View all publications

The state of our land environment

Soil and vegetation cover have been heavily modified by the pressures imposed by human society. While most land users are sustainably managing their soils, and many are now trying to maintain what remains of their native vegetation cover, problems of one sort or another exist in many areas.

Signs of declining soil productivity abound in many parts of the South Island high country and the North Island hill country, as well as in some intensively cropped lowlying areas. Contaminated sites are a problem in some urban and industrial areas. And, in many parts of the country, the native vegetation has become so fragmented that active restoration may be necessary for any protection to be effective.

The state of our soils

Although most problem areas are known to local authorities, soil scientists, and the affected land owners, their overall scale and distribution have not been surveyed. Surveys formerly carried out by government agencies such as the DSIR's Soil Bureau and the Ministry of Works and Development's Water and Soil Division ended when those organisations were disbanded in the mid-1980s (Clough and Hicks, 1992).

The Crown Research Institutes (e.g. Landcare) inherited the old maps and data, but have only now been able to begin updating them. In the absence of recent survey data, the incidence of soil problems must be inferred from the earlier surveys, local examples and case studies (see Table 8.8).

Soil degradation takes many forms. Erosion, or the physical removal of soil, is probably the most serious and the least reversible, but soils are also degraded by nutrient depletion and acidification, which causes them to lose their fertility, by the loss of organic matter (i.e. carbon depletion) and compaction, which causes them to lose their structure, and by contamination with toxic substances which can make them a health risk to humans and other species. Salinisation, which is a problem in some other countries, has not emerged as a problem here.

Some soil degradation is often inevitable with land use, but generally it can be minimised or even reversed by sustainable land management practices (see Box 8.4). The extent of such practices is not known nationally, but surveys in the North Island between 1989 and 1992 indicated that about 80 percent of hill country farmland needed erosion control measures and that measures existed on about twothirds of it (Clough and Hicks, 1992). The same surveys, however, showed that most of the measures are being ineffectively applied and that some 55 percent of the erodible North Island hill country farmland is not receiving adequate treatment.

Box 8.4: Soil conservation techniques

Many of the techniques for reducing or reversing soil degradation are well known. Others have been developed more recently. Nutrient decline can be addressed by applying fertilisers and, in the case of nitrogen deficiency, planting legumes. Acidification can be reduced by applying lime. Various methods exist for combating erosion and soil breakdown. Erosion on cropland can be reduced by minimum tillage, mulching crop residues or harvest stubble rather than burning it, contour ploughing (i.e. going across slopes rather than up and down them), sowing grass in rills and other depressions to slow the flow of rainwater, and planting trees for windbreaks (Hicks, 1995).

On pasture land, the erosion control techniques include maintaining adequate vegetation cover (i.e. avoiding overgrazing and maintaining a dense grass mat through regular applications of fertiliser and grass seed), spaced tree planting on slopes (at distances of up to 12 metres), close tree planting (at densities of 300 stems per hectare or more), reversion (i.e. allowing natural forest to regenerate on erodible land), fencing off river bank grazing strips, planting trees on unfenced banks, controlling grazing by moving stock before pasture becomes depleted, and building debris dams to slow water flows in gullies (Hicks, 1995).

Similarly, techniques to limit compaction include minimum tillage, shallow or variable depth ploughing, retaining crop residues and adding organic matter to the soil, avoiding concentrated pressure on wet soil from machinery or stock, providing drainage systems for susceptible paddocks, avoiding fallow periods where land is left bare for one or more seasons, rotating crops with pasture, and subsoil ploughing to break up already compacted soil (Haynes, 1994).

Erosion

On a human time scale and often on a geological one, soil can be considered a nonrenewable resource because rates of soil formation are so slow. It takes one to four centuries for nature to build a single centimetre of topsoil, and 3,000 to 12,000 years to develop sufficient soil to form productive land (Dailly, 1995). These rates apply to most soil types, though extremes occur beyond both ends of the range.

Volcanic ash, for example, has very fast rates of soil formation. A mere 100 years after the massive 1883 eruption of Krakatau island in Indonesia, soil 25 centimetres deep had formed on a daughter island, Rakata (Dailly, 1995). In North Island New Zealand, soil and vegetation recovered equally as fast after the massive Taupo volcanic eruption of 1,850 years ago (Stevens et al., 1995).

In most cases, however, topsoil formation is much slower and is currently offset by the much faster erosion rates in many areas.

Although vegetation returns within a few years on an eroded site, the growth is generally less productive than before because the underlying soil is thinner and holds fewer nutrients. After hill country slip erosion, for example, pasture production takes approximately 20 years to recover to within 70-80 percent of its pre-erosion levels and some land may never achieve more than 70-85 percent of its pre-erosion potential (Trustrum et al., 1990). In severely eroded areas only a few stresstolerant weeds may be able to survive. After repeated erosion, sites may become barren.

Soil erosion is a natural process which occurs on all land, but has been accelerated by deforestation and unwise land use practices. At least 25 percent and possibly up to 80 percent of the world's agricultural land now has moderate to severe erosion (Pimental et al., 1995a, 1995b; Crosson 1995). About 30 percent of agricultural soils in the United States have been abandoned, largely because of erosion. Croplands are most susceptible because their soil is repeatedly tilled and left without protective vegetation cover. About half the world's pasturelands are also subject to erosion because of overgrazing (Pimental et al., 1995a).

Erosion rates in northern temperate countries (e.g. the United States and Europe) average 13-17 tonnes per hectare per year on cropland and about six tonnes per hectare per year on pasture (Pimental et al., 1995a and 1995b; Crosson 1995). This is relatively low by global standards, but is much higher than the rate in undisturbed forests (less than half a tonne per hectare per year) and greatly exceeds the average rate of soil formation-about one tonne per hectare per year (Pimental et al., 1995a).

Average rates of erosion and soil formation are not known for New Zealand, but a number of studies have shown that more soil has been lost than gained since European settlement (e.g. Hewitt, 1996; Trustrum and Page, 1991). The main forms of erosion have been:

  • surface erosion, which occurs when wind, rain or frost detach soil particles from the surface, allowing them to be washed or blown off the paddock;
  • mass movement erosion, which occurs when gravity combines with heavy rain or earthquakes to cause whole slopes to slump, slip, or landslide;
  • fluvial erosion, which occurs when running water gouges shallow channels (rills) or deeper gullies into the soil; and
  • streambank erosion, which is a special case of fluvial erosion that occurs when banks which have been cleared of tree cover become unstable and collapse because of trampling by livestock and gouging by flood flows.

Surface erosion can occur on any land where bare earth is exposed to wind and rain (e.g. after ploughing, or the burning of crop residues or stubble after harvest, or excessive grazing). If unchecked, surface erosion causes ongoing reductions of crop yield and pasture growth amounting to at least 20 percent and in extreme cases more than 60 percent (Hicks, 1995).

Table 8.8: Estimated incidence of soil degradation on agricultural land.
Region Erosion Compaction Nutrient decline Contamination
Northland local local widespread probable
Auckand negligible local local probable
Waikato negligible local local probable
Bay of Plenty negligible negligible local probable
Gisborne widespread negligible widespread possible
Hawkes Bay local local widespread possible
Taranaki local negligible local probable
Manawatu-Wanganui local local local possible
Wellington local negligible widespread possible
Nelson-Marlborough local negligible widespread possible
West Coast negligible negligible local possible
Canterbury widespread local widespread probable
Otago widespread local widespread probable
Southland local local local probable

Source: Clough and Hicks (1993)

Widespread: occurs in many parts of the region.

Local: occurs in several parts of the region.

Negligible: occurs in few parts of the region.

Probable: site-specific, probably more extensive than reported.

Possible: site-specific, possibly more extensive than reported.

Table 8.9: Erosion susceptible land in New Zealand.
Region Area at risk (hectares) At-risk area being farmed Percentage of farmland that is:
stable erodible
Northland 874,300 521,100 30 70
Auckand 263,300 150,600 52 48
Waikato 945,900 511,700 63 37
Bay of Plenty 682,400 153,500 56 44
Gisborne 697,700 497,300 17 83
Hawkes Bay 967,400 537,900 38 62
Taranaki 341,400 116,000 72 28
Manawatu-Wanganui 1,299,800 809,700 42 58
Wellington 454,300 279,100 42 58
Nelson-Marlborough 1,686,600 528,900 18 82
West Coast 1,865,200 16,700 90 10
Canterbury 3,309,300 2,288,400 11 89
Otago 2,791,800 1,616,800 10 90
Southland 2,202,800 653,300 40 60
New Zealand 18,382,200 8,681,000 32 68

Adapted from Clough and Hicks (1993)

Table 8.10: Areas affected by erosion, 1975-79
Erosion Type North Island South Island New Zealand
Hectares Percent Hectares Percent Hectares Percent
Surface erosion 2,662,600 23 11,150,000 74 13,812,600 52
Sheet 2,117,400 19 8,318,400 55 10,435,800 39
Wind 526,900 5 2,865,800 19 3,392,700 12
Scree 417,200 4 3,259,900 22 3,677,100 13
Mass movement 5,038,200 44 4,602,000 31 9,640,200 36
Soil slip 3,397,000 30 3,615,800 24 7,012,800 26
Earth slip 280,300 3 58,500 <1 338,800 1
Debris avalanche 1,218,900 11 1,603,000 11 2,821,900 11
Earthflow 1,011,500 9 33,300 <1 1,044,800 4
Slump 65,800 <1 44,100 <1 109,900 <1
Fluvial erosion 1,621,900 14 1,440,100 10 3,062,000 12
Rill 13,700 <1 56,400 <1 70,100 <1
Gully 1,157,400 10 803,500 5 1,960,900 7
Tunnel gully 327,600 3 98,700 <1 426,300 2
Streambank erosion 240,400 2 491,100 3 731,500 3

Source: Eyles (1983)

1 Areas are those of mapping units in which the erosion occurs and not of the individual erosion scars. The total area of erosion cannot be derived from this table because different erosion types may occur within the same mapping unit.

Figure 8.9: Erosion in New Zealand, 1975-1979
Textual description of figure 8.9

Sheet erosion primarily occurs in mountainous areas of the South Island and localised areas in the North Island. Soil slip erosion occurs primarily in agricultural areas the North Island and upper South Island. Gully erosion occurs in only a few areas and is scattered through out New Zealand but is common around the East Cape.

Source: New Zealand Land Resource Inventory 1975-79 (Landcare Research)

Most hill slopes steeper than 15 degrees are susceptible to mass movement, and those steeper than 28 degrees generally have a severe potential. Storms are generally the primary triggers (Fransen and Brownlie, 1995). Mass movements initially reduce pasture growth by 40 percent to 80 percent. After they have been regrassed, growth may completely recover but, more often, it remains depressed by 10 percent to 40 percent. Mass movement also damages fences, tracks and drains, and occasionally buildings (Hicks, 1995).

Fluvial erosion can cause rills even on flat land if ground cover is depleted. On sloping land it also causes gullies which can cut deep into the underlying subsoil or undermine surrounding soils. The sediment is washed into streams. Gullies create long narrow scars which do not usually cause much loss of crop or pasture but may disrupt vehicle and stock movements around the farm (Hicks, 1995).

The susceptibility of land to erosion is determined by soil characteristics, geology, and slope. Erosion itself, however, is often triggered by external pressures from climate and land use. Although two-thirds of New Zealand and its farmland are erosionsusceptible (see Table 8.9), areas with potential for very severe or extreme erosion are restricted to unstable hill country. Whether erosion actually occurs in these areas depends on the vegetation cover and land use practices.

The most recent national survey of the actual extent of erosion in New Zealand was carried out two decades ago as part of the NZ Land Resource Inventory (see Table 8.10 and Figure 8.9). Some regions have since been remapped as part of the current NZLRI update, but the only national level statistics still relate to the 1970s. It showed that severe to extreme erosion affected nearly 10 percent of the country, mostly in the mountains and in unstable hill country, such as the South Island high country and the East Coast of the North Island (Eyles, 1983). Although half of New Zealand was affected by surface erosion, and around one-third by slips and other mass movement, much of this was slight. From that survey and subsequent erosion incidents, it is known that North Island areas with serious land instability and erosion include northern Hawke's Bay, East Cape, Northland, the Volcanic Plateau, King Country, Rangitikei, inland Taranaki, and the Wairarapa. South Island areas include Marlborough and coastal Otago.

Surface erosion is significant in Northland and on the Volcanic Plateau. It is also a problem in large areas of the eastern South Island, particularly in North Otago and South Canterbury where soil is picked up by warm dusty northwest winds during droughts and springtime seed bed preparation. A single episode can remove up to 100 tonnes of soil per hectare, or 5 percent of the total (W. Armstrong, 1993). In the South Island high country, surface erosion is caused by the combined effect of frost and wind on denuded soils (O'Connor and Harris, 1991). This has played a significant role in the desertification of around 200,000 hectares of former tussock land which is now bare or covered in hawkweed (Belton, 1991). Surveys have found that in various high country districts anything from 7 percent to 44 percent of the ground is devegetated (Clough and Hicks, 1992).

Recently, Landcare Research scientists estimated the rates of surface erosion in several parts of the South Island by measuring the amount of radioactive caesium (137Cs) in the topsoil. This radioactive material was generated between 1953 and 1976 by global atmospheric atomic weapon testing. It was deposited by rainfall and quickly adhered to surface soil particles. Now, by measuring the distribution of these caesium-bound soil particles, erosion and sedimentation rates can be estimated for the past 40 years. In both South Canterbury cropland and unfarmed Central Otago arid land, the average yearly erosion rates appear to exceed 10 tonnes per hectare (Basher et al. 1995; Hewitt, 1996).

In the Central Otago study, significant surface erosion was estimated to affect about a quarter of the landscape near Earnscleugh. In the eroded parts of this area, the topsoil had declined from an original thickness of at least 9.8 centimetres before the arrival of rabbits and sheep to about 4.3 cm. About 3.4 cm of this loss had occurred in the past 40 years. The remaining topsoil on the most eroded slopes is projected to disappear entirely in the next 44 to 72 years (Hewitt, 1996).

Gullies and slips occur in large areas of Gisborne, Hawke's Bay, eastern Wairarapa, the Volcanic Plateau and inland Taranaki, where erosion rates in steep pasture land have been at least 2 millimetres per year since deforestation 75 years ago (Blaschke et al., 1992). In fact, in the small, steep, Lake Tutira watershed (3,208 hectares), sediment cores reveal an average erosion rate of 14 mm per year in the 110 years since the catchment was cleared for pasture (Trustrum and Page, 1991). A study of slip erosion in two small Hawke's Bay catchments (1,573 hectares) found that 1.3 percent of the total area (20 hectares) was affected, nearly all of it pasture land rather than forested land (Fransen and Brownlie, 1995). It took almost 27 years for slips to be fully recolonised by pasture grasses, which is similar to the 30 year recovery time recorded elsewhere in the North Island (Hicks, 1988).

A study on three hill country properties in Taranaki found that the cumulative effects of erosion amounted to a 20-30 percent long-term reduction in productivity on moderate slopes (around 30 degrees), and a 60 percent reduction on steep slopes (greater than 33 degrees) (Gane et al., 1991). The worst eroded area of New Zealand is the Gisborne-East Coast region. Since the conversion of forest to pasture, 200,000 of its 835,000 hectares have been seriously affected by slips. The most spectacular examples occurred during Cyclone Bola in 1988 (see Box 8.5).

Box 8.5: Erosion on the North Island East Coast

The East Coast region represents only 7 percent of the land area of the North Island, but it contains 50 percent of the worst soil erosion (Miller, 1991). An early example was the Tarndale Slip, one of the largest single examples of fluvial and mass movement erosion in New Zealand. In 1915 Tarndale was a single, narrow gully. Today the slip covers an area of 50 hectares and the soil it has dumped in the gully has raised the bed of the river below by approximately 30 metres (Glasby, 1991). The erosion problems began more than a century ago when European settlers began deforesting the unstable hill country for pasture. Every heavy rain storm resulted in land slips and large amounts of sediment washing into rivers. The infilling of riverbeds by fine silt and gravel was noted from about 1910 onwards, with many river flats and low terraces gradually being overwhelmed since then. As the riverbeds rose, the incidence of flooding increased. By the late 1930s, floods and erosion were causing serious alarm among local residents. Floods in 1938 and 1948 resulted in stopbanks being constructed along the Waipaoa River.

A review of the area in the 1960s led to recommendations for a government-funded soil conservation project whose features would include better land use planning and controls, the retirement of unsuitable land from agriculture, and, most importantly, the afforestation of eroding and erosion prone land (Taylor, 1967). The East Coast Project began a year later. Although the project met resistance from many farmers who did not want pine forests on what they saw as good grazing land, trees were planted on a number of slopes. However, the project conflicted with government policies of the 1970s and early 1980s (such as agricultural subsidies and development loans) which encouraged farmers to bring land that had reverted to indigenous vegetation back into pasture. Policy changes in the mid-1980s ended most subsidies for pasture development and forest planting. In 1987, a government review of the East Coast Project concluded that there was no role for further government support and predicted that commercial forestry interests would probably plant the eroding land.

Three months later, in March 1988, Cyclone Bola hit. Over 72 hours it deluged the region in nearly half a metre of rain, with nearly a metre falling in some locations (Trotter, 1988). The erosion which resulted followed earlier patterns but the scale was unprecedented. The hardest hit area was the 52,000 hectare Uawa catchment, nearly all of which is Class 6,7 and 8 land (Marden et al., 1991). About 60 percent of this was pasture land but a quarter had been planted in exotic forest under the East Coast Project. The cyclone provided an excellent test for both types of land cover. The pasture land lost an estimated 46 million tonnes (ranging from 660 to 3,439 tonnes per hectare), while the forested land fared better, losing an estimated 12 million tonnes (217 to 2,019 tonnes per hectare). The heaviest soil losses on the forested land came from recently planted forests. Land which had been under forest for eight years or more had only a tenth the soil loss rate of pasture land.

A similar result was found in the small Pakuratahi and Tamingimingi catchments near Napier, each of which is nearly 800 hectares in area (Fransen and Brownlie, 1995). Back in 1943, when the area was hit by a severe storm, 83 percent of Pakuratahi and 91 percent of Tamingimingi were in pasture. Pakuratahi, being steeper, suffered close to 300 slips per square kilometre while Tamingimingi sustained about 230. When Bola struck 45 years later little had changed in Tamingimingi catchment, where 93 percent of the land was in pasture and the slip density was 130 per square kilometre. In Pakuratahi, though, pine forests had been planted since the 1970s and, by 1988, 73 percent of the catchment was forested, with an additional 7 percent under native scrub and only 16 percent still in pasture. Pakuratahi's slip density during Bola was just 22 per square kilometre.

In economic terms, Cyclone Bola was one of the worst natural disasters New Zealand has experienced. The cost of property damage is estimated to have been nearly $120 million, and some 1,500 affected landowners received relief payments amounting to $60 million. Smaller erosion events continued after Bola, but neither farmers nor forestry companies showed strong inclinations to begin forest planting in the eroding areas. As a result, the Government in 1992 launched the East Coast Forestry Scheme to once again subsidise the planting of exotic forests on erodible land. The scheme involves the participation of forestry companies, local authorities, and landowners, including the local Māori iwi, and aims to plant some 200,000 hectares of eroding pastoral land over a 28 year period.

Loss of carbon and organic matter

Carbon, the 'molecule of life', is an indicator of the amount of organic matter in the soil. Organic matter maintains soil structure by binding soil particles into small clumps or aggregates. The spaces between these aggregates function as pores which enable water to drain through and roots to penetrate. Loss of organic matter causes the clumps to lose their structure and collapse into the pores, clogging the soil.

The vegetation and soils of natural forests hold 20-30 times more carbon than agricultural lands (Tate, 1993). This is because in native forests, and also tussock, organic material has built up in the soil over thousands of years through the decay of dead roots, leaves, organisms, and animal waste. When land is converted to farmland, this natural recycling system is replaced by a system which regularly removes significant amounts of plant and animal matter as food, fibre or waste. Carbon may also be lost from soil as a result of changes in grassland species caused by prolonged high inputs of nitrogen (Wedin and Tilman, 1996).

Carbon loss is worst in land heavily cultivated for crops. A high proportion of the organic material is removed at harvest. Ploughing causes further significant losses by breaking up the soil, exposing more organic matter to the air and allowing bacteria to decompose it faster. Rather than helping to bind soil particles, the carbon is expelled by the bacteria as carbon dioxide. One New Zealand study found that ploughing alone had caused a 25 percent depletion of soil carbon over 25 years (Harrison et al., 1993). Large decreases in soil carbon have been reported for market gardens in South Auckland and for croplands in Waikato, Manawatu, Canterbury, and Otago (Gradwell and Arlidge, 1971; Cotching et al., 1979; Ross et al., 1982; Hart et al., 1988; Ross et al., 1989; Haynes and Francis, 1990; Shepherd, 1991; Sparling et al., 1992; Tate, 1992).

Pastoral farming has less impact on soil carbon because, although removals are significant (via grass which is converted into animal carcasses, wool, milk etc.), some carbon is returned in animal faeces, decaying grass roots (much denser than crop roots), and their associated microorganisms (Jenkinson, 1988). Carbon losses in hill and high country pastures tend to be associated with erosion rather than plant removals. On the Canterbury Plains, where the dominant farming system is mixed cropping, many farmers have a relatively well balanced rotation in which grazed grassclover pastures are grown for two to five years, followed by two to five years of cereals and crops. As a result, the total organic matter content of the soil remains relatively unchanged through the cropping rotation (Haynes and Francis, 1991). Several studies of long-term grazed and fertilised grass-clover pastures have shown a build-up of soil carbon (Jackman, 1964).

The difference in impact of pasture and cropping on carbon levels was demonstrated in the Manawatu-Wanganui region when pasture was converted to maize fields. Continuous maize cropping caused a rapid decline in soil carbon and, in some cases, a serious deterioration of soil structure. Over a decade, about 20 tonnes of carbon per hectare were lost to the atmosphere from the top ten centimetres of soil (Shepherd, 1991). This was because the maize left insufficient decaying plant residue to replace the large amount of carbon lost in ploughing. The rate of carbon loss was more than double that reported for many North American soils that have been intensively cultivated (Shepherd, 1992).

Compaction and loss of soil structure

While carbon loss is an indicator of declining soil structure, compaction is a consequence of it. When soil structure is lost, and the pores fill in, the soil becomes hard and compact and forms a layer which plant roots, air, and water cannot penetrate. Plant growth declines and bare ground may result. Soil compaction is a primary land degradation problem overseas, but in New Zealand it has only recently gained recognition as a significant problem.

An estimated 2.1 million hectares of mixed pasturecropland and permanent cropland is considered to be at risk (Ross and Wilson, 1983). Compaction is not restricted to carbon depleted soils. Even soils with high organic content can have their structure broken down if heavy loads are applied to them when wet. Heavy machinery and treading by farm animals, particularly cattle, are common causes of this type of soil breakdown.

Compaction can occur at different levels of the soil profile, ranging from surface capping and pugging, through slumping and hard setting of the ploughed layer, to subsoil compaction below the ploughed layer and under wheeltracks and orchard alleys.

Surface capping is common on fine, sandy and silty soils with low organic matter, particularly where the land is subject to continuous cropping or market gardening. Under heavy rain, soil particles which have become detached from the deteriorating soil aggregates form a muddy layer which blocks the soil pores. Upon drying, a dense surface crust forms a rigid cap which may be strong enough to prevent seedlings breaking through (Haynes, 1994).

Pugging, caused by stock treading, is most common on dairy farms but can occur in any wet pasture with large concentrations of animals. It occurs when the soil is so soft that the hooves of grazing animals sink into it, compressing it and blocking the soil pores. The effect tends to be selfperpetuating because partially pugged soil prevents water from draining away. As a result, the soil remains soft and wet for longer, permitting hooves to do more damage at subsequent grazings. Severe pugging in the Manawatu, coastal Otago and Southland, Northland and the Waikato, has reduced annual pasture yields in the affected paddocks by 15-30 percent (Haynes, 1994). However, provided the pugged soil remains under grass and is protected from further animal treading, it recovers its structure much faster than cropped soil compacted by cultivation machinery. Volcanic soils are less susceptible to pugging than the yellowgrey soils which occur in many parts of New Zealand. The latter have a naturally compact subsoil which is slow to drain. As a result pugging is relatively common on the yellowgrey soils, particularly where mobstocking is carried out over winter.

Slumping and hard-setting are one stage worse than surface capping and pugging. They occur in soils that are fine, sandy or silty and have low organic matter with very unstable aggregates. After ploughing, the aggregates slump back to the same density as before and may congeal if they get wet. When the congealed soils dry they set hard into one complete dense mass without cracks which crop roots cannot penetrate and water cannot infiltrate (Haynes, 1994). Slumping and hardsetting can be seen in some soils that have been subject to regular rotary cultivation (e.g. market garden soils) and in arable soils, particularly where row crops (e.g. maize) have been grown continuously. However, hardsetting is not a problem on peats, very sandy soils, or on soils of volcanic origin, where many market gardens are located (Haynes, 1994).

Subsoil compaction occurs when the heavy weight of machinery causes a hardened layer to form below the depth of normal ploughing. Tractors and other wheeled vehicles and equipment, as well as plough soles, disc edges and rotary blades can all cause subsoil compaction in wet soft soils. Cropland, orchards and berry gardens are all susceptible. Subsoil compaction limits drainage and also prevents plant roots from penetrating (Haynes, 1994).

Compaction on maize farms in the Manawatu- Wanganui region has resulted in a rapid and marked loss of production and profitability. Over a ten year period, maize yields have steadily declined by 24-45 percent on poorly drained Kairanga soils, and by 15-30 percent on welldrained Manawatu soils. It is estimated that twothirds of the arable land in the Manawatu-Wanganui region is highly susceptible to severe soil compaction (Shepherd, 1991).

Under Canterbury mixed cropping regimes, a similar deterioration of structure occurs during the cropping phase, but soil structure improves under the pasture phase as a result of the build-up of organic material, earthworm activity etc. Structural deterioration under long-term cropping makes the soil harder to work and more susceptible to wind erosion. Concern has been expressed that current economic pressures are leading to an intensification of cropping at the expense of the pasture phase of the rotation (Haynes and Francis, 1991).

Since the renewed intensification of agriculture in the 1970s and 1980s, examples of serious compaction have been reported, though not systematically monitored, throughout New Zealand from Northland (in horticultural land near Kerikeri), the Waikato (in lowland pasture), Manawatu (in lowland pasture and maizefields), Hawke's Bay (in orchards near Hastings), Canterbury (in cropland), and South Otago and Southland (in land used for winter strip grazing).

Nutrient depletion

Nutrients are lost from agricultural land when plant and animal matter, which would normally decompose and return to the soil, is removed in crops and animal products. The lost nutrients ultimately end up in urban landfills, sewers, surface waters, and other countries. In natural ecosystems nutrient replacement is a gradual process. Under intensive agriculture, fertilisers are needed.

Unlike farming, forestry only requires fertilisers where natural soil fertility is low (on podzolised soils of Northland or the West Coast), where short rotation 'pulpwood' eucalypt forests are grown, or where 'whole trees', including the foliage and small branches, are removed from the site. Most planted forests in New Zealand, many in their fourth rotation, have received no artificial fertiliser input, because most of their nutrient-bearing bark and foliage is left on site.

On pasture, nutrients are lost not only through harvesting, but also when animal waste is deposited in concentrated patches, such as the "camp" area where the animals settle for the night, rather than being evenly redistributed throughout the grazed pasture. This is a significant cause of nutrient decline on the hilltops and slopes of hill country farms (Williams and Haynes, 1991; O'Connor and Harris, 1991). Erosion, leaching, and chemical processes in the soil can also remove nutrients.

To counter such losses, agriculture is very dependent on expensive imported fertilisers containing phosphorus, sulphur, potassium, and, increasingly, nitrogen (Goh and Nguyen, 1991). In the past, nitrogen fertiliser has been used on a small scale in exotic forests (Mead, 1986), but an increasing number of dairy farmers are now applying it (Roberts et al., 1992). Other farms continue to obtain their soil nitrogen from nitrogen-fixing legumes, such as clover (in pasture land) and peas (in crop land).

The removal of fertiliser subsidies in 1984, coupled with a downturn in commodity prices, led to a four year decline in fertiliser use, particularly on hill country properties. Fertiliser consumption began to increase again from 1988, but did not return to peak levels until recently (see Figure 7.11, Chapter 7). Much of the increased fertiliser use has been on dairy farms rather than hill pastures. A review of nutrient inputs and losses on New Zealand's 14 million hectares of pastoral grasslands concluded that phosphorus, nitrogen and sulphur levels were probably sufficient because of high previous levels of fertiliser use and nitrogen fixation (White, 1991). Concern was expressed, however, that potassium levels appeared to be declining and acidity may have increased because of reduced lime applications.

For farms without a buildup of residual soil nutrients, the fertiliser downturn appears to have reduced pasture productivity (Hicks, 1995). A study of one of the worst areas, the South Island tussock lands, found that annual inputs of all four major nutrients were insufficient to replace the losses (O'Connor and Harris, 1991). This is consistent with mounting scientific evidence that nutrient decline is becoming a significant problem on hill country farms throughout New Zealand, particularly in the King Country, East Coast, and Hawke's Bay regions of the North Island, as well as the eastern South Island hill country (Trustrum et al., 1990; Williams and Haynes, 1991; Sheath, 1992).

Soil acidification

Like nutrient loss, acidification reduces the fertility of soil. Acidity is measured on the pH scale, an index reflecting the proportion of hydrogen ions formed when a substance dissolves in water. Soils with a pH of less than seven are acidic, while those above seven are said to be alkaline. Most plants thrive in slightly acid soils, but few can cope with strong acidity (i.e. where pH is below five). Ryegrass and clover, the predominant pasture species, are particularly sensitive, with an optimal pH range of 5.7 to 6.5 (During, 1984).

In some areas, the acidification process is linked to the phenomenon of aluminium toxicity. When soil pH drops in some parts of New Zealand, such as inland Marlborough, essential elements (e.g. manganese) and inessential ones (e.g. aluminium) can become highly concentrated in the soil and inhibit pasture growth (Goh, 1994).

Acidification occurs when the more alkaline elements in the topsoil (e.g. calcium, potassium and magnesium) are removed. This happens when they link up with nitrates or other easily leached molecules which are then washed out of the topsoil by rainfall. The acidic ions that are left behind become more concentrated, gradually lowering the soil's pH. Although acidification is a natural phenomenon (the kauri forests, for example, had very acid soils) it can be accelerated when the topsoil receives large amounts of nitrate (e.g. from animal urine, nitrogen fertiliser or clover growth) or sulphur fertilisers (Goh, 1994).

Present day pH levels have led some experts to suspect that New Zealand's soils are becoming more acidic, while others consider this unlikely. Hard evidence cannot settle the matter because national trends in soil acidity have not been monitored. All that can be deduced from soil maps is that most of our land has the potential to become acidified (Parfitt and MacDonald, 1992). Susceptible soils, mainly on hill country, are widespread in the Northland, Auckland, Manawatu-Wanganui, and Wellington regions of the North Island, and throughout South Island regions.

However, most nitrate leaching does not occur on the susceptible hill country soils. The best clover growth and the heaviest concentrations of animal urine and nitrogen fertiliser are in flat dairying areas rather than hill country. Furthermore, the high rate of nitrate leaching is often countered by the use of lime (calcium carbonate) to reduce soil acidity. Lime is an alkaline powder derived from limestone. Limed agricultural soil is often less acid than the original forest soil that preceded it. However, lime use, along with fertiliser use in general, fluctuates in response to economic trends. Through the late 1980s, the national tonnage of lime application was barely half that required to counter acidification (White, 1991). Lime use has increased in recent years, but so has nitrogen fertiliser use. It is not known whether current applications of lime are sufficient to reverse pasture deterioration on some of the poorer hill country soils.

Chemical contamination of soil

Contamination of land by hazardous chemicals, residues, and waste products is another form of soil degradation that New Zealanders have only recently begun to deal with (Ministry for the Environment, 1992b). Land is considered to be contaminated when hazardous substances are present at concentrations that are likely to pose an immediate or long term hazard to human health or the environment (ANZECC/NHMRC, 1992).

Humans and other animals can be exposed to soil contaminants in a number of ways, including direct contact with the soil, swallowing food or water from contaminated environments, and breathing contaminated dust. Apart from the health hazards, the presence of contaminants can also limit land uses, threaten building structures and services, and reduce land value.

Table 8.11: Industries, typical site activities, and likely soil contaminants.
Industry Sector Examples of Industrial Activities Examples of Likely Contaminants
Chemical Manufacture and use of acids/alkalis, pigments, dyes, fertilisers, pesticides, adhesives, resins, plastics, oils, pharmaceuticals, paints, timber treatment Acids/alkalis, metals, non-metals (e.g. compounds of boron,arsenic, sulphides, chlorides), solvents (e.g. toluene, benzene),chlorinated organics (e.g. DDT, dieldrin, chlordane, PCP); other organic compounds (e.g. phenols, carbamates, organophosphates)
Petrochemical and Energy Oil refineries, tank yards, fuel storage tanks, bitumen manufacture, power stations, oil refining, gasworks Hydrocarbons (e.g. petrol, diesel, polyaromatics, tars), phenols, acids/alkalis, metals, asbestos, fuel and coal products, cyanide and sulphur compounds, other organic and inorganic compounds
Metal industries Iron and steel production, ferro-alloy and metal casting works, metal products manufacture, electrical products (e.g. transformers; batteries), metal coatings(e.g. anodizing and galvanizing), scrap yards, metal recycling, drum reconditioning, electroplating Metals (e.g. Fe, Al, Cu, Ni, Cr, Zn, Cd, Hg, Pb), asbestos, hydrocarbons, PCBs, acids/alkalis, cyanides, inorganics, solvents
Transportation Service stations, engine maintenance shops, railway yards, airports Fuel, hydrocarbons, asbestos, PCBs, pesticides, metals, acids, solvents
Mining, waste disposal Ore extraction, landfills and waste dumps Metals (e.g. Cu, Zn, Pb), inorganics, gases (e.g. methane), cyanides, phenol, PCBs, acids, leachates etc.
Others Ports, tanneries, military land Metals, organic compounds, hydrocarbons, methane, toxic, inflammable or explosive substances
Table 8.12: Some land uses and risk factors associated with exposure to contaminants.
Examples of Land Uses Risk Factors Examples of contaminants of concern
Agricultural soils and domestic gardens; recreation areas Ingestion of contaminated soil (children); adsorption of pollutants by crops and animals; phytotoxicity Arsenic, cadmium, chromium, mercury, lead, free cyanide, copper, nickle, zinc, PAHs, phenols, organochlorines
Residential, commercial and industrial areas Chemical corrosion of building materials and services, fire, explosion Sulphate, sulphide, sulphur, chloride, methane, phenols, mineral oils, ammonia; oily and bituminous substances, coal dust
Construction site Direct contact with pollutants (workers) PAH, phenols, asbestos, oily and bituminous substances
Surface and groundwater Intake of water Phenols, cyanide, sulphate, soluble metals

Many potential contaminants have been identified in oil products, pesticides, fertilisers, industrial chemicals and byproducts, and waste disposed of at landfills. The contaminants include heavy metals (e.g. cadmium, arsenic, zinc, lead, copper, chromium, mercury etc.), organochlorines (e.g. DDT, DDE, PCP, PCBs, dieldrin, aldrin, chlordane, lindane, dioxins etc.), solvents (e.g. industrial cleaning agents), corrosives (e.g. acids and alkalis), hydrocarbons (e.g. oil derivatives such as petrol, diesel, tar and creosote), asbestos, and cyanides.

Contamination is not always limited to a specific site. Contaminants may seep through the soil into groundwater, or be carried to adjacent land and waterways in rainwater or on dust particles. Vapour and gases may emanate from a contaminated site (e.g. volatile hydrocarbon vapours from a service station or methane and carbon dioxide from a landfill site) and may pose a number of hazards (e.g. potential for explosion, odour nuisance, or asphyxiation). Some soil contaminants, particularly heavy metals and some organochlorines, can accumulate in animal tissue, with concentrations increasing as they progress up the food chain. The heaviest concentrations are generally found in top predators, particularly birds of prey, large fish and marine mammals. Human health concerns most commonly focus on the risk of long term toxic effects such as cancer. They are of particular concern because of their insidious nature, in that the harmful effect may only become apparent years after the exposure has occurred. Furthermore, the magnitude of these risks is difficult to estimate accurately.

Investigations in other countries have shown that soil contamination can be caused by a range of industrial activities, including those associated with: chemical manufacturing and use; timber treatment; paint, pesticide and pharmaceutical manufacturing and use; the petrochemical, gas, coal and mining industries; the transport sector; the metal industries (e.g. smelting, manufacturing, recycling); and waste disposal (see Tables 8.11 and 8.12). Agricultural activities can also cause soil contamination (e.g. from pesticides, fertiliser residues, storage areas and dump sites).

The actual extent of land contamination is not yet known, but the number of potentially contaminated industrial sites and landfills is estimated from business and telephone directories and other 'desktop' records to be around 7,800 (Ministry for the Environment, 1992b). No formal assessment has been made of potentially contaminated agricultural and horticultural sites, but the total number could be 4,000 or more (Ministry for the Environment, 1992b).

About 1,580 (22 percent) of the potentially contaminated urban and industrial sites are estimated to pose a high risk of harming human health or the environment. Local authorities are now conducting surveys to systematically identify high risk sites using a rapid hazard assessment methodology (Ministry for the Environment 1993b). Sites surveyed to date include landfills and waste dumps, service stations, oil storage terminals, gas works, pesticide manufacturing plants, dieldrin dump sites, sawmill and timber treatment sites, and some farm and disused mining sites.

Landfills

Landfills receive an estimated 3 million tonnes of waste every year and are likely to continue as the main destination of solid wastes in New Zealand (see Chapter 3). They can pose a risk to human health or the environment because of the contaminants that can escape from them into air, water or soil, even after closure (Ministry for the Environment 1992b).

Gases which escape from landfills can create a risk of methane explosions or asphyxiation by carbon dioxide. In one recent case several Wellington homes had to be evacuated because of gases issuing from a landfill site which had been closed many years previously. This illustrates the fact that the adverse consequences of poor landfill management may take many years to emerge. In contrast, the operators of a number of engineered landfills are taking advantage of the methane emissions by harnessing them as a source of energy.

In the past, most landfills were inadequately managed and many were inappropriately located. Older landfills were often in sites near waterways, or situated in old gravel pits, where leachate could enter the groundwater. Now, under the Resource Management Act, all operating landfills require consents from the local authority. In setting out the environmental conditions that must be maintained at each site the consents draw on national guidelines.

Three sets of guidelines have been developed to cover: the collection of statistical data on landfill wastes; the siting, design, operation and after-care of landfills; and the identification, sources and management options for waste hazardous substances (Ministry for the Environment, 1992c, 1992d, 1994a). The New Zealand Landfill Guidelines are based on the United Kingdom model of landfills as 'waste treatment' facilities (Ministry for the Environment, 1992d). The Waste Analysis Protocol provides a systematic approach to data collection which will enable a better assessment in the future of the environmental effects of landfills and the cumulative level of risk posed by them (Ministry for the Environment 1992c).

At a number of closed landfills, local authorities are conducting leachate monitoring programmes and other investigations to assess the risks they pose and consider management options for high risk sites. For example, Auckland City Council commissioned a review of 85 closed landfills which had been used between 1910 and the late 1970s (Tonkin & Taylor Ltd 1994b). All the landfills were ranked for their potential to harm the environment and human health.

The pre-1950 landfills tended to be small road and quarry infills with limited organic waste. The post-1950 landfills were larger, and had been filled with greater proportions of decomposable matter. In the 1960s and 1970s, some of the landfills in coastal gullies and tidal wetlands had industrial waste dumped in them. Ten sites were singled out for detailed investigations which revealed that:

  • leachate was seeping from all ten landfills, with the youngest sites having the highest levels of Biochemical Oxygen Demand (50 to 200 mg/l), Chemical Oxygen Demand (500 to 1,500 mg/l) and ammonium (up to 100 mg/l);
  • the leachate had very low concentrations of dissolved metals, mostly zinc;
  • volatile (gaseous) and semi-volatile organic compounds (e.g. xylene, phenols etc.) were present in low concentrations in the leachate of young landfills, but not in the point source discharges;
  • significant concentrations of landfill gas were detected at eight of the ten sites, with methane concentrations of up to 65 percent and carbon dioxide concentrations of up to 55 percent at the younger landfills and more variable levels at the older ones; and
  • landfill gas was recorded in buildings at two of the sites with methane levels above the Lower Explosive Limit (LEL) in some floor slab joints and cracks, but not within the larger building areas, suggesting that the standard vent space beneath New Zealand buildings may protect them from gas explosions, fires and asphyxiation risks near disused landfills.

Following the investigation, cost effective management and remedial options were proposed (Tonkin and Taylor Ltd, 1994c). These included the capping of landfills to reduce leachate and seepage, the passive venting of landfill gas, the active and passive venting of particular buildings, and the sealing of floor slab joints. Where leachate discharges occurred near harbours, restrictions were proposed on shellfish gathering and recreational activities. Where environmental and human health effects were minor, the continued discharge of leachate and landfill gas was considered to be a valid option, bearing in mind that the council has an obligation under the Resource Management Act to consider the areas of greatest environmental benefit when targeting its resources.

Sawmills and timber treatment sites

Radiata pine is the dominant timber tree in New Zealand. Being a softwood, it is prone to attack by fungi and insects after felling. Sapstain fungi can discolour the timber surface while other fungi and wood-boring insects can cause decay. Chemical treatment is one of the methods used to protect the timber, though kiln-drying is increasingly used as a non-chemical preservation technique for indoor timber.

The main chemical preparations are anti-sapstain fungicides which are applied to the timber surface immediately after milling, and a range of preservatives which can provide varying degrees of long-term protection. Preservatives in common use since the 1970s include boron (an insecticide), LOSP (an insecticide which is applied as a light organic solvent preparation) and CCA (a heavy metal formulation of copper, chromium and arsenic which acts as a dual insecticide and fungicide).

The first timber preservative used earlier this century was creosote derived from coal tar. With the invention of organochlorine pesticides in the late 1940s and 1950s sawmills made widespread use of pentachlorophenol (PCP) to combat sapstain fungi. Some also mixed PCP with diesel oil for use as a preservative. The mixture was applied under pressure to force it into the wood fibres. Another organochlorine, chlordane, was used in the glue of some manufactured wood products (e.g. plywood), especially those intended for use in tropical regions. Significant use of these organochlorines in timber treatment ceased in 1988 and they were formally deregistered for these purposes by the end of 1991.

Soil contamination has probably occurred to a variable extent at many of the nation's 600 sawmill and timber treatment sites, some of which have now ceased to operate. Contamination would generally have been caused by the cumulative effects of drips, leaks, spills and waste disposal practices during an era when less emphasis was placed on the containment of potentially hazardous chemicals. Contamination tended to occur around treatment and storage areas, and appears to be largely confined to the top 30 cm of soil except possibly in some porous soils (McLaren, 1992; Armishaw et al., 1993; CMPS&F Ltd, 1995). PCP contamination down to 5 metres is known from at least one site (Royds Garden Environmental Services/CMPS&F, 1994). In some cases, water run-off took contaminants to adjacent land, or into waterways and estuaries (Wilcock et al., 1989; Shaw, 1990; Ministry for the Environment, 1992b).

PCP was used by some 70 percent of New Zealand's 400 operating sawmills between 1950 and 1988 (Bingham, 1991a, 1991b). It was also used to a lesser extent in the paper industry, the cultivation of mushrooms and the control of slimes in cooling waters (Shaw, 1990). On 31 December 1991 PCP was deregistered for use as a pesticide and prohibited from importation following confirmation that it contained toxic impurities (dioxins and furans) (Bingham, 1991b).

A National Task Group was established in 1991 to investigate potential site contamination problems (Ministry for the Environment, 1992a). The Group undertook a pilot risk assessment study of the timber processing complex at Waipa, near Rotorua, where PCP had been used over four decades. Significant contamination was found in the vicinity of treatment areas and in associated building dust. The groundwater discharging into the local stream contained high levels of PCP and the contaminants were found at elevated levels in the sediments and biota of Lake Rotorua (CMPS&F Ltd, 1992).

Following the Group's report, a National Steering Committee was established to facilitate actions at national and regional levels among government agencies and industry (Ministry for the Environment, 1993a). A major task carried out by the Committee was the preparation of guidance for the assessment and management of potentially contaminated sites, including the development of clean-up criteria (Ministry for the Environment and Ministry of Health, 1993). Recent initiatives have included:

  • removing contaminated dust and intercepting and treating contaminated groundwater at Waipa (Forestry Corporation of New Zealand, 1993);
  • producing a code of practice for the safe use of timber preservatives and antisapstain chemicals (Occupational Safety and Health Service, 1994);
  • assessing contamination at a former sawmill site in Hanmer Springs (Royds Garden Environmental Services/CMPS&F, 1994);
  • investigations by timber industry owners of a number of sawmill sites; and
  • a collaborative effort between the timber industry, government and technology developers which is currently undertaking treatability trials on soils contaminated with PCP and dioxins.

Service stations and oil storage terminals

Although no published studies exist on the extent of contamination at or near service stations, the sheer number of sites (about 2,600) and the opportunities for spillage and seepage make it likely that petroleum products are a major source of soil contamination. Guidelines have been developed to deal with the hazards associated with storage tanks (Occupational Safety and Health Service, 1992, 1995) and the industry has embarked on an underground storage tank replacement programme. In addition, a set of guidelines for service stations is being prepared by industry, regional councils and central government to give clear guidance on site assessment, management and clean up procedures.

One oil company commissioned an investigation of fifteen bulk storage oil terminals, some of them now closed, to assess baseline contaminant levels (Tonkin & Taylor Ltd 1995). All were built on superficial soil deposits (alluvial and estuarine sands, silts and clays) which, in some cases, were overlain by fill materials. Groundwater at these sites is 2 to 5 metres below ground level. The larger sites date from the 1930s, but have been upgraded with lined tank compounds, improved tank foundations in earthquake sensitive areas, and spill control measures. Contamination was mostly limited to areas associated with separators, drum filling, washing and storage operations, and loading and unloading racks for road tankers and rail cars. No significant off-site environ-mental and human health risks were identified.

Gas works

New Zealand has over 50 redundant town supply gas-works sites. Based on overseas experience, soils and groundwater at these sites are likely to be contaminated by polyaromatic hydrocarbons (PAHs), phenols and sometimes heavy metals. For example, in Napier, cyanide was found on neighbouring properties where contaminated waste from the gas works had been dumped many years before. Clean-up requirements are generally dominated by the cancer-causing chemicals, such as certain PAHs and some heavy metals.

No single remediation technology appears capable of coping with the full range of contaminants, so clean-up efforts to date have relied heavily on removing contaminated material to landfills and on site management procedures to minimise exposure to contaminants. The need for national guidelines was raised following site investigations at Masterton, Wellington, Gisborne, Hamilton, and Napier. The Ministry for the Environment is now developing guidelines to assist the assessment and management of these sites.

Mining sites

The mining industry produces large quantities of waste rock which often contain heavy metals. An estimated 92 sites, most of them no longer being mined, may be contaminated (Ministry for the Environment, 1992b). Nearly 80 percent of these are in just five regions-Otago, Southland, the West Coast, Waikato and Northland. The soil and water surrounding some abandoned and non-working mines often have elevated heavy metal levels. The barren four hectare Tui mine site near Te Aroha, for example, was used from around 1900 to 1973. Heavy metals, such as mercury, lead and zinc leach intermittently from the 100,000 tonne tailings dam above the site and have contaminated water and soil (Ministry for the Environment, 1992b; Morrell et al., 1995).

Mines that are currently operating are closely controlled by requirements under the Resource Management Act, though some are still struggling to resolve earlier mistakes. For example, the huge tailings dam at the Golden Cross mine near Waihi is now known to be on land that is unstable below the surface. The dam holds 3 million tonnes of rock dust mud containing arsenic, cadmium, copper and other metals. Efforts have been made to stabilise the slope and investigations are now taking place into the possible relocation of the tailings.

At the nearby open-cast Martha Hill goldmine efforts to minimise long-term impacts include the restoration of vegetation cover to tailing piles and oxidised waste mounds. So far, 16 hectares have been restored to productive pasture (Mason et al., 1995). The mine's waste water discharges are treated to keep heavy metals traces below the maximum concentration recommended at that time in the Health Department's pre-1995 standards for drinking water (Barker and Hurley, 1993).

Pesticide manufacture

About a dozen pesticide manufacturing sites exist in New Zealand. Little information has been published on the scale of land contamination associated with them. One exception is a disused site in the small coastal town of Mapua, on the Waimea inlet near Nelson. Various pesticides and agricultural chemicals were manufactured and formulated there from 1945 to 1988 by the Fruitgrowers Chemical Company.

Extensive soil sampling revealed widespread contamination by DDT (and breakdown products such as DDE) and more localised contamination by dieldrin (Woodward-Clyde Ltd, 1994). Although no soil samples from the neighbouring Tasman District Council landfill were analysed, the area is assumed to be contaminated because of the history of waste dumping from the chemical sites. Samples from wells show that groundwater is significantly contaminated by pesticides, sulphur, and chlorobenzene. Stormwater analysed in 1993 was contaminated by several organochlorines, organophosphorus pesticides, cadmium, copper, lead, selenium, and zinc.

The stormwater run-off from the Mapua site is considered to be the major cause of the organochlorine contamination in the nearby tidal waters. Organochlorine-contaminated sediment fans out onto the sea-bed and shoreline on both sides of the Mapua site (Fenemor and McFadden, 1996).

The concentrations of the organochlorines and heavy metals in the soils, groundwater, shoreline and seabed sediment, and the tidal water generally exceed acceptable limits. DDT levels in shellfish are many times higher than the acceptable limit, and birds such as bitterns and banded rails, which were once a common sight pecking at the shellfish on the mud flats, have disappeared.

Efforts by the Tasman District Council to coordinate a clean-up of the site were hampered by the extent and severity of the contamination, and by questions of liability because ownership of the site and landfill was in different hands. The Council inherited the landfill, and two companies, Ceres Pacific and Mintech, both owned portions of the original site. In June, 1996, the Minister for the Environment announced that the Government would contribute $1.2 million towards the estimated $2.75 million cost of the Mapua clean-up.

The clean-up operation will seal off both the soil and groundwater on the contaminated site, and prevent continued contamination of the stormwater. A diversion system will be built to redirect groundwater away from the site and the contaminants there, the site will be capped with asphalt to prevent rainwater infiltration, the existing stormwater system will be removed, and the buildings on the site will be either taken away for decontamination, or destroyed. Forty tonnes of chemicals, which had been stored in 250 drums in a locked concrete shed in a fenced off area of the Mapua site, were taken to a secure storage facility at Gracefield, near Lower Hutt, in May 1996 to be held there until the development of an acceptable way of disposing of them, or rendering them safe.

Dieldrin dump sites

Dieldrin is an organochlorine which was withdrawn in the 1960s. In 1961 its use as a pesticide was banned in sheep sprays and dips-along with aldrin, DDT, BHC and methoxychlor. In 1966 its use to control pasture pests such as black beetle and grass grub was also banned. Surplus stocks of the banned pesticides, estimated at 530 tons, were collected from farms around the country by the Department of Agriculture and stored in a number of depots. By March 1963, they had been placed in regional storage centres. By February 1964, approximately 165 tons had been redistributed to land development blocks in Northland, Hawkes Bay, Rotorua, North Canterbury, Canterbury and Southland to be aerial sprayed on pasture (Gibbs 1992).

Some, but not all, of the surplus pesticides were exported from New Zealand in late 1966. What happened to the remainder is still largely a mystery because, although some storage and dump sites have been located, the files and records for most of them seem to have been lost (Loe, Pearce & Associates, 1993, 1994). Where storage and disposal sites have been found, some decontamination activity has been undertaken (Taranaki Regional Council 1993). Contaminated soil and dust have been found at some former storage centres, including four Canterbury sites where the original buildings are still standing (Tonkin and Taylor Ltd, 1994a). A fifth Canterbury site is now covered by a shopping mall. Buried and rusting pesticide containers and evidence of some soil and groundwater contamination have been found at two dump sites in the Southland region (Woodward-Clyde Ltd, 1993b).

Agricultural land

Contamination of rural soils can arise from pesticide and fertiliser residues in fields and orchards, and from chemical spillage and leaching at "hot spots" such as pesticide and fuel storage areas, animal dip and dosing sites, and on-farm landfills. While no area is known to have contamination which is both severe and extensive, soils on some Canterbury and Southland farms have elevated levels of DDT and DDE which require ongoing management, and some farms throughout New Zealand have cadmium concentrations higher than the natural background levels, though still low by international standards (Roberts et al., 1994; Furness, 1996a, 1996b).

In addition, although the number of potentially contaminated rural sites has not been formally assessed, there may be as many as 1,000 storage and dump sites containing unwanted pesticides and herbicides and a further 1,000 private and farm landfills (Ministry for the Environment, 1992b). New Zealand may also have several thousand contaminated sheep and cattle dip sites if the Australian example is any guide. A large number of Australian dip sites that were used before 1965 have high DDT and arsenic concentrations (Ministry for the Environment, 1992b).

Organochlorines

From the 1940s until the 1970s, persistent organochlorines were heavily used in New Zealand agriculture (see Box 8.6 and Table 8.13). DDT was mixed with fertiliser and applied to pasture in a bid to control grass grubs and porina caterpillars. It was also used on lawns and market gardens, parks and sports fields. Its use was restricted in 1970 and was finally banned in 1989. DDT has a halflife of 10 years in dry soils, but its main residue, DDE, is far more persistent, showing little change in soil levels over 20 years.

Table 8.13: Organochlorine toxicity and half-lives.
Organochlorine pesticide Acute toxicity to rats (LD-50) (mg/kg) Toxicity to fish (rainbow trout) (µg/l) Half-life (time needed for 50% decay)
On the surface In the soil
Aldrin 67 2.6 1-18 days 5 years
Chlordane 250 42 27-72 days 5.7-10.6 years
DDT 250-300 8 1 year 10.5 years
Dieldrin 40-87 1.2 2-26 days 2.2-11.4 years
Lindane 88-125 30 1-5 days 0.5-2 years

Source: Bingham (1992)

Relatively low levels of DDT residues are widespread in regions where grass grub and porina are significant pasture pests (Wilcock and Watts, 1993; Roberts et al., 1996). In parts of Canterbury and Southland, paddocks on some farms are still unsuitable for livestock or dairy production because of the DDT residue levels. Soil samples from Canterbury have shown an average DDT residue concentration of 0.27 mg/kg, with wide variation among districts, farms and individual paddocks. The percentage of soils with concentrations above 1 mg/kg was 20 percent in Mid-Canterbury but only 2-4 percent in North, Central, and South Canterbury (Morton and Butcher, 1990).

These residues do not affect crops but may occasionally find their way into meat and dairy products when grazing animals eat soil with their grass. The amount of ingested soil increases in dry periods when grass is shorter and sparser. Residue sampling of Canterbury lambs in drought years during the 1980s found that nearly 40 percent had DDT residues above the European Union's permitted limit, though still within the New Zealand safe tolerance limits (MacIntyre et al., 1989).

A national meat and dairy product monitoring programme run by the Ministry of Agriculture ensures that export products contain low levels of these residues. The 1993/94 monitoring programme tested the fat tissue of 1,206 animals for organochlorines. It found that 656 (54 percent) had low levels of DDE and 17 (1 percent) had trace levels of DDT. However, only one animal exceeded the DDE tolerance level, and none exceeded the DDT level.

The new forms of pesticides made since the mid-1980s are designed to break down quickly, but very little is known about their residue levels in New Zealand soils and their effects on soil organisms. The newer pesticides are also more water-soluble than their fat-soluble predecessors, and thus more likely to move out of the soil into groundwater or surface water courses (Clough & Hicks, 1992).

Box 8.6: Organochlorine contamination

Organochlorines are chemical compounds which contain both carbon and chlorine atoms. Carbon and chlorine are versatile atoms which can combine with many others to form a wide range of molecular structures. At least 1,500 organochlorines are found in naturemore than 100 occur in wood smoke (Abelson, 1994a). Yet this is a small number compared to the 60,000 or so synthetic organochlorines that have been created in laboratories over the past 50 years.

Most organochlorines break down quickly into nontoxic residues which are dissipated in the environment. A small number, however, are highly persistent or longlived (e.g. DDT, DDE, PCBs, PCP, HCBs, dioxins, chlordane, lindane, aldrin, dieldrin). Some of these can kill or cause illness in fish, birds or mammals (see Table 8.13). Overseas research has also linked high concentrations of organochlorines to reproductive abnormalities and immune deficiencies in a number of species, though the interpretation of such data is still controversial (Addison, 1989; Tanabe, 1994; Motluk, 1995; Sharpe, 1995). Even more controversial are claims that low organochlorine concentrations can have similar effects through prolonged exposure or through a combined 'cocktail' effect (Colburn et al., 1996; Wilkinson and Dawson, 1996). The impacts of low to moderate doses on human health are still being debated as researchers investigate their effects on mammalian reproduction, hearing, thyroid function, and cancer susceptibility (Abelson, 1994a, 1994b; Stone, 1994a, 1994b, 1995; Dibb, 1995).

The persistent organochlorines were widely used in many countries between 1940 and 1980 in a variety of products such as: insecticides (e.g. DDT, aldrin, chlordane, dieldrin, lindane), fungicides (e.g. PCP) and electrical transformers and capacitators (i.e. PCBs). Some organochlorines, such as the dioxins (PCDDs) and furans (PCDFs), were created as waste byproducts in the manufacture of other organochlorines. All of these left residues in the environment. Western industrial countries stopped manufacturing many of the persistent organochlorines in the 1970s. Since then, their presence in the environment has fallen to a tenth or less of their former levels. Even so, the traces linger on. Most of us still have low but detectable amounts of DDE, a breakdown product of DDT, in our bodies. PCP and dioxin residues have been found at some New Zealand sawmill sites, and dieldrin residues have been found at former storage and dump sites. New Zealand also has a small number of farms and other sites where soils have elevated organochlorine pesticide residues.

Compared to many northern hemisphere countries, however, the New Zealand environment probably has low organochlorine levels. A recent world survey of 22 different organochlorine compounds in tree bark found generally low concentrations (less than 100 parts per billion) at the three New Zealand sites sampled, though moderate concentrations (100 to 1,000 ppb) of DDT, chlordane and dieldrin were found at one of the two South Island sites (Simonich and Hites, 1995). The Ministry for the Environment has initiated a three year programme to deal with dioxins and other organochlorines of concern, including PCBs, PCP, DDT and dieldrin. Human and environmental data will be collected to establish background levels and assess the significance of any ongoing emissions. The programme will develop National Environmental Standards and guidelines for these substances and appraise the suitability of clean up technologies.

Heavy metals

A survey of heavy metal contamination in pastoral soils was conducted in the early 1990s (Roberts et al., 1992b, 1994). A total of 312 farm sites were sampled in both the North and South Islands. Samples were also collected from 86 natural sites to assess background levels. Five heavy metals were investigated-arsenic, cadmium, copper, lead and zinc (see Table 8.14). Only one of the five, cadmium, was present at elevated levels, though the concentrations were well below the level identified as warranting further investigation (ANZECC/NHMRC, 1992).

At high concentrations, cadmium is a particularly toxic heavy metal, accumulating in the body's tissues and organs, particularly the liver and kidney. Cadmium levels recorded to date do not appear to represent a threat to soil or animal health, but their accumulation in the kidneys and livers of sheep and cattle, particularly older animals, may reach levels exceeding tolerance limits for export markets. As a simple and practical precaution, offal from animals older than 2.5 years is not exported nor sold in New Zealand.

In New Zealand, the cadmium concentration in pasture soils is associated with the use of superphosphate fertiliser. Superphosphate is made from phosphate rock much of which is imported from the island of Nauru. It has natural cadmium concentrations ranging up to 100 milligrams per kilogram (mg/kg), and averaging around 48 mg/kg (Rothbaum et al., 1986; Taylor, 1994). As a consequence, all of New Zealand's agricultural land has received cadmium in direct proportion to the amount of fertiliser applied.

Cadmium has been found to accumulate less in sedimentary soils and more in ash and pumice soils (which tend to bind strongly with phosphorus and therefore have a higher requirement for phosphatic fertilisers). On this basis, much of the central North Island, including Taranaki, Waikato, Bay of Plenty and parts of Gisborne and Hawkes Bay, as well as parts of Northland and Southland would be expected to show elevated levels.

A survey of cadmium accumulation in crops was conducted in South Auckland's prime market garden belt and on mid Canterbury wheat farms (Roberts et al., 1995). It found that cadmium levels were very low. For example, lettuce, potatoes and onions had less than 5 percent of the permissible level and wheat had only 7 percent. This corroborates Ministry of Health findings that the levels of cadmium in the average New Zealand diet are well below the maximum intake recommended by world health authorities and appear to be declining (van Oort et al., 1995b). For females, the cadmium intake was 24 percent of the tolerable level and, for males, it was 29 percent.

At present, the cadmium situation in New Zealand appears to be well managed and measures are in place to ensure it stays that way. Besides the Ministry of Health's food monitoring programme and the withholding of older offal from sale, the members of the fertiliser industry group are also taking steps. The fertiliser industry initially decided to progressively reduce the level of cadmium in fertiliser by a third from the year 2000, but recent technical advances have enabled the reduction programme to get underway in 1997 (Furness, 1996a, 1996b).

Another potential source of heavy metal contamination in rural soils is the use of copper and arsenic compounds to control fungi in orchards and vineyards. Orchards have not been systematically surveyed, but a study in Central Otago found that average concentrations in old orchard soils were above the draft New Zealand guideline of 100 milligrams per kilogram (mg/kg) (Ministry for the Environment and Department of Health, 1992; Morgan and Bowden, 1993). The copper levels were much higher in the orchard soils than in nearby pasture soils (106 mg/kg, compared to 20-30 mg/kg).

In many countries, the major source of heavy metal contamination in agricultural soils is sewage sludge. Sludge spreading is a recent practice in New Zealand. Department of Health guidelines were published in 1992 to control the composition of sludge for land application and prescribe the amounts and method of application. New Plymouth seems to be the only place in New Zealand where sewage sludge is applied to agricultural land, though in some other districts it has been spread, on a trial basis, under planted pine forests.

Table 8.14: Average concentrations of heavy metals in New Zealand rural soils (mg/kg).
Sites Arsenic Cadmium Copper Lead Zinc
Pastoral soil 4.9 0.44 17.7 11.7 68
Natural soil 4.3 0.19 17.0 13.3 65
ANZECC guideline1 20.0-100.0 30.00-20.00 60.0 300.0 200
NZ draft guideline2 10.0 - 30.0-100.0 - -

Source: Roberts et al. (1995)

The state of the vegetation cover

The dramatic vegetation changes of the past 150 years have permanently altered New Zealand's landscape and ecology. Most of the natural landcover and all the exotic landcover are ecologically unstable (that is, unable to maintain their current mix of plant and animal species without human intervention). A degree of 'cultural' stability has been imposed on much of this landcover through measures such as pest and weed control, soil conservation programmes, crop and pasture management, ecosystem protection, and species recovery programmes (O'Connor, 1973). But many areas remain neither ecologically nor culturally stable.

The state of our indigenous forests

Indigenous forests are now largely confined to remote mountainous areas or to lowland fragments. About 60 percent of the surviving indigenous forest is on land with no agricultural potential, approximately 30 percent is in hill country and rangelands with very limited potential, and less than 10 percent is in lowlying areas with good agricultural potential.

The largest remaining stretches in the North Island are in the Kaimai, Urewera and Raukumara ranges, inland Taranaki, the King Country, Mount Taranaki/Egmont, and the Ruahine, Tararua and Rimutaka Ranges. The remaining South Island forests are mainly in the Marlborough Sounds, Westland, Fiordland, the Catlins and Stewart Island.

The forests most affected by deforestation were the kauri and the lowland podocarp hardwoods. The kauri forests were reduced from a preEuropean area of around 1.5 million hectares to just 7,000 hectares of mature forest (a 99.5 percent reduction) with a further 60,000 hectares left to regenerate after logging (Froude et al., 1985). The lowland podocarp hardwood forests were reduced by about 85 percent. Most of the 15 percent that remain are on private land, and only a small proportion are protected. Kahikatea forests were especially affected. Only 2 percent of these wetland podocarps are believed to remain. They were made into scentfree butter boxes for export while their wetland habitats were drained to create more pasture and ultimately more butter.

In general, beech forests were less affected because of their remoter locations and lower timber quality. Beech trees yield less timber than softwoods and are often riddled with small holes created by the native pinhole borer beetle which lives inside them. Upland beech forests were left virtually intact, though lowland beech forests in many areas were cleared for farmland. Chipmills accelerated this clearing process in Southland and Nelson during the 1980s and early 1990s, but this has reduced considerably since the introduction of export control regulations and the passing of the Forests Amendment Act in 1993.

Virtually all of the estimated 4.9 million hectares of publicly owned indigenous forest are now administered by the Department of Conservation. Only 164,000 hectares are not administered by the Department and are available for timber production-152,000 hectares on the West Coast and 12,000 hectares in Southland. The West Coast production forests contain 102,000 hectares of mature loggable forest, 41,000 hectares of regenerating cutover forest, and 9,000 hectares of cleared forest which is being replaced with exotic Tasmanian blackwoods.

The area of private indigenous forest is estimated to be about 1.32 million hectares (Ministry of Forestry, 1988, 1994a). Half of this is on steep land which is not considered loggable for soil and water conservation reasons. A further 40 percent is regenerating from previous logging. Only 124,000 hectares (9 percent) carries mature, harvestable, trees, and even this may be an overestimate because a number of forest owners have recently entered voluntary protection arrangements funded by the Government's Forest Heritage Fund and Nga Whenua Rahui.

A further constraint on logging is the fact that, although private forest owners are still free to clear their forests for farmland or other land uses, logging for timber production is now subject to the indigenous forest provisions of the Forests Act 1949, as amended by the Forests Amendment Act 1993. These require most private timber production after 1996 to be from forests with an approved sustainable management plan or permit. Regenerating forest scrub is still being cleared for dairy production in some areas, but logging and burning are now relatively minor threats to the remaining indigenous forest.

Far greater are the impacts of the browsing mammals, possums, goats and deer, and also invasive plants, such as old man's beard. These attack the forests from within, changing their composition and reducing their biodiversity. This has caused canopy dieback, forest collapse and regeneration failure in increasingly large tracts of native forest (Rogers, 1995).

The few remaining areas of lowland forest also face an additional threat-fragmentation. Having become ecologically isolated as 'forest islands' in a sea of farmland these areas have no adjacent habitat from which to replenish lost populations or species. Their vulnerability is enhanced by the 'edge' effect which increases their exposure to invasive pests and weeds. Because the perimeter, or edge, is relatively large compared to the small habitat area within, exotic species can invade and penetrate fragmented ecosystems more easily than they can invade larger intact stands. The inexorable result is a decline in biodiversity within each forest fragment (Diamond, 1984; East and Williams, 1984; Kruess and Tscharntke, 1994).

The state of our tussock grasslands

The forest fires which raged between 400 and 700 years ago, allowed the tussock to spread from around 1.5 million hectares to almost eight million hectares, or 30 percent of the land surface (see Figure 8.3). While this enormous landscape change had a dramatic impact on forests, soils and biodiversity, its positive feature was that it allowed an expansion in the previously restricted populations of alpine plants and insects.

Most of this expansion occurred in the South Island high country, where tussock replaced mixed podocarp forests, comprised of matai, totara and kahikatea, and open dry forests, containing a great deal of lacebark, kowhai, kanuka and scrub. In the interior, mountain toatoa, totara and scrub species may have been the major vegetation in drier areas. In the North Island the change was most marked on the central plateau which, today, is traversed by the Desert Road and is home to the Army and a thousand or so wild horses. After repeated burn-offs, the forest cover surrounding the volcanoes and Lake Taupo was replaced by 660,000 hectares of tussock grassland.

Since European settlement, the total area of tussock land has decreased towards its original size, though the precise extent of this is partly a matter of definition (see Box 8.7). What can be stated with some certainty is that very little unmodified tussock land exists, as virtually all of it has been burnt many times, and most has been grazed as well. Nearly all the remaining tussock is in the South Island high country, stretching from Southland to Marlborough. The once widespread red tussock grasslands of Southland have all but disappeared, leaving only a few small remnants. The North Island has only about 150,000 hectares of tussock, half of it red tussock in the central plateau.

The reduction in tussock area has been accompanied by a much greater decline in tussock biomass. The attempt to create short pasture for sheep has led to short tussock replacing tall tussock in many areas, followed by a decline in short tussock growth and abundance due to soil degradation, oversowing with other grasses, and the invasion of the stressresistant herbs, scabweed and hawkweed (see Box 8.8).

Tall tussock grasslands in an undisturbed state are naturally resistant to hawkweed invasion, but not to fire and grazing (Kerr, 1991). Hawkweed is not restricted to semi-arid areas. Some of the worst affected tussock lands are in relatively moist range lands with more than one metre of rainfall (e.g. the Clarence Valley in Marlborough). Again, this is land which in the past has been subject to repeated burning and heavy grazing pressure from sheep and rabbits (see Figure 8.10).

Although no native plants are known to have become extinct in the South Island tussock lands, many of the herbs, flowers and fine grasses normally associated with tall tussock have been displaced by short tussock and hawkweeds (Treskonova, 1991). Some species are on the brink of extinction and others are confined to very restricted locations (Working Party on Sustainable Land Management, 1994).

A number of tussock-dwelling animals have disappeared, such as the weka, several reptiles and several invertebrates. The spread of the exotic grass, browntop, in the area near Cass has reduced the amount of tussock and herb cover and caused a dramatic fall in the populations of many native moths (White, 1991). Although no moth species has yet disappeared, extinctions seem inevitable if the exotic grasses continue to spread at the tussock's expense.

In the North Island's central plateau, at least one native plant (Logania depressa) has become extinct. The red tussock itself continues to retreat under pressure from wild horses, military exercises, rabbits and a host of invasive exotic plants, including Hieracium, wilding lodgepole pines (Pinus contorta) and heather (Calluna vulgaris) (Newsome, 1987; Rogers, 1994). The heather was introduced from Europe and deliberately sown in the Tongariro National Park from 1912 to 1921 with the aim of 'beautifying' the tussock and creating a suitable habitat for introduced game birds. Today it is a serious nuisance, its spread having been aided by fire and other forms of vegetation disturbance, particularly near roads, tracks, and in the military area off the Desert Road (Moller et al., 1993).

Box 8.7: How much tussock land?

Although tussock grasslands have clearly declined in the past 150 years, there is some confusion over how much remains today. Some writers give the impression that the surviving tussock lands are limited to the 2.45 million hectares of leased pastoral land in the South Island high country. One widely quoted estimate puts the area at 2.75 million hectares (McSweeney and Molloy, 1984). Other writers seem to suggest that tussock predominates throughout the entire high country, an area spanning some 6.3 million hectares (Working Party on Sustainable Land Management, 1994).

The main database for assessing the national area of tussock land is the New Zealand Land Resource Inventory 197579 (NZLRI)(Hunter and Blaschke, 1986) whose vegetation data was updated and mapped in The Vegetative Cover of New Zealand (Newsome, 1987). A separate assessment of the North Island tussock grasslands was recently undertaken by Rogers (1994). According to the NZLRI, tussock is the 'major' vegetation cover over some 5 million hectares, and a 'minor' part of the vegetation over some 3 million hectares (Hunter and Blaschke, 1986). The vegetation cover map tells a more detailed story. Tussock is listed as a component of at least 10 different vegetation groupings spanning a total of 10 million hectares. Actual tussock grassland makes up a third of this, while the other vegetation categories include tussock in association with other species. In many cases, the tussock has a minor or intermittent presence. The main categories and approximate areas can be grouped as follows:

  • 3.3 million hectares of 'predominant' tussock grassland in which short or tall tussocks are the largest and most abundant plant cover (though significant areas are now hawkweeddominated or interspersed with exotic grasses);
  • 1 million hectares of tussock and subalpine scrub;
  • 900,000 hectares of 'unimproved' pasture in which short tussocks occur in mixed communities of exotic grasses and weeds;
  • 110,000 hectares of mixed forest and tussock; and
  • 4.7 million hectares of intermittent or scattered scrub (e.g. matagouri, sweet brier, manuka and kanuka, gorse, Dracophyllum and Cassinia) in which tussock is sometimes a component.

Nearly all of this area has been modified by burning, grazing and related land pressures. Less than 80,000 hectares, barely 2 percent of the 'predominant' tussock grasslands, are in the care of the Department of Conservation. The Department also administers nearly 650,000 hectares of mixed subalpine tussock and scrub and is acquiring additional small blocks of tussock through consultation with individual land-owners.

Figure 8.10: The South Island tussock lands and main areas of hawkweed infestation.
Textual description of figure 8.10

Tussock land is spread widely throughout the South Island high country and in a small pocket around the North Island central volcanic plateau. In the South Island hawkweed is present mainly on the eastern side of the ranges, inland from Dunedin and all the way north. Hawkweed cover is in pockets, particularly inland from Dunedin, Timaru and Kaikoura.

Sources: Newsome (1986), Parliamentary Commissioner for the Environment (1991)

The central plateau's 660,000 hectares of tussock were nearly halved to 310,000 hectares between 1840 and 1940. Sheep and rabbits had much the same effect as in the South Island tussock lands. Rabbit plagues in the 1920s and 1930s, combined with burning, caused substantial deterioration. This, and also large sheep losses to feral dogs, led to the abandonment of sheep farming in much of the area between 1917 and 1939 (Rogers, 1994).

In the years since then, recurrent burning of tussock for shrub suppression, fodder improvement, and control of wilding pines, plus feral horse grazing and accidental fires caused by military activities, have led to most of the remaining red tussock being replaced by exotic grasses and weeds (e.g. hawkweed). Today, 64,000 hectares survive-10 percent of the North Island's pre-European tussock area-of which about 7,000 hectares are protected (Rogers, 1994).

Box 8.8: The razing and grazing of the native grasslands

After quintupling in area during the period of Māori forest burning, New Zealand's tussock grasslands have been in retreat since the 1850s when European settlers arrived with their merino sheep. The new grazing and burning regime killed off the snow tussock over large areas and allowed the more palatable short tussock to spread above the 900 metre contour. Subsequent oversowing of exotic grasses and invasion by rabbits and and weeds reduced the short tussock as well.

Today, nearly four million hectares of tussock country are classified as pastoral land, though the actual area used for grazing may be considerably less than this. In the 1970s, 2.85 million hectares was leased to 475 runholders by the Office of Crown Lands. By 1990, the number had fallen to 349. Today, approximately 360 runholders lease 2.45 million hectares, with nearly half a million hectares having been 'retired'. Where soil and moisture levels have permitted, about 20 percent of the remaining pastoral tussock has been 'improved' by introduced grasses and regular fertiliser applications. The rest has experienced soil degradation and competition from groundhugging weeds (e.g. scabweeds in the 1930s, Nasella grass and hawkweeds in more recent decades).

Although rabbits and hawkweeds have been named as the cause of the tussock land's decline, scientists now consider that these species are symptoms or, at worst, secondary causes, of land degradation. They have generally become established in areas where overgrazing, burning, and insufficient fertiliser application have so depleted the soil and the tussock cover that only rabbits and hawkweeds, which originally evolved in semiarid conditions, can thrive (Belton, 1991; Kerr, 1991; O'Connor and Harris, 1991; Commission for the Environment, 1991; Working Party on Sustainable Land Management, 1994).

Repeated rabbit infestations and declining productivity have occurred in the tussock pasture lands for over a century. Between 1877 and 1881, for example, 77 Otago sheep runs, spanning more than 600,000 hectares had to be abandoned. Further bad periods occurred in the 1890s, after stock units peaked at around 2 million, and the 1940s, when they had declined to one million. An apparent turn around began in the 1950s as government funding and the new technology of aerial topdressing made it possible to 'improve' pasture by dropping large quantities of fertiliser, grass seed, and poisoned rabbit bait over wide areas. The resulting increase in soil fertility encouraged runholders to increase their stock numbers once more. Stock units reached an unprecedented 2.5 million by the late 1980s. But rabbits and hawkweeds were also on the rise.

It now seems that, in spite of the apparent recovery, the long term trend for the tussock grasslands is one of inexorable decline in both species diversity and production (Kerr, 1991). Of the 3.9 million hectares currently classed as pastoral tussock land, at least two million hectares have been modified by either pasture improvement or soil degradation. These include at least half a million hectares oversown with introduced grasses (Working Party on Sustainable Land Management, 1994), a further half million hectares dominated by hawkweed or bare soil, and 1 million hectares which have a significant and increasing presence of hawkweed (Belton 1991; Kerr, 1991). In the worst affected areas, covering about 200,000 hectares, desertification threatens, with hawkweed and bare soil completely dominating the ground, allowing wind and frost to produce high rates of soil erosion (Belton, 1991; Kerr, 1991).

Positive changes in the high country between 1989 and 1994 included a 9 percent reduction in stock units-50 percent on some farms an expansion of farmers' selfhelp groups (e.g. the 'landcare' programme), and a five year Rabbit Land Management Programme which reduced the percentage of bare ground in a 300,000 hectare area from more than 50 percent to around 30 percent. Such developments, however, are seen by many scientists as insufficient to sustain livestock grazing in much of the South Island tussock land (e.g. Kerr, 1991; Working Party on Sustainable Land Management, 1994). Alternative land uses now being considered are exotic forestry and nature conservation.

The state of our dunelands

Few dunelands bear much resemblance to their original state because of the impacts of stock grazing, introduced plants, sand mining for iron and silicon, and dune-control planting (Newsome, 1987). The original foredunes were home for the native sand-binding plants, pingao and spinifex, and covered a total area of perhaps 60,000 hectares (King, 1984). The backdunes were covered in native scrub and forest.

Today, some 52,000 hectares of foredunes still carry sand-binding grasses and sedgesbut, in most cases, it is the exotic marram grass which dominates. Some 40,000 hectares of backdunes are dominated by exotic lupins and shrubs and, beyond these, more than 200,000 hectares of backdunes are covered in pasture grasses, pine trees, gorse and other exotic species (Hunter and Blaschke, 1986; Newsome, 1987). Of the 305,000 hectares of coastal sand dunes identified in the New Zealand Land Resource Inventory 197579, probably less than 10 percent are still close to their original condition.

Dune control planting occurred early this century after the natural dune vegetation had been destroyed by fire, deforestation, grazing and physical disturbance. The destabilised sand began drifting inland smothering grass on land that had been converted to pasture. Concern about the problem led to the SandDrift Act of 1908. To stop the sand drifts, marram grass was introduced. The marram was more prolific than the native sandbinders, but needed more nitrogen than the dunes could naturally provide, so lupins were also planted.

Lupins, like clover, peas and beans, are among the world's 170,000 species of legumes. Their roots contain bacteria which can absorb nitrogen from the air and convert it into organic forms (e.g. ammonium and nitrates). This organic nitrogen can then be absorbed from the soil by other plants. In the same way that peas are planted to provide nitrogen for other vegetable crops, and clover is planted to sustain the nation's pasture grasses, lupins were planted to sustain the marram grass. For most of this century, the stability of most New Zealand sand dunes has depended on this marram-lupin combination.

A national survey of New Zealand's sand dune and beach vegetation was carried out between 1984 and 1988 by the Botany Division of the former Department of Scientific and Industrial Research (DSIR) (Partridge, 1992; Johnson, 1992). Many of the nation's beaches were visited-608 in total (273 in the North Island, 302 in the South Island, 27 on Stewart Island and six on nearby Ruapuke Island). Despite the high number of beaches in the south, most of the southern dunelands are comparatively small in area. In fact, 80 percent of New Zealand's sand dunes are in the North Island-principally on the west coast and north of the island.

The dunelands were ranked according to four criteria: the diversity of their landforms; the extent to which native sand plants were present; the extent to which the dunes and vegetation had escaped modification; and the extent to which they were free of adventive weeds. Total rankings gave an indication of those areas with the greatest botanical value for conservation.

Unfortunately, the ranking system tended to focus more on the dunes themselves rather than the inter-dune slacks, swamps and lakes which harbour much of the dunelands' biodiversity. Furthermore, a number of dunelands were not surveyed while, for others, the rapid survey approach did not do justice to their complexity and missed many rare species (Rapson, 1996). As a result, the survey probably understates the ecological importance of many dunelands.

Of the 608 beaches, about one third were considered worthy of conservation management, and 53 (9 percent) were identified as dunelands of national importance (Johnson, 1992; Partridge, 1992). A quarter (13) of the top ranking dunelands were in Northland whose far north beaches have the largest assemblage of botanically valuable dune systems in the North Island. The Northland dune lakes shelter threatened species of plants and native fish.

The best of the Northland dunelands is Spirits Bay, an almost pristine sequence of foredune and backdunes grading behind into wetland and forest vegetation. The slack communities which occupy the moist hollows between the dunes are in relatively good condition. All the important native dune species are present and among the common weeds, marram and kikuyu are apparently absent with just a small amount of scattered lupin (Partridge, 1992).

The best of the southern dunelands are in the extreme south, at Stewart Island where nine out of 27 dunelands (33 percent) were rated as being of national importance. They include the near pristine Smoky Bay, a dramatic sweep of beach, native dune, and forest, of high biological and scenic value, and also the 13 kilometre Mason Bay which is outstanding on several counts: its size, the dramatic appearance of its dune landscapes, the diversity of habitats, the rich and largely unmodified plant life, and the presence of brown kiwis and other birdlife. However, marram now dominates Mason Bay's foredunes and grazing by sheep, white tailed deer, red deer, and possums, affects the dune vegetation to some extent (Johnson, 1992).

In the majority of the surveyed dunelands the native pingao and spinifex are dominated by marram grass, usually in association with lupins. Another dune stabilising plant, lyme grass (Leymus arenarius) has also been planted in some dunes, as have radiata pine plantations. Many dune areas have been burnt and some have been invaded by large stands of bracken fern (Pteridium esculentum). Grazing by cattle, sheep, rabbits and hares has also reduced or removed the native vegetation. Direct human pressures have come in the form of housing development, protection works such as sea walls, sand mining, and heavy pedestrian and vehicle traffic, including dune buggies and trail bikes (Partridge, 1992; Hamilton, 1996). The wetlands and slacks between the dunes have been particularly hard hit by the combination of introduced species, pollution, and the drainage of adjacent farmland.

In recent years, two additional pressures have come to bear on the dunelands-the rapid spread of the introduced pampas grass ( Cortaderia selloana), particularly in dunelands throughout the North Island; and the equally rapid spread of lupin blight, caused by the fungus Colletotrichum gloeosporoides (Partridge, 1992). Since 1987 this disease has spread to many areas, reducing lupin populations on open dunes by 90 percent (Douglas et al., 1994). The consequent reduction in nitrogen threatened to kill off the marram and lyme grass and again destabilise the dunes, but it now seems that the lupin populations are recovering and are likely to resume their former abundance.

The state of our exotic grasslands

Well fertilised, oversown, exotic grasses form a dense underground mat of roots that can improve soil structure by increasing the amount of organic matter or humus in it (as roots die and new ones replace them) and by creating root channels that aerate the soil and make it porous. However, exotic grasses also have higher nutrient and moisture requirements than the indigenous grasses, are less tolerant of soil acidity, and have shallower roots.

Without proper management, the exotic pasture grasses are vulnerable to soil degradation, moisture deficits, pests and weeds and soil erosion (Williams and Haynes, 1991; Wedderburn, 1991). In hill country the shallow roots of the exotic grasses are less effective than tussock or forest at retaining soil and water, though 'improved' grassland, with its dense root structure, is far less erosion-prone than unimproved exotic grassland with its sparser roots and vegetation cover.

At present, around two-thirds of New Zealand's farmland (68 percent) is on erosionsusceptible soils, and most of this is exotic pasture in the North Island (Clough and Hicks, 1993). Under current management practices, an estimated 6.1 million hectares of pasture land (53 percent of the entire North Island) is not sustainable because of erosion risk (Eyles, 1993). However, about 3.7 million hectares could be made sustainable by space-planting or close-planting of trees to provide stronger root support for the soil. For the other 2.4 million hectares, however, the only sustainable option is to convert to forest (Eyles, 1993).

Apart from soil degradation, the introduced pasture grasses are also vulnerable to pests and weeds, including native species attempting to reassert themselves. Regenerating native forest is a constant problem for exotic grassland in many areas. It is usually suppressed by grazing and repeated fertiliser applications on well-stocked improved pastures, and by periodic burning on the more sparsely stocked unimproved pastures, aided by the clearing of native seedlings when they start to grow.

Sustaining pasture grasses in many areas, therefore, means maintaining soil condition and moisture availability, and combating pests. Because inputs of fertiliser, lime, irrigation water and pesticides are costly, it is not unusual for pasture condition to deteriorate in times of economic difficulty and improve in more buoyant times. During the 1980s, farmers cut back on fertiliser and lime use to save costs. For most, this caused no immediate loss of pasture condition because enough residual nutrients had accumulated from the previous decades of regular fertiliser application. But in some hill systems, deterioration became apparent.

Although fertiliser use has picked up in recent years, much of this has been increased use of nitrogen fertilisers in dairying areas rather than the replenishment of hill soils. However, on some hill farms where fertiliser applications had been lighter anyway, and nutrient reserves lower, pasture yield has dropped by as much as half over five to ten years, and a gradual reversion to their 1940s condition is already underway (Hicks, 1995).

The state of our croplands

Apart from soil problems, the main threats to crop viability in New Zealand are from pests and weeds, which are controlled by both biological and chemical methods. Lack of moisture and intermittent drought is also a threat in some areas and is controlled by irrigation. At present, the area of horticultural land is expanding, though the overall crop and horticultural area remains relatively small.

Up to 14 percent of New Zealand could support crops (Eyles and Newsome, 1991). At present, less than 2 percent is actually used for this purpose, the rest being used for animal production, such as dairy farming. In recent decades, there has been an intensification of harvesting on some of the crop lands, leading to soil problems, such as compaction, nutrient decline and falling productivity in some areas.

The state of our exotic forests

After slumping in the late 1980s and early 1990s, the rate of new forest planting soared to a record 98,000 hectares in 1994 (Eyre, 1995). This led to planting predictions of 100,000 hectares per annum for the rest of the decade, but more sober analysis by the Forest Research Institute suggests that the rate of new planting until the year 2000 will average around 72,000 hectares per year, falling to 57,000 hectares per year for the next decade (Glass, 1996).

The biggest environmental threats to exotic forests are wind and fire. On hills, erosion is also a risk for a period after harvesting and before new root systems have become firmly established. Other forms of soil degradation, such as nutrient loss and acidification, may limit these forests' sustainability in poor soils, such as sand dunes (Hawke and O'Connor, 1993; Smith et al., 1994). In most areas, however, soil degradation is not believed to pose a serious limit to forest growth and productivity, at least during the first two or three harvest cycles (O'Loughlin, 1995). In richer soils, radiata pine can actually enhance nutrient levels (Davis and Lang, 1991; Hawke and O'Connor, 1993; Smith, 1994).

Because most exotic forests consist of a single dominant species, the impacts of a serious disease or insect invasion (e.g. Asian gypsy moth) could be devastating. On the evidence to date, however, radiata pine appears to have a low susceptibility to disease while vigilant border and pest control operations have, so far, contained Asian gypsy moth arrivals. The insect and fungal pests which are already established incur economic costs but do not threaten the plantations as such. They are controlled through integrated pest management systems combining biocontrols and chemical sprays.

Outside the plantation forests, the vigour and fecundity of pine trees makes them difficult to contain. In some areas they have become aggressive weeds. Wilding pine trees have become established in parts of the tussock country of the South Island and in the North Island's Central Plateau. They have also invaded sand dunes and indigenous forests. At present, the problem is localised but increasing, with some 17,000 hectares affected in the Canterbury Region for example (Canterbury Regional Council, 1995). Despite these impacts, however, the future of exotic forests in New Zealand looks very positive.