Why biodiversity declines as protected areas increase: the effect of the power of governance regimes on sustainable landscapes

Achieving sustainable landscapes that integrate food production with biodiversity conservation remains challenging, particularly in the tropics where most forest clearance results from conversion to industrial agriculture. Land-sparing (delineating protected areas and intensifying agricultural production from developed land) has often been identified as more effective than land-sharing (mixing protection and production in an agro-ecological matrix) for biodiversity in the tropics. Nevertheless, biodiversity decline continues despite protected area expansion meeting global targets under international conventions. We developed a low-order stock-and-flow model to consider this apparent paradox, and used it to structure deliberations on the impacts of the power of governance regimes. The model articulates our shared hypothesis about the basic dynamics of the social–ecological system. We present scenarios that depict plausible biodiversity change over time under three different governance regimes and land-use trajectories. The scenarios raise the possibility that, while land-sparing gives better short-term results for biodiversity, land-sharing may outperform it over time. Two key insights derive from our deliberations. First, the forces that drive forest clearance for development do not necessarily oppose those that drive forest protection; this decoupling helps explain why biodiversity loss continues as protected areas increase. Second, the power of the governance regimes that protect existing forest can be weakened by protected area expansion, through lowering public discourse about risks from biodiversity loss, while the power of governance regimes for development concurrently remain strong; this helps explains why some REDD+ schemes are associated with increasing deforestation. These insights suggest novel leverage points for sustainable tropical landscapes, such as prioritising protected area placement by proximity to active agricultural frontiers, rather than by representative biodiversity or cost-effectiveness; or using area-based conservation targets that include both the extent of protected areas and of other remaining forest habitat. We recommend further investigation of these ideas, and of collaborative conceptual modelling approaches, to explore solutions for sustainable tropical landscapes.