Background

Protected areas cover up to 15.5% of the planet’s land surface and are amongst the most important tool to maintain habitat integrity and species diversity. Unfortunately, despite the increase in coverage, there is considerable debate over the extent to which protected areas deliver conservation outcomes in terms of species populations, habitat coverage, or habitat condition.

Ideally the success of protected areas should be measured in terms of whether they improve condition for biodiversity or habitats compared to a control scenario; often the state before their establishment or in comparable areas outside the protected area boundary. This requires an approach able to document a causal link between conservation actions (e.g. establishment of a protected areas or its management) and the observed outcomes (e.g. improved population trends for species or reduced habitat loss).

Objectives

The primary question of this review is ‘Do terrestrial protected areas maintain natural species populations and prevent habitat loss?’

Methods

Multiple electronic databases, internet engines, and the websites of specialist organizations were searched to identify published and unpublished literature relevant to the review question.

Predefined inclusion criteria were applied to each article included in the review:

Subject population: Spatially referenced units of biodiversity and/or habitat

Intervention: Establishing a protected area.

Comparators: Inside/outside and before/after establishment of protected areas, and differences in interventions.

Outcome: Changes in species abundance or habitat extent or structure

Types of study: Studies describing a trend or spatial difference in populations, or habitat cover, relating to either management or governance of protected areas, were included. Studies without a counterfactual scenario were excluded. Studies where change in outcomes could not be attributed to PA effectiveness were excluded. All factors described by the studies to have influenced the observed changes besides PA effectiveness were recorded.

Main results

In total, 35 articles containing species population time-series and 51 articles covering habitat change were included in the review. All 86 articles linked the primary intervention (protection) and the observed changes in outcomes (populations or habitats) by either comparing inside to outside the PA or before and after their establishment. However, because of the multitude of factors impacting changes in outcomes, we did not attempt to compare effect size across papers, instead recording for each study whether PAs had a) no impact, b) positive impact or c) negative impact on outcomes.

All articles were subdivided into studies based on the number of counterfactual scenarios presented, leading to 42 studies on population trends and 76 studies on habitat change. In the studies focusing on species 31 of the 42 studies reported that protected areas (PAs) were effective in protecting target species populations, when compared with a counterfactual scenario. For habitat change, 60 of 76 studies found that the rate of habitat loss was lower inside PAs when compared with a counterfactual scenario. However differences between study-design across studies as well as important regional and contextual differences (especially in the species studies) precludes us from going beyond vote counting of studies, which might bias results reporting to suggest predominantly positive or negative ones.

Conclusions Implication for policy: For species populations, the low number of studies precludes strong policy recommendations, but we do see a need to make data from monitoring and management programs available, transparent, and standardized.

For habitat protection, the review shows that PAs are an important element of conservation strategies to preserve tropical forests, which was the only habitat for which there was substantial evidence. However, we need to move from a simple understanding of whether PAs are effective or not (which can be established using remote sensing studies) to why they are effective (i.e. how ‘on the ground’ actions influence PA effectiveness, requiring in-situ research), in order to guide PA managers and improve PA performance.

Implications for research: One of the most important conclusions from this review remains the call for systematic reporting and documentation of conservation projects, as well as the inclusion of pressures and responses in the study design of ecological experiments. This includes the need for an improved methodology for the studies of population trends, using BACI (before/after and control/intervention) design to ensure that observed changes can be linked to the human conservation interventions and thus increase our knowledge on what can be done to halt the loss of biodiversity.

Keywords