This systematic review encompasses data from 602 participants in total, with sample sizes ranging from 11 to 187 participants. Where more than one study had utilized data collected from the same samples (Gray and Birrell, 2015; Gray and Thomson, 2016; Adams and Beauchamp, 2018, 2019, 2020), the participants numbers have been included only once in the systematic review’s total.

Three studies were conducted in the United Kingdom (Adams and Beauchamp, 2018, 2019, 2020; Tiplady and Menter, 2021; Moula et al., 2021), one in Ireland (Murphy, 2018), one in Hong Kong (Sobko et al., 2020), two in the United States (Bruni et al., 2017; Arbuthnott and Sutter, 2019) and one in Australia (Gray and Birrell, 2015; Gray and Thomson, 2016).

All included studies varied in terms of their study design. Studies included action research (Murphy, 2018), grounded theory (Adams and Beauchamp, 2018, 2019, 2020), ethnographic-inspired co-production (Tiplady and Menter, 2021), controlled before-and-after designs (Arbuthnott and Sutter, 2019), uncontrolled before-and-after designs (Bruni et al., 2017; Gray and Birrell, 2015; Gray and Thomson, 2016; Moula et al., 2021), and a two-arm randomized control trial (Sobko et al., 2020). Only two studies made comparisons with control groups. Of these, one control group received no intervention (Sobko et al., 2020) and the other (Arbuthnott and Sutter, 2019) had an active control group which engaged in an indoor, school-based intervention (i.e., indoor song writing). The remaining studies did not utilize control groups.

The type of arts-based interventions delivered in nature and outdoor spaces varied in terms of the structure, content, and activities. The most structured intervention was the Play and Grow early environmental education program utilized by Sobko et al. (2020). It involved 15 min of theoretical education on topics such as nature connectedness, followed by 30 min of guided activities with materials found in nature. For example, children were invited to create art with leaves, flowers, or stones. Other activities included practicing awareness of the sounds in nature, growing plants, and suggestions about how to care and protect the environment. Nature “homework” was also encouraged, such as taking care of own gardens, or growing own plants to maximize the frequency of contact with nature. A detailed description of the intervention structure and content has been published in a separate publication (Sobko et al., 2016).

Three interventions implemented multi-modal arts-based approaches which incorporated drawing, sculpturing, song writing, music-making, story-making, filmmaking, dancing, writing, and poetry (Bruni et al., 2017; Gray and Birrell, 2015; Gray and Thomson, 2016; Moula et al., 2021). Additionally, the “Get to Know” program (Bruni et al., 2017) included visits to nature centers, hikes, watching nature films and studying about the environment. The “Touched By The Earth” program (Gray and Birrell, 2015; Gray and Thomson, 2016) focused on ecosystem restoration, while “Eco-capabilities” (Moula et al., 2021) focused on environmental awareness and sustainability. The charity Cambridge Curiosity and Imagination presents a detailed description of the “Eco-capabilities” sessions (CCI, 2021).

Two interventions focused on outdoor music making (Adams and Beauchamp, 2018, 2019, 2020; Arbuthnott and Sutter, 2019). In the first intervention, children were challenged to create music for a ceremonial performance using musical instruments such as drums, djembes, wooden flutes, and bone flutes. Although vocal sounds were allowed, words were not permitted to facilitate emotional expression through music (Adams and Beauchamp, 2018, 2019, 2020). In the second intervention—Songwriting for Nature—outdoor education song writing workshops were delivered, but the sessions also included hikes, campfires, and conversations around the connectedness between nature and music (Arbuthnott and Sutter, 2019).

Two studies used the Forest School approach (Murphy, 2018; Tiplady and Menter, 2021), which followed the cycle of planning, observation, adaptation, and review in each session. Emphasis was given in the relationship between children and nature, as well as in opportunities for children to take supported risks appropriate to the environment and to themselves. Activities included nature-inspired art and crafts, tree climbing and exploration of nature.

Intervention details surrounding number of sessions were diverse, ranging from two sessions (Adams and Beauchamp, 2018, 2019, 2020) to continuous year-long approaches (Gray and Birrell, 2015; Gray and Thomson, 2016; Tiplady and Menter, 2021). The duration also varied from 45 min (Sobko et al., 2020) to one full day (Tiplady and Menter, 2021; Moula et al., 2021). The frequency and intensity varied from two consecutive days (Adams and Beauchamp, 2018, 2019, 2020; Arbuthnott and Sutter, 2019) to once a week (Murphy, 2018; Sobko et al., 2020; Tiplady and Menter, 2021; Moula et al., 2021). In some cases, the description of duration and intensity were not reported (Bruni et al., 2017; Gray and Birrell, 2015; Gray and Thomson, 2016; Murphy, 2018). Specific details of all interventions are presented in Table 6.

Most of the studies recruited participants from primary and secondary schools. In terms of primary schools, Adams and Beauchamp (2018, 2019, 2020) collaborated with six schools in rural locations in North and South Wales. Moula et al. (2021) worked with two East Anglian schools identified as being in areas of poverty and deprivation with minimal cultural provision and infrastructure, and Murphy (2018) with an urban, disadvantaged, multi-denominational school in Ireland. In terms of secondary schools, Gray and Birrell (2015) and Gray and Thomson (2016) collaborated with a school with gifted and talented students in New South Wales, while Arbuthnott and Sutter (2019) worked with a high school in North America; further details about the demographics of this school were not provided. Tiplady and Menter (2021) employed both a primary school and a secondary specialized school in the United Kingdom. Bruni et al. (2017) recruited the participants from schools and youth organizations in South Carolina, while Sobko et al. (2020) from kindergartens and community centers in Hong Kong.

All included interventions took place in nature and outdoor spaces. Adams and Beauchamp (2018, 2019, 2020)) utilized rural locations with a strong prehistoric context, such as fields containing neolithic chambers, but also locations without strong prehistoric context, such as beaches and woodlands. Whilst some sites were close to houses and roads, all were situated in quiet areas with no urban noise pollution. Two studies explicitly divulged settings as national outdoor localities. Gray and Birrell (2015) and Gray and Thomson (2016) carried out their intervention at the Bundanon Trust properties in Australia, and Arbuthnott and Sutter (2019) at the National Wildlife Area, which is situated in south-central Saskatchewan in North America. Two studies utilized the school playgrounds and school gardens or nature resorts (Murphy, 2018; Moula et al., 2021). Tiplady and Menter (2021) delivered the sessions in a community garden with an established woodland and a dedicated Forest School area, while Sobko et al. (2020) delivered the sessions exclusively in public parks. Finally, Bruni et al. (2017) delivered their intervention in various locations; some sessions took place in nature, but some sessions were delivered indoors in aquariums and national history museums.

Two studies implemented the Forest School pedagogical approach (Murphy, 2018; Tiplady and Menter, 2021), which was based on the six fundamental Forest School principles (Forest School Association, 2021):

The former study implemented this approach to teach children the process of art, whilst the latter focused on its application to improve young people’s wellbeing.

The studies conducted by Adams and Beauchamp (2018, 2019, 2020) referred to Waite’s work (2017), who advocated that outdoor learning opportunities are more than simply taking classroom activities outside. In contrast, being outside exposes children in new sounds, such as waves or echoes, which may act both as a valuable resource for musical compositions as well as an emotional stimulus contributing to children’s wellbeing. The authors also referred to work Vygotsky’s (1971) arguing that music can transform our emotional state, and our existential state of being. Another significant concept in this study (Adams and Beauchamp, 2019) was that of spirituality, defined as “an awareness that one is connected to something more, beyond the individual self, but which can be grounded in an existential reality” (de Souza and Watson, 2018, p. 345). Finally, the authors created links to Wilson’s biophilia theory (Kellert and Wilson, 1995), a hypothesis suggesting that humans possess an innate tendency to seek connections with nature and other forms of life, and actualizing this inner tendency leads to both physical and mental health benefits. Sobko et al. (2020) also drew from biophilia theory in their study.

Three studies (Bruni et al., 2017; Gray and Birrell, 2015; Gray and Thomson, 2016; Arbuthnott and Sutter, 2019) drew upon the psychological construct of connectedness with nature (Schultz, 2002; Nisbet et al., 2011) to provide the theoretical framework for their studies. This was defined as the degree to which individuals perceive that they are part of the natural environment (Schultz, 2002). Connectedness with nature involved three key aspects: emotional affiliation to nature, understanding the importance and interconnectedness of all aspects of nature, and seeking regular contact with the natural world (Nisbet et al., 2009). The hypothesis in these studies was that nature connectedness would correlate positively with children’s sense of wellbeing, and the results were therefore analyzed through these lenses.

Finally, Eco-Capabilities (Moula et al., 2021) drew upon Amartya Sen’s work (1993) on human capabilities as a proxy for wellbeing. The term “eco-capabilities” was developed to describe how children define what they feel they need to live a fully human good life through the lenses of environmental sustainability, social justice, and future economic wellbeing (the three pillars of sustainability).

The two most reported outcomes were connectedness to nature and sense of wellbeing, followed by engagement in the learning process.

All studies reported a positive impact on children’s and young people’s connectedness to nature, assessed either through quantitative (Bruni et al., 2017; Gray and Birrell, 2015; Gray and Thomson, 2016; Arbuthnott and Sutter, 2019; Sobko et al., 2020), qualitative (Gray and Birrell, 2015; Gray and Thomson, 2016; Adams and Beauchamp, 2018, 2019, 2020; Moula et al., 2021) and arts-based methods (Gray and Birrell, 2015; Gray and Thomson, 2016; Moula et al., 2021).

Although four studies employed quantitative analysis to assess connectedness or relatedness with/to nature, there was still heterogeneity in the outcome measures used. Two studies utilized the Nature Relatedness Scale (NRS; Nisbet et al., 2009; Gray and Birrell, 2015; Arbuthnott and Sutter, 2019). Sobko et al. (2020) used the Connectedness to Nature Scale (CNS; Mayer et al., 2009), while Bruni et al. (2017) implemented the FlexiTwins game version of the Implicit Association Test (IAT Nature; Schultz et al., 2004; Schultz and Tabanico, 2007).

The impact on sense of wellbeing was only measured in one study (Moula et al., 2021) through the Personal Wellbeing Index—School Children (PWI—SC; Tomyn et al., 2013) yet the results have not been reported yet. Sobko et al. (2020), although not assessing changes in the sense of wellbeing directly, evaluated factors that affect (i.e., stress) or are affected by (i.e., frequency of anger) wellbeing. The authors reported significant reductions in children’s stress level, which was assessed through the Perceived Stress Scale for Children (PSS-C; Cohen et al., 1983), and in the frequency of anger, although it was unclear how this was measured.

Similarly, all other studies reported benefits which are associated with children and young people’s mental health and wellbeing. Adams and Beauchamp (2018, 2020) observed enhanced empathetic skills and emotional regulation. Gray and Birrell (2015) noticed improvements in vitality, energy and joy, and Murphy (2018) reported improvements in children’s happiness, self-esteem, and resilience. The improvements in these studies were evaluated through qualitative methods.

Two studies took a more critical stance toward reporting positive outcomes. Arbuthnott and Sutter (2019) reported only nominal increases in young people’s positive mood, while the negative moods were reduced in both control and intervention groups (i.e., singing outdoors compared to singing indoors). Changes in these outcomes were measured through the short version of the Depression and Anxiety Stress Scales (DASS-21; Henry and Crawford, 2005), and the Positive and Negative Affect Schedule (PANAS; Watson et al., 1988), in which nine items from the Elevating Experience Scale (EES; Huta and Ryan, 2010) were added. Tiplady and Menter (2021) observed improved relationships and children’s self-perception through achieving success, which led to higher levels of independence and internalization of positive self-narratives. However, the authors mentioned that the evidence related to wellbeing was insufficient, partially due to small sample (n = 16), despite children participating in approximately 39 sessions (1 day per week for a school year).

Four studies observed increased engagement and immersion in the learning process (Adams and Beauchamp, 2018, 2020; Murphy, 2018; Moula et al., 2021; Tiplady and Menter, 2021). Specifically, Adams and Beauchamp (2018, 2020) noted that being outdoors resulted in children being more calm and focused, while their listening skills were also improved. Moula et al. (2021) observed additional improvements for children with special educational, learning, behavioral needs, as well as for children who had lower self-esteem and less confidence in their academic skills. The sessions were described as empowering and teachers reported significant improvements in children’s confidence, behavior and focus in the classroom (results reported in Walshe, 2021). Engagement in the learning process was assessed through qualitative methods, such as interviews with teachers, children and young people, but standardized measures were not used in any study.

This section includes some of the less commonly reported outcomes and outcome measures. These were improvements in children’s sense of authenticity (Adams and Beauchamp, 2018), autonomy and agency (Adams and Beauchamp, 2018; Moula et al., 2021), higher awareness of their senses (Adams and Beauchamp, 2018; Moula et al., 2021), and higher sense of belonging in/existence as part of nature (Moula et al., 2021). Sobko et al. (2020) also observed higher recognition and appreciation of biodiversity. The same authors measured children’s level of fecal serotonin and gut microbiota pre- and post-intervention using the ELISA and 16S rDNA amplicon sequencing (García-López et al., 2014; Marizzoni et al., 2020). No significant changes were found in microbiota diversity, but children with decreased perceived stress were found to have significantly higher gut microbiota richness.

To contextualize these results, the quality appraisal was crucial as some of these findings should be interpreted with caution. In the quality appraisal table (Table 7) five studies were classified as “Qualitative” and three studies were classified as “Quantitative.” Although two studies adopted a quasi-experimental design (Gray and Birrell, 2015; Gray and Thomson, 2016; Moula et al., 2021), they only had small quantitative elements and for this reason were classified as qualitative.

Almost all qualitative studies were rated remarkably high in terms of the four quality appraisal criteria: credibility; dependability; transferability; and confirmability. This could be because studies which did not report explicitly the research aims, questions and design (therefore the studies with potentially lower quality) were excluded at the full-text screening stage. Only one qualitative study was rated as “unclear” in terms of its quality (Gray and Birrell, 2015; Gray and Thomson, 2016) because some key details were missing. For example, the duration and frequency of the intervention was not reported, information regarding the researchers’ background was limited, and the interview questions were not publicly available.

All other qualitative studies were rated as high in terms of credibility, for example because verbatim quotes from the participants were used to support the findings, and the results were analyzed independently by more than one researcher. All studies rated high in dependability as they adopted a logical, traceable, and clearly documented research process. Transferability was also rated high because information regarding the reasons for the selection of study participants; demographic characteristics, contextual information, setting location and activities delivered were clearly outlined. Finally, almost all authors reflected on their research background and the impact this might have had on the studies, while interview or focus group questions were publicly accessible, thereby improving the confirmability of the findings.

The quantitative studies were rated significantly lower, partially because most studies were quasi-experimental, therefore methods such as randomization and blinding did not take place. This led to studies being rated as “high” or “unclear” in terms of selection, performance, and detection bias. In the single study that adopted an experimental design (Sobko et al., 2020), information was missing regarding the methods of random sequence generation, allocation concealment, and blinding of the outcome measures. One study also administered the questionnaires during transportation (bus ride) to the workshops and children were expected to complete the questionnaires during that time, increasing the detection bias (Arbuthnott and Sutter, 2019). However, all studies rated low in attrition bias because there were only small attrition rates, and these were reported with justifications. Reporting bias was also rated low in two studies (Bruni et al., 2017; Arbuthnott and Sutter, 2019) but high in Sobko et al. (2020), because necessary information, such as participants’ demographic characteristics, were not reported.

Nature connectedness was the most commonly reported outcome of interest in this systematic review. Being in nature appeared to lead to children and young people feeling gradually, and increasingly more connected to nature. In turn, feeling connected to nature led to sensing that nature was becoming an important part of their identity—what has been defined as pro-environmental identity (Stets and Biga, 2003)—and came alongside higher awareness of environmental issues. This led to pro-environmental behaviors, behaviors in which children and young people were not only caring for the environment but were also taking protective actions, contributing therefore to environmental sustainability. Finally, this led to higher desire in having sustained contact with nature. Figure 2 aims to illustrate this process as understood through the synthesis of the included studies. However, it is worth mentioning that the process of change appeared to be highly complex, therefore the diagram illustrates only one possible direction of travel out of multiple possible directions.

The findings of this review echo previous studies which have found that exposure to the natural world is a strong predictor of pro-environmental attitudes and ecological behaviors (Mayer and Frantz, 2004; Leary et al., 2008; Nisbet et al., 2009; Frantz and Mayer, 2014). Interestingly, connectedness to nature has been found to be an even stronger predictor of pro-environmental behavior in children than environmental knowledge (Schultz, 2002; Schultz and Tabanico, 2007; Gosling and Williams, 2010; Otto and Pensini, 2017; Barrable and Booth, 2020). Similarly, research suggests that the way that connectedness to nature can be achieved is not through learning in theory about the environment, but by being exposed to the beauty of nature, the emotions that arise while being in nature, and with sustained contact (Ryan et al., 2010; Rainisio et al., 2014; Lumber et al., 2017). Hence, the concept of spirituality that was explored in the Adams and Beauchamp (2019) study may be especially important to better understand the process through which connectedness to nature can be achieved, and how it enhances pro-environmental attitudes (Martin et al., 2020).

More specifically, Adams and Beauchamp (2019) found that outdoor music making provoked a sense of interconnectivity and harmony with nature, leading to experiences of extraordinary, transcendent, or so called “spiritual” moments (Hay and Nye, 2006; de Souza and Watson, 2018; Schein, 2017). During these moments, children expressed that they experienced a sense of entering a heightened reality that included bonding with nature and with each other, and it was described as entering a “new” or a “magic world.” Feelings of joy and peace were also generated during these moments. The authors created links with the Hay and Nye (2006) self-transcendence theory, according to which, it is only when we move beyond egocentricity that we can start to relate genuinely with nature and with each other. These immersive experiences resonated with theories of optimal experiences and “flow” (Csikszentmihalyi, 2002), Buber’s (1937) existential philosophy of dialogue, and Gelter’s (2010) concept of “genuine friluftsliv” (translated as “open air living”)—a commitment to celebrating time outdoors regardless of the weather. Most importantly, these findings echo previous literature arguing that it is possible for spiritual experiences to be achieved in childhood, because they don’t require high cognitive abilities or sophisticated language capacity, but universal human awareness (Hay and Nye, 2006). Arts can be the catalyst to inspire these spiritual experiences, thus reinforcing nature connectedness.

The synthesis of the included studies suggests that arts can offer an inclusive medium to increase nature connectivity, make the relationship with nature explicit, understand environmental issues and explore ways to prevent future environmental disasters, leading to environmental sustainability. Figure 3 aims to illustrate how engagement with the arts were found to provide additional benefits to being in nature and outdoor spaces. However, similar to the previous diagram, the process of change appeared to be complex and can occur via different paths that those illustrated in the diagram.

The included studies suggested that engaging with creativity, imagination and sensual experiences while being in nature can increase nature connectivity. Adams and Beauchamp (2019) reported that one of the first noticeable differences observed was a sensual experience of nature that was achieved through music-making; in this case, the freedom of space outdoors appeared to maximize opportunities for creativity and imagination, while also freeing up the institutional normative behaviors that are expected in the classroom (Waite, 2017). The combination of arts and nature also allowed children and young people to explore the significance of the history of places and the “stories of the land” (Stewart, 2008, p. 82). Through taking part in activities that involved emotional stimulation, self-reflection, and exploration of their identity, children and young people gradually started to perceive themselves as part of the environment, and vice versa, the environment as part of themselves (Bruni et al., 2017; Gray and Birrell, 2015; Gray and Thomson, 2016; Arbuthnott and Sutter, 2019). Furthermore, the outdoor sessions raised awareness of environmental issues through the arts and encouraged the exploration of ideas about what can be done to prevent future environmental disasters from happening, thereby reducing eco-anxiety (Bruni et al., 2017; Gray and Birrell, 2015; Gray and Thomson, 2016; Sobko et al., 2020; Walshe, 2021). Bruni et al. (2017) suggested that arts can engage even the children and young people who might otherwise remain disinterested about environmental issues and/or disengaged with educational programs. Finally, through the arts, children and young people can be involved in creative forms of activism, bringing positive environmental change and environmental sustainability (Inwood et al., 2017). This would encourage more people to spend time outdoors and start their own cycle of nature connectedness.

In summary, and to make this process more explicit, the key stages in the cycle of benefits of the arts in nature, as understood through this review, are as following:

Almost all studies reported improved wellbeing for children and young people, even though most did not explicitly use this term. At least 10 benefits to mental health and wellbeing were reported or observed, including improvements in:

Particularly the appreciation of the beauty and the unexpectedness of what can be found outdoors, such as natural sounds and images, appeared to have a positive impact on children’s sense of wellbeing. These findings align with one of the key theories to human-nature relationship, Ulrich’s psycho-evolutionary theory (Ulrich, 1993). Ulrich’s theory posits humans’ innate affiliation with natural environments (Richardson et al., 2018), drawing upon the assumption that natural environments induce positive emotions and feelings. This human-nature connectedness can help children and young people to view themselves as part of a wider ecology which has a positive impact on aspects of wellbeing, such as vitality, creativity, and happiness (Capaldi et al., 2014). Furthermore, connections were made with the PERMA theory of positive psychology (Seligman and Csikszentmihalyi, 2000) which argues that the positive states and traits that children cultivate while in nature (e.g., gratitude, joy, inspiration) contribute significantly to their sense of wellbeing. Seligman and Csikszentmihalyi (2000) have highlighted that the more we enable children and young people to cultivate this state of mind, they happier they can be in the long-term.

The Eco-Capabilities study (CCI, 2021; Moula et al., 2021; Walshe, 2021) developed a novel approach to wellbeing, drawing on Sen’s capabilities work (1993), whereby children were invited to define what is important to them for living a good life through the lenses of environmental sustainability, social justice, and future economic wellbeing. Sen’s human capabilities have been defined as a “a person’s ability to do valuable acts or reach valuable states of being” (Sen, 1993, p. 30). The list of capabilities (Nussbaum, 2000; Biggeri, 2007) include a range of human “functionings” that go beyond the notion of subjective and economic wellbeing (Nussbaum, 2011), aiming to provide opportunities for achieving a state of physical, emotional, intellectual, and existential wellbeing in life (Delors et al., 1996). Among the 101 children who participated in this study, the capabilities that appeared to have been developed the most were the senses and imagination, autonomy, safety, and emotional expression. Based on these, a list of “eco-capabilities” to define children’s wellbeing is currently being developed (Walshe, 2021).

Although the included qualitative research was rated by both reviewers to have high quality standards, this review suggests that quantitative research is still at its infancy. Only five studies with quantitative elements were identified, of which only three had substantial quantitative elements, and only one was experimental. The quality of the single experimental study was also rated as “low” or “unclear.” Considering the limited funding available for research on this area, national and international collaborative efforts would be of high importance to scale up the existing interventions to larger experimental studies, and to increase the recognition around the value of such interventions especially for policymakers.

The included quantitative studies explored different outcomes, or the same outcomes but using different outcome measures. This led to high heterogeneity in the results, thus a meta-analysis was not feasible. The two most reported outcomes were connectedness to nature and sense of wellbeing. The Nature Relatedness Scale (NRS; Nisbet et al., 2009) was the most commonly used outcome measure, although Gray and Birrell (2015) suggested that it did not explore significant concrete questions about how students felt about nature, and therefore other questionnaires should be considered. There was no outcome measure in common with regards to wellbeing. Future studies should consider the implementation of the most valid, reliable, and widely used outcome measures to allow more accurate comparisons with other studies, replication, as well as future meta-analyses. The Salazar et al. (2020) guide to assessing connection to nature is also highly recommended for decisions regarding outcome measures in future research. Arts-based outcome measures as embedded assessment tools may also be beneficial in capturing changes, particularly for children and young people who might find difficult to verbalize their perspectives (Flowers et al., 2015; Staples et al., 2019).

It would be beneficial for future studies to identify the parameters leading to substantially different effects, such as the duration, frequency, or intensity of the interventions. The sessions in all included studies varied significantly, and there was no evidence to suggest that longer sessions led to better outcomes. In contrast, even as little as two sessions found improvements in nature connectedness and experiences of spirituality (Adams and Beauchamp, 2019), although it is unclear how sustained these benefits might be. In this review, there was no evidence regarding the sustained effects of the included interventions. Although it may be challenging to follow-up children and young people who move to different grades or schools, this would help to understand whether long-lasting changes occurred, or whether the reported effects were due to participants’ temporary confidence that their life could be different following the interventions (Younge et al., 2015). As such, more evidence on the number, duration, and frequency of sessions that could lead to substantial and sustained benefits, could increase the public recognition of such interventions from policymakers and their inclusion in national and international guidelines, such as the National Institute for Health and Care Excellence (NICE) guidelines.

Another important element of the interventions was the space that the sessions were taking people. Adams and Beauchamp (2019) observed that outdoor spaces introduced new and unfamiliar stimuli, such as the sounds of waves or echoes from caves, which provided opportunities for creativity and imagination. However, they emphasized that distractive noise presents threats such as, disruption of attention and behavior, masking of important signals, and spurious physiological stimulation (Hatch and Fristrup, 2009). They found that the sonic environment had the greatest impact on children’s creative and spiritual experiences. Other distractions commonly reported in literature is the lack of cleanliness and amenities, which negatively influence the use of green spaces, even when these are available (Akpinar, 2016; Wood et al., 2018). Therefore, optimizing the conditions of space to make it safe, attractive, and welcoming should be prioritized in future research and policies.

Although it is likely that the cost of the included interventions was high, information regarding the cost to design and deliver the interventions was not available in any study. In the interest of increasing funding opportunities for future research, it is important to focus on the cost-effectiveness of arts in nature and outdoors compared to other interventions (e.g., only arts-based or nature-based interventions). This would also help to identify the additional value and mechanisms of change that occur when brining arts and nature together.

Furthermore, the only experimental study directly investigating physical health outcomes (Sobko et al., 2020) suggested that exposure to bacteria from natural environments could positively correlate to behavioral changes. They recommended future studies to explore the links between gut microbiota, connectedness to nature and psychosocial behaviors.

The impact of working with artists should also be explored in future research. Arbuthnott and Sutter (2019) argued that artists played a crucial role in shifting cultural norms and values, such as normalizing “mistakes,” while Gray and Birrell (2015) observed that the direct contact with professional artists who addressed the students as serious artists maximized the benefits of the sessions and supported the development of children’s identity as artists (Hay, 2019). However, given the high financial cost of the interventions, which is often unaffordable for schools, it would be valuable to investigate whether similar benefits could be observed if the sessions were delivered by teachers. Adams and Beauchamp (2018) highlighted that, if the sessions were to be delivered by teachers, less didactic pedagogies would need to be adopted to give students more opportunities for agency, creativity, and to develop higher self-confidence. It is worth noting that Murphy (2018) observed that teachers often lacked confidence in delivering arts-based sessions, outdoor sessions, or both. Yet this lack of confidence could be due to the time constraints that prevent teachers from investing time and effort in the design and implementation of such sessions (Davies et al., 2014). Therefore, future research should focus on what support teachers need from artists and from their schools to achieve the desired effects. Future policies should also recognize the importance of having dedicated time for outdoor arts-making, embedded within the curriculum, to avoid overwhelming the teaching staff.

Finally, through this review it was possible to identify the impact of specific activities on children and young people. Specifically, starting with simple, familiar and fun activities appeared to increase nature connectivity and make children and young people want to spend more time outdoors. Identity-focused and self-reflective activities led to some children gradually perceiving themselves as part of nature, and nature as part of their identity. Activities raising awareness of environmental issues led to higher pro-environmental attitudes and behaviors, while activities exploring ideas to prevent future environmental disasters appeared to increase agency and decrease eco-anxiety. Developing a manual or protocol of practice based on these activities would enable the design of an intervention that can be piloted for larger experimental studies and replicated with larger populations, such as schools, at a national and international level.

This review employed a rigorous methodological design as we conducted extensive searching across multiple databases and we contacted experts in the field to increase our confidence that we have identified the available evidence. We also employed standardized tools, such as the PRISMA guidelines and TIDieR template, to report all stages transparently. The synthesis of the studies in this review have allowed us to develop a framework that illustrates the additional benefits of engagement with the arts in nature and outdoor spaces.

However, it is possible that, although due diligence was taken when carrying out this systematic review, some gray literature, such as research papers that are not formally published in books or journals, may have been missed. Likewise, although a wide range of keywords and search methods was used, it is possible that studies might have been missed if they were inaccurately indexed or categorized in the electronic databases. Considering that the review was limited to English only publications, relevant studies may have been published in other languages.