Thumbnail Image

Modelling the efficacy of in-lake and catchment remediation actions for restoration of a small, eutrophic lake

Land conversion from native vegetation to agriculture has been widespread in New Zealand over the past century. The resulting increases in nutrient runoff have led to eutrophication of many lakes and have resulted in harmful algal blooms, deoxygenation of hypolimnetic waters, and reductions in biodiversity. These problems occur in many of the Rotorua lakes including Lake Okaro, which is the most degraded of the twelve lakes. The lake has been subject to numerous restoration efforts over the past decade (2005-2014). In this study, I linked a catchment and a lake model to address questions around the efficacy of restoration measures. The model was also used to compare the relative benefits of catchment and in-lake restoration actions on lake trophic status. The catchment model INCA (Integrated Nitrogen Catchment model) was applied to the Okaro catchment. The output from INCA was linked to a lake water quality model DYRESM-CAEDYM (DYCD). Catchment model simulations with INCA provided a satisfactory fit to measured discharge values in the two inflow streams (Pearson R=0.77-0.82). The model adequately simulated the major variations in nitrate (NO₃), total phosphorus (TP) and soluble reactive phosphorus (SRP). However, ammonium (NH4) concentrations were not represented well for both the calibration and validation phases of the assessment. Three scenarios were simulated using the calibrated version of INCA. This model was implemented to investigate the effectiveness of the artificial wetland (established 2006) and associated riparian planting by removing it in the catchment model simulation. Another scenario included fertiliser reductions and a third was to convert all pastoral land to dairy farming. The scenario involving the removal of the wetland/riparian area resulted in increases in total and dissolved nutrients. The fertiliser reduction simulations resulted in decreased NH4 and SRP, both of which are a major input to dry-stock and dairy farms. The conversion of all existing pastoral land to dairy farming resulted in a 160% increase in NO3-N concentrations while SRP increased by 500%. The inputs from the INCA model were fed into the DYCD inflow file and compared with the previous analysis performed by Özkundakci et al. (2011). The calibrated model based on INCA inputs represented an improvement over earlier published versions as a result of continuous simulation of catchment inputs and enhanced calibration. The Trophic Level Index (TLI) was used to represent water quality changes in the lake in response to the different catchment management scenarios. The wetland/riparian removal increased TLI, with higher total nitrogen (TN) and total phosphorus (TP) concentrations in the surface and hypolimnetic waters, in addition to increased chlorophyll a concentrations in the surface waters. Average TLI for this scenario was 5.21, slightly greater (i.e. increased trophic status) from the baseline scenario result of 5.17. The fertiliser reduction scenario showed decreases in TN (average 26.9 mg m⁻³) over the seven-year time frame (2005-2012). Concentrations of TP showed minor reductions, averaging 9.6 mg m⁻³. Chlorophyll a concentrations declined indicating lower phytoplankton biomass, particularly lower cyanophyte concentrations. The TLI over the seven-year time frame (5.07) decreased in response to lower nutrient concentrations. The increased dairy scenario resulted in large increases in TN (average 39.4 mg m⁻³) and TP (average 4.6 mg m⁻³). These increases enhanced chlorophyll a levels indicating higher phytoplankton abundance due to increased nutrient availability. The TLI in this scenario was 5.25 compared with 5.17 in the baseline scenario. From 2003-2014 chemical flocculants (alum) and sediment capping agents (Aqual-P®) were applied to Lake Okaro in an attempt to decrease phosphorus (P) concentrations and reduce lake trophic status. To achieve a better understanding of the efficacy of these applications required a model to assess P removal in the surface and hypolimnetic waters. Assessment of alum and Aqual-P was evaluated through a comparative analysis between the original calibrated DYCD model run and the measured data. From 2008-2012 the total average PO4-P removal in the hypolimnion was 81 mg m⁻³. This comparison was followed by a sensitivity analysis of nutrient release parameters in the model based on values in the literature. Reduction of the PO₄ release parameters from 0.016 to 0.008 g m⁻² d⁻¹ resulted in close fit of modelled outputs to measured data for PO₄-P concentrations in the water column. This alteration resulted in an average PO₄-P removal of 108 mg m⁻³ from the hypolimnetic waters over the four-year simulation period from 2008 to 2012. The results from the catchment and lake modelling simulations provide a basis for a detailed examination of the efficacy of internal and external lake remediation applications. Future management techniques should be aimed towards the limitation of external nutrients from the catchment. INCA simulations indicate that an additional artificial wetland installation could attenuate 60-100 kg yr⁻¹ of NO₃₋ N and NH₄₋N as well as removing 5-20 kg yr⁻¹ of TP. This research provides scientists and lake managers with a tool enabling the relative impacts of individual lake management actions. This tool allowed for assessments at a time when multiple actions were being enacted in the catchment and in the lake.
Type of thesis
Mallett, R. J. (2015). Modelling the efficacy of in-lake and catchment remediation actions for restoration of a small, eutrophic lake (Thesis, Master of Science (MSc)). University of Waikato, Hamilton, New Zealand. Retrieved from https://hdl.handle.net/10289/9681
University of Waikato
All items in Research Commons are provided for private study and research purposes and are protected by copyright with all rights reserved unless otherwise indicated.