it could be breached by the earth movements and allow contaminated water to escape. Fortunately, the location problems at the Golden Cross mine appear to be the exception rather than the rule.

Wastewater discharged from the nearby open-cast Martha mine is actually cleaner than required and the mine's tailings piles are being restored to pasture (Barker and Hurley, 1993; Morrell et al., 1995). The mine will be exhausted in a few more years and the whole site rehabilitated. Measures have been taken to minimise any long-term risks of leachate escaping from the tailings beneath the restored pasture (Mathias, 1991; Morrell et al., 1995).

Arguably, the worst case of ongoing environmental damage caused by past mining is at another Coromandel site, the former Tui mine near Te Aroha where lead, copper and zinc ores were exploited and processed on site (Carter, 1982; Morrell et al., 1995). Four years after the mine closed, water feeding from the Tui stream into the town's water supply was found to be contaminated with heavy metals leaching from the tailings heap. The stream was disconnected from the water supply and attempts were made to stabilise the tailings pile. Today, the site's 4-hectare barren dam contains over 100,000 cubic metres of very acidic, sulphide-rich, tailings whose drainage has severely contaminated local streams. No natural plant recolonisation has occurred at the site for more than 20 years and methods for revegetating it are being investigated.

Another mining impact on water is caused by the extraction of aggregate (e.g. sand and gravel) from rivers, beaches and the sea bed. Aggregate provides the raw material for constructing roads and buildings. Aggregate mining can change the natural character of beaches. Although this is not yet a cause for concern, it has the potential to cause problems in areas where there is a high demand for aggregate and no large river beds to supply it (e.g. Auckland's east coast).

Pressures from Pests and Weeds

Alien plants, animals and micro-organisms are an increasing source of pressure on water ecosystems. Among the better established freshwater invaders are the so-called oxygen weeds, Egeria, Elodea and Lagarosiphon, which can almost completely fill a pond, lake or stream, growing up to 6 metres, and crowding out native aquatic plants (Vant, 1987). In some cases, the extent of their growth has been extreme by world standards (Taylor, 1971). A fourth oxygen weed, Hydrilla verticillata, which may be the worst yet, has recently become established in four Hawke's Bay lakes (Champion and Clayton, 1995). Some introduced aquatic plants, such as water hyacinth (Eichhornia crassipes) and Salvinia molesta, are so invasive that they are now regarded as noxious aquatic weeds. Introduced weeds have established themselves in our largest lakes, including Lake Taupo, and our smallest wetlands, such as Pukepuke Lagoon in the Manawatu (see Box 7.8). In many cases, the aquatic plant invasions have been made easier by the clearance of riverbank forests (Howard-Williams et al., 1987).

Boats and water fowl frequently spread weeds from one water body to another. The recent algal invader, water net (Hydrodictyon reticulatum), was first reported in an ornamental pond in Tauranga in 1986, having apparently arrived here with fish or aquatic plants imported by a local hatchery (Hall, 1994). It spread throughout the eastern Bay of Plenty and Rotorua Lakes and became widespread in the Waikato River system. The floating mats formed by this lattice-like green algae spoiled swimming, boating and fishing, and smothered the habitat of desirable aquatic plants, invertebrates, fish, and water fowl. Its decay caused odour problems and oxygen depletion, and the mats clogged pump filters, stock ponds and drainage ditches (Hall, 1994). Fortunately, this particular invader did not come to stay. In 1995, after four years of grappling with the weed, Bay of Plenty Regional Council officials were reported to be baffled by its mysterious decline throughout the region (Ogilvie, 1995a). Spraying and suction dredge trials were cancelled and, a year later, the invader was deemed 'extinct'.

Several introduced fish are also listed as noxious species in the Third Schedule to the Freshwater Fisheries Regulations 1983, including rudd (Scardinius erythrophthalmus) and European or koi carp (Cyprinus carpio) (McDowall, 1990). Catfish ( Amerinus nebulosus) is not designated as a noxious fish, but is regarded as a nuisance. The most widespread fish invaders, however, are probably brown trout (Salmo trutta) and rainbow trout (Oncorhynchus mykiss), which were introduced, with salmon (Salmo spp., Oncorhynchus spp.), as sports fish and are still protected as such because they provide a significant recreational resource. Trout have displaced or become predators of some of our native galaxiid fishes (see Chapter 9, Box 9.9). Another significant problem is the illegal distribution of fish by fishers.

Not all water pests are invaders. Some disease-causing micro-organisms, such as Giardia, Cryptosporidium, Campylobacteria and Entamoeba have probably always been part of the freshwater ecosystem and are only pests to people who swallow them in untreated water. Public water supplies are treated to remove these organisms. Although the quality of treatment varies, few cases of water-borne disease have been traced to public supplies.

Coastal waters may be more susceptible than land-based waterways to unnoticed invasions. One that has been noticed is Colpomenia durvillaei, a brown sac-like seaweed that was first seen in 1980 at Leigh and has since been found in Hawke Bay and in Wellington Harbour where it appears to be a very successful coloniser of space, potentially competing with native seaweeds (Nelson, 1994). A widespread animal import is the Asian date mussel, Musculita senhousia, which was first recognised here around 1980 and is apparently still spreading, forming mats on the sea floor over hundreds of square metres which smother other species.

The impacts of invaders cannot always be predicted. A seaweed called Codium fragile tomentosoides is known as the 'oyster thief' because of the damage it inflicted on North American oyster fisheries. It was first noticed here in 1975 but, 20 years on, has not become a serious weed problem. A widespread animal invader, the Pacific oyster (Crassostrea gigas), arrived in the Auckland area in the early 1970s and reached Cook Strait by 1980. It is now the basis of New Zealand's oyster farming industry (Nelson, 1994).

Most of the successful invaders come from cool water environments in the northern hemisphere. Warm-water invaders sometimes arrive here but are less successful in establishing themselves. In warm La Niña years (i.e. 1970-71, 1973-75, 1988-89) the East Australian Current, which bathes the northern part of the North Island, brings warm-water algae, protozoans and animals to north-eastern New Zealand, depositing them in rocky habitats where they usually die out when temperature and other conditions return to normal (Nelson, 1994).

In contrast to these natural invaders, the cool-water species often arrive from the Northern Hemisphere in human-made structures, particularly the ballast water of foreign ships. This phenomenon is not as well monitored in New Zealand as in Australia where scientists have identified with reasonable certainty at least 14 exotic species of fish, crustaceans, molluscs, worms, seaweeds and toxic algae that have arrived recently in ballast water (Jones, 1991). Ballast prevents ships becoming top-heavy when their cargo holds are light. Ships began using sea water for ballast, in place of sand and boulders, around 1880. Sea water ballast is known to contain many marine organisms, spores and eggs (Nelson, 1994).

Whaling ship ballast water last century is believed to have brought the red tidal seaweed, Chondria harveyana, from Tasmania to Porirua, and the brown algae, Chnoospora minima , from tropical waters to Port Underwood (Nelson, 1994). These species have not spread, but a more recent ballast water invader is spreading fastthe prolific Japanese seaweed, Undaria pinnatifida. This very large brown kelp which grows up to 3 metres in length was first discovered in Wellington Harbour in 1987 and has reached Timaru, Oamaru, Port Chalmers, Lyttelton and Picton. It is an important food species in Asia where it is grown in vast marine farms. Its impact in New Zealand is not known at this point, though its aggressive competition for space may crowd out some native seaweeds and the organisms that depend on them for food and shelter (Nelson, 1994). Ballast water was also a suspected, though unconfirmed, source of at least some of the toxic algae involved in recent shellfish poisoning episodes (see Box 7.12).

Marine invaders are known to 'hitch-hike' on natural rafts, such as large floating seaweeds, and on human-made rafts, such as cut logs, glass floats, fishing gear, plastic objects (e.g. bottles), ship hulls, oil rigs and other large structures such as the Maui gas platforms which were towed from Japan in 1975 and the clip-on pieces of the Auckland Harbour Bridge (Nelson, 1994). One of the tiniest invaders, a Herpes virus which kills pilchard fish by damaging their gills, swept through our coastal waters in 1995. It appears to have arrived in dead pilchards which were imported as longline fishbait from Australia. The same virus had already killed large numbers of Australian pilchards and may have reached Australia in fishbait imported from the Americas. The epidemic appears to have been short-lived.

Pressures from Climate Change

The water cycle is driven by temperature. Temperature determines the rate at which water evaporates from land and sea, the moisture carrying capacity of the air, the melt rate of polar ice and alpine snow, and the air pressure differences that give rise to the world's winds. Climate scientists therefore expect significant changes in climate patterns if greenhouse gas pollution continues to force up global temperatures (see Chapter 5).

Such temperature-induced climate changes could alter New Zealand's rainfall patterns and wind currents, speed up snow and glacial melt rates and water evaporation rates, and affect river flows, water temperatures and sea levels. Although Mosley (1988) found no apparent impact of rising temperatures on New Zealand river flows, possible future impacts include decreasing winter rainfall, and increasing summer rainfall in western and northern parts of the country (Ministry for the Environment, 1994). Increased coastal erosion may also result from rising sea levels (Hicks, 1990).

Despite the uncertainty over the likely impacts of global climate change on New Zealand's water, there is no doubt that the intermittent climate changes associated with the El Niño phenomenon do affect water supplies in different parts of the country (see Box 7.2). Because it is a recurrent natural phenomenon, El Niño-related climate change is generally considered as something quite separate from long-term global climate change. However, some scientists are now suggesting that the recent increase in El Niño events may itself be a symptom of global warming (Kerr, 1994; Wuethrich, 1995).