Community-managed conservation areas can be an effective strategy for conserving biodiversity while still allowing local people to benefit from the sustainable use of natural resources. However, in many parts of the world, economic and political shocks challenge the ability of communities to protect biodiversity while also meeting their basic needs.

In Madagascar, threats to both people and biodiversity are severe, due to high rates of poverty and food insecurity, coupled with unsustainable rates of deforestation. Since its independence in 1960, the island country has suffered from repetitive political crises, but the one from 2009 to 2014 was the most severe and prolonged in the nation’s history, exacerbating both humanitarian and environmental issues. Given that much of the world’s biodiversity is concentrated in politically fragile nations like Madagascar, understanding whether community managed conservation areas are as resilient as state-managed protected areas to a crisis is essential if we want to avoid species extinctions and the loss of ecosystem services.

In our recent paper, we explored the impacts of the Madagascar political crisis, and its immediate aftermath, on the performance of community managed forests. We compared deforestation rates in Community Forest Management (CFM) and protected areas administered by Madagascar National Parks (MNP), as well as all forests nationally, before, during, and after the political crisis. In Madagascar, CFM were established starting in the late 1990s with the objective of preventing forest loss, while also providing a sustainable flow of forest products to benefit local communities. MNP protected areas are much older and were established to protect forests and biodiversity, not to provide livelihoods for local populations, except from tourism in some protected forests. Unfortunately, MNP, CFM, and forested areas all experienced deforestation before, during, and after the political crisis (Figure 1). Rates of deforestation were higher in CFM than MNP in all three time periods.

Unlike previous research showing spikes in illegal logging (especially of rosewood and other precious hardwoods) during the crisis, we found that deforestation rose significantly only in the last year of the crisis and for several subsequent years (2014-2017) before returning to (or slightly above) pre-crisis levels. To our knowledge, this post-crisis deforestation spike has not been previously reported and, to date, cannot easily be explained by us or other experts working in the region.  In the paper we propose several theories, and we hope that future research will explain this worrisome phenomenon and explore whether other conservation hotspots suffering political crises experience similar patterns.

Our research also underscores the careful analysis required to investigate these kinds of questions.  In our case, CFM differ systematically from MNP in ways that are expected to increase pressure; they are smaller, more fragmented, and are located closer to roads and human population centers (Figure 2). If not controlled for, these confounding factors would bias estimates of deforestation outcomes in CFM and MNP, making it difficult to tease out the effects of the political crisis.   To address this, we used quasi-experimental methods: statistical matching to control for differences in location and other time-invariant characteristics, and an event study estimator to control for differences in time-variant variables such as climate and commodity prices. This enables us to make a more robust ‘apples to apples’ comparison, as well as study the effects of the crisis over time.

After controlling for confounding factors, we found that CFM and matched areas within MNP both performed poorly during the crisis years, but that their performance was not statistically significantly different from one another until the end of the crisis (Figure 3). In other words, CFM and MNP were equally vulnerable to the effects of the crisis. In the post-crisis period, when deforestation rates spiked nationally, CFM performed significantly worse than matched areas within MNP. This indicates that when deforestation pressures intensified as the economy began recovering from the crisis, community management withstood forest loss pressures less well than MNP did. This is a discouraging finding, as there was hope that community-managed conservation would effectively protect forests in Madagascar.

In summary, we found that the crisis and post-crisis dynamics negatively impacted Madagascar’s forests under both CFM and MNP designation. Community managed areas were especially vulnerable to the post-crisis spike in deforestation pressure, likely in part due to insufficient and inconsistent financial support. Given extremely high rates of poverty and food insecurity, it is understandable that such areas face especially intense pressures, as there is no financial incentive for protecting them. Clearing forests for agriculture and charcoal production was likely a way communities met their basic needs, but at the cost of biodiversity. Similarly, intensifying economic pressures can make it difficult for local people to restrict or prevent logging for the lucrative timber trade by external actors. The implication of these findings for conservation is that in under-resourced nations like Madagascar, community forests will need much more external support from government, international NGOs, or multilateral banks to avoid further forest clearing and associated biodiversity loss.