Quantifying the checks and balances of collaborative governance systems for adaptive carnivore management.
Recovering or threatened carnivore populations are often harvested to minimise their impact on human activities, such as livestock farming or game hunting. Increasingly, harvest quota decisions involve a set of scientific, administrative and political institutions operating at national and sub-national levels whose interactions and collective decision-making aim to increase the legitimacy of management and ensure population targets are met. In practice, however, assessments of how quota decisions change between these different actors and what consequences these changes have on population trends are rare. We combine a state-space population modelling approach with an analysis of quota decisions taken at both regional and national levels between 2007 and 2018 to build a set of decision-making models that together predict annual harvest quota values for Eurasian lynx (Lynx lynx) in Norway. We reveal a tendency for administrative decision-makers to compensate for consistent quota increases by political actors, particularly when the lynx population size estimate is above the regional target. Using population forecasts based on the ensemble of decision-making models, we show that such buffering of political biases ensures lynx population size remains close to regional and national targets in the long term. Our results go beyond the usual qualitative assessment of collaborative governance systems for carnivore management, revealing a system of checks and balances that, in the case of lynx in Norway, ensures both multi-stakeholder participation and sustainable harvest quotas. Nevertheless, we highlight important inter-regional differences in decision-making and population forecasts, the socio-ecological drivers of which need to be better understood to prevent future population declines. Synthesis and applications. Our work analyses the sequence of decisions leading to yearly quotas for lynx harvest in Norway, highlighting the collaborative and structural processes that together shape harvest sustainability. In doing so, we provide a predictive framework to evaluate participatory decision-making processes in wildlife management, paving the way for scientists and decision-makers to collaborate more widely in identifying where decision biases might lie and how institutional arrangements can be optimised to minimise them. We emphasise, however, that this is only possible if wildlife management decisions are documented and transparent.