Modelling the hydrological impacts of land use change and integrating cultural perspectives in the Waikouaiti Catchment, Otago New Zealand

Changes to land use in a catchment impacts the quality and quantity of water as well as affecting Māori relationships with water. However, cultural effects are often considered in isolation to the more easily quantifiable effects on water quality and quantity. As such, the aim of this research is to explore how scientific and cultural data can be integrated in a freshwater management setting to inform decision making in the Waikouaiti catchment, Otago. The impacts of land use change on flow regimes and water quality were modelled using the Soil and Water Assessment Tool (SWAT) and the Catchment Land Use for Environmental Sustainability (CLUES) model. A cultural stream health assessment survey was undertaken by whānau members to articulate cultural perspectives. The SWAT model was calibrated for baseline conditions at two locations in the catchment; Bucklands and Lawsons. Monthly Nash Sutcliffe efficiency values for the calibration periods were 0.68 and 0.83.Water quality was modelled at Bucklands and Orbells. Total nitrogen (total N) and total phosphorus (total P) loads and concentrations for the baseline were overestimated at Orbells but reasonable at Bucklands. Six land use simulations were run in SWAT and CLUES. The cultural stream health assessments were conducted over a ten month period at two culturally significant locations; Bucklands and Orbells, individual scores were averaged and compared to mean flow on the sampling day. Replacing existing mixed native and exotic forestry with woody biomass (scrub) at Lawsons resulted in a significantly altered flow regime; 7-day mean annual low flows (7-day MALF) were increased by 68% and peak flows by 45%, with the highest proportion of surface runoff compared to other land uses. In contrast, 7-day MALF decreased by 53% in a fully afforested scenario whilst the impacts of partial conversion to tussock or low producing grassland were smaller; ~12% increases in annual yield primarily due to lower evapotranspiration. At Bucklands (predominantly low producing grassland) the largest decreases in mean annual runoff were from native forest (42%) and tussock (26%), decreases in both scenarios were a result of high evapotranspiration and low surface runoff, groundwater and lateral flow. 7-day MALF and peak flows decreased by 86% and 73% under the native forest and tussock scenarios whilst peak flows increased by 17% in the intensive dairying scenario which had a high proportion of total runoff from lateral and surface flow. The intensive dairying scenario was predicted to have the most profound effects on water quality at both locations; total N loads more than doubled and total P loads doubled compared to the baseline although there was no change in sediment loads. The lowest total N loads were modelled in the native forest scenario whilst woody biomass and tussock yielded the lowest sediment loads. Cultural satisfaction was not exclusively determined by flow as an unsatisfactory score in one cultural theme did not always correlate with an unsatisfactory score in all the other themes for that surveying day. Proportionately more themes were scored unsatisfactory at Bucklands than at Orbells. Although more days with flow less than 0.6 m3s-1 were scored unsatisfactory, themes were scored poorly across all flow. The themes with the most unsatisfactory scores at Bucklands were Wai Māori and health whereas at Orbells the landscape and overall theme scored poorly more frequently. By integrating hydrological and water quality modelling with cultural data a broader understanding of the impacts on cultural health and the suitability of various land use scenarios for aquatic life was achieved. This approach provides a template to meet legislative objectives and the outcomes are applicable in local decision making