Effects of spatial autocorrelation and sampling design on estimates of protected area effectiveness

Estimating the effectiveness of protected areas in reducing deforestation is useful to support management decisions. This information helps underpin whether to invest in better management of areas already protected or to create new ones. Statistical matching is commonly used to assess this effectiveness, but regional differences in protection effectiveness and the presence of spatial autocorrelation are frequently overlooked. We assessed methods to estimate the effectiveness of protected areas, using Colombia as a case study. We employed statistical matching to account for confounding factors in park location, and controlled for spatial autocorrelation to determine statistical significance. We compared the performance of different matching procedures - ways of generating matching pairs at different scales - in estimating protected area effectiveness. Differences in matching procedures affected the performance of matching, resulting in different estimates of the effectiveness of protected areas. Independent matching resulted in the greatest covariate similarity between matching pairs. On average, 95% of the variables in each region were balanced with independent matching, whereas 24% of variables were balanced when using the method that performed worst. The best estimates suggest that average deforestation inside protected areas in Colombia was 40% lower than in matched sites. Protection significantly reduced deforestation but its effect differed among regions; protected areas in Caribe were the most effective, while those in Orinoco and Pacific were ineffective. Our results demonstrate that accounting for spatial autocorrelation and using independent matching for each subset of data is necessary to draw inference about the effectiveness of protection. Not accounting for spatial autocorrelation can distort the assessment of protection effectiveness, increasing Type I and II errors and inflating the effect size. Our methodology allows improved estimates of protection effectiveness across scales and under different pressure conditions, and can be applied to other regions to effectively assess the performance of protected areas. Article impact statement: Spatial autocorrelation and sampling design can distort estimations of protected area effectiveness at reducing forest loss. This article is protected by copyright. All rights reserved