However, root zone loss estimates alone don't allow us to manage to water quality limits at the catchment-scale.
Root zone losses are just the beginning since the National Objectives Framework that forms part of the Government's Freshwater Reforms, defines water quality bands and national bottom lines, not in terms of kg/ha/yr losses from the root zone, but as concentrations in rivers, lakes, wetlands and estuaries (in mg/L).
The challenge is to establish the causal link between the water quality observed in these water bodies and the sources of nutrients.
Where the nitrogen has come from, when it was lost from the root zone and to what extent it has been reduced before arriving at the water body.
This depends on the transport and transformation processes occurring in the subsurface between the bottom of the root zone and the water body.
To identify the source areas of nitrogen arriving at a particular water body, we first need to understand the subsurface water flow paths (e.g. artificial drainage, interflow, shallow and deeper groundwater) and groundwater- surface water exchanges.
Modelling of the subsurface hydrological system helps to define the boundary of the groundwater catchment that contributes both water (and the nitrogen it carries) to a water body.
Recent investigations have clearly demonstrated that groundwater catchments do not neces- sarily match the topography of defined surface water catchments.
Getting the boundary right is particularly important where special regulations are put in place in a catchment feeding an iconic water body like Lake Taupo, while less stringent rules apply to neighbouring catchments.
Subsurface lag times are the second issue to consider when trying to link observed water quality levels with the intensity of land use.
Depending on the relative importance of the various flow paths and the size of the subsurface reservoirs, the current water quality could largely be due to recent land use intensity (e.g. Southland) or the legacy of historical land use (e.g. the North Island's central plateau).
Water dating and modelling research during the last decade has substantially improved our understanding of lag times but substantial uncertainties remain.
Thirdly, it is critical to quantify any nitrogen reduction along the subsurface flow path between the bottom of the root zone and the water body of interest.
Particularly the big alluvial aquifers of the eastern provinces like Canterbury and Hawke's Bay where a dilution of nitrogen-enriched water from agricultural land on the plains mix with near-pristine mountain water resulting in a substantial reduction in nitrate
concentration.
This 'dilution effect' helps to meet concentration limits. In contrast, not only the concentration but also the mass of nitrogen is reduced in groundwater systems where denitrification occurs.
The substantial role of this ecosystem service is recognised and utilised by many European drinking water supply companies and regulatory authorities.
However, systematic research here into groundwater denitrification was only initiated in the past decade so we have only a limited understanding of where and to what extent groundwater denitrification occurs.
In summary, root zone loss estimates play an important role in our land management, but they are just the beginning with regard to evidence-based management of our water resources.
To defensibly establish the causal link between past and current land management and water quality, we need to understand the subsurface transport and transformation processes occurring in the catchment area of the water body.
Adequate research investment targeted at establishing novel measurement and modelling capability would be required to achieve this goal.
