Getting to grips with wildlife research by citizen scientists: What role for regulation?

Wildlife research by citizen scientists, involving the capture and handling of animals, provides clear scientific benefits, but also potential risks to animal welfare. We explore debates about how best to regulate such work to ensure that it is undertaken in an ethical manner. We focus on the UK as a case study, drawing on qualitative research and stakeholder engagement events. We show that because trapping and marking of certain species requires minimal licensing, training and justification, some argue for increased formal regulation to minimise risks to animal welfare. However, others have reflected on the already complex regulatory landscape affecting wildlife research, and have expressed concern that introducing additional formal regulations could potentially make citizen science working with wildlife more difficult. Informal regulation could therefore offer a preferable alternative. We set out three steps that could be taken to open up conversations about ethics and regulation of wildlife-focussed citizen science, in the UK and elsewhere: (a) take stock of wildlife-focussed citizen science in terms of numbers and harms to animal welfare; (b) assess the state of formal regulations and consider reforms; and (c) consider informal regulations as alternatives or additions to formal regulations.

. Approaches that lean towards a democratic approach might closely resemble participatory action research (PAR), which involves collaboration between academic researchers and participants to examine and improve a problematic situation for the participants (Eitzel et al., 2017;Kindon et al., 2009). Developing this still further, what has been termed 'extreme citizen science' might involve scientists acting as facilitators who help participants to achieve their goals; or, the entire process could be carried out without professional scientists (Haklay, 2013). Citizen science projects therefore potentially have multiple goals, such as scientific data collection, fostering public engagement with nature or science, and social change.
Use of the term 'citizen science' may also vary cross-culturally, in part because the term may not be the most appropriate to describe indigenous peoples' knowledge and involvement in scientific data collection (Eitzel et al., 2017). Indigenous knowledge holders might be important 'citizen scientists', such as in participatory mapping (Kimura & Kinchy, 2016) and wildlife and environmental monitoring projects (Flora & Andersen, 2016;Hoyte, 2017) aimed at helping indigenous groups access territorial and resource rights. Indigenous knowledge may also be sought for environmental research and management, such as in the field of ethnobiology (Anderson et al., 2012).
However, as Eitzel et al. (2017) observed, the term 'science' may not always be appropriate or acceptable to these groups, although indigenous knowledge development might also be regarded as an example of science in its own right (see for example Giraldo Herrera, 2018). Eitzel et al. (2017) also observed that referring to indigenous peoples as 'citizens' may in some cases be inadvisable due to the legacies of colonialism.
For our purposes, we use the term 'citizen science' with reference to the participation by members of the public in wildlife research, irrespective of the more specific goals of the project or the citizen scientists' level of involvement in determining the research methods, goals, and outcomes. By 'members of the public' we mean people who are not employed, working towards qualifications, or in possession of formal qualifications in wildlife research. We recognise that these terms and working definitions may not be appropriate for discussing the involvement of indigenous peoples in scientific knowledge production. However, our paper primarily focuses on the UK, where the term 'citizen science' is widely used (Eitzel et al., 2017) and which lacks a recent history of colonisation.
The field of astronomy perhaps boasts the highest profile citizen science projects, many of which have used the platform Zooniverse to collect volunteer observations of celestial objects and phenomena (e.g. Faherty et al., 2020). Alongside astronomy, wildlife research tends to dominate in the public profile of citizen science, perhaps because such work lends itself so readily to public participation (Dickinson et al., 2010;Roy et al., 2012). While citizen science focussed on wildlife takes many forms, including surveys that involve minimal disturbance, we write specifically about activities involving handling or other direct interaction-for example, trapping, marking and the use of tracking devices-with free-ranging wild animals.
Hereafter, we refer to such work simply as 'wildlife-focussed citizen science'. Examples of such projects in the UK include the National Nathusius' Pipistrelle Project (Bat Conservation Trust, 2020) and ringing schemes of the British Trust for Ornithology (BTO;Robinson et al., 2019), which both enlist experienced ringers to provide species survey data. Another example is the Mammal Society's research comparing methods for monitoring harvest mice (Micromys minutus) and other small mammals, which involved live trapping undertaken by non-professional volunteers (Poulton & Turner, 2009).
Our goal in this perspectives piece is to stimulate conversations about the regulation and ethics of wildlife-focussed citizen science, using the UK context as a detailed case study. As Rasmussen and Cooper (2019) pointed out in their introduction to a special issue of Citizen Science: Theory and Practice devoted to ethics, there is no obvious drive from citizen scientists or regulators to talk about ethics in citizen science (at least not in the USA where these co-authors reside). Despite this, they noted that focussing on ethics not only shows a willingness to undertake a 'healthy assessment of the field', but is also sensible given that we should expect problems to arise at some point, as they would in any field. For example, while citizen science has not yet seen any major scandals involving researcher misconduct, such as the fabrication of data, there is always the potential for them to occur, especially when working practices have not been fully formalised. There have not to our knowledge been any scandals regarding animal welfare in wildlife-focussed citizen science, and citizen scientists may often possess a greater degree of skill and experience in trapping, handling, and marking animals than paid professionals with whom they collaborate (see Eitzel et al., 2017;Pocock et al., 2014). However, because wildlife-focussed citizen science involves direct engagement with animals, it potentially poses risks to animal welfare. Like other biodiversity conservation work, it may also involve important ethical considerations, including how to balance biodiversity conservation, the welfare of individual animals, and the values and rights of indigenous peoples such as customary hunting rights (Wehi & Lord, 2017) and non-Western understandings of animal welfare and ethics (Callicott, 1994;Doerfler & Peters, 2006;Linzey & Linzey, 2018). Such considerations arise, for example, in the case of subsistence hunting of narwhals (Monodon monoceros) and other species of conservation concern by indigenous groups in the Arctic, where indigenous hunters may be involved in recording the presence of wildlife or catching animals for tracking (Flora & Andersen, 2016). Following other similar work exploring ethics in citizen science (see other contributions to the 2019 Citizen Science special issue, and Resnik et al., 2015;Riesch & Potter, 2014), we therefore see a need to engage in discussions about ethics and regulation in wildlife-focussed citizen science.
We focus on ethics and regulation together because ethical standards inform the development of regulations, and regulations are in turn are deployed as tools for promoting ethical behaviour. For example, laws are deployed as instruments for protecting animal welfare, although in general domestic animals have received more attention in terms of welfare regulations than free-living wildlife (Dawkins & Bonney, 2008;Harrop, 1997Harrop, , 2011Harrop, , 2013. This pattern arguably follows a broader trend of wild animal welfare being relatively neglected as a subject of attention among wildlife researchers (e.g. see Cattet, 2013) and philosophers (e.g. see Paquet & Darimont, 2010).
Wildlife law is also shaped by cultural practices, with Harrop (1999Harrop ( , 2013 arguing that the development of wild animal welfare and conservation law has been protracted by hunting interests, with hunting with hounds, for example, only banned in the UK in 2004. Wildlife law in countries with a history of colonisation may also fail to account for indigenous subsistence and customary hunting, such as in Aotearoa (the Māori name for New Zealand; Ruru et al., 2017;Wehi & Lord, 2017). Thus, laws are not only conceived of as tools for promoting ethical standards (such as humane treatment of animals), but are also shaped by philosophical, cultural, and historical approaches to ethics.
Regulations intended to promote or enforce ethical behaviour can be thought of as having varying levels of formality, ranging from 'command and control' regulations involving law and government enforcement, through to more informal and non-binding forms of regulation (Gorwa, 2019). Examples of the latter include: voluntary codes of conduct, such as those intended to promote ethical behaviour in companies (Gorwa, 2019;Hodges, 2015); internally developed and encouraged standards, within professions (e.g. the legal profession: Wendel, 2001) or institutions, such as via the promotion of 'cultures of care' within animal research facilities (Greenhough & Roe, 2018); pressure from communities affected by corporate behaviour, such as in response to pollution (Pargal et al., 1997); and the use of behaviour-change tools such as covert messaging to structure people's choices towards a desired outcome, commonly known as 'nudge' (Baldwin, 2014;Jones et al., 2014;Whitehead et al., 2019).
Both formal and informal regulations have been proposed for encouraging ethical behaviour in citizen science. For example, Rasmussen (2019) suggested the use of best practice guidelines and training resources to help to prevent research misconduct, and Guerrini et al. (2018) proposed changes to laws and the creation of codes of conduct for protecting participants and intellectual property, and promoting scientific integrity in citizen science (see also Cooper et al., 2019). While some attention has been paid to the ethical and regulatory challenges faced by professional wildlife researchers (e.g. Cooke et al., 2016;Curzer et al., 2013;Russow & Theran, 2003;, few commentators have reflected on how formal and informal regulations work, and ought to work, for citizen scientists undertaking work with wildlife. This situation could, in part, reflect a reality that much citizen science receives little formal regulatory oversight (Rasmussen, 2019, p. 5). For example, Rasmussen (2019, p. 2) observed that citizen science lacks an overall approach to ensuring research integrity, in part because its 'decentralized, open-access ethos' means fewer formalised organisational processes such as institutional ethical review. In other words, citizen science may face less oversight than the work of professional scientists, in part because citizen science is inherently a decentralised activity involving members of the general public.
We reflect on the opportunities and challenges of regulating wildlife-focussed citizen science from a UK perspective. While some details are therefore necessarily country specific, we use them to reflect on questions and proposals that are applicable to citizen science internationally. The UK offers a particularly interesting case study because regulation of animal (including free-living wildlife) research in the UK is often described as particularly comprehensive.
We base our examination of the UK context on personal knowledge and experience. S.J.R. is a professional wildlife researcher and a member of a university Animal Welfare Ethical Review Board This piece, and the panel discussion on citizen science regulation, were inspired by the ways in which regulation and ethics in wildlife-focussed citizen science emerged as key themes in the qualitative research and workshop on non-laboratory research. In particular, research and workshop participants expressed a range of views on whether there is a need for further regulation of hands-on work with free-ranging animals by citizen scientists (e.g. capturing, handling, and marking animals), given that some work (depending on the species) requires minimal licensing, training, and justification in the UK. For example, some small mammal trapping using Longworth traps, and trapping of some larger species such as red foxes (Vulpes vulpes), does not require licences. Some argued for the merits of extending the A(SP)A to cover a greater proportion of wildlife work by citizen scientists. Others reflected on the already complex legal landscape affecting wildlife research in the UK, and the risk that over-regulation would reduce the amount of citizen science taking place, suggesting that further formal regulation is undesirable. If no formal regulation was introduced, the question is then: How might we encourage, and potentially enforce, ethical behaviour in the absence of regulation? Our goal in this piece is not to argue from either of these positions, but rather to consider how and why these questions have arisen, some proposals for answering them, and the implications of these proposals. In addition, we propose three steps that could be taken to progress conversations about formal and informal regulation of wildlife-focussed citizen science in the UK and internationally.

| REG UL ATI ON OF WILDLIFE-FO CUSS ED CITIZEN SCIEN CE: THE KE Y SOURCE S OF DEBATE IN THE UK
To begin, we briefly describe wildlife research legislation in the UK and the key debates surrounding it (see Harrop, 1997;Lane & McDonald, 2010 for further details on these laws). These laws, and relevant points about them in relation to capture and handling of wildlife by citizen scientists, are summarised in Table 1.

First, the A(SP)A (1986) is implemented and enforced by the HO.
The A(SP)A regulates work with animals (vertebrates and cephalopods) undertaken for a 'scientific or educational' purpose, which meets or exceeds the 'lower threshold', defined (since 2013) as causing the animal 'a level of pain, suffering, distress or lasting harm equivalent to, or higher than, that caused by inserting a hypodermic needle according to good veterinary practice'. While science is not explicitly defined within the A(SP)A or its associated guidance, specific activities are excluded from its remit. In the most recent revision of the A(SP)A these include 'the ringing, tagging or marking of an animal' or other procedures undertaken 'for the primary purpose of enabling an animal to be identified, provided that it causes only momentary pain or distress (or none at all) and no lasting harm', procedures commonly used by citizen scientists working with wildlife (for above see Home Office, 2014, pp. 10-12). In addition, the A(SP)A does not cover the capture of wild animals, although guidance indicates that animals must be captured by a 'competent person using a method which does not cause the animal avoidable pain, suffering, distress or lasting harm' (ASRU, 2016, p. 22). Similar purpose-and harm-based criteria restrict the scope of local laws related to animals in science across Europe, following the introduction of EU Directive 2010/63/EU on the protection of animals in science (Lindsjö et al., 2019;NORECOPA, 2017).  as well as its mass and size (Vandenabeele et al., 2012(Vandenabeele et al., , 2014. Considering obvious impacts on avian locomotion, such devices potentially affect other aspects of avian biology such as breeding performance and foraging success to name but two (see Geen et al., 2019;Vandenabeele et al., 2011 for reviews). For these reasons, the pragmatic '3% rule' (i.e. tags should be at most 3% of the body mass of the bird on which they are deployed) is used widely in field ornithology but is no guarantee that deployment is free from tag-induced adverse effects (Geen et al., 2019). Recent research has suggested that tags weighing only 3% of avian body mass may result in a 4.67%-5.71% increase in energy expenditure during flights by seabirds (Vandenabeele et al., 2012). Thus, determining the circumstances under which tagging birds might exceed the threshold for inclusion under the A(SP)A is a complex task (see Wilson, 2017 (Harrop, 1999).
Even species that are not protected under the WCA or other wildlife laws are covered by the AWA, which focusses on both preventing animal cruelty and ensuring that the welfare needs of an animal are met (see also Lane & McDonald, 2010 (Harrop, 1999, p. 694). One might also look to other cultural factors as influencing wildlife laws. Birds have featured prominently in arts, folklore, spirituality, healing practices, and other customs around the world (Tidemann & Gosler, 2010). For example, for Māori the now-extinct huia (Heteralocha acutirostris) was especially revered, with their feathers worn by chiefs of distinction during specific ceremonies and when entering battle (Houston, 2010). In the UK, birds are regarded as an especially 'charismatic' taxon (Lorimer, 2015).
However, the cultural heritage and contemporary public perception of birds may vary by species; Eurasian magpies (Pica pica), for example, have historically been viewed as thieves and harbingers, and today tend to be less popular than other garden birds in the UK (see Amongst our research and event participants, concern has been directed at forms of trapping perceived as particularly high risk. One mammalian example is the use of Longworth traps for capturing small mammals, with shrew mortality ranging from 10% to 93% (Shonfield et al., 2013). Indeed, trapping in general is widely perceived as the most stressful part of research for a wild animal, with the level of stress experienced by the animal likened to being pursued and captured by a predator (Wilson et al., 2019).
Negative effects may also persist after capture. For example, myopathy, which is caused by a build-up of lactic acid in muscles and is particularly common in large mammals such as deer (Cervidae) following prolonged pursuit and handling, may take as much as a week to manifest and therefore cannot be detected during capture (Lane & McDonald, 2010 (Eden, 2012) or assessing the effects of repeated capture on fish welfare (Thorstad et al., 2020), so the distinction between recreation and citizen science can be hazy. In essence, the concern expressed by research participants was that some projects lacked sufficient benefits (e.g. contributions to the theoretical and applied scientific knowledge base) to outweigh the harms (e.g. to individual animal welfare). This worry links with a broader concern about citizen science-that if data quality is poor, projects may not be jus- One concern is that introducing still more formal regulation to what is already a permit-heavy area may not be ideal. Paul and Sikes (2013)

| MOVING FORWARDS
To conclude, we highlight three steps that we propose should be taken to progress discussions of ethics and regulation in wildlifefocussed citizen science, as illustrated in Table 2.
We propose that these three steps may also be applicable to other countries, although the nature of each step may differ. For example, countries with indigenous peoples will need to consider indigenous rights and management responsibilities in considering any formal regulatory reforms affecting citizen science. Still, broadly these steps could be applied internationally with effectiveness. We do not offer a strict timeline for these steps but we suggest that where possible they are best undertaken in the order specified in Table 2. They should begin as soon as possible; in the UK our research has highlighted a desire among many research and event participants to resolve outstanding issues as a matter of priority.
However, each step involves a slightly different set of actors: step 1-SNCOs and citizen science groups; step 2-government regulators; and step 3-citizen science societies and groups, institutions engaged in citizen science (e.g. NGOs) or which administer citizen science projects (e.g. Zooniverse: www.zooni verse.org), professional researchers, and citizen scientists themselves where participatory approaches are adopted. Because a variety of actors are involved, citizen science communities cannot accomplish all three steps on their own, and may be restricted to implementing only step 3 without support from government regulators.
We now describe in greater depth the justification for each step, and what resolutions could involve. The first step we propose is to take stock of the range and number of procedures currently undertaken by citizen scientists with wildlife (Table 2) Encourage institutional or sector-wide standards (e.g. via dissemination of best practice guidelines and training resources), and take a participatory or process-oriented approach (e.g. via creation of a mnemonic) to encourage reflection on animal welfare and promote harm-benefit assessment would likely be doing work classified as 'low risk' in terms of animal welfare harms and public concern. To place this proposal in context, the BTO reported that 982,858 birds were ringed by licensed ringers in 2018  Even if such statistics were available, we would still need to determine whether introducing formal regulation to citizen scientists would be desirable, hence our proposed second step of assessing the state of formal regulations and considering reforms (Table 2).
This step comes with several issues and challenges. One is that citizen scientists would be unlikely to be directly involved in formal regulatory reform; decisions about citizen science would therefore be made in a top-down manner. This issue resonates with broader discussions around power dynamics in citizen science, such as the risk that citizen scientists will receive insufficient credit for their work (see Resnik et al., 2015;Riesch & Potter, 2014). Furthermore, resolving this step requires consideration of who would be tasked with regulation, and how this would affect their needs in terms of staffing and expertise, and their ability to regulate effectively.
In addition, revising formal regulations would involve deciding on how to balance the perceived benefits of wildlife research carried out by citizen scientists-for example, encouraging public engagement with science and nature, and contributing to science and conservation-with the risks, particularly compromises to animal welfare. To an extent, this consideration reflects a more general problem of balancing species and ecosystem conservation with individual animal welfare (e.g. Beausoleil, 2014;Hayward et al., 2019;Paquet & Darimont, 2010;Ramp & Bekoff, 2015), and in particular how best to balance these concerns in formal regulation (Harrop, 2010).
One possible outcome of this second step is the conclusion that informal regulation is more desirable or effective than formal regulation for encouraging best practice and ethical behaviour in cit- whereby they follow the letter, but not the spirit, of the law (p. 2).
Other scholars have highlighted a risk that regulations enforced through auditing become merely a 'tick box' exercise, with the audit process becoming decoupled from the qualities it is meant to assure and becoming an end in itself (e.g. Escobar & Demeritt, 2017;Power, 1999;Strathern, 2000). Meanwhile, Hodges (2015) argued that formal regulations may be ineffective because they are based on flawed understandings of human behaviour, such as the principle of deterrence (i.e. that punishment for undesirable behaviour will prevent similar behaviour in future) and the idea of humans as 'rational actors' who function in isolation. Formal regulation's effectiveness may also be limited given that it tends to focus on deterring negative, rather than encouraging positive, behaviour. For these reasons, Hodges (2015) advocated for a holistic approach to encourage ethical behaviour in which formal regulatory systems are combined with a broader range of approaches, with the goal of encouraging corporations to 'achieve and exceed compliance' (p. 703) rather than simply deterring negative behaviour.
Informal regulatory mechanisms may therefore also play an important role in encouraging ethical behaviour, hence our proposed third step that informal regulations be considered as alternatives or supplements to formal regulations (Table 2). For instance, in the realm of animal research a 'culture of care' is increasingly encouraged. While the culture of care concept is mentioned in EU regulations and the A(SP)A guidance documents (ASRU, 2017;Greenhough & Roe, 2018), it is not strictly tied to formal regulation, and is perhaps best described as a kind of 'workplace atmosphere' in which care is encouraged for both animals and staff, and legal regulations are not just met but exceeded (Greenhough & Roe, 2018, p. 12). In other words, a culture of care is primarily a product of informal regulation through professional and institutional standards rather than formal regulation through laws and enforcement. Rasmussen (2019) made similar proposals for preventing misconduct amongst citizen scientists. Among other ideas, Rasmussen (2019, p. 5) suggested enlisting citizen science organisations to disseminate key resources, run tutorials, and encourage their members to follow best practice guidelines, with the goal of promoting a 'culture of research integrity' within citizen science. However, one potential weakness of this type of informal regulation is that institutional and professional cultures may develop harmful as well as positive norms (Wendel, 2001).
In addition to institutional or professional values and standards, covert manipulation of behaviour is also sometimes employed as an informal regulatory mechanism for encouraging desirable behaviour. 'Nudge' describes an approach of structuring people's choices to lead them towards a desired outcome, such as supplying information and reminders, placing limitations on choices, and using covert framing strategies to encourage people to make healthy lifestyle choices (see Baldwin, 2014 This is perhaps where more participatory forms of informal regulation should come to the fore. For instance, informal regulation can take the form of communities boycotting or protesting the work of companies, which may in turn shape corporate behaviour (Pargal et al., 1997). Taking a more participatory approach would also resonate with the more 'democratic', PAR-oriented, and 'extreme' strands of citizen science, which encourage public participation in shaping the goals and processes of science (Cooper & Lewenstein, 2016;Eitzel et al., 2017;Haklay, 2013;Riesch & Potter, 2014). More participatory approaches to promoting ethical standards could involve engaging with citizen scientists in deciding on harms and benefits of research from the outset. It could also involve encouraging citizen scientists towards a process rather than a result, such as directing them towards harm-benefit analysis rather than a specific position on the circumstances under which animal welfare outweighs the accumulation and application of scientific knowledge. This approach might be understood as resembling the A(SP)A which, while offering guidance on the kinds of factors to be considered in harm-benefit analysis, does not contain a predetermined position on the circumstances when scientific benefits outweigh harms to animal welfare, or vice versa. Thus, formal regulation can also potentially be process oriented and flexible. To give an example of participatory, process-oriented informal regulation, during a workshop on non-laboratory research we discussed the idea of creating a simple tool such as a mnemonic that could be disseminated to remind citizen scientists to think carefully about regulation, animal welfare, and harms and benefits associated with their research activities (Animal Research Nexus, 2019). This kind of informal regulation could help to promote consideration of ethics, harms, and benefits within citizen science communities while ensuring that citizen science could still readily proceed. It could also maintain the 'democratic' ethos of some citizen science via the process of co-producing ethical standards.
To conclude, we have argued in this piece that there is a need to talk about ethics and regulation in wildlife-focussed citizen science. In a context where trapping and marking certain spe- research, it may be preferable to encourage citizen scientists towards a process of ethical evaluation rather than a result. Some wildlife-focussed citizen science (e.g. ringing birds under the BTO's authority) is already thoroughly and formally regulated, but some others with little formal regulation might already encourage ethical behaviour in various ways, such as offering best-practice guidelines (e.g. a booklet on live trapping small mammals published by the Mammal Society: Gurnell & Flowerdew, 2006).
Despite this, we think it is nonetheless advisable to encourage and promote conversations about ethics and regulation in citizen science, as a way of undertaking a 'healthy assessment of the field' (Rasmussen & Cooper, 2019, p. 1). We have proposed three steps that could be taken to progress these conversations, both in the UK and internationally (Table 2).

ACK N OWLED G EM ENTS
We gratefully acknowledge support from the Wellcome Trust, which Finally, we wish to thank our AnNex colleagues and POLEs research participants for their ongoing support.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

AUTH O R S ' CO NTR I B UTI O N S
A.P. led the writing of the manuscript; B.G., J.L. and S.J.R. critically commented on and contributed text to the manuscript; R.D. and all authors were involved in conceiving the arguments in the manuscript and approved the manuscript for submission; A.P. and B.G.
were involved in data collection for the AnNex research, and in organising the stakeholder events that shaped the arguments in the manuscript; J.L., R.D. and S.J.R. presented at both events.

DATA AVA I L A B I L I T Y S TAT E M E N T
By agreement with the Wellcome Trust and research participants, anonymised interview transcripts will be deposited in the UK Data