Stakeholder discourse and opinion towards a charismatic non‐ native lizard species: Potential invasive problem or a welcome addition?

1. Analysis of discourse between stakeholders is becoming increasingly recognised for its importance in resolving conflicts of opinion regarding complex environmen‐ tal issues such as the human‐mediated spread of invasive non‐native species—one of the major drivers of biodiversity loss world‐wide. Species’ attributes, stake‐ holders’ level of knowledge, perceptions of threat, attitudes towards intervention and nature values all have subjective influence on opinion, often creating highly opposed interests and perspectives that can create barriers preventing effective management. 2. Using a Q method approach towards analysis of subjective opinion among stake‐ holders, this study aimed to identify emerging viewpoints regarding the presence of Common Wall Lizards ( Podarcis muralis ) in the UK—an introduced, non‐native species with which there are high levels of human interaction but low levels of knowledge regarding potential negative ecological impacts. It explores the ways in which different stakeholder groups (i.e., public, land managers, conservationists) might share views and the reasoning behind shared or opposing discourse be‐ tween


| INTRODUC TI ON
The interaction between nature and society has never been more complex, politicised or researched (Aitken, 2012;Biermann & Mansfield, 2014;Foss, 2018). This is particularly evident in relation to the causes and consequences of global loss in biodiversity, frequently regarded as one of the most pressing environmental challenges currently facing humanity (Skogen, Helland, & Kaltenborn, 2018). Loss in biodiversity has negative impacts on both ecosystem function and ecosystem services, ultimately threatening human well-being (Cardinale et al., 2012;Naeem, Chazdon, Duffy, Prager, & Worm, 2016). Human activities are the predominant drivers behind biodiversity loss (e.g. over exploitation, land use change, introduction of invasive species) (Maxwell, Fuller, Brooks, & Watson, 2016) and, as such, there are a multitude of associated socio-environmental issues that are often defined by highly opposed interests and perspectives amongst stakeholders (i.e. economical, political, ecological, cultural, social) and that prevent unilateral solutions to the wider problem.
Indeed, the discourse surrounding issues of INNS appears to be beset not only by the inherent uncertainty that surrounds ecological knowledge of biological invasions, their management and environmental implications (García-Llorente et al., 2008), but also by highly varied and subjective comprehension of emotive terms such as 'nonnative', 'alien', 'exotic', 'pest', 'invasive species', 'ecological impact' and the connotations that arise through social representation of these concepts as a whole (DEFRA., 2009;Essl et al., 2018;Tassin & Kull, 2015). The way in which people become familiar with complex ecological concepts also has an important influence on the development of opinions and formation of perceptions. Non-scientists and scientists have varied frames of reference, and make comparisons with existing personal understanding from other domains to formulate their comprehension of the complex (and relatively new in public conscience) ecological concept of INNS (Fischer et al., 2014;. Differences between viewpoints based on these utilitarian, moralistic, humanistic or naturalistic values thus have potential to cause significant obstacles to management operations (e.g. public opposition to plans for eradication of the American grey squirrel in Italy where the species' poses a threat to native red squirrel [Bertolino & Genovesi, 2003]). The myriad factors that might shape opinion and ultimately result in contesting viewpoints can be understood within a conceptual framework of hierarchal influence.
Within that framework, attitudes and opinion are defined by an individual's core values (i.e. stable mental constructs that transcend specific situations and represent personal needs according to enduring beliefs) and perceptions of risk (judgments of potential hazards influenced by heuristic rules and social context that simplify complex 5. These findings will be useful in identifying discordant attitudes and areas of potential contention between stakeholders that may arise in consideration of management decisions regarding non-native species more widely. The holistic method of interpreting the analysis gives insight into how and why stakeholders may have formulated certain viewpoints. This in turn could help conservation managers identify ways in which to appreciate and work with subjective influences on stakeholder perceptions in order to best communicate the complex challenges and opportunities presented by non-native species.

K E Y W O R D S
discourse analysis, invasive non-native species, non-native species, Q method, socioenvironmental issues, stakeholder opinion

Box 1 Definition of terms
The term 'non-native species' (NNS) is the equivalent of 'alien species' as used by the Convention on Biological Diversity (CBD) to describe species that occur in the wild (following introduction) beyond their natural geographic range. INNS or 'invasive non-native species' (the equivalent of 'invasive alien species' or 'IAS' are broadly defined as species whose introduction and/or spread threaten biological diversity or have other unforeseen impacts. concepts and may misrepresent reality; Estévez, Anderson, Pizarro, & Burgman, 2015).
It has become increasingly recognised that discourses involving stakeholders should be a focus of analysis in environmental governance studies. These discourses are important for facilitating cooperation between diverse actors aimed at resolving complex environmental problems (Hagan & Williams, 2016;Hajer & Versteeg, 2005). Indeed, understanding of the role of discourse between actors of different status, and the inherent subjectivity that shapes personal perception and opinion driving the policymaking process has become a burgeoning science all of its own (Andersen, Schulze, & Seppel, 2018;Frate & Brannstrom, 2017).
Discourse analysis has received much attention in the environmental sciences, particularly in politically charged arenas of response to climate change (Foss, 2018;Lansing, 2013), renewable energy (Mukonza, 2017;Rennkamp, Haunss, Wongsa, Ortega, & Casamadrid, 2017), land use change and sustainable development (Cook, Pieri, & Robbins, 2004;Soini & Birkeland, 2014;Walder & Kantelhardt, 2018). Identification of social perspectives through such analysis may avoid conflict and barriers to planned strategies and suggest socially acceptable solutions for their implementation (Frate & Brannstrom, 2017;Mayett-Moreno, Villarraga-Flórez, & Rodríguez-Piñeros, 2017). Arguably, the most emotive and polarized discourse within environmental science can be found in regard to contention over issues of wildlife management. Such conflicts are often characterised by arguments over scientific truth claims and the addition of ethical/animal welfare dimensions as typified by such issues as the culling of badgers (Meles meles) to control the spread of bovine tuberculosis in the UK (Price, Saunders, Hinchliffe, & McDonald, 2017), lethal control of wild deer populations, and other 'pest' species (Dandy et al., 2012;Wallwork & Dixon, 2004;White et al., 2003). However, uncertainty borne from knowledge gaps and deficiencies in data often pervades understanding of such environmental issues. Thus, in the absence of, and perhaps sometimes in despite of (Gozlan et al., 2013) empirical evidence, actors involved in associated discourses necessarily construct their viewpoints and understanding based on subjective influences (i.e. scientific reasoning/theory, personal experience, historical/cultural bias, economical, ethical values, heuristic judgement) (Davidson, Cambell, & Hewitt, 2013).
Studies investigating stakeholder discourse have done so largely through standard interview and questionnaire techniques and quantitative analysis of participant responses aimed at providing a representative overview of attitudes held (Bremner & Park, 2007;García-Llorente et al., 2008;Hoyle, Hitchmough, & Jorgensen, 2017;Poudyal, Bowker, & Moore, 2016). Other authors have employed qualitative techniques such as focus groups to delve deeper into the reasoning behind the formation of particular viewpoints, enabling an interpretation of the data that allows for a broader understanding and recognition of important patterns and themes within the discourse (Dandy et al., 2012;Selge, Fischer, & van der Wal, 2011). One method that combines both qualitative and semi quantitative analysis of subjective opinion among stakeholders is Q method (Brown, 1993;Yang, 2016). Although frequently used in analysis of discourse within the environmental sciences (Brannstrom, 2011;Hagan & Williams, 2016;Lansing, 2013;Walder & Kantelhardt, 2018) the method has seen very little application in regard to discourse on NNS (Falk-Petersen, 2014).
Here, we use a Q method approach to investigate the discourse surrounding the presence of Common wall lizard Podarcis muralis in the UK-a species that has established numerous populations in the South of England following introduction from mainland Europe and is thriving in habitats ranging from coastal cliffs of Dorset to highly urbanised areas of West Sussex (Michaelides, While, Zajac, & Uller, 2015). The presence of wall lizards poses a potential risk to native lizards through contest and transmission of pathogens, and possible wider impacts on invertebrate communities (Foster, 2015), although to date there is only anecdotal evidence to suggest negative impacts are apparent (Mole, 2010).
Being locally abundant and gregarious in behaviour the lizards are frequently encountered by members of the public, particularly in areas of high public footfall such as the seafronts of Bournemouth and Eastbourne. No control measures are currently in place to manage existing populations, and further ecological research into established populations is needed to assess viability and justification for management (Foster, 2015). This situation presents an interesting opportunity to investigate people's attitudes towards the lizards in the face of high levels of interaction but potentially low levels of knowledge regarding possible negative ecological impacts. Through use of a holistic Q method approach this study aims to identify emerging viewpoints regarding the presence of wall lizards, and explores the ways in which different stakeholder groups (i.e. public, land managers, conservationists) might share views and the reasoning behind shared or opposing discourse between groups. Such insight will be useful in identifying discordant attitudes and areas of potential contention between stakeholders that may arise in consideration of management decisions regarding NNS more widely. In addition, the analysis will also help illustrate how people reason their subjective views regarding complex ecological concepts in general. Specifically, we ask: How and why do stakeholders group in their opinions towards the lizard introduction? What does the discourse in this case study tell us about perceptions and attitudes towards management of introduced species more generally?

| MATERIAL S AND ME THODS
All Q studies are reconstructive and characterized by two key features. Firstly, the collection of data is done in the form of Q sorts (Hagan & Williams, 2016;Watts & Stenner, 2012). This is typically done by presenting people with a sample of statements (or items) about the given topic, which is referred to as the Q set. The selected participants, or P-set, are then instructed to rank-order the statements from their personal point of view on a score sheet. Participants work with their subjective interpretation of the statements thus revealing their subjective viewpoint in the final rank order of statements. Secondly, these Q sorts are factor-analysed to establish different patterns ('discourses'). Unlike in standard survey analysis this factor analysis is not aimed at establishing patterns across individual characteristics such as age, gender and class, but rather patterns within and across individuals by focusing on their discursive understanding of a particular issue (Hagan & Williams, 2016;Watts & Stenner, 2012).

| Q set design
The Q set was developed from statements generated from several  (Table 1).

| Participants
Twenty-six participants across seven stakeholder groups (Table 2) took part in the Q sorts in the summer of 2017. Participants from the land manager group were invited to take part through having had previous contact with the lead researcher in granting consent to conduct field work at wall lizard sites in Dorset and East/West Sussex as part of wider ecological research into the introduction.
Although employed in environmental management/conservation officer roles, their specific knowledge of the wall lizard introductions was not assumed. The public group consisted of participants who again had previously been engaged with during ecological field work in Devon, Somerset and West Sussex during 2016-2017 and were known to at least be aware or have direct experience of wall lizards in their local area (e.g. has them present in garden, encounters them frequently). Members of the public with no experience of the lizards were not considered to take part in this Q sort as many of the statements in the Q set would have no personal relevance to such a group, thus making interpretation and sorting of statements problematic (see Q sort Methods). Participants forming the environmental advisory group were invited to take part based on their expertise in the field of INNS and national biosecurity-their specific knowledge of the wall lizard introductions was also not assumed. for the land manager and public groups as it was considered there might be greater scope for variation in experience and opinion within these sectors compared to other groups. Participants were given the same background information before the Q sort regarding the nonnative status of wall lizards in the UK, but no other information was given about the consequences of their presence. Instructions were also careful to point out that none of the statements were to be considered as 'scientific fact', and merely represent a point of view.

| The Q sort
The Q sort was administered either in person or online using software developed by Pruneddo (2013). Thirteen participants conducted their Q sorts in the presence of the lead researcher. The Q set was given to the respondent in form of a deck of randomly numbered cards. Each card contained one of the 76 statements from the final Q set. The participant was first instructed to sort the deck into three piles; 'identify/ agree', 'neutral/undecided' and 'do not identify/disagree', depending on his/her personal point of view. This initial 'sort' served to familiarize the participant with the nature of the statements and start formulating their own viewpoint. Thereafter, the respondent was instructed to sort out the statements on a score sheet with a pyramidal, or 'quasi-normal', sorting distribution, ranging from 'strongly disagree' (−6) to 'strongly agree' (6). The sorting distribution was pre-arranged; the whole Q set had to be allocated a ranking relative to one another within this distribution (Watts & Stenner, 2012) (see Figure 1). Participants were TA B L E 1 The final Q set' of statements (and their ID numbers) representing the overall discourse surrounding the presence of Podarcis muralis in the UK

ID Statements 1
The more wall lizards the merrier 39 There is a bias against alien species 2 Wall lizards are much prettier to look at than our native lizards

40
We see less sand lizards due to loss of habitat so the wall lizards are a nice substitute It is good to see wall lizards appear to be thriving here 9 Wall lizards feel almost out of place here 47 It is a pleasure and a privilege to observe wall lizard behaviour

10
Wall lizards really shouldn't be here 48 I find wall lizards intriguing and interesting.

11
These wall lizards are invading the UK 49 Wall lizards are a great subject to photograph

12
Prefer if wall lizards were in their own environment best suited to their well-being

50
It's great for the kids to have wall lizards in the garden

13
If wall lizards don't cost anything (financially) then their presence is not a problem 51 I don't mind that wall lizards are not native to the UK

14
Wall lizards co-exist quite happily with common lizards on the continent so there's no reason why they wouldn't here.

52
No problems with wall lizards being here as far as I am aware

15
Wall lizards seem harmless enough.

53
Nice to see wall lizards but I'd rather see a native lizard

16
I feel wall lizards are not competing with our native species.

54
The wall lizards always give a topic of conversation when they come out in the warmer weather

17
I cannot see that wall lizards are likely to become a problem

55
It's a shame people don't consider wall lizards when repairing/pointing walls

18
Wall lizards are not detrimental to our gardens 56 I would make special trips to see wall lizards

19
I think wall lizards should be protected here 57 Don't know how I feel about wall lizards being here 20 I am always unhappy about ANY introductions, which are usually bad news for native species.

58
I am concerned there are likely to be more currently unrecorded wall lizard populations.

21
The presence of species which did not arrive here under their own steam is a concern 59 I wish we had wall lizards in the garden. Can I have some

22
It would be a shame if native lizards got edged out of our landscape by wall lizards.

60
The wall lizards provide the only chance we get to see lizards

23
I don't want wall lizards to damage native species 61 I don't want wall lizards to be removed.

24
I'm concerned about Wall Lizards affecting native Sand Lizards and Common Lizards both in terms of competition and pathogens.

62
It's really nice to have this wildlife (wall lizards) in our garden

25
I do wonder if global warming is a factor to wall lizards being here

63
No strong feelings about wall lizards one way or another 26 I have nothing against wall lizards but would like to see some scientific studies investigating their impacts

64
Classifying wall lizards according to our standards of whether or not they should be here is not useful

27
It is worrying that wall lizards are surviving in a colder climate than they are used to.

65
Many non-native species have become established here, wall lizards are just another

28
We should stop wall lizards spreading if possible 66 The wall lizards add to the character of our garden and village (Continues) encouraged to rearrange the position of statements until they were satisfied their placement represented as closely as possible their personal point of view. Each Q sorting was combined with discussion with the researcher during the process where participants were asked to elaborate on his/her point of view, explain the most salient statements and discuss whether there were any themes not represented by the items in the Q set. Participants (n = 13) completing their Q sort online (Q-software [Pruneddo, 2013]) followed exactly the same procedure by dragging and dropping items into their desired arrangement in the fixed distribution. Follow-up discussions were conducted via email.

| Ethics statement
This research received ethics approval from the Biological Sciences Faculty Research Ethics Committee, University of Leeds. Written consent was secured in advance of every Q sort undertaken in person. Those participating online consented by proceeding to the start screen following the introductory brief. The following statement was included in the introductory brief: I agree to complete this online/in-person questionnaire for research purposes and that the aggregate anonymous data derived from this questionnaire may be made available to the general public in the form of public presentation, report and journal article.

| Statistical analysis
A total of 26 Q sorts were intercorrelated and factor-analysed using the dedicated computer package PQMethod (Schmolck, 2002). Factors were extracted (centroid analysis) and rotated using an initial varimax rotation followed by additional by-hand adjustments to ensure the maximum number of participants could be included within the Q sort groupings across factors and to bring said groupings 'into focus' (Brown, 1993;Watts & Stenner, 2012).
In this case, an anticlockwise rotation of −2° was applied to factors 1 and 2. The common criterion in deciding how many factors to retain for rotation is for the eigenvalue of each factor to be greater than 1 (Addams, 2000;Brannstrom, 2011 Conceptually, Q sorts that load significantly on a particular factor (i.e. factor defining Q sorts) do so because they exhibit a very similar sorting pattern and therefore share a distinct viewpoint in respect to the presence of wall lizards in the UK. These defining Q sorts were then merged to form a single idealisedtypical Q sort for each factor called a factor array. The factor array looks like a single complete Q sort and is calculated by a procedure of weighted averaging (i.e. defining Q sorts are given more weight in the averaging process since they better exemplify the factor (Watts & Stenner, 2012; Figure 1). Factor arrays then provided the basis for interpretation of each individual factor by means of a careful and holistic inspection of the patterning of items in each factor array using a crib sheet system (Stenner, Cooper, & Skevington, 2003;Watts & Stenner, 2012).
Crib sheets enable factor arrays to be systematically organised allowing identification of the important issues about which a given viewpoint is polarized, and how the viewpoint is polarized relative to the views defining the other factors. The interpretation aimed to uncover, understand and fully explain the viewpoints captured by each factor and thus shared by significantly loading participants. Credence was given to correct interpretation by insights gained from the open-ended discussion held with participants during or after (online) their sort was completed.
Comments made by participants are quoted where they clarify the interpretation.

| RE SULTS
Each of the factors extracted from the analysis has been given a summary title and a textual interpretation of the subject viewpoint which the factors express. To ease interpretation, numbers in brackets have been included to refer to the statement number (in bold; see Table 1) and the array score. For example, (8, −3) refers to a score of minus three (relatively strong disagreement) given to statement eight for the particular viewpoint/factor.
The factor analysis revealed three lines of discourse relating to the presence of wall lizards in the UK. These three factors together explained 53% of the study variance. Twenty-three of the 26 Q sorts loaded on one or other of these three factors. Factor loadings of ±0.31 or above were significant at the p < 0.01 level (Brown, 1980).
However, due to several confounding Q sorts (Q sorts with significant loading on more than one factor) in the solution using this loading parameter, the level of significant factor loading was raised to >0.5 ( Factor characteristics are also summarised in Table 3. In this threefactor solution, factor 3 had a negative correlation with factor 1 (−0.16) and factor 2 (−0.37). Factor 1 was positively correlated with factor 2 (0.79), indicating commonalities between the constituent defining Q sorts (Table 2).

| Factor 1: 'Innocent until proven guilty'
This viewpoint identifies a lack of personal knowledge and stresses need for evidence of specific impacts of wall lizards on native fauna (29,6; 26,5; 45,5; 34,4; 57,1) in order to be able to express strong positive or negative feeling towards the lizards (63,0). Despite this lack of knowledge it is considered that, in principal, the introduction and presence of wall lizards in the UK is not a good thing (10,3; However, there is a feeling that in the case of wall lizards their ecological role is far more pertinent in making a judgment on their presence than the species origins alone (34,4), and that there is an TA B L E 3 Rotated factor matrix and factor characteristics following Q sort analysis pertaining to attitudes towards introduced wall lizards in the UK. Values in bold indicate significant loading for given factor. Asterisk denotes Q sorts that did not load on any factor Q SORT Factor 1 'Innocent until proven guilty'

Factor 2 'Precautionary, informed concern'
Factor 3 'The more the merrier!' 3,−4) and thus control of the species (on condition of proven impact) may be acceptable to this group.

| Factor 2: 'Precautionary, informed concern'
Whilst similar to factor 1 in many ways, the factor 2 perception var-

| D ISCUSS I ON
In this study, we used Q methodology to identify and describe stakeholders' perspectives towards the established presence of a non-native lizard species introduced to the UK. We were able to extract three clearly defined viewpoints on the species' introduction, which term: 'Innocent until proven guilty', 'Precautionary informed concern' and 'The more the merrier'. These perspectives reflect both differences and commonalities in stakeholder perceptions and opinion regarding the species' presence, which are discussed here with reference to the conceptual framework for understanding social perceptions towards NNS described by Estévez et al. (2015).
In comparing and contrasting the three viewpoints, we identified four key areas of disagreement, specifically 'acceptance of wall lizards', 'concern about the ecological threat posed by wall lizards', 'attitudes toward NNS in general' and 'opinion towards management/ control of the species'. These four areas of disagreement are inextricably linked and may arise due to variation in the level of knowledge (actual or perceived) and uncertainty of actors within each group regarding the ecological impact posed by wall lizards. Accordingly, theory holds that when faced with insufficient information individuals will process this limited information and develop their judgement heuristically (Trumbo, 2002). Levels of knowledge (or lack thereof) and uncertainty are thus likely to be the main drivers shaping all subsequent opinion within each viewpoint (i.e. whether or not the lizards are welcomed, support for management/control) as individuals rely on heuristic value methods to reach their conclusions and assessment of risk (Davidson et al., 2013;Kahan et al., 2012). Defining themes and values between groups are discussed below.

| How and why do stakeholders group differ in their opinions?
Both the 'Innocent until proven guilty' and the 'Precautionary, informed concern' views appear to build their opinions by drawing on a source of theoretical knowledge of species introductions and the associated potentially negative ecological implications. This is not surprising considering the majority of actors (84% across both views) expressing these views come from an environmental science background and are likely to be familiar, to varying extent, with concepts and terminology regarding NNS . As a result, both views express elements of a precautionary approach to NNS introductions conforming to some degree with guiding principles of best practice that permeate through such professions (DEFRA, 2003). There are however significant differences in perceived knowledge and levels of uncertainty between the two groups.
The reluctance for the 'Innocent until proven guilty' view to make a judgment in the case of wall lizards is due to self-confessed personal knowledge gap regarding the species' impacts, suggesting participants were more likely to evaluate the species via heuristic methods (Davidson et al., 2013;Trumbo, 2002). This makes sense considering the majority of the participants holding the 'Innocent until proven guilty' view come from backgrounds conferring a broad ecological knowledge rather than being specialist in a particular field (i.e. ecological consultants, land managers). Interestingly, the 'Innocent until proven guilty' view also hints at a deviation (at least where wall lizards are concerned) from the normative assumption that all NNS should be treated as a potential threat (Rejmanek & Simberloff, 2017;Simberloff et al., 2011). Instead, this discourse is inclined to agree that there is a pervasive bias against NNS and that detailed case-specific knowledge is a prerequisite for judgements about whether a particular species' should be labelled a threat or not (Davis et al., 2011;Guerin, Martín-Forés, Sparrow, & Lowe, 2018;Van Der Wal, Fischer, Selge, & Larson, 2015;Warren, King, Tarsa, Haas, & Henderson, 2017). This finding is very much the basis for the viewpoints' 'Innocent until proven guilty' label and is the traditional and assumed approach in empirical scientific research and, within this, impact studies (Davidson et al., 2013;Mapstone, 1995).
Knowledge gaps and uncertainty are recognised as important factors in predicting concern about NNS impacts (Gozlan et al., 2013;Verbrugge et al., 2013). Biosecurity experts assessing the risk posed by aquatic non-native species tended to assign lesser concern about potential impacts when faced with little information and other uncertainties about species traits (Davidson et al., 2013). In this study, when faced with great uncertainty over impacts, the 'Innocent until proven guilty' view also appears to use another heuristic influence to form its perceptions of the lizard introduction. This appears to be formed on the relationship between the positive value (benefits) exponents place on engagement opportunities with the lizards and the perceived ecological risks of the species. Judgement based on this emotive 'affect heuristic' follows theory of a negative correlation between risk and benefit in the decision-making process (Finucane, Alhakami, Slovic, & Johnson, 2000; i.e. positive feelings garnered from opportunities for engagement with wall lizards translates to lower perceived risk).
In contrast, the theoretical knowledge drawn upon to form the 'Precautionary, informed concern' viewpoint appears to be supported by heuristics based on direct experience and/or specialist knowledge leading to less uncertainty and more concern about the potential impacts of wall lizards on native fauna, thus informing a more hard-line precautionary approach to the introduction than that of the 'Innocent until proven guilty'. In this case, negative feelings towards the wall lizards arising from observed or anecdotal evidence of ecological impact has resulted in an opposite pattern to 'Innocent until proven guilty', whereby negative feeling and experience translates as higher perceived risk (Finucane et al., 2000). Those holding the 'Precautionary, informed concern' view are also those most likely to have specialist knowledge of invasion biology and/or reptile ecology (i.e. environmental advisory, conservation NGO, academia) and be in professions where standards (in this case the precautionary approach) are often ingrained, creating a filter through which the individual perceives risk, often unconsciously and through conformance with policy (Sjoberg, 2002). 2008). The same study also identified two different conservation professional groups defined by slightly divergent viewpoints very much akin to those of the 'Innocent until proven guilty' and 'Precautionary, informed concern' groups emergent in this study.
Similarly, both studies found a small subset of the public (nature aware/nature tourists) having a shared perception of NNS with conservation professionals. Other studies have also found little divergence between public and professional views on NNS and go further to argue that although the content of their thoughts might diverge slightly, ecological professionals and the lay public essentially share the same structure of thought about the natural environment in general (Fischer et al., 2014;Van Der Wal et al., 2015). Our findings do not entirely support this, as the 'More the merrier' viewpoint was held by a majority of lay public participants who clearly constructed their opinions from different values and influences to those in other groups.

| What does the discourse in this case study tell us about perceptions and attitudes towards management of introduced species more generally?
The discourse emerging in this study provides an insight into attitudes regarding the potential management of an introduced species' when faced with little knowledge or evidence of negative impact.
Our findings are in concordance with higher levels of knowledge (of biological invasions, ecological principles) being associated with increased support for NNS management options (Bremner & Park, 2007;García-Llorente et al., 2008)-in this case adoption of either the precautionary ('Innocent until proven guilty', 'Precautionary, informed concern') as opposed to the 'hands off' ('More the merrier') approach. Perceptions of risk, abundance and detrimental impacts have also been seen to strongly inform participants' attitudes towards management of NNS , with some authors suggesting that these factors, rather than non-nativeness, have the greatest influence on judgment (Estévez et al., 2015;Gobster, 2011;Van Der Wal et al., 2015). In our analysis, the origins of the wall lizards per se certainly appear to be of less concern in this overall discourse compared to perceived risks and impacts. This is most apparent in the 'More the merrier' viewpoint where the foreign origins of the lizards have no bearing on the perception of there being no negative ecological impacts, leading ultimately to the 'hands off' approach. Meanwhile the 'innocent until proven guilty' group are more concerned with the potential impacts of non-native species, but appear to evaluate on a case-by-case basis rather than assuming that all non-native species are inevitably detrimental (Davis et al., 2011). Although beyond the scope of this discussion, our findings are indicative of the wider discussions on the dichotomy of native species-good, non-native species-bad and the deeper arguments regarding ideologies of 'nativeness' and 'naturalness' (Goodenough, 2010;Low, 2007;Rejmanek & Simberloff, 2017;Schlaepfer, Sax, & Olden, 2012).
The wall lizards evidently have a charismatic appeal, as shown by the items of consensus amongst the three viewpoints (Table 3).
Attractiveness and charisma are well recognised as key influences on peoples attitude towards to NNS (Bremner & Park, 2007;Fischer et al., 2014;Verbrugge et al., 2013), and can thus have significant bearing on support for species management operations (Estévez et al., 2015). Although, as is the case in this study, others have found that whilst those from ecology backgrounds appreciate these values they often regard them as ambiguous attributes that are not necessarily a legitimate criterion to which to base their judgments about species management . In this study it is the 'More the merrier' viewpoint again that appears most influenced by these attributes, perhaps because the wall lizards are amongst the most conspicuous of wildlife regularly encountered by this group.

| Limitations
It is important to recognise that the Q methodology and subsequent analysis can provide only a limited description of the prevailing discourses surrounding a topic at a given time and place, and that the resultant interpretation is not directly transferable. For example, notably different discourses may be evident surrounding introductions of P. muralis elsewhere (i.e. Vancouver Island, BC, Cincinnati, USA).
Whilst every care was taken to remove ambiguous statements from the Q set and make the Q sort as intuitive as possible to all participants, we recognise that this is difficult to achieve when studying responses of such a heterogeneous group of participants. As a consequence there is the possibility that some statements may have been perceived as having little relevance to an individuals' personal experience and would thus have been difficult to rank objectively.
Our Q set of 76 statements could also be considered fairly large.
Although this lends confidence to having obtained a comprehensive representation of the available discourse, engagement with the Q sort process may have been adversely affected as a result. This may be particularly true for those Q sorts conducted electronically in the absence of the lead researcher. In addition, the simple 'drag and drop' method of sorting statements used by the software may have reduced the diligence applied to rearranging such a large number of statements into the final sort arrangement.

| CON CLUS I ON S AND WIDER IMPLIC ATIONS
The holistic method of interpreting the analysis gives insight into how and why stakeholders may have formulated certain viewpoints regarding their perceptions towards introduced NNS. This in turn could help conservation managers identify ways in which to work with these subjective influences in order to best communicate the dichotomies and complexities surrounding the introduction of NNS, with an aim towards a more informed and balanced discourse. Our analysis of the discourse highlights three interesting issues that relate not only to the wall lizard introduction, but approaches to NNS and their management more generally. Firstly, there is significant variation between stakeholder groups regarding the presence and management of NNS. The analysis flags early signs that opposing views between a subset of the public and decision-makers has potential to present obstacles should management of the species ever be considered justifiable and practical.
Indeed, with the majority of the UK wall lizard populations being found in residential or busy public areas, any operation involving the lizards is likely to be met with some opposition from those with a protectionist view that may not be equipped, or care, to reconcile with views of conservation managers (Temple, 1990). Furthermore, finding such strong positive sentiment towards the lizards amongst a subset of the public holding a possessive view about the wildlife they frequently encounter is illustrative of a mindset that could facilitate the spread of a charismatic NNS where ecological impacts are not known, or are perhaps considered but disregarded.
Secondary human movement of wall lizards from established populations is the most likely pathway for rapid range expansion of wall lizards into new areas of the UK (Foster, 2015;Michaelides et al., 2015). Secondly, the discourse analysis illustrates that awareness of the wider concepts of NNS (invasion ecology) is lacking, particularly amongst the general public, and suggests that in the public domain scientific evidence alone may not be sufficient to inform perceptions of risk. In this regard, engagement with a characterful, conspicuous NNS may provide useful opportunities not only to educate the public on the ecological concepts and dichotomies associated with NNS and INNS, but also to promote wider interaction with nature and general interest in conservation. On another level, our analysis shows that communication of scientific evidence between scientists and conservation managers implementing policy is wholly necessary in order for personal judgment of NNS risks and management decisions to be made. Finally, the discourse highlights that a softer view of NNS, one that does not assume negative impacts in the face of limited evidence, is held personally by some professional's in contention to the policy of a precautionary approach that governs their profession.

CO N FLI C T O F I NTE R E S T
Nothing to declare.