Executive summary

Auckland Council (AC) is currently developing a Unitary Plan that will establish default limits for the allocation of surface water across the Auckland region. The objective of these limits is to provide a specific level of environmental protection, while enabling out-of-channel water use at specified levels of availability and reliability. To achieve this, limits must include at least minimum flows (the flow below which no water can be abstracted) and total allocation (the total quantity of water that can be abstracted upstream of any location). 

Default limits can be defined using “rules of thumb” that are based on hydrological indices, such as the mean annual low flow (MALF). This is the approach that is taken by the proposed National Environmental Standard for Ecological Flows and Water Levels (proposed NES; MfE 2008). Rules of thumb are easily applied, but there are two main disadvantages. Firstly, the consequences for both environmental values and out-of-channel water users (i.e., availability and reliability) are not clearly articulated, making justification of the rules difficult. Secondly, the consequences are spatially variable because flow regimes and the relationship between environmental protection and flow are spatially variable.

The Environmental Flows Strategic Allocation Platform (EFSAP) provides a method to evaluate the consequences of setting different water resource use limits across all parts of a catchment or region, including those for which detailed information is not available. It integrates scientific tools to enable the concurrent evaluation of consequences for instream habitat and reliability of supply for out-of-channel water uses, accounting for the interaction between the flow regime, minimum flow and total allocation limits at all locations. 

In this study we used EFSAP to simulate the consequences of various potential sets of limits (i.e., minimum flows and total allocations) for all river and stream reaches in the Auckland region with a mean flow greater than 10 L s-1. A range of alternative scenarios, both more environmentally conservative and more resource use enabling than the proposed NES rules, were simulated.

The indicator species selected for analysis were banded kokopu (Galaxias fasciatus), shortfin eel (Anguilla australis), longfin eel (Anguilla dieffenbachia) and common bully (Gobiomorphus cotidianus). These were selected based on their presence and value in the Auckland region. EFSAP is based on the analysis of flow duration curves (FDCs). In this study we used the annual FDC and the March FDC to represent the average consequences and the consequences under the most restrictive summer conditions respectively.

Under the proposed NES default limits for small rivers (minimum flow 90 per cent MALF, total allocation 30 per cent MALF) the spatial patterns of reliability at the management flow (where partial restrictions begin) are similar for both the annual and March FDCs. For the annual FDC, overall variability throughout the region is relatively low, with the majority of locations having reliability of between 80 per cent and 95 per cent at the management flow. However, for the March FDC, while the spatial patterns are similar, the magnitude of variation between locations in the region is much greater. 

Overall regional variability in reliability at the minimum flow (i.e., where abstractions must cease) was low for the annual FDC with most locations having a reliability of between 90 per cent and 95 per cent. For the March FDC, regional variability was greater, with a range for the majority of locations of 70 per cent to 100 per cent reliability, but the median reliability was similar between both at 93.4 per cent for the March FDC and 94.5 per cent for the annual FDC.

The spatial pattern of consequences for instream physical habitat was similar for all four indicator species that were modelled. The median loss of habitat for the annual FDC was - 6.0-6.5 per cent for banded kokopu, shortfin eels and longfin eels, and slightly higher at - 8.2 per cent for common bullies. For the March FDC, the median loss of habitat was -7.0 per cent for common bullies and-5.0-5.5 per cent for the other three indicator species.

To improve equitability for stakeholders, we attempted to define spatially discrete management units with relatively uniform outcomes for each of the values. However, no significant differences were found between classes based on geology, land cover, stream order or river size. Consequently, it was concluded that there was insufficient spatial differentiation at the regional scale being considered to justify the definition of different spatial management units within the region.

A range of allocation scenarios were simulated for the Auckland region for both the annual and March FDCs. The consequences for each value for the full range of simulated scenarios were then summarised in a decision space diagram that encompasses minimum flows ranging from 10 per cent to 100 per cent of MALF and total allocation limits that range from 10 per cent to 150 per cent of MALF, each in 10 per cent increments. 

Once objectives have been set for each value, the decision space diagrams can be used to determine which combination of limits satisfies each objective. Once the subset of limits that satisfy the objective for each individual value have been defined, they can be combined to find the set of limits which meet all objectives. In some cases, the defined objectives for all values will result in a combination of limit options that overlap. Water resource managers therefore have the choice of defining limits that satisfy all objectives. However, in some circumstances there will be no combination of limits that satisfies all objectives. In this situation, a compromise has to be found between the different values until an acceptable combination of limits can be agreed upon. The decision space diagrams can assist in this trade-off process by illustrating to stakeholders and resource managers how limits interact with each other and the relative consequences of alternative
management decisions.

It must also be recognised that EFSAP does not evaluate all values that may be relevant for a given location. It also does not explicitly consider flow variability, or the temporal sequencing on flows. It is also based on the assumption that instream physical habitat at low flows is limiting. These factors must therefore be considered when determining the most appropriate combination of limits. However, despite these limitations, EFSAP provides a robust and defensible approach to evaluating the relative merits of different combinations of limits and therefore will allow AC to more transparently communicate and set water resource limits that meet their nominated objectives.

Auckland Council technical report, TR2012/043

July 2012