4 Conclusions
During the period of rapid development of understanding of the limnology of New Zealand lakes (1970–1990), many empirical relationships were revealed showing increases in lake productivity and eutrophication of lakes due to increases in nutrient enrichment. Much of the early research focused on point sources discharges (eg, Lakes Rotorua and Rotoiti, Bay of Plenty), and the feasibility and time frames required for restoration after diversions of these. More recent concerns have focused on diffuse nutrient run-off due to catchment development and changes in land use. In particular, recently developed models of nutrient run-off strongly suggest that lake water quality will be lower in developed catchments than in undeveloped catchments. The analysis in this study has shown that all key water quality parameters and TLI indices were poorer in pasture catchments than in either native or exotic forest catchments. Most of these degraded rural lakes were lowland lakes.
The patterns in water quality seen with land use are consistent with our understanding of factors responsible for lake degradation. Catchment development is known to increase nutrient loads to lakes in tributaries and ground water, and to lead to increased phytoplankton productivity, decreased water clarity and changes in dominant algal species. Large, deep lakes take a long time to respond to increased nutrient loads associated with catchment development as a result of their long residence times, and hence show greater initial resilience to nutrient inputs than shallow lakes. However, large lakes do suffer extensive degradation if nutrient enrichment continues, and are more difficult to restore than shallow lakes because of longer residence times. There is evidence of degradation occurring in some of our larger lakes where catchments have been developed (Vant and Huser, 2000), although that is not apparent in the data analysed in this report. Degradation is likely to occur in other large lakes if nutrient inputs increase (eg, as a result of conversion from native cover to pasture).
Lack of knowledge of the range of lake water quality in pre-European times is often raised as an issue when assessing deterioration caused by human activity. The extent to which some lakes may have had naturally eutrophic conditions is an important consideration in lake water quality studies. The best evidence on the pre-European or pristine state comes from palaeolimnological studies of lake sediments. Existing palaeolimnological evidence in New Zealand suggests that some shallower lakes may have had moderately high nutrient concentrations, but this is probably less widespread than often claimed. Early botanical records from shallow lakes in the Waikato, for example, indicate very high water clarity in pre-European times. In our opinion, the eutrophic, supertrophic and hypertrophic conditions in the current state are largely the result of human activity.
The regression tree analysis suggests that climate change is likely to be an important issue exacerbating deterioration in lake water quality in future. Temperature featured as an important factor throughout the regression trees, and this strongly suggests that if lakes become warmer, their sensitivity to land use in the catchment will also increase.
The limited number of small lakes in monitoring programmes compared to the number of small lakes in New Zealand, and the greater degradation of small lakes than large lakes, suggests that they are likely to be under-represented in the national picture of lake water quality in New Zealand. There has also been much less research carried out on nutrient cycling and food web structure in small lakes than in shallow lakes, so there is less understanding of how they respond to nutrient inputs. Future research issues for these lakes include the nature of internal nutrient regeneration, the importance of polymixis and short-term stratification patterns, and the effects of nutrient enrichment on their food-web structure.
As part of this project, we also collected dissolved oxygen data from councils. However, oxygen profile data were restricted to a small number of regions and lake types and there were not enough records for broad-scale geographic, land-use and lake-type comparisons. It should be stressed that dissolved oxygen is a key indicator of degradation in lakes, and more extensive data sets would be valuable for understanding lake status and trends. Profile data will continue to be important for deeper monomictic lakes, but shallow polymictic lakes are likely to show diurnal oxygen cycles with nocturnal de-oxygenation accompanying eutrophication. The recent development of oxygen sensors with data-logging capacity makes nocturnal monitoring and monitoring over longer time periods more feasible.
Overall, it can be seen that New Zealand still retains a large number of relatively oligotrophic lakes, although most of these are in high-altitude areas with low human pressures. Eutrophication processes have nevertheless evidently resulted in extensive increases in nutrient status and biological productivity on a wide scale, and the worst water quality is strongly associated with pasture in the catchment. Although much of the early water quality science in lakes focused on the effects of point source discharges, diffuse catchment run-off is clearly implicated as a major cause of the current condition of lakes throughout the country. Application of nutrient run-off models in conjunction with improved understanding of in-lake processes in lakes of different sizes will be critical to understanding, managing and reversing these processes. Restoration will need to focus on minimising loads throughout catchments, and understanding the time scales required for water quality to recover in relation to residence times and internal nutrient regeneration. The examples of improved water quality in lakes with restoration plans in place show that restoration is feasible, although it is usually a prolonged and expensive exercise.
