2 Groundwater quality status 1995–2006
2.1 Categorisation of monitoring sites based on groundwater quality status
Daughney and Reeves (2005) developed a hierarchical classification scheme for the NGMP in order to simplify the interpretation of groundwater quality data at the national and regional scale (see Figure 1). This classification scheme assigns each monitoring site to a category and sub-category on the basis of measured concentrations of 15 water quality parameters (see Table 5). Each site is assigned to one of two main categories, where category 1 includes all oxidised (oxygen-rich) groundwaters and category 2 includes all reduced (oxygen-poor) groundwaters.
At the second level of classification, category 1 is divided into two subcategories. Subcategory 1A includes all oxidised groundwaters that show evidence of human influence in the form of elevated NO3-N, K, SO4 and/or Cl, whereas subcategory 1B represents oxidised groundwaters that show little or no such impact. Note that human activities can influence groundwater chemistry in other ways (aside from the elevation of NO3-N), some of which might occur in reduced groundwater. However, any other styles of human influence are not evident in the NGMP data set and therefore are not accounted for in the categorisation scheme developed by Daughney and Reeves (2005). Note also that subcategory 1B groundwaters may be recharged primarily from rivers, whereas subcategory 1A groundwaters may have a higher proportion of recharge from rain (which allows accumulation of salts and NO3-N during passage through the soil zone).
Categorical assignments are based only on groundwater chemistry and do not specifically consider factors such as well depth, aquifer confinement or lithology. However, Daughney and Reeves (2005) showed that these factors are usually related to the categorical assignments. For example, at the third level of classification, six subcategories are defined. These subcategories reflect differences in water type in subcategories 1A and 1B, possibly associated with aquifers of different lithology, and differences in extent of reduction in category 2, possibly associated with degree of aquifer confinement.
Figure 1: General characteristics of status-based groundwater quality categories defined by Daughney and Reeves (2005)
One major limitation of this classification scheme is that it does not include a category for groundwater with marine character (eg, high salinity), simply because the NGMP includes few coastal monitoring sites that are influenced by salt water intrusion. The development of a new national scale classification scheme is beyond the scope of this investigation, and so due to the lack of a dedicated category, groundwaters with low, moderate, high and very high marine chemical signature are assigned to subcategories 1A-1, 1A-2, 2A and 2B, respectively. In the following discussion we identify cases where this limitation might be important.
Despite any limitations, the hierarchical classification scheme provides a useful way of summarising the status of groundwater quality across New Zealand. Table 6 lists the total number of monitoring sites that could be categorised as part of this investigation, and the percentage that fall into each groundwater quality category (site-specific categorical assignments are given in Spreadsheet 1). In some regions, such as the West Coast and Taranaki, relatively few SOE monitoring sites could be categorised, and so the conclusions made below may not apply to the region as a whole. Figures 2 and 3 display the site-specific categorical assignments in map form, to clearly highlight important regional differences in groundwater quality. As stated above, the categorical assignments are based only on groundwater quality and do not specifically account for factors such as well depth or aquifer confinement. However, there are logical relationships between these factors and groundwater quality categories (see section 4).
Table 5: Parameters used to assign monitoring sites to the status-based groundwater quality categories defined by Daughney and Reeves (2005), and their median concentrations in each category, based on the NGMP data set
Table 6: Percentage of SOE monitoring sites in each region assigned to each status-based groundwater quality category, relative to the total number (n) of sites for which a categorical assignment could be made
|
Council |
1A-1 |
1A-2 |
1B-1 |
1B-2 |
2A |
2B |
|---|---|---|---|---|---|---|
|
ARC |
4.2% |
20.8% |
0.0% |
4.2% |
33.3% |
37.5% |
|
EBOP* |
1.9 |
11.5 |
0.0 |
38.5 |
26.9 |
21.2 |
|
ECAN |
35.6 |
16.1 |
28.5 |
13.5 |
4.9 |
1.5 |
|
ES |
14.0 |
53.5 |
2.3 |
18.6 |
7.0 |
4.7 |
|
EW |
4.9 |
53.4 |
4.9 |
29.1 |
7.8 |
0.0 |
|
GDC |
5.3 |
1.3 |
0.0 |
0.0 |
25.0 |
68.4 |
|
GWRC |
10.0 |
24.3 |
1.4 |
24.3 |
18.6 |
21.4 |
|
HBRC |
8.0 |
10.0 |
10.0 |
14.0 |
38.0 |
20.0 |
|
MDC |
4.3 |
0.0 |
39.1 |
26.1 |
30.4 |
0.0 |
|
MWRC |
3.2 |
16.1 |
0.0 |
16.1 |
29.0 |
35.5 |
|
NRC* |
14.6 |
31.7 |
0.0 |
7.3 |
36.6 |
9.8 |
|
ORC |
23.0 |
14.9 |
21.8 |
10.3 |
24.1 |
5.7 |
|
TDC* |
43.8 |
25.0 |
12.5 |
0.0 |
18.8 |
0.0 |
|
TRC |
0.0 |
14.3 |
0.0 |
14.3 |
28.6 |
42.9 |
|
WCRC |
0.0 |
0.0 |
25.0 |
75.0 |
0.0 |
0.0 |
|
n |
158 |
191 |
120 |
149 |
154 |
126 |
|
%n |
17.6 |
21.2 |
13.4 |
16.6 |
17.1 |
14.0 |
* Regions in which categorical assignments may be biased by groundwater under marine influence.
Figure 2: Assignment of North Island SOE monitoring sites to status-based groundwater quality categories
Figure 3: Assignment of South Island SOE monitoring sites to status-based groundwater quality categories
Across New Zealand, for all monitoring sites for which a categorical assignment could be made (n = 898), 69% fall into category 1, indicating that the groundwater is oxidised, whereas the remaining 31% of sites fall into category 2, indicating that the groundwater is oxygen-poor (Figure 4).
As stated above, sites in category 1 are potentially at risk for elevated concentrations of NO3-N, which persists in oxidised groundwater, but the sites in category 2 are not, because NO3-N would tend to be removed by the natural process of microbial denitrification. On the other hand, sites in category 2 are potentially at risk for elevated TDS and high concentrations of Fe, Mn, NH4-N and/or As, but the sites in category 1 are not.
Figures 2 to 4 indicate that the SOE monitoring sites in several regions of the country display a predominance of oxygen-poor (category 2) groundwaters, notably in Gisborne, Auckland and Manawatu-Wanganui, but also western Northland, coastal Bay of Plenty, northern Hawke’s Bay, south Wairarapa, and some parts of Otago and Southland (in the latter case in relation to lignite deposits). Conversely, monitoring sites in several parts of the country reveal a predominance of oxygen-rich groundwaters, including Waikato, northern Wairarapa, Tasman, Canterbury and the West Coast. Thus the particular groundwater quality risks associated with oxidised versus reduced aquifers are to some extent region-specific.
Figure 4: Proportion of monitoring sites considered in this investigation that fall into category 1 (oxidised groundwater potentially at risk for elevated NO3-N)
The categorisation scheme indicates that over half the monitoring sites that are at risk for elevated NO3-N currently show some evidence of human influence. In other words, across New Zealand, of all of the sites assigned to category 1, just over half (56%) fall into subcategory 1A (39% of sites overall) (see Figure 5). The groundwater quality at sites in subcategory 1A displays evidence of human and/or agricultural impact in the form of elevated concentrations of NO3-N, K, Cl and/or SO4. The remaining sites in category 1 fall into subcategory 1B (30% of sites overall), indicating that they show little or no evidence of anthropogenic impact.
Of SOE sites that are at risk of elevated NO3-N, the highest proportions that currently show human influence are found in Auckland (86%), Southland (76%) and Waikato (63%). Note that the categorisation scheme used here may mistakenly assign groundwater with marine character to subcategory 1A (see above), and so the proportions for regions such as Northland (86%), Tasman (85%) and the Bay of Plenty may be about 10% too high.
The categorisation suggests that all of the oxidised groundwaters in Gisborne are already affected by human activity, but this conclusion is based on just five monitoring sites and so may not be representative of the region as a whole. The regions that show the lowest proportions of affected sites are West Coast (0%), Marlborough (6%) and Bay of Plenty (26%). Overall, of all sites that might potentially show human impact, this analysis shows that there are clear regional differences in the proportion of monitoring sites that currently show such impact (subcategory 1A) compared to those that do not (subcategory 1B).
Figure 5: Proportion of category 1 monitoring sites considered in this investigation that fall into subcategory 1A (oxidised groundwater that currently shows some level of human or agricultural impact)
The categorical assignments also reveal several variations in groundwater chemistry within individual regions (see Figures 2 and 3). Some examples of the more significant sub-regional scale variations in groundwater chemistry, and hydrogeological explanations for them, are as follows.
Auckland
Many of the SOE monitoring sites in the central and northern parts of the Auckland region tap deep confined aquifers (Crowcroft and Smaill, 2001). These aquifers contain highly reduced groundwaters that are assigned to category 2. On the other hand, some monitoring sites in central and southern parts of the Auckland region tap shallower unconfined basalt aquifers. These aquifers often contain oxidised groundwaters assigned to category 1.
Manawatu–Wanganui
Little information is available about the confinement and lithology of aquifers in this region, but there are clear differences in groundwater quality in the central-northern and southern (Horowhenua) parts of this region (Bekesi and McConchie, 1999; Taylor et al, 2001). The fact that the former are almost all assigned to category 2 and the latter are in category 1 suggests a major difference in hydrogeological environment.
Wellington
Monitoring sites in the Wellington region are either on the Kapiti Coast, in the Hutt Valley, or in the Wairarapa, and because these aquifer systems are distinct they all have distinct groundwater quality (Morgan and Hughes, 2001). The Kapiti Coast aquifers are generally oxidised, and so most monitoring sites are assigned to category 1. The Hutt Valley aquifers are characterised by a progressive decrease in redox potential along the flow path (Downes, 1985), and so monitoring sites near the recharge zone are oxidised (category 1), whereas those further down-gradient are more reduced (category 2). There is a similar gradation in redox potential along the major groundwater flow path in the Wairarapa (Morgenstern, 2005), and so monitoring sites in the northern Wairarapa are usually assigned to category 1, whereas those further south are assigned to category 2.
Marlborough
Many of the monitoring sites in the Marlborough region tap into the gravel aquifers of the Wairau Plains. There is a distinct gradation in groundwater quality across the Wairau Valley, such that groundwater close to the river has chemistry essentially the same as the river itself (category 1B) and groundwater further away shows more evidence of rainfall recharge and the accompanying accumulation of salts and NO3-N (category 1A). There is also a gradation in groundwater chemistry parallel to the axis of the Wairau Valley, with groundwater in deep coastal wells usually being more reduced. The southern valleys also tend to contain clay-bound aquifers with slow-moving groundwater, and so many monitoring sites in this part of the region fall into category 2.
Canterbury
The gravel aquifers of this region have been described extensively elsewhere (see Brown, 2001). Unlike many gravel aquifers in New Zealand, the groundwater in the Canterbury aquifer system generally remains oxidised even after many years and significant travel distance along the flow path, and so most monitoring sites in the region are assigned to category 1. Groundwater recharge from rivers is important in many areas of the Canterbury Plains, which explains the dominance of subcategory 1B sites near rivers (similar to the Wairau Plains in Marlborough).
Some distance away from the rivers, a greater proportion of groundwater recharge comes from rain. During passage through the soil zone this recharge water accumulates salts and NO3-N. Thus many of the monitoring sites in between the major rivers show more evidence of human or agricultural impact and are assigned to category 1A. The region around Oamaru is characterised by limestone aquifer systems with distinctive groundwater chemistry.
2.2 Quantification of groundwater quality status at the national scale
The categorisation scheme described above provides an overview of groundwater quality in New Zealand, but it does not quantify the magnitude of any specific groundwater quality issues. For example, the categorical analysis indicates that over one-third of the monitoring sites considered in this investigation currently show evidence of human influence. But how severe is this influence for each of the parameters of concern?
This information can be provided by national medians and percentiles. In this investigation the calculated percentiles are compared to two water-quality guidelines: the Drinking-water Standards for New Zealand (DWSNZ) (Ministry of Health, 2005) and the Australia and New Zealand Environment Conservation Council (ANZECC) guidelines for fresh and marine water quality (Australia and New Zealand Environment Conservation Council, 2000). The DWSNZ defines health-related maximum allowable values (MAVs) and aesthetic guideline values (GVs) related to taste, odour or colour. The ANZECC guidelines define trigger values (TVs) based on specified protection levels for aquatic ecosystems. This study uses TVs that correspond to the 95% protection level for freshwater ecosystems. The ANZECC guidelines also define TVs for stock drinking-water, which are referred to in some sections of this report. Comparisons to both water-quality standards are performed on a per-parameter basis, to determine the number and percentage of monitoring sites at which calculated medians exceed the relevant MAVs, GVs or TVs.
National medians for most parameters (see Table 7) are very similar to previously reported values across the NGMP (Daughney and Reeves, 2005). National medians for major elements are intermediate between the global average for river water and the global average for groundwater (Turekian, 1977; Hem, 1985; Langmuir, 1997), often being closer to the former (see Figure 6).
Table 7: Calculated national percentiles and maximum values for groundwater quality parameters, with global average concentrations for river water and groundwater given for comparison
Figure 6: National percentiles for selected groundwater quality parameters, including (a) major elements, (b) minor elements and microbiological indicators, and (c) trace elements
Notes: The centre line in each box represents the median. The boxes extend from the 25th to 75th percentile, and the whiskers from the 5th to the 95th percentile. White and black circles represent the global average for river water and groundwater, respectively (Turekian, 1977; Hem, 1985). White and black stars represent the aesthetic GV and the health-related MAV, respectively (DWSNZ). Black crosses indicate the ecosystem TV (ANZECC).
The percentage of sites that exceed a threshold MAV, GV or TV is shown in Table 8. This analysis indicates that there are only a few nationally significant groundwater quality issues, including NO3-N, Fe and Mn, microbiological parameters and salinity, as discussed below.
Table 8: Percentage of monitoring sites at which median concentrations exceed water-quality guidelines
Nitrate
From a groundwater quality perspective, NO3-N, with a national median concentration of 1.3 mg/L, is the compound of most concern. For comparison, previous studies have provided estimates of 0.3-1.0 mg/L for median NO3-N concentration in unaffected groundwaters in New Zealand (Burden, 1982; Morgenstern et al, 2004; Daughney and Reeves, 2005). Of the 956 sites at which a site-specific median could be calculated, 4.9% exceed the MAV based on the DWSNZ (11.3 mg/L), 10.3% exceed the TV for ecosystem protection based on the ANZECC guidelines (7.2 mg/L), and none exceed the ANZECC TV for stock drinking-water (400 mg/L). A previous study defined a threshold value of > 1.6 mg/L (about twice the estimated background, or around quarter of the TV) as a “probable” indicator of human influence (Daughney and Reeves, 2005). The same study defined a threshold of > 3.5 mg/L (about four times the estimated background, or around half the TV) as an “almost certain” indicator of human influence.
Based on these threshold values, at 46% of all sites for which median NO3-N concentration could be determined, the evidence for human influence is “probable”, and at 31% of sites the evidence of human influence is “almost certain”. These percentages are comparable to the groundwater quality categorisation described in section 2.1, which indicated that 39% of the monitoring sites considered in this study currently show some level of human impact (category 1A).
It is important to recognise that low concentrations of NO3-N do not necessarily correspond to pristine groundwaters (subcategory 1B), but may indicate reduced groundwaters (category 2) where NO3-N has been removed by microbial respiration. In other words, a low nitrate concentration doesn’t mean that groundwater isn’t polluted or never was polluted, but could instead indicate that all evidence of the pollution has been erased by natural microbial respiration. It is instructive to compare the national median nitrate concentration (1.3 mg/L, above) to the median nitrate concentration in only oxidised groundwater (ie, where nitrate would persist if present). For all monitoring sites in category 1 (1A-1, 1A-2, 1B-1, 1B-2), the 5th, 25th, 50th, 75th and 95th percentiles are 0.14, 0.85, 2.8, 5.7 and 12.5 mg/L for NO3-N, respectively (n = 611). The sites with elevated NO3-N concentrations are found in many regions of New Zealand, especially Waikato, southern Manawatu-Wanganui (Horowhenua), Canterbury and Southland (Appendix 1).
Iron and manganese
The calculated national medians for Fe and Mn are 0.03 and 0.01 mg/L, respectively, which is in good agreement with previously reported values (Rosen, 2001; Daughney, 2003; Daughney and Reeves, 2005). Elevated concentrations of dissolved Fe and/or Mn can impart an unpleasant taste to drinking-water and can lead to staining and clogging of pipes, and so the DWSNZ includes aesthetic GVs of 0.2 and 0.04 mg/L for Fe and Mn, respectively. Due to risks to human health and freshwater ecosystems, Mn has a MAV of 0.4 mg/L and a TV of 1.9 mg/L (there is no MAV or TV for Fe).
Of the sites at which median concentrations could be determined, 27% exceeded the GVs for Fe and 33% for Mn, 15% exceeded the MAV for Mn, and 2% exceeded the TV for Mn. Note that the water-quality guidelines relate to dissolved concentrations of these elements; dissolved concentrations, especially for Fe, are often two to five times less than total concentrations (Daughney, 2003). Elevated concentrations of dissolved Fe and/or Mn generally arise from natural microbial respiration in oxygen-poor aquifers (category 2), which are found in many regions of New Zealand. Elevated total concentrations of Fe and/or Mn may be caused by this same process, but may also arise from corrosion of the well casing or reticulation system. Of the sites for which median dissolved Fe or Mn can be calculated, 16% and 31% exceeded the aesthetic GVs, respectively (data not shown). Groundwaters with high concentrations of Fe and/or Mn are found in many regions of New Zealand, especially Gisborne, Auckland and Manawatu-Wanganui, but also in western Northland, coastal Bay of Plenty, northern Hawke’s Bay, south Wairarapa, and some parts of Otago and Southland (see Appendix 1).
Microbiological parameters
The SOE groundwater monitoring programmes run by many regional councils include one or more microbiological indicator parameters. Of all sites considered in this investigation, a median value for at least one microbiological indicator could be determined at 520 sites (either E. coli, enterococci, faecal coliforms or total coliforms). The microbiological parameter with the highest median (MaxMicro) exceeds the health-related MAV (1 cfu/100 ml) at 20% of sites, and exceeds the ANZECC TV for stock drinking-water (100 cfu/100 ml) at 2% of sites.
Note that the distribution of MaxMicro is heavily skewed: the majority of results at most sites are near or below the detection limit (1 cfu/100 ml), but occasionally a much higher result is reported. The rare elevated microbiological counts might reflect a flooding event at a site with poor well-head protection, or perhaps contamination during sampling. For comparison, previous regional surveys have reported exceedance levels of 9-60% (Sinton, 2001). Note also that E. coli is the only microbial indicator parameter that is actually considered in the New Zealand drinking-water standards. All other microbial parameters listed above are considered to be proxies for E. coli. For those regions that provided microbiological indicator data for this investigation, the greatest proportions of sites that exceed the MAV are found in Northland, Southland and Canterbury (Appendix 1).
Salinity
There are aesthetic GVs under the DWSNZ for Cl, Na, SO4 and TDS, but these guidelines are related to taste. High concentrations of these parameters (and high salinity in general) usually indicate that the groundwater is older and more chemically evolved (see section 1.3). High-salinity groundwaters are found in certain regions, notably Gisborne but also in northern Hawke’s Bay, central Manawatu, south Wairarapa, Auckland (especially deep aquifers), and south Canterbury around Oamaru (see Appendix 1).
However, high salinity is not a pervasive issue in New Zealand aquifers: the relevant aesthetic GVs are exceeded at only 2-4% of all sites at which site-specific median values could be calculated for Cl, Na, SO4 and/or TDS. In some coastal aquifers high salinity arises from saltwater intrusion, which is a serious issue for groundwater resource management in regions such as Northland, Bay of Plenty and Horowhenua.
Other parameters
There are several other groundwater quality issues that can arise in selected regions or aquifers, but they are not pervasive at the national scale.
Arsenic
Arsenic concentrations in New Zealand groundwater are typically less than the health-related MAV (0.01 mg/L). However, of all sites at which median arsenic concentrations could be determined (n = 157), 10% exceeded the MAV. The hydrochemical behaviour of arsenic is governed by the behaviour of Fe and SO4. As a result, arsenic is usually only present at significant concentrations in cases where the redox potential of the groundwater falls within a narrow range. In oxidised groundwaters Fe is insoluble, and the arsenic tends to be removed from solution via co-precipitation with Fe-oxide minerals. In slightly more reduced groundwaters, as Fe is solubilised by natural microbial respiration, arsenic is also released into solution. In highly reduced groundwaters SO4 is also removed by microbial respiration. This leads to the formation of sulphide, which tends to cause the removal of arsenic from solution through the precipitation of sulphide minerals.
Trace metals
Trace metals in New Zealand groundwater are generally present at low concentrations and thus do not pose a risk to human health. It is difficult to determine whether or not trace metal concentrations in groundwater pose a serious threat to ecosystem health. This is because, for many heavy metals, the ANZECC TV is near or below the detection limit of many analytical methods. Additional research with more sensitive sampling and analytical methods may be necessary to assess the threat that trace metals in groundwater pose for New Zealand’s aquatic ecosystems.
Ammonia
Of the 924 sites at which a site-specific median could be calculated for NH4-N, 7.1% exceeded the aesthetic GV of 1.5 mg/L and 9% exceeded the TV of 0.9 mg/L. Elevated concentrations of NH4-N are only found in oxygen-poor groundwaters (category 2) and so tend to co-occur with elevated Fe and Mn, but do not co-occur with elevated concentrations of NO3-N. As for Fe and Mn, elevated concentrations of NH4-N are confined to certain regions of the country, including Gisborne, Manawatu-Wanganui, Auckland and some parts of Hawke’s Bay.
Phosphate
The DWSNZ does not define a MAV or GV for PO4-P, and the ANZECC guidelines do not define a generic TV, but the latter standard includes a TV that is specific to “slightly disturbed” ecosystems in New Zealand (for upland and lowland rivers the TV is 0.009 and 0.01 mg/L, respectively). From a groundwater perspective, 0.01 mg/L represents a relatively low concentration of PO4-P, equivalent to the 25th percentile reported in this investigation (see Table 7). Thus it is possible that the discharge of phosphate-rich groundwater into a stream or river may pose a significant threat to its ecosystem, but the number of locations in New Zealand where this is presently occurring cannot be determined with the data used in this study.
pH
The DWSNZ states that for aesthetic reasons pH should be between 7.0 and 8.5. Of all sites for which a median pH value could be determined (n = 952), 64.2% were less than 7.0 and 1.5% were above 8.5. The fact that the majority of New Zealand groundwaters have pH less than 7.0 is typical of the global situation, where most groundwaters have a pH between 6.5 and 7.5 (Langmuir, 1997). Although in some regions of New Zealand groundwater pH might be a problem for water supply, it should not be considered a pervasive environmental issue.
