As the title suggests, this post contains an online copy of my MA thesis which I submitted in mid-2018 (see here for a citable PDF). My views on this issue have updated significantly since pursuing my PhD on welfare biology. As soon as I have submitted my doctoral thesis I will link it here!
“Are God and Nature then at strife,
That Nature lends such evil dreams?
So careful of the type she seems,
So careless of the single life;That I, considering everywhere
Alfred, Lord Tennyson, 1850.
Her secret meaning in her deeds,
And finding that of fifty seeds
She often brings but one to bear”
Canto 55, ‘In Memorium’
Abstract
This thesis concerns the moral significance of wild animal welfare and discusses some practical methods of intervention on their behalf.
In chapter two of this thesis I defend the functionalist method of mental state inference and examine relevant empirical research to establish the affective welfare capacities of small animals which are not often considered sentient. Chapter three investigates the circumstances in nature which cause wild animals to suffer, which are the limited availability of natural resources and evolutionary adaptions that have arisen in response. In chapter four I examine relevant axiological and normative theories that consider animals who suffer in the wild, defending my view of affective sentience as the only morally defensible criterion for consideration. Chapter five includes several welfare considerations to be accounted for under circumstances of practical wild animal intervention. And chapter six concludes by investigating the welfare impact of certain current and speculated future methods of wildlife intervention, including suggestions for future research.
Acknowledgements
Thank you to Marc Davidson, for supervising this thesis, and Oscar Horta, for providing extensive feedback throughout its progress.
1 Introduction
1.1 Background
Historically, non-human animal (hereafter, ‘animal’) advocacy has taken the scope of human exploitation of animals as the be-all and end-all for their normative consideration, ignoring the welfare issues that result from animals living in nature. This position incorrectly assumes exactly what the average wild animal life entails, and that even though they may sometimes suffer under naturally occurring circumstances (of which include predation, starvation, disease, among others), wild animals generally have quite happy and good lives. They are lives that are worth living, and are thus excluded from our immediate moral consideration. In addition to these beliefs, it is often held that intervening in nature is something that is intrinsically wrong, or consequentially undesirable. Even though animals may suffer enormously in the wild it is not our place to aid them, and we ought to leave nature alone in its idyllic state.
In recent years, however, these assumptions have been challenged by animal advocates such as Oscar Horta, Brian Tomasik, and others, on the grounds of speciesism. Despite this, little research has yet been made to practically consider the welfare of wild animals within significant pre-existing wildlife interventionist strategies (such as those which already occur to benefit humans and the environment). There has furthermore been little research to determine how intervention might take place to directly benefit wild animal welfare, taking into account the practical hurdles that are associated with large-scale intervention. Both of these are arguably required for the proposal of a wild animal welfare-considering applied ethic, to give practical guidance on future interventionist measures.
1.2 Thesis Introduction
In this thesis, I shall make the case for practically considering wild animal welfare according to the negative conditions they experience in nature. My central argument shall consist of the following five premises, which I shall address in the second, third, and fourth chapters of this thesis (respectively);
P1: ‘A wide variety of animals are capable of experiencing significant negative welfare over the course of their lifetimes’
P2: ‘Natural conditions (ie; resource scarcity, reproductive strategy) tend to create excess suffering and little happiness for wild animals’
C1: ‘The experience of negative welfare can be expected to outweigh the experience of positive welfare for the lives of most wild animals’
P3: ‘We have good reasons to oppose the suffering of wild animals’
C2: ‘We ought to intervene in natural processes to improve wild animal welfare’
Given the applied consideration of wild animals related to my argument, this thesis heavily involves the use of factual analysis and descriptive argumentation to justify its normative conclusions.
In the latter half of this thesis containing chapters five and six, I shall propose a set of act-based welfare-promoting considerations to be used in cases of applied ethical judgements relating to wild animals. These include our epistemic uncertainty of other animals being sentient and the qualitative extent of their experienced welfare, welfare impacts resulting from some of the biologically determined habits of wild animals (such as those underlying predator/prey relationships), and variable public attitudes concerning welfare-directed wildlife intervention (these, I argue, tangibly affect our applied obligations toward wild animals). In line with these welfare considerations, I shall examine current and speculated future used methods of wildlife intervention with significant welfare implications, of which include baiting, poisoning, and biological control of wild animals. Chapter seven shall conclude with a short review of the arguments expressed per-chapter in this thesis and a light commentary on its findings.
2 On Suffering
The aim of this chapter is primarily to establish the first premise (P1) of my argument, which states that ‘a wide variety of animals are capable of experiencing significant negative welfare over the course of their lifetimes’. To do this, I shall first in §2.1 introduce the notion of suffering which I shall refer to for the duration of this thesis. §2.2 shall thereafter propose a system for assessing mental states in animals such that their experience of suffering may plausibly be inferred. Then, in §2.3, I shall investigate relevant empirical research on animal cognition to show that at the very least we have good reasons for thinking that considerably more animals possess affective welfare states than we would initially posit.
2.1 Defining Pain and Suffering
To begin, it is useful to clarify my argument’s use of the term ‘suffering’, as different writers can be seen to use the word in remarkably different contexts. Classical evidence based medicine tends to describe suffering in the third-person perspective, invoking dualism about the mind and the body, whereas first-person phenomenological accounts advocate for ‘emboddied’ approaches in which there is a crucial continuity between the two (Bueno-Gómez, 2017, p.3 & 6). As my argument is primarily sentiocentric, I shall consider only the phenomenal aspects of suffering – relevant physical descriptions only matter insofar as they have an impact on sentient minds.
Similarly, I shall define the term ‘sentience’ in terms of having affective welfare states. Yew-Kwang Ng (1995, p.258) points out that the commonplace definition of sentience as consisting of mere phenomenal perception includes beings that are both “capable of the feeling of being happy and/or unhappy”, and “beings just capable of perceiving things without having any affective feeling”. The moral difference between these two is tantamount to the moral difference between a mouse and a stone; the former has welfare-affecting interests for which we can normatively consider (namely; the interest not to suffer, and to experience positive welfare states) while the latter does not (Singer, 2015, p.37). Simply put, “[i]f a being is not capable of suffering, or of experiencing enjoyment or happiness, there is nothing to be taken into account” (Singer, 1974, p.108).
It may be argued in response to this claim that such non-affectively sentient entities as motor engines also have morally considerable interests; namely, to be regularly oiled – an absurd conclusion, thereby demonstrating the argument’s invalidity (Sagoff, 1984, p.298). But as Singer responds, we ought rather consider these to be mere functions of various entities. Interests come from subjective entities who are able to will something for their own sake, so the possible function of a motor engine to be well oiled can be understood as no more than the projected interest of, say, a car mechanic (see §4.1.2 for a more in-depth defence of this position).
From this, it is clear what our basic description of suffering will be; to suffer is to experience, first-person, states of negative welfare. But suffering plausibly comes in many forms – most immediately, the state of physical pain. Droege and Braithwaite (2014, p.80) argue for three elements present in human pain experience, which include the nociceptive registration of noxious stimuli and subsequent reflex, the “conscious hurtfulness” which usually results from this interaction, and the “self-conscious understanding of the pain’s significance”. Other aspects of suffering, however, may have little to do with pain experience at all, of which include “[s]ocial problems like poverty, social exclusion, forceful social inclusion, […] existential and personal problems like grief and stress; conditions like nausea, paresthesia [sic], a non-painful illness, anxiety or fear” (Bueno-Gómez, 2017, p.2).
Pain can therefore be seen as a form of suffering, given that is not necessary for all such instances of its occurrence. However, it is plausibly the most frequently occurring state of suffering giving its inferred evolutionary continuity throughout the animal kingdom (as opposed to the other more complex suffering states that humans often experience). And the capacities for other animals to experience complex states of suffering are plausibly different if not reduced in comparison to our own (we make this assumption based on relevant differences in our respective neuroanatomical configurations, which are likely responsible for if not closely associated with one’s mental capacities). Pain thus appears to be the most empirically reducible suffering state for observation, so I shall treat it as the primary criterion for which we infer these states in animals. This, also, to avoid unnecessarily anthropomorphising animals in terms of human states and capacities (Buckner, 2013, p.855). But pain states are too (negative) emotional states, which are themselves conscious states of the mind (given that phenomenal experiences are conscious experiences). So “[t]he question regarding suffering in animals, therefore, primarily addresses the issue of whether animals have a consciousness” (Bermond, 2003, p.79).
2.2 On the Nature of Mental States
Inferring mental states in nonlinguistic animals is a notoriously inconclusive task due to the absence of any agreed upon behavioural or neurological markers of consciousness (Shea & Heyes, 2010, p.95) (Clayton, 2017, pp.227-228). In the past, theory of mind research in animals has attempted to identify their mental states and capacities in terms of the phenomena that is usually associated with the presence of such states and capacities in humans. Analytical behaviourism is a well-known flawed example of this, as it claims the identification of mental states as behaviour dispositions. The immediate difficulties of holding this view, however, make it an implausible theory of mind that few still advocate (for a more detailed account of this, see Grush, 2002, pp.272-273).
An alternative theory of mind to consider – psychophysical type-type identity theory (‘brain state theory’ for short) – claims to identify mental states as specific neurological configurations, by which mental state identifications are made by correlating “first‐person reported mental states [with] the brain states observed by third‐person empirical means” (Van-Gullick, 2009, pp.2-3). Accordingly, we might disqualify animals from experiencing pain if they lack integral components of the human nervous system that are associated with and responsible for pain perception (C-fibre firing, for example). This position does not account for the multiple realizability of mental states (by which more than one neurological state may be responsible for the same mental state), which is damning insofar as it does not support theory of mind in most nonhumans with a sufficiently dissimilar neuroanatomy.
A third position that has been proposed as a solution to this problem is functional state theory which proposes that we identify mental states as functional states of the mind. This is, as first argued by Hilary Putnam (1975, p.431), because the link between mental states and physical states is not a matter of type-identity from which we can claim empirical reduction. To make the claim ‘pain states are brain states’ (as the brain state theorist might) assumes two premises; that both properties are “(a) associated with a spatio-temporal region; and (b) the region is one and the same in both cases” (ibid). Accordingly, the phrase ‘temperature is mean molecular kinetic energy’ identifies type with type, whereas it is unclear if and how a pain in one’s foot can be identified completely as the accompanying brain state (they may be correlated, but this does not imply that they are invariably so) (ibid, p.432).
2.2.1 The ‘Myth’ of Animal Suffering
Despite the prevalence of functionalism in mainstream science, there remain some who continue to advocate for the type-identification of mental states, and fewer still who take it to such extremes as to deny conscious states in large vertebrate animals. Bob Bermond (2003, pp.83-84) is one such commenter, who argues that according to a perceived lack of emotional capacities in animals, “claims for suffering in animal species, other than anthropoid apes and possibly dolphins, are incorrectly substantiated”. As a brain state theorist, he rejects two common methods of reasoning suffering states in animals; (1), that an animal’s behaviour accurately represents their internal states (to the degree that mental state extrapolations are possible), and (2), that animal suffering can be inferred with respect to “similarit[ies] between the structure of the sensory nervous system of […] animal species with the human sensory nervous system” (ibid, p.82). Another sceptic of animal suffering, Peter Harrison (1991, p.25), additionally rejects (3), that pain is a useful evolutionary adaption with “no radical discontinuity between human[s] and other species”. In the following, I shall respond to each of these three positions.
To address (1), Bermond critiques several research strategies which are commonly used to infer the presence of pain states in animals. The first two of these incorrectly assume “that the experience of emotion may be assumed to be present if one or more of the other components of the emotional process can be shown to be present”, such as the accompanying behavioural and physiological responses to noxious stimuli (Bermond, 2003, p.81). But the relationship between behaviour/physiology and mental experience is not a necessary one. Pain experience may occur in absence of its commonly accompanying physiological elements; an animal may sustain considerable tissue damage with little to no pain behaviour, and likewise, “pain behaviour without any sensory experience has been described in spinal cord lesioned patients” (ibid) (Harrison, 1991, p.23). A third research strategy similarly posits the demonstrability of operant conditioning in animals as evidence that they have affective states of welfare. To this, Bermond cites evidence that for operant conditioning in humans to succeed, “conscious awareness of neither the conditioning process itself nor of the changes which occur as a result of the conditioning are necessary” (Bermond, 2003, p.81). Therefore, he claims that it is possible if not likely that animal responses to reward-based tasks are unconscious and lacking the affective feeling that we prescribe to them.
The foremost problem of these responses is that they critique the inference to mental states as behavioural dispositions, which is not necessarily valid as a critique of the functionalist’s method of inference. In contrast to behavioural and brain state theories of mind, functionalism seeks to identify mental states by such observed phenomena that is claimed to be associated with their occurrence, since “similarities in the behaviour [and physical details] of two systems are at least a reason to suspect similarities in the[ir] functional organisation” (Putnam, 1975, p.437). So, in addition to supporting the multiple realizability of mental states, this approach takes into account any relevant evidence that counts toward theory of mind hypotheses. As such, the fact that pain states sometimes are disassociated from their usually accompanying behavioural and physiological states “does not force us to concentrate [only] on the nervous system rather than on an animal’s behaviour” (Rivas, 2015). In addition, Bermond’s charge against operant conditioning loses strength when one considers that a large portion of human mental processes are unconscious, yet on this basis we would not conclude that humans lack affective states. It is thus unclear how, as Bermond claims, “[t]hese results demonstrate that the possibility of conditioning may never serve as proof for consciousness or emotional experiences” in animals (Bermond, 2003, p.81). Rather, this seems only to show that animals, like humans, may engage in unconscious cognitive processes as well as conscious ones (Ng, 1995, p.262).
On the second position (2), Bermond and Harrison depart from one another in their approach. Bermond, a brain state theorist, puts forward that “[t]he right neocortex and the prefrontal neocortex are both prerequisites for emotional experience”, thereby disqualifying the inference to animal suffering made on neuroanatomical grounds (Bermond, 2003, p.83). Yet his argument, “founded on an outdated concept of the neural localization of mental functioning”, relies primarily on the supposed inadequacy of behaviour as evidence in theory of mind inference, from which he claims follows a behaviourist account of mental states (Rivas, 2015). But as we have shown, behaviour as evidence for mental states does not necessarily imply the sole use of behaviour for identification of such states. Such theories as functionalism successfully navigate the problems raised by Bermond and do not require the denial of affective states in nonhumans.
In contrast, Harrison’s argument claims that “there is no necessary logical [nor contingent] relation between propositions about mental states and propositions about physical states” (Harrison, 1991, pp.27-28). The former of these claims can be easily refuted as it fails to differentiate between humans and animals in its scope – ought we, too, claim that human victims of torture cannot feel pain? (House, 1991, pp.376-377). For the latter, Harrison claims that “despite such well-established connections between observable brain structures and more elusive mental states, it would be rash to attempt to predict the mental states of individuals on the basis of the presence or absence of certain structures” (Harrison, 1991, p.28). Rather, Harrison himself appears to base his argument on the functional identification of mental states, citing evidence of individuals who possess such working states and capacities despite their lack of the neurological structures that are assumed to be necessarily responsible for their presence. For instance, some humans though lacking a visual cortex have been shown to possess “above average visual perception”, and likewise so in the case of birds whose brains do not include a visual cortex yet surely possess some structures that are (by functional equivalence) capable of processing visual stimuli (ibid, pp.29-30).
Despite this method of mental state identification, Harrison’s argument remains at odds with the animal suffering hypothesis since “the psychology of pain affords instances in which […] certain mental states [may be produced] in the absence of the relevant physical structures” (ibid, p.29). Simply, the presence of neurological structures cannot be used to infer mental states such that any other explanation may be ruled out. But perhaps our strongest inference to animal suffering is based on evolutionary continuity – ought we not expect that pain capacities arise in other species given its obvious fitness benefit? Rather, Harrison argues, the fitness benefit arises not in the negative mental experience of pain, “but [in] the behaviour which is elicited when the damaging stimulus is applied” (ibid, p.32).
In a reply by Ian House, it is pointed out that if pain behaviour is the sole aspect present in an animal’s response to noxious stimuli, then Harrison must explain how operant conditioning takes place without the apparent incentive of being able to experience affective welfare (House, 1991, p.377). Harrison’s argument therefore seems to misunderstand the function of pain experience, which he later claims as being exclusive to beings with a “rational, decision-making mind”, for which only humans qualify (as per his cartesian reasoning) (Harrison, 1991, p.38). This claim is at odds with the evolutionary hypothesis that “[p]ain may be expected in organisms whose survival can be augmented by the experience”, which requires empirical investigation to substantiate or refute according for different animals (Lockwood, 1987, p.76). Harrison’s argument furthermore conflates pain as a psychological capacity with pain as a physiological response mechanism, which are two functionally separate responses to noxious stimuli.
2.3 Inferring Pain States in Animals
Physiologically, the bodily reflex that is elicited from one’s contact with potentially harmful stimuli precedes the mental experience of pain. This is evident when we consider their respective functions; the nociceptive response is unconscious and results in one’s immediate withdrawal from harm, while the negative mental experience of pain creates a conditioning incentive for future avoidance of such stimuli (Elwood, 2011, pp.175-176). This separation may lead to some semantic misunderstandings of ‘pain’ and ‘pain perception’, as seems likely in Harrison’s assessment of pain behaviour in animals. If we are thus to use behaviour as evidence for mental rather than physiological states of pain, then we ought to consider such evidence with respect to its functional role as a mental experience – avoidance learning.
2.3.1 Evidence of Vertebrate Suffering
While it is clear that most large mammals possess capacities for avoidance learning (House cites the use of negative reinforcement training in dogs as an example of this – see 2011, p.377), the strength of our inference is reduced as we begin to consider the lives of smaller and less-complex animals. In small mammals such as rats, repeated exposure to predators has been shown to induce memory impairment, learning disability, and other such anxiety-related behaviours reminiscent of post-traumatic stress disorder in humans – for up to 115 days post-stress (Zoladz, 2008, pp.142-143). A strong argument may also be made using the evolutionary proximity of other mammals to humans as evidence for their suffering states. As mammals ourselves, we share closer similarities in our behaviour and internal physiology to other mammals in comparison to other more distant animal classes. Using comparative reasoning, we might therefore claim a higher probability on our being able to infer the presence of pain states in such animals with evolutionary proximity to ourselves (this point becomes important in §5.1.2). This form of evidence, however, becomes progressively weaker the further one strays from the evolutionary path of linguistically communicable animals (for whom we can be near certain possess affective sentience; humans, for our purposes).
For other such vertebrates as fish there is “strong evidence of the existence of homologies in neurobehavioral functions” to mammals in their learning response to adverse stimuli, which is plausibly caused by their experiencing the negative mental experience of pain (Hurtado-Parrado, 2010, p.675). Fish possess capacities to flexibly alter their behaviour in response to noxious stimuli as they have been shown to overcome their learned avoidance behaviour to obtain food while starving (despite their lacking a neocortex, the brain structure thought responsible for executive function in mammals), and “[d]opaminergic systems essential to reward conditioning […] been identified in the fish forebrain” (Sneddon, 2015, p.970., Droege & Braithwaite, 2014, p.91 & 93). Fish also have been shown to possess the same kinds of nociceptors and nerve fibres as those responsible for immediate and lasting pain in humans and other mammals, and furthermore, their behavioural response to anaesthetic is consistent with the hypothesis that their mental experience of pain becomes subdued (which is conditional on being able to experience negative emotional states to begin with).
2.3.2 Evidence of Invertebrate Suffering
While it is not often contended that many vertebrate species are able to experience pain states, claims made on invertebrate suffering are frequently met with scepticism resulting from their reduced size and internal complexity to humans, and also from their radical difference in evolutionary development. From this it is difficult to say whether they even possess a subjective character of experience to begin with, much less whether they have affective welfare states and what those states are. Thomas Nagel (1974) describes this difficulty in relation to the subjective facts of what it is like to be a bat. Such facts are by nature of our distinctly human cognitive capacities (resulting from our difference in internal structure and physiology) inaccessible; the resources of my mind limit me to imagining what would it be like for a human to behave as a bat but do not allow my knowing what it is actually like for a bat to be a bat (ibid, p.439). It is therefore sensical that people tend to assume small insects (like ants) to be no more than rule according living reading machines (to use the Wittgensteinian notion), since we are confined to the limitations of our own imagining capacities, and the lives of ants seem utterly alien and robotic to how we live as humans.
But some invertebrates are nonetheless given legislative consideration for their welfare. In article one clause three of the European Union directive on the protection of animals used for scientific purposes (2010, p.34 & 39), cephalopods are included alongside non-human vertebrates “as there is scientific evidence of their ability to experience pain, suffering, distress and lasting harm”. If it is true that such invertebrates as cephalopods – of which include octopodes, squid, cuttlefish (among others) – experience such suffering states as ‘pain, distress and lasting harm’, then we are presented with two scenarios; (1), that the capacities which support these states evolved independently at a ‘turning point’ in cephalopod evolution, or (2), that there is a link in the genetic heritage of vertebrates and invertebrates such that suffering states may be reasonably expected both cases. But in response to (1), we ought not consider differences between animal species to be definitive but rather progressive. The term ‘species’ itself is an arbitrary term used only to differentiate between groups of sufficiently genetically similar organisms, and the relative differences that are identifiable between these groups are a matter of token rather than type (as established per our functionalist reasoning). This is not to say mental states can always be inferred on the basis of their occurrence in genetically related animals, only that their presence may be a matter of degree rather than of kind.
If we are to employ this form of reasoning, then the first group of invertebrate animals we should consider being suffering-capable are other molluscs that are closely related to cephalopods – a phylum that includes such spanning classes as Bivalves (mussels, clams) and Gastropods (limpets, snails) of which there are an estimated 20,000 and 100,000 extant species (respectively) (Haszprunar & Wanninger, 2012, p.R511). Sneddon (2015, p.971) and Elwood (2011, p.177 & 182) report the presence and behavioural accompaniment of nociceptors in these three classes of mollusca, and so too in crustaceans, another aquatic invertebrate class belonging to the phylum arthropoda. And while cephalopods possess dopaminergic systems of hormone dispersal, crustaceans have their own stress hormone – crustacean hypoglycemic hormone – which observably plays the same functional role in response to bodily harm as cortisol does for many vertebrates (ibid, pp.178-179). Avoidance learning behaviour has also been observed in such crustacean species as crab, crayfish, and hermit crab, showing capacities that are consistent with the explanatory hypothesis that negative mental states like pain motivate behavioural change.
Such avoidance learning capacities have also been observed in non-aquatic arthropods like fruit flies, whose avoidance learning capacities allow for their discrimination between specific odours depending on their prior experiences of odour association with painful electric shocks (ibid, p.179). The application of pain-relieving analgesics to fruit flies have also been shown to inhibit their response to noxious stimuli – findings which have been replicated in other such arthropods as cockroaches – further supporting the hypothesis of pain experience as a method of disincentivisation (as opposed to some other non-conscious avoidance mechanism) (Knutsson, 2015, p.16). Even further, earthworms belonging to the phylum annelida “possess B-endorphins and enkephlins which, by functional analogy, suggest the capacity for pain” (Lockwood, 1987, p.76).
But further research is required into the behavioural flexibility of invertebrate animals in response to harmful situations. Notably, Jonathan Birch (2017, p.7) proposes three particularly credible indicators of animal pain – self-delivery of analgesics, motivational trade-offs, and conditioned place avoidance – that to my own knowledge have not been shown in invertebrate species other than cephalopods. However, detailed empirical research investigating the possible affective sentience of non-cephalopodic invertebrates is in short supply, so we ought not treat the lack of evidence as evidence of the absence of such flexible capacities (Clayton, 2017, p.231) (Chan, 2011, p.327). Thus, although it can be argued that none of what we have considered so far definitively proves the presence of suffering states in invertebrates such that any other explanation may be ruled out, we can neither claim the inverse, and most of the relevant supporting evidence that has been considered is consistent with the hypothesis that invertebrates do experience suffering states. So, it may “be more prudent to assume differences in degree rather than in kind as the default position, and seek evidence that this is not the case” (Corballis, 2013, p.152).
2.4 Conclusion
In this chapter, we first described the functional role of pain and suffering for the purposes of our thesis argument, and then proposed a method of mental state identification as per that role. Functionalism is furthermore consistent with how we infer mental states in nonlinguistic beings as it takes into account any relevant evidence that counts towards theory of mind hypotheses (and not only specific kinds of evidence, as in the case of competing type-theories of mental state identification). On this basis, we investigated relevant empirical research on the affective capacities of nonhumans to justify the premise (P1); ‘a wide variety of animals are capable of experiencing significant negative welfare over the course of their lifetimes’.
3 Wild Animal Suffering
The aim of this chapter is to build two more additional premises upon (P1), which are (P2); ‘natural conditions (ie; resource scarcity, reproductive strategy) tend to create excess suffering and little happiness for wild animals’, and (C1); ‘the experience of negative welfare can be expected to outweigh the experience of positive welfare for the lives of most wild animals’. To do this, I shall first in §3.1 establish the importance of wild animal welfare, and then in §3.2 I shall present and discuss some of the reasons we have for believing that most natural lives are negative (in terms of their experienced welfare).
3.1 The Importance of Wild Animal Suffering
From our findings in the previous chapter, we can make the reasonable assumption that a great many animal species – both vertebrate and invertebrate – experience affective welfare states, and therefore possess tangible interests which can be taken into moral account. For now, let us at least consider their welfare as a precautionary measure (I shall expand upon the relevance of sentio-uncertainty in §5.1.2). For such animals as vertebrates, the commonly expected result of this observation (and the one that is adopted by most animal rights activists) is that we re-examine our domestic relationship with the animals already used by humans for food, entertainment, and experimentation, among other such areas of human-animal interaction (Tomasik, 2015, p.134). Little consideration is given to the lives of invertebrates such as insects, for most of the reasons that were discussed in the previous chapter to do with undervaluing their sentience, and because humans rely more on the resources that are provided by large vertebrates (although this view may change upon realisation of the number of invertebrate organisms that are required to produce one unit of a derived product relative to their mammalian counterparts; approximately 5000 silkworms are killed per kilogram of silk produced, compared to less than one sheep per kilogram of wool – see Tomasik, 2016). Far more pressing, however, is the extent to which animals suffer in the wild.
Before we discuss the specific ways in which wild animals suffer, let us consider the ways in which nature and the lives of wild animals are idealised by humans. The traditionally adopted view of animal ethics is that moral injustice begins and ends with the subjection of nonhuman (usually vertebrate) animals. Although it may be noticed from this perspective the presence of extreme suffering in the lives of wild animals (see the recent controversy surrounding a Dutch rewilding project located in the east of Amsterdam where thousands of animals starved to death – Barkham, 2018), it is frequently assumed that a natural life is a good one, which is to say that wild animal lives are averagely happy and worth living (either in terms of their welfare value or in terms of their afforded rights). This belief is also assumed by non-sentiocentric views of intrinsic valuing, which often include theories held under the umbrella of environmentalism (Horta, 2010b, pp.74-75). Some of these consider value in the existence of living entities in nature, some in ecosystems, and others in the continuation of natural processes (see §4.1.2.1 for a more in-depth analysis of these positions). Others find value in nature as a source of nourishment and emotional fulfilment, and it is frequently assumed that on this basis the lives of wild animals – though surely with their hardships – are happy and worthwhile.
One of the reasons people tend to significantly underestimate the welfare prospects of wild animals results from a skewed perspective of what the average animal life entails, and indeed, what the average animal life actually is. Misrepresentative examples of wild animals usually include large adult mammals like dolphins, elephants, lions, and gazelle – animals that may suffer from such causes as predation, “thirst, hunger, and disease, but who may also have several sources of pleasure in their life” (ibid, p.76). But basic estimates place the total number of wild mammals in the live population range of 1011 to 1012, which is in the lower range in comparison to estimations on other vertebrate classes (Tomasik, 2009). Amphibians and reptiles may separately range from 1011 to 1014, whilefish are likely the most prosperous vertebrates with a conservatively estimated range of 1013 to 1015+ (based on preliminary investigation – more research is required to confirm these figures). Moreover, the significant majority of animals that exist are invertebrates, with arthropodic insects existing in the range of 1018 or 1019 and with “species richness […] so great that, to a near approximation, all organisms can be considered to be insects” (Gullan & Cranston, 2010, p.2).
This increase in magnitude ought not be underestimated – numbers multiplied by powers of ten increase exponentially such that the number 1013 contains one hundred sets of the number 1011, and so on. Based on these preliminary figures, the notion that such animals as mammals – and large mammals at that – accurately represent an average in the pool of currently existent wild animals is woefully misguided.
3.2 Evolutionary Disvalue
During the process of natural selection, only the adaptions which help to maximise the transmission of genetic information are selected for, which can occur either directly via reproductive strategy, or indirectly by supporting individual survivability (note that for the sake of simplicity, we will assume that the gene-centred theory of evolution is the case). As such, there is no selective benefit to welfare optimisation in and of itself – positive and negative welfare states can be viewed as no more than “tools for the self-regulation of the homeostasis of organisms and for other aims that maximize the transmission of their genetic information” as non-affective “[c]onsciousness as such does not contribute to fitness” (Horta, 2015, p.21) (Ng, 1995, p.264). Presently it is also worth noting that in this section and for the remainder of this thesis my discussion of reproductive strategy in terms of ‘K’ and ‘r’is a debated oversimplification in the field of wild animal population dynamics. However, because my usage of the term is taken only to describe the fecundity values of different animals (‘K’ as and minimising ‘r’ as maximising offspring count) and their associated traits, and not the otherwise debated implications of these terms, I will assume them to be the case for simplicity’s sake, noting that their possible future refutation does not count as a refutation to my argument in whole.
3.2.1 The State of Nature
In the oft-repeated words of Thomas Hobbes, life in the state of nature without societal infrastructure is “nasty, brutish, and short” (or for the purposes of our argument, consists of extreme negative welfare) (Leviathan, i. xiii. 9). Horta (2015, p.21) notes that this arises because of two circumstances; the limited availability of resources in nature and the existence of evolutionary processes, which together “cause many beings to come into existence, for whom there will not be enough resources”. Competition arises, creating a pressure for selected animals to evolve mechanisms which benefit their survivability in the state of nature (to the extent that the energy expenditure from maintaining these new adaptions is offset by the fitness benefits in allowing an animal to better navigate their environment). Species which heavily invest in these adaptions are often called K-stategists (where ‘K’ represents the carrying capacity of the local environment), as they prioritise efficiency of environmental resources to produce fewer offspring, each with a greater invested energy to increase their chances of survival to the point of reproduction (Pianka, 1970, pp.592-593). K-strategists are most common within the vertebrate subphylum, particularly among mammals and birds (Horta, 2015, p.23).
Sources of disvalue for K-strategists include most notably predation, which usually involves a period of intense suffering followed by death. The period of time spent suffering may range from seconds to minutes depending on the method of slaying, and in outlier cases may even last hours or days. For example, natural predation is often characterised by the image of “a lioness hunting her prey” (Tomasik, 2015, p.135). For such prey as impala with small and vulnerable necks (relative to the jaw size of the lioness) death will probably occur within seconds, resulting from crushed vertebrae at the base of their skull. Zebra on the other hand possess large and muscular necks that cannot be so easily bitten through, and therefore must suffer the slow and painful death of asphyxiation (which in this case is estimated to last some 5-6 minutes).
Other such predators as hyena “tear off chunks of ungulate flesh one bite at a time […] [w]ild dogs disembowel their prey, venomous snakes cause internal bleeding and paralysis over the course of several minutes, and crocodiles drown large animals in their jaws” (ibid). Here it is worth noting that most welfare trade-offs in nature are almost certainly inequivalent, insofar as the positive welfare that can be expected to be gained by killing and consuming another animal does not support the amount of suffering that is caused as a result of this act (Torres, 2015, p.42). Also, and because of obvious limitations in our experiential knowledge, it is unclear for how long a fatally wounded animal is affectively sentient before their mental death.
Not all acts of predation cause immediate death. Kelp gulls have been seen to mortally wound seal pups by attacking and consuming their eyeballs (Gallagher et al., 2015, p.412). Other animals have evolved mechanisms to produce intensely painful venoms; anecdotal accounts by human victims of platypus attacks report that their venom causes “immediate and excruciating pain”, which sometimes “develops into a long-lasting hyperalgesia that persists for days or even months” (de Plater et al., 2000, p.1340). And in cases where prey survive an attack we might expect psychological trauma to develop (assuming the presence of a sufficiently complex emotional system). Earlier we reviewed evidence of PTSD-like symptoms in mice, and since a learned fear of predators may help to augment survival, it seems likely that many other animals experience such lasting harms (note again that it is difficult to evaluate the extent of these harms when they occur because of phenomenological limitations in nonlinguistic inference).
Though less immediate, another significant source of negative welfare for many vertebrates can be found in ways they suffer and die through other non-predatory means. Starvation, dehydration, malnutrition, natural disaster, and adverse living conditions are all commonplace in the state of nature, and wild animals are severely limited in their options for relief of these harms, frequently resulting in slow and painful deaths (Horta, 2015, p.19). One study reports that in high density populations of Soay sheep, more than 90% of lambs and 70% of yearlings die during the winter season, and elephants face a mortality rate ranging above 50% during the first three years of their lives (Animal Ethics, 2016) (Pearce, 2015, p.158).
Other factors that contribute to such high mortality rates in wild animals are disease and parasitism. Bovine tuberculosis (TB) is a bacterial infection which affects a variety of animals including cattle, pigs, dogs, deer, and notably; brushtail possums (this will become important in §6.1.2) (Corner, 2001, p.13). Its effects vary between species, but in possums “it is principally a respiratory infection, with 85% of tuberculous possums having lung lesions” (ibid, p.19). TB also causes swollen lymph nodes to appear, “which may rupture through the skin to form draining sinuses” (ibid). These conditions almost certainly involve significant suffering, particularly given the lifespan before death of possums infected with TB; a median of two months (beginning at the first detection of clinical signs), with one possum recorded to survive a lengthy 15 months while diagnosed with the disease.
TB is merely one example of a wild animal affecting disease demanding welfare consideration, some others include salmonellosis, haemorrhagic disease, and leptospirosis – each of which occurs in a variety of strains with relative (often high) rates of fatality (for a more extensive list, see CFSPH, 2018). Ordinary parasitic diseases, by contrast, rarely kill wild animal populations, though they are estimated to increase an animal’s mortality rate in the wild by 2.65 times in comparison to regular unaffected animals, a figure subject to change depending on “geography, host clade, parasite clade, and many other factors” (Ray, n.d, §3.2.2). Parasitoids, however, necessarily kill their hosts, and often using violent and grotesque methods. Microparasitic diseases such as rabies and mange affect a variety of wild vertebrate animals, while most other parasitoids primarily affect invertebrate populations. Notably, there exists a wasp species that after being laid inside the larvae of butterflies and moths, will proceed to eat their host from the inside out (ibid, §3.2.2.1). In a letter to his colleague responding to this fact, Charles Darwin (1860) recognises the prevalence of suffering in nature, writing; “There seems to me too much misery in the world. I cannot persuade myself that a beneficent & omnipotent God would have designedly created the Ichneumonidæ with the express intention of their feeding within the living bodies of caterpillars”.
3.2.2 Disvalue in Reproductive Strategy
So far, we have mainly considered the disvalue experienced during the lives of most primarily K-selected wild animals. These animals plausibly experience significant suffering (for reasons which we have already described), but we might also infer several sources of pleasure in their lives. Just as negative mental states like pain may serve a purpose in teaching animals to avoid harmful stimuli, we can expect that positive mental states serve to reinforce fitness-enhancing behaviour. Mating is central to gene transmission, so the animals that survive to the point of being able to fulfil their species’ reproductive requirements probably do experience significant positive welfare (Ng, 1995, p.270). We might also consider the satisfaction of other interests for longer living K-selected animals as sources of positive welfare that make it difficult to conclude, with certainty, that their lives contain sufficient negative welfare to be ‘not worth living’. This is consistent with certain deontological accounts of wild animal suffering, which argue that “as terrible as predation is, it is nonetheless compatible with K-strategists’ collective flourishing […] so long as the amount of suffering and premature death does not exceed a certain threshold” (Johannsen, 2016, p.338). However, K-strategists represent only a small minority of the living animals that are likely to exist at any given time.
In nature, the most populous species are those that tend to produce large numbers of offspring at once, each with a minimal invested energy to ensure their individual survivability; these animals are often called r-strategists (where ‘r’ represents the “maximal intrinsic rate of natural increase” of a species) (Pianka, 1970, p.592). It ought to again be noted that when describing animals in terms of ‘K’ and ‘r’ we consider them to be useful guidelines rather than definitive classifications, since differences in reproductive strategy are a matter of degree rather than kind. The problem of suffering arises when we consider that to maintain a stable population size, only one offspring member per reproducing adult must survive to be able to reproduce themselves (and thereby, continue gene transmission). For r-strategists with large progeny sizes, this means that the majority of offspring that come into existence “starve to death, are eaten by other animals, or die for other reasons that usually entail a great deal of suffering” (Horta, 2015, p.23).
Determined by the relevant environmental conditions present during natural selection, a species may evolve toward maximising either ‘K’ or ‘r’ with minimal investment in the alternate, or with a mixture of both. For example, the clutch size of birds in living in Europe and India range in the 1’s and low-10’s with a general estimated average of 3-4, meaning that although birds are primarily considered to be K-strategists, their reproductive strategy employs some r-based elements (Lack, 1947, pp.336-346). The litter size of mammals tends to range in the 1’s and low-10’s per reproductive cycle, with a proposed upper limit of 28 reported in the naked mole rat (Ng, 1995, p.270) (Segelken, 1999). Though depending on the frequency of their reproductive seasons, some mammals may give birth to 100’s or even 1000’s of offspring over the course of their lifespan (Tomasik, 2015, pp.139-140) (Junk & Da Silva, 1997, p.410). Reptiles from the Amazon floodplains include turtles, caimans, and snakes, whose offspring number in the high-10’s and low-100’s per reproductive cycle, while others like the common wall lizard have been recorded to lay up to 570 eggs (ibid, p.413) (Johannsen, 2017, p.337). Amphibians like the common European frog deposit eggs in the lower-1000’s (which may occur 2-3 times per season in a given area), and the common cane toad may lay 8,000–25,000 eggs per clutch (Rastogi et al., 1983, p.243) (Tyler, 1998, p. 116).
Land based invertebrates including many arthropods vary greatly in their fecundity values because of their species richness. Some, like the European earwig, exhibit signs of K-strategy by providing maternal care to their offspring both pre- and post-hatching (which range in the mid-10’s), while others show an extremely low lifetime fecundity of fewer than 10 offspring (Koch & Meunier, 2014, pp.2-3) (Nygård, 1995, pp.39-40). Most other insect species tend to be very r-selected, “[laying] between one or two hundred and a couple of thousand eggs” (Brueland, 1995, p.41). At a high fecundity, non-social insect species has been recorded to lay eggs in the low-10,000’s, while eusocial insect species such as bees and ants produce eggs in the range of 100,000’s to low 1,000,000’s (ibid, p.42). Such numbers increase further as we examine the reproductive habits of marine species, both vertebrate and invertebrate. Commonly fished species such as the Atlantic salmon have been shown to lay eggs in spawns of up to 18,000 (Baum & Meister, 1971, p.764). Likewise, the Alaskan polluck may spawn 1,000,000 eggs per reproductive cycle, Atlantic cod can produce up to 5,000,000, and for the Atlantic bluefin tuna as many as 10,000,000 eggs may be produced in a single spawn (WWF, n.d). From my own analysis, the highest recorded fecundity of any species – vertebrate or invertebrate – appears to be the greasy grouper, which has been shown to reproduce in excess of 340,000,000 eggs per individual (Jeyaseelan et al., 1998, p.91).
Let us now consider the individual welfare of these animals. In comparison to the r-strategists we have just considered, K-strategists like mammals and birds can be said to face fair odds of surviving into adulthood, making it less clear and more objectionable that the lives of surviving individuals are welfare-negative. However, for lives that last mere hours or minutes before ending in a probably painful death (as is the case for most primarily r-reproducing animals’ offspring), the probability that they will encounter any significant welfare-positive experience that is sufficient to compensate for the otherwise welfare-negative experiences in their short lives is small. This leads us to think that their lives are averagely welfare-negative. And pairing the above fecundity values corresponding to animal class/phylum groupings with the figures which we have already considered on the estimated living members of each, there exists a clear trend by which primarily r-maximising animal groups greatly outnumber K-strategists (Horta, 2010b, p.80). Thus, even assuming that most K-strategists live happy and fulfilling lives, the overwhelming numbers of welfare-negative r-strategist lives give us good reasons to preliminarily conclude (C1), that ‘the experience of negative welfare can be expected to outweigh the experience of positive welfare for the lives of most wild animals’.
For most oviparous r-strategists with a spawn size as we have just considered, it is possible that not all offspring will hatch with the emotional capacities belonging to an adult member of their species which allow for the experience of extreme suffering. This is relevant insofar as we treat affective sentience as the defining criterion for moral inclusion. One may therefore reject the case for consideration of newly-hatched offspring on the basis that it is qualitatively worse for a viviparous new-born to, say, be eaten alive shortly after coming into existence (which assumes that viviparous births, being generally more K-selected and with a greater energy expenditure per offspring, are born with more complex emotional experiences at their disposal). But this is not necessarily the case – precociality may be even more expected in r-strategists than K-strategists owing to their greater need for premature self-care and their lack of maternal guidance. Furthermore, many fish and amphibians (and many species of invertebrates) already possess a functioning nervous system at and before the time of their hatching, while for others these capacities “develop rapidly with experience of independent life” (EFSA, 2005, p.38).
This shows us that although it is possible that many of the animals who die premature deaths resulting from r-strategy are not affectively sentient of their deaths, an important percentage of them probably are, and there is a non-negligible possibility that most if not all possess some degree of individual welfare (this assumes that we already have good reasons for believing in the affective capacities of a regular member of their species, as per the functionalist method of mental state inference established in chapter two). However, further research is required to determine the extent to which we can infer affective sentience in premature animals, according to relevant species-specific facts which also need be determined (for example; does nociception occur during the embryonic stage of an animal?). Other such yet-uncertain welfare-relevant factors include averages for infant mortality rates and length of life per-surviving individual.
3.3 Conclusion
In this chapter we first established the case for considering wild animal welfare, which included debunking widely held beliefs about what the average wild animal life entails, and then analysed the extent to which wild animals ordinarily suffer under naturally occurring circumstances. In this analysis we considered two general reproductive strategies which favour either ‘K’ or ‘r’ to maximise transmission of genetic information, and the commonly occurring welfare issues that are associated with each. r-strategists are of particular note as they are many times more populous than their ‘K’ counterparts, have many times more offspring (and correspondingly, higher rates of infant mortality), and live generally shorter lives. Given that most r-selected newborns can be expected to live relatively short and painful lives, and given the usual limited availability of natural resources which may act as sources for positive welfare in their lives, it is reasonable to conclude that these animals “live lives whose overall levels of wellbeing are negative” (Horta, 2010b, p.79). And since r-strategists account for almost all existing animals, we may plausibly conclude that suffering dominates enjoyment in nature many times over.
4 Wild Animal Ethics
In the previous two chapters we have established three premises: (P1); ‘a wide variety of animals are capable of experiencing significant negative welfare over the course of their lifetimes’, (P2); ‘natural conditions (ie; resource scarcity, reproductive strategy) tend to create excess suffering and little happiness for wild animals’, such that (C1); ‘the experience of negative welfare can be expected to outweigh the experience of positive welfare for the lives of most wild animals’. But by itself, this does not carry any express normative invocation for action. We must first establish (P3); that ‘we have good reasons to oppose the suffering of wild animals’, and (C2); ‘we ought to intervene in natural processes to improve wild animal welfare’. The first of these premises shall primarily be derived in §4.1 in which I argue for sentiocentrism against competing ecocentric and anthropocentric views (which usually disallow natural intervention to improve animal welfare). And the second, in §4.2, where I shall defend the interventionist thesis as a logical consequence of most sound sentiocentric moral theories.
4.1 Morally Considering Wild Animal Welfare
4.1.1 Sentience-based Arguments
The view that wild animal suffering bears sufficient moral weight to justify intervening in natural processes can be arrived at from a number of normative theories which take into account what is good or bad for sentient individuals. Theories of distributive justice with a maxim principle (for example; egalitarianism, sufficientarianism, prioritarianism) all disvalue the suffering-rich lives of non-mating r-strategists, as do most other utility-based moral theories (Horta, 2010b, p.79). Theories of wellbeing which take into account elements other than hedonism may also be found to derive this conclusion.
For the desire fulfilment theorist, the short length of most animal lives make the accomplishment of any of their significant desires plausibly unlikely, and the manner of their deaths violates their held desire not to be eaten, or to die in other such painful ways. And while the judgement of basic goods held by an objective list theorist to benefit human wellbeing may not be the same for animals who have different preferences and requirements, it is difficult to imagine that a short and suffering-rich life does not violate at least one of the rights held on their own species-specific list (for example, pleasure is defined as an objective good for humans on account of its qualitative first-person properties, which we may reasonably expect to be the case for other animals – see Rice, 2013, pp.206-207). It may be argued by an objective list theorist that ‘life’ is the only basic good that can be afforded to individual animals (leading them to adopt a biocentric view of what is valuable), but this would require having a good enough argument to justify excluding animals from the many other goods that are afforded to humans (which, as per our rejection of anthropocentrism in §4.1.2.2, is problematic).
Furthermore, rights-based approaches such as we have already considered by Johannsen (2017, p.338) argue that the circumstances faced by r-strategists in nature disallow them “the opportunity to experience the enjoyment associated with reaching a stage in life where one has learned to competently manage” – to flourish in their natural environment (Horta, 2013, p.118). While K-strategists may be born disposed to face the challenges found in their natural environment which make their lives worth living, the same cannot be said for animals who are deprived of this chance to begin with. Intervention in nature is thus required to ‘even the odds’.
My own moral position (which I shall advance over the next chapter) is act-centred and thus consistent with most welfare-maximising consequentialist approaches to suffering reduction, putting forward general considerations on wild animal welfare that ought to be made for current and future interventionist strategies. One of these considerations that is currently worth noting argues to spend fewer resources differentiating between the many ethical positions one may take to establish the importance of wild animal welfare, and place a greater focus on practical methods of intervention. This is because for the time being, most sentience-considering arguments identify the disvalue experienced by wild animals to be of fundamental moral concern while others at least partially consider their welfare, and any such methods of intervention likely to be considered in the near future will not differ significantly from one another so as to warrant the promotion of a single moral approach. Thus, the welfare-maximising consequential approach to wild animal suffering reduction that I shall advocate in this thesis is sufficient for making ethical judgements, and it is not necessarily tied to the premise (P3) that ‘we have good reasons to oppose the suffering of wild animals’.
4.1.2 Rejecting Sentiocentrism
One method of rejecting (P3) (and thereby, C2) is to deny the axiological relevance of affective sentience by adopting either an ecocentric position which assumes consideration for “entities such as natural processes, ecosystemic relations, ecosystems, biocenoses and species” (as concepts), or an anthropocentric position, which values the interests of humans above the interests of other sentient beings (according to some morally defensible boundary separating the species) (Cunha, 2015, p.52). Furthermore, these beliefs must be held either independent to or overriding of sentiocentric valuing systems, such that the claim to intervene in natural processes to benefit wild animals can be considered unjustified. A good definition of speciesism that is proposed by Horta (2010a, p.5) describes it as “the unjustified disadvantageous consideration or treatment of those who are not classified as belonging to one or more particular species”. The first of these views I shall now argue commits speciesism by positing a misguided notion of moral relevance, while the second is usually speciesist in its discrimination of sentient beings.
4.1.2.1 Ecocentric Valuing
Rather than taking into account the individual welfare of sentient beings, ecocentric views place moral importance on holistic concepts such as ecosystems – which can sometimes contain affectively sentient beings with interests, but also entities which are probably not affectively sentient (if sentient at all) like plants, microorganisms, among other such living entities. Notably, interests are said to exist for geographical features such as landscapes, mountains, and bodies of water – this is taken as an “axiological thesis about the kind of things which are intrinsically valuable” (Paez, 2015, p.168). Such ecocentric views that are held for metaphysical reasons posit intrinsic value in natural entities, but an ecocentric position may also be defended by recognising the value of such entities instrumentally, only in terms of their worth to morally considerable beings (like humans or other animals). Adopting an intrinsic view of ecocentric value, the interests of naturally occurring phenomena are taken to promote their own preservation and conservation rather than the sometimes individually held interests of their constituent sentient members, which constitutes a “thesis about the result of the balance of reasons when intrinsically valuable entities are affected” (ibid, p.168).
For the first of these, we are presented with a nonspecific metaphysical definition of value that derives its significance from human-independent features of natural occurrence. Aldo Leopold (1949, pp.224-225) summarises this view in his attempt to establish a land ethic; “[a] thing is right when it tends to preserve the integrity, stability, and beauty of the biotic community. It is wrong when it tends otherwise” – thereby encapsulating a second thesis of non-interference. Furthermore, Leopold advocates for our understanding of morality as a feature arising from continuing evolutionary and biological processes (Nelson, 1998, pp.743-744). By recognising our own species to be a part of a larger biotic community, he argues to extend our ethical consideration to other natural entities and concepts (such as we have already described).
Certain other philosophers defend this premise only insofar as it relates to the moral consideration of (affectively) non-sentient living things. Joel Feinberg (1974, p.51) writes; “[Plants] are not ‘mere things’; they are vital objects with inherited bio-logical propensities determining their natural growth. Moreover, we do say that certain conditions are ‘good’ or ‘bad’ for plants, thereby suggesting that plants, unlike rocks, are capable of having a ‘good’”. And taking a holistic view, Callicott (1980, p.320) argues for the instrumental valuing of individual organisms belonging to ecosystems, which are modulated according to their relative importance or cruciality within such ecosystems. Callicott therefore defends a context-sensitive view of morally considering animals which, rather than taking into account their individual welfare, places importance only on the role that they play for the interests of intrinsically valuable entities (which he posits on the same basis as Leopold). But this notion of attributing interests to non-affectively capable entities is problematic.
Recall our discussion in chapter two in which Singer argues against the notion that such entities as motor engines possess interests (Sagoff 1984, p.298). Natural processes, biocenoses, and ecosystems (among other such concepts) cannot be said to possess interests in the fullest sense of the term simply because they lack the first-person subjective qualities that are associated with affective sentience, which is to say that any such value that is derived from the satisfaction of their supposed interests cannot be experienced on their own behalf. And given, even, that there is something that it is like to be a plant and that some are sentient (as per many popular theories concerning mentality and consciousness); our lack of good evidence supporting their having individual interests or preferential states, such as the evidence we take to infer emotional states in animals, means it is more plausible that “[t]heir welfare is necessarily zero, just like nonsentients” (see §5.1.2 for further discussion on this topic) (Ng, 1995, p.258).
These entities might be said to possess functions (as is argued by Callicott, 1980, p.328). For example, natural selection is a process that tends to maximise gene transmission among organisms, and birch trees usually flourish in conditions of adequate sunlight and rainfall. But given that there plausibly exists no responsible being for whom the fulfilment or accomplishment of these functions is better or worse by, we must place them in the same category as, say, the processes which determine the physical laws – the only relevant difference being that the former is contingent upon naturally occurring circumstances of evolutionary development. And since natural law cannot be used to infer or derive moral prescriptions (lest we are willing to grant that any imaginable contingent process carries normative weight), it cannot be logically inferred that value exists in and of itself for non-sentient entities. There are, of course, beings for whom the continuation of natural processes is something worthwhile, but this is a separate discussion to whether environmental value can be derived intrinsically.
As a final reason to reject the metaphysical definition of environmental value, Horta (2010b, pp.83-84) claims that the theory’s followers usually only suppose moral requirements insofar as they do not conflict with already held anthropocentric interests (to varying degrees, given the anti-environmental impact of most human activities). He notes that few choose to accept the undesirable conclusion of culling most living humans to benefit the environment, and that “[t]his can help us to see why views that value wholes more than individuals are very questionable” (Horta, 2015, p.28). I disagree; there exist many consequential reasons for why, assuming ecocentrism to be a sound moral theory, it would be detrimental to promote the repugnant conclusions which result from the argument’s reasoning – not least of which because many would see it as an adequate refutation of the theory as a whole. As we shall discuss in §5.2.1, public perception ought to be realised as a significant hurdle for the success of a given applied ethic, so the reasons given by ecocentric theorists for ignoring or disvaluing ‘human extinction’ as an accepted goal of the movement has no bearing on the reasoning of the underlying argument (which we have, for other reasons, found to be unsatisfactory).
The second argument for ecocentrism in any case explicitly accounts for conflicting anthropocentric values as it identifies environmental value instrumentally, only in terms of its value for morally considerable beings (like humans). Ecocentrism is thus useful in “realization that […] ecosystems constitute the ‘life-support system’ for humans” and other animals (Kopnina et. al., 2018, p.111). Additionally, humans (and plausibly, many other animals) gain significant pleasure from the aesthetic features that are present in nature. But given our analysis in §3, very few animals within the scope of those that exist would identify with ecocentric values given their negative welfare consequences, so it is likely only humans (among few other K-strategists) that could present this argument. And, as I shall now argue, most forms of anthropocentrism are unjustifiable.
4.1.2.2 Anthropocentric Valuing
The reasons we have for rejecting human-centred valuing systems are well described in Peter Singer’s ‘All Animals are Equal’ (1974, p.103), in which he argues that because the features that make humans morally considerable are not species-exclusive, there exists no moral basis for excluding nonhuman animals who possess those same features under “the basic principle of equality that most of us recognize should be extended to all members of our own species”. Singer makes two arguments against anthropocentrism; equality – a moral idea based on equal consideration of interests – does not depend on species membership (or other such arbitrarily-defined groupings), and that suffering – which is not an exclusive state to humans – is something that we ought to take into normative account. Anthropocentrism relies on claims about the factual differences that exist between humans and nonhumans to justify discrimination between these two groups, which are usually in conflict with these two arguments.
Horta (2010a) responds to an argument that is made by Mary Midgley on this basis. Midgley’s argument claims that the factual differences between humans are insignificant to the degree that it is not possible to justifiably discriminate on grounds of race, creed, or sex. Species, however, are a significant grouping, and it follows that the factual differences which separate humans from other animals result in different interests (creating a factual inequality if we are to provide equal treatment between them). Two problems immediately arise in Midgley’s reasoning. First, it is not the case that the factual differences between humans are so small as to eliminate differences in their factual interests. Horta notes that differences in skin colour measurably affects one’s chances of contracting skin cancer, and the immunological differences that exist between different human groups may give rise to a factual imbalance in their respective medical needs and requirements (ibid, p.8). Remember our earlier observation that the differences that exist between organisms are a matter of degree rather than kind – species classification is useful only insofar as it allows for the general group identification of sufficiently genetically similar organisms.
The second and more pressing difficulty with Midgley’s argument arises in its implications. Even granting as she does that all humans are averagely the same in terms of their interests, basing an equality on the factual differences that exist between groups of humans would allow for the possibility of justifiable discrimination were such differences to emerge (say, as an outcome of recent scientific investigation). And last, it is not apparent that morality itself can be reduced to the factual features of morally considerable beings, particularly given that our reasons for morally considering such beings to begin with are conceptual (for example; we don’t realise suffering as a type-property that is exclusive to the human mind – here Midgley’s reasoning takes a similar form to Bermond’s brain state argument as discussed in §2.1.1). This leads us to conclude that “[e]quality is a moral ideal, not a simple assertion of fact”, in which our equal consideration of the differences that exist between organisms takes into account the significance of their interests, rather than their factual contents (Singer, 1974, p.106) (Horta, 2010a, p.9). Factual circumstances are relevant insofar as they affect one’s held interests, but they are not the sole determinants of consideration.
The only defensible form of anthropocentrism (see §5.1) takes into account the interests of humans above the interests of nonhumans according to our relative epistemic uncertainty of other animals actually possessing the morally relevant features which we have better reasons to believe exist in humans. This, I admit, can be considered a case of very slight justifiable discrimination on the general basis of species membership, but given the many other welfare considerations which I shall also propose in chapter five, it is far insufficient to warrant the rejection of most other sentient being’s welfare to benefit humans.
4.2 Other Arguments Against Natural Intervention
Many of the arguments which reject natural intervention employ abductive reasoning as an ‘inference to the best explanation’, leading one to take on one of the valuing positions which we have already argued against. Such forms of inference tend to assume the premise that there is something morally worse about animals being affected by humans than animals being affected by natural processes, assuming that their welfare value remains the same in both cases. A possible conclusion that may be derived from this reasoning is that mere welfare does not account for that ‘immoral something’ which separates these two cases of animal suffering. Callicott (1980, p.336) describes this as “the transmogrification of organic to mechanical processes”, by which there is something unnatural and therefore immoral about our treatment of animals in such domestic practices as, say, farming and experimentation. Or it may be argued reductio ad absurdum that the premise (C2) of our argument to intervene is so farfetched that there must be a logical inconsistency in the premises (Sagoff, 1984, pp.303-304). But it is clearly disputable that there is something by itself immoral with intervention, particularly if we employ consequence-based reasoning to establish our sentiocentric position. And furthermore, to make an abductive inference rejecting natural intervention we require an alternative, more plausible explanation of what is considerable, which as per our discussion in §4.1.2 appears does not exist. Let us, then, consider some of the objections that may be made on the basis of what is better for sentient beings.
4.2.1 Consequentialist Arguments Against Intervention
A popular consequence-based argument in rejection of natural intervention is that we cannot reasonably foresee the outcome of our actions while interfering with nature and may, either as a direct result of our actions or by accidentally altering loosely connected causal relations in ecosystems, increase the overall suffering that exists in nature. This argument cites our epistemic limitations about nature as good reasons to believe that systematic intervention “would be futile at best and ecologically disastrous at worst” (Johannsen, 2017, p.336). When considering the extent of his animal liberation ethic, Peter Singer (2015, p.326) writes; “The short and simple answer is that once we give up our claim to “dominion” over the other species we should stop interfering with them at all. We should leave them alone as much as we possibly can. Having given up the role of tyrant, we should not try to play God either”. His argument claims that because of past failures to be able to accurately predict the consequences of interventionist measures, and because of the extent that animals are already harmed by humans, intervening to improve the lives of wild animals is not normally required (note that Singer’s view on intervention has since changed – see PEI, 2016).
The second of these claims is, as we have already established in chapter three of this thesis, untrue; the amount of animal suffering in nature dwarves that which occurs domestically. And to the first, it is simply not the case that we are unable to conduct research and collect evidence to support our knowledge of ecosystemic relations and their associated welfare issues. It is likely that some interventionist strategies carry “considerable probabilities of triggering ecological catastrophes” which may result in widespread surplus suffering (Torres, 2015, p.39). But with sufficient research to promote the mitigation of these anticipated harms to the best possible degree, such strategies can be employed to produce desirable consequences. Even such moderate forms of intervention as providing food, water, and basic medical care to wild animals can be expected to reduce their suffering to such degrees that, giving a proportionate consideration of their interests to the interests of most humans who are regular recipients of humanitarian aid, the utilisation of resources for expected welfare gains can be considered many times more efficient (Paez, 2015, p.167).
Horta (2010b, p.84) furthermore points out that natural intervention for anthropocentric and ecocentric aims are already commonplace, and with a success rate that does not warrant their arrest. And last, Mikel Torres (2015, p.39) notes that “uncertainty affects equally the interventionist and the anti-interventionist theses; it neither supports nor opposes intervention as such”. To base an anti-interventionist argument on the possible negative welfare consequences that might accidentally result from intervention incorrectly assumes that we already have sufficient knowledge to say that non-intervention will result in better consequences, such that intervention cannot be justified – but evidentially this is not the case. For these reasons, we must accept that on the basis of welfare consideration alone, intervention may be permissible.
4.2.2 Deontological Arguments Against Intervention
As we have already discussed in the case of Donaldson and Kymlicka’s (2011) flourishing argument, intervention presents itself as a moral duty to ensure that sufficient individual animals are given the opportunity to flourish, and possibly also for some to form sovereign communities. They even grant that intervention is possible “starting with small-scale pilot projects, in order to build up our knowledge about how to re-engineer nature so as to reduce suffering overall” (ibid, p.164). Other deontological positions, however, may reject the interventionist thesis for wild animal suffering on account of competing rights and duties.
Clare Palmer’s contextual argument holds that although we have negative duties not to harm wild animals on the basis of their status as morally considerable beings, we neither have any positive duties to intervene on their behalf, since we have no positive duties to assist anyone – regardless of species membership. Rather, we have special obligations to help others, which “emerge from certain morally-relevant contexts, histories, relations and entanglements” (Palmer, 2015, pp.206-207). Parents, for example, have special obligations to their children because of their participatory roles in bringing into existence beings that are morally considerable, not on account of the morally relevant features themselves. On this view, natural intervention is simply not required in absence of a morally-relevant entanglement with wild animals. Rather, it is permissible per choice of a moral agent.
Arguing on the grounds of sentience, Palmer’s argument attempts to avoid including the lives of wild animals as legitimate recipients of our moral attention by proposing a set of context-sensitive rules that do not take into account species membership, thereby seeming to avoid the mistakes that would otherwise result in it committing speciesism. But her notion of special obligations is problematic, both for practical ethics and for equally considering the interests of different species. Consider first the vagueness of the term ‘moral entanglement’. Are the circumstances in which one’s life becomes ‘entangled’ with another ethically tractable? Palmer responds by claiming that large-scale entanglement (for instance; anthropogenic climate change) results in different kinds of special obligations, but that in her paper she lacks the space to properly explain their nature (ibid, p.208). But granting that it is possible to practically navigate circumstances of moral entanglement, we have other good reasons to reject the contextual view.
First, the prevalence of r-strategy in nature means that not all animals care for or even about their offspring as humans do. Only a comparatively small number of animals (usually K-selected) offer support for their young, and “[a]lthough parental care is assumed to increase offspring survival, filial cannibalism, the consumption of one’s own viable offspring, commonly co-occurs with parental care” (examples of which include species of rodent, bird, spider, beetle, and most extensively documented – fish) (Klug & Bonsall, 2007, p.886). Thus, it is either not the case that there are special obligations between most other animal species that exist, or those obligations are being constantly violated without repercussion (which we might expect, given that most sentient beings are not moral agents).
Catia Faria (2015, p.212) also responds to Palmer’s dismissal of general duties and obligations; “[e]ven conceding that parents have special obligations towards their own children, it does not follow that they do not have reasons to assist other children in need”. Likewise, her argument grants that it is morally permissible to ignore the welfare of human groups who are, say, victims of natural causes, or other such harms that occur without causal relations to a morally culpable agent (ibid, p.213). To consider the lives of sentient beings in equal terms that are altered under circumstances of special obligation, as Palmer does, evidently reduces the value of one’s life to the small chance of being born in the right circumstances, which does not seem like a truly equal basis for consideration. To otherwise claim that special obligations emerge for equality-based reasons (as her argument must do to successfully navigate these objections), there must be other justifiable reasons for differentiating between wild animals and non-causally related humans (ibid, p.215). And since there are not, Palmer’s argument either must include wild animals in its scope (thereby strengthening the case for intervention) or admit to committing speciesism.
4.2.3 Virtuous Arguments Against Intervention
A third relevant perspective in the discussion of wild animal suffering that is sometimes neglected concerns our treatment of other animals according to the virtues which we hold as humans. As a defender of animal consideration based on moral virtue, Rosalind Hursthouse (2011, pp.131-132) puts forward that compassion (which encompasses our consideration for the welfare of beings who are affectively sentient) is a particular expression of the virtue of love, which “not[ing] Kant’s sapient observation […] needs to be tempered with the virtue of respect”. She therefore advocates for the virtue of respectful love, taking into account the rights of others to decide how to live for themselves, despite the potential conflicts which may arise with regard to their individual wellbeing. This view may be held both toward individual sentient beings and holistic concepts (as are advocated for moral consideration by Leopold in §4.1.2.1), as it embraces the non-anthropocentric aspects of both systems (ibid, pp.132-133).
With regard to wild animals, the virtue of respectful love argues for non-interference on a similar basis to Donaldson and Kymlicka’s flourishing argument, and it is likewise susceptible to the same criticism. The concept of autonomy that is central to Kant’s philosophy of respect is seemingly separate from merely being able to act and respond with intent (as in the case of affective sentience) – it is considered the freedom for “laws to be made and laid down by oneself”, which (to a degree) requires moral agency (Johnson & Cureton, 2018, §10). But it, again, seems implausible to claim that high fecundity r-strategists fulfil this criterion, particularly in the early days, hours, or even minutes of their lives, which is the period in which most of these animals usually die.
Hursthouse (2011, p.133) furthermore claims that by nature of anthropocentric involvement, wildlife intervention for animal welfare ought to be considered “yet another manifestation of anthropocentric arrogance”, despite its consistence with the virtue of compassion by itself. And since this seemingly competing virtue is argued to consist of respectful compassion, it is difficult to reconcile the theory in such a way as to provide a prescription on intervention. On this basis, Torres (2015, p.48) rejects Hursthouse’s argument as a sufficient objection to the interventionist thesis, but notes that “these problems do not disqualify virtue ethics as an adequate and stimulating approach to moral problems”. Taking a bolder stance, Knutsson and Munthe (2017, pp.10-11) put forward character-centred accounts of virtue as complementary to consequential or deontological approaches which, rather, take into account material outcomes or duties, rights, among other things. This view is additionally consistent with the welfare consideration which I proposed in §4.1.1.
4.3 Conclusion
This chapter defended the axiological relevance of affective sentience as the criterion for moral considerability and established the widespread consideration of wild animal welfare held across sentiocentric normative theories, to establish the premise (P3); ‘we have good reasons to oppose the suffering of wild animals’. Next, we considered and rejected sentiocentric reasons for not intervening in nature to assist wild animals, such that the premise (C2); ‘we ought to intervene in natural processes to improve wild animal welfare’ is justified.
5 Welfare Considerations
Earlier in §4.1.1 we proposed one consequentialist welfare consideration for practical intervention in nature (given C2 of our argument), which is based on the general consensus held across sentience-based theories in ethics that wild animal suffering is something that is highly disvaluable in itself (or at the very least, that we have good reasons to consider their status as recipients of suffering-causing natural processes to some degree). But from a strictly welfare-considering approach is it not clear exactly how we ought to intervene, particularly given our partial ignorance of the causal relations that exist between ecosystemic processes and the long-term welfare impacts that might result from intervention. Thus, in this chapter I aim to propose a general nonspecific welfare-maximising calculus for taking into account welfare-relevant factors pertaining to the lives of wild animals. These factors, which I shall propose in §5.1, can be used to calculate the estimated welfare impacts of intervention on an individual and population-wide basis, which I shall discuss in §5.2. The latter of these, it is worth noting, has more bearing on significant interventionist undertakings while the former applies only to isolated cases in moral decision-making. And since what we are proposing is an applied ethic, I shall continue in §5.2 to consider practical measures related to the likelihood of any given interventionist strategy being accepted and implemented under a set of democratically held social principles.
5.1 Consequentialist Welfare Considerations
While discussing non-sentiocentric views of moral consideration in §4.1.2, we described speciesism as “the unjustified disadvantageous consideration or treatment of those who are not classified as belonging to one or more particular species” (Horta, 2010a, p.5). Horta also notes that “a justified prescription according to which only the members of some species could enjoy certain goods or benefits would not be speciesist”, but later argues against the usual claims to moral supremacy made for humans in relation to other animals (ibid, p.6). Although I shall not present an argument which explicitly favours human interests over nonhumans (the inverse may, rather, be shown), I believe that there are morally relevant features which correspond with species membership that are capable of justifying discrimination between the interests of sentient beings. Such features, I argue, are determined on the basis of their estimated welfare impact on non-morally autonomous agents and are not, as Knutsson and Munthe (2017, p.10) might otherwise claim, related to judgements of their character (by which my argument ought not be understood as an attempt to discriminate between individual beings for reasons that are of their own fault). This act-considering view maximises welfare by advocating for an equal consideration of interests, and in such cases as there are conflicts of considered interests, it endorses the outcome which we can predict carries better net-welfare consequences. And although I have already described the folly in treating species as definitive classifications for different animal groups, I shall for the sake of simplicity treat them as generally useful classifications which help to identify the presence or absence of certain welfare-relevant features between individual animals.
5.1.1 ‘All Animals are Equal’
The most immediate welfare-affecting features to consider between differing animals are those which affect their capacities to inflict suffering upon each other. Natural predators, for example, happen to be born into bodies which require them to inflict suffering and death upon other organisms for their continued survival. Because of this, the welfare impact for their continuing to exist is largely negative in comparison to some of their herbivorous counterparts (when considering only the effects of their diet and psychological dispositions towards violence). Another factor related to the estimated welfare impact of sentient beings existing in nature is, notably, their fecundity value, by which we ought to consider the reproductive habits of a member of a species as an act which also affects their welfare impact. Because most animals are unable to be able to make autonomous moral decisions (for example; not to mate, or to hunt and kill other sentient beings), and because the welfare calculus that I propose is act-based, I advocate for the allocation of general nonspecific values which detail the predicted average outward welfare impact of an animal belonging to a particular species (this value shall be represented as ‘O’, while ‘x’ represents a species).
But Ox only represents the welfare value that can be expected to be experienced by other sentient beings as a causal result of x’s average existence and does not take into account the personal welfare that is experienced by x themselves (let’s call this value ‘P’). Say, for example, the satisfaction of interests for the lioness we considered in §3.2.1 to kill and consume an adult impala totals 3 welfare points (that are experienced on the lioness’ behalf). Let us also make the conservative assumption that the impala’s experience of being killed and eaten in this manner totals -6 welfare points, so that it is at least twice as disvaluable as the lioness’ experience. Although we estimate a greater welfare impact on behalf of the impala, it is not the case that the positive welfare experienced by the lioness has no bearing on the total welfare outcome of this isolated case (unless one is willing to assume an asymmetrical welfarist position that places greater importance on minimising suffering than it does on maximising pleasurable states). We derive -3 welfare points (Ox + Px), which leads us to conclude that this act (and remember, not the lioness herself) is unjustified and thus we have reasons to intervene.
But nonetheless, the lioness is an integral component to this exchange, and to future similar welfare exchanges that will occur during her lifetime. Given that her minimum daily requirement of meat totals 5 to 8.5 kilograms and that she lives 10 to 15+ years, we can calculate that the normal surviving lioness consumes from 18250 to more than 46537.5 kilograms of meat in her lifetime (with a mean of 32393.75) (Sunquist & Sunquist, 2002, p.292 & 297). For such small prey as impala (30kgs), this averagely amounts to 1079.79 deaths per surviving lion, while for such large prey as buffalo (500kgs) only 67.78 deaths are required. Using the conservative welfare figures posited in the previous paragraph, an average surviving lioness may result in anywhere between -203.34 to -3239.37 welfare points over the course of her lifetime (which only takes into account the disvalue that is caused by predation, and assumes that the negative welfare value is equal for the deaths of both impala and buffalo).
Given these figures alone (and without considering the welfare-impact caused by disruptions in ecosystemic relations), we are led to the conclusion that it would be better for such predators as lions not to exist. Other factors which one would need to consider for a complete welfare estimation include litter size (which for lions in the wild is dependent on geographical location, averagely ranging from 1.7 to 3.3) and cycle frequency, infant mortality rate (male lions often commit infanticide, and the fierce competition for food causes as many as 28% of cubs aged between six and twelve months to starve to death), and the general frequency of positive welfare encounters in their lifetimes (which we might infer on the basis of interest satisfaction, and welfare estimations on the lives of their prey were they to escape unscathed) (ibid, pp.295-296). The length and relative pain that is experienced during an animal’s death is another important factor to consider in relevant welfare calculations, particularly for the lives of numerous r-selected offspring whose welfare has significant bearing on the Ox value of reproducing r-strategists. It is also worth noting that our relative uncertainty of the cases which contain welfare and our uncertainty of qualitative similarity of experienced welfare states between animals tangibly affects the net-expected result from our welfare-considering calculus. Thus, we ought to morally account for uncertainty within our framework.
5.1.2 Adjusting for Uncertainty
5.1.2.1 Sentio-uncertainty
For the first of these two uncertainties – that of affective sentience – let us return to where our argument left off in §2.3. We concluded with the premise (P1); ‘a wide variety of animals are capable of experiencing significant negative welfare over the course of their lifetimes’, which we base on the available evidence that exists on the welfare capacities of various animals. Assuming simply that all animals are capable of experiencing affective states – a form of ethical extensionism that is often adopted by animal advocates under conditions of epistemic sentio-uncertainty – takes the form of a ‘precautionary principle’ in which it is better to “assum[e] animals have consciousness and feelings on a ‘just in case they do’ basis alone” (Bradshaw, 1998, pp.111-112) (Carter, 2005, p.22). Other theorists like Jonathan Birch (2017, p.4) argue for a modified version of the precautionary principle which takes into account practical measures related to the legislative protection of sentio-uncertain animals. And last, some advocate for an ‘all or nothing’ approach that assumes animal sentience in the absence of conclusive evidence showing that animals are not sentient (although it is unclear what kind of evidence would count toward this claim). Peter Singer (2015), for example, constructs an analogical argument for animal pain inference based on our recognition of pain in humans (which is, too, an inference – see ibid, p.41). He claims that “there are no good reasons, scientific or philosophical, for denying that animals feel pain. If we do not doubt that other humans feel pain we should not doubt that other animals do so too” (ibid, p.46).
But this position is problematic, not least of which because Singer’s examples tend only to include large mammals and not small invertebrates for whom we are much less certain of being sentient. This is because the prescription of affective sentience is not a categorical matter. If our aim is to maximise the predicted experienced welfare of morally considerable beings, then we cannot adopt an all-encompassing ethic for inferring who these beings actually are, lest we ignore the considerable differences that separate our reasons for assuming sentience in different beings to begin with. The simple fact that we have reasons for thinking that a dog and also a fruit fly are affectively sentient does not warrant the conclusion that their sentience probability values are equal to one another. We plausibly have better reasons supporting our inference to dogs’ affective capacities (which include their close evolutionary continuity to humans, behavioural flexibility, and semblance in nervous system structure) than for fruit flies (for whom evidence is limited to showing their capacities in nociceptive reflex avoidance, basic operant conditioning, and response to analgesics) (Elwood, 2011, pp.177-179).
But neither does our current legislative method of ‘drawing the line’ solve this issue for then we risk not considering beings who actually are affectively sentient, due either to a lack of research pointing out their affective capacities, or because the research that already exists does not pass an arbitrary threshold which we ourselves define. Simply, “uncertainty of sentience cannot justify either lack of consideration or equal treatment”, and ‘draw the line’ approaches to consideration ought to be rejected (Chan, 2011, p.325). A compromise to this apparent crossroads in welfare-considering ethics (and other sentiocentric ethical theories) that is proposed by Kai Chan attempts to account for sentio-uncertainty by explicitly including it within a given normative theory. Chan argues to modulate duties in accordance with the general nonspecific probability values we have of different beings possessing a point of view, treating epistemic uncertainty as a tangible welfare-affecting factor that influences the moral decision-making process. By adopting this stance, we “give moral patients exactly their due based on available evidence, acknowledging the considerable uncertainty therein” (ibid, p.329).
In an analogy to show the utilitarian benefits which result from this form of reasoning, Chan asks us to imagine that there exists a rare disease called ‘lineitis’ which may or may not have affected eleven people, and that we only have three pills for its treatment. And as sometimes occurs in ordinary circumstances of medical triage, none of these eleven patients we can be definitively sure carry the disease. Rather, we have uniform estimations of probability (0, 0.1, 0.2, …, 1) that are conferred from the available evidence of each patient’s visible symptoms, genetic compatibility with the disease, pre-existing physical health (among other such relevant factors). Given a ‘draw the line’ approach with, say, a reasonable threshold of 0.5 probability value for having the disease, pill distribution will result in an efficiency value of approximately 0.75, thereby effectively treating 2.25 people. If we, rather, modulate the distribution of pills according to our relative uncertainty values, then the three pills can be expected to effectively treat 2.7 people with a pill efficiency value of 0.9. And given, also, that this approach scales both in terms of numbers and uncertainty, “[i]n the face of highly variable costs, benefits and probabilities, modulating responsibilities based on probabilities can have an immense pay-off” (ibid, p.330).
Chan proposes the application of this uncertainty framework on nonhumans under conditions of uncertain sentience, but does not attempt to provide sound normative justification for his use of the ‘Golden Rule’ as a guiding principle for moral decision making (which states ‘do unto others as we would have them do unto us, were we in their position’ – see ibid, pp.330-331). Although I do not intend to deny the golden rule as a valid method of normative guidance for considering the lives of wild animals, it is possible (as with many other rule-based frameworks) that one’s moral prescriptions under this principle may come into conflict with the otherwise welfare-maximising approach. So, for the argument laid out in this chapter, we shall not endorse it. Let us, rather, consider some of the other possible welfare-relevant areas of uncertainty beyond sentience that may be accounted for under this framework.
5.1.2.2 Qualitative Uncertainty
At a first glance, claims made to equally consider the welfare of different animals appear to value their interests equally (assuming other welfare-relevant factors which we have considered remain equal). Given our earlier example considering a dog and a fruit fly, assuming that the evidence contributing to our probability value for affective sentience (Sx) totals 0.85 in both cases, it may seem as if we have no good reason for morally discounting the pain state of the fly to the dog (since in both cases we can identify the same categorical interest). But more than just being uncertain of the presence of affective capacities in either species we are also uncertain as to the extent of their qualitative states, since emotional experiences carry comparable degrees of intensity. The mere fact that receiving a paper cut on your finger and having that same finger amputated are both states of pain does not warrant equal consideration between these two cases. We must likewise assume that for members belonging to different species, with considerable differences in their qualitative experiences resulting from differences in physical configuration, the pain resulting from, say, a crushed leg is not measurably equal such that equal consideration is warranted.
Because qualitative differences in the experience of welfare states can only be fully realised subjectively, the process of making inferences according to objective facts about different organisms (behavioural responses, physical-chemical systems etc…) is considerably more difficult and disputable than making mere inferences on the presence of such affective capacities to begin with. But this does not mean that such inferences cannot be made at all. Just as we take the presence of a more developed neural system in the dog as greater evidence for their being able to experience affective states of welfare, those attributes might also be used to infer richer and more intense welfare experiences which demand greater consideration in cases of moral decision-making. While it is true that neural structures do not necessarily correlate with mental experiences, evidence of the greater emotional capacities of more developed nervous systems may possibly support the inferential claim that dogs, with cerebral brain cell count of 527.91 × 106, experience negative welfare to a greater extent than fruit flies, whose brain cells number 250,000 (note that considerably more research and argumentation would be required to sufficiently back this claim) (Jardim-Messeder et al., 2017, p.6) (Shimada, 2005, p.145). One could, therefore, construct a good argument for valuing the welfare of a member of one species over a member of another by pointing out evidence of more developed welfare capacities, but since this a less precise and impactful welfare-affecting factor we ought not consider it the basis for moral judgements to be made, particularly for judgements which extend to include large groups of morally considerable beings.
5.1.3 Proposing a Welfare Biology
In all, considerable research is required to facilitate the study of wild animal welfare, which Yew-Kwang Ng (1995, p.256) terms as a ‘welfare biology’. Such as how conservation biology conducts empirical research under the tenets of ecocentrism, welfare biology assumes the consideration of affective sentients such that the study of their welfare states is justified. And likewise, the outcome of research conducted under welfare biology is not itself normative; moral prescriptions require normative premises to be derived (take for example the generalised welfare calculus that I promote in this chapter, or the other positions considered in §4.1.1) (ibid, p.260). The study of welfare in descriptive terms makes it a positive study, subject to scientific scrutiny, and the role of welfare judgements within normative theories does not mean that the two are necessarily tied (just as anthropocentric empirical research results are not necessarily tied to the normative anthropocentric prescriptions that these results may support).
It might be objected to the idea of welfare biology as a legitimate field of scientific inquiry on the basis that subjective experiences are difficult to objectively measure. Welfare judgements made on nonlinguistic animals (as discussed in §2.2) are notoriously debated, and require a less certain method of inference to, say, conservation biology (of which judgements may often be simple matters of empirical measurement. Recording and comparing the atmospheric impact of carbon dioxide versus methane per unit of emission is one example of this). But as per our functionalist criteria of mental state inference, it isn’t the case that there exists no basis whatsoever to make claims on animal welfare states. Donald Griffin (1994, p.6) argues that “the difficulty, or even the impossibility of conveying to others the exact nature of something does not rule it out of existence or deprive it of significance”, and Ng (1995, pp.258-259) furthermore adds that “[m]any accepted sciences (e.g. archaeology, cosmology, meteorology) also deal with variables difficult to measure precisely”. Hence, it seems perfectly reasonable to treat welfare as a measurable criterion for empirical investigation.
5.2 Applying these Considerations
Given the welfare-relevant factors considered in this chapter so far, a very basic calculus to estimate the net-welfare impact value of a single entity would look something like this;
(Px * Sx) + Ox
In this calculus, P represents x’s experienced welfare, which we modestly infer from their apparent interest satisfaction/dissatisfaction, but which may plausibly vary in light of new research detailing the extent of affective capacities in different entities. Note that I use the term ‘entities’ rather than ‘animals’ because S, which represents x’s probability value of affective sentience, applies to every being for which we have reasons to believe experience affective welfare states. As Chan (2011, p.327) notes, “we do not know that sentience is dependent upon particular physical structures (e.g., nerve cells), or that it cannot arise from sensory feedback processes based on hormones (which plants possess)” – it is therefore possible that plants possess sentience to some degree. In my view, however, it is extremely unlikely that plants are affectively sentient, which is accounted by the fact that we have very few reasons to infer their affective capacities, resulting in a low S value. And because we multiply plants’ S value by P for their few inferable possible welfare (ie; sunlight, water, adequate nutrition), their individual considerability value remains markedly low (subject to change given relevant new welfare research). And last, because we are promoting an act-centred approach which can be used to determine the approximate welfare impact value of any given entity, O value represents x’s outward welfare bearing on other morally considerable beings (which must take into account the Px * Sx value of all affected beings within the scope of this calculation).
Relating to our earlier example in which an adult lioness accounts for -203.34 to -3239.37 predatory welfare points over the course of her life, these numbers assume certain conditions of sentience and conservative estimations of personally experienced welfare on behalf of prey. They are thus by no means definitive and must be altered according to the relative probability values we derive from future welfare biology research. As we have already discussed, there are complications for the circumstances in which this formula can be applied, which include the difficulty of conducting welfare-relevant research on wild animals and in attempting to measure and compare values that are subjectively experienced, and which are sensitive to variable factors (like those specific to geographical location, which include competition for natural resources, prey saturation…etc). However, the extent to which large scale suffering can be observed as a result of natural processes – even when only taking into account very basic welfare-relevant factors (like clutch size, cause of death…etc) – means that welfare biology it is a positive study, separate to the normative conclusions which can later be derived from its findings.
It is also important to emphasise exactly how much welfare biology research will be required to even generally determine these values, given the number of and causal relations between contributing natural processes and entities. For calculations regarding specific acts like predation this may be easier to manage, but when calculating the welfare impact resulting from x’s average lifespan one must try and predict all of the welfare-relevant habits associated with their average existence. We have already described the high mortality rates of lion cubs, but as per §5.1.2 it remains to be seen whether an act as arbitrary as viviparous birth fundamentally alters our probability value of an animal’s sentience, such that we would grant the same affective capacities to a new-born as to a fully developed adult member or their species. Lion cubs in particular are noted to be extremely precocial at and around the time of their birth, appearing “lifeless for three days, until by the roaring of their father and by breathing in their face they are quickened. […] [W]ithout controversy they are senseless for a certain space after their whelping” (Sunquist & Sunquist, 2002, pp.295-296).
Another matter more difficult to deduce relates to the welfare impact resulting from the presence of wild animal populations (as will be required for large-scale interventionist undertakings). For this there are a multitude of factors that must be included within a relevant welfare calculation, which include the numbers and welfare values of those animals that are directly affected by population x, the indirect ecosystemic impact that may predictably result from x’s intervention (and which bears upon the lives of future sentient beings to exist), among many others which have not yet been determined. It is for these reasons that the promotion of welfare biology as a legitimate field of scientific inquiry is essential, so that in the long term we are better able to determine the nature of our moral obligations towards wild animals and act surely on their behalf.
5.2.1 Considering Extinctionism
It is no surprise that as a solution to the problem of wild animal suffering, some (Vinding, 2016, p.4., Studebaker, 2013., and myself; Soryl, 2017) have considered the position of extinctionism – that of wildlife anti-natalism – as a possible method of ensuring the discontinuation of suffering-predominant lives. By merely not reproducing, a single adult lioness can be expected to prevent her male cub from facing 7/8 odds of not surviving into adulthood and a female chinook salmon can prevent her few thousand-sentient offspring from a survival rate of <1/100 (these numbers progressively worsen as we consider the high fecundity figures of §3.2.2) (Vinding, 2016, pp.5-6). Magnus Vinding goes so far as to argue that by failing to consider the lives of wild animals who face these dismal rates of survival into adulthood, we (humans) are guilty of committing speciesism, as such rates would not be deemed acceptable for members of our own species. And to deny that it would not be better for children to be born facing 7/8 or <1/100 odds of surviving into adulthood, where the alternative is a painful death, one does “not tak[e] seriously the victims whom these parents are sure to create” (ibid, p.8).
Because nonhumans cannot be reasoned with to ignore their own reproductive urges, the extinctionist position holds that humans (moral agents) must act on behalf of the welfare of wild animals that have yet to be born (moral patients) by preventing their existence. On a population-considering scale, this reasoning demands that we actively work to prevent the thriving of species with disvaluable reproductive strategies, which include most wild animals. In continuation to (C2) of my thesis argument so far – ‘we ought to intervene in natural processes to improve wild animal welfare’ – the case for extinctionism is thus presented as follows;
(t)P4: ‘Species extinction is the most impactful course of action to reconcile the overwhelming prevalence of negative welfare over positive welfare in nature’
(t)C3: ‘We are morally required to advocate for species extinction’
(where ‘t’ represents ‘tentative’)
It is first worth noting that (P4) is for the time being uncertain given our current lack of knowledge in and epistemic limitations about welfare-related ecosystemic analysis, which may be overcome given sufficient welfare biology research in the future. Because of this, we cannot yet know if (C3) can be theoretically derived from the reasoning we have so far presented in this thesis. But even assuming (P4) to be the case, we have reasons for rejecting (C3) as a pragmatic consequence of our argument given its practical untenability. Simply, it is not the case that for an applied ethic (as we are proposing), which must take into account real-world circumstances, intentional species eradication would be permitted to occur under democratically elected social principles for welfare-specific reasons. And furthermore, the possible practical implication of (C3) under different social circumstances does not constitute a problem per se with our argument, only if one were to claim (C3) as a valid response to current circumstances, which I deny. Because of this, and for other applied reasons related to the widespread promotion and application of wild animal welfare-considering ethical principles, we are justified in practically rejecting extinctionism in favour of other moderate interventionist measures.
This applied reasoning can be likened with my objection to Horta’s rebuttal of ecocentric valuing in §4.1.2.1; assuming sound moral reasoning, it is pragmatically very unfavourable to promote the repugnant conclusions of one’s argument if practical implementation stands as one of its aims. In the case of ecocentrism, the theoretical consequences that result from many of its corresponding normative theories promote human extinction, but this by no means supports the claim that human extinction is a practical consequence of its reasoning. Vinding (2016, p.7) himself admits “the conjecture I have argued for above is mainly a theoretical rather than a practical one”, which helps to show that there is an important difference between these two arguments – a difference that must be realized for circumstances of real-world implementation. Though (C3), it should be noted, presents a theoretical case for rejecting our strictly welfare-maximising ethic if extinctionism can be shown to be a logical consequence resulting from our argumentative reasoning (practical restrictions aside).
I am not disposed to argue explicitly against alternative sentiocentric normative theories which promote wild animal welfare consideration and do not result in extinctionism (as put forward in §4.1.1, these positions do not differ significantly in their pragmatic prescriptions so as to promote a single approach above others). Rather, in response to this seeming objection I would remind the reader of my reasoning in §4.2 to reject reductio arguments as always-valid indicators of argumentative inconsistency. Simply, if (P4) were proven to be the case, the immediate repugnance of (C3) does not present us with a case (at all) to reject our reasoning simply because we find it unfavourable for ourselves. To say otherwise would unjustifiably value human interests (ie; our interests to reject extinctionism) over the interests of the other animals our argument claims to consider (we are, after all, presenting a sentiocentric case for moral consideration rather than one that is anthropocentric). Acknowledging this as a possible theoretical outcome of my argument is sometimes called ‘biting the bullet’ – I disagree. If we promote non-anthropocentric reasoning then there is no bullet to bite; any derived outcome of our axiology can, of course, be disputed for various normative reasons, but not for reasons related to our view of what is considerable (unless one is willing to supplant another axiological theory in its place). In any case, as we cannot yet make informed estimations on whether or not extinctionism even is a theoretical result of our welfare-maximising reasoning, we ought to withhold judgement.
5.3 Conclusion
This chapter proposes an approach to wild animal welfare consideration, one that takes into account certain concerns that can be consulted when determining how to best intervene in nature to benefit wild animal welfare. These concerns include our epistemic uncertainty of other animals being affectively sentient and the qualitative extent of their experienced welfare, welfare impacts resulting from some of the biologically determined habits of wild animals, and last, variable public attitudes concerning welfare-directed wildlife intervention (as is required to consider for the proposal of an applied ethic). As only the last of these affects the normative prescriptions that are derived from our welfare impact calculus, we ought to rather consider it as a general measure for assessing possible methods of wildlife intervention than as a factor directly relevant to moral acts themselves – which I shall expand upon in the next chapter.
6 Practical Intervention
Given that we have now established a framework for considering the welfare of wild animals (taking into special account their circumstances in nature), the aim of this chapter shall be to contemplate practical methods of wildlife intervention on their behalf. To begin, in §6.1 I shall discuss in greater detail the practical considerations that are relevant to my proposal of an applied animal ethic, which requires that I review some of the currently employed practices of ecocentric and anthropocentric wildlife intervention to generally determine their estimated welfare impact. In response, §6.2 shall consider the proposed future method of gene drive intervention – by which wild animal populations may be engineered to modify, suppress, or eliminate predetermined genetic traits – as an alternative to the practices considered in §6.1. Given the welfare implications of this technology, it would be beneficial to consider the role of welfare biology within its current early stages of research, and to promote consideration of animal welfare for speculated future interventionist measures. And last, §6.3 shall shortly reflect upon the role of animal welfare within current and future scientific research.
6.1 Considering Pre-existing Methods of Wildlife Intervention
As we discussed in §4.1.2, two chiefly considered valuing systems are ecocentrism and anthropocentrism, which are reflected by most public attitudes toward environmental modification. It is, indeed, often assumed that only these two positions have something to say about how we act towards the environment, whether we see it as a means to our own ends or as an end in itself. Or, theories which value sentient beings individually sometimes assume that a natural life is a good one, leading them to adopts a pseudo-ecocentric position on a similar basis to anthropocentric environmentalism (note that as per our reasoning in chapter three, this position is unjustified). And though our argument in chapter four has been against non-sentiocentric systems of environmental intervention, its realistic result is not going to be a widespread change of democratically adopted social principles and interventionist methods (just as how P4 is irrelevant to the proposal of an applied ethic because of its practically implausible consequences). Thus, rather than attempting to make moral prescriptions independent to the dominant ecocentric and anthropocentric positions of wildlife intervention, the welfare-advocating position ought to try and navigate these theories by recognising their presence as serious drivers of natural intervention, and possibly align with them when it is possible to predict that the welfare consequences justify doing so.
Consider as an analogy the arguments that are made in favour of act versus rule consequentialism. While the former applies to isolated cases of moral decision making and requires, sometimes overwhelmingly, consideration for every possible relevant outcome of a particular act, rule-based consequentialism recognises that humans are not always perfect moral agents able to determine moral obligations on a case-by-case basis and proposes general guidelines which seek to better overall outcomes by applying to every moral agent. Rule consequentialism therefore considers and manages real-world limitations which in practice affect the predicted net-outcome of all cases considered, and is thus arguably a stronger applied ethic than its counterpart. If we likewise acknowledge the prevalence of ecocentrism and anthropocentrism in public opinions of environmental intervention, a better applied ethic would attempt to circumstantially reconcile sentiocentrism with these competing views if doing so would result in better short and long-term net-welfare outcomes.
Furthermore, the danger of ignoring pre-existing public opinions about topics as debated as environmental intervention give us good reasons not to approach it without heed. Intervention made on the sole basis of welfare-maximisation may sometimes be contrary to ecocentric aims (for example; reducing the biodiversity of extreme r-strategists if the species is considered endangered or native to a certain region), so it is likely that unheeded attempts will negatively affect public opinions related to the role of sentiocentrism in wildlife management. We can assume that this will push back or delegitimize future welfare-guided intervention, which will have a tangible long-term effect on net-considered welfare. An instructive example concerns the 1997 illegal introduction of the rabbit haemorrhagic disease into the New Zealand ecosystem, in which a group of farmers smuggled the calicivirus from its escaped quarantine in Australia and spread it to target a large area of the South Island in order to reduce rabbit population density (O’Hara, 2006, pp.120-121). The disease is now endemic to New Zealand, and despite the continued anthropocentric benefits which result from its prevalence within existing wild rabbit populations, the unconstitutional nature of its release has been considered by many to be an act of bioterrorism (a label which carries strong legislative penalties). Thus, an outward respect for current practices and beliefs concerning intervention should be included within an applied sentiocentric ethic.
6.1.1 Ecocentric Intervention
Current methods of ecocentric intervention with welfare impacts include most notably pest control, in which non-native animals are routinely culled in order to reduce their population size (and thus, to help preserve their containing ecosystem). In New Zealand, considered to be a leading country in biosecurity (though also one whose ecosystem has been significantly affected by non-native animals), nearly 50 resident vertebrate species are officially listed as pests (Esvelt & Gemmell, 2017, p.1) (Littin et al., 2004, p.3). These animals, ranked in terms of their economic and conservational costs, include possums, stoats & ferrets, rats, among other less common species like “rabbits, feral pigs, feral goats, deer and hares” (ibid, p.4). Methods of eradication with their associated welfare issues include the following;
Figure 1: From ‘Animal welfare and ethical issues in island pest eradication’ in Island Invasives: Eradication and Management. Proceedings of the International Conference on Island Invasives (Cowan & Warburton, 2011, p.419)
Non-lethal trapping methods like cage and leg traps are possibly very suffering intensive, despite not causing immediate death. Box traps may be used to catch several animals at once, but Mason and Litten (2003, p.17) note that they sometimes result in stress-induced cannibalism in the time before the traps are checked and its containing animals are either released off-site or destroyed. The time it takes for live-traps to be checked may vary – leg hold traps “[b]y [New Zealand] legislation […] must be checked within 12 hours of sunrise on the day after they are set”, and in the UK, sticky board traps holding rodents require that they are checked at least once per day (DOC, 2004, p.48). But live-traps are not always so routinely checked, and even in the cases that they are, the fear and trauma resulting from their panic for escape may lead animals to undergo acts of self-mutilation (Mason & Litten, 2003, pp.16-17).
Kill-traps are sometimes used due to their perceived added humaneness and reduced labour requirement, but they are noted as being less capture-efficient and more cumbersome to set up than their live-counterparts (DOC, 2004, p.48). New Zealand regulations permitting the use of snap-traps, for example, “require traps to cause a loss of palpebral reflexes in under three minutes” – from this we can assume that for effective kills, animal suffering only lasts a mere few minutes (Mason & Litten, 2003, p18). But this assumption only holds for cases in which the trap functions as intended which does not always happen, as there is evidence to show that anywhere between 5% and 14% of caught animals remain alive and injured for a substantial period of time (as they are not regularly checked, injured animals may spend days or even weeks suffering). Electrocution traps are likewise believed to be ‘intensely painful’, though it is unclear for how long suffering is experienced in these cases before death. In all, significantly more research is needed to determine the welfare-relevant features which would allow us to compare trapping-induced suffering to the suffering that these animals would otherwise experience in the wild, such that a welfare-impact calculation can be made to justify their use.
Biological control is another regularly used method of pest-control (as seen in our earlier example of rabbit haemorrhagic disease), and it is one that is certainly very disvaluable for its affected animals. Similar to our analysis of trapping, species and disease-specific research would be required to make informed claims on the justifiability of these methods, which include the use of salmonella, myxoma, and cholera to target specific animals (Gherardi & Angiolini, 2007, p.15). In addition, biocides are used to control large numbers of vertebrate animals through poisoning with bait being distributed from the ground when the targeted terrain is easily traversable, otherwise spread aerially (DOC, 2004, p.32). Of the fifteen poisons that are currently registered for use against mammal pests in New Zealand, sodium monofluoroacetate (‘1080’) is by far the most used, accounting for 80% of the product’s global demand (ibid, p.24) (PCE, 2011, p.22). Although 1080 does not present the most serious welfare concerns when compared with other previously applied poisons (strychinine and phosphorous), welfare research on its use shows us that retching, vomiting, spasms and seizures are common for affected possums – the primary target for its distribution in New Zealand (Littin et al., 2009, p.716). It is furthermore speculated that they, like humans, experience epigastric pain and possibly hypothermia before falling unconscious and, after approximately 11.5 hours of consuming poisoned baits, dying (through interference of their TCA cycle causing severe energy deprivation and a build-up of cirate) (ibid, p.710, 714 & 717).
6.1.2 Anthropocentric Intervention
Some of the methods of ecocentric pest control that were considered in the last section also apply in cases of anthropocentric intervention; trapping and hunting, for example, may occur for reasons that are both anthropocentric and ecocentric. And possum culling occurs for reasons that possums endanger native New Zealand wildlife, but also because they are the primary carriers of TB (which, if left unchecked, would result in an estimated loss of $5 billion (NZD) in agricultural productivity over the course of ten years) (DOC, 2004, p.16). There are also cases in which anthropocentric and ecocentric reasoning for intervention are at odds with each other; deforestation and overfishing are two examples of disagreement. But the methods we have considered so far chiefly apply to mammals, which as we have already established are far fewer in number to other animal groupings. Although pre-existing methods of intervention for non-mammals – particularly non-vertebrates – occur far less frequently, there are some cases where animals like insects interfere with agricultural productivity, or act as carriers for human-affecting diseases, where intervention takes place. In such cases insecticides are often used as they are generally cheap and effective, though from my analysis of the literature there exists little research on its associated welfare issues.
Autocidal control methods are also particularly effectual in this regard; sterile insect technique, for example, has often been used to promote infertility among populations of targeted insects (Gherardi & Angiolini, 2007, p.15). This method “relies on the mass rearing, sterilization, and release of large numbers of [male] insects” to artificially reduce the reproductive fitness of a given population, which has led to the successful local eradications of the screwworm fly, the fruit fly, and the tsetse fly in certain areas (Thomas et al., 2000, p.2474). Sterile insect technique also poses as a more ethical alternative to traditional use of insecticides, though this fact remains unclear given the ways in which sterile insects are bred, which are mass-rearing facilities in which females are prematurely culled and the males exposed to harmful levels of radiation to promote their reproductive degradation. Given that this method tends to maximise the number of insect lives that are brought into existence for release, and because insect releases are periodical (thus involving the rearing of many insect generations whose lives contain little more than radiation therapy and basic nourishment – if that), it seems unlikely that sterile insect technique as we have described presents a particularly welfare-considering case for insect population control.
Recent developments in genetic engineering have also meant that, rather than reducing the reproductive potential of affected males, artificially induced self-limiting genes leading to the premature death of offspring have been made able to spread throughout insect populations (Braverman, 2017, p.16). Although these systems have been proven more efficient than sterile-based systems, they plausibly involve more suffering by promoting the short and stunted growth of insect offspring (rather than merely promoting infertility). Though as in many of the cases which we have considered, further welfare biology research is required to make informed claims on the justifiability of one method over another. We can, however, speculate on the welfare impact resulting from the application of possible future biocontrol methods that are still in development. Namely; gene drive systems, which have been proposed as a plausible alternative to the methods of wild animal population control that we have so far considered in this thesis.
6.2 Gene Drives: A Nonlethal Solution to Wildlife Management?
Despite the focus of most research on gene drives, their spread within populations is a naturally occurring phenomenon in which certain genes evolve “better-than-even [chances] of being inherited” (transmission rates have been demonstrated to artificially exceed 90% – see Alphey, 2016, p.149) (Esvelt, 2017, p.21). These systems are unique in that they bypass the traditional rules of Mendelian inheritance in which a gene has a 50% chance of being passed to an offspring, and are thus not immediately subject to the forces of natural selection that would otherwise remove their presence from the gene pool (Champer et al., 2016, pp.146). Gene drive systems can infiltrate populations despite the possible reproductive disadvantage that results from their spread, and the spread of their carried edited gene. Within the field of wildlife management and natural intervention, this makes it possible to edit the gene pool of entire animal populations, possibly globally (assuming that there is a sufficient genetic familiarity between members of this grouping).
Figure 2: From ‘Rules for Sculpting Evolution’ in Gene Editing, Law, and the Environment: Life Beyond the Human (Esvelt, 2017, p.22)
Gene drives can be identified under two categories; modification drives, which can be used to change the phenotype of a particular species (for example; a gene drive may be used to favour the spread of ‘blue’ eye colour within the gene pool of a given population), and suppression drives, which are capable of reducing or eliminating certain animal populations (ibid, pp.147-148). Most notable of the available drive systems are homing based drives, which can fulfil either role by severing a selected section of the corresponding allele (in the opposite chromosome to the one carrying the drive system), which causes the cell to use the DNA of the introduced gene as a template for its subsequent repair. When introduced into the germline (reproductive) cells of an organism, homing based systems such as CRISPR Cas-9 can remove and replace segments of DNA with artificially engineered segments to bias gene transmission to their offspring, which continues until either a mutation has evolved to stop its spread, or the gene drive is manually halted via external intervention (Esvelt, 2017, p.21 & 23). Ordinary gene drive systems it is worth noting “are not permanent on an evolutionary timescale”, nor even in the short term depending on the fitness disadvantage of the introduced gene (Esvelt et al., 2014, p.9). Thus, the risk of accidental harm resulting from the use of a gene drive is considerably smaller than we might initially think.
Other variations of gene drives include the Medea and traditional Underdominance systems, each of which carries its own associated advantages and disadvantages (Champer et al., 2016, p.157). Given its greater efficiency, effectivity, and precision, however, the CRISPR system is currently the most researched gene drive system for public release as it has already been implemented in four unique species; yeast, fruit flies, and two species of mosquitoes (Braverman, 2017, p.16). And furthermore, localisable drives – drives which are able to be limited to selected environments (and which are noted to be the only gene drive systems likely to receive public release – see Esvelt, 2017, p.32) – are already in development.
Daisy drives are an example of this, which are created by splitting the components of a singular gene drive into a number of serially dependant drive systems, introducing a kind of genetic ‘expiry date’ on the implemented gene, which can be modified according to the purpose of the drive. (ibid, pp.31-32). The downside of these solutions is that they may not result in a species’ complete genetic restoration (as the edited gene remains suppressed in the DNA of the species). However, they do allow for complete phenotypical restoration to take place, which for practical purposes can be said to achieve the same outcome (just as two identical cakes with the same ingredients may each contain slight variations in their recipes). Although this may seem like a relatively inconsequential trade off given the scope and power of gene drives, current efforts in research aim at engineering daisy drives which operate with underdominance ‘quorum’ systems, possibly allowing for locally confined population modification to take place without leaving behind a genetic ‘footprint’, which some publics might otherwise oppose (Min et al., 2017, p.8).
6.2.1 Modification Drives
Considering the spread of gene drives among wild animal populations, modification drives seem like the most plausible system that can be used to immediately benefit animal welfare. Earlier, in §3.2.1, we considered a wide range of diseases and afflictions with significant animal welfare impacts, including TB, for which there also exist strong anthropocentric reasons to eradicate the disease (given its estimated cost of $60 million (NZD) per annum to control and its longstanding impact on animal agriculture – Norton, 2001, p.93). Research shows us that it is possible to engineer TB-resistant genes specific to cattle (SP110), giving us reason to suspect that by conducting similar pathology-related genetic research into possums, a similar gene may be found and then possibly paired with a gene drive system to spread TB-resistance within possum populations (Tuggle & Waters, 2015, p.3854).
For ecocentric intervention, targeted gene flow has been proposed as a method to oppose the spread a contagious strain of facial cancer among populations of Tasmanian devils (Kelly & Phillips, 2016, p.261). The method suggests taking and rearing individuals from a naturally occurring disease-free population to introduce into the wider gene pool of disease-affected devils, assisting the flow of cancer-resistant alleles (as would otherwise occur under regular circumstances of natural selection). Similarly, spread of the amphibian-infecting fungus ‘chytridiomycosis’ may possibly be counteracted by assisting the flow of genes that are resistant to specific strains of the disease (ibid, pp.261-262). In each of these cases gene drives may be used to considerably accelerate the process of disease eradication by promoting ‘genetic immunisation’ for animals susceptible to infection (Champer et al., 2016, p.147). And other than the obvious ecocentric benefits which result from this form of intervention, we can expect that animal welfare will be significantly improved.
Far more speculative and distant, we might also consider the possible implications of modifying animal genes to directly improve their welfare. On the assumption that positive and negative welfare states are no more than tools which benefit the transmission of genes (which we tentatively accept), future research may uncover genetic bases for, say, pain experience in animals, or other such welfare states that have evolved to benefit their fitness. Though seemingly farfetched, research in humans shows that qualitative accounts of pain intensity correspond to the presence or absence of the gene ‘SCN9A’, “provid[ing] additional evidence that human pain experience is under genetic influence” (Reimann, 2010, p.5151). Likewise, recent identification of a mutation within the ZFHX2 gene which causes congenital perceptual insensitivity to harmful stimuli in humans has been replicated in transgenic mice, who demonstrate a likened reduced sensitivity (though not to the same extent in humans due to the specific circumstances in which genes function for different organisms) (Habib et al., 2018, pp.373-374). Nonetheless, this research shows us that it is at least possible to conduct welfare research on animals, leaving the door open to possible future gene drive intervention to affect wild animal welfare capacities.
Another speculated use of gene drives to improve wild animal welfare that has been proposed by Kyle Johannsen (2016, p.341) argues to reduce the clutch size of selected primarily r-reproducing animals, such that their offspring numbers are fewer and thus, less suffering can be expected to occur from wild animal reproduction. While this is a plausible application of CRISPR-based gene drives (given sufficient research on the genes that are associated with reproduction), an immediate objection to this reasoning is that the habits of different animals have been fine tuned to the circumstances of their evolution. By disrupting these habits, a species may accidentally become extinct by not producing enough offspring to counteract their environmental pressures, which in turn may lead to the extinction or increased suffering of dependant predators. Johannsen, a rights-based theorist, responds by suggesting the possible engineering of plants to obtain the same nutritional properties as a predator’s otherwise would-be prey; “[g]iving predators an alternative food supply would reduce the extent to which they must rely upon hunting r-strategist infants, thereby helping to ensure that (former) r-strategists remain well adapted after being genetically modified” (ibid, p.342).
While it is possible that genetically reducing the general fecundity values of highly r-selected animals may reduce their net-experienced suffering, the argument to create plant-based predators seems strategically unlikely for two reasons. First, it does not take into account the role of ecocentrism in wildlife management practices, which (as per our reasoning in §6.1) ought to be considered when proposing realistic methods of intervention. And secondly, Johannsen’s argument oversimplifies the nature of predator-prey relationships. Merely making plants available to predators that are sufficient to satisfy their nutritional requirements by no means guarantees that those predators will be disposed to eat them. Because behavioural dispositions tend to be the result of natural selection, we ought to assume, instead, that predators with millions of years of evolutionary conditioning to respond to specific phenomena (for example, a mouse running across an open field) will not realise plants as a food source (unless those same predators are able to flexibly alter their behaviour, in which case a change of food habits may occur on an individual basis, though cannot be expected to supplant predation for entire populations of predators).
There is furthermore little evidence that such plants could be developed for all or even for some predators to function as a viable food source equal to or greater than prey animals. And supplementing gene drives to change the food habits of carnivores, as Johannsen posits, does not take into serious account the significant diet-related physiological differences that exist between carnivores and herbivores. These include “differences in digestive enzymes, shortened digestive tracts, changes in amino acid dietary requirements, and alterations to taste bud sensitivities”, among others, all of which would need to be genetically sequenced and then built into a CRISPR-based gene drive system, then after constantly maintained through regularly scheduled ‘reinforcement’ drives (to prevent the accumulation of drive-resistant mutations as would inevitably occur) (Kim et al., 2016, p.2). In all, given our current lack of understanding ecosystemic relations and their associated welfare impacts, Johannsen’s suggestion seems a tad too speculative to be considered as a practically plausible interventionist strategy – though his argument does succeed in furthering the debate about welfare-guided intervention, which ought to be encouraged. For the time being, at least, I argue to chiefly consider intervention as it relates to pre-existing methods and practices.
6.2.2 Elimination Drives
Much of the press surrounding elimination drives have described as ‘the extinction invention’ because of their possible use to spread infertility within populations, or to bias sex ratios leading to the eventual global population decline of a species (Regalado, 2016). But as per our earlier comments on daisy-drive systems, this is not necessarily the case. And even supposing that an accidental release of global elimination drive systems was to occur, “[a] single unauthorized release is exceedingly unlikely to suffice”, due to the ease at which drive resistant mutations evolve within species to prevent their spread (Esvelt, 2017, p.25). And in the case that a drive system was somehow created to overcome this obstacle, there are a variety of preventive measures one may take to halt a drive’s unintended spread (for example; engineering reversal drives to combat the edited phenotypical alteration of the original drive system) (ibid, p.23).
Assessing the welfare impacts resulting from the use of elimination drives is, like our argument in §5.2.1, currently extremely speculative and subject to future welfare biology research. But given the obvious ecocentric and anthropocentric benefits which result from its implementation in areas such as pest control, elimination drives have already been considered by many (Sandler, 2017, p.46., Esvelt & Gemmell, 2017, p.1., Hill et al., 2017) to be used in non-native population eradication as a replacement of the methods considered in §6.1.1. Given that the Royal Society Te Apārangi of the New Zealand government has already begun to consider its use in pest eradication, we can expect that at the time gene drive systems become implemented, elimination drives will become a primary focus of their use.
In a technical paper accompanying a public discussion document on the Royal Society’s website, the organisation anticipates future scenarios of gene drive intervention for pest control. Vespula wasps are another non-native species in New Zealand with negative ecocentric consequences that are proposed for future consideration over the course of gene drive research and development (RSTA, 2017, pp.6-8). And although they (currently) lack the adequate welfare biology research to be able to make definitive claims on the justifiability of locally eradicating their populations, preliminary research on their dietary habits shows that they commonly feed on arthropods like spiders, giving us basic reasons to favour their non-being (Harris & Oliver, 1993, p.10). Other such animals that are considered pests are possums, stoats, and rodents, which also have strong predatory tendencies. For now, though, further research is required to determine the general welfare-relevant habits of these animals such that we are able to conduct a basic welfare-impact calculation for their eradication.
Ecocentric cases which we can otherwise already advocate for with relative confidence include elimination of the New World screwworm and the significant reduction of desert locusts to prevent swarming (which are considered as potential targets for the use of daisy quorum drives – see Min et al., 2017, p.3). In these cases, it is apparent that the estimated welfare impacts resulting from gene drive intervention are welfare-positive, given the suffering-intensive survival habits of the screwworm (which, like the Australian sheep blowfly endemic to New Zealand, include eating and infesting the flesh of live animals – see RSTA, 2017, p.8), and the reproductive disvalue caused by the rampant spawning of desert locusts. Causes like these with clear animal welfare benefits ought to be considered early on in research, to promote their genetic sequencing and serious consideration for intervention if and when such drive systems become publicly adopted.
6.3 Future Directions in Wildlife Intervention
It is clear from our thesis argument so far that significantly more welfare biology research is needed to determine the justifiability of differing interventionist strategies, and also that such research has strong legislative consequences with more immediate welfare benefits. In a report commissioned by a Technical Expert Working Group (2003, p.6) to determine the scope of the European Union directive on the protection of animals used for scientific purposes, they propose the following;
“Inclusion of any invertebrate species in this Annex should only occur on the basis of sound scientific evidence as to their sentience and ability to feel pain, as assessed by a Scientific Committee of experts appointed by the EC […] Insufficient evidence is available at the present time to consider the inclusion of any invertebrate species other than cephalopods.”
Thus, welfare biology research ought to be considered a primary focus for wild animal welfare advocates.
It is also worth noting that discussion of wild animal welfare in relation to gene drives – particularly at this early stage in the technology’s research and development – may help to promote its future consideration alongside ecocentric and anthropocentric views for targeted areas of implementation. Somewhat interestingly, Kevin Esvelt – a leading gene drive researcher and one who is partially responsible for the creation of gene drive technologies – expresses his concern for wild animal welfare in a recent interview, in which he claims “[e]xistence in the wild is basically unmitigated pain and suffering” (Braverman, 2017, p.8). His views toward natural intervention are to “get the ball rolling” on gene drive research, to later facilitate discussion on attitudes towards nature and the welfare of wild animals, which he hopes to include in the future of gene drives. This view is reflected in the arguments expressed in his own publications, which call for ‘open, public-consulting’ gene drive research given their high controvertibility value, and thus, high chances of future delays in implementation resulting from negative public perception of their study ‘behind closed doors’ (Esvelt, 2017, pp.28-30). Esvelt’s strategy looks to appeal to the long-term promotion of welfare consideration within practices of wildlife management which, given current rare circumstances of wild animal consideration, seems like a positive contribution to wild animal welfare.
6.4 Conclusion
Foremost, this chapter emphasises the requirement for further welfare biology research to benefit animal welfare consideration in both the short and the long term. We continued some of our discussion in chapter five by considering the role of pre-existing attitudes towards wildlife management as tangible barriers against welfare-guided intervention, and analysed some of the specific methods that stem from these attitudes. Then we speculated upon possible future applications of biotechnologies like gene drives as an alternative to current methods of wildlife intervention to benefit wild animal welfare, even if they are intended for ecocentric or anthropocentric aims. We also considered possible methods of welfare-intended intervention which may prove fruitful in the long-term.
7. Conclusion
In closing, my thesis has argued that wild animals are equally legitimate recipients of our moral consideration to non-wild animals, and has investigated the role of applied ethical judgements to practically consider their welfare.
In chapter two of this thesis I defended the functionalist method of mental state inference and examined relevant empirical research to establish the affective welfare capacities of small animals which are not often considered sentient. This led to my argument in chapter three, where I investigated the circumstances in nature which cause wild animals to suffer, which are the limited availability of natural resources and evolutionary adaptions that have arisen in response. In chapter four I examined relevant axiological and normative theories that consider animals who suffer in the wild, defending the view of affective sentience as the only morally defensible criterion for consideration. As per my proposed act-based framework for normative judgement, chapter five included several welfare considerations that ought to be accounted for under circumstances of practical wild animal intervention. Chapter six concluded my argument by investigating the welfare impact of certain current and speculated future methods of wildlife intervention (following from chapter five’s recommendations), including suggestions for future research.
The most important take-away from the research I have presented in this thesis is that significantly more descriptive and empirical research is required in the areas that I have addressed to best achieve long-term benefits in animal welfare. Only after establishing wild animal welfare evaluation as a positive science will we have access to the requisite information to make justified and informed prescriptions for wildlife intervention on behalf of animal welfare. It is possible, otherwise, that much of the current strategic discussion on wild animal welfare consideration will be rendered irrelevant in lieu of future welfare biology research. Thus, to achieve better net-welfare consequences in the future, it is essential that we promote welfare biology chiefly as a descriptive study, thereafter as a normative study, rather than conflating the two. Simply, “[i]t is long overdue that biologists look more at welfare than at fitness” (Ng, 1995, p.274).
8. Bibliography
Animal Ethics, 2016, Population dynamics and animal suffering. Available from: http://www.animal-ethics.org/population-dynamics-animal-suffering/#sdfootnote4sym [7 May 2018]
Barkham, P., 2018, ‘Dutch rewilding experiment sparks backlash as thousands of animals starve’, The Guardian, 27 April. Available from: https://www.theguardian.com/environment/2018/apr/27/dutch-rewilding-experiment-backfires-as-thousands-of-animals-starve [4 May 2018].
Baum, E.T. & Meister, A.L., 1971, ‘Fecundity of Atlantic Salmon (Salmo salar) from Two Maine Rivers’, Journal of the Fisheries Research Board of Canada, vol. 28, no. 5, pp. 764-767.
Bermond, B., 2003, ‘The Myth of Animal Suffering’ in Armstrong, S. & Botzler, R., (eds), The Animal Ethics Reader, pp. 79-85. Routledge: London.
Birch, J., 2017, ‘Animal sentience and the precautionary principle’, Animal Sentience, vol. 2, pp. 1-16.
Bradshaw, R.H., 1998, ‘Consciousness in non-human animals: adopting the precautionary principle’, Journal of Consciousness Studies, vol. 5, no. 1, pp. 108-114.
Braverman, I., 2017, ‘Gene Drives, Nature, Governance: An Ethnographic Perspective’, in Gene Editing, Law, and the Environment : Life Beyond the Human. University at Buffalo School of Law Legal Studies Research Paper No. 2017-006. Available from: https://ssrn.com/abstract=3032607 [6 June 2018].
Brueland, H., 1995, ‘Highest Lifetime Fecundity’, in University of Florida Book of Insect Records, pp. 41-43.
Buckner, C., 2013, ‘Morgan’s Canon, meet Hume’s Dictum: avoiding anthropofabulation in cross-species comparisons’, Biology & Philosophy, vol. 28, no. 5, pp. 853–871.
Bueno-Gomez, N., 2017, ‘Conceptualizing suffering and pain’, Philosophy, Ethics, and Humanities in Medicine, vol. 12, no. 7, pp. 1-11.
Callicott, J.B., 1980, ‘Animal Liberation: a Triangular Affair’, Environmental Ethics, vol. 2, no. 4, pp. 311-338.
Carter, A., 2005, ‘Animals, Pain and Morality’, Journal of Applied Philosophy, vol. 22, no. 1, pp. 17-22.
Center for Food Security & Public Health [CFSPH]., 2018, Animal Disease Information. Available from: http://www.cfsph.iastate.edu/DiseaseInfo/ [7 May 2018].
Champer, J., Buchman, A. & Akbari, O.S., 2016, ‘Cheating evolution: engineering gene drives to manipulate the fate of wild populations’, Nature, vol. 17, pp. 146-159.
Chan, K., 2011, ‘Ethical Extensionism under Uncertainty of Sentience: Duties to Non-Human Organisms without Drawing a Line’, Environmental Values, vol. 20, no. 3, pp. 323-346.
Clayton, N. S., 2017, ‘Episodic-Like Memory and Mental Time Travel in Animals’ in J Call, (ed), APA Handbook of Comparative Psychology: Vol.2, pp. 227-243. American Psychological Association: Washington.
Corballis, M., 2013, ‘The wandering rat: response to Suddendorf’, Trends in Cognitive Sciences, vol. 17, no. 4, p. 152.
Corner, L. A. L., 2001, Bovine Tuberculosis in Brushtail Possums (Trichosurus vulpecula): Studies on Vaccination, Experimental Infection, and Disease Transmission, Ph.D thesis, Michigan Bovine Tuberculosis Bibliography and Database. Available from: http://digitalcommons.unl.edu/michbovinetb/16 [27 April 2018]
Cowan, P. & Warburton, B., 2011, ‘Animal welfare and ethical issues in island pest eradication’, in Veitch, C.R., Clout, M.N. & Towns, D.R., (eds), Island Invasives: Eradication and Management. Proceedings of the International Conference on Island Invasives, IUCN: Gland, Switzerland., CBB: Auckland, New Zealand. pp. 418-421.
Cunha, L.C., 2015, ‘If Natural Entities Have Intrinsic Value, Should We Then Abstain from Helping Animals Who Are Victims of Natural Processes?’, Relations: Beyond Anthropocentrism, vol. 3, no. 1, pp. 51-63.
Darwin, C., 1860, Darwin Correspondence Project. Available from: https://www.darwinproject.ac.uk/letter/DCP-LETT-2814.xml [17 May 2018].
de Plater, G.M., Milburn, P.J. & Martin R.L., 2000, ‘Venom From the Platypus, Ornithorhynchus anatinus, Induces a Calcium-Dependent Current in Cultured Dorsal Root Ganglion Cells’, Journal of Neurophysiology, Vol. 85, No. 3, pp. 1340-1345.
Department of Conservation [DOC]., 2004, The Use of 1080 for Pest Control. Available from: http://www.doc.govt.nz/nature/pests-and-threats/methods-of-control/1080-poison-for-pest-control/the-use-of-1080-for-pest-control/ [10 June 2018].
Directive 2010/63/EU of the European Parliament and of the Council of 22 September 2010 on the protection of animals used for scientific purposes, 2010, Official Journal of the European Union.
Donaldson, S. & Kymlicka, W., 2011, Zoopolis: A Political Theory of Animal Rights. Oxford University Press: New York.
Droege, P & Braithwaite, VA., 2014, ‘A Framework for Investigating Animal Consciousness’, in Lee, G., Illes, J. & Ohl, F., (eds), Ethical Issues in Behavioral Neuroscience, pp 79-98. Springer: London.
Elwood, R. W., 2011, ‘Pain and Suffering in Invertebrates?’, ILAR Journal, vol. 52, no. 2, pp. 175-184.
Esvelt, K.M., 2017, ‘Rules for Sculpting Ecosystems: Gene Drives and Responsive Science’, in Braverman, I., (ed), Gene Editing, Law, and the Environment : Life Beyond the Human, pp. 21-38. Taylor and Francis: London.
Esvelt, K.M., Smidler, A.L., Catteruccia, F. & Church, G.M., 2014, ‘Emerging Technology: Concerning RNA-guided gene drives for the alteration of wild populations’, eLife. Available from: https://doi.org/10.7554/eLife.03401 [9 June 2018].
European Food Safety Authority [EFSA]., 2005, Aspects of the Biology and Welfare of Animals Used for Experimental and Other Scientific Purposes, Animal and Welfare Scientific Panel. Available from: http://ec.europa.eu/environment/chemicals/lab_animals/pdf/efsa_scientificreport.pdf [12 April 2018].
Faria, C., 2015, ‘Disentangling Obligations of Assistance. A Reply to Clare Palmer’, Relations: Beyond Anthropocentrism, vol. 3, no. 2, pp. 211-218.
Feinberg, J., 1974, ‘The Rights of Animals and Unborn Generations’, in Blackstone, W.T. (ed), Philosophy & Environmental Crisis, pp.43-68. The University of Georgia Press: Georgia.
Gallagher, A.J., Staaterman, E. & Dreyer, N., 2015, ‘Kelp gulls prey on the eyes of juvenile Cape fur seals in Namibia’, African Journal of Marine Science, vol. 37, no. 3, pp. 1-4.
Gherardi, F. & Angiolini, C., 2004, ‘Eradication and control of invasive species’, EOLSS /UNESCO. Available from: http://www.eolss.net/ebooks/sample%20chapters/c12/e1-67-08.pdf [11 June 2018].
Griffin, D., 1994, Animal Minds. University Of Chicago Press: Illinois.
Grush, R., 2002, ‘The Philosophy of Cognitive Science’, in Machamer, P. & Silberstein, M., (eds), The Blackwell Guide to the Philosophy of Science, pp.272-289. Blackwell Publishers: Massachusetts.
Gullan, P.J. & Cranston, P.S., 2010, The Insects: An Outline of Entomology, 4th edn. Wiley Blackwell: West Sussex.
Habib, A.M., Matsuyama, A., Okorokov, A.L., Santana-Varela, S., Bras, J.T., Aloisi, A.M., Emery, E.C., Bogdanov, Y.D., Follenfant, M., Gossage, S.J., Gras, M., Humphrey, J., Kolesnikov, A., Le Cann, K., Li, S., Minett, M.S., Pereira, V., Ponsolles, C., Sikandar, S., Torres, J.M., Yamaoka, K., Zhao, J., Komine, Y., Yamamori, T., Maniatis, N., Panov, K.I., Houlden, H., Ramirez, J.D., Bennett, D.L., Marsili, L., Bachiocco, V., Wood, J.N. & Cox, J.J., 2018, ‘A novel human pain insensitivity disorder caused by a point mutation in ZFHX2’, Brain, vol. 141, no. 2, pp. 365-376.
Harris, R. & Oliver, E., 1993, ‘Prey Diets and Population Densities of the Wasps Vespula Vulgaris and V. Germanica in Scrubland-pasture’, New Zealand Journal of Ecology, vol. 17, no. 1, pp. 5-12.
Harrison, P., 1991, ‘Do Animals Feel Pain?’, Philosophy, vol. 66, no. 255, pp. 25-40.
Haszprunar, G & Wanninger, A., 2012, ‘Molluscs’, Current Biology, vol. 22, no. 13, pp. R510-R514.
Hill, K. (chair), Johnston, J., Byrom, A., Hackwell, K. & Hania, J., 2017, Editing Our Genes: Pest Control, radio program, RNZ, Wellington, New Zealand.
Horta, O., 2010a, ‘What is Speciesism?’, Journal of Agricultural and Environmental Ethics, vol. 23, no. 3, pp. 243-266.
Horta, O., 2010b, ‘Debunking the Idyllic View of Natural Processes: Population Dynamics and Suffering in the Wild’, Revista Iberoamericana de Estudios Utilitaristas, vol. 17, no. 1, pp. 73-90.
Horta, O., 2013, ‘Zoopolis, intervention, and the state of nature’, Law, Ethics, and Philosophy, vol. 1, pp. 113-125.
Horta, O., 2015, ‘The Problem of Evil in Nature: Evolutionary Bases of the Prevalence of Disvalue’, Relations: Beyond Anthropocentrism, vol. 3, no. 1, pp. 17-32.
House, I., 1991, ‘Harrison on Animal Pain’, Philosophy, vol. 66, no. 257, pp. 376-379.
Hursthouse, R., 2011, ‘Virtue Ethics and the Treatment of Animals’, in Beauchamp, T.L. & Frey, R.G., The Oxford Handbook of Animal Ethics. Available from: www.oxfordhandbooks.com [9 June 2018].
Hurtado-Parrado, C., 2010, ‘Neuronal mechanisms of learning in teleost fish’, Univ Psychol., vol.9, no.3, pp.663–672.
Jardim-Messeder, D., Lambert, K., Noctor, S., Pestana, F.M., Leal, M.E., Bertelsen, M.F., Alagaili, A.N. & Mohammad, O.B., 2017, ‘Dogs Have the Most Neurons, Though Not the Largest Brain: Trade-Off between Body Mass and Number of Neurons in the Cerebral Cortex of Large Carnivoran Species’ Frontiers in Neuroanatomy, vol. 11, no. 118, pp. 1-18.
Jeyaseelan, M.J.P., Ramanathan, N., Sundararaj, V., Venkataramanujam, K. & Devaraj, M., 1998, ‘Manual of fish eggs and larvae from Asian mangrove waters’, United Nations Educational, Scientific and Cultural Organization. UNESCO Publishing: Paris.
Johannsen, K., 2017, ‘Animal rights and the problem of r-strategists’, Ethical Theory and Moral Practice, vol. 20, no. 2, pp. 333-345.
Johnson, R. & Cureton, A., 2018, ‘Kant’s Moral Philosophy’, The Stanford Encyclopedia of Philosophy, (Spring 2018 Edition), Edward N. Zalta (ed.). Available from: https://plato.stanford.edu/archives/spr2018/entries/kant-moral/ [26 May 2018].
Junk, W.J. & Da Silva, V.M.F., 1997, ‘Mammals, Reptiles and Amphibians’, in The Central Amazonian Foodplain, pp.409-417. Springer: New York.
Kelly, E. & Phillips, B.L., 2015, ‘Targeted gene flow for conservation’, Conservation Biology, vol. 30, no. 2, pp. 259–267.
Kim, S., Cho, Y.S., Kim, H.M., Chung, O., Kim, H., Jho, S., Seomun, H., Kim, J., Bang, W.Y., Kim, C., An, J., Bae, C.H., Bhak, Y., Jeon, S., Yoon, H., Kim, Y., Jun, J.H., Lee, H.J., Cho, S., Uphyrkina, O., Kostyria, A., Goodrich, J., Miquelle, D., Roelke, M., Lewis, J., Yurchenko, A., Bankevich, A., Cho, J., Lee, S., Edwards, J.S., Weber, J.A., Cook, J., Lee, H., Manica, A., Lee, I., O’Brien, S.J., Bhak, J. & Yeo, J.H., 2016, ‘Comparison of carnivore, omnivore, and herbivore mammalian genomes with a new leopard assembly’, Genome Biology, vol. 17, no. 211, pp. 1-12.
Klug, H. & Bonsall, M.B., 2007, ‘When to Care for, Abandon, or Eat Your Offspring: The Evolution of Parental Care and Filial Cannibalism’, The American Naturalist, vol. 170, no. 6, pp. 886-901.
Knutsson, S., 2015, The Moral Importance of Invertebrates Such as Insects, Master’s thesis, University of Gothenburg.
Knutsson, S., Munth, C., 2017, ‘A Virtue of Precaution Regarding the Moral Status of Animals with Uncertain Sentience’, Journal of Agricultural and Environmental Ethics, vol. 30, no. 2, pp. 213–224.
Koch, L.K. & Meunier, J., 2014, ‘Mother and offspring fitness in an insect with maternal care: phenotypic trade-offs between egg number, egg mass and egg care’, BMC Evolutionary Biology, vol. 14, no. 25, pp. 1-9.
Kopnina, H., Washington, H., Taylor, B. & Piccolo, J.J., 2018, ‘Anthropocentrism: More than Just a Misunderstood Problem’, Journal of Agricultural and Environmental Ethics, vol. 31, no. 1, pp. 109-127.
Lack, D., 1947, ‘The Significance of Clutch‐size’, Ibis, vol. 89, pp. 302-352.
Leopold, A., 1949, ‘The Land Ethic’, in A Sand County Almanac, pp.201-226. Oxford University Press: New York.
Littin, K.E., Mellor, D.J., Warburton, B. & Eason, C.T., 2004 ‘Animal welfare and ethical issues relevant to the humane control of vertebrate pests’, New Zealand Veterinary Journal, vol. 52, no. 1, pp. 1-10.
Littin K. E., Gregory N. G., Airey A. T., Eason C. T. & Mellor D. J., 2009, ‘Behaviour and time to unconsciousness of brushtail possums (Trichosurus vulpecula) after a lethal or sublethal dose of 1080’, Wildlife Research, vol. 36, no. 8, pp. 709-720.
Lockwood, JA., 1987, ‘The Moral Standing of Insects and the Ethics of Extinction’, The Florida Entomologist, vol. 70, no. 1, pp. 70-89.
Mason, G. & Littin, K.E., 2003, ‘The Humaneness of Rodent Pest Control’, Animal Welfare, vol. 12, pp. 1-37.
Min, J., Noble, C., Najjar, D. & Esvelt, K., 2017, Daisy quorum drives for the genetic restoration of wild populations, PREPRINT, bioRxic. Available from: https://doi.org/10.1101/115618 [11 June 2018].
Nagel, T., 1974, ‘What is it like to be a bat?’, The Philosophical Review, vol. 83, no. 4, pp. 435-450.
Nelson, M.P., 1998, ‘Aldo Leopold, Environmental Ethics, and The Land Ethic’, Wildlife Society Bulletin, vol. 26, no. 4, pp. 741-744.
Ng, Y.K., 1995, ‘Towards Welfare Biology: Evolutionary Economics of Animal Consciousness and Suffering’, Biology and Philosophy, vol. 10, no. 3, pp. 255–285.
Norton, S., 2001, Bovine Tuberculosis in the Brushtail Possum (Trichosurus Vulpecula): Behaviour and Development of an Aerosol Vaccinator, Ph.D. Thesis. Available from: https://www.massey.ac.nz/massey/fms/Colleges/College%20of%20Sciences/Epicenter/docs/SolisNortonMVS.pdf?56399FEF5AD8DEA0AEF6CC6B898E92C9 [8 June 2018].
Nygård, B.C., 1995, ‘Lowest Lifetime Fecundity’, in University of Florida Book of Insect Records, pp. 39-40.
O’Hara, P., 2006, ‘The illegal introduction of rabbit haemorrhagic disease virus in New Zealand’, Rev Sci Tech, vol. 35, no. 1, pp. 119-123.
Paez, E., 2015, ‘Refusing Help and Inflicting Harm. A Critique of the Environmentalist View’, Relations: Beyond Anthropocentrism, vol. 3, no. 2, pp. 165-178.
Palmer, C., 2015, ‘Against the View That We Are Normally Required to Assist Wild Animals’, Relations: Beyond Anthropocentrism, vol. 3, no. 2, pp. 203-210.
Parliamentary Commissioner for the Environment [PCE]., 2011, Evaluating the use of 1080: Predators, poisons and silent forests. Available from: http://www.pce.parliament.nz/publications/evaluating-the-use-of-1080-predators-poisons-and-silent-forests [8 June 2018].
Pearce, D., 2015, ‘A Welfare State for Elephants? A Case Study of Compassionate Stewardship’, Relations: Beyond Anthropocentrism, vol. 3, no. 2, pp. 153-164.
Princeton Environmental Institute [PEI], 2016, “The Suffering of Wild Animals: Should we do anything about it, and if so, what?” Peter Singer, YouTube video. Available from: https://www.youtube.com/watch?v=1VYYHLH990A [14 June 2018].
Pianka, E.R., 1970, ‘On r- and K-Selection’, The American Naturalist, vol. 104, no. 940, pp. 592-597.
Putnam, H., 1975, ‘The Nature of Mental States’, in Mind, Language and Reality: Philosophical Papers, pp.429-440. Cambridge University Press: Cambridge.
Rastogi, R.K., Izzo‐Vitiello, I., Meglio, M., Matteo, L., Franzese, R., Costanzo, M.G., Minucci, S., Iela, L. & Chieffi, G., 1983, ‘Ovarian activity and reproduction in the frog, Rana esculenta’, Journal of Zoology, vol. 200, no. 2, pp. 233-247.
Ray, G., n.d, Parasite Load and Disease in Wild Animals. Available from: https://was-research.org/paper/parasite-load-disease-wild-animals/#Effects_on_mortality [7 May 2018]
Regalado, A., 2016, ‘The Extinction Invention’, MIT Technology Review. Available from: https://www.technologyreview.com/s/601213/the-extinction-invention [11 June 2018].
Reimann, F., Cox, J.J., Belfer, I., Diatchenko, L., Zaykin, D.V., McHale, D.P., Drenth, J.P., Dai, F., Wheeler, J., Sanders, F., Wood, L., Wu, T.X., Karppinen, J., Nikolajsen, L., Männikkök, M., Max, M.B., Kiselycznyk, C., Poddar, M., te Morscheg, R., Smith, S., Gibson, D., Kelempisioti, A., Maixner, W., Gribble, F.M. & Woods, C.G., 2010, ‘Pain perception is altered by a nucleotide polymorphism in SCN9A’, PNAS, vol. 107, no. 11, pp. 5148-5153.
Rice, C.M., 2013, Defending the Objective List Theory of Well-being, Ratio, vol. 26, no. 2, pp. 196-291.
Rivas, T., 2012, The Denial of Consciousness in Non-Human Animals. Available from: http://txtxs.nl/artikel.asp?artid=696 [12 April 2018].
Royal Society Te Apārangi [RSTA]., 2017., The use of gene editing to create gene drives for pest control in New Zealand, New Zealand Government. Available from: https://royalsociety.org.nz/major-issues-and-projects/gene-editing-in-aotearoa/ [11 June 2018].
Sagoff, M., 1984, ‘Animal Liberation and Environmental Ethics: Bad Marriage, Quick Divorce’, Osgoode Hall Law Journal, vol. 22, no. 2, pp. 297-307.
Segelken, H.R., 1999, Counting mole-rat mammaries and hungry pups, biologists explain why naked rodents break the rules. Available from: http://news.cornell.edu/stories/1999/08/naked-mole-rats-break-rules-nursing-study-shows [11 May 2018].
Shea, N & Heyes, C., 2010, ‘Metamemory as Evidence of Animal Consciousness’, Springer, vol.25, no.1, pp. 95-110.
Shimada, T., Kato, K., Kamikouchi, A. & Ito, K., 2005, ‘Analysis of the distribution of the brain cells of the fruit fly by an automatic cell counting algorithm’, Physica A, vol. 350, no. 1, pp. 144-149.
Singer, P., 1972, ‘Famine, Affluence, and Morality’, Philosophy & Public Affairs, vol. 1, no. 3, pp. 229-243.
Singer, P., 1974, ‘All Animals Are Equal’, Philosophic Exchange, vol. 5, no. 1, pp. 103-116.
Singer, P., 2015, Animal Liberation : The Definitive Classic of the Animal Movement, 40th Anniversary Edition, Open Road Media. Available from: ProQuest Ebook Central [4 April 2018].
Sneddon, L.U., 2015, ‘Pain in aquatic animals’, Journal of Experimental Biology, vol. 218, no. 7, pp. 967-976.
Soryl, A., 2017, Wild Animal Suffering and the Ethics of Extinctionism: Considering Widespread Anti-Natalism as a Solution to Wild Animal Suffering. Available from: https://investigationsinappliedphilosophy.wordpress.com/2017/12/02/wild-animal-suffering-and-the-ethics-of-extinctionism-considering-widespread-anti-natalism-as-a-solution-to-wild-animal-suffering/ [5 June 2018].
Studebaker, B., 2013, Considering Extinctionism. Available from: https://benjaminstudebaker.com/2013/04/29/considering-extinctionism/ [5 June 2018].
Sunquist, M. & Sunquist, F. (2002). ‘Lion’, in Wild cats of the World, pp.285-305. University of Chicago Press: Chicago.
Technical Expert Working Group for the Revision of Directive 86/609/EEC [TEWG]., 2003, Review of EU Directive 86/609/EEC. Final Report from the Scope Sub-Group.
Thomas, D.D., Donnelly, C.A., Wood, R.J. & Alphey, L.S., 2000, ‘Insect Population Control Using a Dominant, Repressible, Lethal Genetic System’, Science, vol. 287, no. 5462, pp. 2474-2476.
Tomasik, B., 2009, How Many Wild Animals Are There? Available from: http://reducing-suffering.org/how-many-wild-animals-are-there/ [4 May 2018].
Tomasik, B., 2015, ‘The Importance of Wild Animal Suffering’, Relations: Beyond Anthropocentrism, vol. 3, no. 2, pp. 133-152.
Tomasik, B., 2016, Insect Suffering from Silk, Shellac, Carmine, and Other Insect Products. Available from: http://reducing-suffering.org/insect-suffering-silk-shellac-carmine-insect-products/#Number_of_bugs_per_unit_product [14 June 2018].
Torres, M., 2015, ‘The Case for Intervention in Nature on Behalf of Animals: a Critical Review of the Main Arguments against Intervention’, Relations: Beyond Anthropocentrism, vol. 3, no. 1, pp. 33-49.
Tuggle, C.K. & Waters, W.R., 2015, ‘Tuberculosis-resistant transgenic cattle’, PNAS, vol. 112, no. 13, pp. 3854-3855.
Tyler, M.J., 1998, Australian Frogs. Penguin Books: London.
Van Gulick, R., 2009, ‘Functionalism’, The Oxford Handbook of Philosophy of Mind. Available from: http://www.oxfordhandbooks.com/view/10.1093/oxfordhb/9780199262618.001.0001/oxfordhb-9780199262618-e-8 [11 April 2018].
Vinding, M., 2016, The Speciesism of Leaving Nature Alone, and the Theoretical Case for “Wildlife Anti-Natalism”, Rakuten kobo. Available from: https://www.kobo.com/us/en/ebook/the-speciesism-of-leaving-nature-alone-and-the-theoretical-case-for-wildlife-anti-natalism . [2 June 2017].
World Wildlife Fund [WWF]., n.d, Priority species, database. Available from: http://wwf.panda.org/knowledge_hub/endangered_species/ [10 May 2018].
Zoladz, P.R., 2008, An ethologically relevant animal model of posttraumatic stress disorder, Ph.D Dissertation, University of South Florida.
Citation style in accordance with the University of Western Australia, Harvard
Word Count: 23279