Averting degradation of southern Ontario wetlands due to hydrologic alterations associated with development

Abstract

Policy statements in Ontario express a desire to prevent degradation that threatens significant wetlands. Yet loss of wetland features and functions continues. Better approaches and tools need to be applied to support planning and management and reverse the trend of incremental wetland degradation. This paper synthesizes existing knowledge to: (1) suggest some requirements for demonstrating that activities will not have a negative impact on a wetland feature or its functions due to hydrologic alterations; and (2) expose some flaws in reasoning with respect to wetland hydrology that could lead to poor decisions. Studies to demonstrate “no negative impact” should be required to confirm the plausibility of the conceptual hydrologic model of the wetland and consider the full range of hydrologic alterations that may be ecologically relevant. Evidence-based arguments, with sound scientific underpinnings, are needed to show that there will be no hydrologic changes, that ecological functions are not sensitive to changes, or that changes can be controlled to prevent negative impact. Better collaboration between disciplines will help to avoid results based on faulty or incomplete knowledge.

La règlementation en vigueur dans la province de l’Ontario (Canada) reflète la nécessité de prévenir la dégradation des milieux humides d’importance. Pourtant, cette dégradation, et la perte des fonctions écologiques des milieux humides, se poursuit. De meilleures approches et outils d’intervention sont nécessaires pour soutenir l’aménagement des territoires et inverser cette tendance. Cet article synthétise les connaissances existantes pour (1) proposer des critères permettant démontrer que les activités prévues n'auront pas de répercussion néfaste sur les milieux humides ou sur leurs fonctions hydrologiques, et (2) exposer quelques failles dans le raisonnement à l’égard de l'hydrologie des milieux humides qui pourraient mener à de mauvaises décisions. Il est proposé que les exigences imposées aux études environnementales incluent un modèle hydrologique conceptuel du milieu humide ainsi que l'examen de la gamme complète des modifications hydrologiques qui pourraient avoir un impact écologique. Des arguments fondés sur des preuves, reposant sur des bases scientifiques solides, sont nécessaires pour montrer qu'il n'y aura pas de changements hydrologiques, que les fonctions écologiques ne sont pas sensibles au développement prévu, ou que les modifications qui seront apportées au milieu puissent être contrôlées pour éviter les répercussions néfastes. Il doit être largement reconnu que les milieux humides peuvent dépendre de l’eau souterraine même si cette contribution est faible dans le bilan hydrologique du milieu humide. Également, les bassins de gestion des eaux pluviales ne contrôlent pas l'augmentation du volume d’écoulement se déversant dans les milieux humides en raison de l'urbanisation dans leurs bassins hydrographiques. Une meilleure collaboration entre les disciplines permettrait d’éviter les conclusions erronées ou incomplètes.

Introduction

Wetlands provide enormous value to society and sustain a level of biodiversity virtually unsurpassed by any other ecosystem on the planet. They provide habitat (e.g. Klove et al. 2011) for native plants, invertebrates, amphibians, fish, birds and mammals, including many threatened and endangered species (e.g. Gopal 2009). Wetlands improve water quality (e.g. Blackwell et al. 2009) through sedimentation, filtration and biogeochemical transformation processes. Some wetlands provide water storage capacitance within the landscape, attenuating the response of the groundwater table (McLaughlin and Cohen 2013), providing opportunities for groundwater recharge, and reducing downstream flooding (e.g. Maltby and Acreman 2011). Wetlands can provide carbon sequestration (e.g. Mitsch et al. 2015) and moderate local climate through evapotranspiration (McLaughlin and Cohen 2013).

Wetlands are critical to watershed management in Ontario (Ducks Unlimited 2001). Conversely, Bedford (1999, 775) identified the importance of watershed-scale management for wetlands, highlighting the cumulative effects of human activities on wetlands: “significant loss of wetland area, disproportionate loss of some wetland types, degradation of remaining wetlands, and a consequent decrease in the diversity of native wetland types and species.”

It has been estimated that natural, inland wetland areas have declined globally by 69–75% since 1900 (Davidson 2014). Snell (1987) reported that by 1982, the original wetland area in southern Ontario had been reduced by 68% with higher losses, on the order of 90%, in southwestern Ontario. A more recent analysis of wetland conversion in southern Ontario found that wetland loss has continued uninterrupted (Ducks Unlimited 2010). Whereas losses prior to 1982 were largely attributed to agriculture, the more recent study identifies urbanization as a significant factor, particularly within the Golden Horseshoe which encompasses Toronto and other densely populated areas around the western end of Lake Ontario (Ducks Unlimited 2010). There are, in fact, many human activities that cause disturbances which, individually or cumulatively, lead to wetland degradation. Climate change presents yet another threat to southern Ontario wetlands.

The Mixedwood Plains ecozone, in southern Ontario, includes geological features like the Niagara Escarpment and the Oak Ridges Moraine. The varied and complex glacial deposits in southern Ontario result in equally complex local groundwater flow patterns (Rivera 2014) and varied groundwater–wetland interactions. Most of the remaining wetlands are small; swamps are the most common wetland type (87%) and peatlands (bogs and fens) are the rarest (1%) within the Mixedwood Plains ecozone (Ontario Ministry of Natural Resources and Forestry [OMNRF] 2015). Coastal wetlands associated with the Great Lakes also occur in southern Ontario, but are not the subject of this paper.

Nobody has purposely planned for the level of wetland loss and degradation in Ontario. On the contrary, the language in the Ontario Provincial Policy Statement (PPS, issued under the Planning Act and updated in 2014) suggests the need to plan for maintaining, even restoring, natural heritage systems including wetlands.

Wetland losses serve as an example of environmental degradation due to what Odum (1982, 728) described as the “tyranny of small decisions.” They do not represent a conscious choice of a preferred solution but rather the outcome of hundreds of small decisions with respect to land use and water management. Better approaches need to be applied to stop the trend of incremental degradation of wetlands and move onto a path towards wetland restoration. Evidence-based arguments, with sound scientific underpinnings, are needed to demonstrate that neither a given activity, nor the cumulative effects of multiple activities, will result in loss of wetland functions.

The purpose of this paper is to synthesize existing science, tools and experience which can be applied in the near term to improve outcomes for wetlands in areas that are currently experiencing intense development pressures. Experience includes visiting wetlands that have had poor, post-development outcomes, and reviewing dozens of studies intended to evaluate the effects of proposed developments. However, specific wetlands and reports are not referenced in this paper. Practitioners perform their roles, in both preparing and reviewing studies, conscientiously. There are many reasons for the poor outcomes, some of which become clear with the benefit of hindsight. The paper provides a rationale for the emphasis on hydrological alterations, and describes knowledge and approaches that can be applied and further developed to support land use and water management decisions with better outcomes for wetlands. The paper also presents some misconceptions that can lead to poor outcomes for wetlands, such that they may be avoided or recognized.

Ecological functions: what needs to be demonstrated?

The Ontario PPS (2014) specifies: “Development and site alteration shall not be permitted on adjacent lands … unless the ecological function of the adjacent lands has been evaluated and it has been demonstrated that there will be no negative impacts on the natural features or on their ecological functions” [emphasis in original]. Negative impact in regard to significant wetlands is defined as: “degradation that threatens the health and integrity of the natural features or ecological functions for which an area is identified due to single, multiple or successive development of site alteration activities” (Ontario PPS 2014). In Ontario, provincially significant wetlands (PSWs) are identified through the application of the Wetland Evaluation System, which is a process that ranks the relative value of wetlands (Ontario Ministry of Natural Resources [OMNR] 2014). It does not indicate the vulnerability of PSWs to various development pressures (OMNR 2014).

Wetland functions

A brief introduction to wetland ecological functions and the factors controlling these functions will establish a basis for the discussion of hydrologic alterations and what is required to demonstrate that an activity will have “no negative impact”.

Wetland functions are the physical, chemical and biological processes occurring within a wetland. Sheldon et al. (2005) also describe these processes as interactions among the different components of the wetland and its landscape. It is recognized that these interactions occur at scales ranging from microscopic to global. Every process that occurs could be described as a separate function (Sheldon et al. 2005). A function can also be a grouping of many environmental processes.

Functions are considered a means of dealing with a subset of ecosystem attributes and processes rather than confronting the complexity of the ecosystem as a whole (Brinson 2009). It is appropriate to group environmental processes that are related and occur on similar temporal and spatial scales (Sheldon et al. 2005). Hanson et al. (2008), providing a Canadian perspective, adopt similar categories of functions to others (Sheldon et al. 2005; Brinson 2009; Maltby 2009), including hydrological, biochemical and habitat functions.

A critical point is that not all wetlands perform the same functions. Wetlands providing similar functions may not achieve the same level of performance with respect to those functions (Sheldon et al. 2005; Maltby 2009). Negative impacts can be associated with a loss of functions (even if others are gained) or with a decline in performance with respect to a function.

Wetland functions provide products and services, which have value to humans (Hanson et al. 2008). The value of the benefits provided by wetlands and other ecosystems is sometimes referred to as “natural capital” (Maltby 2009). Since value is a societal perception, it can change over time even when wetland functions remain constant, and can vary from place to place. Wetland benefits accrue to different stakeholders at different scales (Mitsch and Gosselink 2000a). For example, carbon sequestration becomes especially relevant when the cumulative influence of many wetlands over large geographic areas is considered (Brinson 2009). Analyzing wetland values requires understanding a different set of factors and the use of different methods from those required to assess wetland functions (Sheldon et al. 2005). Whereas the definition of ecological function in Ontario’s PPS extends to products and services as well as processes, the scope of this paper does not extend beyond processes.

Controls on wetland functions

Wetland functions are controlled by a number of environmental factors, both within the wetland boundary (site scale) and within the broader landscape (landscape scale). Any factor that influences how well a function is performed by a wetland can be considered a “control” or “driver” (Sheldon et al. 2005). Bedford (1999) recognized that a hydrogeologic perspective promotes a view of wetlands as elements of the landscape connected to larger-scale surface- and groundwater systems.

The position of a wetland in the landscape affects the quantity of inflows from various sources which may have distinct and important geochemical characteristics. It controls the degree to which a wetland is open to hydrologic and biological fluxes with other systems, including urban and agricultural systems (Mitsch and Gosselink 2000a). Human activities that alter the regime of natural disturbances or initiate new disturbances can affect controls and, hence, the structure and functioning of wetlands. These activities, or the disturbances they cause, may be referred to as “stressors.” For example, urbanization is a stressor that can change the pathways along which water moves to a wetland, disturbing the frequency, timing, amounts and/or quality of the water delivered.

Ecological indicators may show a proportional response (from simple linear to highly non-linear) to a change in a controlling factor, or they may show a step or threshold response, whereby minimal change occurs until a threshold is reached (Murray et al. 2003; Poff et al. 2010). In many cases, co-occurrence of multiple stressors can be expected. It may be that a few stressors can explain most of the ecosystem response, but teasing these apart from relatively unimportant factors may not be a trivial exercise (Wenger et al. 2009).

It is important to consider that stressors are not independent – they can interact. For example, vegetation communities may be better able to withstand hydrologic conditions outside preferred ranges with an optimal nutrient regime than with nutrient conditions near the boundary of their preferred range (Wheeler et al. 2004).

Hydrology plays a principal role in wetland functions – for example, through effects on the sources and amounts of water inputs and through effects on biogeochemical processes (Hill and Devito 1997). There are other controls on wetland functions, such as the size and distribution of habitat patches in the surrounding landscape. Fragmentation – reduction in habitat area or changes in the spatial configuration of what remains – has negative consequences for wetland functions (Sheldon et al. 2005).

Hydrologic regime of wetlands and metrics with ecological relevance

Demonstrating that there will be no negative impact on wetlands from hydrologic alterations is necessary. It is, of course, recognized that this alone may still be insufficient to demonstrate no negative impacts on wetland functions.

Wetlands are dynamic ecosystems. The hydrologic regime, characterized by intra- and inter-annual variability, plays a large role in the biotic composition, structure and function of wetland ecosystems. Even slight changes in hydrologic regime can result in alteration of wetland processes (e.g. oxidation of organic soils, nutrient dynamics) and species composition (Nilsson and Svedmark 2002; Acreman and Miller 2007).

The hydroperiod, or seasonal water level pattern, of a wetland has been described as a wetland’s “hydrologic signature” (Mitsch and Gosselink 2000b). Water level fluctuations reflect a change in the amount of water stored within a wetland in response to differences between the quantity of inflows and outflows over time. Certain aspects of the hydroperiod may be particularly important for some wetland functions (e.g. duration of inundation in spring for amphibian breeding; duration of soil aeration for many wetland tree species). The hydroperiod illustrates the intra-annual variability of conditions, with daily or monthly values plotted as means or typical ranges. The frequency and duration of high and low conditions have important influences (Warner and Rubec 1997), both positive and negative, on wetland ecosystems as well.

A number of studies have shown that mean conditions alone are insufficient to explain vegetation distribution. Other measures, such as magnitude and duration of high water level during the growing season, are often a stronger determinant (Loheide and Booth 2011). Duval et al. (2011) found duration of saturation at the beginning of the growing season to be an important metric in a study of southern Ontario fens. Prolonged initial saturation likely allows sedges, which have extensive aerenchyma for transporting oxygen to their roots, to complete most of their annual life cycle early in the growing season, thereby outcompeting more robust herbs and shrub species that require greater peat aeration for growth (Duval et al. 2011).

Richter et al. (1996) proposed the Indicators of Hydrologic Alteration (IHA) approach to statistically characterize the within-year and between-year temporal variability in hydrologic regimes using metrics that quantify the magnitude, timing, duration, rate of change and frequency of water conditions. The metrics include monthly mean conditions but also other conditions, such as high and low levels of various durations. A time series of daily water conditions is needed for analysis. The time series should be long enough to include cycles of dry and wet years. Hydrologic models can be used to generate the long-term time series provided that data is available for model parameterization and calibration.

In contrast to streams where emphasis is on the flow regime, in many wetlands it is the water level regime that is important. In systems relying on surface expressions of groundwater and surface water, depth of inundation is usually the major controlling factor for vegetation (Wheeler et al. 2004; Eamus et al. 2006). For systems reliant on subsurface expressions of groundwater, the depth to the groundwater table is often the focus. In the latter case, the inherent assumption is that water table depth is an adequate descriptor of the soil water regime (Wheeler et al. 2004). It is related to the true controlling variables and studies have shown it to be an adequate surrogate (Wheeler et al. 2004), although other factors, such as soil properties, can have an important influence (Lowry and Loheide 2010).

Where the knowledge exists, the hydrologic requirements for specific ecologic functions should be identified (i.e. the acceptable range of values for specific hydrologic metrics). Our knowledge of hydro-ecological relationships is growing, but quantitative, process-based understanding remains elusive (Hunt and Wilcox 2003), and it is not appropriate to focus on the needs of a few species, where knowledge exists, to the exclusion of other species and wetland processes in general. Use of a broad set of metrics that characterize aspects of the hydrologic regime including the magnitude, duration, timing, rates of change and frequency of water levels may be helpful in this regard.

Predicting hydrological changes and implications for ecological functions

Hydrologic changes may be caused by a multitude of activities such as aggregate extraction, groundwater and surface water takings, and urban development and stormwater management practices. Understanding how a wetland works in a hydrological sense and how activities, at both site and landscape scales, will affect the wetland water regime are starting points for impact assessment. Where potential hydrologic changes are identified, it must be shown that the wetland feature and its ecological functions are not sensitive to the changes or that the changes can be controlled to prevent a negative impact. It is critically important to recognize the full range of hydrologic changes that may have ecological relevance. The possibility for autogenic feedbacks that may amplify or dampen changes in the hydrologic regime of a wetland should be recognized (Waddington et al. 2015).

Building sound conceptual understanding of hydrology

Acreman and Miller (2007) identify the need for a consistent and robust assessment framework for diverse and, often, complex wetland systems. Chief among the elements of such a framework is building a sound conceptual understanding of the processes controlling a wetland’s water regime, which includes the application of methods such as the wetland water balance to test and refine the understanding.

It is appropriate to begin with a conceptualization of how different components of the hydrologic cycle interact with the wetland. The preliminary conceptualization can be based on available information. Recognition of hydrologic wetland typology may be helpful in identifying water transfer mechanisms. For example, the Ramsar Convention Secretariat (RCS 2010) adopts a hydrological wetland typology based on landscape location, and provides cross-sectional illustrations showing water transfer mechanisms that may be operative in different wetland types. In the Canadian Wetland Classification System, wetland forms are distinguished on the basis of features such as surface morphology and patterns, while recognizing that hydrological processes resulting from water transfers are important determinants of these forms (Warner and Rubec 1997).

The conceptualization should include spatial and temporal variability in the presence or dominance of water transfer mechanisms. In a study of calcareous fens, Duval and Waddington (2011) found high within-wetland spatial variability in hydrologic and biogeochemical factors. This serves as an important reminder that different areas within a wetland may have different landscape-scale controls and critically important site scale factors.

Temporal variability should consider not only the mechanisms that transfer water into and out of the wetland, but also the water stored within the wetland. Imbalances between inputs and outputs to a wetland are reflected by changes in water stored on the surface or in the subsurface.

The outcomes of such an analysis may include diagrams showing water transfer mechanisms, for a wetland or different zones of a wetland, and under various conditions, such as dry and wet periods. The analysis may include hypotheses with respect to the relative magnitudes and importance of water transfer mechanisms under various conditions. For many wetlands, a key tool to confirm and refine the hydrological conceptualization will be a wetland water balance.

Wetland water balances

Balancing inputs, outputs and changes in storage provides a quantitative test of hydrologic understanding. An approximate balance does not confirm the hydrologic understanding, but does confirm that it is plausible (Acreman and Miller 2007). If an approximate balance is not achieved, a potentially significant water transfer may have been omitted, or not estimated accurately, or the cumulative errors in measurement may be too high (although if errors cancel out, this may not be recognized).

This analysis requires each term of the water balance equation to be independently estimated. This may involve calculation using available data (e.g. to estimate evapotranspiration), modelling (e.g. to estimate catchment runoff) or measurement (e.g. to estimate surface outflow). Terms should not be estimated as the residual of the water balance (Winter 1981) because the estimate will include the accumulated errors (of measurement and omission). Although simple in principle, quantifying a wetland water balance is rarely straightforward. Instrumentation necessary for direct measurement or more rigorous estimates of wetland evapotranspiration is costly. Quantification of water transfers, such as groundwater exchanges, requires some field data. Wetlands often have numerous, indistinct channels providing connections with watercourses, such that even surface flows can be difficult to measure.

Explicit definition of the spatial boundaries for the wetland water balance, including the vertical dimension, is a key to sound analysis. The analysis should cover time periods when different transfer mechanisms are operative. Insights can be gained by performing water balance analyses for different periods (e.g. wet and dry periods) and examining higher resolution data (e.g. diurnal water level fluctuations).

Quantification of the water balance will be uncertain; explicit presentation of uncertainties can help to avoid poor decisions (Winter 1981). It is preferable to estimate the uncertainty associated with each water transfer mechanism (e.g. by using different methods to quantify a term) so that efforts to refine the analysis can be directed to better measurement of more uncertain terms (Acreman and Miller 2007). The conceptual hydrological model can be refined through these efforts to “close” the water balance.

Other techniques for confirming and refining the conceptual understanding

Thermal and geochemical approaches can be used to refine hydrological understanding. For example, in a study of a fen in a Danish river valley, Johansen et al. (2011) initially identified alternative conceptualizations of the interactions of a wetland with components of the hydrologic cycle, specifically groundwater flows. They were able to reject one of the alternatives based on the pH of water in various locations.

In a second example, Rains et al. (2006) were able to refine a conceptual model of vernal pool hydrology using electrical conductivity measurements. They found that vernal pools were not simply isolated depressions that ponded due to direct precipitation and drained largely due to evapotranspiration. Conductivity did not increase over time, indicating that water flowed through the vernal pools providing an ongoing source of fresh water.

In many cases, more sophisticated surface water, groundwater or integrated models can help to improve understanding, and in some cases may be necessary tools. A sound conceptual understanding will inform model selection and focus data collection efforts. If wetland water levels are tied to groundwater level fluctuations which are controlled by processes outside the wetland boundaries, a wetland water balance will not be an adequate tool. In such cases, the water table is not fluctuating in response to imbalances between inputs and outputs to the wetland, but rather is controlled by broader, landscape-scale processes.

It is a challenge for a single model to represent landscape-scale controls (e.g. regional or intermediate groundwater flow system) and site-scale controls (e.g. wetland soils and local geological heterogeneities). Nevertheless, models can be essential tools for confirming that the conceptual model is plausible, and for subsequent analysis of the potential effects of a proposed development.

Predicting hydrologic changes

Potential effects of development or site alteration

With a sound understanding of how the wetland works in a hydrological sense, it is possible to examine how a proposed activity will affect the wetland water regime. Which water transfer mechanisms are likely to be affected, and which may be affected? How will they be affected – magnitude, timing, quality? Could the proposed activity change the types of transfers or mechanisms controlling exchanges (e.g. outflow governed by unsaturated flow rather than saturated flow processes)? Could the change of land cover on adjacent lands affect the micrometeorological conditions controlling wetland evapotranspiration?

Hydrologic alterations

The wetland water balance (i.e. spreadsheet-type analysis) can be expanded to examine development (and mitigation) scenarios. This can reveal how sensitive the wetland water levels may be to changes in various inflows and outflows. Note that, in southern Ontario, it is common to use rainfall runoff models of the catchment to estimate inflows to a wetland from overland runoff (and to predict changes resulting from development). This is a means of estimating one of the terms of the wetland water balance, not a model that includes the wetland as described below.

It may be appropriate to expand the time period and increase the temporal resolution of the water balance analysis. For example, changes in surface water storage determine the hydroperiod for inundated wetlands. To ultimately connect the assessment of hydrologic alterations to ecological responses, it may be important for the analysis to be able to determine changes in the depth and duration of ponding (in days). So, the time period and resolution for the analysis may be informed by the ecological functions of the wetland.

Wetland studies have used a broad range of modelling tools. A number of wetland modelling studies have linked multiple models or utilized integrated models (e.g. Gasca and Ross 2009; Rayburg and Thoms 2009; Thompson et al. 2009). The conceptual hydrologic understanding of the wetland will assist in the selection of an appropriate model. Different combinations of surface, unsaturated zone and saturated zone processes may be important in various wetlands, and models capable of coupling these processes offer advantages. However, data are needed for model parameterization, calibration and validation. It may be difficult to satisfy the data requirements of a more sophisticated model.

Changes in wetland water levels on the order of 0.1 to 0.3 m may result in changes in the vegetation communities of some wetlands (e.g. Wheeler et al. 2004; Aldous and Bach 2015); this is well below the accuracy considered to be acceptable in most modeling applications. Regardless, models can be effective tools for understanding the types and relative magnitudes of changes associated with proposed development and mitigation scenarios. In other words, a model may not be able to predict the absolute change in water levels for a particular development scenario with the desired certainty; however, it may be useful to indicate that scenario A is preferable to scenario B.

Long-term meteorological data sets can be used with models to simulate wetland water levels over longer time periods. The model can be run for pre- and post-development conditions, and changes to the hydrologic regime can be identified. The effect of a proposed development on hydroperiods for typical, wet and dry years can be examined. The metrics characterizing the hydrologic regime (e.g. IHA parameters) can be compared for pre- and post-development conditions. Mitigation scenarios can also be examined.

Sensitive model parameters can be varied over a reasonable range of values, and the effect on the results can be examined. The magnitude of predicted hydrologic alterations can be put into perspective by comparing them to the results of sensitivity analyses.

A wetland’s water level regime may not be sensitive to changes in water transfer mechanisms that make a small volumetric contribution to the water balance. However, a wetland’s water quality may be highly sensitive to such changes. For example, an inflow of groundwater may not be important for the quantity of water it delivers to the wetland, but it may have much higher concentrations of certain chemical constituents (than precipitation and surface inflows) and thus be essential for delivering the mass of these constituents and maintaining wetland water quality. Small additions of groundwater may also degrade water quality in a wetland.

Implications for ecological functions

Ideally, predicting hydrological changes would not be the end point of an impact assessment. An ability to predict the ecological responses to hydrologic alterations would support better decision making and allocation of resources. There is a very solid, general knowledge of how ecosystem processes, structure and function depend on hydrology. However, models that directly predict ecological responses to various types and degrees of hydrologic alteration are not readily available (Poff et al. 2010). Information exists on the specific hydrologic needs for some wetland species and processes but does not exist for a broad range of functions. The deviation beyond the range of hydrologic variability characteristic of a system which can be tolerated is also largely unknown.

Conceptual ecohydrological models that link hydrologic changes and wetland processes (e.g. retention and transformation of nutrients) and biota which may respond can be valuable. Their development also promotes interdisciplinary knowledge exchange and collaboration, and can be initiated early in the process of evaluating development impacts. They can help to identify potential autogenic feedbacks that may influence the ecological responses to hydrological alterations due to development. An interpretation that there will be no negative impact to a wetland feature or its functions, despite hydrologic alterations, must be evidence based. Scrutiny of this evidence is merited; when wetlands are lost, or functions are lost, through development, these losses are often irreversible (Bedford 1999; Mitsch and Gosselink 2000a).

In summary, to demonstrate that a proposed development will not result in hydrologic alterations with negative impacts to a wetland feature or its associated ecological functions, it is necessary to convey how the wetland works in a hydrological sense and how activities, at both site and landscape scales, will affect the wetland water regime. Verification of wetland water balances or models will provide evidence of the adequacy of the understanding of wetland hydrology. Where potential hydrologic changes are identified, evidence-based arguments are needed to show that the wetland feature and its ecological functions are not sensitive to the changes or that the changes can be controlled to prevent a negative impact. Conceptual ecohydrological models can help in the identification of hydrologic changes, and temporal and spatial scales, with ecological relevance.

Inadequacies in evaluations

This discussion focusses on two areas: challenges related to understanding the role of groundwater, and understanding the limitations of stormwater management practices. Incomplete or faulty knowledge, and/or flaws in reasoning, can result in misleading or unsound arguments. These arguments may be accepted, resulting in poor decisions.

Misconceptions related to the role of groundwater

There has been an acceleration of activity in the area of hydrogeoecology in recent years (e.g. Hancock et al. 2009; Klove et al. 2011) and many excellent studies of groundwater–wetland interactions (e.g. van der Kamp and Hayashi 2009). However, the body of knowledge is quite specialized, and contributions to close the gap between research and practice in this area are needed to prevent impacts to groundwater-dependent wetlands. Several conceptual errors related to the role of groundwater, which have been observed in wetland studies, provide structure for the following discussion.

Wetland is predominantly supplied by surface water

It is often determined that wetlands are predominantly supplied by surface water. This does not, however, mean that these wetlands are not dependent on groundwater. Wetlands may be dependent on groundwater even where the net groundwater contribution to the annual wetland water balance is small.

Groundwater inflows may be a small proportion of the total inflows to a wetland on an annual basis, but may comprise a substantial proportion during the dry season. This dry season supply may be important to prevent accelerated organic soil oxidation (Mitsch and Gosselink 2000b) and to sustain evapotranspiration rates (Lowry and Loheide 2010) which have a moderating effect on local climate. Even small amounts of groundwater during low-water conditions may be significant (Acreman and Miller 2007).

Despite small volumetric contributions, groundwater may be a concentrated source of chemical constituents, such as dissolved calcium, which influence the distribution of vegetation and the suitability of the habitat for some rare wetland flora (Duval et al. 2011; Johansen et al. 2011).

There may be cases in which the volumetric inflow of groundwater is substantial, but is approximately balanced by groundwater outflow over most of the year. A change in the amount of groundwater flow-through may not affect wetland water levels, but may affect wetland water quality. For example, groundwater flow-through may prevent some vernal pools from becoming increasingly concentrated as water is removed via evaporation (Rains et al. 2006).

No seeps are observed and/or the aquifer heads are below the wetland surface

Neither an absence of visible seepage on the wetland surface nor a piezometric head which is below the wetland surface provides sufficient evidence that groundwater fluxes are unimportant to the wetland water balance. The timing of the observation(s) is important. In some wetlands, groundwater discharge to the wetland surface may occur, but only for a short period of time in most years.

Groundwater flow is three dimensional. Water may move laterally into the subsurface soils of the wetland from an adjacent aquifer. A vertical component of flow to deliver groundwater to the subsurface soils of the wetland requires the hydraulic head at depth below the wetland to be above the water level in the wetland, but not necessarily above the ground surface in the wetland.

Eamus et al. (2006) recognize Type II groundwater-dependent ecosystems (GDE) which are dependent on the surface expression of groundwater, but also Type III GDE, which are dependent on the subsurface presence of groundwater. The presence of groundwater, above the extinction depth, provides a source of water to replenish soil moisture depleted through evapotranspiration (Gasca and Ross 2009; Lowry and Loheide 2010). The subsurface presence of groundwater can influence linkages with uplands and overland runoff processes (Todd et al. 2006). It can also influence the flow and distribution of water within a wetland; this has implications for the effectiveness of mitigation measures that rely on diffuse discharges of water to replace natural exchanges.

The presence of fine-grained deposits limits exchanges

The location of aquifers and lower hydraulic conductivity deposits, relative to a wetland, is important to the development of the conceptual hydrologic model. Low-hydraulic-conductivity units can restrict subsurface exchanges. However, there are several points that merit further consideration.

The hydraulic conductivity of clay tills (e.g. Halton Till in southern Ontario) at a particular location depends on the degree of fracturing. Hydraulic conductivity can be considerably higher in the zone of oxidation above the average position of the groundwater table. When the groundwater table is high, there can be increased interactions between wetlands and shallow groundwater. This can also provide subsurface hydraulic connectivity between wetlands (van der Kamp and Hayashi 2009) at certain times of year. In general, it is the site-scale hydraulic conductivity which is relevant to quantifying groundwater exchanges with wetlands, and this should be taken into account when selecting testing methods and interpreting results.

Some wetlands are located in complex hydrogeological settings. In tills that are predominantly fine grained, groundwater may find higher conductivity pathways to a wetland in the presence of hydraulic gradients driving flow towards the wetland. Sampling may or may not intercept these paths; sampling can confirm the presence, but not the absence, of higher hydraulic conductivity materials. The interpretation that a continuous low-hydraulic-conductivity layer exists, restricting exchanges, should be consistent with the depositional history of the geologic materials.

Finally, it is well established that as the volumetric water content drops, the unsaturated hydraulic conductivity of fine-grained materials surpasses that of coarse materials (e.g. Fetter 2001). Hence, fine-grained deposits can enhance the transport of water from the groundwater table to the root zone of wetland vegetation. Similar processes can also influence the rate at which the depth of ponding declines; as soil moisture is depleted by evapotranspiration of vegetation surrounding a ponded area, it can be replenished by water from the wetland pool.

Misunderstanding the function of stormwater management ponds

Maintaining the proportions of supply from different sources and the pathways along which water is delivered to wetlands in order to avoid alterations to the magnitude, timing and quality of inflows to a wetland is a tall order. In many cases, proposed mitigation measures are best efforts to fulfill the desire for development and protection of natural features. That wetlands continue to be degraded suggests that our best efforts are not good enough. Advances in mitigation measures, based on a clear understanding of what is needed as well as an appreciation of the limitations of existing approaches, must be pursued.

Urban developments, for which there is intense pressure in many jurisdictions in southern Ontario, are a clear threat. In some cases, extreme and clearly unacceptable hydrologic changes have occurred, with disastrous consequences for wetlands. These failures should have been avoidable with the current state of knowledge, analysis tools and mitigation measures. Wetlands can experience too little water in cases where runoff is bypassed around a wetland, or they can receive too much water. The discussion focusses on the latter case, which arises from the misunderstanding that stormwater management ponds are designed to prevent (all) hydrologic alterations such that there will be no ecological effects.

Stormwater management objectives have evolved. In addition to the historical objective of controlling flooding, objectives related to addressing the water quality and erosion of receiving streams are now recognized. Stormwater management should also mitigate changes to the water balance of a site, helping to maintain groundwater and baseflow characteristics. Reducing the volume of runoff also contributes to achieving flood, erosion and water-quality control objectives (Bradford and Gharabaghi 2004).

Stormwater management measures capable of mitigating changes to the water balance are essential elements of stormwater management plans in some jurisdictions. Other jurisdictions continue to rely heavily on “end-of-pipe” ponds to manage the hydrologic effects of urbanization. Stormwater ponds provide storage and allow the volume of inflows to be discharged over a longer time period at a lower flow rate. However, they ultimately release the volume of runoff received with little modification (Graham et al. 2004). In other words, ponds are effective for peak flow control but not for volume control.

Urbanization results in much higher volumes of runoff. Even where infiltration practices are used to achieve pre-development levels of infiltration, runoff volumes are expected to be higher because there is less evapotranspiration from urbanized catchments (Burns et al. 2012). An increased volume of runoff from urban development(s) to a wetland can have a major effect on the wetland water balance, particularly when the development occupies a large portion of the wetland’s catchment. The sensitivity of the wetland water level regime to increased inflows depends on whether increased outflows also occur.

Isolated wetlands do not have surface water outlets and are particularly vulnerable. The amount of water stored in the wetland, and hence the water level, can increase (year after year) until the elevation of an outlet is reached. Increasing depths of inundation will result in major shifts in vegetation communities, and a cascade of other ecological responses.

Non-isolated wetlands may also be vulnerable. For example, a small stream passes through a silver maple swamp and discharges to a larger stream. The area contributing flow to the small stream is developed and a stormwater pond is constructed to manage the stormwater before release to the stream. However, the increased volume of inflows to the swamp is not balanced by increased outflows. Discharges from the pond result in water levels in the small stream which are high enough to spill over the banks and into (lower) wetland areas, with no surface pathway to the outlet. The stormwater pond is functioning as designed – reducing peak flow rates to pre-development levels. However, these peak flow rates occur more frequently after development. In the case of this wetland, the more frequent flooding results in a cumulative volume of water entering the (lower) wetland areas which cannot be removed through infiltration and evapotranspiration. Soils now remain saturated year-round and the vegetation, including the tree species, responds to this change. In this case, the design objectives were not correctly defined. Where a wetland and its ecological functions need to be protected, managing changes to the water balance of a development site is a particularly important objective. It should be recognized that additional, more specific, design criteria are often needed to protect wetlands.

There is too often a gap between the disciplines; the ecologists perhaps assuming “hydrology” is the domain of the water resources engineers and the engineers perhaps assuming that the “wetland” is the purview of the ecologists. Rather than passing the baton between members of the team, with unstated assumptions about who is responsible for what, there needs to be a strong collaborative effort to develop design criteria that, if met, can prevent the loss of wetland features and their functions.

Conclusions

The purpose of this paper was to synthesize existing science, tools and experience which can be applied in the near term to improve outcomes for wetlands in areas that are currently experiencing intense development pressures. Action on several fronts could help to reverse the trend of incremental wetland degradation in southern Ontario. For example, Ontario’s Provincial Policy Statement focuses on Provincially Significant Wetlands (PSWs). Wetlands that are not PSWs may still be important (i.e. provide services of value). Further, many small wetlands are unevaluated and these systems, too, provide services, particularly when their aggregate contributions are considered on a watershed scale. There are also research needs – for example, to advance our understanding of ecological responses to hydrologic alterations, and to advance hydrologic models and analysis tools to improve capabilities for predicting small hydrologic changes that may have ecological relevance.

Still, there are ways to improve outcomes for wetlands in the near term, using existing science and tools. Sound, evidence-based arguments, with rigorous scientific underpinnings, can be provided and demanded. For example, it is important to recognize the range of roles that groundwater can play, even if it makes a small volumetric contribution to the wetland water balance. It is also important to recognize the limitations of stormwater management ponds and the potential need for specific design criteria for stormwater management systems, in order to prevent impacts to wetlands and their ecological functions. Other misconceptions should be identified, so that they too can be recognized and avoided by others.

Although it may not be sufficient to demonstrate that a proposed development will have no negative impact on a wetland or its ecological functions, a hydrological analysis will often be required. A sound understanding of a wetland’s hydrology is needed to support claims that a development (with or without mitigation practices) will result in no hydrological alterations. There are ways to carry out wetland water balances, and make use of other tools (e.g. models and geochemical analyses), to demonstrate a good understanding of wetland hydrology. It is important to recognize that a range of hydrologic alterations may have ecological relevance (rather than, for example, just mean monthly conditions). There should be no illusion that these are simple analyses to perform. On the contrary, it can be difficult to collect data and model wetland systems. Nevertheless, simpler analyses that do not achieve desired outcomes must be challenged.

Water resources practitioners can collaborate with ecologists to identify ecologically relevant hydrologic metrics, and to identify potential ecological responses to hydrologic alterations. Development of a conceptual ecohydrological model for a wetland may be a useful way to structure interdisciplinary work. Interdisciplinary efforts are much more likely to reveal the processes that can explain observed responses in wetlands and support development of predictive tools. Where knowledge is lacking with respect to ecohydrological relationships, mitigation measures may need to prevent the alteration of a range of hydrologic metrics characterizing the timing, duration and frequency of water levels. Advances in the design of mitigation measures to achieve a broader set of objectives will be another key to preventing wetland degradation. It is important to learn from successes and failures to ensure the latter become a rare exception.

Acknowledgements

Review of the literature, which was essential to this paper, was done for a variety of other projects funded by the Ontario Ministry of Natural Resources, and the Grand River, Toronto and Region and Credit Valley Conservation Authorities. Thanks also to the many individuals who have provided documents which have been helpful in understanding the state of practice in Ontario, and to the reviewers who provided comments on the paper.