• About this site|
• Site index|
• Contact us

Draft Guidelines for the Selection of Methods to Determine Ecological Flows and Water Levels

Acknowledgements

Executive Summary – Recommendations

1 Introduction

2 Rivers

3 Lakes and Wetlands

4 Groundwaters

5. References of Main Text and Appendices

Appendix 1: Relationship between Total Allocation and Ecological Flow Requirements in Rivers

Appendix 2: Effects of Flow Regime on Stream Ecology

Appendix 3: Methods Used in Ecological Flow Assessments of Rivers

3 Lakes and Wetlands

3.5 Description of individual methods

Lake and wetland water levels are determined by the difference between inflows and outflows (including precipitation and evaporation) at any point in time and the starting level. An understanding of basin shape is essential for any calculations of water level. Therefore any changes to inflow or outflow regimes (units as m³/s) will be reflected in lake or wetland levels. Water levels may be strongly influenced by groundwater inflows. The discussion below centres on levels but recognises that the inflow-outflow regime is central to setting ecological flows or water levels.

3.5.1 Historical time series analysis

Framework for use (Tables 3.5 and 3.6)

These methods are based on time series records of water levels and are the simplest and easiest to apply. Analysis of historical time series is based on water levels records and uses a statistic to specify maximum and minimum levels. The statistic could be the average level, the length of time a level must be within a specified range, or the length of time a level may exceed the minimum and maximum levels. A minimum of five years’ data would be needed to avoid biases due to natural inter-annual variability. There are much longer water level records available for most of the larger lakes in New Zealand.

The aim of historical analyses is to maintain the levels within the historical range, or to avoid the level regime from deviating largely from the natural regime. The underlying assumption is that the ecosystem has adjusted to the levels regime and that changes from this will cause reduction in the biological state (abundance, diversity, etc) proportional to the change in level. It is usually also assumed that the natural ecosystem will only be slightly affected as long as the changes in level are limited and the water body maintains its natural character. It is implicitly assumed that the ecological state cannot improve by changing the natural water level regime.

The most detailed example of application of rules for lake level management in New Zealand has been the development of the lake operational guidelines for Lakes Manapouri and Te Anau (Mark and Kirk 1987). A 35-year record was used to establish the natural range. The operational guidelines define high, main and low operating ranges, and restrict the amount of time the water level is allowed outside the main range. Long time series records were used for the calculation of lake level duration profiles in the analyses of levels in Lake Taupo (ECNZ 1995; Mighty River Power 2001) and Lake Rotoaira (Genesis Power Limited 2000). These profiles record the percentage of time that the lake has different levels and provide an excellent visual and quantitative summary of the extent of level change.

Bathymetric lake maps are important; there is limited ability to predict the extent of level change anticipated or the effects of changing levels without bathymetric data.

The relationship between level change and biological response in lakes and wetlands is of course not linear. In shallow water, small fluctuations have a much greater effect than the same fluctuations in deeper water (Walker and Coupland 1968). For this reason it is recommended that the historical time series analysis be carried out by an expert panel if the levels will vary by > 10% of the natural regime in summer and > 30% in winter, unless there is a quantified relationship between biological response and levels.

3.5.2 Expert panel

Framework for use (Tables 3.5 and 3.6)

Expert panels may be used to provide assessments of effects of changes to flows and levels in situations where the lake and wetland values are low to medium and where the available information on the waterbody is sparse. Expert panels usually comprise interested parties as well as ‘experts’. Expert panels may inspect the system and consider the suitability of suggested ecological levels or flows. If records are available, the panel will consider the proposed ecological water levels or flows and particularly the timing of these in relation to the combined knowledge of life cycle requirements of the organisms most likely to be affected. The suitability of ecological water levels or flows for many aquatic biota may be assessed by the panel provided panel members have relevant experience. Particularly important is the ability of the panel to recognise the boundaries of the system and understand the non-linear nature of the biological responses to level changes.

The panel may decide ecological water levels or flows. The panel should justify this decision and identify information needs for future decisions on ecological water levels or flows. It is recognised that results will be dependent on the objectivity of the experts.

3.5.3 Site and catchment mapping

Framework for use (Table 3.6)

This is particularly important for wetlands, which may not have distinct boundaries with catchments. Site and catchment (at least immediate surrounding catchment) mapping is needed to define the approximate boundaries of the wetland, soil types, the approximate water depths, location of open water and channels, and if possible major vegetation types. All of the above determine the values (eg, distribution of flora and fauna) and are affected by water levels and the periodicity of water level change. Mapping can be done by ground-truthing aerial photographs, working with soil and land cover maps (eg, LCDB2 (Terralink International Ltd), or satellite imagery. Landcare Research has now completed a delineation of wetlands across New Zealand for the Department of Conservation, to be available for this purpose.

3.5.4 Wetland record sheet

Framework for use (Table 3.6)

This method is provided in the Handbook for Assessing Wetland Condition (Clarkson et al 2003), developed to allow condition assessments for wetland environments in New Zealand, and was based on the major pressures affecting wetland condition (drainage, nutrient enrichment etc). It included a specific section on hydrological modification that is directly relevant to the type of water abstraction that would occur with flow variation. The Handbook covers in detail the types of hydrological damage that occur in wetlands and their effects on wetland values, and how to assess them. The record sheet is a starting point, and hence minimum requirement, for assessing effects of flow modification on wetlands.

The method (Clarkson et al 2003) requires completion of a table to classify a wetland in a wider context including evaluation of the value of a wetland in terms of type, regional distribution, rarity. The method requires a suitably qualified person with a basic knowledge of wetland systems. A site visit is essential but the requirement for data is minimal as long as the assessment is carried out by a qualified analyst.

The Handbook is linked to other tools developed for wetland management in New Zealand, including the wetland classification scheme (Johnson and Gerbeaux 2004), and the forthcoming Department of Conservation’s lake and wetland prioritisation (Waterbodies of National Importance, WONI) that will provide a multi-factor prioritisation of wetlands in terms of conservation value (contribution to rare habitats and species).

The Wetland Record Sheet component of the methodology is a simple summary of the condition, components and pressures on a site that can be easily used after relatively little training, and suitable for the ‘low to medium’ level. Other more detailed components of the methodology are discussed below.

3.5.5 Habitat analysis in drawdown zone

Framework for use (Tables 3.5 and 3.6)

Habitats refer to the substrate type and water depth. Examples in lakes are rocky platforms, rocky reefs, cliffs, sand, gravel and mud in relation to water depth. In large lakes, wave action usually results in the formation of wave-cut platforms below water level and beaches at and above water level, where substrate type permits (Mosley 2004). In wetlands substrates may be peaty or mineral, with a variety of particle sizes, and water depths may range from areas that are only occasionally flooded to areas that are permanently submerged.

The proportion of lakeshore occupied by various habitat types and their depth distributions need to be quantified. Where sensitive habitats occur, their proportional loss or gain at various lake levels is calculated. Because lake edges are seldom uniformly sloping, the first stage in an analysis of habitats is from a hypsographic curve on which lake levels can be plotted. Wave-cut platforms near the lake surface imply that significant loss of lake area and hence lake littoral communities may occur for relatively small changes in lake level. Water level variation along cliff faces, in contrast, may have little biological impact.

Similar issues arise in wetlands, where community composition and species diversity of plants as well as animals vary across sites with different substrates and water depths.

Many of the values in Tables 3.1 and 3.2 will be impacted by water level change if this change coincides with areas of gentle slope thereby removing significant areas of habitat. Calculations may be based on:

• rarity of habitat type impacted by the level change; for rare habitats any loss may be considered unacceptable

• proportion of habitat type impacted by the level change; for common habitats less than 20% loss may be considered ‘Low impact’ and more than 30% loss may be considered ‘High impact’

• ability of the biological communities in the habitat to migrate up or down with water level. Some plants that provide habitat structure are highly adaptable to water depth variations (eg, characean algae) and some, mainly those that require exposure for flowering at some stage in the life cycle and therefore live only in water depths < 1 m are highly intolerant of deep water. Many shallow-water plants may survive long (weeks’) exposure to air in winter, but are intolerant of short (days’) exposure in summer.

This assessment needs to be made by an experienced practitioner.

3.5.6 Species-environment models

Framework for use (Tables 3.5 and 3.6)

Lake and wetland communities can be categorised into four major groups that respond (often in concert) to changes in water levels. These are aquatic macrophytes and periphyton, aquatic macroinvertebrates, fish, and birds.

Aquatic macrophytes are fundamentally important in lake-edge structure and in defining the characteristics of wetlands. These may be submerged, emergent or free floating (de Winton and Schwarz 2004). Assessments in relation to water level change require two stages.

The first stage in a macrophyte assessment is a biodiversity analysis, including:

• number of species, particularly native species

• identification of rare species

• analysis in the depths that will be impacted by an ecological level or flow regime.

The second stage is an assessment of the depth distribution of the communities within the water body. Methods are summarised by de Winton and Schwarz (2004). The LakeSPI method for water quality assessment (Clayton et al 2002) may also provide the necessary information. With this combined information a proportional loss (gain) of macrophyte species and of macrophyte-dominated communities can be calculated by overlaying the proposed water level regime on the existing regime using either direct transects and/or the hypsographic curve as outlined in Section 3.5.5. An expert judgement is then required as to whether this loss is significant: there are no current quantitative methods to define importance and there is the potential for the plants (depending on the species) to adjust by upward or downward migration to new levels.

Macroinvertebrates occupy the middle levels of aquatic food webs. Their numbers, distributions and life cycles are often critically dependent on water levels, particularly in shallow water where drying may occur with water level changes.

The wetlands and the littoral zone in most lakes include a variety of habitats, thus necessitating a large number of replicate samples and sites to encompass spatial variability. Additionally, habitats within the lake littoral zone can be extremely diverse (eg, macrophytes, boulders, fine sediment), requiring the use of several sampling methods. The following four methods are recommended by Kelly and McDowall (2004) for macroinvertebrate community assessments in lakes:

• sweep netting

• benthic grabs

• coring

• Hess/Surber sampler.

The pros and cons of the use of these are given by Kelly and McDowall (2004). There are other methods in the literature (eg, detailed diver assisted analyses) but they are generally more time-consuming and will be needed only for very detailed work if warranted. Sampling needs to be conducted to obtain a relationship of benthic macroinvertebrates to water depth and then the same process is applied as for macrophytes above.

Fish communities that will be affected mostly by water level variations are those that utilise or inhabit the littoral zone of lakes. For example, bullies may use plants in the littoral zone as spawning sites, and fish in spawning streams may be affected by variations in lake levels where the spawning stream enters a lake across a delta.

Quantitative assessments of littoral fish communities include:

• seining

• fyke netting

• fish trapping

• gill netting

• electric fishing.

For details of the methods see Kelly and McDowall (2004) and references therein.

There are no native herbivorous freshwater fish in New Zealand so an initial analysis of changes to the water level regime can be obtained from the effects on macrophytes where these provide habitat for fish, and macroinvertebrates where these constitute the majority of fish food.

Quantitative bird counts in lakes and wetlands may be made by a number of techniques, but variability in numbers due to regular migrations between water bodies in a given area mean that good assessments require at least seasonal and preferably intra-seasonal counts.

The assessments need to be made for species directly associated with habitats that may change as a result of water level variability, eg, species directly dependent on the littoral zone for food or moulting and breeding shelter such as ducks, swans and pukeko (Williams 2004). Some shallow water bodies attract wading birds, and spectacular seasonal changes in abundance can be found where migrant species utilise lake margins (eg, Lake Wairarapa: Williams 2004).

3.5.7 Wetland hydrological condition assessment and model change

Framework for use (Tables 3.5 and 3.6)

The method is described in the Ministry for the Environment’s ‘Handbook for Monitoring Wetland Condition’ (Clarkson et al 2003). It is an extension of the Wetland Record Sheet (Section 3.5.4) supplemented by a data set from the wetland, that includes details of plant species presence and height as well as physical and chemical parameters measured in the field or from laboratory analyses for a number of plots in the wetland. This provides a more robust base for scoring wetland condition, and defines the factors controlling the habitat for biota more precisely. Wetland scores are assigned to a set of indicators using a systematic comparison and evaluation process based on expert knowledge.

The methodology (Clarkson et al 2003) can be used by any wetland manager after some training from wetland experts.

3.5.8 Water balance models

Framework for use (Tables 3.5 and 3.6)

Water balance models that relate the change in lake or wetland storage and therefore levels to inflows and outflows are relatively straightforward to compute. With knowledge of bathymetry, these can be predictive. Water balance models are therefore the next step from habitat or species-environment models as these can provide the predictive component for lake levels on which the habitat or species-environment models can be made.

The method therefore requires both hypsographic information and a time series of inflows and outflows including precipitation and evaporation. Running a simple inflow-outflow model requires keeping a balance between hydraulic inputs and outputs, and how any imbalance causes lake levels to change. It can be carried out over a wide range of timesteps by a simple spreadsheet approach.

Ungauged rivers and groundwaters entering lakes and wetlands are a potential difficulty. In some lakes and wetlands, evaporation may be significant (usually when surface area to volume ratio is high and in dry windy conditions (eg, Lake Ellesmere). Measurement of evaporation relies only on indirect methods, eg, by pans or energy balance models.

Changing the inflow-outflow-lake level regime will also change the lake residence time with considerable potential flow-on effects to lake ecosystems as discussed in Section 3.5.9.

3.5.9 Residence time – water quality modelling

Framework for use (Tables 3.5 and 3.6)

The inflow-outflow regime for a given lake or wetland may affect water level and residence time which in turn affects water quality. Simple empirically derived regression models can be used to estimate long-term (or equilibrium) nutrient and chlorophyll concentrations from values of nutrient loadings and lake residence time. Application of these models should be a first step in an assessment of effects on water quality of altering inflow-outflow regimes; it requires knowledge of nutrient loading rates and existing residence times of a lake (Ryding and Rast 1989). The models provide information and predictions on in-lake nutrient concentrations with changing inflows and out flows. Other empirical relationships have been derived relating chlorophyll to in-lake nutrients, as shown in the New Zealand examples in Pridmore et al (1985).

Consideration of residence time in water quality is not applicable when the residence time is less than approximately 10 days: the system can then be considered a riverine environment. In terms of residence time considerations alone, a rule of thumb might be that if the change in residence time is less than 10%, then residence time is not a useful parameter.

The models, although relatively simple empirical approaches, require some information on nutrient loading and lake water quality. They are applicable when residence times are > 10 days and < 10 years. Experience in limnology is required to develop and apply these models although they are available in the literature (eg, Ryding and Rast 1989).

3.5.10 Detailed local wetland delineation

Framework for use (Tables 3.5 and 3.6)

Section 3.5.3 above described how remote delineation of wetland environments, using mapping and GIS techniques, is an important step in any assessment of effects of flow change on wetlands. For wetlands with medium to high values, or where the potential change to the water level is medium to high, the boundaries of the wetland environment with the terrestrial catchment are usually an important issue; this often requires a more precise local on-site delineation than is possible from remote techniques. This is because the significance of the wetland community and its susceptibility to hydrological modification are often greatest near the margins (margins are often species-rich and rare communities). Methods for onsite wetland delineation were initially developed by the United States Army Corps of Engineers, and allow the nature and extent of the wetland and its constituent communities to be determined quantitatively. The methods involve a range of hydrological, soil property and vegetation composition techniques, and can be applied to data collected using the methods described in Sections 3.5.4 and 3.5.7. In order to be applied objectively, only qualified wetland experts can carry out this method. The basic principles are described in American publications such as Tiner (1999); and many of these methods have been modified for New Zealand conditions by local wetland practitioners.

3.5.11 Bank stability and geomorphology

Framework for use (Table 3.5)

One of the major effects of change in water level in lakes is impairment of lakeshore stability, resulting in shoreline collapse and loss of beach sediments. This problem has received considerable attention in New Zealand, as collapses of Lake Manapouri shorelines in 1972 were an important driver behind the lake operational guidelines for Lakes Manapouri and Te Anau. Geomorphological methods for determining whether a change in flow regime and lake level is likely to cause such problems are well established, but are complex and must be carried out by qualified and experienced engineers. Kirk and Henriques (1986) and Kirk et al (2000) provided detailed examples of how to carry out these methods.

3.5.12 Full ecohydrological assessment

Framework for use (Table 3.6)

Given that the extent of wetlands is so reduced in New Zealand, any proposed hydrological alteration that has potential to affect a wetland of medium to high value would need to be assessed with the greatest caution; and with thorough, objective methods for predicting effects of change in response to lowering or raising of water tables. Detailed ecohydrological assessments, which model the distribution and productivity of natural communities in relation to hydroperiod, are the most robust approach. Yet they have not often been applied because of their high cost and need for a high level of scientific expertise. Ecohydrology is an interdisciplinary approach in which hydrologists, soil scientists, and biologists use hydrological data and species distributions to identify gradients and patterns in hydrology-species responses, recognising that even small differences in hydrology can have large effects on biological values when water levels pass critical values. The methodology involves monitoring networks of hydrological instrumentation (dipwells, piezometers and capacitance probes) placed along ecological gradients to characterise hydrology-species composition relationships, and make predictive models for effects of change in hydroperiod. Browne and Campbell (2005) give a recent New Zealand example of a detailed ecohydrological study for wetland management purposes. This approach clearly requires a high level of expertise and should always involve qualified wetland experts.

3.5.13 Microtopographic survey

Framework for use (Table 3.6)

Wetland communities, and therefore wetland structure and function, are critically dependent on small (often cm scale) changes in topography that may mean the presence or absence of standing water, channelised flows and aerated versus non-aerated soils. Microtopography is therefore a key factor related to hydrology that promotes the development of vegetative structure and composition, and biogeochemical functions. Microtopography requires accurate field mapping and survey. If a full wetland microtopographic map cannot be produced for cost or logistic reasons, then survey cross-sections need to be chosen that best reflect the water level issues that face the wetland on a case-by-case basis. Microtopographic work is usually accompanied by detailed soil profiling. Soil profiles are carried out by coring or, if the soils are relatively dry at the time of study, soil pits. Alteration of wetland levels in peat soil-dominated areas may have a significant effect. Drying of peat-dominated soils invariably results in soil shrinkage and general lowering of soil profiles around the affected wetland area (eg, Lake Poukawa in Hawkes Bay).

Microtopography involves transect elevation measurements using survey equipment. Several organisations in New Zealand have expertise in microtopography. Rapson et al (2006) provide a recent New Zealand example showing the importance of microtopography and how to design a microtopographic analysis in relation to a hydrological gradient.

3.5.14 Wave action assessment

Framework for use (Table 3.5)

Wave action assessments are required when lake levels change to the extent that new shorelines are formed. These assessments are often closely linked with bank stability and geomorphology (Section 3.5.11).

Data requirements include an outline of the lake shoreline, bottom bathymetry, and information on wind speed and direction. Wave action assessments have been carried out for example on Lake Taupo to simulate effects of lake level changes on shoreline erosion (Hicks et al 2000), and this method can also be applied to assess the effects of waves on littoral zone vegetation and biological communities. The method involves application of a wave hindcasting model coupled with a wave refraction model. The underlying science and methodology may be found in the Coastal Engineering Manual (http://www.veritechinc.net/products/cem/index.php).

3.5.15 Water clarity assessments

Framework for use (Table 3.5)

Water clarity is influenced by lake levels in two ways.

First, lake levels influence sediment re-suspension; suspended sediment effects are most important in lakes where the shorelines are gently sloping and/or where much of the lake bed is influenced by wave action. Changing lake levels in deep lakes with steep sloping shores will have a relatively small influence on lake clarity (and hence ecosystem effects) when compared with lake level changes in shallow lakes with gently sloping shores. Changing inflows may alter the sediment loading and hence clarity of lakes, as is the situation for Lake Manapouri and Lake Coleridge where diversions of sediment-rich water are controlled to minimise inputs to the lakes.

Second, lake levels influence lake algal growth. The effect of lake level change on water clarity via changes in algal (phytoplankton) concentrations is indirect. If the change in water level regime alters the primary drivers of phytoplankton dynamics, nutrients temperature and light, then it will influence phytoplankton biomass and hence water clarity.

The method involves considerable field work to measure light penetration both inshore and offshore in different wind and wave conditions, and at different times to derive relationships between clarity and sediment and phytoplankton concentrations. The latter may be predicted from their residence time models (Section 3.5.9) and wave action assessments (Section 3.5.14).

An assessment of lake level influences of this nature requires considerable knowledge of lake ecosystem processes and a good knowledge of the relationships of sediment and phytoplankton biomass and clarity relationships which vary from lake to lake (Hamilton et al 2004).

3.5.16 Temperature modelling

Framework for use (Table 3.5)

The temperature of a river inflow determines the amount of heat that the river delivers to, or removes from, a lake; and also the depth in the lake to which the inflowing water will sink. This latter effect is generally of greater concern, because of the substances that may be transported by the inflow (oxygen, nutrients, suspended particulate matter); and because of the control that thermal stratification exerts on vertical mixing and transport of these substances within the lake.

Application of a 1D hydrodynamic model that simulates vertical thermal structure and predicts insertion depths for inflows is helpful to understand the dynamics of river-lake interaction.

Outflow dynamics can influence in-lake thermal structure in deep artificial reservoirs with offtakes at depth, but this will not be considered here.

An example for the inflows from the Tongariro Power Development Scheme into Lake Taupo using a 1D model is given in Spigel et al (2005) and for Lake Rotoiti using a 3D model is given in Stephens (2004).

3.5.17 Groundwater / surface water interaction

Framework for use (Tables 3.5 and 3.6)

Groundwater inputs are an important component of the hydrology of wetlands and shallow lakes, and changes in groundwater input potentially have significant effects on nutrient inputs and hence ecological character. Rates of groundwater discharge are known to be important in sustaining productivity of the littoral vegetation of lakes and of wetland communities. Groundwater discharges from catchments into wetlands and lakes can be estimated from methods such as piezometer clusters, Darcy calculations and salt balances. The groundwater section of this document has further details of methods for estimating groundwater discharge.

3.5.18 Hydrodynamic water quality models

Framework for use (Table 3.5)

A comprehensive process-based lake model may be used if a risk of adverse effects is identified and the results from empirical residence time modelling, clarity and temperature modelling (Sections 3.5.9, 3.5.15, 3.5.16) does not provide sufficient information to assess ecological water levels and flows. Such modelling requires extensive supporting data including climate, inflows and the in-lake conditions.

The first stage is the implementation of a lake hydrodynamic model. Such a model provides necessary background information on thermal structure, mixing, and water movements that can be used to underpin water quality and aquatic ecosystem models. A suite of hydrodynamic models for different purposes and lake types are available (see Hamilton (1999) for a discussion of models and their applications).

In a combined hydrodynamic-water quality ecosystem model, the degree of spatial and temporal resolution provided by the hydrodynamic model controls the overall resolution of the simulation. Hence a 1D thermal stratification model will limit water quality predictions to profiles that represent lake-wide averages. If 2D or 3D effects are of interest or cannot be ignored (as in very large shallow lakes) then 2D or 3D hydrodynamic models must be used in combination with a water quality model.

Following calibration and verification of a hydrodynamic model, the second stage is the coupling of this to a lake water quality model. It is probably best to run the hydrodynamic and the water quality models as a coupled system. To capture any interactions between the physics, chemistry and biology, it is also possible to run the models in an uncoupled mode: output on physical factors from the hydrodynamic model are then used to control the movement and mixing of water quality components in later, separate water quality model runs. Process-based water quality models simulate interactions between lake inflow-outflow regime, lake physical dynamics, water quality and biological components.

An example of such a modelling exercise in relation to changing inflows is given in Hamilton et al (2005) for Lake Rotoiti, simulating effects of a flow diversion into and out of the lake.

[ Previous|Next