Rethinking Our Approach to Protected Areas for Conservation

Written by Justine Atkins

Over the last fifty years, there has been progressively more widespread recognition that species’ biodiversity is rapidly declining. This is a huge problem, and not only ethically: biodiversity also has crucial economic returns such as ecotourism and promoting ecosystem resilience to climate change and invasive species. It is now well-established that the overwhelming responsibility for this decline rests firmly on our shoulders. Therefore, humans must change the way in which we interact with the environment.

One of the key ways in which we have responded to this ecological crisis is through the establishment of protected areas. These areas of land or ocean are sectioned off and restricted from human use, (theoretically) protecting the ecosystems within them from negative anthropogenic impacts such as deforestation and hunting. At least four international treaties have been established with the aim of protecting a representative example of all ecosystems and species types that exist in the world today [1]. Most recently, the Convention on Biological Diversity (CBD) set targets of protecting 17% of terrestrial area and 10% of the world’s oceans by 2020, to be specifically achieved through the establishment and expansion of strategically designed and managed protected area (PA) networks.

Perhaps surprisingly, global PA coverage is actually moving steadily towards these targets. Unfortunately, equally surprising is that biodiversity continues to decline despite this increased investment in conservation. The disconnect between the potential and realized impact of these reserves has led scientists to begin questioning the efficacy of PAs as a strategy for conserving biodiversity. This shift in perspective has, in turn, forced researchers to look more closely at how the effectiveness of PAs is assessed. Past evaluations have proven inconclusive, demonstrating both the huge benefits and significant shortcomings of protection. For example, many populations of large mammals within Africa’s reserves are still showing declines, while, on the other hand, raptor species in Botswana are much more abundant within PAs than outside these areas.

Why are there such contrasting outcomes? The answer involves several complex and interacting issues. Firstly, there is a wide variety of ways in which PAs can intervene in the environment — each with its own set of costs and benefits. Marine PAs, for example, can save declining fisheries stocks but are likely to negatively affect species living outside these areas as fisherman move their activities to neighboring areas. Secondly, establishing a PA has complicated socioeconomic impacts; these can range from being helpful, such as providing jobs, to harmful, such as forcing indigenous people off their land. Thirdly, because we lack a unified framework, past assessments of the consequences of any particular protection method have had to rely on “before-and-after” style analysis. This method is problematic because it fails to account for what would have happened to, for example, the biodiversity in a PA if that area had not been put under protection. Making direct comparisons to a baseline of conditions is crucial in science and is also referred to as the use of “control groups” (Box 1).

Box 1. Impact evaluation
Impact evaluation (IE) is a method of assessing the potential or realized consequences of a conservation policy (e.g. protected areas). IE involves the use of scientifically rigorous paired comparisons to assess the effects of an intervention. Unlike performance measurement, which monitors changes in ecological and socioeconomic indicators (such as number of species within a given area) over time, evaluating the impact of a protected area also accounts for changes that might have occurred in that area even in the absence of protection. In this way, we can get a picture of the transformations that have occurred within a protected area that are i) the direct result of the establishment of protection and ii) simply due to natural fluctuations in ecosystem characteristics over time and space.

Several closely related experimental techniques are key to the IE method:

- Control groups are groups of test subjects or sites which very closely resemble the subjects or sites receiving an experimental treatment (for example, a clinical trial of a new prescription drug) but are not themselves subject to the treatment. These groups ‘control’ for factors beside the treatment that could influence the outcome.

- Matched pair experimental design directly compares each ‘treated’ site or subject with a matching ‘control’ site or subject.

- Counterfactual is a quantified assessment of what would have happened if there had been no intervention in an area, or, to continue with the drug trial example, if no medication had been given to a sick patient.

A great ecological experiment that incorporates these techniques is found at La Selva research station in Costa Rica in which many comparative experiments are being carried out using plots of intact primary forest paired with plots of land that are, in all ecological aspects (e.g. elevation, gradient, size), very similar but were cut down or burnt different numbers of years ago (e.g. 10, 20, etc.) for a variety of purposes.

Recognizing the urgency of this problem for biodiversity conservation, the prominent journal Philosophical Transactions of the Royal Society B (Phil Trans for short) emphasized the need for a revised methodology of PA assessment in a recent special issue. As a way forward, this collection of articles proposes and presents several applications of a new control group-oriented technique called impact evaluation (Box 1). Impact evaluation (IE) is a growing field in conservation science. Like previous assessment strategies, IE measures the effects of an intervention (such as building a new PA). Unlike before, however, IE also explicitly considers what would have happened without any intervention, described by researchers as the “counterfactual” (Box 1).

The Phil Trans articles convincingly argue that considering the counterfactual is the only way to truly quantify how protected areas affect biodiversity, ecosystem conservation, and human welfare. Collectively, the authors show that this new method is crucial given the limited budgets in conservation. If implemented on a broad scale, IE could allow for much greater payoff in protected area development than is currently being observed.

While this goal of minimizing biodiversity loss and maximizing socioeconomic benefits may seem ambitious, there is already empirical evidence that suggests such a goal is within reach. So far, IE research has concentrated on measuring PA effectiveness in relation to changes in deforestation rates and species loss. With a baseline of unprotected areas for comparison, researchers have specifically quantified how the impact of a PA changes according to variation in environmental and socioeconomic characteristics. The Phil Trans issue documents how this approach is applied effectively in areas as varied as the Brazilian Amazon and freshwater systems in Northern Australia. Results of IE show, for example, that PA management has led to a greater reduction in the spread of an invasive mimosa plant in Kakadu National Park than would have been observed in the absence of a PA.

Great Barrier Reef
The Great Barrier Reef on the eastern coast of Australia is one of the largest marine protected areas in the world. While protection has allowed this ecosystem to remain a highly complex and biodiverse environment, this status and the future of the reef’s marine life remain under increasing threat. In large part, this is because it was initially judged ‘effective enough’ to have only a small percent of the area as specific ‘no-take’ zones, with commercial fishing and oil and gas exploration still allowed in many regions. This approach, now subject to ongoing review and revision, calls into question current methods for measuring the effectiveness of protected areas. Photo credit: Jurgen Freund / NaturePL

Socioeconomic effects are also more readily identified within the IE framework. The debate over PAs and poverty is long-running and controversial. In large part, this disagreement is because there is weak and inconclusive quantitative evidence of the impact of PAs on people. People instead rely on highly subjective assumptions and inconsistent anecdotal findings. However, under the direction of IE, conservation scientists can recommend reserve strategies that more accurately describe the direct benefits and costs of PAs for human welfare. For example, in Bolivia, an impact-based assessment of PAs provided empirical support for earlier qualitative findings that PAs are in fact not linked with poverty traps. Using pre-, mid- and post-implementation IE, researchers found similarly counterintuitive results in relation to the socioeconomic impacts of marine PAs in Indonesia.

It is understandably difficult to see why such an approach has not been the established practice for many years. Unfortunately, those in a position to transition to using IE in the context of PAs have little incentive to do so. In an increasing publication-centered field, academics are reluctant to commit to projects evaluating the impact of conservation initiatives because i) generally, prestige journals prefer to focus on core science and are less likely to publish such work and ii) funding for this line of research is limited. A recent opinion piece in Conservation Magazine from the directors at the Wildlife Conservation Society highlights this paradox.

On a more hopeful note, international funding agencies could be a lifeline for the facilitation of IE. These organizations not only have monetary means, they are also key stakeholders in that they control a vast proportion of the funding for PAs and would benefit greatly from the decreased cost: benefit ratio that impact evaluation could deliver.

In light of the continuing rate of species’ extinctions on our planet, it is crucial that we invest in preserving biodiversity. Protected areas have the potential to be highly valuable conservation tools, but to achieve this potential, we need a good way of objectively assessing the effectiveness of PAs. The Philosophical Transactions issue clearly demonstrates that widespread uptake of IE could fulfill this need and re-establish protection-based strategies as a cornerstone of conservation science.


[1] The Stockholm Declaration (1972), World Charter for Nature (1982), the Rio Declaration at the Earth Summit (1992), and the Johannesburg Declaration (2002).


Justine is a first-year PhD student in the Ecology and Evolutionary Biology department at Princeton University. She is interested in the interaction between animal movement behavior and environmental heterogeneity, particularly in relation to individual and collective decision-making processes, as well as conservation applications.