Enhancing source water protection and watershed management: Lessons from the case of the New Brunswick Water Classification Initiative

Abstract

Source water protection varies by locale, and approaches and experiences are accumulating in response to concerns about drinking water safety. Learning lessons and transferring them from experiences elsewhere is a well-established practice for addressing water governance challenges. In response to the need to enhance source water protection policies and initiatives and a growing interest in modes of governance in which government and non-government actors collaborate, this research investigated and derived lessons from the Water Classification Initiative in New Brunswick, Canada. The research specifically aimed to describe the development of the initiative, analyze structural relationships among actors involved in the initiative and describe the successes and challenges experienced. Investigation of the Water Classification Initiative illustrates how key aspects of source water protection identified in the literature (e.g. watershed as a focal scale, collaborative approaches, incorporation of science and local knowledge) can be incorporated into policy, how capacity may be built or constrained in the context of government-led collaborative approaches, and how social network analysis offers a powerful tool to understand interactions among those involved in a policy process. Learning from these insights offers an opportunity to advance the development of new approaches as well as to enhance existing source water protection policies.

Abstract

La protection de l’eau de source varie selon la localité, et les approches et les expériences en réponse aux soucis quant à la sécurité de l’eau potable accumulent. Apprendre les leçons et les transférer des expériences ailleurs est une pratique bien établie pour aborder des défis de la gouvernance de l’eau. En réponse au besoin d’améliorer les politiques et les initiatives de la protection de l’eau de source et à un intérêt croissant pour les modes de gouvernance dans lesquels le gouvernement et les acteurs non gouvernementaux collaborent, cette recherche a examiné et a tiré des leçons de l’Initiative de classification des eaux au Nouveau-Brunswick, Canada. La recherche a visé spécifiquement à décrire le développement de l’Initiative, à analyser les relations structurelles entre les acteurs qui participent à l’Initiative et à décrire les succès obtenus et les défis rencontrés. L’enquête sur l’Initiative de classification des eaux illustre comment : les aspects clés de la protection de l’eau de source identifiés dans la littérature (ex. le bassin versant comme une échelle focale, des approches collaboratives, l’incorporation de la science et les connaissances locales) peuvent être incorporés dans la politique; la capacité puisse être développée ou contrainte au contexte des approches collaboratives dirigées par le gouvernement, et l’analyse du réseautage social offre un instrument puissant pour comprendre les interactions entre ceux qui participent dans un processus politique. Apprendre de ces aperçus offre une occasion d’avancer le développement de nouvelles approches aussi bien que d’améliorer les politiques de la protection de l’eau de source existantes.

Introduction

Protecting source waters is imperative for humans and ecosystems. Source waters encompass all surface and ground resources from which present or future drinking water originates (O’Connor 2002; Ivey et al. 2006a; Simms et al. 2010). The emphasis on safe drinking water is expected as it is a Millennium Development Goal (MDG). Although the recent achievement of providing safe drinking water to over 88% of the world’s population is impressive, millions of people still lack access and keeping drinking water safe remains an important and substantive challenge (UNICEF/World Health Organization 2012). Safeguarding drinking water requires the establishment of multiple barriers (source water protection, treatment, distribution, monitoring, contingency response) and the protection of source waters offers a critical initial impediment that reduces risks and costs effectively (O’Connor 2002; Hrudey and Hrudey 2004; Hrudey et al. 2006; Ivey et al. 2006a; Plummer et al. 2010). According to the World Health Organization (1996), “source protection is almost invariably the best method of ensuring safe drinking-water and is to be preferred to treating a contaminated water supply to render it suitable for consumption.”

Despite consensus on its importance and much success in the provisioning of safe drinking water, source water protection remains a formidable challenge for several reasons: nonpoint sources of water pollution in particular persist and have defied solution (Mandelker 1989; Caruso 2000; United States Environmental Protection Agency (USEPA) 2012); integration is required between planning the uses of land and water (National Research Council 2000; FitzGibbon and Plummer 2004; Plummer et al. 2011); competing interests are held by diverse stakeholders (National Research Council 2000; Simms et al. 2010), and capacities of institutional arrangements and governance approaches vary considerably (Ivey et al. 2006a, 2006b; Timmer et al. 2007; Patrick 2009; Plummer et al. 2010; Simms et al. 2010). Source water protection is broadly understood and generally defined as “watershed and aquifer management for the protection of drinking water supplies and is operationalized through land-use management programs with the specific goal of protecting drinking water sources against contamination” (Patrick 2009, 209).

As a form of environmental decision-making influenced by contextual social and biophysical processes, the approaches used for source water protection vary by locale (Ivey et al. 2006b; Patrick 2008; Simpson and de Loë 2011). Experiences with source water protection policies and initiatives are accumulating quickly in North America due to persistent and pressing concerns about drinking water safety. For example, the Clean Water Act (33 U.S.C. § 1251 et seq) in the United States provides regulatory structure for water pollution broadly and is central to the efforts by the USEPA concerning source water protection specifically (USEPA 2012). The USEPA (1991, 1993) early on developed a five-step process for wellhead protection that is now a well-documented template in the source water protection literature (Young et al. 2009). The New York City Watershed Memorandum of Agreement is a novel example of effective source water protection utilizing a suite of regulatory and non-regulatory mechanisms with a watershed focus (National Research Council 2000; Pires 2004).

In Canada, attention to drinking water safety has been thrust into the public consciousness nationwide by chronic drinking water issues in First Nations communities as well as judicial inquires associated with contaminated water supplies in North Battleford, Saskatchewan (Laing 2002) and Walkerton, Ontario (O’Connor 2002). Research by FitzGibbon and Plummer (2004) and Simms et al. (2010) captures the diversity of source water protection initiatives, tools and approaches across Canada. Prompted by the tragic events of Walkerton, where seven people perished, the province of Ontario has embarked upon “a radically new approach to governance for drinking source water protection” (Plummer et al. 2011, 3). The Safe Drinking Water Act (2002) and the Clean Water Act (2006) were established to reflect the multi-barrier approach, with the latter setting forth a province-wide watershed-scale planning process. Ontario’s approach to source water protection is thus mandated by the government and operationalized through regional multi-stakeholder “source protection committees” tasked with assessing the threats to source water and creating a plan to protect drinking water sources. New Brunswick’s Water Classification Initiative (WCI) is another example of government-mandated source water protection, with the Water Classification Regulation (2002) establishing classifications for surface waters at a watershed scale (New Brunswick Department of Environment and Local Government (NB DELG) 2002a) and encouraging an ongoing watershed-based management approach.

Several of the above examples of source water protection serve also to illustrate the increasing interest in and use of collaborative approaches in watershed-based natural resources management (Selin and Chavez 1995; Conley and Moote 2003; Innes and Booher 2010). There are important benefits from collaborative planning approaches, including but not limited to increased legitimacy of decision-making, increased knowledge and resource sharing, reduced conflict and building of social capital (Conley and Moote 2003; Imperial 2005; Trachtenberg and Focht 2005); however, it is important to note that these approaches also have potential weaknesses (e.g. political derailment, lack of skills to undertake the process) (Wollondeck and Yaffee 2000; Walker and Hurley 2004) and do not constitute a panacea for source water protection.

Learning lessons from experiences elsewhere is a well-established practice for addressing water governance challenges, such as source water protection, and can be expedited through policy transfer (Swainson and de Loë 2010, 2011). Largely located in political science and international development scholarship, several labels (e.g. policy transfer, emulation, lesson drawing) are used to broadly refer to “a process in which knowledge about policies, administrative arrangements, institutions etc. in one time and/or place is used in the development of policies, administrative arrangements and institutions in another time and/or place” (Dolowitz and Marsh 1996, 344). Investigations of source water protection have endeavoured to identify facilitating and constraining factors (de Loë et al. 2005; Ivey et al. 2006a; Patrick 2008; Patrick et al. 2008), showcase the range of approaches and initiatives to inform others (FitzGibbon and Plummer 2004; Simms et al. 2010), and identify lessons from specific experiences (e.g. National Research Council 2000; Timmer et al. 2007; Young et al. 2009; Benson et al. 2012). From their specific work with policy transfer and water governance, Swainson and de Loë (2010) conclude that

[T]he opportunity to learn from water policy experiments and experiences in other jurisdictions, at multiple scales and levels of governance, has never been greater. However realizing the benefits of policy transfer will require that practitioners and policy makers understand the factors that shape transferability, and develop the capacity to utilize lessons appropriately in their own contexts.

In response to the need to learn from water policy experiments and enhance source water protection policies and initiatives in other jurisdictions, this research investigates the WCI in New Brunswick, Canada. The research specifically aims to (1) describe the development of the policy initiative, (2) illustrate and analyze the structural relationships among the government and non-government actors, and (3) identify successes and challenges experienced by the actors involved. The WCI was selected because it offers a longstanding policy initiative for source water protection on a watershed basis and at a provincial scale and also illustrates the growing interest in new modes of water governance in which government and non-government actors collaborate to solve water management challenges.

Methods

The case study method was selected to investigate the New Brunswick WCI. Yin (2009, 2) explains that the case study is an appropriate research design “when conducting exploratory research focusing on a contemporary phenomenon within a real life context especially when the boundaries between phenomenon and context are not clearly evident and over which the investigator has little or no control.” While a variety of case study types exist, the intent in this research was to conduct a collective case study. Stake (1995, 3) explains that, in collective case studies, “each case study is instrumental … but there will be important coordination between the individual studies.” The New Brunswick WCI was therefore selected because it offers a longstanding policy initiative for source water protection (the “case”) on a watershed basis (a collection of watershed organizations involved in the initiative) and at a provincial scale. It also illustrates the growing interest in new modes of water governance, in which government and non-government actors collaborate to solve a problem or take advantage of an opportunity (Rogers and Hall 2003) and provides an opportunity to investigate a process that has been completed as a source for lesson-drawing for other source water protection processes currently underway and those being planned or conceptualized.

While the New Brunswick WCI was the main unit of analysis, a nested case study structure was used to undertake the research. This approach was devised to be congruent with the organization of the WCI on a watershed basis, and the objectives of the research. A list of the watersheds in New Brunswick where the WCI was being, or had been, initiated was provided by the NB DELG. Twenty-four watersheds were identified on this list and, of those, 19 had a watershed organization associated with them. All 19 of these watershed organizations were contacted and invited to participate, and 15 agreed, for a response rate of 79%. While the entire population (n = 25) did not participate, the participation rate was high among those watersheds where the WCI process had been completed. These 15 watersheds formed the sub-units of analysis within the New Brunswick WCI.

Multiple techniques were employed to collect and analyze data relating to the case: documents and interviews provided the data for the study, and content analysis and social network analysis served as the analytical techniques. The methods described below are presented by data source.

Documents provided a first source of data. Searches conducted for WCI documents regarding each specific watershed resulted in 16 reports; 13 retrieved from internet searches, and three retrieved from the NB DELG. Among these, the 15 provisional water classification documents were particularly informative because they were created by each watershed organization during the WCI process. These documents describe the watershed, the WCI process including all monitoring and data collection efforts, and the water classifications selected for each water body within the watershed, and, in some instances, also offer a description of an action plan to maintain or improve water quality. The provisional water classification documents were reviewed using a coding key to capture the timeframe in which the water classification was completed, the activities undertaken in order to classify water bodies and the post-classification action plan activities proposed.

Semi-structured interviews provided the second and main source of data. Participant identification began with background searches and interviews with NB DELG employees with detailed knowledge of the WCI to identify key contacts in participating watersheds. The key contacts identified were contacted by researchers. In cases where those contacted felt someone else was better suited to respond, referrals were made and researchers contacted the alternative contacts. One representative of each watershed organization involved in the study and two employees from the NB DELG (n = 17) participated in semi-structured interviews during the month of June 2011. The first part of the semi-structured interviews included questions focused on the history and process of the WCI (for the NB DELG respondents only), the watershed organization (where applicable), and the network of people and organizations they interacted within specifically for exchanging knowledge and/or collaboration for the purposes of the WCI, with separate free recall questions asked for each interaction type. Knowledge exchange was defined as sharing of information or consulting for an opinion, while the definition of collaboration required “working together on a project/task with a shared goal and reciprocal relationship.” The NB DELG employees and watershed organization representatives were asked to identify as many individuals, organizations, agencies, departments and businesses they could think of that fit within one or both reasons for interaction. Data from the first part of the semi-structured interview were used for social network analysis. Social network analysis is a useful tool for understanding and illuminating the structure and actors involved in interactions, and provides insights into who was involved in the WCI and the nature of their involvement. An ego network approach was employed for this study; that is, respondents were asked to provide information about all of the actors with whom their watershed organization interacted for the purpose of the WCI. However, due to the high response rate, a network analysis of those organizations involved in the WCI from 2002–2011 was possible with the acknowledgement that a few watershed organizations (four) did not participate.

Responses regarding interactions with others for knowledge exchange and/or collaboration were entered into MS Excel using a matrix structure and a binary code that indicated an interaction (“1”) or no interaction (“0”) between actors. The resulting matrices were imported into social network analysis software, Ucinet 6 (Analytic Technologies, Inc.), where the structure of each network (knowledge exchange and collaboration) was analyzed. The following analyses were undertaken: the size and diversity of the network of each respondent organization for the purpose of understanding how many and with what types of other actors the watershed organizations and the NB DELG interacted; the structure of the network of respondent organizations to understand the interactions (including direction) among organizations actively involved in the water classification process, and the roles of individual organizations within the networks (particularly those in positions of influence) to understand which actors may be considered as central, or influential, in the WCI process. Finally, an analysis of variance (ANOVA) was conducted to identify statistically significant differences in network size among two distinct types of watershed organizations: those that were created expressly for the purpose of the WCI and those that existed prior to the initiative. For a visual assessment of the network data, Netdraw 2 (Analytic Technologies, Inc.) was used. A follow-up email was sent to all respondents with their own network data only (i.e., their “ego” network) to ensure the network data collected was accurate.

Questions in the second portion of the interview were asked only of watershed organizations and focused on the resources used and activities undertaken by each watershed organization as a result of the WCI as well as the benefits and challenges of the WCI. Data from these questions was recorded by the interviewer and analyzed using content analysis (Krippendorff 2004). Responses recorded by the interviewer (and translated to English when provided in French) were coded in two rounds: the first was a round of inductive, open coding to identify themes present in responses; the second was a round of axial coding that fit responses within emergent themes from round one and refined themes. A single coder completed this work and thus intercoder reliability was not a concern. Tabulations of the number of occurrences of each theme were conducted and absolute frequencies are presented (Krippendorff 2004). Illustrative quotations from responses for each theme were identified and transcribed (and translated to English when in French) from audio recordings of these interviews to ensure accuracy.

Findings

Development and experience with the Water Classification Initiative in New Brunswick

The requirements for drinking water quality in New Brunswick follow the Drinking Water Quality Guidelines by Health Canada (2012). The New Brunswick Clean Water Act (1989) was established to protect the quality of drinking water (NB Government 1989) and to provide a framework for its management in the province (NB DELG interview). Several regulations were put in place in relation to land use and drinking water quality, including: Water Well, Watercourse and Wetland Alteration, Protected Area Exemption, Fees for Industrial Approvals, Potable Water, Wellfield Protection Area Designation Order and the Watershed Protection Area Designation Order. The Water Classification Regulation (2002) complements these land use-based regulations (NB DELG interview). It was developed through a multi-jurisdictional review of water classification programs and is most closely modeled after Maine’s water classification structure for rivers, streams and lakes (Barbour et al. 2000). The NB DELG refined the approach through a series of pilot tests in selected watersheds throughout the 1990s.

The Water Classification Regulation established water quality classes, associated water quality standards, and administrative processes and requirements (NB DELG 2002a). The purpose of the regulation was to “set goals for surface water quality and promote management of water on a watershed basis” (NB DELG 2002a, 1). It set out a five-step process, the WCI, through which water bodies were to be classified. While the Water Classification Regulation provided the overarching regulatory framework for classifying water bodies, the WCI is the process by which the watershed organizations undertook the classification. First, stakeholders were identified and involved. Second, water quality information was gathered by the stakeholders. Third, land and water use information was assembled. Stakeholders, using information gathered in steps two and three, then classified water bodies within the watershed according to six available classifications that ranged from “outstanding natural waters” to “acceptable water quality” (Table 1). Each classification was based on standards dictating suitable uses, prohibited activities and standards for aquatic communities, dissolved oxygen, bacteria and trophic status (see Table 1) (NB DELG 2002a). The classification could reflect the current condition of the water body or a set water quality goal. Once classified, the final step in the stakeholder process was to create and implement an action plan to sustain or achieve the water quality goals.

Table 1. Levels of water classification (NBDELG 2002a, 10–11).

Class	Suitable uses	Aquatic community standards	Dissolved oxygen standards	Bacteria standards	Standards from trophic status (lakes, ponds and impoundments only)	Prohibited activities
O (Outstanding natural waters)	habitat for aquatic life; primary and secondary contact activity; other appropriate uses	the aquatic life shall be as naturally occurring	the concentration of dissolved oxygen shall be as naturally occurring	the faecal coliform organisms and E. coli shall be as naturally occurring	the trophic status shall be as naturally occurring	release of a contaminant; creation of a new mixing zone; release of a contaminant into a mixing zone; significant withdrawals
AP (Designated drinking water supplies)	raw drinking water (treated or untreated); uses permitted under the Watershed Protected Area Designation Order (WPADO)	the aquatic life shall be as naturally occurring	the concentration of dissolved oxygen shall be as naturally occurring	E. coli shall be as naturally occurring; the total coliform organisms shall be as naturally occurring	the trophic status shall be as naturally occurring	see the Watershed Protected Area Designation Order
AL (Lakes not classified as O or AP)	habitat for aquatic life; primary and secondary contact activity (see glossary); other appropriate uses	the aquatic life shall be as naturally occurring	for cold water species: > 9.5 ppm (early life stages) and > 6.5 ppm (other life stages); for warm water species: > 6.0 ppm (early life stages) and > 5.0 (other life stages); for estuarine waters: > 80% saturation	the faecal coliform organisms and E. coli shall be as naturally occurring	the trophic status shall be stable or naturally changing; the water shall be free of algae blooms that impair use as habitat for aquatic life, or use for primary or secondary contact activity	direct discharge of a contaminant that is not being released, or any increase in the volume or concentration of a contaminant that is being directly discharged, on the date of commencement of the Regulation; creation of a new mixing zone.
A (Excellent water quality)	as habitat for aquatic life; primary and secondary contact activity; other uses that will not prevent the standards from being met	the aquatic life shall be as naturally occurring	for cold water species: ≥ 9.5 ppm (early life stages) and ≥ 6.5 ppm (other life stages); for warm-water species: ≥ 6.0 ppm (early life stages) and ≥ 5.0 (other life stages); ≥ 80% of saturation in estuarine waters	E. coli shall be as naturally occurring	the trophic status shall be stable or naturally changing; the water shall be free of algae blooms that impair use as habitat for aquatic life, or use for primary or secondary contact activity	creation of a new mixing zone; release of a contaminant into a mixing zone
B (Good water quality)	as habitat for aquatic life; primary and secondary contact activity; other uses that will not prevent the standards from being met	releases shall not cause adverse impact to the aquatic community in that the receiving water shall be of sufficient quality to support all indigenous aquatic species without detrimental changes to resident biological community	for cold water species: ≥ 9.5 ppm (early life stages) and ≥ 6.5 ppm (other life stages); for warm-water species: ≥ 6.0 ppm (early life stages) and ≥ 5.0 (other life stages); ≥ 80% of saturation in estuarine waters	the faecal coliform organisms shall be less than 14 per 100 mL for estuaries with identified shellfish beds, and E. coli shall be less than 200 per 100 mL for all other watercourses (geometric mean of a minimum of five samples in a 30-day period)	the trophic status shall be stable or naturally changing; the water shall be free of algae blooms that impair use as habitat for aquatic life, or use for primary or secondary contact activity
C (Acceptable water quality)	as habitat for aquatic life; secondary contact activity; other uses that will not prevent the standards from being met	releases that may cause some changes to the aquatic community are permitted if the receiving water is of sufficient quality to support indigenous fish species and maintain the structure and function of the resident biological community despite the releases	for cold water species: ≥ 9.5 ppm (early life stages) and ≥ 6.5 ppm (other life stages); for warm-water species: ≥ 6.0 ppm (early life stages) and ≥ 5.0 (other life stages); ≥ 80% of saturation in estuarine waters	the faecal coliform organisms shall be less than 14 per 100 mL for estuaries with identified shellfish beds, and E. coli shall be less than 400 per 100 mL for all other watercourses (geometric mean of a minimum of 5 samples in a 30-day period)	the trophic status shall be stable or naturally changing; the water shall be free of algae blooms that impair use as habitat for aquatic life, or use for primary or secondary contact activity

The NB DELG approached the WCI from a watershed basis and sought collaboration with watershed organizations (NB DELG 2002b). In some situations, watershed organizations already existed, and they took on water classification. In instances where non-governmental organizations (NGOs) existed but did not strictly focus on the watershed scale, an opportunity was presented to adjust their mandate and work on the WCI. In the remaining watersheds where no appropriate organization existed, the NB DELG identified and convened stakeholders using the Outreach and Partnership Initiative for the purpose of working through the WCI process. Throughout development and implementation it was explicit that “public involvement is a cornerstone of the Water Classification Regulation” (NB DELG 2002b, 1). Although government-mandated like Ontario’s source water protection, the process and context of the WCI were very different.

A key milestone for each watershed organization involved in the WCI process was the development of a provisional water classification document. The provisional water classification document describes in detail the procedures followed and actions taken in relation to classifying water bodies within the watershed. All watershed organizations agreeing to participate in the research completed a provisional water classification document and most were submitted between 2000 and 2005. The average amount of time reported to complete the process and submit the provisional classification document was 2.5 years. Seven watershed organizations elected to revisit their water classification, and all who did increased the classification of at least some of the water bodies within their watershed (e.g. classification was changed from “B” to “A”).

Analysis of the provisional documents illuminates the range of techniques and degrees of stakeholder involvement within the WCI. Ways in which the watershed organizations involved stakeholders included: hosting general meetings and open houses, sending letters out to the community, posting newsletters, gathering information from the public during meetings and including stakeholder input in the action planning. More collaborative engagement of stakeholders identified by watershed organizations included organizing volunteer programs to monitor the water quality in their watershed, incorporating stakeholder feedback into the classifications, having stakeholder approve action plans and actively recruiting volunteers to assist with implementation. Stakeholders engaged by watershed organizations included individuals, municipal governments, public and private organizations, government agencies and others.

Action plans, as part of the provisional water classification documents, were not required but were nonetheless drafted by 12 of the 15 watershed organizations that participated in the study. Action plans included ecological management items such as riparian restoration (Kennebecasis Watershed Restoration Committee) and improved forestry practices (Shediac Bay Watershed Association), and social and educational items such as efforts to promote sharing of information and ongoing collaboration (Petitcodiac Watershed Alliance), and community newsletters and volunteer events (Hammond River Angling Association). The watershed organizations varied in their methods for finalizing and implementing the plans. Nine indicated they had recruited volunteers to help implement the action plans, and 10 stated that they had implemented their plan.

Although the water classifications developed were intended to gain regulatory stature, none of the submissions to the NB DELG had been adopted as enforceable water quality standards and interviewees reported that the status of the water classifications was unknown at the time the interviews were conducted. This issue appeared to be at the forefront of many watershed organization representatives in a 2011 workshop where individuals expressed frustrations about status of the WCI (Baird et al. 2011). According to a recent news report at the time of writing, the status of the WCI remains “under review” (Canadian Broadcasting Company (CBC) 2013b).

The structural relationships among Water Classification Imitative actors

As watershed organizations worked through the WCI process, they drew upon other stakeholders to obtain information and for collaboration. The following section reports the structures of those connections and the roles of specific actors within the networks of knowledge exchange and collaboration during this time. The structure of the entire network of both types is that of a star-type graph, where few actors are highly connected in the centre of the graph, but the majority of the actors are situated in the periphery and are highly disconnected (Wasserman and Faust 1994).

Knowledge exchange

Most watershed organizations exchanged knowledge with a range of actors during the WCI process. The average size of a single watershed organization “ego network,” or the actors directly connected to a single watershed organization, was 16.5 actors. In some cases, watershed organizations accessed knowledge from commonly-used sources (indicated by those grey nodes in the centre of the social network diagram in Figure 1). These sources include several government departments, NGOs, and educational institutions. In many cases, however, watershed organizations relied upon several unique sources of knowledge not accessed by other watershed organizations, as is clear from the large number of nodes connected to a single watershed organization in Figure 1. Often, these sources were local or held information of special interest to a particular watershed; for example, local citizens, media and industry were often identified.

Figure 1. Knowledge exchange network. Black nodes indicate watershed organizations; the white node indicates the New Brunswick Department of the Environment and Local Government (NB DELG); grey nodes indicate non-respondents. Node size indicates scale of organization from local to international.

Focusing specifically on the respondents (watershed organizations and the NB DELG) (Figure 2), structures of communication during the WCI process become clear. Centralization of the network was 93%; this measure indicates unevenness in connections among actors in the network where 100% indicates a single central actor and 0% indicates an identical number of connections among all actors. The knowledge exchange network exhibited a highly centralized structure, with the most central actors being the NB DELG, identified by all watershed organizations, and the watershed organizations still engaged in the WCI process, who reported exchanging knowledge with all respondents. Although the watershed organizations exchanged knowledge with several sources outside the respondent network, there was a low level of connectivity among watershed organizations (Figure 2), and all but two ties among watershed organizations were not reciprocal. It is interesting to note that the NB DELG held a position of considerable power in terms of the knowledge exchange network structure during the WCI process, connecting otherwise disconnected watershed organizations. The other central actor in the network, the Miramichi River Environmental Assessment Committee, engaged in the WCI process much later and was able to draw on knowledge and experiences from others; thus, it held a unique position and advantage in the network, albeit one of less power than the NB DELG.

Figure 2. Respondents’ knowledge exchange network. Black nodes indicate watershed organizations; the white node indicates the New Brunswick Department of the Environment and Local Government (NB DELG).

Collaboration

The collaboration network was substantially smaller with fewer connections among actors in the network. The average ego network for watershed organizations was eight nodes, approximately half the size of the knowledge exchange network. This is not surprising as a greater effort is required to meet the conditions for collaboration than for knowledge exchange. Collaborations occurred more often with educational institutions, government departments and NGOs, but the network still demonstrated considerable diversity with the public, industry and other groups represented in several ego networks.

The respondent network exhibited some surprising structural qualities (Figure 3). Each watershed organization identified the NB DELG as a collaborator; however, the NB DELG did not report collaborations with the watershed organizations. As a result, the NB DELG was a highly central actor in the network without acknowledging that central role. Also surprising was the lack of collaboration among watershed organizations. While the knowledge exchange network reported some connections among watershed organizations, those connections did not constitute collaborative ties. Similar to the knowledge exchange network, the Miramichi watershed organization still engaged in the WCI process regarded all connections with other watershed organizations as collaborative and, due to the difference in time between the completion of the provisioning documents by others, none of them identified collaborative ties with this watershed organization.

Figure 3. Respondent collaboration network. Black nodes indicate watershed organizations; the white node indicates the New Brunswick Department of the Environment and Local Government (NB DELG).

Ego networks

The ego networks of watershed organizations were compared using ANOVA between those that existed prior to the WCI (n = 7) and those that were created for the purpose of working through the WCI process (n = 7). The Miramichi River Environmental Assessment Committee ego network was not included in the analysis as they were still engaged in the WCI process during the study. The mean ego network size for the knowledge exchange and collaboration networks was not significantly different between these two watershed organization types (p = 0.89 and p = 0.51, respectively).

Discussion of the social network findings

The social network analysis highlighted the multi-stakeholder approach the NB DELG promoted for the WCI. The analysis usefully reveals how the structure of the networks and the attributes of the actors connected to watershed organizations built capacity in two main ways during the WCI process. First, large networks were fostered for the purpose of working through the WCI process. These networks included linkages among scales and sectors. Cross-scale linkages have been promoted as “solutions to sustainability of community-based management” (Adger et al. 2005). The inclusion of diverse actors in the WCI brings local and expert knowledge together. Although local knowledge may have played a subordinate role in the WCI process (van Tol et al. in press), the cross-scale and cross-sector linkages increase the capacity of watershed organizations through the joint generation of local technical knowledge, identified as lacking in other source water protection initiatives (Ivey et al. 2006b). Second, the collaboration network showed a strong multi-sector, multi-scale dimension to the WCI for each watershed organization. Despite the range of contextual differences (e.g. industrial, economic, developmental, environmental and social) that can hinder inter-organizational collaboration and the process more generally (Brummel et al. 2012), all of the watershed organizations that participated in the study successful completed the WCI process and classified the water bodies within their jurisdiction.

The social network analysis also reveals some puzzling, neutral and less positive outcomes. A puzzling result was a lack of reciprocity among respondents (watershed organization representatives and the NB DELG) in the collaboration network. Collaborative ties were indicated where two actors worked together toward a common goal – a definition that is explicitly reciprocal. However, the NB DELG did not acknowledge collaborating with any watershed organizations, while every watershed organization indicated a collaborative link to the DELG. Some insight into why this may have occurred may be gained from the language used by the DELG in the WCI guidance documents provided to watershed organizations. The documents describe a “partnership” between government and watershed organizations, and focus on the watershed organizations as the administrative unit for undertaking the task (NB DELG 2002b). A provincial government presentation at a workshop held in New Brunswick to disseminate the results of this study described the partnership largely in terms of one-way information flows from watershed organizations to the government (Baird et al. 2011). Watershed organizations, conversely, reported that they collaborated with the provincial agency. This disparate view of the process may have had impacts on the outcomes and ongoing relationship between watershed organizations and the DELG. Leach (2006) highlights the critical importance of transparency in the relationship between public agencies and other stakeholders involved in collaborative management, and Brummel et al. (2012) caution that mandated collaborative stakeholder organizations may dissolve after the specified task is complete, reducing the capacity for long-term management.

Successes and challenges of the Water Classification Initiative

Interviewees were asked to reflect upon the successes and challenges of the WCI. The results presented in this section give voice to the participants and reflect the themes that emerged from the qualitative analysis.

Successes

Respondents to the interviews identified several benefits and successes that came about because of the WCI process. The first theme or category addressed the building of knowledge and collection of baseline data. Ten respondents described how the WCI helped their watershed organisation build knowledge of their watershed, in terms of both quality and quantity of information. Watershed organization respondents stated that the WCI increased their water quality monitoring capacity, and gave them the opportunity to add water quality information to their database:

For us it’s not a dead end, because we are using that document as a go forward tool anyway, we’re still using it and we’re still owning it. Just because the regulation hasn’t been put into place doesn’t mean that you can’t use the water quality data you’ve collected and try to maintain and contain and monitor on our own accord that data and water quality. It’s been a benefit in that way in which we’ve now gained a much greater understanding of water quality, not just habitat, but water quality within that habitat, within our watershed. [Be]cause we had already gone through a full, an extensive habitat assessment of our watershed, but we didn’t have water quality data, so now we’ve been able to [assess] the quality … to have a good understanding of our watershed. (Respondent from Kennebecasis Watershed Restoration Committee)

In particular, the development of a baseline set of data for the watersheds was often mentioned (n = 8) as a valuable benefit of the WCI. For example, a respondent from the Société D’aménagement de la Rivière de Madawaska et du Lac Témiscouata stated, “The first thing is the awareness of the situation of their water. The quality of the water was one thing … it really validated that dimension there” (translated from French).

A second theme concerned the increased capacity of watershed organizations. More than half of the respondents (n = 8) expressed that one of the big successes of the WCI was that it helped to create capacity for further work within their watersheds. The WCI was identified to build capacity by enabling the creation of watershed organisations where there were none previously. Where watershed organisations did exist, watershed organization respondents identified that the WCI was a catalyst for the maturation of the watershed organization (respondent from the Meduxnekeag River Association) and focused the efforts of the watershed organization (respondent from the Nashwaak Watershed Association).

The provisional water classification documents, and the action plans that arose from the WCI process, also served as a starting point, or catalyst, for future actions for several respondents (n = 6). Each watershed organization respondent who identified increased capacity as a benefit of the WCI also noted the development and availability of baseline data for each watershed, and these benefits are highly connected: for example, a respondent from the Meduxnekeag River Association stated, “[The WCI] is a catalyst for other things … [it provided an] opportunity to increase capacity and gain a better quantitative understanding of the watershed”; and a respondent from L’Association des Bassins Versants de la Grande et Petite Rivière Tracadie remarked, “[t]he successes, well [the provisional water classification document] left us, it went to the organization, but we still use the classification to continue to put the action plans in place each year” (translated from French).

A government-mandated collaborative process can be successful and persist. In the New Brunswick WCI, many of the watershed organizations tasked with classifying waters within their watershed boundaries recognized the value in the work they undertook. Although the water classifications are not yet legally recognized, several watershed organizations continued to engage in watershed-based monitoring and protection/restoration activities. In creating watershed organizations, or re-purposing existing associations, and providing them with resources to engage in the WCI, the NB DELG created capacity – through knowledge generation and synthesis – for watershed-based action within the province (Ivey et al. 2006b).

Challenges

Respondents also identified challenges that arose during their experience with the WCI process. Themes that emerged strongly from the qualitative analysis of interview transcripts include a lack of information and momentum, concerns about capacity (including funding) and lack of enforceable regulations.

Several watershed organization respondents (n = 9) stated that the provincial government was not forthcoming with them, or the public, in terms of sufficient information about the WCI once the provisional water classification documents were completed. By extension, these respondents expressed that it was their perception that the WCI was of low importance to the provincial government. One respondent from the St. Croix International Waterway Commission cited little to no transparency for how the provincial government was planning to implement and enforce the Water Classification Regulation. Other respondents (n = 3) expressed that they were suspicious of the government because of the lack of progress towards establishing enforceable regulations. The respondents interpreted this as lack of progress as the WCI no longer being a priority for the DELG; two statements from respondents explaining this follow:

[L]egitimately I think there [are] some legal issues [that] have to be addressed but I think they can be addressed. I just think that the priorities of the provincial government have shifted with shifting parties. And that happens a lot across provinces I’m sure; with shifting governments, priorities shift. And until that comes back to a top priority again we are going to be struggling against this battle. (respondent from the Kennebecasis Watershed Restoration Committee)

[The ministry] claim[s] it’s a problem with the regulations. But they have not defined what that is and in fact the new group within the New Brunswick environmental network, the water caucus, fired off a letter to the minister recently expressing concern about the lack of activity, lack of action. And suggested that we sit down with her and discuss what the issues are … (respondent from the Meduxnekeag River Association)

A second theme in the category of challenges, expressed by several respondents (n = 7), was the reduction of capacity due to less funding availability after the creation of the provisional water classification documents. For example, a respondent from the Coalition des Bassins Versants de Kent stated, “[T]he obstacles are that it’s tough to keep the momentum” (translated from French). Without secure funding, watershed organizations must apply for funding from other sources, which is not always reliable. Several watershed organization respondents (n = 4) stated that they do not have the required financial capacity to continue working on the WCI in their watershed:

Funding is the biggest challenge … Watershed groups were almost totally funded through the Department of Environment through the bottle return environmental trust fund. [It] used to be that Watershed Groups [were] the only Environmental Groups out there. Now since those funds have become available, environmental groups [have] pop[ped] up all over the place. Everyone competes for that money, and even universities get a chunk of that money now. So there’s a whole lot more hands out looking for funding. The Department of Environment tends to try to give everyone a piece of the pie, which means that the foundation for the watershed groups is not there anymore … Challenge for us is to find new funding sources, right now. Or try to figure out what hoops you have to jump through to get funding through the Department of Environment … (respondent from the Petitcodiac Watershed Alliance)

The capacity of the provincial government was also identified as a challenge by some respondents (n = 3). Staff turnover and changing priorities for government departments were cited as two significant issues in maintaining momentum and capacity within the provincial government to move forward with the water classifications. The lack of communication between the different Ministries was identified by two respondents.

The final theme to emerge within the category of challenges addressed the lack of enforceability through the Water Classification Regulation (n = 4). Respondents expressed considerable frustration as they had worked towards this expectation for over a decade and after much effort realized that they would not be able to influence water quality enforceable standards in their watershed. Even respondents who did not explicitly identify enforcement of regulations as a challenge (n = 6) commented on the lack of progress within the provincial government for this regulation and have corresponding questions about their capacity:

Well, the biggest frustration of course is that [the WCI] has not been implemented. We’ve been working on this process for many many years now and it’s been a massive project, a valuable project, an important project. It would be important to see it actually come into fruition and come into full use to really take full advantage of it. I mean it has been good because it has given us structure; we know a lot more about our watershed after having gone through that process. And it’s a good internal document, it’s certainly providing us with guidelines and targets for projects and project activity, but it just doesn’t have the teeth it should have at this time. (respondent from the Hammond River Angling Association)

The frustrations described by watershed organization representatives speak to the vulnerability of watershed organizations to political changes (Kallis et al. 2009), specifically to the period of uncertainty about the legal status of the WCI and the financial challenges. Other studies in collaborative governance describe similar challenges: Ivey et al. (2006b) describe a lack of legal capacity by local governments to protect source waters in Alberta, and highlight the potential to compromise existing strategies in place at the local level as a result of this lack of capacity. Timmer et al. (2007) identify financial capacity of organizations to engage in protection projects and maintain a balanced budget as one of five capacity indicators for successful source water protection. More broadly, the Global Water Partnership (2000) underscores the importance of political commitment to making the necessary investments in management of water resources for sustainable water resources management, and the importance of appropriate management instruments including regulations. The concerns about the lack of enforceability of the water classifications persist: recent news stories reported by the CBC (2013a, 2013b) highlighted the concerns about industrial activities on water quality in watersheds where provisional water classifications may have promoted the maintenance of good water quality, had the Water Classification Regulation been enacted.

Conclusions: Some lessons for source water protection from the New Brunswick Water Classification Initiative

Protecting source waters is an important and formidable challenge. While contextual factors that shape approaches to source water protection vary, a considerable opportunity exists to learn from experience and policy experiments in other locales (Swainson and de Loë 2010, 2011). This research investigated the WCI in New Brunswick. The WCI offered an opportunity to gain insights from a longstanding provincial policy initiative undertaken on a watershed scale which coincides with growing interest in new modes of water governance. The in-depth case study described the development of the policy initiative, examined the structural relationships among the government and non-government actors, and identified successes and challenges experienced by those involved.

In drawing upon the in-depth investigation of this case study, three lessons or insights stand out to inform the development of source water protection policy in other places, including source water protection policy and regimes beyond Canada. Practitioners from across Canada dealing with source water protection governance challenges identified the insufficient nature of simple descriptions and called for accounts to include detailed information to discern applicability to their situation (de Loë and Murray 2012). Lesson one stems from the detailed description and analysis of the case. The WCI confirms widespread assertions in the source water protection literature about the importance of the watershed scale (e.g. National Research Council 2000; Pires 2004), the need to emphasize collaboration (e.g. Hardy and Koontz 2008; Plummer et al. 2011; Simpson and de Loë 2011), and opportunities afforded by combining “science” infused with local knowledge (e.g. Ferreyra and Beard 2007; van Tol et al. in press). It also offers a useful illustration as to how this suite of key aspects may be incorporated into the design and implementation of source water protection policy. The second lesson gleaned from the case study concerns the role of government and the ability to build capacity through the policy process. The need for research related to government (or state)-led/mandated collaborative environmental management is established (Koontz et al. 2004; Brummel et al. 2012). The New Brunswick WCI case provides important insights into the extent to which source water protection planning and action can be catalyzed through policy alongside supportive efforts to engender engagement such as outreach, access to information and funding. Perceptions of the non-governmental actors involved speak to both the benefits and challenges experienced with the WCI. These highlight enabling and constraining factors for policy makers elsewhere. Perhaps most importantly, the WCI process and all those involved built considerable capacity across New Brunswick and around watershed organizations. In so doing it reinforces work by Timmer et al. (2007) in the Annapolis Valley of Nova Scotia concerning key factors which build capacity for source water protection, such as leadership at multiple scales, education and networks.

A third lesson from this research concerns the incorporation of social network analysis in future source water protection studies. The use of social network analysis in natural resource management is increasing (Bodin and Crona 2009); however, to the best of our knowledge, the analytical technique has not been employed in relation to source water protection in Canada. The results of the social network analysis in the WCI in New Brunswick indicate the importance of considering and paying particular attention to relationships among actors. A novel approach in identifying the types of interactions among both formal and informal actors, in the context of source water protection, offers insights into the patterns and type of interactions that occurred during the WCI, including an explicit demonstration of cross-level and cross-sector linkages. The specific social network analysis methodology employed provided a valuable opportunity to identify social capacity building and the nuances of interactions between the NB DELG and watershed organizations. Future research efforts to analyze or experiment with source water protection policy may benefit from empirically investigating interactions among actors, including the quality thereof.

Acknowledgements

The authors express their appreciation to the individuals, organizations and agencies who participated in this research project. We also are grateful to Dr. Cynthia Stacey and Paul Wilson for their kind assistance, support and insights. The Canadian Water Network and the Social Sciences and Humanities Research Council of Canada are gratefully acknowledged for their financial support.