Managing the Messy Middle: Toward a Typology of Dilemmas in Nature-Based Recreation

4.1.1 Leave no trace

Resource management scholars across disciplines have identified a recurring tension wherein recreators inadvertently degrade the ecosystems that support their experiences, even when their recreational behaviors are motivated by appreciation for the resource. For example, social scientists have written at length about the incentives that lead to overfishing in recreational commons, while natural scientists have focused on the ecological consequences of that overfishing (Jackson et al., 2001; Wilen, 2006). We refer to this pattern as the LNT dilemma. The degradation associated with LNT is exacerbated when resource user preferences align with natural infrastructure vulnerabilities. In extractive contexts, hunters and fishers may target a single trophic level, reducing biomass and transforming ecosystems in ways that hinder recovery (Chang et al., 2017; Weijerman et al., 2018).³ In non-extractive contexts, features that attract resource users (e.g., crevices for climbing, lagoons for boating, or cliffs for photographing) often coincide with sensitive ecosystems (e.g., cliff-dwelling plants, nursing dolphins, or nesting Golden Eagles) (Bomanowska et al., 2014; Shawky et al., 2020; Spaul & Heath, 2016).

It is worth noting that because we coded cases that contained symptoms of failure the resource users discussed in each study were limited to actual or potential recreators. However, another important group of resource users are residents of gateway communities—towns adjacent to or overlapping National Parks and other public lands used for recreation. It is important to clarify that “use” refers to either active or passive use values. There are many different subgroups within the Resource User node (discussed at length in Anderies et al., (2019)) and many flows along link 1. Consequently, the benefits and costs associated with system failures are unequally borne. Recreational users and some community members that directly interact with the natural infrastructure (e.g., in a volunteer custodial capacity) actively use the resource, while other residents only enjoy passive use benefits (e.g., tax revenues from tourism). Regardless of how they interact with the resource, gateway residents are all potentially harmed by the natural infrastructure degradation associated with recreation (i.e., increases in litter or reduced aesthetic quality), but are also often documented engaging with the natural infrastructure in a volunteer custodian role (Smith & Zakaria, 2026).

LNT also emerges when slow exogenous changes in resource user preferences or natural infrastructure resilience create new problems or undermine existing management solutions. For instance, rising demand for camping and a growing preference for dispersed tent setups have led to more campers and the creation of ‘satellite campsites,’ both of which, for example, damage flora (natural infrastructure), at Boundary Waters Canoe Area Wilderness (Eagleston & Marion, 2017).

Managers can address LNT through links 4, 5, or 6. Link 4 interventions modify natural infrastructure directly, through habitat restoration (e.g., pollutant remediation to support overfished species) (Weijerman et al., 2018), spatial separation of resource users and natural infrastructure (e.g., beach nourishment to create isolated seabird nesting sites or wildlife overpasses to help game species evade hunters) (Burger & Niles, 2014; Chang et al., 2017); and behavioral deterrents (e.g., covering satellite campsites with rocks) (Eagleston & Marion, 2017).

Link 6 interventions use information or sanctions to reduce resource user pressure on natural infrastructure. Two types of information resource managers frequently employ are restrictions on tools (e.g., axes, high-impact fishing gear) (Eagleston & Marion, 2017; Weijerman et al., 2018) or on access (e.g., banning climbing hardware in sensitive areas) (Bomanowska et al., 2014). However, these measures are only effective when enforceable. For example, illegal hunting persists in remote Chinese villages due to weak enforcement (Chang et al., 2017).

Two other link 6 interventions include behavioral bans (e.g., species-specific harvest moratoriums to prevent harvesting down the food web (Chang et al., 2017; Weijerman et al., 2018)) and spatial restrictions (e.g., “no stopping zones” to protect active eagle nests or banning boats near the mouth of the lagoon on Samadai Reef to allow nursing dolphins to escape (Spaul & Heath, 2016)).

Education campaigns, which could be used to encourage some resource users to moderate their own impacts, are yet another link 6 intervention. These campaigns are more likely to succeed when natural infrastructure is charismatic and resource user impacts visible (e.g., a single lingering pedestrian can cause Golden Eagle nest abandonment and chick death) (Spaul & Heath, 2016) than when effects are cumulative and less observable (e.g., the fitness of nursing dolphin pairs is only appreciably impacted over repeated encounters) (Shawky et al., 2020). Where resource users face insufficient private incentives to abide by link 6 interventions (e.g., mandates or education), link 5 interventions like monitoring and enforcement may be necessary. Monitoring and enforcement, modify the resource user-natural infrastructure interaction as it is happening. For example, park rangers could enforce “no stopping zones” by establishing patrols and herding people through. However, monitoring and enforcement are often limited by budget constraints, diffuse access points, and social norms that shield violators. For example, Miscou Island and the forests of Xishuangbanna Dai Autonomous Prefecture (Xishuangbanna) are large with numerous official and unofficial access points, making them difficult to monitor for illegal ATV riding or hunting, respectively (Chang et al., 2017; Hogan et al., 2021). In such cases, enforcement technologies may offer a viable solution. For instance, managers at Samadai Reef might require GPS trackers on boats, negating the norms that keep human monitors from reporting illegal boating activities.

4.1.2 Hell is other people

Recreation scholarship has long documented conflict arising from crowding, norm clashes, and heterogeneity among user groups (Hammitt & Schneider, 2000). We call this trend the HOP dilemma, in which the presence or behavior of fellow resource users, rather than environmental conditions, drives the decline in experience quality. While HOP can be as simple as congestion caused by resource users with homogeneous preferences, we find the harm associated with HOP is often exacerbated by behavioral or value-based heterogeneity. HOP can emerge suddenly, as seen during the COVID-19 pandemic in Brazil, where beachgoers suddenly became hazards to one another (Pereira et al., 2021), or gradually, through repeated negative interactions. At the St. Anthony Sand Dunes (SASD), an influx of newcomers with risky ATV behaviors degraded the recreational experience for legacy riders, prompting their exit and accelerating a transition to dangerous norms (Hughes & Paveglio, 2019). Similar dynamics occur when resource users clash over incompatible values (e.g., recreational vs indigenous fishers) or activities (e.g., hikers vs mountain bikers), eroding empathy and fostering resentment or even violence (Brown, 2016; Nguyen et al., 2016).

Similar to LNT, residents of Gateway communities are a resource user group that is absent from the cases we coded but is relevant to HOP. Recreators flocking to high-value landscapes contribute to seasonal congestion, as well as lasting changes to community culture and increased housing prices (Rumore & Stoker, 2023; Smith & Zakaria, 2026).

Managers can address HOP by: 1) limiting the number of resource users being concentrated, 2) facilitating spatial or temporal dispersion, or 3) mediating the harms of proximity. Limiting concentration may involve direct restrictions, such as closures or capacity limits (soft human made infrastructure through link 6, enforcement through link 5), which are most effective for short-term exogenous shocks like pandemics. For persistent threats, indirect methods like closing parking lots (hard human made infrastructure through link 4, or enforcement through link 5), changing access rules (e.g., disallowing beach access via public transportation), or implementing gate fees can reduce resource user density—though only where access is monitorable (Pereira et al., 2021).

Facilitating dispersion can involve hard human made infrastructure (link 4), such as spaced beach umbrellas, or modifications to resource users’ decision processes around access (link 6), like encouraging vehicle use to spread resource users across a site. Managers can spatially and temporally allocate access by resource user group (e.g., separating indigenous from recreational fishers and hikers from bikers) when threats correlate with observable traits. However, legal constraints (from public infrastructure providers) and ecological constraints (from natural infrastructure) often limit this approach. For instance, the Land Reform Scotland Act of 2003 makes all paths multi-use, which means managers at Cairngorms National Park (Cairngorms) cannot legislate right-of-way to hikers or bikers (Brown, 2016). Similarly, a Supreme Court ruling limits the ability of fishery managers to separate First Nation and recreational salmon fishers on the Fraser River (Nguyen et al., 2016). This type of scalar diversity among environmental management agencies is a common challenge. Because public infrastructure providers often set mandates for several management agencies without specific attention to system context, those mandates may complicate on-the-ground management as is seen in the Cairngorms and Fraser River examples. The fishery case illustrates a further complication—certain natural infrastructure attributes may make spatial zoning impractical. Salmon migrate up-river from the sea to spawn, so access rights to salmon and other non-stationary natural infrastructure must be zoned both spatially and temporally, complicating the manager’s task.

Information-based interventions (link 6) can also help resource users self-disperse, provided spatial or temporal zones are identifiable and monitorable. Managers at Cairngorms could install trail cameras and publicize live trail counts to help hikers and bikers avoid each other (Brown, 2016). However, in open systems like SASD, monitoring capacity is limited and resource users may find it difficult to employ regional resource user counts to avoid busy areas (Hughes & Paveglio, 2019).

To mitigate the harms of proximity, managers can regulate behaviors through speed limits or safety rules (link 6, and monitoring through link 5), or involve long-time users in governance (resource users = public infrastructure or an active link from resource users to public infrastructure via consultative processes or similar) to formalize legacy norms into site-specific policies (Hughes & Paveglio, 2019). Information, like trail etiquette campaigns or live trail counts, may also reduce conflict, especially when it highlights shared values, emphasizes common enemies, or explains the origins of others’ access rights (Brown, 2016; Nguyen et al., 2016). However, both laws and information require enforcement or buy-in. In Brazil, for example, skepticism surrounding COVID-19 undermined compliance with social distancing rules (Pereira et al., 2021).

4.1.3 Don’t poke the bear

Recreation scholars frequently document cases of resource users putting themselves in peril to satisfy their needs for sensation- or thrill-seeking (Haegeli & Pröbstl-Haider, 2016). We classify these instances as examples of the DPB dilemma, which arises when resource users are drawn to inherently risky experiences that offer higher U but expose them to harm from natural infrastructure. This trend is evident on Japan’s Numameguri Hiking Trail (NHT), where hikers seek brown bear encounters (Kubo & Shoji, 2016), and in Western Australia, where beachgoers walk on a sandbar despite known drowning risks (Gstaettner et al., 2017). In both cases, prohibiting the activity outright would reduce U, especially for thrill-seeking resource users.

Managers can intervene by restricting hazard access with soft human made infrastructure (e.g., trail closures or sandbar access bans) or hard human made infrastructure (e.g., trail gates or barricades along the sandbar) that restrict access to hazards (link 5) (Gstaettner et al., 2017; Kubo & Shoji, 2016). However, some managers may consider the reductions in U from such prohibitions to be unacceptable. Instead, managers might use narrative and information campaigns (link 6) to help resource users accurately assess and prepare for their personal risk levels. For example, narrative campaigns that highlight fatalities among strong swimmers may counter the cognitive dissonance of poor swimmers who use the presence of children and other seemingly vulnerable individuals to justify walking the sandbar (Gstaettner et al., 2017). Similarly, publicizing high-risk periods (e.g., the sandbar during high tide) or promoting safer practices (e.g., hiking in large groups in bear territory) can help resource users recreate more safely (Gstaettner et al., 2017; Kubo & Shoji, 2016).

4.1.4 Can’t get there from here

Many studies document how access to natural spaces is unevenly distributed across social groups. We refer to this recurring pattern as the CGT dilemma (Ghimire et al., 2014; Lemieux et al., 2012). This dilemma emerges when certain resource user groups face barriers to accessing natural infrastructure, often due to language, culture, or socioeconomic status. These barriers, while sometimes addressable, can also reinforce inequities. For example, Chinese immigrants in Vienna were hesitant to visit the Wienerwald Biosphere Reserve (WBR), because promotional materials used the term “hiking,” which they associated with strenuous activity, instead of their preferred activity, “taking walks in nature (Höglhammer et al., 2019). At Lake Superior, older, lower income, and first-time visitors were less able to adapt to climate-related weather shocks, making them more likely to exit the system (McCreary et al., 2019).

To address CGT, managers can include underrepresented groups in decision making (resource users = public infrastructure), ensuring outreach materials and access strategies reflect diverse needs and preferences. Recruiting community guides and co-designing promotional content with migrant resource users, as suggested by Hӧglhammer et al. (2019), can improve salience and trust.

When barriers to access are tied to sociodemographic factors, managers should assess what adaptations these groups can realistically adopt. For instance, lower income resource users may lack resources to temporally substitute or invest in technology adaptations, while older resource users may resist new technologies or be unable to physically adapt to extreme conditions. In either case, low-cost and low-tech risk mitigation strategies (e.g., shaded rest areas, gear sharing cooperatives) may help retain access for these vulnerable groups.

4.2.1 Counter-clockwise dilemmas

These dilemmas often stem from LNT-related challenges. For example, in Alaska, differentiated harvest rules for First Nation and recreational fishers—based on their distinct impacts on salmon stocks—triggered conflict over fairness, a Type A dilemma (Nguyen et al., 2016). Here, identity-based heterogeneity is both a policy target and a source of social tension.

Heterogeneity can be instrumental in determining system outcomes, whether or not it is targetable by policy. At NHT, trail closures after bear sightings disproportionately reduce U for non-locals, who value bear viewing more (Kubo & Shoji, 2016). In this case, differentiating information or rules by local status could improve outcomes. At Lake Superior, however, weather alerts disproportionately deter older and lower-income resource users (McCreary et al., 2019). These sociodemographic characteristics lead to differentiated outcomes but are not an obvious policy target. However, the Minnesota Department of Natural Resources—whose mandates include providing information and technical assistance—could provide gear or promote low-cost adaptation strategies—interventions that would be especially beneficial to older and lower-income resource users—to all Lake Superior recreators.

4.2.2 Clockwise dilemmas

Clockwise dilemmas often arise when interventions to support resource users—especially in CGT or HOP contexts—lead to natural infrastructure degradation. For example, creating ski trails or installing climbing routes involves directly modifying natural infrastructure (Type C) (Bomanowska et al., 2014; Carello et al., 2018), while building marinas increases resource user pressure on natural infrastructure by facilitating the spread of aquatic invasive species (Type D) (Bomanowska et al., 2014; Martínez-Laiz et al., 2019).

Clockwise dilemmas can also intensify HOP. At SASD, the erosion of legacy ATV norms by newcomers led to increased resource user harm (Hughes & Paveglio, 2019). Similarly, long-standing tensions between hikers and bikers at Cairngorms caused bikers to “disengage” and abandon their courteous use norms, degrading all resource users’ experiences (Brown, 2016). These cases illustrate how a collective action feedback loop that mediates or prevents some primary dilemma might be “hidden” to managers until an exogenous or endogenous force transforms norms and thus the nature of the clockwise feedback, rendering it problematic.

5.1.1 Resource user heterogeneity

Resource user heterogeneity—differences in recreation mode, values, knowledge, culture, or adaptability—can generate externalities by fueling conflict, reducing empathy, and complicating management. For example, tensions between First Nation and recreational fishers (Nguyen et al., 2016) or between legacy and newcomer ATV riders (Hughes & Paveglio, 2019) illustrate how differing values and norms can escalate dilemmas. Vulnerable groups, such as older or lower-income recreators (McCreary et al., 2019) or underserved immigrant communities (Höglhammer et al., 2019), may also be disproportionately affected by shocks or excluded from recreational benefits.

Targeting sources of resource user heterogeneity—when possible—can pose a significant challenge. For example, swimming ability and knowledge of sandbar dynamics—which loosely correlate with local status—affect drowning risk for sandbar walkers, but enforcing access bans only for non-locals would be impractical (Gstaettner et al., 2017). Even when policy can be tailored, other latent sources of heterogeneity may hinder efficacy. On the Fraser River, fishing rules are differentiated by the source of tension (indigenous status), and therefore raise new concerns about fairness that perpetuate inter-group violence (Nguyen et al., 2016).

Heterogeneity also shapes the distributional outcomes of management interventions. At NHT, non-locals who value bear sightings more are disproportionately affected by trail closures (Kubo & Shoji, 2016). Recognizing these asymmetries is essential for equitable and effective management.

5.1.2 Dilemma visibility

Even visible heterogeneity may be overlooked if linked to an invisible dilemma. For example, had managers not noticed the underrepresentation of immigrant resource users at WBR (i.e., if CGT had remained invisible), they might not have recognized cultural characteristics (visible heterogeneity) as relevant policy targets (Höglhammer et al., 2019). Managers can only act on visible dilemmas, so understanding what determines visibility is critical. We discuss how three factors—monitoring capacity, speed of emergence, and presence of mediating feedbacks—drive visibility in the following subsections.

5.2.1 Link 4: Modify natural infrastructure

Link 4 interventions directly modify natural infrastructure and include installing hard human made infrastructure to deter harmful behaviors (e.g., putting debris on satellite camp pads), reversing degradation (e.g., replacing eroded soil at official camp pads), and enhancing resilience (e.g., supporting overfished species through pollution remediation).

We identify two insights about link 4 interventions. First, they often trigger clockwise secondary dilemmas (e.g., grooming ski trails or installing climbing rings led to degraded natural infrastructure). Accordingly, managers must consider how resource users will perceive and respond to any changes to natural infrastructure. If interventions designed to redirect resource user efforts aren’t binding or lack buy-in, they may fail. For example, beachgoers in Pará ignored spaced umbrellas and tables (hard human made infrastructure) meant to encourage distancing (Pereira et al., 2021).

Second, even minor modifications can degrade natural infrastructure. For instance, the managers who installed marinas in the Mediterranean Sea or who groomed ski trails at Cucumber Gulch certainly did not intend the invasive species incursions that followed their actions (Carello et al., 2018; Martínez-Laiz et al., 2019). Data collection and modeling can help anticipate such unintended outcomes.

5.2.2 Links 5 and 6: Mediate the resource user – natural infrastructure feedback

Link 5 and 6 interventions aim to limit resource user impact on natural infrastructure by limiting access or intensive use (i.e., partially enclosing the commons), restricting technologies, or separating resource users from vulnerable natural infrastructure.

Partial enclosure of a recreational commons can involve limiting resource user quantity, pricing recreation, or limiting individual resource user’s impacts. Quantity-based interventions may be direct (e.g., harvest caps, access quotas) or indirect (e.g., limiting parking or access modes, as in Pereira et al. (2021)) and come with their own set of challenges. For instance, first-come-first-serve systems may allocate access inefficiently, favoring those with easier access over those who get more U from access (Holzer & McConnell, 2014). Tradable harvest tags (e.g., Chong et al., 2024; R. J. Johnston et al., 2007; Jungers et al., 2023) can address this access-U mismatch.

Price-based interventions include congestion pricing, access fees (Holmes & Englin, 2005; Richer & Christensen, 1999), or per-unit fees on intensive use (Jungers et al., 2023). Economists have long favored these interventions for the theoretical efficiency gains that come from rationing resources to those with the highest marginal value (Holzer & McConnell, 2014). However, price-based interventions can run up against issues of political palatability or equitable access.⁸ System-specific models can help managers assess feasibility and fairness.

Technology restrictions—which limit an individual resource user’s pressure on natural infrastructure—are common in extractive recreation (e.g., barbless hook rules, magazine limits), and increasingly in non-extractive contexts (e.g., e-bike bans) (Mitterwallner et al., 2021). However, these and other policies meant to limit individual impact (e.g., daily harvest limits in recreational fisheries) do not cap total impact, and may need to be paired with price- or quantity-based access restrictions to effectively limit pressure on natural infrastructure (Cox et al., 2002). Moreover, high-tech users are not always the most impactful. For example, Spaul and Heath (2016) found that pedestrians disturbed nesting eagles more than ATV riders. Therefore, recreation managers considering technology restrictions should carefully consider whether such policies will be effective given the resource user and natural infrastructure attributes of their systems.

Finally, managers may use hard human made infrastructure or soft human made infrastructure to separate resource users from vulnerable natural infrastructure. For instance, the New Jersey Department for Environmental Protection maintains a fence around key habitat for migrating shorebirds in the Brigantine Natural Area (Burger & Niles, 2014), and hunting in ecologically-important protected areas is illegal in Xishuangbanna (Chang et al., 2017). However, these interventions require buy-in or enforcement. In Xishuangbanna, hunters disregard bans and use information-sharing to avoid detection. Link 6 strategies, discussed next, may help build compliance.

5.2.3 Link 6: Appeal to resource users

Link 6 interventions influence resource user behavior through information, incentives, access restrictions, and gear restrictions. These strategies can improve compliance and equity, but their effectiveness depends on message design and context.

Information campaigns can raise awareness and promote safer or more sustainable behaviors. For example, McCreary et al. (2019) suggest providing adaptation support to retain older and lower-income resource users at Lake Superior, while Martínez-Laiz et al. (2019) recommend educating Mediterranean boaters about invasive species to encourage proper hull cleaning. However, information alone may be insufficient, especially when behavior change is costly. In such cases, pairing education with incentives (e.g., subsidies for hull cleaning or fines for noncompliance) may be necessary.

Cognitive dissonance can undermine information campaigns. At Mersey Point, for example, even weak swimmers rationalize risk by blaming drownings on others’ poor skills. Therefore, Gstaettner et al. (2017) suggest using counter-narratives (e.g., stories of strong swimmers drowning) to improve message salience.

Narratives and information can also foster empathy between resource user groups (Brown, 2016; Nguyen et al., 2016). However, to avoid unintended consequences, managers should survey resource users or involve them in message design. For example, Hӧglhammer et al. (2019) suggest consulting with resource users to draft multilingual, culturally resonant outreach at WBR, while Hughes and Paveglio (2019) recommend involving legacy users in governance at SASD. However, co-management must be approached carefully. In Xishuangbanna, using local informants for enforcement backfired, incentivizing retaliation. Instead, Chang et al. (2017) propose transferring hunting governance from national agencies to local villages to improve compliance.

While not addressed in our cases, narratives of care and stewardship and targeted information can also be used to encourage resource users to undertake conservation initiatives. For example, volunteers commonly preserve natural infrastructure used for recreation by cleaning up trash or maintaining trails (Propst et al., 2003). If system managers understand the motivations that might drive these volunteer actions, they can use information via link 6 to stimulate volunteer maintenance of recreational systems via link 1. Similarly, recreators can volunteer as citizen science observers, collecting information about the natural infrastructure through link 1 and sharing that knowledge with resource management agencies through link 6 (Rowe et al., 2021).

Finally, public infrastructure may limit access to protect vulnerable, high-demand natural infrastructure. In extractive contexts, this often involves licenses or quotas. In non-extractive settings, gate fees or visitor caps are common (e.g., at U.S. National Parks). Few cases in this analysis discussed limiting extensive use. This omission may suggest an implicit assumption that resource users should not be prevented from accessing natural infrastructure, whether for reasons of ethics or political palatability. Nevertheless, its widespread application in practice suggests access rationing remains a key tool—albeit one that must be balanced against equity and political feasibility.