"Unfortunately, the evidence base for the effectiveness of protected areas in reducing deforestation and forest degradation is weak (as is the evidence base for most conservation interventions)," Paul Ferraro of Georgia State University told environmentalresearchweb. "Unlike in fields such as public health, much of the empirical work in conservation science fails to control for confounding factors that can mask or mimic the impacts of conservation programs and policies."

Around 11% of global land surface is currently protected in this way. According to Ferraro, protected areas are frequently located on low productivity lands, a characteristic that reduces the likelihood of human disturbance even in the absence of legal protection. "We wanted to show how one could control for these confounding factors and why such control can make a huge difference in our understanding of protected area effectiveness," he said.

Together with colleagues from the International Food Policy Research Institute, Duke University, the University of Alberta, and Costa Rica’s Centro Agronómico Tropical de Investigación y Enseñanza, Ferraro analysed the impact of protected areas on deforestation in Costa Rica between 1960 and 1997.

In 1960 the country had around three million hectares of forest: by 1997 more than one million hectares had been cleared and around 900,000 hectares put under legal protection. This system of protected areas has been widely praised, and Costa Rica is spearheading moves to get avoided deforestation credits included in international climate change conventions.

"Measuring this 'avoided deforestation' is difficult because it is unobservable: the forests were in fact protected, and we cannot go back in time and re-run the last four decades of deforestation without protected areas," explained Ferraro.

By using matching methods to allow for factors such as land use productivity, distance to forest edge, distance to roads and distance to nearest major city, the team believes it has shown how to quantify the amount of avoided deforestation.

The new technique calculated that 11.1% of areas in the study that had been protected before 1979 would have been deforested by 1997 if the protection hadn’t been in place. In contrast, conventional methods implied that 44% of the plots would have been deforested.

"Approaches popularly used in the scientific literature to measure avoided deforestation substantially overestimate it because they fail to recognize that protected areas are not randomly assigned across the landscape," said Ferraro. "Establishing a protected area is politically contentious, and thus protected areas are often located where there is the least resistance. Such locations usually correspond to areas that are not at high risk of deforestation. Thus the protected areas have a smaller effect on land use patterns than they might have otherwise."

Ferraro believes nations interested in reducing deforestation may not achieve as much as they hope unless they pay more attention to where they locate their protected areas. "Our results suggest that protecting ecosystems and their services in the future may require investments that are substantially different, in size and in nature, from those made in the past," he explained.

The researchers say their results also imply that recent efforts in conservation planning to jointly consider benefits, costs and measures of threat of conversion are warranted.

"Sellers of avoided deforestation credits in a REDD scheme have an incentive to erroneously posit high counterfactual deforestation rates, and thereby claim emission reductions that did not actually take place," said Ferraro. "Analyses like ours can help REDD program designers avoid such adverse outcomes by revealing the design characteristics that affect forest conservation program success."

Now the researchers plan to apply the methodology to other nations to build up a global evidence base for protected area effectiveness, to "move beyond looking at protected areas as a monolithic concept and clarify how different kinds of protected areas, and different ways of locating them in the landscape, affect deforestation and local human welfare", and to measure the impacts of protected areas on humans that live in neighbouring communities, an issue that is "subject to contentious international debate but for which there is little credible evidence to support the opposing views".

The team reported the results in PNAS.