What is valued in conservation? A framework to compare ethical perspectives

Perspectives in conservation are based on a variety of value systems. Such differences in how people value nature and its components lead to different evaluations of the morality of conservation goals and approaches, and often underlie disagreements in the formulation and implementation of environmental management policies. Specifically, whether a conservation action (e.g. killing feral cats to reduce predation on bird species threatened with extinction) is viewed as appropriate or not can vary among people with different value systems. Here, we present a conceptual, mathematical framework intended as a tool to systematically explore and clarify core value statements in conservation approaches. Its purpose is to highlight how fundamental differences between these value systems can lead to different prioritizations of available management options and offer a common ground for discourse. The proposed equations decompose the question underlying many controversies around management decisions in conservation: what or who is valued, how, and to what extent? We compare how management decisions would likely be viewed under three different idealised value systems: ecocentric conservation, which aims to preserve biodiversity; new conservation, which considers that biodiversity can only be preserved if it benefits humans; and sentientist conservation, which aims at minimising suffering for sentient beings. We illustrate the utility of the framework by applying it to case studies involving invasive alien species, rewilding, and trophy hunting. By making value systems and their consequences in practice explicit, the framework facilitates debates on contested conservation issues, and complements philosophical discursive approaches about moral reasoning. We believe dissecting the core value statements on which conservation decisions are based will provide an additional tool to understand and address conservation conflicts.


30
Perspectives in conservation are based on a variety of value systems. Such differences in how 31 people value nature and its components lead to different evaluations of the morality of 32 conservation goals and approaches, and often underlie disagreements in the formulation and 33 implementation of environmental management policies. Specifically, whether a conservation 34 action (e.g. killing feral cats to reduce predation on bird species threatened with extinction) is 35 viewed as appropriate or not can vary among people with different value systems. Here, we 36 present a conceptual, mathematical framework intended as a tool to systematically explore and 37 clarify core value statements in conservation approaches. Its purpose is to highlight how 38 fundamental differences between these value systems can lead to different prioritizations of 39 available management options and offer a common ground for discourse. The proposed 40 equations decompose the question underlying many controversies around management decisions 41 in conservation: what or who is valued, how, and to what extent? We compare how management 42 decisions would likely be viewed under three different idealised value systems: ecocentric 43 conservation, which aims to preserve biodiversity; new conservation, which considers that 44 biodiversity can only be preserved if it benefits humans; and sentientist conservation, which aims 45 at minimising suffering for sentient beings. We illustrate the utility of the framework by applying 46 it to case studies involving invasive alien species, rewilding, and trophy hunting. By making 47 value systems and their consequences in practice explicit, the framework facilitates debates on 48 contested conservation issues, and complements philosophical discursive approaches about 49 moral reasoning. We believe dissecting the core value statements on which conservation 50 decisions are based will provide an additional tool to understand and address conservation 51 conflicts. Value systems consider more or less inclusive communities of moral patients, defined as the 72 elements with intrinsic or inherent value towards which humans, considered here as the 73 community of moral agents, are considered to have obligations (in the following, for simplicity, 74 we refer to the community of moral patients as the moral community; Table 1). Moral 75 communities can include only humans (anthropocentrism), to further incorporate sentient beings 76 (sentientism), living beings (biocentrism), and collectives (such as species and ecosystems; 77 ecocentrism) (Table 1, Figure 1). The definition of moral communities can also be influenced by 78 additional elements (such as spatial elements in the case of nativism), and, at the assessor level, 79 by personal experience. These value systems underlie different sets of explicit or implicit 80 normative postulates, i.e. value statements that make up the basis of an ethic of appropriate 81 attitudes toward other forms of life, which, in turn, can form the basis of different conservation 82 approaches (Soulé 1985; Table 1). If the normative postulates of different value systems diverge 83 (and excluding considerations that moral reasoning, experience, etc., may change one's value 84 system), conflicts can arise between different groups of stakeholders whose members share 85 common moral values (Crowley et al. 2017). In particular, conservationists who value 86 biodiversity per se [as defined initially by Soulé (1985), called hereafter 'traditional 87 conservation' (Table 1) In the following, our aim is to conceptualize and decompose value systems in an explicit, and 95 potentially (but not necessarily) quantifiable, fashion using a common mathematical framework, 96 and to explore their repercussions for the perception of conservation management actions by 97 stakeholders with different value systems. We argue that doing so allows for explicit comparison 98 between these perceptions to identify sources of potential conflicts. First, we recapitulate four 99 archetypal value systems in environmental affairs and relate them to different conservation 100 philosophies. Since identifying commonalities in the perspectives of different parties is key in 101 conflict management (Redpath et al. 2013), we then introduce a formal framework to 102 conceptualise these value systems, and examine how it can be applied to clarify different 103 perspectives. Finally, we discuss opportunities for identifying commonalities between different 104 value systems that may enable identifying widely acceptable solutions to otherwise polarising 105 issues.

VALUE SYSTEMS AND CONSERVATION PRACTICES
In practice, the separation between anthropocentrism, sentientism, biocentrism, and ecocentrism 150 is blurry, and values given to different species may vary under the same general approach. For 151 example, biocentrism can range from complete egalitarianism between organisms, i.e. 152 universalism (Table 1), to a gradual valuation resembling sentientism. These four value systems 153 can also interact with other systems that use other criteria than the intrinsic characteristics of 154 individuals to define the moral community. For example, nativism is a system that values 155 organisms indigenous to a spatial location or an ecosystem over those that have been 156 anthropogenically introduced. Nativism can therefore interact with any of the four systems 157 presented above to alter the value attributed to a species in a given context. To account for the different elements that can be combined to create the concept of value, in the 166 following, we distinguish between 'intrinsic', 'inherent', and 'utilitarian', value (our definitions; 167 Table 1). Intrinsic value is the value possessed by an individual or collective as defined by one of 168 the systems above, and is therefore independent of context. Intrinsic value is based on objective 169 criteria such as cognitive ability. The choice of a criteria may be subjective, but the value is 170 independent of the assessor once the criteria has been defined. This has been termed "objective 171 intrinsic value" by others (Sandler 2012). Inherent value is the value of an individual, species or 172 ecosystem that results from the combination of its intrinsic value and context-specific and 173 subjective factors (note that other scholars have used 'inherent' differently, e.g. (Taylor 1987, 174 Regan 2004); here it corresponds to what has also been termed "subjective intrinsic value"; 175 (Sandler 2012) Conservation practices can historically be divided into three main categories, closely related to 194 specific systems of moral valuation (Mace 2014). At one extreme, a 'nature for itself' (Table 1) 195 view mostly excludes humans from the assessment of the efficacy of conservation management 196 actions ( Figure 2). This ecocentric perspective is the foundation of traditional conservation as 197 defined by Soulé (1985), and relies on the four following normative postulates: "diversity of 198 organisms is good," "ecological complexity is good," "evolution is good," and "biotic diversity 199 has intrinsic value" (Soulé 1985). It historically underlies widely-used conservation tools, like 2). These values encompass ecosystem services that help sustain human life (Bolund and 215 Hunhammar 1999) or economic assets (Fisher et al. 2008), and can rely on the assessment of 216 species and ecosystem services in terms of their economic value (Costanza et al. 1997), which 217 can be considered as the most general form of utilitarian value, and has also been termed 218 economism (Norton 2000 Figure 2). It has been argued that such an anthropocentric perspective will, by  The 'One Health' approach, endorsed by the Food and Agriculture   244   Organization, the World Health Organization, and the World Organisation for Animal Health   245 also acknowledges the interdependence between the state of ecosystems, human health, and 246 zoonoses (Gibbs 2014). The difference between people and nature and new conservation 247 approaches therefore lies in the fact that it merges anthropocentric and ecocentric systems, rather 248 than considering that the latter will be addressed by focusing on the former (see Section "Nature 249 despite/for/and people" below for details). consequentialist perspective. We therefore consider an objective-driven type of conservation. 272 Our purpose is not to argue about the relevance of consequentialism vs. deontology, or on the 273 place of virtue ethics in conservation. Rather, we consider that, from a management perspective, 274 conservation necessarily includes objective-driven considerations. A better understanding of how 275 and why objectives can differ between stakeholders as a result of their value systems is therefore 276 useful to anticipate and manage potential conflicts. Although some participants of the discourse 277 will be more receptive to discursive than mathematical conceptualisation, we argue that defining 278 concepts as mathematical terms can make differences in value systems and their normative 279 postulates more explicit and transparent, which will be beneficial when used with appropriate 280 stakeholders, even when these terms would be hard to quantify in real life. A mathematical 281 formulation can be seen as a logic way to express relationships between different elements. proposed a mathematical framework for assessing the environmental impacts of alien species. 286 This work was highly influential in the conceptualisation of biological invasions (being cited 287 over 2,000 times until April 2021 according to Google Scholar), rather than by its direct 288 quantitative application. We also acknowledge that this approach has specific limitations, which 289 are discussed below. Our mathematical formalisation conceptualises the consequences of environmental management 292 actions. As we develop below, these consequences will be defined differently depending on the 293 value system, but can be understood generally as the consequences for the members of the moral 294 community. Under anthropocentrism, these will be consequences for humans; under sentientism, 295 these will be consequences for sentient individuals; under biocentrism and ecocentrism, these 296 will be consequences for biodiversity. We argue that our mathematical formalisation can account 297 for these different value systems (see Appendix S1 for an extension to ecocentrism beyond 298 species and considering wider collectives, i.e. ecosystems), while also accounting for cultural 299 and personal contexts. These consequences C can be conceptualised as a combination of the 300 impact of an action on the different species or individuals involved and the value given to said 301 species and individuals under different value systems as follows: is a function (e.g. mean, maximum, etc.) of the impact (direct and indirect) resulting 306 from the management action on all individuals of species s, V s is the inherent value attributed to 307 an individual of species s (as described above), N s is the abundance of species s, and a 308 determines the importance given to a species based on its abundance or rarity (and enables to 309 account for the importance of a species rather than an individual, see below). and biocentrism, whose characteristics (sentience and life) are defined at the individual level. As 326 a result, impacts on larger populations would weigh more on the consequences. As a decreases 327 towards 0, the correlation between the value of a species and its abundance decreases. For a = 0, 328 the consequence of a management action becomes abundance-independent. For a < 0, rare 329 species would be valued higher than common species (or the same impact would be considered 330 to be higher for rare species), for example due to the higher risk of extinction. And for a > 1, 331 disproportionate weight is given to abundant species, which are often important for providing 332 ecosystem services (Gaston 2010    Over the past decade there has been some debate between proponents of traditional conservation, 457 and those of new conservation (Table 1) Traditional conservation is based on an ecocentric value system and seeks to maximize diversity 465 of organisms, ecological complexity, and to enable evolution (Soulé 1985). For simplification, 466 we will consider an extreme perspective of traditional conservation, championed by 'fortress 467 conservation' (Siurua 2006, Büscher 2016, i.e. excluding humans from the moral community. 468 To capture these aspects, consequences C in Equation 1 can be more specifically expressed as 469 follows: Assigning a stronger weight to rare species (a < 0) accounts for the fact that rare species are New conservation considers that many stakeholders ("resource users", Kareiva, 2014) tend to 487 have an anthropocentric value system, and that conservation approaches that do not incorporate 488 such a perspective will likely not succeed at maximizing diversity of organisms ( biodiversity can indeed support the provision of ecosystem services to humans. Such an 522 approach will necessarily distinguish between "useful" species and others, and impacts will be  (Table 3). However, since compassionate conservation is not based on 577 consequentialism, it uses different criteria to assess the appropriateness of conservations actions 578 (but see (Wallach et al. 2020) for responses to some criticisms). Our purpose here is not to 579 discuss the relevance or irrelevance of virtue ethics for conservation (see (Griffin et  Although V(E s ) should be measured in an objective fashion, many factors may influence the 610 relationship between the inherent value and the emotional capacity of a species. For example, 611 high empathy (Table 1) from the observer will tend to make the distribution uniform, whereas 612 anthropomorphism and parochialism (Table 1)  It has been argued that sentientism and ecocentrism are not fully incompatible (Varner 2011). 646 The relationship between biodiversity and animal suffering can be formalised more clearly using 647 the traditional conservation and the sentientist Equations 2 and 4, to explore if the same 648 management action can minimize the consequences evaluated using the two equations (see also 649 Appendix S2 for the application of the framework to theoretical cases). The main difference with   The issue of spatial and temporal scale also warrants consideration (Table 3). In the case of a 721 species that may be detrimental to others in a given location but in decline globally, the spatial 722 scale and the population considered for evaluating the terms of Equations 1 to 4 is crucial to 723 determine appropriate management actions. Similarly, management actions may also result in a 724 temporary decrease in welfare conditions for animals, which may increase later on (Ohl and Van 725 der Staay 2012), or the impacts may be manifested with a temporal lag. In that case, determining 726 the appropriate time period over which to evaluate the terms of Equations 1 to 4 will not be 727 straightforward. Impacts will be of different importance depending on whether they occur in the 728 short-or long-term, especially since long-term impacts are harder to predict and involve higher 729 uncertainty. Discount rates (Table 3)  In the following, we present three case studies where conservation actions have either failed, had 761 adverse effects, or were controversial, and we explore how our framework can help to identify 762 normative postulates underlying these situations. Although these case studies have been 763 discussed at length in the articles and reports we cite, we argue that our framework helps capture 764 the different components of the controversies in a more straightforward and objective fashion 765 than using a discursive approach that might require either emotionally loaded language or more 766 neutral but less understood neologisms.
ሻ are higher than without management, 811 due to the increase in V(E gs ) and I(S gs ). The application of our framework therefore allows to 812 clarify a discourse whose perception could otherwise be altered because of techniques such as 813 appeal to emotion.
. In that case, assuming for simplification 824 the same suffering through euthanasia for grey squirrels as red squirrels suffer from the grey 825 squirrels, and the same value to individuals of each species (i.e. avoiding nativism), the mere in abundance would lead to a higher value of C without management. This  (Table 3). Culling animals might be acceptable in some cases, but 872 might not be if these individuals were purposefully introduced, which may lead to considering a 873 sentientist perspective. hidden commonalities between seemingly antagonistic stances. We hope that this framework can 957 foster fruitful debates and thus facilitate the resolution of contested conservation issues, and will 958 ultimately contribute to a broader appreciation of different viewpoints. In an increasingly 959 complex world shaped by human activities, this is becoming ever more important.  Table1. Glossary of terms as they are used for the purposes of this paper.

Term Definition
Anthropocentrism (strong) Value system that considers humans to be the sole, or primary, holder of moral standing, and therefore the concern of direct moral obligations.
Non-human species are considered only to the extent that they affect the statisfation of felt preference of human individuals (Norton 1984, Rolston 2003, Palmer et al. 2014).

Anthropocentrism (weak)
Value theory in which all values are "explained by reference to satisfaction of some felt preference of a human individual or by reference to its bearing upon the ideals which exist as elements in a world view essential to determinations of considered preferences" (Norton 1984). That is, the value of an individual or species is not only exploitative, but incorporates human experience and the non-utilitarian relationship between humans and nature.
Anthropomorphism "The attribution of human personality or characteristics to something non-human, like an animal, object, etc." (Oxford English Dictionary n.d.).

Biocentrism
Value system considering all living beings as the concern of direct moral obligations (Rolston 2003, Palmer et al. 2014).

Collectivism
Value system in which a group or collective has a higher value than the individuals that compose it (Wallach et al. 2018 Consequentialism "An ethical doctrine which holds that the morality of an action is to be judged solely by its consequences" (Oxford English Dictionary n.d.).
Convergence hypothesis "If the interests of the human species interpenetrate those of the living Earth, then it follows that anthropocentric and non-anthropocentric policies will converge in the indefinite future" (Norton 1986).

Deontology
A normative ethical theory considering that "choices are morally required, forbidden, or permitted" (Alexander and Moore 2016).

Ecocentrism
Value system considering that species, their assemblages and their functions, as well as more broadly ecosystems, rather than individuals, are the concern of direct moral obligations (Rolston 2003, Palmer et al. 2014).
Empathy "The quality or power of projecting one's personality into or mentally identifying oneself with an object of contemplation, and so fully understanding or appreciating it." (Oxford English Dictionary n.d.).
Empathy will influence the inherent value given to individuals from other species. Relational values are not present in things but derivative of relationships and responsibilities to them." (Chan et al. 2016).

Sentience
The ability to experience phenomenal consciousness, i.e. the qualitative, subjective, experiential, or phenomenological aspects of conscious experience, rather than just the experience of pain and pleasure (Allen and Trestman 2017).

Sentientism
Value system considering sentient beings as the concern of direct moral obligations (Rolston 2003, Palmer et al. 2014).

Speciesism
Value system in which some species are considered to have a higher value than others, for various possible reasons (Singer 2009). Speciesism is often used to refer to the superiority of humans, which is a specific expression of speciesism as considered in this paper.

Suffering
Negative emotion, sometimes called emotional distress, experienced by sentient beings, and which can result from different causes, including but not limited to physical pain (Dawkins 2008, Farah 2008).
Traditional conservation Discipline aiming at preserving biological diversity through the management of nature, and based on four value-driven normative postulates: "diversity of organisms is good," "ecological complexity is good," "evolution is good," and "biotic diversity has intrinsic value" (Soulé 1985). Traditional conservation is rooted in ecocentrism.
Utilitarian value Value given to an individual or collective by humans, based on its utility.
Virtue ethics Ethical doctrine that emphasizes the virtues, or moral character as the reason for action (Hursthouse and Pettigrove 2018). -If all species must weigh the same, a = 0.
-If rare species should be given more importance, a < 0.

Factor
Influence on variables and outputs in Equations 1 to 5

Biotic interactions
The impact or suffering of individuals from one species can be caused by individuals from another species, either through direct or indirect interactions. Management actions can therefore also have non-trivial indirect impacts on some species.
Capacity to provide ecosystem services The presence of a specific species may increase the fitness/welfare of other species through the ecosystem services it provides. Since these effects can be difficult to quantify explicitly, the value of such species may be increased in Equations 1 to 4 to account for them.
Discounting rate Rate at which impacts that occur in the future lose importance.

Impact quantification and commensurability
How the impacts of management actions are quantified is also dependent on value systems, as some impacts (such as death) may be considered incommensurable to others (such as suffering).

Responsibility from previous actions
Previous human actions on certain species, such as reintroduction of domesticated species or the introduction of alien species can change the perception of the public and therefore change the inherent value attributed to these species, or change the morality of an action, in addition to obviously having an impact on these species.

Spatial scale
The spatial scale will change the abundance N and the number of species considered. As a result, a management action that is more beneficial than another at small scale may not be such at a larger scale, and reciprocally. Additionally, the spatial scale can change the inherent value of species, for example under nativism, or because of the range of cultures that are considered. This uncertainty corresponds to a moral dilemma due to a conflict between the desire to have a 1339 small positive impact for the species with the larger value and abundance, and the desire to avoid 1340 a very negative impact for the species with the lower value and abundance for aii. For aiii, the 1341 dilemma is due to a conflict between the desire to avoid a small negative impact for the species 1342 with the higher value and abundance, and the desire to have a very positive impact for the 1343 species with the lower value and abundance. b) The species with higher value V high has the lower 1344 abundance N low . If impacts are different between the two species, the opposition between V and 1345 N will most likely generate moral dilemmas (C ? ). If V high × N low > V low × N high , bii is equivalent 1346 to aii, and to aiii otherwise (and biii is equivalent to aiii, and to aii otherwise), but because value 1347 and abundance have different units, it is difficult to determine for which value and abundance 1348 V high × N low = V low × N high . Therefore, an additional moral dilemma arises due to a conflict 1349 between the desire to avoid a negative impact for the larger population and the desire to avoid a 1350 negative impact for the species with the higher value.