The dark side of nature experience: Typology, dynamics and implications of negative sensory interactions with nature

1. The human health benefits of direct sensory interactions with nature (hereafter direct human– nature interactions) are increasingly recognised. However, these interactions can also have various negative health and well-being impacts on people, some of which may be severe. Compared to positive ones, there has been relatively little investigation of such negative direct human– nature interactions beyond the medical literature, and what has been done is widely scattered across disciplines. 2. Here, we provide an overview of the typology, characteristics and dynamics of negative direct sensory interactions with nature and suggest management implications and future research directions. 3. We highlight the breadth of forms that negative direct human– nature interactions occur, and evidence that the incidences of some have recently grown rapidly in many parts of the world. 4. Our review also suggests that more intense negative direct human– nature interactions


| INTRODUC TI ON
Direct sensory interactions with nature (hereafter direct humannature interactions) are an important part of many people's lives, and the health and well-being benefits that often result are increasingly recognised (Bratman et al., 2019;Hartig et al., 2014;Keniger et al., 2013). There is broad consensus that a regular 'dose' of nature contributes to improved physical, psychological and social wellbeing (Bratman et al., 2019;Shanahan et al., 2015) and, especially in higher income societies, policymakers and practitioners have made considerable efforts to design and implement strategies and programs aimed at increasing people's direct engagement with nature (Shanahan et al., 2019).
Alongside these benefits, direct human-nature interactions can also have various negative health and well-being impacts on people, some of which may be severe (e.g. being attacked by large mammals, being stung by venomous wasps; Barua et al., 2013;Gulati et al., 2021;Marselle et al., 2021;McPhee, 2014;Penteriani et al., 2016;Soulsbury & White, 2016). Compared to positive ones, there has been relatively little investigation of such negative direct human-nature interactions (see Box 1 for their definition) beyond the medical literature, and what has been done is widely scattered across disciplines and journals (Barua et al., 2013;Methorst et al., 2020;Soga & Gaston, 2022;Soulsbury & White, 2016). Indeed, while much attention has been paid to the health benefits derived from direct human-nature interactions (e.g. Bratman et al., 2019;Hartig et al., 2014;Keniger et al., 2013), their negative impacts have been little addressed in much recent discussion of such interactions (Marselle et al., 2021;Methorst et al., 2020). Moreover, since positive and negative direct human-nature interactions have traditionally been investigated by researchers from different disciplines and perspectives (the bulk of the consideration of positive interactions has occurred outside of the traditional medical domain, including from psychology, environment and public health perspectives), and therefore, how the two are linked remains poorly understood. The limited understanding of negative direct human-nature interactions in this broad context is problematic because (i) these interactions entail substantial economic and social costs (Gilbert et al., 2017;Gulati et al., 2021;Kasturiratne et al., 2017); (ii) in many parts of the world, there has been an increasing trend in some kinds of negative direct human-nature interactions (see Figure 1); (iii) it limits the attention paid to actions that reduce such interactions and improve human and animal welfare; and (iv) encouragement of positive direct human-nature interactions may fail to consider the potential simultaneous increase in the magnitude of the negative ones that people may experience.
Here, we provide an overview of the typology, characteristics and dynamics of negative direct human-nature interactions and suggest management implications and future research directions.
Throughout, we take a global perspective, considering the wide range of negative direct human-nature interactions that can occur within different contexts, including both higher and lower income countries as well as urban and rural settings (see Table 1 for a list of examples). To date, beyond the medical literature, the negative impacts of nature on human health and well-being have been discussed mainly in the context of 'ecosystem disservices' or 'human-wildlife conflict' (Barua et al., 2013;Shackleton et al., 2016). However, while negative direct human-nature interactions are concerned with direct sensory interactions of people with nature (see Box 1), this is often not the case in these two contexts. Indeed, studies on ecosystem disservices and human-wildlife conflict have often focused on more generic economic or societal costs generated from nature that are not directly related to sensory interactions with it (e.g. crop weeds and pests, livestock damage, fire hazards; Barua et al., 2013;Shackleton et al., 2016). Also, as commonly in its definition humanwildlife conflict focuses mainly on interactions between humans and animals that have negative consequences for both parties (e.g. human injury or death caused by large carnivores), many negative direct human-nature interactions (especially those that result in

BOX 1 Definition of negative direct human-nature interactions
We consider negative direct human-nature interactions as 'a contact with nature by people through sensory systems that results in adverse health outcomes' (Soga & Gaston, 2020). Following the use in personalised ecology Soga & Gaston, 2020;Soga & Gaston, 2022), we define nature as 'all living organisms and ecosystems, excluding those that are not self-sustained'; we recognise that some prefer much broader definitions and that these may also include abiotic and cultivated/ domesticated components (Ducarme & Couvet, 2020). We consider health as a state of complete physical, psychological and social well-being and not merely the absence of disease or infirmity (WHO, 1948). Negative consequences for human health are thus not simply the occurrence of physical injury or fatality but include the elicitation of negative emotions (e.g. fear, disgust, sadness).
Under these definitions, there are a wide range of negative direct human-nature interactions, such as suffering from physical injury from bites, stings or attacks by animals, allergic or toxic reactions from contacting plants and physical injury during an accident occurred in natural environments (e.g. slip, fall), encountering a type of animal that causes psychological phobia, hearing sounds of animals that evokes fear and perceiving bad smells of animal carcasses or faeces left in natural environments (see Table 1).
These sensory interactions could occur consciously or subconsciously, and will chiefly take place through tactile, visual, auditory and olfactory sensory channels (see Section 2 for a more detailed discussion). In principle, our definition of direct human-nature interactions could also include interactions through the sense of taste, although we will not discuss these further because for most people the associated health outcomes are normally derived from dead organisms, which therefore lie beyond our definition (for present purposes) of nature. Since the focus of this review is on 'sensory' interactions with nature, we will not discuss interactions with viruses or microorganisms (e.g. bacteria) or parasitism (e.g. roundworms), for which these are much less relevant. Obviously, these other interactions with nature are nonetheless vitally important for human health (Hay et al., 2004;Messina et al., 2019). F I G U R E 1 Global rise in negative direct human-nature interactions. Examples include: (a) the number of attacks by brown bears in developed countries has increased approximately fourfold over the last 16 years (Bombieri et al., 2019); (b) the number of snakebites in Mexico has increased approximately 1.5 times over the last 13 years (Chippaux, 2017); (c) the number of shark bites in Australia has increased approximately tenfold over the last 100 years (West, 2011); (d) the number of stings by venomous caterpillars in Brazil has increased approximately seven times over the last 12 years (Chippaux, 2015); (e) the number of scorpion stings in northeastern Brazil has increased approximately threefold over the last 8 years (Araújo et al., 2017); and (f) the number of deaths caused by fall-related accidents in the Austrian Alps has increased approximately 1.5 times over the last 9 years (Faulhaber et al., 2017). Note that the figures shown represent the total, rather than per capita, number of negative human-nature interactions. psychological impacts on humans through non-tactile sensory channels; see Table 1) have been little considered in previous discussion of such conflict (Barua et al., 2013;Nyhus, 2016). Against this background, we think that this review will be useful to researchers, as well as policymakers, in facilitating a broad view and in-depth understanding of negative direct human-nature interactions and helping develop a more balanced view of the benefits and costs derived from direct interactions with nature. For this overview, we did not perform a formal systematic literature search because research on the topic of negative direct human-nature interactions is too fragmented and transdisciplinary for such an approach to be applied. The material we draw on was identified through a purposefully broad search (using a combination of Web of Science, Google Scholar and Google) of both peer-reviewed literature and other sources.

| FO RM
As described in Table 1, negative direct human-nature interactions take diverse forms. They occur in a range of settings from highly urbanised to remote areas, from terrestrial to aquatic environments, and in both indoor and outdoor spaces. In this section, we outline some of the major types of negative direct human-nature interactions and discuss how they occur and the organisms and ecosystems responsible for them.

| Tactile interactions
Arguably, the most obvious form of negative direct human-nature interactions is tactile contact with animals that can harm humans.
Numerous species from a broad range of taxonomic groups are known to be more or less a source of negative physical health outcomes such as physical injury (e.g. being attacked by large carnivores), allergic reactions (e.g. being bitten by mosquitoes) or toxic reactions (e.g. being stung by venomous wasps; see Table 2). Some of the well-reported groups of animals contributing to these interactions include annelids (e.g. leeches), arthropods (e.g. scorpions, spiders, ticks, crabs, moths [esp. caterpillars], ants, bees, hornets, centipedes), cnidarians (e.g. jellyfish, sea anemones), echinoderms (e.g. sea urchins, sea stars), molluscs (e.g. octopus, cone snails), fish (e.g. stingrays, sharks), reptiles (e.g. iguanas, snakes, turtles, alligators), birds (e.g. kites, crows, gulls) and mammals (e.g. cats, bears, monkeys; Bachleda, 2001;Burnett et al., 1996;Habermehl, 2012;Iverson & Skinner, 2006;Rowland & Eipper, 2019). The majority of these negative tactile interactions are caused by animals obtaining food resources or in the action of so doing (e.g. mosquito bites, tick bites, jellyfish stings), or protecting themselves, their territories or offspring from perceived threats from humans (e.g. large carnivore attacks, snakebites). Negative tactile interactions with animals in the worst case result in the death of humans, which is normally caused by animals that have particularly strong physical TA B L E 1 Examples of negative direct human-nature interactions according to the origin of nature and sensory pathways  (Brookes, 2007;Schmidlin, 2009). There are also many plant species that contain rash-inducing chemicals such as nettles, poison ivy and poison oak and dogwood (Bachleda, 2001;Modi et al., 2009). Despite such a potentially widespread occurrence of negative tactile interactions with plants, they have received remarkably little attention compared to those with animals. It thus seems important to develop further knowledge and understanding of the magnitude and spatial and temporal dynamics of negative tactile interactions with plants.
Negative tactile interactions with nature are not only caused by particular organisms, but also generated from ecosystems more generally. In terrestrial settings, for example, mountain-based activities such as hiking and mountaineering are associated with a risk of injury which may range from minor, to disabling, to even death (e.g. slip, fall or avalanche-related accidents; Faulhaber et al., 2017;Gatterer et al., 2019). A similar risk also exists with freshwater or marine environments (Bessereau et al., 2016;Peden et al., 2016).
Indeed, in Japan alone, for example, there are estimated to be c. 3000 and 1500 accidents annually during mountain and maritime leisure activities, respectively (National Police Agency, 2022).

| Audio, olfactory and visual interactions
In discussions of direct human-nature interactions, most attention has so far been paid to tactile interactions with nature (Barua et al., 2013;Gulati et al., 2021;Marselle et al., 2021;Methorst et al., 2020). However, there are also many negative direct humannature interactions that occur through non-tactile sensory channels (i.e. visual, auditory and olfactory), and these can occur at a much higher frequency and over a much wider range of conditions (see Section 3).
Encountering wild animals that are potentially harmful to humans, and environments that can harbour these organisms (e.g. dark  Soga & Gaston, 2022;Ulrich, 1993 hitting deer with a car), they might feel distressed, fearful or sad (c.f., Pagani et al., 2008). Of course, this can also be the case when people encounter organisms or environments that are damaged regardless of the cause (natural processes or anthropogenic impacts).
Many animals emit sound to communicate information such as mating potential and territory size, or in response to potential threats, which can sometimes result in unpleasant auditory experiences for humans. For example, some insect, frog, bird and mammal species (e.g. cicadas, pigeons, crows, gulls, foxes) can create loud or repetitive noises, especially when in groups and during the breeding season, which can lead to a severe noise nuisance, especially in urban areas (Gu et al., 2012;Kalnicky et al., 2019;Kung et al., 2015;Lyytimäki, 2014;Temby, 2004;    inhalation of which has been reported to cause sinus irritation, sneezing, headache and even loss of consciousness (Nassini et al., 2012).
Negative olfactory interactions with nature can also be generated from ecosystems. For example, animal carcasses or faeces left in greenspace can often be a major source of odour nuisance in urban areas (e.g. faeces of raccoon dogs in domestic gardens).

| OCCURREN CE
The many forms of negative direct human-nature interactions can be classified along three key dimensions: frequency, intensity and consistency ( Figure 2).

| Frequency
Within a given human population, some negative direct humannature interactions occur frequently and others rarely (Figure 2).
High-frequency interactions might include, for example, being bitten by a mosquito and suffering from noise nuisance caused by birds.
Low-frequency interactions might include being stung by a stingray or falling down a mountain. Of course, the relative frequency of occurrence of negative direct human-nature interactions likely varies depending on the human population under consideration (Soga & Gaston, 2020). Nevertheless, in the main it is likely that the negative direct human-nature interactions that a person experiences in a space and during a certain period consist of many kinds of lowfrequency interactions and a few kinds of high-frequency ones, although this assumption has not yet been formally tested.
As well as there being variation in the frequency with which negative direct human-nature interactions occur, there is also variation in how often an individual person experiences these interactions.
This is because the opportunities that people have to experience negative direct interactions with nature are generally not uniformly distributed across the population but are biased heavily towards certain groups of people or sectors of society (Chaves et al., 2015;Dunham et al., 2010). For example, those working in the agriculture, forestry and fisheries sectors normally have greater opportunity to experience nature compared to other groups of people, and thus they often experience negative direct interactions with nature more frequently (Quandt et al., 2013). It is also known that, in more urbanised societies, those having regular nature experiences tend to be rare, and a high proportion of direct human-nature interactions are disproportionately experienced by a small proportion of the human population Soga et al., 2021). This implies that, at least in urbanised areas, most people have no or very low numbers of negative direct interactions with nature and a small proportion probably experience the majority of these interactions.

| Intensity
Intensity is the degree of negative health impacts that a person receives from direct nature interactions (Figure 2), which can be

| Inter-relations
The frequency, intensity and consistency of negative direct humannature interactions are likely closely interrelated in several ways. For example, more intense negative direct nature interactions are likely to occur less frequently compared to less intense ones. Likewise, more intense negative direct nature interactions are more likely to be more consistent; they normally occur through the tactile sense.
Determining the relationship between the three dimensions of negative direct human-nature interactions would be an important step in understanding the dynamics of these interactions and their consequences for human health and well-being.

| DYNAMIC S
The occurrence of negative direct human-nature interactions varies at multiple spatial (from local to regional, and national to international) and temporal scales ( Over the last few decades, there has been a growing trend in some kinds of negative direct human-nature interactions such as snakebites (Chippaux, 2017), shark bites (Chapman & McPhee, 2016;McPhee, 2014), attacks by large carnivores (Bombieri et al., 2019;Penteriani et al., 2016), invertebrate interactions (Araújo et al., 2017;CDC, 2019;Chippaux, 2015) and fall-related accidents in mountains (see Figure 1). The so-called 'expansion of negative experience' (Soga & Gaston, 2022) is found in both developed and developing F I G U R E 2 Examples of negative direct human-nature interactions across three dimensions to their typology (frequency, intensity and consistency). Examples include (a) being bitten by a mosquito, (b) being bitten by a snake, (c) being stung by a wasp, (d) perceiving noise nuisance caused by crows, (e) being stung by a jellyfish and (f) encountering a type of organism that can evoke a psychological phobia (e.g. insects). Note that the relative occurrence frequency of negative direct human-nature interactions likely varies across regions with different climatic, environmental and geological conditions, and thus the classification system in terms of frequency may vary by region. Images adapted from iStock.

| MITI G ATI ON S TR ATEG IE S
In principle, negative direct human-nature interactions can be mitigated by four major strategies (see Table 3). The first and perhaps most straightforward approach is to reduce inappropriate human behaviours that can lead to negative direct human-nature interactions (e.g. feeding and approaching wild animals, going for a nature walk without proper equipment or skills; Penteriani et al., 2016), which might be achieved by various educational and social marketing programmes aimed at increasing people's knowledge about nature F I G U R E 3 Evidence demonstrating the importance of the four drivers of negative direct human-nature interactions: (a) the abundance of nature potentially harmful to people, (b) the behaviour of animals towards people, (c) the abundance of humans, and (d) the behaviour of humans towards nature. Examples include that (a) tiger attacks on humans in the Chandrapur District of Central India occur mainly in and near forest that is the main habitat of tigers in this region (see Dhanwatey et al., 2013 for more details about how predicted occurrence was calculated), (b) mosquito preference for humans varies widely across Africa and increases significantly with human population density (see Rose et al., 2020 for more details about the preference scores), (c) large carnivore attacks on people in North America and Europe occur mainly from late spring to early autumn, because many people participate in outdoor recreational activities during this period (Penteriani et al., 2016), and (d) around half (47.6%) of the large carnivore attacks reported in North America and Europe were associated with certain human behaviours that may have contributed to the probability of suffering an attack (Penteriani et al., 2016). (Abrams et al., 2020). The second strategy is to curb the behaviours of animals that are potentially harmful to humans. For example, reducing the intensity of artificial feeding in national parks and the opportunities to obtain human-provided food in urban and suburban areas is likely to contribute to limiting levels of wildlife habituation to humans (Marley et al., 2017). The third strategy is to suppress people's use of nature (e.g. recreational or agricultural activities) during periods when or in places where there is a high risk of experiencing negative direct human-nature interactions (Linnell et al., 2005).
The fourth, and often last resort, strategy is to reduce the abundance of nature potentially harmful to people, either through more or less selective direct population reduction (e.g. trapping, hunting), or through habitat reduction (e.g. vegetation clearance, removal of temporary water pools where mosquitoes breed).
Of course, the feasibility and effectiveness of the four strategies will depend strongly on the specific socioeconomic and cultural background of the regions concerned. For example, in many regions (especially those of particular importance for biodiversity conservation), it would not be desirable to undertake extermination of wild animals or clearing of vegetation (e.g. Colman et al., 2014).
Likewise, these actions normally require significant financial resources, and thus they would often not be employed in areas where socioeconomically disadvantaged groups of people live (Bateman et al., 2021). Also, in areas where ecotourism is the major industry, suppressing people's use of nature (i.e. recreational activities) is not easily achievable for economic reasons. Furthermore, in urban areas where improving public safety and well-being is considered to be of predominant importance, people may be more inclined to accept TA B L E 3 Examples of mitigation strategies to reduce some major negative direct human-nature interactions. Note that this adopting strategies and policies aimed at reducing the abundance of nature potentially harmful to them (Bateman et al., 2021;Cahill et al., 2012). There is no 'one size fits all' solution to manage negative direct human-nature interactions and careful consideration is required when choosing among mitigation strategies.
Some of the above-mentioned approaches can decrease the intensity and frequency of positive direct human-nature interactions substantially, which may, in turn, reduce the health and well-being benefits that people can obtain from nature (Bratman et al., 2019;Hartig et al., 2014;Keniger et al., 2013). For example, preventing or restricting recreational activities in urban greenspace or national parks is likely to reduce people's opportunity and motivation to experience positive nature interactions in these environments (Bratman et al., 2019). Similarly, cutting trees that produce flowers with unpleasant smells is likely to decrease the positive impacts on people's health and well-being that interactions with these organisms can produce (e.g. psychological benefits of viewing trees through a window; Soga et al., 2021).

| FUTURE RE S E ARCH D IREC TI ON S
There are several key areas in which research efforts could be focussed to improve the understanding of negative direct human-nature interactions (see Box 2 for a list of priority research questions). First, it is important to determine the scale, patterns and drivers of the recent documented increases in some kinds of negative direct human-nature interactions (Figure 1). It remains unclear how common and widespread such trends are across countries and regions with different socioeconomic, cultural and environmental contexts. Indeed, it is often pointed out that many negative direct human-nature interactions, especially more intense ones, have been under-reported because they are not witnessed, occur in poor countries, occur during illegal activities (e.g. poaching), or perhaps out of fear of bad press for popular vacation destinations (Koziarski et al., 2016;Ricci et al., 2016;Tchoffo et al., 2019).
Hence, the existing evidence is probably just the tip of the iceberg and the actual number of negative direct human-nature interactions, and their associated costs to society, might be much higher.
Second, it is crucial to understand how positive and negative direct human-nature interactions are related. Importantly, at least in more affluent societies, it is increasingly apparent that more intense nega-  (Liu et al., 2021;Shackleton et al., 2017;Vink et al., 2011). Likewise, it might be possible that loss of some predator species due to anthropogenic activities can result in the increase in the abundance, and activity level, of wildlife that can cause negative health impacts for humans (Enari, 2021  unfavourable feelings towards wildlife (see Liu et al., 2011). An important direction for future research would be to determine the relative importance of and possibly interactive effects of positive and negative direct human-nature interactions in shaping people's probiodiversity attitudes and behaviours.

| CON CLUS IONS
With the ongoing growth of human populations and the associated increase in demand for natural resources, negative direct humannature interactions are expected to escalate worldwide (Figure 1).
Global climate change and urbanisation are also likely to accelerate this trend (Backe et al., 2021;Rose et al., 2020;Sonenshine, 2018;Zacarias & Loyola, 2019). It seems thus vitally important to develop a more comprehensive and detailed understanding of the nature and dynamics of negative direct human-nature interactions and enhance the ability to best manage them for improved human health and well-being. We argue that to advance this goal requires scientists to focus more attention on the negative personal, social and economic impacts resulting from direct human-nature interactions.
This is, of course, not to say that nature experiences are bad and detrimental to humans. Rather, we argue that there needs to be a balanced view of the benefits and costs of direct human-nature interactions and a clear recognition of the inevitable trade-offs that potentially exist between the two. From a conservation view point, it will also be crucial to explore how negative direct human-nature interactions can be mitigated with a minimal impact on biodiversity, and how conservation and restoration programmes influence the occurrence of these interactions. Determining ways of maximising beneficial outcomes of direct human-nature interactions, while minimising negative consequences for both humans and nature, is a major challenge in the 'Anthropocene'.

AUTH O R S ' CO NTR I B UTI O N S
Both authors conceived the work, drafted the manuscript, edited and approved the final version.

ACK N OWLED G EM ENTS
We

CO N FLI C T O F I NTE R E S T
K.J.G. and M.S. are the Editor-in-Chief and an Associate Editor of People and Nature, respectively, but took no part in the peer review and decision-making processes for this paper.

DATA AVA I L A B I L I T Y S TAT E M E N T
No new data were generated or analysed in this study.