Multi-scale estimation of the effects of pressures and drivers on mangrove forest loss globally

Abstract

Human activities that threaten ecosystems often vary across small spatial scales, though they can be driven by large-scale factors like national governance. Here, we use two decades of data on global mangrove deforestation to assess whether landscape-scale indirect pressures – cumulative impacts, population density, mangrove forest fragmentation, the global human footprint – and management responses (protected areas) are related to rates of mangrove loss, and whether the impacts of these activities vary by nation. By integrating rates of loss at different spatial scales into a Bayesian hierarchical model, we also assess whether national-scale patterns in mangrove loss are predicted by national regulatory quality. Globally, less fragmented forests had lower rates of mangrove loss. We observed variability among nations in the effect of pressures and management responses on mangrove loss. National regulatory quality mediated how pressures and management interact to influence mangrove loss. Protected areas had a greater benefit for slowing mangrove loss rates in countries with low, rather than high, regulatory quality, ostensibly because countries with higher regulatory quality have greater protection of mangroves outside of protected areas. High population densities were also associated with greater mangrove loss, but only in nations with low regulatory quality. We suggest that efforts to protect mangrove forests will benefit from developing solutions that consider national context and address differences in the effect of pressures and cumulative impacts. Our model can also be applied to other globally threatened ecosystems to understand how variation in local context can affect national-scale conservation outcomes.