As he proposed conservation science as a new discipline, Michael Soulé (1985) defined its mission as providing « principles and tools for preserving biodiversity ». As they try to promote change to limit the impacts of human activities on ecosystems, conservation scientists are not only interested in improving the state of knowledge on ecosystem functioning but are also increasingly struggling with the social dimensions of conservation policies and practices (Mascia et al., 2003). Indeed, one of the crucial issues in managing conservation issues (such as implementing a strategy for the long-term conservation of Tanzanian elephants, or for the restoration of coastal ecosystems) lies in the absence of an organizational centre or « unity of agency » (Mermet, 2013 ). Rather, we typically find a complex interplay of multiple managers and stakeholders who act in a fragmented, divisive, often competitive, even adversarial way on ecological systems. The crux of the challenge thus consists in designing, implementing and managing viable forms of organized collective action (Crozier, Friedberg, 1980) suited to such complex and unstable problems and contexts. In their efforts to do so, conservation scientists have been developing a large variety of « calculative practices » (Porter, 1995; Power, 1996; Miller, 2001) in order to inform decision making and management processes. Such practices include for instance ecological indicators informing on the level of fish stocks, performance metrics to measure the effectiveness of ecosystem restoration, Red Lists of threatened species useful for setting conservation priorities, ecological footprints maps informing cities of their impact on natural resources, economic valuations of the contributions of ecosystems to human well-being etc. As the number and scale of conservation issues keep increasing, the ever-increasing efforts to develop such information systems have only mixed results to show. They raise a whole set of difficulties hindering their implementation or limiting their effectiveness. For instance, many questions remain unsolved regarding the ability of such calculative practices to represent appropriately the diversity of worldviews and values attached to biodiversity (Allenby, 2005; Chan et al., 2012). This often leads to the use by decision-makers of contested ecological metrics, which can undermine decisions-making processes. Furthermore, conservation scientists who invent and produce most ecological indicators have often little connections with those who actually use them in practice. As a consequence, a lot of the information produced is not in a format suited to address the variety of concrete situations that managers and decision-makers have to deal with (Turnhout et al., 2007). There is now a crucial need for new conceptual and practical approaches to address such unresolved questions that stand between innovations in calculative practices and the roles they could play in improving biodiversity management. In this paper, we defend the idea of treating these questions as being essentially questions of accounting. The crux of our argument is that mobilizing accounting research, and more notably the knowledge accumulated by three decades of interdisciplinary accounting research (Baxter and Chua, 2003; Ahrens et al., 2008; Chapman et al., 2009), could bring decisive insights to researchers and operators who attempt to design and use calculative practices to inform the collective management of ecosystems. As an example, we shall discuss a specific set of tools recently developed by conservation scientists in order to provide different sets of information (biophysical, economic, trade-offs etc.) on the consequences of landscape planning and management choices in the ability of ecosystems to provide services (such as biodiversity habitat, carbon storage, water purification) (Kareiva, 2011). These tools have so far been applied in real decision-making contexts around the world (Ruckleshaush et al., 2014; McKenzie et al., 2014). We show that the current questions that the promoters of such approaches are confronted with would benefit in several ways from being addressed as accounting issues. More specifically, we draw on interdisciplinary accounting research to insist on the importance, as one studies calculative practices used in the realm of ecosystem conservation, of focusing not only on their scientific, technical or legal dimensions, but more decisively on their social and organizational dimensions (Hopwood, 1983; Roberts and Scapens, 1985). This calls for two complementary steps. The first is to elucidate the implicit conceptions of organized action on which rest current uses of evaluative information for ecosystem management. In the paper we shall examine three of them we believe dominate the current state of the art. We refer to them as respectively the “Rational model”, the “’As is’ model” and the “Participatory model”. For each of them, we show their contribution, but more pointedly the limitations to their ability to support effective action for change. The second step aims at overcoming such limitations by mobilising other, more explicit, theoretical frameworks to enrich our questions and our answers on the production, formatting and use of information in the context of organized action for ecosystem conservation. In the paper we elaborate on three of them: Boltanski and Thévenot ‘s “On Justification: Economies of Worth” (2006), Bruno Latour’s “Politics of Nature” (2004) and our own “Strategic Environmental Management Analysis” (Mermet et al., 2005; Mermet, 2011). Each of these frameworks introduces its own, consistent set of specific questions of accountability, heterogeneous metrics and modes of valuation, the role of accounts, the place of accountants etc. in the context of ecosystem conservation. We encourage readers to dig deeper into the questions and answers that these theoretical frameworks can spark off and organise and, eventually to explore other perspectives that could shed light on the organizational dimensions of collective ecosystem management. Finally, we point out what this perspective can bring both to biodiversity conservation researchers and to accounting researchers. To the first, insights from accounting research will help guide them through the complex challenges of organized action to improve biodiversity management. We also show that adopting an accounting lens in conservation science should go much further than current “ecosystem accounting” approaches which are exclusively centred on one aspect of the whole issue, namely the problem of standardisation of biophysical measurements and economic valuation methods (Boyd and Banzhaf, 2007; Weber, 2011; Zurlini et al., 2010; Edens and Hein, 2013; ). To accounting researchers, we hope to demonstrate that ecological questions can call for other types of answers than those developed since the 1980s in Social and Environmental Accounting research (Owen, 2008; Hopwood, 2009; Brown and Dillard, 2013; Parker, 2011; Richard, 2012). We do acknowledge the rich legacy of the latter, but by taking the step to complement accounts centred on firms by other, new types of accounts centred on ecological systems, we think accounting research can explore new conceptual and methodological ground and address today’s burning ecological and organizational issues.