Small-scale fisheries are increasingly conceptualized as complex social–ecological systems (Folke et al., 2005; Basurto et al., 2013). Management of complex social–ecological systems requires knowledge to monitor resource availability, make decisions about their allocation, and respond to feedbacks from the system at multiple social and ecological scales (Berkes et al., 2003; Robinson and Berkes, 2011). Typically, no one agency, organization or group is likely to have the full range of knowledge needed for effective management of complex social–ecological systems (Berkes et al., 2003; Gelcich et al., 2006). Thus, there is a pressing need to develop environmental governance models, the structures and processes by which people in societies make decisions and share power with respect to the environment (Folke et al., 2005), which can allow partnerships and cooperation among agencies and stakeholders to confront the management of such complex systems (Berkes, 2009; Gelcich et al., 2010; Robinson and Berkes, 2011). A possible way to deal with environmental governance of complex marine social–ecological systems is to incentivize shifts towards the development of new multi-level, polycentric forms of governance, partly through government-designed decentralization (Ostrom, 2010). ‘Polycentric’ connotes many centres of decision making, that are independent of each other (Ostrom et al., 1961). Polycentric systems are thus characterized by multiple governing authorities at differing levels rather than a monocentric unit (Ostrom, 1999). Each unit within a polycentric system exercises considerable independence to make norms and rules within a specific domain (such as in a family, a firm, a local government, a state or province, or a region; Ostrom et al., 1961). In theory, with such approaches, problems associated with non-compliance, power inequalities and inappropriate discrimination can be better addressed, and major investments made towards information and innovations (Ostrom, 2010). Although no governance system is perfect, research has shown how polycentric systems can have considerable advantages given their mechanisms for mutual monitoring, learning and adaptation over time (Ostrom et al., 1961; Galaz et al., 2012). Rather than relying on a single type or level of governance, a shift towards polycentrism would be