Selection of valued ecosystem components in cumulative effects assessment: lessons from Canadian road construction projects

Abstract

Valued ecosystem component (VEC) selection is a core component of cumulative effects assessment (CEA) and gives direction to impact analysis, mitigation and monitoring. Yet little is known about CEA VEC selection practices. This paper examines 11 Canadian road infrastructure project CEAs completed between 1995 and 2011 to determine how VEC selection in CEA is performed, and whether these practices are sensitive to the linear project development context. Document review and semi-structured interviews reveal an absence of VEC selection guidance, late timing of cumulative effects considerations in impact assessment, lack of sensitivity in CEA VEC selection to the unique, linear nature of the road construction projects and a general lack of insightful, creative approaches to CEA VEC selection – ones that better reflect potential impacts to social and economic aspects of the environment – despite it being shown to be a values-driven, subjective process. There is a clear need for regional databases to support consistent CEA VEC selection processes, and the development of CEA-specific VEC selection guidance.

Keywords::

• environmental impact assessment
• cumulative effects assessment
• valued ecosystem component selection
• linear development
• road construction

1. Why study valued ecosystem component selection?

In recent years, there has been no shortage of poor practice reviews of cumulative effects assessment (CEA) in the impact assessment (IA) literature. Numerous authors have described procedural and substantive shortcomings of CEA including: ill-defined scope, scale and methodologies (Benson 2003; Duinker & Greig 2006; Gunn & Noble 2011, 2012; Morgan 2012); inadequate data to support analysis and conclusions (Cancer & Kamath 1995; Johnson et al. 2011; Noble & Gunn 2013); superficial interpretations of the nature of cumulative effects (Canter & Ross 2010; Duinker & Greig 2006; Folkeson et al. 2013) and gaps between predicted and actual cumulative effects (Johnson et al. 2011). The anxiety over CEA performance is just one facet of a broader and deeper concern for IA effectiveness worldwide; something that has been called into question by a past president for International Association for Impact Assessment (Fuggle 2005) as well as IA academics and practitioners (e.g. Duinker & Greig 2006; Cashmore et al. 2010; Bina et al. 2011), and is arguably reflected in legislative ‘streamlining’ in countries such as Canada (Gibson 2012; Owens 2012).

However, while IA effectiveness research intensified significantly in the first decade of the twenty-first century (e.g. Cashmore et al. 2004; Elling 2009; Cashmore et al. 2010), particularly in the areas of strategic assessment (e.g. Thérivel & Minas 2002; Retief 2006; Jha-Thakur et al. 2009; van Buuren & Nooteboom 2009; van Doren et al. 2013) and project-based environmental IA (e.g. Cashmore et al. 2004), similar research on CEA effectiveness has not. Evaluative frameworks for IA effectiveness have been introduced (Sadler 1996; Wood 2003) and tested (Baker & McLelland 2003) and SEA effectiveness research has examined process transparency and quality; tiering of assessment to ‘lower’, more concrete program and project decisions (van Doren et al. 2013); as well as integration of assessment with decision-making (e.g. Fischer & Gazzola 2006; Retief 2006). In contrast, very few studies of CEA effectiveness exist as yet, despite it being a leading topic in IA in the last two decades (Morgan 2012) and the numerous sessions devoted to it at recent IAIA conferences (e.g. Canter et al. 2013; Sadler & Canter 2013).

Valued ecosystem component (VEC) selection is one possible focal point for CEA effectiveness research. Introduced by Beanlands and Duinker (1983) in Canada, VECs were recommended as a means to avoid the inefficient ‘count everything’ approach that characterized early IAs (Kennedy & Ross 1992; Treweek 1999). They were also meant to direct the focus of assessors to a more holistic view of ecological health and integrity (Harwell & Gentile 2006; Barnes et al. 2010; Nunes 2010) rather than simply the effects of a particular development on ‘just water’ or ‘just vegetation’, for example (Beanlands & Duinker 1983). Similar processes exist outside the Canadian practice, e.g. the European Commission guidance on cumulative effects. A review of 50 UK environmental impact statements (EISs) by Cooper and Sheate (2002) also revealed that the CEA studies in the projects examined were centered on VECs. But while the central role of VECs in both project-based IA and its corollary, CEA (Canter & Ross 2010; Johnson et al. 2011), has long been established, surprisingly little research has been done to examine the principles, processes and rationales applied to VEC selection. This is in spite of the fact that CEA performance reviews have emphasized the importance of good scoping (of which VEC selection is a core determinant) (Bérubé 2007; Canter & Ross 2010; Senner 2011), and that various authors have pointed out that VEC selection is a key problem area in IA practice (e.g.: McCarty & Power 2000; Dowlatabadi et al. 2004; Ball et al. 2013). For example, Bérubé (2007, p. 103) noted that VEC selection for CEA presents a ‘first category’ challenge to the IA process.

Insight into VEC selection processes in IA, and CEA particularly, is surprisingly rather limited, and guidance is scant. A scan of the literature reveals some attention to the values that influence VEC choices, characterization of stressors affecting VECs, and application of the VEC concept within different assessment frameworks (see, for example, Lohani et al. 1997; Harwell & Gentile 2006). In a recent review of project-based IAs in southern Saskatchewan, Ball et al. (2013) found that aquatic VECs were usually selected to address a proponent's exposure to liability and penalty under federal legislation (e.g. Fisheries Act and Species at Risk Act) and to reflect organizational mandates of stakeholder agencies involved in the process. But in general, the principles, processes, and rationales applied to VEC selection in CEA have not been intensively studied and are not well understood (McCarty & Power 2000; Dowlatabadi et al. 2004; Bérubé 2007). In addition, almost nothing is known about specific sectoral challenges to VEC selection such as might exist for unique development contexts such as linear projects.

The road construction sector provides an excellent context within which to investigate VEC selection in CEA for at least two reasons. First, is the likelihood of major new roads and highways (and other linear developments) to induce significant cumulative social, economic and environmental effects. There is general agreement that road construction is a leading cause of adverse effects on VECs such as wildlife (Forman & Alexander 1998; McGarigal et al. 2001; Tricker 2007), and that regional VEC sustainability will be impossible if the cumulative effects of multiple road projects are ignored (CEQ 1997; Wasike 2001; Swor & Canter 2008). Second, what is learned about VEC selection in a road construction context is germane to CEA of other linear development projects (such as oil and gas pipeline corridors, electric utility transmission rights-of-way, and rail corridors), since they share many similar physical characteristics and similarly contribute to cumulative landscape and habitat fragmentation. The extensive geographic reach and longitudinal life-cycles of such projects can have particularly serious consequences, especially when constructed in previously undeveloped natural resource regions (Jaeger et al. 2007).

Despite that CEA is well-positioned to address the complex environmental changes resulting from linear (road) projects (e.g. McGarigal et al. 2001; World Bank 2005) – and that in Canada, at least, necessary legal structures have been in place to support CEA for many years – to our knowledge there has never been a detailed investigation of VEC selection processes as pertains to CEA, or within a specific development context. Given the need to both expand the IA effectiveness conversation into the CEA domain and to better understand VEC selection, we investigate how a ‘typical’ VEC selection process unfolds within CEA using comprehensive study IAs of major road projects in Canada as a basis for the study. Specifically, we ask: (i) ‘What VECs are chosen?’; (ii) ‘By what process and according to what rationale?’; and (iii) ‘Are VEC choices reflective of cumulative effects concerns particular to a linear project development context?’ This study will possibly help to improve CEA performance and spark effectiveness research in this area, and contribute to a more complete understanding of the practice of VEC selection. Section 2 describes the methods, study area and selected road construction cases. Results are presented in Section 3 followed by discussion in Section 4. Section 5 presents our key findings and recommendations for improving VEC selection in CEA.

2. Methods and cases

CEA was required by the Canadian Environmental Assessment Act 1995 (the Act) for over 15 years, and is also a requisite of the Canadian Environmental Assessment Act 2012; its recent replacement. The 1995 Act (s.16) stipulated that all projects deemed ‘likely to have significant adverse environmental effects’ are subject to CEA, including those designated as comprehensive study IAs. When ‘triggered’ by the legislation, comprehensive studies allowed for larger, complex development activities including road construction projects to be thoroughly evaluated by a panel of experts in consultation with the proponent, affected communities and other interested parties. The former Comprehensive Study List regulation2 (Schedule 29) specifically required that ‘an all-season public highway that will be more than 50 km in length and either will be located on a new right-of-way or will lead to a community that lacks all-season public highway access.’

Between 1995 and 2011, a total of 11 comprehensive studies involving road construction projects were completed. These projects (listed in Table 1) form the basis of investigation for this study: it is too early to assess similar projects under the more recent 2012 Act. Figure 1 shows the location of each project within Canada.

Table 1 Details of the selected comprehensive study road construction IAs.

No	Project name	Proponent (Provincial jurisdiction)	Year of CSR/EIS	Road length (km)
1	407 East Transportation Corridor	Ontario Ministry of Transportation	2011	70
2	Athabasca Seasonal Road	Saskatchewan Highways & Trans	1996	180
3	Completion of Highway 35 between Saint-Jean-sur-Richelieu and the US Border	Dept of Transport Quebec	2008	38^a
4	Greenville to Kincolith Road	BC Ministry of Transp. & Hwys	1999	24^b
5	Lake Winnipeg East Side Road Project	The Manitoba Floodway and East Side Road Authority	2011	156
6	Local Access Road – Highway 58, Fox Lake and Garden River	The Little Red River Cree Nation (AB)	2007	64.8
7	New Route 2 Trans-Canada Highway Perth-Andover to Woodstock	New Brunswick Dept. of Transp.	2004	70.7
8	St Theresa Point/Wasagamack Airstrip and Connecting Road	Manitoba Transp. & Govt. Services	2001	28^c
9	Trans Labrador Highway Project – Phase III	NL Dept. of Works, Services and Trans	2003	250
10	Waskaganish Permanent Road	Waskaganish Band Council (QC)	1999	102
11	Wollaston Lake Road	Saskatchewan Highways & Transp.	2007	100

a The project encroached into a migratory bird sanctuary.

b Road provided access to communities which lacked an all-season highway.

c Project included a runway greater than 1500 m in length.

Figure 1 Canadian road infrastructure projects subject to ‘comprehensive study’ environmental IA between 1995 and 2011.

Data collection proceeded in two stages. First, a document analysis of 11 comprehensive study reports (CSRs) (and/or EISs) prepared for the road construction projects was performed. This was guided by a review template which focused on characteristics of VECs chosen, rationales for their selection, determination of those selected for CEA, and whether or not the road construction context influenced VEC selection. The data were then extracted into a spreadsheet for subsequent coding and analysis. A major challenge with the CSR analysis was the inconsistency in the structure of the reports. However, to ensure that the emerging themes are grounded in the original data, the EISs were consulted for further details when information from the CSRs alone was inadequate.

Next, semi-structured interviews were conducted with 22 individuals directly involved in the 11 projects. Those interviewed include project proponents (n = 5), federal responsible authorities (n = 3), consultants (n = 7), provincial government representatives (n = 3) and Canadian Environmental Assessment Agency project managers (n = 4). Using a semi-structured interview schedule comprising open-ended questions, qualitative data were gathered from interviewees regarding: the types of VECs typically selected for project IA and CEA; VEC selection procedures; the roles of key actors and their values and rationales for VEC selection; and tools used to aid VEC decision-making. The interviews also gathered opinions on the process effectiveness and deficiencies with respect to linear projects, including sensitivity of VEC selection processes to cumulative effects issues. The data were coded by theme and, where applicable, themes were described quantitatively according to frequency of occurrence. Data collected at this phase was particularly used to investigate the ways CEA VEC selection is practiced as this is less evident in many of the project documents.

3. Results: VEC selection for CEA of linear projects

3.1 What becomes a VEC in CEA and why?

3.1.1 How are VECs chosen in CEA?

Based on evidence from both the interviews and the document review, many of the environmental components used as VECs in CEA are the product of the parent project assessment; in fact it is quite difficult to distinguish the two VEC selection processes from one another. In project VEC selection, project proponents in conjunction with responsible authorities initially identify broad environmental concerns based on the terms of reference for the assessment. These concerns are then filtered through experts' (scientific) opinions and, with further inputs from the general public and the affected community, VECs are selected with the geographically local or temporally immediate environmental impacts of the project in mind. Any VECs that are predicted to sustain significant adverse effects following mitigation are then carried forward as VECs in the CEA for further analysis3. This is known as residual effects analysis (BBOP 2009). It is reported that the practice of proponents/consultants relying on previously established project VEC lists to guide CEA VEC lists (via residual effects analysis) is very common. One respondent provided an account of a CEA in which the consultants simply ‘lifted’ the entire VEC list from a previous project and applied it directly within the CEA for their own project. The respondent adds that this approach typifies the level of subjectivity some consultants adopt in VEC decisions for CEA, and clearly does not take into consideration the context of the new development proposal.

Interviewees also report that environmental legislation and regulations are a key influence in designating CEA VEC. According to the terms of reference of many of the projects, responsible authorities and affected government agencies are expected to contribute to VEC selection processes, as the law establishes protections standards for many environmental components. For example, one interviewee from the Canadian Environmental Assessment Agency stated:

…one of the things that we consider sometimes is fish and fish habitat because there is a legal requirement to compensate effects on fish and fish habitat. When you know the project effects on fish and fish habitat, you could consider for example that there will not be any residual effects of the project because it would have been compensated… because there is a regulation or an Act that requires compensation of those effects.

A provincial agency representative explains: the ‘review team members are the most important people selecting the VECs and they all have regulatory requirements and mandates for the different government agencies.’

On the whole, interviewees tended to agree that regulators and responsible authorities are the most influential in defining VEC selection criteria and determining final CEA VEC lists. However, more than half of all interviewees (n = 12)4 expressed concern that ‘the ecosystem approach is being circumvented by just focusing on certain parts of the environment or certain species or issues’ and that VEC selection is based on a great deal of ‘speculation or judgment calls’, with little attention given to holistic ecosystem considerations. One respondent from Aboriginal Affairs and Northern Development Canada remarked:

that is one of the big things we are having problem with right now… we are looking at a more holistic, wider scope, wider line (which) is quite difficult for a lot of proponents and a lot of government …and just because we draw an imaginary line on a map doesn't mean that the environment stops at the line.

Even though public commentary in VEC selection remains very influential (n = 15), most case informants view VEC selection as a ‘value-ridden’ and ‘highly subjective’ process (n = 12) based on negotiation (as opposed to scientific evidence) conducted with little regard to the specific context of the project.

3.1.2 What kinds of VECs are chosen?

The rationale that most often underlies selection of VECs is perceived ecological importance (n = 11) (see Figure 2). Perceived social (n = 7) and economic (n = 7) importance of an environmental component also figures prominently in VEC selection processes, as does cultural importance (n = 8), which is often related to concerns about the effects of proposed projects on indigenous cultural survival. Less frequently cited rationales for CEA VEC selection include traditional use values (n = 4), educational value (n = 2), scientific value (n = 1) and human health value (n = 1).

Figure 2 VEC selection rationales identified in road construction projects. Note: Frequency in the table depicts number of cases in which a value is cited.

Figure 3 illustrates that CEA VECs are indeed heavily influenced by project VEC lists; it compares project VECs and CEA VECs identified in the CSRs of the 11 road construction projects examined. Of the 51 project-specific VECs selected in the 11 projects studied, the most common were: fish and fish habitats (n = 10), vegetation and vegetation communities (n = 8), wildlife and wildlife habitat (n = 8), species at risk (n = 7) and water quality (n = 6). VECs such as current use of land and resources by Aboriginal people, geomorphology and geology, and wetlands were also common: each was identified in five of the projects. In comparison, 36 of the 51 project-specific VECs also served as VECs for CEA on the 11 project assessments studied, the most common of which were fish and fish habitats (n = 6), species at risk (n = 6), vegetation and vegetation communities (n = 6), wetlands (n = 5), and wildlife and wildlife habitat (n = 5). Water quality and birds/migratory birds were also selected as CEA VECs in four of the road construction projects studied.

Figure 3 Project-specific VECs and CEA VECs identified in road construction projects.

Interestingly, Figure 3 reveals some inconsistencies in terms of how VECs are defined, i.e. what is labeled as a VEC varies from project to project. For example, moose, grizzly bear, ungulates, caribou and furbearer are named as VECs for certain assessments, but these can also be broadly classified as simply wildlife, which itself appears as a VEC in eight project-specific assessments (and six CEAs). Furthermore, what is identified as a VEC indicator in some projects is also designated as a CEA VEC in other projects. For example, grizzly bear and Vancouver Canadageese which were designated CEA VECs in the Greenville to Kincolith project, were, in certain project assessments, identified as proxies or indicators of wildlife and bird VECs, respectively. Figure 3 also reveals ambiguous characterization of certain non-biophysical CEA VECs. The exact meaning and nature of CEA VECs such as: community life; resource use and users; socio-economic environment; and cultural environment are all rather vague and not explicitly defined in the CSRs in which they were used.

This may be related to the finding that in most cases, the term ‘VEC’ is not explicitly defined (it was in just five of the 11 cases). When the term is defined, however, those definitions appear to be quite consistent with one another (see Table 2). Definitions provided seem to emphasize (1) features, aspects or attributes of the environment (social, cultural, economic or environmental) and (2) the importance or value of such features to humans or society.

Table 2 Definition of the term VEC provided in project CSRs.

Project	VEC Definition
407 ETR	Specific features or attributes of the environment that are considered to be important for regulatory reasons, or because of their social, cultural, economic or ecological value
Athabasca^a	None given^b
Completion of Hwy 35^a	Any part or aspect of the environment that is considered important by the proponent, the public, scientists, government or any other administrative entity involved in the assessment process
Greenville to Kincolith	None given^b
Lake Winnipeg ESR	None given^b
Highway 58	Components of the environment (biophysical, socio-cultural, and economic) that are valued by society
New Route 2 Trans-Canada^a	Components of the environment that may be affected by the project, and are of scientific, ecological or cultural importance
St Theresa/Wasagamack	None given^b
Trans Labrador (Phase III)	None given^b
Waskaganish	Component of the natural ecosystem which is perceived to be of social, cultural, economic or ecological significance as identified through consultation with affected people and through scientific opinion.
Wollaston Lake Road^a	None given^b

a The term ‘valued social component’ (VSC) was also used in these projects but none provides a definition different from the VEC concept adopted.

b No definition of ‘VEC’ is provided in the CSR.

3.2 Does CEA VEC selection practice reflect the linear project development context?

3.2.1 Do the assessment boundaries influence CEA VEC selection?

A great majority of interviewees 73% (16 of 22) indicate that consideration of spatial and temporal scale of the road development did not influence CEA VEC selection. One of the project proponents explains: ‘We did not adjust the spatial scale… we set the spatial scale then we look at the VECs, so geographically, no adjustment; temporally only some adjustment I would say when you are looking at cumulative effects.’ For static VECs (e.g. vegetation and vegetation communities), spatial boundaries for CEA analysis were typically synonymous with the project boundaries. In contrast, for highly mobile VECs (e.g. deer population) ‘an ecologically sensible boundary’ (e.g. a watershed) was often selected in order to examine the relative importance of each CEA VEC regionally. Interviewees noted that the receiving environment can register the potential effects of roads at multiple spatial scales (e.g. effects can be experienced across multiple watersheds, or wildlife population ranges).

Many interviewees (seven of 22) believe that setting a temporal scale in CEA is a rather more difficult task than setting geographic boundaries. Roads' growth-inducing effects in particular complicate temporal boundary setting for CEA VEC analysis. Understanding what the addition of roads means to a region (e.g. access to hunting, reduction in cost of resource extraction, increased interest in forestry or mining etc.) and what that might mean to different communities in terms of cumulative effects remain a challenge. Many respondents suggested that the longer service life of a road or any other linear project, in particular, can complicate the determination of appropriate VECs for assessing cumulative effects.

3.2.2 Does the staging of project decisions influence CEA VEC selection?

Respondents report that the timing of key project decisions creates a clear challenge for CEA VEC selection, which typically takes place long after project feasibility studies, and other key project influences:

We looked at routes and the reason for the project at the project definition stage i.e. is there a reason to pursue this development? Those are the questions you ask at that stage before you get into the environmental IA. By the time you get to environmental assessment stage, you are too late for that kind of question (on CEA VEC selection). Those questions happen early on or should happen early on.

The Canadian EA Act requires the consideration of ‘alternative means of carrying out a project that are technically and economically feasible and the environmental effects of any such alternative means’ (section 16d). Based on the interview data and document review, alternative means are often discussed in the context of route description, but none of the project EISs/CSRs (or CEAs) explicitly discusses the ecological and/or socioeconomic variables that would have influenced the choice of a preferred routes and by implication, how that decision influences the nature of VECs selected. In all cases, project feasibility studies and the assessment processes are differentiated and do not appear to connect in any significant way. One interviewee points out the challenge: ‘roads cause problems and whether they are social or environmental, those problems are well known. Everybody turns a blind eye to those issues (yet) we need to put them into the context of cumulative effects from the beginning.’ Although none of the study participants explicitly link initial routing considerations with the IA process, some interviewees suggest the need to anticipate the cumulative environmental effects of projects in the project design and description phase, prior to the formal assessment process:

…it will allow us [EA practitioners] to work the with the planners, the designers, the engineers to modify the project design to basically mitigate for adverse (cumulative) effects at the project design stage before it even happens, so that we don't have to go and do that after the effects. So the biologist provides this feedback to the planners and the designers and the engineers to allow them to tweak the design of the project to accommodate the interests of those species.

Addressing cumulative effects early in the design and route selection process may have great impact on the CEA VECs eventually selected (e.g. Baxter et al. 2001).

Similarly, many project documents do not explicitly clarify whether cumulative effects formed part of the considerations in the scoping phase of the IA. When interviewees were probed about the omission, most (12 of 22) said that cumulative effects were in fact part of the scoping but did not influence VEC decisions. This is because project VEC choices are typically centered on the local project environment, and not necessarily the project's regional context and this is the scale at which many cumulative effects must be addressed.

The time between when decisions on project VEC selection are taken and when the CEA is carried out (often following residual effects analysis) places an additional burden on the assessors to revisit initial environmental concerns when deciding CEA VECs. The following quote, which typifies comments on the subject, explains the reason why cumulative effects are not a consideration early on in IA processes:

Generally, it (the CEA VEC list) was narrowed down after the general IA. When we did our scoping document, we went over a list of VECs for direct impact of the project. That was probably available and we saw public comment on it. We didn't base the (cumulative) analysis of the project on the VECs; we only considered the ones that have residual adverse effects at the end of it…

Many interviewees believe that early CEA consideration (i) was simply omitted even though necessary; (ii) would not ‘bring anything new’; and (iii) is not feasible, as CEA is ‘a federal requirement’ which would not normally come up during project scoping, particularly for projects that started under provincial regulations such as the 407 East Transportation Corridor and the Lake Winnipeg East Side Road. However, in the few projects where cumulative effects were in fact considered early on during the project scoping phase, interviewees indicated that doing so helped to protect against inadequate baseline data collection during project IA, which was later helpful in the CEA.

3.2.3 Is there guidance for CEA VEC selection in a linear project development context?

While CEA practice has benefited from guidance documents such as those issued by Canadian Environmental Assessment Agency (Hegmann et al. 1999) and the US Council on Environmental (CEQ 1997); guidance for CEA VEC selection for linear development projects is entirely absent from these. Not one interviewee mentioned of any guidelines that were consulted for assistance, nor did any of the project documents reviewed. A provincial government official remarked that because ‘there hasn't been a kind of consistent approach,’ similar projects completed within the same region usually adopt different sets of VECs. This complicates the capability ‘to determine what level of detail or what level of direction’ is required to ensure stakeholders are consistent in their choice of VECs. Another interviewee (a Canadian Environmental Assessment Agency official) commented: ‘I know that's the issue that we have when we ask colleagues here “do you have any good example to provide this proponent?” and basically none of us has good example for CEA.’ Guidance for watershed CEA that makes selection of certain VECs compulsory across projects has been suggested (Ball et al. 2013). Unfortunately such guidance has not been developed yet, and neither has similar guidance for VEC selection for linear projects.

3.2.4 Does data availability affect CEA VEC selection?

Data availability plays a key role in identifying regional VECs especially when long linear projects are involved. Respondents (13 of 22) highlighted the absence of a regional database for key environmental components as a drawback to CEA VEC selection: ‘There are things that are definitely measurable but there is no regional information about them.’ Lack of accessibility to data is also often a stated problem, not only in terms of willingness to share available data, but also in terms of their use-ability. A proponent provided insight into accessibility challenges:

There may be different proponents in the past that have information that a new proponent needs to do that cumulative IA and other proponents may be unwilling to share that data. Let me say some are willing to share data but so many proponents make it in so many different formats that may not be that useful when you try to combine all that data to make sense of it.

Most respondents asserted that funding for (regional) data collection is a major challenge: ‘Going out to collect such data gets to be very expensive and proponents typically don't want to (collect) data that they don't see as practically useful to their operations’ was the remark of a provincial government official. Another provincial government official asked: ‘…who funds both the baseline data collection and the use of those data in the cumulative effects arena?’ Even if it is possible to aggregate all necessary data, interviewees caution that CEA and CEA VEC selection must account for the variation in spatial characteristics over the entire road length during analysis of cumulative effects; this is both time-consuming and error-prone.

Linear projects are fraught with unique assessment challenges, partly due to the different scales that assessors need to work at. As earlier noted when discussing assessment boundaries, many respondents express concern over the appropriate scale for CEA VEC selection. There can also be significant differences among linear projects in terms of the length of proposed development, local versus regional environmental conditions and community concerns. One interviewee, a consultant, suggests that a complete overhaul of the current approach to data management with regard to linear project assessment is necessary, and any restructuring should involve a regional approach, especially with respect to addressing cumulative effects:

Frankly, there is no data available on a regional scale. The availability of regional database on key VECs was critical. That is the big challenge… There are things that are definitely measurable but there is no regional information about them. We couldn't put it in a more regional context to see the potential cumulative effects or whether they are significant or not.

In general, the interview data and document analysis suggest that there is a failure in project IA to link CEA VEC selection to the unique attributes of linear road construction projects, or to the unique socio-political development context that influences each project.

4. Discussion

4.1 Overlap of CEA VECs and VEC indicators should be addressed

The blurred distinction between VECs and VEC indicators is evident from the results (see Table 2). This observation is consistent with findings from a South Saskatchewan watershed CEA study wherein certain environmental components were found to serve variably as both a VEC and a VEC indicator (e.g. ground water quality as an indicator of surface water quality), depending upon the project (Ball et al. 2013). This inconsistency raises the question of what exactly should be considered as VEC, and what is an appropriate VEC indicator. The importance of having a consistent set of VECs across related projects and their connection to relevant and measurable indicators have been stressed in recent studies (e.g. Ball et al. 2013, Sheelanere et al. 2013). Inconsistencies at the project level will add confusion to any regional assessment exercise, which typically collates or adds to existing empirical knowledge gathered at the project level.

Proper distinction between VECs and indicators is especially important for linear projects, because, unlike non-linear projects such as mining and oil and gas development, the effects of linear projects can vary considerably geographically, particularly for projects that cut across different ecosystems. VECs and VEC indicators must reflect this complex spatial and temporal context of linear development projects. For road construction projects, which are regional projects by nature due to their generally vast spatial scale, CEA VECs should be both regionally and locally relevant. VEC indicators may reflect different spatial scales as well, as appropriate for analysis and monitoring of VEC health.

4.2 More emphasis needed on social and economic CEA VECs

The dominant emphasis on biophysical VECs in both project IA and CEA is evident in this study and in other literature (e.g. McAfee & Malouin 2008; Gallaugher & Wood 2009). This observation holds across all projects reviewed. For example, in the St Theresa Point/Wasagamack IA, all project VECs were biophysical. Likely, this is explained by the way ‘environment’ is defined in the CEA Act 1995 – which was in force when the 11 projects in the study were conducted – and that is primarily biophysical.5 Social and economic impacts were to be considered only insofar as they were affected by a change in the biophysical environment.

Also evident from the study is that VEC choices are rarely tied to a single rationale, but that the most dominant rationale is perceived ecological significance. Even rationales for VEC inclusion such as scientific value are considered to be proxies of ecological significance in some ecological literature (e.g. Hay et al. 1996). A sustainability rationale – where ecological, social and economic VECs are treated more equally in VEC selection processes (provided they are affected by the proposed development) might be a more effective approach, particularly for linear project development such as roads given their obvious impacts to the social and economic conditions in a region.

It is apparent that road construction causes as much adverse impact to the social environment as the biophysical one (United Nations 2001) and that, in addition to ecological issues, impacts to social and economic growth (e.g. jobs, improved accessibility, amenities) or ills (e.g. alcohol consumption, accidents, health issues) should be factored into CEA VEC decision-making. This study reveals that there is a fair degree of subjectivity in VEC selection processes, meaning that while some VECs are selected in response to legislative and/or regulatory requirements, many VECs are chosen based upon expert opinion and/or public pressure. Thus, there is opportunity in the discretionary portion of VEC selection processes to consider additional social and economic VECs: this might be more appropriate to evaluating the cumulative impacts of linear development projects, including road construction projects.

4.3 Timing of cumulative effects considerations in IA is too late

The interviews examined whether VEC selection practices were enhanced through early consideration of cumulative effects issues, as advocated in literature (e.g. Baxter et al. 2001). Several studies have argued that CEA is ‘what truly matters’ (e.g. Senner 2011) which presupposes that any assessment should be dominated by identification of cumulative effects issues and a focus on cumulative effects analysis. However, about half of the respondents (10 of 22) revealed that, for the projects in which they were involved, discussion of CEA came up only after the initial assessment of impacts to project VECs was performed. Most view early consideration of CEA as necessary, but this does not often happen in practice, in no small part due to the absence of guidance.

This is a startling finding, given the clear direction that early consideration of cumulative effects constitutes good practice (Treweek 1995; Baxter et al. 2001). More disturbing is the fact that project IA for road construction projects seems to be highly reactive, i.e. the preferred routing alternative is identified before the assessment is conducted. This constrains opportunities to consider VEC sustainability from a regional or cumulative perspective – and whether regional cumulative effects issues should influence project design and/or routing alternatives. It is perhaps a bit of an ‘egg before chicken, or chicken before egg’ conundrum. However, given the powerful influence of roads to induce long-term social, economic and biophysical change in a region, it is important that the potential cumulative effects of a proposed road, and all of its alternatives (including the ‘no go’ option) are identified in the pre-assessment phase during feasibility studies.

This approach could assist the early identification of potential social and economic CEA VECs and their susceptibility to road impacts, which can then be addressed in baseline data collection, and considered in greater detail during trend analysis in CEA processes. It could be especially beneficial in cases where a proposed road construction project is expected to increase traffic flow significantly, particularly in regions without previous road access. For example, a single new highway can lead to hunting pressure and habitat fragmentation; exponential demand for social, recreational and municipal services; and numerous other effects (Wasike 2001; Tricker 2007).

4.4 IA boundaries must be flexible to facilitate CEA

The boundaries used in CEA are generally synonymous with those initially adopted for the IA, and those are typically established with intent to facilitate analysis of direct effects in the immediate project area. This was made clear in the interviews, as 73% (16 of 22) of interviewees stated that boundary adjustment was not necessary when considering impacts to CEA VECs. Inappropriate boundary setting, especially for CEA, can either mask or magnify the level of exposure to a VEC (Hegmann et al. 1999; João 2002; Noble 2008). For instance, while some VECs such as wildlife population may be present or abundant at the north end of a proposed highway, they may not be present at all at the south end of the same road, some 50 or more kilometers away. Assessment boundaries should be defined by VEC sustainability, and not simply by the anticipated direct effects of a project or its physical footprint (Hegmann et al. 1999; Duinker & Greig 2006; Canter et al. 2013).

Different scales of analysis can make VECs seem more or less critical in an assessment context (Thérivel & Ross 2007; Noble 2008). In one of the projects – the Greenville to Kincolith project – an interviewee noted that the selection of the grizzly bear as a CEA VEC was based on considering a scale different from the traditional ‘local study area versus regional study area’ perspective that usually drives VEC decisions. Rather than focusing on these spatial classifications, ‘individual bear spatial scale below the local study area’ was used to identify different bear populations, which warranted their consideration as a CEA VEC. Where this level of detail in terms of VEC selection and boundary setting is not observed, the scale of analysis could either be too narrow (confined to the defined local project area) or too broad (scaled up to the defined regional project area) to facilitate thorough CEA VEC selection processes and/or sound decision-making about impacts to CEA VECs. In the case of evaluating the cumulative impacts of linear development projects, important information about VEC sustainability may be missed if the boundaries of the assessment are too rigid (Mulvihill & Baker 2001) or ubiquitously applied, rather than that being sensitive to individual VEC context.

4.5 The need for CEA VEC selection guidance and regional databases

At least two authors have noted that inadequate data contributes to process challenges in CEA VEC selection (Bérubé 2007; Connelly 2008). IA is, generally speaking, beleaguered by data insufficiencies (Duinker & Greig 2006); collection of data which bear little or no relevance to CEA (Fuggle 2005); and shortage of funds to collect extensive data (Duinker & Greig 2006). The participants in this study noted similar challenges. Some interviewees suggested the need for regional-scale databases to enhance CEA VEC selection and cumulative impact analysis. We believe that regional databases, organized according to natural regional boundaries (e.g. watersheds, eco-regions, river basins) or designated regional resource development regions (where these exist), could in time remedy some of the challenges associated with cost intensive project-specific data collection and enhance usability due to better consistency in reporting formats.

Guidance for CEA VEC selection appears to be non-existent at this time. Not only are general guidelines needed, sector-specific guidance is also desirable, particularly for classes of projects that share similar physical features and potential impact profiles, as many linear projects do. At minimum, guidance should address ways in which regional cumulative effects concerns can be parlayed into better, more inclusive CEA VEC lists.

5. Conclusion

The goal of this study was to develop insight into the process of CEA VEC selection, as viewed through the lens of the road construction sector. It was found that CEA VEC selection processes are not sensitive to the unique nature of road construction projects, or the linear development context generally. Cumulative effects considerations come late in the IA process after a preferred route is identified and project VEC list has already been determined; thus CEA VEC lists are generally not sensitive to broader, regional cumulative effects issues and trends that may already be in play. Instead they typically mirror the VEC lists used for direct effects analysis, and mainly consist of any biophysical VECs that are expected to experience adverse residual impacts, assuming successful implementation of mitigation and environmental management programs. There is clearly room for more insightful, creative approaches to CEA VEC selection – ones that better reflect potential impacts to social and economic aspects of the environment – since these are values-driven, subjective processes and not only directed by legislation and regulations.

Biophysical VECs receive far more attention than non-biophysical VECs in both project IA and CEA, although it is well-known that the induced socio-economic impacts of linear developments have on surrounding landscapes and communities can be considerable (e.g. Noble et al. 2011): particularly in less intensively developed regions (Jaeger et al. 2007). There is a clear need for regional databases to support consistent CEA VEC selection and data collection processes, yet none were known to study participants. There appears to be no guidance or support for CEA-specific VEC selection, despite that CEA guidance documents do exist (e.g. Hegmann et al. 1999).

The results of this study are broadly transferable. Even though the focus of our analysis was on 11 road construction projects in Canada, road construction is common all over the world. Indeed linear projects are common elements of nearly every industrial project, as they must be connected with the energy sources, natural resources sources, and communities that sustain them. Beyond Canada, the core challenge is to transform existing practice from a scenario in which project feasibility studies and the assessment processes are differentiated, to one where the former can reinforce the later in determining what VECs are appropriate in evaluating cumulative effects. For this reason, closer attention to CEA VEC selection processes is appropriate, and the development of sector-specific and/or region-specific guidance is warranted. We believe improved CEA VEC selection will enhance CEA efficacy, and IA efficacy more generally.

Acknowledgements

We thank the Social Sciences and Humanities Research Council of Canada for funding this research.

Notes

^1. Email: [email protected]

^2. Most projects (including road infrastructure) triggered by this legislation are now classified in the new 2012 Act as physical activities under ‘environmental assessments by responsible authorities’.

^3. A more detailed account of VEC selection process elements for project-specific IA, and relatedly CEA, is provided in Olagunju and Gunn (2013).

^4.n = number of cases.

^5. Section 2(1) of the 1995 Act defines environment as ‘the components of the Earth, and includes (a) land, water and air, including all layers of the atmosphere, (b) all organic and inorganic matter and living organisms, and (c) the interacting natural systems that include components referred to in paragraphs (a) and (b)’