Should we feed wildlife? A call for further research into this recreational activity

Independent human–wildlife feeding interactions (i.e., the feeding of wildlife by the public outside of organized ecotourism activities) represent an increasingly common way in which humans and wildlife are engaging with one another. It is important to determine what effects these interactions are having on the wildlife involved in order to ensure that optimum coexistence scenarios are being achieved, however the nature of these interactions makes them notoriously difficult to study. Extrapolation from ecotourism activities has suggested detrimental impacts on the health and fecundity of the populations involved. Moreover, recent findings indicate that bold individuals may actually gain reproductive advantages from this excess food supply, driving artificial selection of risk‐taking behaviors within human‐dominated landscapes. Humans may, therefore, be unknowingly manipulating wildlife populations, forcing them into unnatural states, and potentially decreasing future viability. Here, we outline key literature pertaining to the potential impacts of these self‐led interactions on wildlife and address the need for further research into the associated effects. Due to the associated safety risks to the humans involved in these interactions, and by applying the precautionary principle until further research can be performed, we recommend that management actions be employed to actively reduce their occurrence. We address current management practices in use and make recommendations for further research to adapt and improve them. Ultimately, we make a call for further research addressing two fundamental key areas: (i) to explore the effects of these interactions on the wildlife involved, across different species and habitats experiencing this phenomenon, with emphasis on the potential role of artificial selection, and (ii) to work to improve the management practices currently employed to reduce the occurrence of these interactions, at least until such time as the effects of these interactions on both humans and wildlife have been thoroughly disentangled, with the overall goal of improving coexistence.


K E Y W O R D S
artificial selection, human-wildlife coexistence, human-wildlife conflict, media communication, recreational activity, wildlife feeding, wildlife management 1 | HUMAN-WILDLIFE INTERACTIONS: A UBIQUITOUS PHENOMENON Human-wildlife feeding interactions do not represent the first or only medium for contact between humans and wildlife species, though they do represent one of the most popular forms of human-wildlife contact in modern times. Throughout evolutionary history, mankind has experienced continuous contact with wildlife (Bhatia et al., 2020). This relationship is highly dynamic, ranging from periods of tolerance to conflict. Specifically, humans have historically alternated between phases of appreciation, reverence, utilization, acceptance, and retaliation, with this variation in our regard for wildlife driving how we interact with them (Bhatia et al., 2020;Ingold, 2002;Treves & Naughton-Treves, 1999). This is ultimately driven by the context in which we find ourselves in contact with wildlife species, which can vary enormously between different communities depending on spatial, geographical, and temporal factors.
For example, conflict scenarios arise when the actions and movement of humans and wildlife result in adverse impacts on one another (P atru- Stupariu et al., 2020;Soulsbury & White, 2015). People living in rural or agricultural landscapes may experience conflict with local wildlife populations, with negative impacts for both the humans and wildlife involved. This conflict may occur due to the killing of livestock by wild predators (Butler, 2000;Holmern et al., 2007) or due to the damage and loss of crops by foraging mammals and birds (Conover, 1994;Kale et al., 2012;Naughton-Treves, 1998). These occurrences result in economic losses for farmers and often lead to the persecution of the wildlife involved (Cerri et al., 2017;Gordon, 2018). Similarly, in urban landscapes, wildlife may be regarded in a negative light as they scavenge through human refuse and waste (Newsome et al., 2015;Newsome & Van Eeden, 2017), resulting in disease spread and nuisance behaviors such as property damage or aggression (Guenther & Shanahan, 2020;Soulsbury & White, 2015). Humanwildlife relationships, however, are not always negative; species also provide important ecosystem services through pest control (Greenberg et al., 2000;Soulsbury & White, 2015) and waste disposal (Abay et al., 2011;Soulsbury & White, 2015), resulting in more positive public opinion. Wildlife has also played an important role in cultural and religious beliefs throughout history, and different species are regularly depicted in religious texts and in myths and legends (Methorst et al., 2020). For this reason, many communities and ethnographic groups may gain a sense of inspiration, spirituality, or connection by living in proximity to cwildlife species. Outside of religious and cultural circumstances, many people also report positive emotional responses from being near or seeing wildlife, including awe, excitement, and pleasure (Dou & Day, 2020;Farber & Hall, 2007;Soulsbury & White, 2015).
As urbanization has resulted in a global shift toward city-dwelling, people have begun to report a sense of disconnect from nature, termed the "extinction of experience" (Lopez et al., 2020). This has coincided with a shift toward public demand for interactions with wildlife (Curtin, 2005). This desire for contact with the natural world has resulted in increased interest in wildlife-based activities. For example, there has been a documented rise in interest in bird watching (CBI, 2021;Lopez et al., 2020), increased engagement with wildlife viewing and photography (Orams, 1996), demand for more and larger species in zoos (Mooney et al., 2020), appearances of wildlife cafés (McMillan et al., 2021), and rapid continued growth in the ecotourism sector, which is estimated to reach a value of over US$385 billion by 2028 (Million Insights, 2022).
One activity associated with ecotourism that has gained international traction and popularity is the feeding of wildlife (Newsome & Rodger, 2013). This occurs both indirectly, for example through baiting in order to provide better viewing opportunities, and directly, for example by allowing tourists to feed the animals in order to facilitate closer contact (Dou & Day, 2020). However, this "provisioning" of wildlife as part of a tourism activity is highly controversial (Dou & Day, 2020;Dubois & Fraser, 2013a). This is due to evidence that wildlife may become reliant on these artificial food sources and that these sites may attract abnormally high concentrations of wildlife (Higginbottom, 2004;Newsome & Rodger, 2013). These concerns have resulted in it being banned in several areas (Dobson, 2006;Dou & Day, 2020). However, this has not prevented it from filtering into the public domain, and it is becoming increasingly common for the wider public to engage in wildlife feeding activities independently.

| THE RISE OF INDEPENDENT HUMAN-WILDLIFE FEEDING INTERACTIONS
The feeding of wildlife as a self-motivated, independent activity is not a new occurrence; for example, the feeding of garden birds has been a universally popular activity for many decades (Jones, 2011;Robb et al., 2008). However, wildlife feeding has now expanded to encompass multiple species across various habitats and countries ( Figure 1); performed opportunistically at roadsides, within public spaces and parks, and on private land (Dubois & Fraser, 2013a). Species reported as targets for these interactions include chipmunks (Dubois & Fraser, 2013a;Marion et al., 2008), primates (Baker & Dubois, 2012;Pragatheesh, 2011), racoons, skunks, coyotes, deer (Dubois & Fraser, 2013a), dingoes (Burns & Howard, 2003), and bears (Dubois & Fraser, 2013b), among many others.
While it is possible that these activities gained public attention due to their popularization as an ecotourism activity, potentially stimulated further by the rise of social media (Pagel et al., 2020), these independent feeding interactions are markedly different in how they function from ecotourism feeding. For example, ecotourism activities typically include a level of supervision and organization, meaning that appropriate food items can be provided and inappropriate behaviors can be reprimanded or controlled (Clua, 2018;Orams, 1996). Despite this, even within ecotourism, as mentioned, the provision of food is now often prohibited (Dobson, 2006;Dou & Day, 2020). In independent feeding, there is no control of food provision or human behavior, therefore the consequences could be vastly different from those documented in ecotourism activities. However, due to the nature of their occurrence, the effects of selfmotivated feeding activities are difficult to unravel. Therefore, their impacts are virtually unknown, leaving little consensus on whether to view the practice as positive or negative (Soulsbury & White, 2015). Robb et al. (2008) has urged for research exploring whether these interactions may be driving artificial selection processes, with the goal of unraveling their role in wild populations. This has prompted the push to study how wild, freeliving animals respond to, and are affected by, these unmonitored activities (Goumas et al., 2020), which has motivated our scoping review.

| METHODS
This is a scoping study that reports on the literature pertaining to human-wildlife interactions, specifically independent feeding interactions, with a focus on identifying key concepts and gaps in our current knowledge of this field. In line with the exploratory nature of this form of F I G U R E 1 Images depicting individuals recreationally hand-feeding a variety of species across different habitat types, including squirrels (credit: cottonbro), river dolphins (credit: Lucia Barreiros Silva), macques (credit: Rachel Claire), seabirds (Yura Forrat), and kangaroos (credit: Valeriia Miller). All free-to-use images were sourced using www.pexels.com). study, key terms were identified and searched using the Web of Science database in early November 2022.
Two broad searches were organized in line with our two main research areas of interest, both of which are subdivided into two key questions each: 1. The ecological aspects: What are the drivers and dynamics of independent human-wildlife feeding interactions? What are the associated effects on the targeted wildlife? 2. The human and management aspects: How are these interactions currently being managed? Is this management effective or how could it potentially be improved?
The final key terms searched for the first topic were "Wildlife AND feeding AND tourist OR wildlife AND artificial feeding" with the dropdown set to "All Fields" for each term. This search produced 467 papers, which were then manually screened for relevance. All irrelevant papers and duplicates were removed, resulting in a final total of 36 papers for review. Previous reviews of humanwildlife feeding activities have encompassed a range of feeding types, including organized ecotourism activities, baiting for tourism, and supplementary food stations set up for management purposes. However, here we focused explicitly on literature regarding independent wildlife feeding, therefore search results pertaining to supplementary feeding for management and ecotourism activities were removed. The key terms searched for the second topic were "Management AND human AND wildlife AND feeding AND interactions OR managing AND tourist AND feeding AND wildlife" with the dropdown set to "All Fields" for each term. This produced 198 papers, which was reduced to 16 after initial screening, with 3 additional papers being added due to subsequent keyword screening on Google Scholar for publications not available on Web of Science, bringing the total to 19.
These terms were selected as in initial screening they encompassed phrases and terminology most commonly used in papers on the topic of independent wildlife feeding and its associated management. Notably, we still incorporated the term "tourist" into our search regarding the effects of feeding, as independent feeding often occurs across popular tourist sites, such as national and urban parklands. If a point was briefly made in a paper, but it was mentioned that it was expanded upon in a paper that did not appear in the aforementioned literature search, then this paper was then manually searched and examined accordingly. This happened if a paper referenced results from studies on conservation supplementary feeding or organized ecotourism activities, which serves to again highlight to lack of literature on independent feeding interactions alone. This is also indicative of how often researchers must extrapolate from these alternative types of feeding studies due to the deprivation of studies directly addressing independent feeding interactions. In fact, both the 36 papers addressing the effects of independent feeding and the 19 papers addressing their management were published from 2002 to 2022, with only one outlier regarding the behavioral impacts of independent feeding being published prior to this in 1994. These studies, which focus on independent feeding interactions as opposed to organized tourism activities or supplementary feeding by wildlife managers, are relatively few considering the 20-year timeframe across which they occur.

| Behavioral dynamics in targeted wildlife
It has been shown that wild individuals that engage regularly in feeding interactions, both in the context of tourism interactions and independent interactions, eventually show changes from their natural behaviors. For example, they may display increased aggression, both toward other wildlife species (Wrangham, 1974) and toward humans (Burns & Howard, 2003;Soulsbury & White, 2015). Human presence may also result in increased stress levels, with knock-on physiological impacts (Ellenberg et al., 2007). The availability of human-provisioned food may also result in changes to natural activity patterns (Senigaglia et al., 2020) and decreased home range size (Craighead et al., 1995). Yet, the majority of what we know about these effects of wildlife feeding are from studies based on ecotourism feeding activities and other managed forms of feeding. Therefore, it is of key importance that future studies aim to unravel the effects explicitly of independent feeding activities, in order to determine what the true effects are.
Additionally, understanding what drives engagement by wildlife in these independent interactions in the first place remains a significant gap in our knowledge.
Traditionally, a wild animal's likelihood to engage with humans has been viewed as a conditioning process (Senigaglia et al., 2020). Researchers argued that willingness to interact was the direct result of the amount and type of interactions that the animal had been exposed to previously. This could be broadly categorized into avoidance (i.e., aversion to a stimulus), habituation (i.e., the waning of a response to a repeated stimulus), and attraction (i.e., the movement toward a stimulus) (Whittaker & Knight, 1998). However, recent research has established the presence of natural inter-individual variability, which can drive behavior in different ways (Ascensão et al., 2014;Carrete & Tella, 2011;Forsman & Wennersten, 2016). For example, bolder personality types are associated with increased use of, and better adaptation to, urbanized areas (Honda et al., 2018). It has also been shown that, certain species may gain advantages from proximity to humans (Carlson, 1985) and that prey species may take advantage of a "human-shield" when utilizing human-dominated landscapes which are often unavailable to large predators (Berger, 2007;Muhly et al., 2011). Bolder personality types are also associated with riskier behaviors in response to predator presence (Quinn & Cresswell, 2005). Therefore, individuals within single populations of prey species may utilize humanshields in different ways depending on personality and associated perception of risk. As approaching humans may be perceived as a high-risk activity for wildlife (Miller et al., 2001), this calls into question whether there is also natural variation in likelihood to engage with humans for food within targeted populations (Goumas et al., 2020).
In other words, while wild animals may all have the potential to become habituated to human presence, we need to explore whether a collection of habituated individuals all behave the same way in terms of engagement in feeding interactions (Donaldson et al., 2012), or whether natural disparity exists. Natural variation in engagement with feeding interactions has recently been identified in wild fallow deer , with only a bolder subgroup of the overall population repeatedly approaching people for food offerings. This subgroup, therefore, has access to a vastly different diet than other individuals in the population that perform more natural foraging behaviors, with associated impacts on physiology and reproductive outputs (see section below and references therein). However, as of yet, whether this variation in behavior exists within other single populations or in other species has not been explored.
If feeding processes favor one species over another, or if certain individuals within a single species or population are favored, then this could have long-term impacts on populations in the form of artificial selection (Robb et al., 2008). If these selection pressures favor one personality type (i.e., the bolder individuals) over others, resulting in loss of overall variation in personality types across a population, then there could have long-term consequences for the plasticity and resilience of the population (Wolf & Weissing, 2012). Notably, however, there may also be additional factors that could influence individuals' likelihood to engage in interactions, which must also be explored in order to get a fuller understanding of what drives wildlife engagement with humans for food. These may include temporal effects, for example as natural food sources vary in availability over the course of the year, or life history effects, such as injury or loss of maternal care. The nature of independent feeding interactions makes it difficult to track entire free-living populations exposed to the same opportunities and pressures, but it is of fundamental importance that variation in engagement with these interactions be further assessed.

| Potential effects on wildlife physiology and reproduction
If variation in engagement with human feeding is occurring, then it calls into question whether this difference in resource utilization, and therefore diet, is having a knock-on effect on individual physiology and reproduction, as is evidenced by Griffin, Haigh, Amin, et al. (2022). If these knock-on effects are occurring, and this impacts individual fitness, then artificial selection is at work. However, it may be selecting for or against this, potentially bolder, behavioral group depending on whether it is advantageous or detrimental to the health and reproductive fitness of the subset of individuals involved.
There are a number of potential physiological impacts associated with the feeding of wildlife by humans. However, these have been primarily extracted from studies exploring the effects of provisioning as a form of conservation management or on controlled ecotourism activities (Robb et al., 2008), which are not a good model for independent feeding studies. Additionally, results are often conflicting (Orams, 2002). For example, studies on ecotourism feeding have reported negative health impacts, including the development of skin lesions, excessive fat deposits and parasite burdens, stomach ulcers, and decreases in overall health due to malnutrition (Brookhouse et al., 2013;Murray et al., 2016). Independent feeding of ruminants has been shown to result in papillae proliferation in the rumen, likely due to increased acid production (McLaughlin et al., 2022). Ecotourism feeding studies have also reported a high risk of injury to wildlife (Christiansen et al., 2016;Donaldson et al., 2010;Dubois & Fraser, 2013a) and pathogen transmission (Murray et al., 2016). Moreover, the decrease in home range size associated with any human-driven feeding also increases the risk of inbreeding Newsome & Van Eeden, 2017). Some ecotourism studies have also reported decreased fitness, reproductive success, and maternal care in certain sites, resulting in higher levels of offspring mortality (Fitzpatrick et al., 2011;Foroughirad & Mann, 2013;Orams, 1997). Loss of genetic diversity in populations that isolate themselves to feeding sites has also been documented (Afonso et al., 2021).
Contrary to these reports of negative impacts, supplemental feeding, when performed using best practice as part of conservation management programmes, can have a positive effect on the health and weight of individuals in targeted populations (Murray et al., 2016). For example, it has been linked to the advancement and extension of breeding seasons (Orams, 2002), can enhance reproductive success in times of limited resources (Dubois & Fraser, 2013a), and may aid over-winter survival (Robb et al., 2008). A similar increase in fecundity has also been reported in some species that feed from human waste in urban areas (Beckmann & Lackey, 2008), though this has led to concerns about topdown impacts on other species and vegetation (Soulsbury & White, 2015). Additionally, any increased fecundity may also be negated by human-caused mortality (Beckmann & Lackey, 2008), for example, through lethal control if an animal becomes dangerously habituated (Dubois & Fraser, 2013b) or through hunting activities if habituation results in decreased vigilance (Goumas et al., 2020).
These positive effects may also result in the artificial selection of certain (i.e., bolder) behavioral types, which may impact population viability in the long-term . If engagement with feeding interactions is not random, and instead is driven by behavioral type, then they may also only be occurring in a subset of the population. Recent studies have indicated that independent feeding of wildlife can impact the weight and growth rates of offspring (Griffin, Haigh, et al., 2023;Griffin, Haigh, Amin, et al., 2022) (Figure 2) and that the resultingly larger neonates have increased survival rates . This highlights the potential that artificial selection may be at work and we may involuntarily be manipulating wild populations. It is of fundamental importance that this be further investigated through research across other species and populations. Goumas et al. (2020) states that there is a "need for long-term studies on free-living, wild animals to fully understand the causes and ecological consequences of variation in [wild individuals'] responses to human cues." They conclude by stating that the effects of humans on wildlife are "likely to be substantial" and that "a detailed understanding of these effects is key in implementing effective conservation strategies and managing human-wildlife conflict." This prompt is directly relevant to independent human-wildlife feeding interactions, in which we need to unravel what is driving them and what the associated effects on targeted individuals, populations, and ecosystems are.

| The human element of independent feeding interactions
Understanding the dynamics and impacts of these independent human-wildlife feeding interactions is fundamental in informing their management. Further to this point, whether or not the independent feeding of wildlife results in negative impacts on the wildlife species involved needs to be explored further. Until such a time as this has been effectively discerned, employment of the precautionary principle constitutes enough F I G U R E 2 Images depicting visitors to Phoenix Park, Dublin, recreationally feeding a female fallow deer mother (Dama dama) (left) and a neonate fallow deer fawn hidden in the undergrowth (right) (Credits: Laura L. Griffin). Fawns remain hidden for the first few weeks of life, during which time they are regularly visited by the mother for feeding bouts, until they have reached an appropriate size to join the herd. motivation to instigate management aiming to reduce their occurrence (Orams, 2002). Additionally, these interactions constitute a social-ecological system, with drivers and implications for both the wildlife and humans involved. As mentioned, people initiate these interactions due to associated mental health benefits and feelings of joy resulting from close proximity to nature (Farber & Hall, 2007;Soulsbury & White, 2015). However, there are also many potential risks to the humans involved that may exacerbate conflict with the associated wildlife and, therefore, underline the need for management actions aiming to reduce risky close contact human-wildlife interactions.
The feeding of wildlife may result in wild animals associating people with food. Ecotourism feeding studies have highlighted that wild individuals may begin to display aggression toward humans or initiate pushy interactions, putting people at risk of injury (de S a Alves et al., 2012; Orams et al., 1996;. Even where injuries do not occur, these behaviors can manifest in wildlife attempting to engage with reluctant humans. This is something that has also gained recent media attention as an effect of independent feeding interactions (Holmes, 2022;Smith, 2021), and could potentially exacerbate human-wildlife conflict through human fear responses, potentially resulting in calls for the removal of the animals involved (Soulsbury & White, 2015). Additionally, as highlighted by the concept of OneHealth, the physical health of both wildlife and humans is interdependent and close contact activities that may result in disease transfer between the two should be minimized (Decker et al., 2010). Finally, if feeding of wildlife does result in exponential growth of the population, then this may directly result in increased property damage, wildlife road collisions, or disease prevalence in the population or transfer to domestic animal populations due to increased contact rates. This can lead to the resentment of resident wildlife species by proximal human communities (Guenther & Shanahan, 2020;Soulsbury & White, 2015). Human perceptions of independent wildlife feeding, the psychological drivers behind it, and how the aforementioned ramifications may alter people's attitudes toward wildlife are topics that are gaining increasing attention. Large-scale, funded projects are recently becoming established to explore these topics, and it is important that research continues to be developed to fill these gaps in our knowledge (Wellcome, 2020).

| The management of independent feeding interactions
Ultimately, the potential impacts on both the humans and wildlife involved in independent feeding interactions mark them as a high-risk source of human-wildlife conflict. This has resulted in many authorities, wildlife managers, and ecologists calling for management to reduce their occurrence (Dubois & Fraser, 2013a). Wildlife feeding can be viewed as a process that is only acceptable where it can be monitored and controlled (Dubois & Fraser, 2013a), that is, where restrictions can be directly maintained by employees through the use of a regulatory framework, for example, a system of progressively harsher punishment for infractions (Orams, 1996), as may be the case during ecotourism activities.
The management of independent feeding, however, is a more complex matter. For example, controlled environments can be monitored and regulated through visitor number restrictions, limits on time slots, limits on noise levels, and the implementation of permits and licenses (Orams, 1996). These are not feasible tools in the management of independent feeding as it can occur at any time, in any area, or even on someone's own private property. As a result, studies that actively test controls aiming to reduce the occurrence of these interactions are limited, with our keyword search only producing six papers, all focusing on signage and aversive conditioning of wildlife (Table 1). The remaining 13 papers included reviews, papers assessing the opinions of stakeholders and motivators for feeders, and additional considerations for managers. Notably, while all 19 papers addressed the management of independent wildlife feeding, only 15 of these addressed it exclusively, with the remaining papers also including other feeding interactions, such as provisioning in wildlife tourism. This highlights the need for research that aims to actively test, and improve, management controls aiming to reduce engagement in these independent feeding interactions.
Food aversion or repellent treatments have been shown to successfully reduce the levels of engagement with humans in treated wild individuals (Conover, 1999;Johnson et al., 2022). However, this does not reduce human attempts to initiate feeding interactions, requires manpower, and requires direct and repeated sessions with all wild individuals in a population. For this reason, management campaigns aiming to reduce the occurrence of independent feeding interactions are typically reliant upon communicating with the public, and usually from a distance due to limited manpower, for example through signage. In recent years, this communication has seen a shift toward public education as opposed to simple statements of prohibition (Dubois & Fraser, 2013a;Orams, 1996;Orams & Hill, 1998;White & Gehrt, 2009).
The reasoning for this shift toward an educational framework in management is clear. First of all, simply explicitly stating that people should not feed wild animals is rarely effective under any circumstances (Marion et al., 2008;Newsome & Rodger, 2008). This aligns with findings from other human-wildlife interaction studies, where survey respondents reported a marked preference for information relating to education over simple statements of regulations (Rimmer et al., 2013). Secondly, there is now a push, whether that be in regional spaces, zoos, or national parks, to incorporate learning and facilitate attitude changes in public audiences (Orams, 1996). This engenders long-term change in human behaviors and may facilitate further information spread by the audience reached. This also aligns with the modern philosophy aiming to shift human-wildlife interactions from conflict to coexistence scenarios, ergo generating positive relationships with wildlife (Bhatia et al., 2020). In the case of independent human-wildlife feeding interactions, as the viewing of wildlife is linked to mental health benefits (Fuller et al., 2007;Maller et al., 2006), this sense of connection with nature should be promoted, while also enabling visitors to make informed decisions on why not to engage in other inappropriate behaviors, such as feeding (Guenther & Shanahan, 2020). In fact, educating the public in this manner can result in a sense of stewardship of the area, enabling them to inform other members of the public on why not to feed the wildlife, and thereby generating a state of self-management (Bhatia et al., 2020).
In recent years, extensive research has focused on exploring ways in which to improve upon the communication tools that are most commonly used in relation to managing wildlife feeding, with the ultimate goal of making them more effective. These primarily include traditional management tools, for example, signage, posters, and ranger talks, with studies focusing on testing audience responses to different designs and messaging (Ballantyne & Hughes, 2006;Marion et al., 2008). The findings of these studies are important, however, these forms of public communication continue to rarely, if ever, succeed in entirely eliminating wildlife feeding T A B L E 1 List of key papers revised in our scoping study, categorized by the topic that each individually addresses. We have summarized whether each tested the efficacy of a management tool or whether they only specifically addressed independent human-wildlife feeding interactions, without grouping them with ecotourism provisioning or management feeding (i.e., yes/no). (Clark et al., 2015;Parkin, 2001). It is, therefore, of fundamental importance that we continue to identify barriers to the success of these tools, or "controls," in order to advance public education and reduce the occurrence of feeding interactions. Lischka et al. (2018) state that "there is growing recognition that interdisciplinary approaches that account for both ecological and social processes are necessary to fully address human-wildlife interactions." It is apparent that there is a need for social-ecological integration in the management of human-wildlife feeding (Leopold, 1989;Lischka et al., 2018), especially as humans are the ones driving the initial establishment of these risky interactions (Smith et al., 2008). In ecotourism feeding studies, there has been a recent focus on exploring the social perspective of provisioning activities, as motivations may vary and impact engagement with regulations depending on visitors' backgrounds (Patroni, Day, et al., 2018;Ziegler et al., 2018). It is likely that the successful management of independent feeding interactions is also influenced by these social parameters, meaning that blanket educational statements may have varying effects depending on the audience. Notably, even when people are aware of the fact that feeding is banned, against regulations, or even illegal, some individuals persist in this behavior (Cunningham-Smith et al., 2006). Additionally, studies have shown that those who engage in potentially damaging interactions in tourism activities often also rate animal welfare as being of key importance to them (Senigaglia et al., 2020). This indicates that there is a disconnect between the understanding of the impacts of feeding and the drivers motivating people to feed. Therefore, we must consider the norms, values, and community culture of different socio-demographic groups within education-based projects, and determine which groups are not responding effectively to the management tools being used (Decker et al., 2010;Lischka et al., 2018) (Figure 3, adapted from Lischka et al. (2018)).
Factors of influence, as outlined in Figure 3, may be derived from either individual attributes or from external influences on the individual, all of which culminate in generating wider-scale social norms for communities. In terms of individual attributes, this could include certain socio-demographic characteristics that impact beliefs, learning, and attitudes pertaining to wildlife feeding. One factor that warrants consideration is location of  Lischka et al. (2018). This diagram describes the overall SES model of Human-Wildlife Interactions (HWIs), including external influences on individual behavior for both humans and wildlife (in yellow) and individual attributes which also affect behavior (in blue). Dashed arrows indicate feedback among all levels of external influences in the social and ecological systems. This model demonstrates the importance of the background of an individual as a driver in these HWI and the need to consider variation in these attributes as a potential barrier to the success of management actions. residency, that is, individuals' situation to the site where feeding is occurring. For example, one may be inclined to believe that locals and residents ought to be more responsive to feeding management due to place attachment (Patroni, Day, et al., 2018;Tonge et al., 2013;Tonge et al., 2015). Conversely, however, locals and residents may generate a sense of ownership of local wildlife that may result in a disregard for management that is attempting to change their established behaviors (Dubois & Fraser, 2013a;Dubois & Fraser, 2013b;Griffin, Haigh, Conteddu, et al., 2022;Kretser et al., 2009). When variation in behavior between groups occurs, feeding management may need to be adjusted to appropriately target the audiences present. Due consideration must also be given to factors such as age and nationality, as associated circumstances and experiences may manifest as variation in norms and attitudes pertaining to wildlife feeding. This is closely tied to the second aspect mentioned, that is, external influences, which suggests that the groups, societies, and institutions that people are involved with may also influence their attitudes toward engaging with wildlife (Lischka et al., 2018).

| Developing new tools to boost campaigns
As well as focusing on ways to improve existing management controls, there is also the definitive need to develop new tools. This is particularly important for two reasons: 1. due to the lack of complete success when current controls are used on their own (Cunningham-Smith et al., 2006) and, 2. campaigns inevitably experience interruptions and, therefore, may require boosts.
Efforts to manage human-wildlife interactions and conflict may move between periods of success and failure, often linked to engagement levels by select stakeholder groups (Messmer, 2000;Moswete et al., 2012). Occasionally, wildlife management may be paused; for example, if resources need to be allocated elsewhere or simply are not available for a period of time (i.e., because of war (Kanyamibwa, 1998), loss of funding and associated personnel (Rhodes Jr, 2013), and pandemics/epidemics (Miller-Rushing et al., 2021;Zhongming et al., 2021)). For associated reasons, research is usually also paused during these periods, meaning that the effects of these interruptions are very rarely analyzed. Typically, after these periods, managers will actively aim to boost or "re-kickstart" previous campaigns, which may require the employment of new tools (Miller-Rushing et al., 2021). The management of human-wildlife feeding interactions remains remarkably understudied in this area. It is, therefore, vital for researchers to explore both the impacts of interruptions on these management campaigns, as well as develop new tools for management which can assist either after these periods or with management in general.
One relatively new tool that is being used to improve human-wildlife relationships is the structured use of the media, referred to as media or marketing campaigns. This has gained traction in the field of illegal wildlife trade as researchers aim to both gather information on and change human purchasing behaviors (Feddema et al., 2020;Kachen & Krishen, 2020). It has also been noted as of key importance in the spreading of information regarding garbage management to prevent conflict with bears (Beckmann & Lackey, 2018). Media campaigns may include the use of both traditional and social media, with particular emphasis being put on the role that social media may play in changing behaviors (Ayeh et al., 2012;McGloin & Eslami, 2015;Zeng & Gerritsen, 2014). Indeed, people are increasingly using social media as their primary source of information (Guenther & Shanahan, 2020), and may often even get the idea to feed wildlife from it (Pagel et al., 2020). Ergo, wildlife managers may benefit from using this tool to share educational information rather than having it promote detrimental activities.
Media communication also opens the opportunity to include eye-catching imagery and videos, allowing messaging to potentially gain more traction than that communicated through traditional controls such as signage (Guenther & Shanahan, 2020). This allows wildlife managers to introduce "narrative-based communication" (Guenther & Shanahan, 2020). This may serve to better influence cognitive hierarchies, which are typically difficult to change due to long-term establishment (Kretser et al., 2009). Additionally, unlike other more traditional forms of management, the use of social media allows managers to gain direct feedback on what messaging and imagery performs best (i.e., through clicks and overall engagement). With traditional management tools, researchers are reliant upon in-person surveys and questionnaires, which may be biased in terms of the groups willing to respond and also the truthfulness of responses (Murray, 1999;Rybar & Bielikova, 2016). The use of social media campaigns has gained some initial success, marking these campaigns as a potentially invaluable in managing independent feeding interactions (Griffin, Nolan, et al., In press-b), but it requires continuous testing and feedback in order to truly be optimized and have its validity thoroughly assessed.

| FUTURE RESEARCH
As discussed, it important that we continue to ensure that the relationship between humans and wildlife is kept mutually beneficial, or at least benign. Therefore, until we have fully uncovered what impacts humans are having on the wildlife involved in independent feeding interactions, it is inadvisable that this form of feeding be continued. Ergo, it is fundamental that the effects of feeding interactions and their management be carefully tested across multiple species and locations, in order to provide a fully-informed body of literature on this subject.
There are several key areas that could be developed to fill current gaps in the literature regarding independent feeding interactions. First of all, the role of individual animal behavior and how this may be driving artificial selection must be further assessed across different populations of wildlife (Merrick & Koprowski, 2017;Swan et al., 2017). Secondly, whether feeding behavior changes over the course of the year, and the associated effects that this could have, must also be addressed. For example, there is the potential that humans approaching to feed directly may interrupt mating behaviors during set mating seasons, which could result in delayed impregnation and, therefore, potentially have impacts on offspring survival (Fairbanks, 1993). Additionally, engagement levels from wildlife may change over the course of the year as forage and prey naturally varies (Luccarini et al., 2006;Putman et al., 1993), which could result in temporally specific effects upon those individuals. Thirdly, the role of life history effects (Kemp et al., 2006;Maspons et al., 2019) in driving or changing interaction behaviors must also be assessed to determine whether there is an identifiable causative factor in engagement. Furthermore, longitudinal studies must be performed in order to determine whether the offspring of individuals that interact regularly also display this behavior later in life. This would also enable the tracking of whether interaction behaviors become more predominant in a population and impacts overall natural behaviors. Finally, the road to human-wildlife coexistence is not complete until such time as that which is stimulating conflict is entirely alleviated. Therefore, until such time should occur, it is of paramount importance that research identifying ways to effectively reduce risky human-wildlife feeding interactions be continued and perfected.