3 Cumulative Effects Assessment

3.1 Cumulative effects

The cumulative effects assessment (CEA) is rooted in natural resource management and environmental impact assessment processes (Halpern and Fujita, 2013) and has appeared in the legislation in certain countries for decades, including the United States (National Environmental Policy Act 40 CFR 1508.7, 1969), Canada (Canadian Environmental Assessment Act S.C. 1992, c. 37, now the Impact Assessment Act 2019 22 (1)(a)(ii)), Australia, and various European countries (Halpern and Fujita, 2013). The scientific and grey¹ literature thus abounds with publications providing definitions, implementation frameworks, best practices guides, and guiding principles for cumulative effects assessments in environmental impact assessments (Hegmann et al., 1999; Krausman and Harris, 2011; e.g. Peterson et al., 1987). Therefore, a wide variety of definitions and approaches exists (Duinker et al., 2013; see Therivel and Ross, 2007).

In its broadest sense, the Canadian Council of Ministers of the Environment (Conseil canadien des ministres de l’environnement, 2014) defines cumulative effects as a “change in the environment caused by multiple interactions among human activities and natural processes that accumulate across space and time.”* Cumulative effects assessment is defined as “a systematic process of identifying, analyzing and evaluating cumulative effects”*.

These definitions point to the systematic nature of cumulative effects and their assessment, i.e., management of ecosystems and the interactions that structure them as a whole, including human aspects, as related to ecosystem-based management (Christensen et al., 1996; Leslie and McLeod, 2007; Rosenberg and McLeod, 2005). However, cumulative effects are defined in more restrictive terms in policy, legislative, and regulatory texts (Jones, 2016). The 1992 Canadian Environmental Assessment Act defines cumulative effects as follows:

“[…] the cumulative environmental effects [of a project] that are likely to result from the project in combination with other projects or activities”.

Cumulative effects assessment processes are thus typically conducted from a project perspective rather than a systemic perspective. A number of scientists thus call for the use of regional approaches to the cumulative effects assessment (e.g. Dubé, 2003; Duinker and Greig, 2006; Jones, 2016; Sinclair et al., 2017). Moreover, Canada’s Minister of the Environment and Climate Change recently authorized a regional assessment of the St. Lawrence River area on July 15, 2021, in response to a request by the Mohawk Nation of Kahnawà:ke under the Impact Assessment Act. According to Sinclair et al. (2017), a regional cumulative effects assessment can be defined as follows:

”[A Regional Effects Assessment (REA)] is an [environmental assessment] whose primary or sole defining feature is its regional scope and its focus on understanding the interactions between human activities and the natural world. This means that in just about all aspects other than its spatial limitations, an REA should be comprehensive and integrated.”

Similar to a project-based approach within environmental impact assessments, the regional approach focuses instead on the total effects (effect-based) and viability of environmental receptors of interest, commonly referred to as valued components (Beanlands and Duinker, 1983; Sinclair et al., 2017). Several constraints limit its practical application, especially the need to use environmental measures – e.g. variation in the mortality rate of a species as a function of an environmental stressor – to establish the effects of one or more stressors on a valued component. These types of measures are particularly difficult to obtain in cumulative effects assessments; they also necessarily make assessments reactive since effects must be observed in order to establish a meaningful link between the presence of a stressor and the degradation of a valued component (Dubé, 2003). An ideal approach would instead make it possible to combine effects-based and stressor-based cumulative effects assessment approaches (Dubé, 2003; Sinclair et al., 2017).

3.2 Method

A study by Halpern et al. (2008) took a major step toward a systemic and spatially explicit assessment of the cumulative effects of human activities on the world’s oceans by assessing the cumulative effects of 17 environmental stressors on 20 types of marine ecosystems. This study showed that few environments remain unmarked by human activities and that most ecosystems are affected by multiple environmental stressors. The updates published by the same group in 2015 (Halpern et al., 2015) and 2019 (Halpern et al., 2019) also show that cumulative effects on the oceans are increasing globally. In addition to being used worldwide, this method has been used many times in different regions of the globe to characterize cumulative effects, including in California (Halpern et al., 2009), the Arctic (Afflerbach et al., 2017; Andersen et al., 2015), and the Canadian Pacific Ocean (Ban et al., 2010; Clarke Murray et al., 2015b, 2015a; Singh et al., 2020). At the time of writing, an assessment using a similar approach is underway for the Placentia Bay, Newfoundland and Labrador pilot area as part of Transport Canada’s national initiative Cumulative Effects of Marine Shipping. Fisheries and Oceans Canada is also conducting an assessment for the Maritimes Region and is investigating the general applicability of the approach to characterize cumulative effects on Canadian oceans.

3.2.1 General model

The method developed by Halpern et al. (2008) allows for the combination of the diversity of valued components, stressors, and vulnerability of each valued component to each stressor. The method requires three types of data: 1) the mapped presence or absence of the valued components (\(CV_i\)), 2) the spatial distribution and relative intensity (i.e. normalized between 0 and 1) of the environmental stressors considered (\(S_j\)), and 3) the relative vulnerability of each valued component to each stressor (\(\mu_{i,j}\)) These data are then incorporated into a grid made up of cells of homogeneous size characterizing the study area. Cumulative effect predictions (\(E_C\)) are calculated for each cell (\(x\)) of the grid by summing up all individual stressor effects over the set of valued components:

\[E_{C_x} = \sum_{i=1}^n \sum_{j=1}^m CV_{i,x} * S_{j,x} * \mu_{i,j} \label{eq1}\]

This method proposes the calculation of a relative indicator of cumulative effects to predict the risks associated with the effects of multiple environmental stressors on valued components. The term “relative” is central to understanding the proposed assessment method. An absolute indicator would identify a change in the state of valued components resulting from cumulative environmental stressors, such as a decline in the beluga whale population in the St. Lawrence Estuary in response to cumulative environmental stressors. A relative indicator would instead allow for a comparison of the various environmental stressors according to their intensity within the region studied and their effects on the valued components. The results must therefore be interpreted as a probability of risk of the marine vessel activities to the valued components considered. The results of the cumulative effects assessment using this approach are presented in section 5.4 of the report.

The calculated cumulative effects predictions can be broken down to assess the relative share of the effects of a single or multiple stressors on one or more valued components (Figure 3.1). For example, predictions of cumulative effects could be broken down to explore cumulative effects on all marine mammals, on a single species, or to identify regions where marine mammals are most at risk in the study area. A complete exploration of all stressor-valued component combinations is possible (Figure 3.1); this offers the ability to analyze different management scenarios and lead to clear and focused recommendations in order to optimize and prioritize management efforts in a region of interest (Halpern et al., 2015). It also offers a flexible and quantitative approach that can integrate different types of data that are sometimes difficult to compare, such as biophysical data with qualitative knowledge (Halpern et al., 2008, 2015; Halpern and Fujita, 2013).

Figure 3.1: A fictitious example of spatial assessment of cumulative effects using the methodology suggested by (Halpern et al., 2008). The assessment begins by delineating a study area of interest (A). A picture of the study area is then made by characterizing the distribution of environmental stressors (B) and the valued components (C) allowing the assessment objectives to be reached. The summation of all the environmental stressors allows us to identify the environments that are most exposed to cumulative stress, i.e. cumulative exposure (D). The sum of valued components, on the other hand, allows us to identify the environments in the study area where a higher number of valued components overlap (E). By combining the distribution of environmental stressors and valued components along with the vulnerability of the valued components to the environmental stressors, a relative assessment of individual effects is obtained (F). It is possible to assess the impact of all the environmental stressors on a single valued component (G); similarly, it is possible to assess the impact of a single environmental stressor on all valued components (H). Finding the sum of all the individual impacts provides the relative assessment of cumulative effects incorporating all combinations of environmental stressors and valued components (I).

3.2.2 Partial models

The general cumulative effects assessment model can easily be broken down into partial models to explore simplified portions of the assessment. We therefore present three additional indicators allowing the results of the assessment to be explored.

3.2.2.1 Cumulative stressors and hotspots

Considering only the spatial distribution and relative intensity of the environmental stressors considered (\(S_j\)) provides an assessment of the cumulative stressors \(E_{S,x}\) (Beauchesne et al., 2020), which corresponds to the total of all the environmental stressors \(S\) in each cell \(x\) considered for the assessment:

\[E_{S_x} = \sum_{j=1}^m S_{j,x} \label{eq2}\]

The cumulative stressor assessment of environmental stressors identifies the sites that are most exposed to the potential effects of the environmental stressors considered for the cumulative effects assessment. It does not provide an effects assessment, since it only considers the stressors, and their intensity and distribution. However, it does allow us to assess the environments that are most likely to be affected by the stressors in our study area.

In order to limit an over-representation of stressors composed of several subcategories, it is also possible to normalize the intensity values of a stressor category by the number of sub-categories that compose it. For example, a stressor consisting of 10 categories would have a higher weight than a stressor consisting of only one category. To normalize the intensity values, the intensity of these subcategories need only be divided by the number of subcategories, e.g. for a category composed of 10 subcategories, the intensity is divided by 10.

Using the cumulative stressor assessment, it is also possible to obtain an assessment of cumulative stressor hotspots, which identify environments where environmental stressors co-occur at high relative intensities (Beauchesne et al., 2020). The cumulative hotspots \(E_{H_x}\) is calculated in each cell of the study grid and corresponds to the sum of stressors whose intensity is contained in their respective \(80th\) percentile:

\[E_{H_x} = \sum_{j=1}^m \mathbb{1} (S_{j,x} \; \epsilon \; P_{80, S_j})\]

where \(P_{80, S_j}\) is the \(80th\) percentile of stressor \(j\). The results of normalized and non-normalized cumulative stressors and cumulative hotspots are presented in section 5.1.

3.2.2.2 Cumulative valued components

Considering only the mapped presence or absence of the valued components considered (\(CV_i\)) provides an assessment of the cumulative stressors \(E_{S,x}\), which corresponds to the total of all environmental stressors \(CV\) in each cell \(x\):

\[E_{CV_x} = \sum_{i=1}^n CV_{i,x} \label{eq3}\]

The cumulative footprint assessment identifies the environments with the highest content of valued components. As with the cumulative stressor assessment, it does not provide an effects assessment, since it only considers the valued components. However, it does allow us to assess the environments that are most important for the valued components in our study area. The presence of valued components can also be normalized by the number of valued component categories to avoid over-representation of valued components composed of multiple categories (see section 3.2.2.1). The results of normalized and non-normalized valued components are presented in section 5.2.

3.2.2.3 Cumulative exposure

The intersection of the assessment of the stressors and the cumulative valued components provide the third partial model that we are presenting in addition to the overall model, i.e. cumulative exposure. The cumulative exposure \(E_{E,x}\) is the product of the intensity of the environmental stressors \(S_{j,x}\) and the presence of the valued components \(CV_{i,x}\) in each cell \(x\):

\[E_{E_x} = \sum_{i=1}^n \sum_{j=1}^m CV_{i,x} * S_{j,x} \label{eq4}\]

Cumulative exposure provides an assessment of environments where there is greater overlap of valued components and environmental stressors. Although this metric does not predict the effect of stressors on valued components, it does identify environments where the valued components are most likely to be subjected to the effects of the environmental stressors considered. Cumulative exposure also makes it possible to delimit all the environments where cumulative effects could occur, since stressors and valued components must overlap in order to detect an effect; as such, this metric captures a significant portion of cumulative effects even though it does not incorporate the vulnerability of valued components to stressors. As with the cumulative valued components and stressors, cumulative exposure can be normalized by the number of stressor and valued component categories to avoid over-representation of valued components composed of multiple categories (see section 3.2.2.1). The results of normalized and non-normalized cumulative exposure are presented in section 5.3.

3.2.3 Environmental stressors

In the context of the cumulative effects assessment, environmental stressors from human activities are typically characterized using the footprint of the activities themselves as an index of stressor intensity (Beauchesne et al., 2020; e.g. Halpern et al., 2019). As such, the work of characterizing environmental stressors is aimed at a direct characterization of marine vessel activities. Furthermore, each marine activity is characterized by a unique footprint and requires an appropriate approach to properly capture its intensity and distribution. For example, a method for characterizing navigation would not be appropriate for characterizing dredging activities. Individual approaches were thus used to appropriately characterize the different environmental stressors. In addition, some environmental stressors can be divided into subcategories with different spatial footprints or different effects on the valued components. For example, ferries do not use the same sailing routes as oil tankers, and trawling does not affect ecosystems in the same way as gillnetting. We have thus divided certain stressors into subcategories. The characterization of environmental stressors is part of the study area picture and is presented in section 4.2 of the report; this section presents the data and approaches used to characterize environmental stressors in the study area.

The diversity of environmental stressors considered for the cumulative effects assessment also requires data transformation; indeed, marine traffic intensity expressed as a number of navigation transits cannot be directly compared to dredging activity intensity in \(m^3\) of dredged sediment. All stressors are thus normalized between 0 and 1 to obtain a relative intensity value and thus allow a comparison between different types of stressors. The 99th quantile of the stressor intensity distribution was used as an upper threshold for normalizing extreme values (Halpern et al., 2019). The navigation categories are normalized together so as not to assign higher relative intensity values to vessel types with lower traffic. For example, individual normalization would result in a similar maximum intensity for a navigation category with a maximum traffic of 100 vessels per \(km^2\) and another with a maximum traffic of 1000 vessels per \(km^2\). A joint normalization thus makes it possible to keep these differences between types of vessels. Environmental stressor data is also log-transformed to avoid underestimation of intermediate value stressor intensity (Halpern et al., 2019).

3.2.4 Valued components

valued components are typically characterized by presence-absence in the study grid, although some may be characterized by a continuous variable. For example, marine mammals are characterized based on the likelihood of sighting a marine mammal rather than by presence-absence. This allows us to assign greater importance to environments where marine mammals are frequently sighted. As with the environmental stressors, the valued components can also be divided into subcategories to better capture their diversity. The characterization of valued components is part of the study area picture and is presented in section 4.3 of the report; this section presents the data and approaches used to characterize valued components in the study area.

3.2.5 Vulnerability

cumulative effects assessments require the assessment of valued component vulnerability to environmental stressors considered (Kappel et al., 2012; Teck et al., 2010). In general, the vulnerability of a valued component is defined based on its exposure (i.e., the likelihood that a habitat will be subjected to a stressor), its sensitivity (i.e., the degree to which a habitat will be affected by a stressor), and its adaptive capacity or resilience (i.e., the ability of a habitat or the elements composing it to recover from a disturbance) to stressors (Halpern and Fujita, 2013; e.g. Metzger et al., 2005; Teck et al., 2010).

A number of methods are described in the scientific literature to assess the vulnerability of natural attributes to natural or anthropogenic disturbances (Wilson et al., 2005). For example, Foden et al. (2011) used empirical recovery time values from the literature as a marker of benthic habitat vulnerability in the UK to apply the method of Halpern et al. (2008). However, the significant gaps in the empirical data available allowing us to characterize all valued component-stressor combinations curtail the use of a similar approach (Halpern et al., 2007; Teck et al., 2010). This type of knowledge can be particularly difficult to obtain, and entire research teams typically focus on the vulnerability of a single valued component to a single stressor; think of the vulnerability of marine mammals to underwater noise. Furthermore, while the vulnerability of certain valued components to certain environmental stressors is well documented and would allow for a robust assessment of individual environmental impacts, this type of knowledge is rarely – if ever – available for all valued component-stressor combinations yet is a necessity for conducting a cumulative effects assessment.

A qualitative approach using expert opinion and/or document review is thus typically used to generate a matrix of relative vulnerability scores for all valued components and stressors included in the cumulative effects assessment (Halpern et al., 2007; Kappel et al., 2012; Teck et al., 2010). Using a method in which experts are consulted also enables us to have the benefit of expertise and knowledge that would otherwise not be available to support management and decision-making (Teck et al., 2010; Wilson et al., 2005). Criteria for assessing the exposure, sensitivity and adaptive capacities of the valued components are thus generally qualitatively selected and assessed. For example, exposure may depend on the spatial magnitude and frequency of a stressor; the sensitivity of a species may be defined by the effects of a stressor on its reproduction; adaptive capacities may be influenced by the vulnerable status of a certain animal population. These criteria may vary depending on the valued components selected, since criteria applicable to a species (e.g. Maxwell et al., 2013) will not necessarily be applicable to a habitat (e.g. Teck et al., 2010). The criteria may thus vary from one group of valued components to another; indeed, this is the case for the various valued components considered for the assessment of this pilot project.

A qualitative vulnerability assessment is aimed at assessing each criterion selected to obtain a numerical rank per criterion. These individual assessments are then combined and normalized to obtain a relative vulnerability score ranging from 0 – insensitive – to 1 – very sensitive – for all “valued component-stressor” combinations. The following sections describe how the vulnerability of valued components to stressors was assessed in this cumulative effects assessment. The section 4.4 presents data, approaches and results of the assessment of valued component vulnerability to environmental stressors considered for the cumulative effects assessment of marine vessel activities in the study area.

---

1. Grey literature refers to the material produced by various public, commercial or industrial bodies, subject to intellectual property rules, that is not checked by the scientific peer review process.↩︎