• About this site|
• Site index|
• Contact us|
• Jobs

Lake Water Quality in New Zealand

Acknowledgements

Executive summary

1 Introduction

2 Methods

3 Results and discussion

4 Conclusions

References

Appendix 1: Glossary of common limnological terminology used in this report

Appendix 2: Morphological and Catchment Characteristics of the 121 Lakes

Appendix 3: Maps of Water Quality Parameters

Appendix 4: Map of Water Quality Trends (1990-2006) based on Percent Annual Change (PAC) using the Burns et al (2000) Methodology

3 Results and discussion continued

3.2 Land-cover relationships

Catchment descriptions allocated 55 lakes to the predominantly pasture class, 39 to the native class, 22 to exotic forest, and three to the alpine class. Only two lakes (Lake Pupuke and Hamilton Lake) were allocated to predominantly urban catchments, and this class was therefore excluded from figures comparing land use (Figures 5 to 8). Differences in TN, TP, Z_SD and chla between the land-cover classes (Figure 5) were highly significant. TN, TP and chla were all extremely low in the alpine class. Median values for TN, TP and chla were all lower for the native class than in the exotic and pasture classes. Although the median values of TN, TP and chla for the pasture class were similar to the exotic class, there were many more extreme high values in the pasture class than in other classes, and all of the lakes with very high concentrations of TN, TP and chla were in the pasture class. Z_SD was also affected by land use, being highest (ie, clearest water) in the native class, followed by exotic, and then pasture. (Note that there were no Z_SD data for the alpine class.)

The average TLI statistics derived from these parameters also differed between land-use classes (Figure 6). Median TLI values were highest for the pasture and exotic classes (both eutrophic), with the extreme values (most supertrophic and all hypertrophic lakes) in the pasture class. Native category lakes were predominantly oligotrophic or mesotrophic, and alpine lakes had the lowest TLI values.

Pasture lakes with high water quality (TLI < 3) were exclusively Canterbury high-country lakes, where there is little fertiliser application and low stock density. Supertrophic and hypertrophic pasture lakes almost without exception were those with very high pasture (>50%) and low native (< 20%) cover.

The inorganic nutrient (NH₄-N, NO₃-N and DRP) concentrations were excluded from the TLI methodology primarily because of the large number of very low values found for these parameters in monitoring programmes, and because they can be highly variable. This pattern was confirmed in the current data set across all land-use categories, although occasional high concentrations of inorganic nutrients were observed in both the exotic forest and pasture categories. Median values of NH₄-N were ≤ 30 mg/m³ for all classes, but values as high as 1 g/m3 occurred in a few pasture lakes (see Figure 7). NO₃-N (data not shown) was very low throughout (< 20 mg/m³ in all but a very small number of lakes). Median DRP values were low for all lakes (see Figure 7). They were similar for both native and pasture, very low (usually below detection limits) in alpine lakes, and most elevated in the exotic forest lakes. Both DRP and TP were highest in the exotic forest lakes, suggesting that either many exotic forests occur on soil with naturally high phosphorus yields, or that forestry practices are elevating P inputs to these lakes.

Notes: Horizontal lines within boxes show median values, boxes show 25–75% data ranges, whiskers show 5–95% ranges, and pluses outliers outside the 5–95% ranges.

Annual averages of concentrations of inorganic nutrients were frequently below detection limits for the lakes in both native and exotic forests, but were usually well above detection limits in the pasture class. Inorganic nutrients represent the bioavailable fractions of N and P, so this suggests that nutrient concentrations in pasture lakes are frequently in excess of those required for algal productivity, although this varies seasonally.

Conductivity and turbidity also differed among land-use classes (see Figure 8). For conductivity, median values were similar across classes but there was a strong trend for high conductivity values in the pasture class (high conductivity values are usually associated with higher trophic status). The highest turbidity values were also in the pasture class, which is consistent with known impacts of catchment development on lake turbidity. Relatively high turbidity values in many alpine lakes are associated with glacial flour.

The median values from the box plots in Figures 5 to 8 are summarised in Table 5, showing the increase in concentrations of water quality parameters, and the decrease in clarity (Z_SD) from alpine to native to exotic forest to pasture classes. Maps of water quality parameters are given in Appendix 3.

Table 5: Median values for water quality parameters in the four main land-cover classes (number of lakes in each class shown in brackets)

Notes: Total number of lakes for each parameter ranged from 68 to 121 due to differences in monitoring programmes; na = no data available.

Figures 9 to 11 show significant regressions between water quality parameters and climate, land cover and lake morphology characteristics. Because many of the relationships are non-linear, Figures 9 to 11 are plotted using best-fit functions (linear, logarithmic, etc) for each relationship, as stated in the figure notes. Regressions were judged significant if they satisfied both p < 0.05 and r² > 0.20; this applies to all regressions shown in Figures 9 to 11.

Given the patterns seen in the cover classes it was not surprising that TLI was also significantly related to the degree of the catchment retaining native cover and the amount of pasture in the catchment (Figure 9). The effects of native and pasture land cover on water quality parameters are summarised in Table 6. Low native cover and high pasture cover correlated particularly strongly with increased TLI and the four TLI parameters, and less strongly with increased NH₄-N and turbidity. Only NO₃-N, DRP and conductivity were not correlated with land use. Native cover was more strongly correlated with TLI parameters than pasture cover, consistent with the identification of lower water quality associated with other land uses such as exotic forest and urban cover.

The scatter of points in Figure 9 is not surprising given that dominant cover is a coarse measure of land use. Lakes with similar pasture cover have widely different stocking rates and fertiliser use. Very few lakes with over 40% native cover had eutrophic (TLI > 4.0) values, and all of these had significant proportions of pasture or exotic forest in their catchments.

Only 10 lakes had more than 5% urban development, and TLI was over 4 for all these lakes. The two predominantly urban lakes both had eutrophic TLI values (4.3 for Pupuke and 4.6 for Hamilton Lake), but these lakes were less eutrophic than the other eight lakes with over 5% urban cover, all of which had high pasture cover (35–70%) contributing to their water quality. Hence, pasture may have a stronger effect on trophic level than urban cover, provided the urban development lacks point source discharges and has reticulated sewer systems, as is the case with Lake Pupuke and Hamilton Lake. Urban development may have a much greater effect if its waste treatment is by septic tanks or has other features such as stormwater discharges that allow nutrient run-off to lakes (eg, Gibbs, 1991).

It should be noted that, given the context of a broad national comparison of lakes varying widely in climate, area and depth, the correlations in Figure 9 and Table 6 are highly significant. Together with the following regressions in Figures 10 to 11, they explain most of the total variation in lake water quality, and the interactions between these environmental drivers is examined further by the multiple regression that follows.

Table 6: Spearman rank correlations for water quality parameters (three-year averages from 2004–2006) vs pasture land cover and native land cover

Note: Significant correlations are shown for *p < 0.05, **p < 0.01, ***p < 0.001.

Correlations are negative for native cover and positive for pasture.

Lake altitude and temperature were the other parameters that showed significant correlations with TLI (see Figure 10). High-altitude lakes are largely microtrophic and oligotrophic (TLI < 3), and most of the eutrophic, supertrophic and hypertrophic lakes are lowland lakes.

TLI increases strongly with temperature (Figure 10). Some of this relationship is due to the generally greater degree of catchment development in lakes in warm regions than in cold regions. However, Figure 10 shows that there are also numerous mesotrophic and eutrophic cold lakes, such as Lakes Emma, Clearwater and Johnson, all of which have high proportions of pasture in their catchments. The cold eutrophic lakes in Figure 10 are also almost all small, shallow lakes.

Figure 11 shows that deep lakes generally have higher water quality than shallow lakes. Note that there is considerable variation in water quality in the shallow lakes in the data set, largely due to the differences in land-use patterns already discussed. No significant regressions were found with the other physical descriptors lake area, fetch (lake distance over which prevailing wind blows), and residence time.

[ Previous|Next