The French network of PAs is composed of different types of protection elaborated at the national, European, or international scales: regulatory protection, land acquisition strategy, and contractual protection. The French regulatory network is constituted of NNRs (0.33% of the national territory, Fig. 1), regional nature reserves (0.08% of the national territory), biotope protection areas (0.3% of the national territory), national parks (0.66% of the national territory), and biological reserves2 (0.08% of the national territory). Considering the area covered by NNRs compare to other PAs and the fact that along with national parks and biological reserves, they are the only instrument allowing for the active management of species and habitats, NNRs represent a key tool in France to protect biodiversity.

Fig. 1.

French procedure for the creation of a national nature reserve. NCNP: National Council of Nature Protection; SC: standing committee.

(0.15MB).

NNR creation followed a procedure summarized in Fig. 1. An NNR project was sent to the standing committee (SC) of the National Council of Nature Protection (NCNP)3 in charge of validating the “scientific interest” of the project. Then, the Ministry of the Environment referred the matter to the local state authorities in charge of the area to proceed with a public inquiry and consultations. A new version of the NNR project was then submitted to the NCNP for a final non-binding opinion on the project. Finally, after ministerial consultation, the decree for the creation of the NNR was signed by the Minister of the Environment.

In this article, we focus mainly on the activity of the SC of the NCNP, which appears to be central in the decision-making process. The NCNP is an expert body that was reformed in 1978 to strengthen the representation of scientific institutions and nature conservation groups. Before the reform, SC meetings brought together three to five scientists (i.e. holding a teaching or research position in a university or research institution) who were specialized in ecology, botany, or zoology. After the reform, the SC was composed of ten members: seven scientists of whom four were also members of local or national nature conservation groups, one agricultural representative, and two representatives from the Ministry of Agriculture and Equipment, respectively. The SC is thus a relevant institution to understand the role of scientists and expertise in this particular decision-making process involving different actors, sectors, and interests. To better comprehend how scientific statements were integrated into the decision-making process leading to NNR creation, we studied the archives of NCNP meetings from 1970 to 1985 (see Appendix S1 for data availability and content).

Our analysis was performed on 176 NNR projects, which led to the creation of 72 of the current NNRs. To assess the impact of scientific statements in the decision-making process, we used the detailed record of each NCNP meeting to determine the different variables that might influence NNR creation. From 1970 onwards, these meetings follow more or less the same procedure: a presentation of the “scientific interests” of the NNR project by a rapporteur (NCNP member or invited guest), possible threats to the site, insights into the local situation, and possible issues raised by the creation of the NNR, followed by a discussion among the SC members. A first qualitative analysis allowed us to identify three major components of the scientific interest: “biodiversity” (i.e. species or ecosystem richness) (yes/no), “remarkable species or ecosystems” (i.e. rare, endemic, or endangered) (yes/no), “representative species or ecosystem” (i.e. species or ecosystem characteristic to a region) (yes/no). We added another variable representing the “overall scientific interest” (i.e. the sum of the previous scientific features, with a score between 0 and 3). We also assumed that the policy process could be influenced by the existence of a threat (i.e. a development project in or around the project area that could alter the ecosystem) (yes/no), the level of conflict raised by the NNR project (i.e. number of stakeholders mentioned as against the presented NNR projects: 0; 1; 2 or more), the NCNP reform (pre-reform/post-reform), and the date of discussion at the SC (NNR projects were divided into four categories: 1970–1975; 1975–1978; 1978–1980; 1980–1985).

The influence of these factors was analyzed by logistic regression (glm procedure using a logit function with a binomial error). To assess the effect of the NCNP reform in the decision-making process, the interaction with NCNP reform and the other variables was also tested. Considering that “biodiversity”, “remarkable” and “representative species and ecosystems” were not independent from the “overall scientific interest” and that the NCNP reform and the date of discussion were not independent, we performed logistic regression on four different models and showed only the model that best fit the data (based on AIC).

To better interpret these quantitative results, we also performed a more detailed qualitative analysis on the archives of three NNR projects (Fig. 2; Table 1; see Appendix S1 for data availability). We chose three seemingly different projects to cover the diversity of mechanisms responsible for the success or the failure of a project. Archives are composed of the scientific dossier, letter exchanges between stakeholders, legal documents, and results of the consultations (i.e. public inquiry, local and ministerial consultations). We specifically examined: (i) the type of actors involved in the procedure, (ii) their relations, (iii) the arguments used against or for the project, and (iv) their methods and actions to influence the result of the procedure.

Fig. 2.

Map representing the current natural nature reserves (NNRs) in France. NNRs are shown in red, while the three case studies used for the qualitative analysis are indicated. The NNRs located in the Val de Munster area on the map are part of more recent NNR projects than the one studied in this article.

(0.62MB).

The main characteristics of the projects are summarized in Table 2.

In the first model (Table 3), time and level of conflict significantly influenced NNR creation. The likelihood that an NNR project was finalized decreased with time. Moreover, there was a significant increase in the likelihood of NNR creation in areas with more conflict (i.e. at least two actors opposing the project). Other factors were non-significant and were removed from the final model.

In the second model (Table 3), NCNP reform and its interaction with the existence of a threat significantly influenced NNR creation. The likelihood that a project led to NNR creation decreased after the NCNP reform. The first model that best fit the data would suggest that the significant effect of the NCNP reform on NNR creation was likely due to the effect of time. However, after the reform, the likelihood that a project led to NNR creation increased with the existence of a threat on the site. The level of conflict only had a positive marginal effect on the likelihood of NNR creation. Other factors were non-significant factors and were removed from the final model.

The variable “overall scientific interest” was not significant in the two other models (data not shown).

The NNR creation procedure has barely changed since the studied period. However, the relationships, responsibilities, and balance of power between the actors have evolved over time. Most of the NNR projects studied here were created before the French decentralization law in 1982–1983. As this reform gave local authorities greater power and responsibility over land-use planning, this only reinforce their importance on NNR creation (see below). Some authors have also highlighted how environmental policy instruments used by worldwide state authorities have evolved from regulatory to incentive-based approaches (Jordan et al., 2003), which necessarily modify people's perceptions about regulatory instruments and thus their acceptance of them. This evolution also concerned PAs, while new non-regulatory instruments have been created to protect such sites, including sites of community interest based on the European Habitats Directive or the “blue and green ecological network.” Acknowledging these differences, we can nevertheless discuss the general mechanisms most likely to influence the procedure for PA creation today.

Our results show that the positive opinion of the NCNP was required for the creation of an NNR, which suggests that expert views were taken into account in the early phase of the NNR creation procedure. However, the scientific validation of a project was far from sufficient for the creation of an NNR. After acknowledging the scientific interest of the project, its success was independent of its ecological relevance. No specific or combination of scientific features significantly increased the probability of success of an NNR project (Table 3). According to our case studies, the positive outcome of a project was rather determined by the local balance of power. During the negotiations, opponents most often defended extreme positions – i.e. abandoning the project – not by refuting the scientific arguments but rather by contending other matters such as the violation of property rights, the loss of elected officials’ independence, or economic development constraints. It would thus seem that scientific relevance and sociopolitical matters were examined separately and in different places and times during the procedure. Considering that only 45%4 of the 159 projects that received a positive expert opinion were successful, the second part of the process seems to be decisive.

Our results thus reveal a two-step process. First, an expert body selects the NNR projects that deserve protection from among the proposals. Second, another selection is made based on social, political, and economic interests. Even though this process may not be optimal, as it fails to prioritize more relevant projects based on the expert body's opinion, it reveals that scientific opinion is still taken into account during the process. In our specific case, scientific interest operates as the primary filter. This means that all the current NNRs were considered to be sufficiently interesting by the experts. As revealed by the recent analysis of expert opinions (Chassé, in press), this filter values projects according to their taxonomic diversity (i.e. rare, endangered, or endemic species as well as species and ecosystem richness). While the taxonomic approach may be criticized (see, for instance, Cadotte et al., 2011; Rosauer et al., 2017) and the biological models improved, this is not, in our view, the most important issue at stake.

Our analysis of the decision-making process reveals that the limiting factor was not the scientific relevance of the projects or the availability of biologically interesting projects. The most important filter was the second phase that selects projects according to other criteria. Improving the biological models would thus only replace older projects without increasing their chance of success. This observation is particularly pertinent in order to overcome the “research-implementation gap.” We are in agreement with Toomey et al. (2017) who considered that this gap suffers from a misconception of its problems and solutions. It is often perceived as the result of a lack of communication and the inability of intermediaries to translate science into policies. This is regularly accompanied by calls for a more evidence-based conservation approach. In our opinion, this reveals a misconception about what is happening in the “real world.” First, it ignores research from different fields of the social sciences (for a review of the arguments, see Toomey et al., 2017). Second, our analysis of a particular decision-making process strongly supports the fact that more evidence would not lead to better conservation. To overcome this issue, we agree with several authors who maintain that the scientific community must favor “informed opportunism” (Game et al., 2011; Knight and Cowling, 2007; Noss et al., 2002). This approach does not mean abandoning systematic conservation planning but rather acknowledging that other considerations (e.g., social, economic, political) may also be legitimate and should be taken into account. Maintaining a simplistic vision of the decision-making process – i.e. the non-use of scientific opinions accompanied by a lack of political will – prevents reflections about other possible solutions, which can, in our opinion, improve the success of NNR projects.

Our results indeed provide insights into the factors influencing NNR creation. Increasing the percentage of successful projects by identifying and addressing these factors is of primary importance to improve the efficiency of the PA network for biodiversity conservation. Our quantitative analysis identified time, level of conflict, and after the NCNP reform, the existence of an on-site threat as the main factors influencing NNR creation (Table 3). In the following section, we discuss the potential mechanisms that explain their effects and several directions to improve the PA creation process.

First, the effect of time is most likely explained by the well-known issue of scarce resources being available for conservation policies (Bottrill et al., 2008). In France, the main income source for the NNR budget is the state budget. If public expenditure remains unchanged, the success of a new project decreases the budget per NNR. In a context of highly constrained public spending, this is probably responsible for the limited growth in the number of NNRs. While inadequate funding for PAs was found to be a detrimental factor for biodiversity conservation within PAs (Coad et al., 2019; Watson et al., 2014), our results also stress that the lack of public expenditure slows down the extension of the PA network. This shows once again the urgent need to increase and find innovative sources of funding for PAs (Watson et al., 2014).

Second, the positive effect of a high level of conflict surrounding NNRs is counterintuitive and somewhat difficult to interpret. It highlights that despite a possible bias toward lands with lower productivity, there are still conflicts over land use. Even if the case of the Val de Munster NNR shows that too much conflict can lead to the abandonment of the project, our quantitative inquiry also shows that these conflicts can be overcome. We assume that the success or failure of a project is less related to the number of actors opposing it than to their identity and strength (e.g. ability to use personal contacts). In our opinion, this last point is critical and offers some opportunities for conservation strategies. The failure of the Val de Munster NNR and the success of the Marais de Bruges and Plan de Tuéda mostly depended on the efforts of local authorities. In the first case, their opposition was explicitly mentioned in the abandonment of the project. In the second, the desire of the local authority of Bruges to create an NNR was rewarded despite the financial cost. Finally, the Plan de Tuéda was required by the state as an offset measure for a ski development and thus received the support of the local authority. Considering that PAs, especially in France, are frequently located in areas with disagreements over land use, the support of local authorities lends greater weight to the PA projects and helps to overcome conflicts. This result contributes to the call for providing insights into social attributes that matter in conservation planning processes (see, for instance, Ban et al., 2013; Knight et al., 2010; Pasquini et al., 2010).

In systematic planning, there is a growing recognition about the importance of assessing the social features of a given area. In our opinion, the social proxies that are currently used in such analyses are too general (see, for instance, Whitehead et al., 2014) and gathered without knowing if they really matter during the decision-making processes. Even if social data collection is both time and money consuming (Knight et al., 2010) and with varying factors depending on the type of PA or the spatial scale (Guerrero et al., 2010; Pasquini et al., 2010), it is still necessary. By revealing the importance of local authorities in NNR creation, our research provides an interesting proxy to use for implementing the “informed opportunism” approach and tracking conservation opportunities. Conducting similar research is needed to help identify more relevant variables and improve the success of PA implementation.

Finally, the positive effect of the existence of a threat provides another interesting perspective about the decision-making processes that should be taken into account. This suggests that after the NCNP reform, NNR creation seemed to follow to an “ad-hoc process” based on an impact-driven approach. NNRs were more frequently located in areas subject to direct threats on biodiversity such as urbanization, wetland draining, or natural resource overexploitation. While the previous NCNP included almost exclusively scientists from Paris and its surroundings, the inclusion of regional nature conservation groups within the NCNP probably guides such strategy. This result implies that the procedure was often used as a tool to stop development projects. Considering the issue of scarce resources for conservation (see above), it necessarily diminishes the opportunity to build a science-based network. However, a purely scientific strategy of planning tends to ignore biodiversity impacts outside of the priority-set boundaries. If other instruments such as the environmental impact assessment procedure exist to ensure biodiversity conservation in these more common areas, their effectiveness is widely debated (Bigard et al., 2017; Weissgerber et al., 2019). Although NNRs were initially not designed to fulfill this role, we believe that the systematic conservation planning of PAs should not limit the possibility of initiating other actions to conserve biodiversity. In our opinion, such impact-driven strategy contributes fully to limit biodiversity loss and needs to be integrated in the implementation of the “informed opportunism” approach.

Our results have important implications for the conservation community, because they suggest that the production of conservation-oriented knowledge is far from sufficient to reverse the current trend of biodiversity loss. The reasons for the existence of a “research-implementation gap” are far more complex than the simple mistranslation of science into policy. The lack of data on conservation decision-making processes leads to the widely shared misconception about the problem and its solution, which can be counterproductive. More than ever, the causes of this gap need to be better understood and addressed. By applying and sharing the kind of research that we recently called for (Chassé et al., 2020), we hope to highlight the importance of social sciences for biodiversity conservation and demonstrate the value of an interdisciplinary approach.