Integrating space, spatial tools, and spatial analysis into the human dimensions of parks and outdoor recreation

Applied Geography 38 (2013) 76e85

Contents lists available at SciVerse ScienceDirect

Applied Geography journal homepage: www.elsevier.com/locate/apgeog

Integrating space, spatial tools, and spatial analysis into the human dimensions of parks and outdoor recreation J. Adam Beeco a, *, Greg Brown b a b

Clemson University, 137 Lehotsky Hall, Clemson, South Carolina 29634, USA University of Queensland, Australia

a b s t r a c t Keywords: Protected areas Parks GPS tracking Landscape values Space Time

The analysis of space and the use of geographic information systems (GIS) have long been important to natural resource applications. More recently, social scientists have been exploring ways to integrate spatial concepts with social science data related to natural resources for theoretical, practical, and methodological reasons. This trend is particularly evident with research in park and protected area (PPA) management and outdoor recreation. The purpose of this paper is to present an updated review of how space has been incorporated into PPA research, integrate concepts and methods, identify gaps, and propose future directions for research. Overall, this review suggests that the incorporation of spatiallyrelated social science data is advancing the field PPA research in an effective and viable way. Ó 2012 Elsevier Ltd. All rights reserved.

Park and protected area (PPA) management seeks to balance human use and influence with the protection of ecosystems and natural resources. Most public lands within the U.S. provide recreational access for society while at the same time trying to preserve natural resources for future generations. This dual mandate has created issues for public land managers seeking to sustainably manage such places. Public land managers and researchers focused on PPAs recognize and dedicate much of their efforts to addressing contradictions that arise from trying to provide use opportunities that inherently impact resources. For the past four decades, the concept of carrying capacity has been used as the primary conceptual basis for managing this dual mandate in PPAs. Carrying capacity has been defined as “the level of use beyond which impacts [on the biophysical resource and experiential quality] exceed levels specified by evaluative standards” (Shelby & Heberlein, 1984, p. 441). Carrying capacity at its most basic level seeks to identify the number of people and accompanying use types an area can accommodate without degrading the resource upon which the experience is dependent (Whittaker, Shelby, Manning, Cole, & Haas, 2011). How much and which types of use can take place without the deterioration of the resource and visitor experience is dependent on spatial and temporal variables. As such, carrying capacity is a spatial construct which requires understanding of the space available and its uses. Spatial variables have been underexplored * Corresponding author. Tel.: þ1 864 979 1926; fax: þ1 864 656 2226. E-mail address: [email protected] (J.A. Beeco). 0143-6228/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.apgeog.2012.11.013

within the concept and practice of carrying capacity and related planning frameworks (e.g., Visitor Experience and Resource Protection [VERP; National Park Service, 1997]; Limits of Acceptable Change [LAC; Stankey, Cole, Lucas, Peterson, & Frissell, 1985]). Where visitors recreate has been considered for recreational planning (Gobster, Gimblett, & Kelly, 1987); however, newer spatial technologies may offer unique, effective, and better approaches to determining recreation use distribution and incorporating this data into management techniques. Understanding the spatial context of both ecologically-based measurements and social data are needed to maintain a quality experience for visitors and adequate protection of resources. Space is a platform where different types of data can be integrated, including economic, ecological, and social data (Goodchild & Janelle, 2004). Specifically, spatial mapping and geographic information systems (GIS) are valuable planning tools when balancing multiple use claims on natural resources (Vries & Goossen, 2002). However, integrating spatially-related social science data into GISbased PPA planning is challenging because of measurement issues. In particular, social science data are rarely location specific and analysis of social science data are often difficult to integrate within spatial planning models (McIntyre, Moore, & Yuan, 2008). Despite these difficulties, the need to incorporate spatial data in planning is evident because recreational experiences in protected areas are a spatially-conditioned process. Páez and Scott (2004) review four spatially-conditioned processes that reveal a relationship between space and human-related phenomena. Three of the four of these processes influence the visitor distribution in PPAs.

J.A. Beeco, G. Brown / Applied Geography 38 (2013) 76e85

 Spatial diffusion. Diffusion is a geographical pattern of spatial distribution from concentration to dispersion. Recreational use in parks is highly concentrated in certain areas (e.g., parking lots) but highly defuse in other areas (e.g., trails; D’Antonio et al., 2010).  Spatial segmentation. Segmentation is the partitioning of a formerly homogeneous region (protected area) into two or more sub-regions (e.g., primitive, rustic, or concentrated). The Recreation Opportunity Spectrum (ROS) and related zoning management practices are an example of spatial segmentation within PPA management (Manning, 2011).  Spatial interaction. A spatial interaction process is evident when one area of space affects other areas. Yogi Berra’s oxymoronic quote “Nobody goes there anymore. It’s too crowded,” exemplifies the spatial interaction of visitor use in PPAs. PPAs are generally known for specific features many visitors seek to experience, while other visitors seek to experience areas that are less crowded. A fourth spatially-conditioned process that is highly relevant to PPA planning is impacts to the resource. While the amount of visitor use impacts the resource, impacts are also dependent on other spatial factors independent of use levels. For example, the recreational impacts on trail systems demonstrate the importance of spatial considerations when understanding visitor use. On trails, visitors are likely to encounter varying levels of trail degradation; however, volume or type of use is likely to appear consistent. In other words, while use is consistent across a space (i.e., the trail), impacts vary across that same space. This has led researchers to identify that impacts along a single trail vary according to spatial dimensions (e.g., landform and trail design; Weber, 2007). Despite these four logical spatial processes, the examination of space from a visitor use perspective has generally been limited. While other fields, such as urban planning, have made a swift adoption of spatially-related social science data for planning and analysis (Páez & Scott, 2004), researchers of PPAs have lagged behind. This is perhaps largely due to the type of data needed for analysis in each field and the ease of access to relevant spatially-related social science data. For example, census data is readily accessible and useful for urban analysis, yet its utility is limited for PPAs. However, in the past decade park and conservation area research has seen a growth in the integration of spatial-related social science data and analysis. The objectives of this paper are to identify how social science theory has approached visitors’ spatial movement, review some of the more pertinent attempts of PPAs to incorporate spatiallyrelated social science data, outline some of the conceptual and analytical difficulties with the inclusion of spatially-related social science data, and address the importance of incorporating spatially-related social science data into current management frameworks. First, this paper will review some of the theoretical attempts to conceptualize visitor travel patterns including typologies, space-time budgets, and landscape values. Next, this paper will examine some of the current methods for mapping visitor use: 1) the utilization of GPS for visitor tracking (D’Antonio et al., 2010; Hallo, Manning, Valliere, & Budruk, 2005); 2) the spatial modeling of recreation terrain suitability indexes (borrowed from conservation biology; Kliskey, 2000); and 3) the mapping of recreational impacts (this is actually the mapping of ecological indicators, so it is only loosely considered a social variable). Finally, we will discuss the current state of the field with respect to the integration of spatial considerations into recreation carrying capacity models. Space and recreation use The importance of space and time in PPA management has been recognized (Berkes & Folke, 1998). Understanding the locations of

77

visitors, their travel routes, and the amount of time spent at these locations are some the most basic, but relevant data on recreation (Hallo et al., 2012). The spatial extent of visitor use has impacts both on the physical resource (Hammitt & Cole, 1998) and the social experience (Manning, 2011) of visitors to PPAs. However, much of the theoretical attempts to conceptualize visitor behavior come from tourism research. Nevertheless, researchers of recreation and tourism management have had mixed success identifying variables that consistently account for differences in the spatial patterns of visitors (Shoval & Isaacson, 2010). There is, however, consistent conceptual support that landscape characteristics and spatial behavior are connected (Brown & Reed, 2009; Kliskey, 2000; Shoval & Isaacson, 2010). Spatial diffusion, segmentation, interaction, and spatiallyrelated impacts all suggest that recreation in PPAs is a spatiallyconditioned process. Additionally, early PPA research suggests that visitor use concentration and related impacts can be understood in the concept of ‘nodes and linkages’ (Manning, 1979). Nodes and linkages suggest that visitors concentrate in specific destination areas, known as nodes (e.g., waterfalls, campsite, river put-ins), that are connected by trails or roads (linkages). Furthermore, uneven spatial distribution is nearly a universal finding in user distribution research (Manning, 2011). As such, researchers have attempted to understand and conceptualize visitor use in a number of ways. These include non-spatial and spatial typologies (Lew & McKercher, 2006; Shoval & Isaacson, 2010), space-time budgets (Fennel, 1996), and landscape values (Brown, 2005). Typologies. Devising tourist typologies has long been an interest of tourism researchers. Starting with Cohen’s (1972) four-fold typology of the drifter, explorer, individual mass tourist, and organized mass tourist, researchers have generally approached tourist typologies from both psychological (Plog, 1972) and sociological (Cohen, 1972) perspectives. Tourist typologies have also focused on destination choice and travel style. Plog proposed that travelers’ personalities fell on a spectrum ranging from more confident, independent, and curious allocentric to more insecure, dependent psychocentric personalities that prefer familiar destinations and take part in package tours (1991). As the literature has grown, more complex typologies have been proposed and sub-categories have been developed within existing typologies. For example, twenty different travel styles were identified by Park, Tussyadiah, Mazanec, and Fesenmaier (2010). They condensed the most common travel styles into ‘Sight Seeker,’ ‘Family Person,’ and ‘Beach Bum’ such that 99.9% of all respondents chose at least one of these three as among their top travel personalities. For theoretical and practical purposes, tourism researchers have continued to group tourists into various types and categories. However, the majority of these studies have been based on nonspatial data, while fewer studies have dealt with the spatial activity of tourists (Shoval & Isaacson, 2010). Therefore, combining spatial movement patterns with non-spatial visitor characteristics provides a powerful and insightful tool for understanding tourist behavior and how psychological and sociological typologies affect or align with travel patterns. Traditional non-spatial typologies classify tourists based on their personality, interaction with the destination, or other characteristics (e.g., Cohen, 1972; Plog, 1972). Spatial typologies, however, group tourists according to how they “consume the space” at the destination. Although tourist typologies can be spatial or nonspatial, the two are not necessarily mutually exclusive. Research on spatial tourist typologies can use non-spatial data (e.g., demographics) to explain the difference in people’s travel patterns. Likewise, spatial data can also increase the breadth and depth of nonspatial typologies by adding an understanding of travel behavior, as well as validate or expand upon existing tourist typologies.

78

J.A. Beeco, G. Brown / Applied Geography 38 (2013) 76e85

Lew and McKercher (2006) modeled visitors’ spatial activities to conceptualize different travel patterns. They suggested two different spatial models: territory-based and linear. The territorybased model refers to the distance a visitor travels away from a primary location (e.g., hotel). The territory-based model can then be related to visitor characteristics (i.e., motivations, travel style, knowledge of destination). Lew and McKercher (2006) suggest that linear models are strongly influenced by the destination’s design and infrastructure. As such, the linear model aims to differentiate between actual travel patterns including point-to-point patterns, circular patterns, and complex patterns. However, these typologies were deductively designed and not based on empirical study. More recently, a separate study sought to test a psychometric travel style’s (wanderers and planners) influence over actual visitor travel patterns (Beeco, Hallo, & Manning, in press-a). This study found that the tourist typology of wanderers or planners did not affect visitor travel patterns. Parks Victoria developed a typology of visitors to national parks with strong implications for spatial travel patterns and use of facilities within the park system (Zanon, n.d.). The typology describes seven categories of park visitors: nature admirers, urban socials, trail users, passive users, activity centrics, access made easy, and country vacationers. Urban socials and trail users are most likely to use urban parks while nature admirers, country vacationers, and trail users are dominant user groups in regional parks. Spatially, trail users will focus their park use activities along established trail routes while urban socials will seek out suitable locations for large social gatherings. Parks Victoria also commissioned a separate study of visitors to Port Campbell National Park to develop a visitor typology based on park travel patterns (see Arrowsmith & Chetri, 2003). Using GPS technology, the study monitored visitor movements within the park. The spatial use data was combined with socio-demographic information to generate a typology of four groups: the single groupie, the international couple, the elderly couple, and the local family. The study found that singles and international couples stayed longer and visited more park nodes than elderly couples and local families. While the validation of park and tourist typologies that examine on-site visitor travel patterns is still needed (Beeco, Huang, et al., in press-b; Shoval & Isaacson, 2010), these typologies suggest potential influences of travel behaviors in PPAs. For example, groups that differ by motivations or activity may also vary in travel patterns. Space-time budgets. Time is an important aspect to tourism and recreation, including both the time needed for leisure and the time budget during a leisure experience (Fennel, 1996). A time budget is described as “a systematic record of a person’s use of time over a given period” (Shoval & Isaacson, 2010, p. 31). The extension of this concept to the space-time budget appear early in research of activity patterns, because any point in time is associated with a point in space (Anderson, 1971). Fennel (1996) expanded the space-time budget concept to suggest the longer a group spends in a certain area (zone) the “greater the implications to that area regarding social, environmental, and economic” (p. 816) impacts. Fennel’s model also adopts the economic theory of core-periphery, suggesting that visitors likely begin a trip to an area at the core (a locale’s central economic area for the development and distribution of goods and services) and depending on time and pressure, venture to see more in the periphery (an area directly dependent on the core) of the locale. A space-time budget requires participants (typically tourists) to complete time-space diaries (Note: the short comings of each of these methods are discussed at length in other texts [D’Antonio et al., 2010; Hallo et al., 2012; Shoval & Isaacson, 2010], but include issues of accurate recall, a lack of precision, and difficulty

reading maps). This model and other space-time budget research have revealed the important relationship between time budgets and spatial movement of visitors (e.g., Shoval & Isaacson, 2010). Additionally, as Fennel (1996) suggests, there is a direct link between space, time, and visitor impacts (social, environmental, and economic) to an area. While time-space models have only focused on tourism destination, they would make a logical and valuable contribution to the understanding of visitor travel patterns and experiences in PPAs. Landscape values mapping. Landscape values mapping is perhaps the most common form of social spatial data used for PPA planning and management. Landscape values mapping is an interpretive approach to spatial mapping (McIntyre, Yuan, Payne, & Moore, 2004) that provides an “operational bridge” to mapping place attachment (Brown, 2005, p. 19; Brown & Raymond, 2007). Typically, landscape values data are collected in point form. These points represent values that participants hold for specific areas of land (e.g., aesthetics, recreation, economic, ecological). These spatial data can be displayed individually, but more commonly, collectively to identify areas which are generally valued, allowing for the examination of where people most likely visit (Brown & Reed, 2009). Perhaps the greatest importance of landscape values data is the richness it provides researchers and managers. The values associated with specific places can provide insight into visitors’ desired management strategies. For example, if an area is identified as being ecologically important, it is probably not suitable for intense development. More specifically, landscape values can be used to identify the compatibility of controversial recreational activities such as motorized use in protected area locations (Brown & Reed, 2011). To assist in this planning process, Brown and Reed (2009) identified five domains for analyzing landscape values: 1) the relationships between different landscape values; 2) the relationships between landscape values and activities and/or policies with forest management; 3) the modeling of compatibility with forest plans; 4) the relationship between landscape values and biophysical impacts; and 5) the relationship between landscape values and actual public use. Recently, landscape values mapping has been used in planning for national parks (Brown & Weber, 2011), national forests (Brown & Reed, 2009), conservation areas (Brown & Weber, 2012a, 2012b; Pfueller, Xuan, Whitelaw, & Winter, 2009), scenic byways (Brown, 2003), allocation of protected areas (Raymond & Brown, 2006), as an addition to traditional survey research (Brown, 2005), tourism planning (Brown, 2006; Brown & Weber, 2012a, 2012b), and recreation management (McIntrye et al., 2004). According to Alessa, Kliskey, and Brown (2008) “landscape values have an implicit or potential set of prospective human activities and specific land uses” (p. 37) which can assist in PPA planning processes. Public Participation GIS (PPGIS) is often the mechanism used for collecting landscape values data. PPGIS is used to gather information on landscape values to support natural resource related planning processes by relating planning information to the public, expanding stakeholder representation, facilitating a better understanding of information through visualization of extent and impacts, and evaluating and weighting alternatives used in graphical interfaces (Sieber, 2006). PPGIS is also used to gather local knowledge of spatial characteristics of areas using GIS technologies (Sieber, 2006). In short, PPGIS helps protected areas by increasing the amount and diversity of public participation, and provides place-specific information for planning (Brown & Reed, 2009). Overall, typologies, space-time budgets, and landscape values are some of the more established theoretical attempts to conceptualize visitor travel patterns in tourist destinations and PPAs. These conceptualizations of visitor travel patterns have all

J.A. Beeco, G. Brown / Applied Geography 38 (2013) 76e85

contributed to an understanding of how space is related to visitor behavior. While the theoretical conceptualization of visitor travel patterns has been useful, it has been severely limited by the difficulty of accurately measuring visitor travel patterns. GPS tracking of visitor use. GPS tracking of visitor use has been shown to alleviate many of the problems associated with traditional methods for measuring visitor travel patterns (D’Antonio et al., 2010; Hallo et al., 2005, 2012; Shoval & Isaacson, 2010). GPS tracking of visitor use has increasingly been used in PPA research over the last five years. Recently, GPS-based measurements were used for backcountry visitor monitoring to measure use behavior and use intensity (D’Antonio et al., 2010). Visitor behavior, specifically off-trail use, waiting times at destination points, and average visit times were calculated to inform management of visitor experience and resource impacts. Additionally, point density analyses were used to identify areas of high and concentrated use. Another study at the Blue Ridge Parkway in Virginia also utilized GPS tracking of visitors and revealed that GPS methodology could be used in a complex network of roads and sites (Hallo et al., 2012). GPS tracking of visitor use has also been used to better understand recreational conflict between visitors and wildlife at Lake Umbagog Wildlife Refuge in New Hampshire (Beeco, Hallo, et al., in press-a, chap. 6). Specifically, GIS data layers were used to identify suitable habitat for nesting Common Loons (a federally listed species of concern and state threatened species in NH). The Common Loon is highly sensitive to disturbances, so visitor interactions with Common Loon nesting sites are a major concern of managers. GPS units were distributed to visitors of Lake Umbagog to track use patterns. GPS data were then overlayed with suitable habitats to better understand the locations of potential humanloon conflicts. Perhaps the greatest benefit of GPS methodology is the ability to link travel patterns with survey data. This enables researchers not only to understand spatial distribution, use intensity, and patterns, but also some of the visitor characteristics, desires, and motivations measured in traditional visitor use surveys (Arrowsmith & Chetri, 2003; D’Antonio et al., 2010; Hallo et al., 2005). For example, by linking GPS and survey data, researchers can examine how motivations influence travel patterns or how activity type (e.g., mountain biking, hiking) influences travel distance. A study conducted in Acre, Israel connected tourist GPS data with visitor sociodemographic information (Tchetchik, Fleischer, & Shovel, 2009). The most novel aspect of this study was the creation of polygons (created in GIS) over the study area that were used to segment differing areas of use. Each polygon represented specific areas within a tourist destination. GPS data revealed that 40% of the visitor groups never left the first area upon entry to the site. Therefore, visitors were segmented into groups that stayed in this area and groups that went outside the entry area. Other factors such as total time at sites were also included (Tchetchik et al., 2009). The connecting of GPS data with survey data was also demonstrated (as discussed above) by testing tourist typologies of wanderers and planners on the Blue Ridge Parkway, Virginia (Beeco, Huang, et al., in press-b), which found that psychometric survey data did not match actual travel behavior. Overall GPS tracking of visitor use in protected areas is seen as an accurate, effective, and cost effective method for measuring visitor distribution. However, significant advances can be made with respect to using GPS data to better understand recreation impacts, influences of visitor travel patterns, and integration with protected area planning. Recreation simulation models. There is a field of experimental spatial research that simulates the behavior of recreationists in high use natural environments rather than actual monitoring of visitor behavior. For example, RBSim is a computer program that uses GIS

79

to represent the physical environment where virtual, autonomous human agents simulate human behavior within a geographic space (see Gimblett, Itami, & Richards, 2000; Itami & Gimblett, 2000). The purpose of simulating visitor behavior is to provide insight into the probable effects of a given management scenario on visitors’ flows and movements, and indirectly, visitor satisfaction. The major value of simulation modeling is the incorporation of both time and space as a condition of a larger process (e.g., visitor encounters), so you can run alternative testing to estimate how different factors affect the larger process. For example, Hallo and Manning (2010) used simulation modeling to estimate how an increase in vehicle traffic would impact the acceptability of the scenic drive along Ocean Drive in Acadia National Park. This specific example focused on time, while space was broken into zones to examine when and where standards were being violated. While computer simulation has the capacity to integrate multiple physical and social variables to predict visitor behavior, the simulated travel patterns within PPAs must ultimately be ground-truthed with real-world spatial monitoring. In other words, ‘garbage in, garbage out.’ Recreational terrain suitability index. Land use suitability mapping is an analysis which assists in determining the best location for a variety of developmental actions given a set of goals and other criteria (Lyle & Stutz, 1983). Suitability mapping is a technique mostly associated with GIS that determines the appropriateness of a specific space in terms of a physical or socioeconomic factor (i.e., a specific use; Hepner, 1984). Suitability mapping has been used extensively in conservation biology (Kliskey, Lofroth, Thompson, Brown, & Schreier, 1999), city and regional planning (Steinitz et al., 2003), and natural resource planning (Brown & Reed, 2011; Reed & Brown, 2003). Suitability mapping is also beginning to emerge in recreation planning, although it has seen limited use. Since certain landscape characteristics are preferred for specific recreation activities, there is a relationship between landscape characteristics and resource activities (Kliskey, 2000). For example, Kliskey (2000) completed a recreation suitability analysis to identify specific areas for optimal snowmobile use which led to the development of the Recreation Terrain Suitability Index (RTSI). Specifically, RTSI maps resources based on user attitudes or preferences for landscape characteristics. This methodology is broken down into 4 stages (Kliskey, 2000). First, preferred landscape characteristics are identified for each recreational activity (e.g., what are the slope preferences for snowmobilers). Second, the landscape characteristics are converted to spatial data. Third, the spatial criteria are weighted to represent the preferences of landscape characteristics per user group. Fourth, suitability overlay maps are used to identify the spatial extent of terrain suitability for each user group. RTSI mapping can be used to understand the spatial characteristics of recreation suitability at different scales (i.e., micro or macro) and for different activities. For example, in the snowmobiling suitability example, landscape features such as openness, road access, remoteness, slope, snow conditions, and topographic position were used to determine suitability within a single watershed (Kliskey, 2000). RTSI has also been used to examine the development of semi-primitive and rural outdoor recreation opportunities within an entire park (Gabriela, 2006). Gabriela (2006) used a different approach to inform the RTSI model. Identification of suitability variables was accomplished through a literature review rather than attitudinal scales. However the suitability variables were similar to Kliskey’s (2000), including topography, landscape characteristics, accessibility, and infrastructure. The analysis revealed that 4% of the area was highly suitable, 31% moderately suitable, 41% barely suitable, and 24% not suitable at all for semi-primitive and rural outdoor recreation. Terrain suitability

80

J.A. Beeco, G. Brown / Applied Geography 38 (2013) 76e85

mapping also allows recreation activities to be considered with other spatially-oriented management objectives such as timber harvesting and wildlife preservation to provide a “tool of considerable value when applied to multiple use management” (Kliskey, 2000, p. 42). Spatial tools and analysis in parks and outdoor recreation There are several notable issues with the current state of GIS and spatial data integration into park and outdoor recreation planning. Operationalizing and accurately measuring socially related spatial variables is perhaps the most difficult aspect of integrating social science data into GIS-based planning models because this type of data is rarely location specific (McIntyre et al., 2008). Yet, this is the area where social scientists have perhaps improved the most by including landscape values, RTSI, and GPS tracking of visitor use. One issue with the research to date is that most visitor spatial data have been analyzed at the descriptive rather than inferential level. This deficiency is most evident in the analysis of travel patterns (Hallo et al., 2012). Most studies have not ventured beyond visual analysis such as mapping (Chancellor & Cole, 2008), point densities (D’Antonio et al., 2010), overlays (Beeco, Hallo, et al., in press-a, chap. 6), or the mapping of measures of central tendency distributions (Hallo et al., 2012). This has led to an understanding of the effects of travel patterns (Chancellor & Cole, 2008) but not what influences travel patterns. Even simpler spatial analysis tools in ESRI ArcMapÒ have seen little use. For example, viewshed analysis could be used to assist in the spatial zoning of recreational activities by identifying hotspots of potential conflict and interaction between groups. Zonal statistics could be used to identify patterns of use or impacts within watersheds or different parts of a trail system. Mapping of preferences similar to RTSI on a micro scale could also help assist managers identify the best areas for specific types of use. The mapping of spatial social data is often analyzed only by visually identifying patterns and structures such as point densities to determine value “hotspots” (Alessa et al., 2008). This type of analysis often involves subjective judgments. For example, Alessa et al. (2008) defined landscape value hotspots as point densities that fell within the upper third of the kernel density range. Alessa et al. (2008) suggest that more robust approaches are needed, such as defining a threshold that statistically defines hotspots in density estimations. There is recent research that seeks to quantify spatial social data into landscape metrics similar to those used in landscape ecology (Brown & Reed, 2012). These metrics quantify human perceptions of place to describe the composition and configuration of landscapes from a social perspective. This information can provide management decision support, for example, by modeling which potential management activities are compatible with the values located in different places (Reed & Brown, 2003). Metrics can also be calculated based on different spatial attributes such as activities, experiences, management preferences, and facilities’ needs, among others. These metrics have the capacity to identify the suitability of various recreational activities with a range of social landscape conditions using methods similar to RTSI, except with social data. Recreational terrain suitability analysis could be extended to integrate social landscape metrics as suitability criteria. Spatial data can also assist researchers at a level beyond descriptive statistics and mapping (i.e., visual analysis, point densities). Interpolation and inferential measures can also be conducted using spatial data (Davis, 2002; Wong & Lee, 2005). These other spatial statistics have seen less use in analyzing social data, but are more widely used in geostatistics. A key concept in geostatistics is the regionalized variable. Regionalized variables have

“characteristics intermediate between a truly random variable and a variable that is completely deterministic” (Davis, 2002, p. 416). In essence, regionalized variables have some spatial dependence in distributions. Interpolation methods are appropriate for regionalized variables. Interpolation methods such as kriging (simple, ordinary, or universal) and inverse distance weighting are common geostatistics tools which have seen little attention with social science data. The advantage of interpolation methods is in estimating data values at unknown locations. Specifically, kriging is a generalized linear regression technique that does not require spatial independence or random samples (Davis, 2002). Kriging does require prior knowledge of the spatial covariance between data. This is retrieved from semivariance. The semivariance is the measure of spatial dependence between observations (Davis, 2002). Semivariance is most often displayed as semivariograms, a plot of the results. However, it can be difficult to adapt geostatistical techniques to the structure of social science data. To date, Alessa et al. (2008) has used the most sophisticated techniques when comparing point density mapping methods with spatial interpolation. In seeking the appropriate treatment of spatial social data (specifically landscape values data), they compared simple point density and kernel point density. The authors identified simple point density as the number of point features within a defined area, while kernel density mapping was described as fitting a smoothly curved surface around each point, which produces a circular area with a certain search radius. Alessa et al. (2008) also discussed the merit of interpolation methods such as inverse distance weighting and kriging. The authors found their social data (landscape values) did not adhere to the rigorous statistical criteria of kriging, but found that social data could be used in inverse distance weighting. However, interpolation maps were found to be “spurious and non-intuitive” (Alessa et al., 2008, p. 31). Davis (2002) states that “because geostatistics is an abstract theory of statistical behavior, it is applicable to many circumstances in different areas of geology and other natural sciences” (p. 254). The value that geostatistics holds for social science is currently unclear. This is due to the conceptual and statistical validity of estimating values at unknown locations. While physical characteristics, such as soil and rock type, can be estimated at unknown locations, social values, such as landscape values or estimating use from GPS tracking in trail systems, do not follow these same distributional trends. However, in certain GPS applications where visitor movements are not confined (e.g., open water) interpolation methods may still hold some value. For example, kriging could estimate use density for an entire lake based on a sample of GPS visitor tracking. Understanding the spatial dependence of data may still offer valuable information even when kriging is not appropriate for social science data. Examining the spatial dependence of GPS or landscape value point data provides a better understanding of the data. Specifically, a cluster of points typically reflects some type of behavioral change or important landscape feature. Furthermore, this spatial dependence may be of particular value for recreation ecology data. If the spatial dependence of impact data is high despite varying topographical considerations (i.e., trail slope and trail grade), this may reflect that use intensity influences the impact more than trail design. Additionally, the spatial dependence creates correlated errors within data, which needs to be properly addressed for accurate statistical inferences. Space in carrying capacity and experience management There is also a need for improved planning frameworks that integrate GIS and advanced tracking technologies for both tourism and PPA management. While attempts at these planning

J.A. Beeco, G. Brown / Applied Geography 38 (2013) 76e85

frameworks have been made, spatially-oriented concepts, models, and theories which can serve as a foundation to planners are still lacking (Boers & Cottrell, 2007; Dredge, 1999). Furthermore, many of the attempts at spatial integration have large limitations, making them less useful for broad generalizations and uses in other applications. For example, the sustainable tourism infrastructure planning (STIP) framework tried to incorporate carrying capacity standards by relying on ‘forest cover’ and ‘slope gradient’ as indicators (Boers & Cottrell, 2007). Forest cover and slope gradient are easily modeled spatially (thus measurable), but not manageable characteristics, a requirement for high-quality indicators of carrying capacity (Manning, 2011). Dredge (1999) argues that despite considerable advancements in tourism planning frameworks there is a distinct lack of spatial concepts, models, and theories that land use planners can draw. We argue this is also true for PPA planning frameworks, specifically recreational carrying capacity. Carrying capacity has generally drawn criticism in the fields of tourism and ecosystem management for its lack of consideration of the complex interaction and interdependencies between resources and stakeholders (Strickland-Munro, Allison, & Morre, 2010), yet it is the predominant framework for managing recreation within PPAs (Manning, 2011). Carrying capacity is linked to Hardin’s (1968) ‘tragedy of the commons.’ Hardin uses national parks as an example of tragedy if human uses of common resources are not managed in a sustainable way. Carrying capacity recognizes that there are limits on the environment’s ability to sustain different types of uses (Whittaker et al., 2011). “Where spatial claims for different functions often exceed the available amount of land” (Vries & Goossen, 2002, p. 5) spatial mapping and GIS functions can be an effective planning tool. These spatial claims are at the heart of many conflicts within PPAs. Yet, the predominant management concept of PPAs (i.e., carrying capacity) has often neglected this spatial dependence. Within the original four types of carrying capacity (i.e., ecological, physical, facility, and social), only physical capacity addressed “space parameters” (Shelby & Heberlein, 1984, p. 443). However, these space parameters were limited in their extent, referring to people per square foot or people per acre rather than considering space more holistically. This caused the physical concept of carrying capacity to be less useful to recreation managers (Wagar, 1974). This may be, in part, because space and movement are often so fundamentally obvious it is often overlooked (Lew & McKercher, 2006). Current carrying capacity frameworks consider only three dimensions: resource (ecological), experiential (social), and managerial (similar to facility), without space parameters seen as an individual dimension (Manning, 2007). Maximizing the space utility of natural area resources is of growing importance as recreation increases and the availability of land remains stagnant. A spatial context can provide a mechanism for integrating the products of various social sciences fields, particularly when natural resources and physical processes of the landscape are involved (Goodchild & Janelle, 2004). For example, spatial areas can be zoned for specific management objectives including recreation (e.g., concentrated, primitive), timber harvesting, or protection of sensitive habitats. The Recreation Opportunity Spectrum (ROS) is perhaps the first planning tool to consider spatial extent within an entire management area (Clark & Stankey, 1979). The ROS concept has been “adopted” within carrying capacity related frameworks including VERP and LAC (Manning, 2011, p. 196). However, ROS only considers types of recreational opportunity settings (e.g., primitive, rustic, concentrated) and their spatial distributions. This is a macro level analysis of recreational opportunities and excludes other

81

considerations, such as recreational demand, intensity of use, recreational impacts, and natural resource demands (e.g., timber harvesting). Overall, spatial approaches to recreation have focused on opportunities rather than suitability for specific activities (Gabriela, 2006). One recent approach for park planning that is conceptually similar to ROS has visitors identify the spatial locations of park experiences and perceived environmental impacts using public participation GIS methods (Brown & Weber, 2011). Mapped visitor experiences and perceived impacts can be used to identify park experience zones similar to the way that ROS experience classes are identified by managers. The fundamental difference is that ROS represents a top-down expert approach to PPA zoning whereas participatory GIS relies on empirical visitor experiences as the foundation for identifying PPA zones. These top-down and bottomup approaches to spatial zoning could be integrated in a PPA planning process. At the micro level, few attempts have been made to incorporate spatially-related social science data into recreational planning and carrying capacity frameworks such as VERP and LAC. The measuring and mapping of recreational impacts is perhaps the most advanced form of spatially-related social science data; however, these data are measurements of ecological impacts and not a true social science measure with the exception of the participatory GIS methods described in Brown and Weber (2011) where visitors identify perceived environmental impacts. Incorporating spatial considerations into recreational planning frameworks is perhaps the most direct contribution spatial data can make to visitor management of PPAs, particularly at a micro scale. A starting point for this incorporation could be to focus on how spatial data can produce a better understanding of visitor use distribution and travel patterns. Spatial analysis may also help with ‘positive people management’ rather than regulations and control of the ‘people problem’ (Burch, 1984). For example, spatial use information can reduce negative social encounters by zoning according to group preferences. Additionally, measures such as trail density (for which the U.S.D.A. Forest Service already has applied standards) or landscape fragmentation by trails could also assist in the recreational carrying capacity frameworks. For a more concrete example, this paper argues that even a simple visual analysis across three different studies areas using GPS tracking of visitor use reveals a clear trend. Specifically, an examination of Beeco, Hallo, et al. (in press-a), chap. 6, D’Antonio et al. (2010), and Hallo et al. (2012) suggests there are three different types of recreational systems when considering visitor movements: open (open water or trail-less area), semi-open (trail-less snow covered terrestrial landscape), and closed (a system that restricts use to linear patterns). These different systems can drastically influence visitor behavior. Open systems, which can be conceptualized as open water, do not restrict visitor movements, which allows for endless movement patterns (Fig. 1a). Semi-open systems, such as a cross country ski area, allow visitors to travel where they please, but are restricted by natural landscape features (Fig. 1b). Closed systems are more traditional terrestrial recreation areas that restrict visitor use to trails, roads, or specific sites (Fig. 1c). Even a coarse visual analysis can identify the different patterns of these three systems. These different types of recreational systems may influence the amount of visual or physical encounters which might affect use distribution or permitting systems. Geographical representations of spatially-related social science data Spatial data consist of two different types of data: vector and raster. Vectors are represented by points, lines, and polygons.

82

J.A. Beeco, G. Brown / Applied Geography 38 (2013) 76e85

Fig. 1. Three different types of recreational systems based on travel options: (a) Open recreational system Lake Umbagog, NH (Beeco, Huang, et al., in press-b, chap. 6); (b) Semiopen system at Grand Tetons National Park during the winter season (D’Antonio et al., 2010); (c) Closed recreational system at Blue Ridge Parkway (Hallo et al., 2012).

Collectively these features can represent an endless variety of shapes (Ormsby, Napoleon, Burke, Groessl, & Bowden, 2008). Raster data is often used for variables that have no distinct shape but have measurable values in space, such as rainfall, slope, and temperature (Ormsby et al., 2008). Specifically related to social science data, much of the attention has focused on vector data (i.e., GPS tracking of visitor use and landscape values), while RTSI uses combinations of vector and raster data to create spatial models (Kliskey, 2000). When considering vector data, combining each of the three types creates 10 different combinations (Table 1). Table 1 shows how spatially-related social science data has been used in past research and the possibilities for future study. Generally speaking, GPS data is limited to points and lines depending on the intent of the research question and type of analysis. Modeling visitor flows has not been completed in 3-D although tracking scuba divers with 3-D modeling would provide a very interesting and unique assessment of travel patterns. Landscape values and other participatory GIS attributes such as recreational activities have been mapped using both points and polygons. The use of either approach will converge on a collective spatial ‘truth’ provided there are enough observations, but the degree of spatial convergence varies by attribute type and the quantity of data collected (Brown & Pullar, 2012). Identifying points is generally easier for participants but the data demands for point collection are considerably higher than for polygons. Discussion and future considerations Future PPA planning frameworks at the landscape scale should attempt to integrate social, ecological, infrastructural, and economic factors that show explicit or implicit spatial relationships. Place-based conservation planning by the New Zealand Department of Conservation provides an illustrative example of the explicit consideration of larger-scale spatial information into the PPA planning process (Brown & Weber, 2012a, 2012b). Site specific, or recreational planning frameworks should also include the spatial dimensions of social, biophysical, and managerial factors that inform decisions. One example is proposed research by the New South Wales National Parks and Wildlife Service to have mountain bike and horseback riders identify site-specific locations of track

qualities related to personal motivations in an internet-based GoogleÒ mapping application (I. Wolf, personal communication, October 10, 2012). But for spatial frameworks to functionally exist, quality social spatial data is needed. We suggest three research needs for incorporating spatially-related social science data into the human dimensions of parks and outdoor recreation. First, additional piloting and evaluation of spatial methods and decision-support tools are needed for parks and protected areas. A valuable first step would be to expand use of existing spatial models, tools, and statistics. One example is the potential inclusion of landscape value mapping in visitor surveys conducted by the Visitor Services Project, a national human dimensions research program that has been surveying national park visitors since 1982 (see https://psu.uidaho.edu/c5/vsp). According to VSP staff, the program is considering a pilot survey that would ask visitors to identify the spatial location of values within the national parks (L. Le, personal communication, October 25, 2012). Of the current types of spatially-related social science data collected in PPAs, few have been explored beyond the most basic GIS tools. To be effective, spatial social science data must demonstrate its value to planning and management decision-support systems. Additionally, exploring the applicability of geostatistics could also further the field in ways currently unexpected. Foundationally, researchers should seek to identify ways to incorporate spatial concerns into current management frameworks such as LAC and VERP. Second, research is needed that uses space to foster a better understanding of how humans interface with the landscape. Understanding what influences visitor flow (e.g., travel patterns) and resource flow (e.g., timber harvesting area) is key to understanding how humans engage, utilize, and impact resources within a protected area. Typologies (spatial and non-spatial), simulations, landscape values, and space-time budgets all further this effort. For example, an adaptation of space-time budgets would be a valuable contribution to the PPA research for two notable reasons. First, space-time budgets could further the understanding of how available time affects travel patterns. For example, local visitors to a PPA may use the space in a way similar to a local park (e.g., walk their dog, go for a jog) limiting their visit time. However, non-locals may use the PPA as a destination, spending the entire day at the site, and possibly traveling further and seeing more sights than locals.

J.A. Beeco, G. Brown / Applied Geography 38 (2013) 76e85

83

Table 1 Geographic combinations of spatially-related social science data. Geographic combinations

Description

Human dimensions examples

Potential studies

Points on points

Visitor travel patterns around specific attractions (Beeco, Huang, et al., in press-b)

Points on lines

Visitor travel patterns on primary and secondary roads (Beeco, Huang, et al., in press-b). Mapping values and activities along major river corridors (Pfueller et al., 2009). Locating scenic byway qualities (Brown, 2003).

Points on polygons

Visitor travel patterns over specific habitat types (Beeco, Hallo, et al., in press-a, chap. 6). Mapping experiences and impacts inside national parks (Brown & Weber, 2012a, 2012b). Mapping values and preferences inside national forests (Brown & Reed, 2009). Identifying values inside “Wilderness” areas (Brown & Alessa, 2005).

Lines on polygons

Mapping extent of formal and informal trail networks (Marion & Wimpey, 2011; Manning, Jacobi, & Marion, 2006)

Mapping recreational boating networks on lakes.

Lines on lines

Mapping highway segments with outstanding qualities (Brown, 2003)

Visitor travel routes over roads or trails. Mapping density of recreational use along rivers, roads, trails.

Polygons on polygons

Mapping place attachment on public lands using polygons (Black & Liljeblad, 2006).

Measuring campsite areal extent versus campsite design (Marion, 1995). Mapping values, experiences, or other attributes on public lands using polygons.

Mapping travel patterns of scuba divers.

Points in 3-D

Lines in 3-D

Mapping recreational activities along wild/scenic rivers. Mapping locations of trail impacts.

Trail slope and alignment model for predicting trail conditions (Marion, Wimpey, & Park, 2011)

Mapping routes used by hang-gliders. Mapping scenic overflight paths.

Polygons in 3-D

Mapping floors within a historic home.

3-D objects in 3-D

Trail system within an underground cave system.

Arrowsmith and Chetri’s (2003) findings were similar to this idea. Second, space-time budgets could also give great insight into how visitor movements influence the experience. For example, a visitor carrying a GPS unit (with time stamp) could also be asked to carry a recorder and verbally express their experience while noting the

time. This could directly link time, space, and other social values such as preferences, feelings of conflict or crowding, experiences of solitude or accomplishments, landscape values, response to interpretive programs, and numerous other constructs. When considering resource flow, spatial approaches assist in managing multiple

84

J.A. Beeco, G. Brown / Applied Geography 38 (2013) 76e85

use forests where protection of specific habitat, timber harvesting regimes, and fish stocking must all be balanced with recreation. Spatial models could be used to effectively identify the relationships between visitor use and other management goals. Third, research is needed that uses space to better understand the relationship between use and impacts to the biophysical resource. As visitors interface with the landscape, impacts to the resource will inevitably occur. Mapping impacts (such as trail impacts) and visitor use distribution (such as GPS tracking) could provide valuable insight to how visitors affect the landscape and how the landscape affects the visitors. Furthermore, spatial analysis could engender a better understanding of the effects of concentrated versus dispersed recreation impacts or how activity type (e.g., horseback riding) affects trail conditions throughout a system. Finally, spatial modeling may provide the ability to predict impacts with commonly used data such as digital elevation models (DEMs) that would allow managers to identify areas of potentially high impacts with limited use of resources. While the incorporation of spatial parameters in the current application of recreational carrying capacity frameworks could substantially advance the field of recreation management, it must be acknowledged that future research will likely be constrained by agency budgets that are limited for human dimensions research. Future research featuring spatial information must not only show decision-support value, but it must be cost effective. The use of the internet and mobile device technology combined with public participation GIS (PPGIS) methods and volunteered geographic information systems (VGI), may assist agencies in collecting spatial information from PPA visitors that is cost effective. The fact that some recreation management agencies are now experimenting with these tools is encouraging. Space may very well be the new frontier for both practical and theoretical human dimensions research in parks and outdoor recreation.

Acknowledgments The authors would like to thank Dr. Jeffrey Hallo for a presubmission review of the manuscript.

References Alessa, L., Kliskey, A., & Brown, G. (2008). Social-ecological hotspots mapping: a spatial approach for identifying coupled social-ecological space. Landscape and Urban Planning, 85, 27e39. Anderson, J. (1971). Space-time budgets and activity studies in urban geography and planning. Environment and Planning, 3, 353e368. Arrowsmith, C., & Chetri, P. (2003). Port Campbell National Park Patterns of use: A report for the development of visitor typology as input to a generic model of visitor movements and patterns of use. Melbourne, Victoria, Australia: Parks Victoria Visitor Research, Retrieved from. http://www.srnr.arizona.edu/wgimblett/ ArrowSmith%20and%20Chetri%20GPS%20final%20report.pdf [11.07.12]. Beeco, J. A., Hallo, J. C., & Manning, R. GPS as a method for tracking visitors in natural environments. In J. M. Baas, & R. Burns (Eds.), Best practices in recreation resource planning: A resource guide for planners: Using GIS applications. State College, PA: Venture Publishing, (in press-a). Beeco, J. A., Huang, W-J., Hallo, J. C., Norman, W., McGehee, N. G., McGee, J., et al. Spatial analysis of travel routes of wanders and planners using GPS tracking. Tourism Geographies, (in press-b). Berkes, F., & Folke, C. (Eds.), (1998). Linking social and ecological systems: Management practices and social mechanisms for building resilience. Cambridge: Cambridge University Press. Black, A., & Liljeblad, A. (2006). Integrating social values in vegetation models via GIS: The missing link for the Bitterroot National Forest. Final report JFSP project no. 042-1-114. Missoula, MT: Aldo Leopold Wilderness Research Institute. Boers, B., & Cottrell, S. (2007). Sustainable tourism infrastructure planning: a GIS supported approach. Tourism Geographies, 9, 1e21. Brown, G. (2003). A method for assessing highway qualities to integrate values in highway planning. Journal of Transport Geography, 11(4), 271e283. Brown, G. (2005). Mapping spatial attributes in survey research for natural resource management: methods and applications. Society and Natural Resources, 18(1),1e23.

Brown, G. (2006). Mapping landscape values and development preferences: a method for tourism and residential development planning. International Journal of Tourism Research, 8(2), 101e113. Brown, G., & Alessa, L. (2005). A GIS-based inductive study of wilderness values. International Journal of Wilderness, 11(1), 14e18. Brown, G., & Pullar, D. (2012). An evaluation of the use of points versus polygons in Public Participation Geographic Information Systems (PPGIS) using quasiexperimental design and Monte Carlo simulation. International Journal of Geographical Information Science, 26(2), 231e246. Brown, G., & Raymond, C. (2007). The relationship between place attachment and landscape values: toward mapping place attachment. Applied Geography, 27(2), 89e111. Brown, G., & Reed, P. (2009). Public participation GIS: a new method for use in national forest planning. Forest Science, 5(2), 166e182. Brown, G., & Reed, P. (2011). Values compatibility analysis: integrating public values in a forest planning decision support system. Applied Spatial Analysis and Policy . http://dx.doi.org/10.1007/s12061-011-9072-x. Brown, G., & Reed, P. (2012). Social landscape metrics: measures for understanding place values from Public Participation Geographic Information Systems (PPGIS). Landscape Research, 37(1), 73e90. Brown, G., & Weber, D. (2011). Public Participation GIS: a new method for use in national park planning. Landscape and Urban Planning, 102(1), 1e15. Brown, G., & Weber, D. (2012a). A place-based approach to conservation management using Public Participation GIS (PPGIS). Journal of Environmental Planning and Management . Brown, G., & Weber, D. (2012b). Using Public Participation GIS (PPGIS) on the Geoweb to monitor tourism development preferences. Journal of Sustainable Tourism . Burch, W. (1984). Much ado about nothing e some reflections on the wider and wilder implications of social carrying capacity. Leisure Sciences, 6(4), 487e496. Chancellor, C., & Cole, S. (2008). Using geographic information system to visualize travel patterns and market research data. Journal of Travel & Tourism Marketing, 25, 341e354. Clark, R., & Stankey, G. (1979). The Recreation Opportunity Spectrum: A framework for planning, management, and research. USDA Forest Service Research Paper PNW-98. Cohen, E. (1972). Towards a sociology of international tourism. Social Research, 39(1), 164e182. Davis, J. (2002). Statistics and data analysis in geology. New York: John Wiley and Sons, Inc. Dredge, D. (1999). Destination place planning and design. Annals of Tourism Research, 26(4), 772e791. D’Antonio, A., Monz, C., Lawson, S., Newman, P., Pettebone, D., & Courtemanch, A. (2010). GPS-based measurements of backcountry visitors in parks and protected areas: examples of methods and applications from three case studies. Journal of Park and Recreation Administration, 28(3), 42e60. Fennel, D. A. (1996). A tourist space-time budget in the Shetland Islands. Annals of Tourism Research, 23(4), 811e829. Gabriela, N. (2006, October). A GIS-based approach of recreation terrain suitability mapping for resource use management in Apuseni Natural Park. Proceedings of the second International Conference on Environmental Research and Assessment, Bucharest, Romania. Gimblett, H. R., Itami, R. M., & Richards, M. (2000). Simulating wildland recreation use and conflicting spatial interactions using rule-driven agents. In H. R. Gimblett (Ed.), Integrating GIS and agent based modeling techniques for understanding social and ecological processes. Oxford: Oxford University Press. Gobster, P. H., Gimblett, H. R., & Kelley, B. B. (1987). Modeling forest recreation policy alternatives: a geographic information systems approach. Proceedings of the Second Annual International Conference, exhibits and workshops on Geographic Information Systems. Goodchild, M. F., & Janelle, D. G. (2004). Spatial integrated social science. Oxford: Oxford University Press. Hallo, J. C., Beeco, J. A., Goetcheus, C., McGee, J., McGehee, N. G., & Norman, W. C. (2012). GPS as a method for assessing spatial and temporal use distributions of nature-based tourists. Journal of Travel Research, 51(5), 591e606. Hallo, J., & Manning, R. (2010). Analysis of the social carrying capacity of a National Park Scenic Road. International Journal of Sustainable Transportation, 4, 75e94. Hallo, J., Manning, R., Valliere, W., & Budruk, M. (2005). A case study comparison of visitor self-reported and GPS recorded travel routes. Proceedings of the Northeastern Recreation Research Symposium e GTR-NE-326, USA, pp. 172e177. Hammitt, W. E., & Cole, D. N. (1998). Wildland recreation: Ecology and management (2nd ed.). New York: John Wiley & Sons. Hardin, G. (1968). The tragedy of the commons. Science, 162, 1243e1248. Hepner, G. F. (1984). Use of value functions as a possible suitability scaling procedure in automated composite mapping. Professional Geographer, 36(4), 468e 472. Itami, R. M., & Gimblett, H. R. (2000). Intelligent recreation agents in a virtual GIS world. In Proceedings of complex systems 2000 Conference. November 19e22, 2000. Dunedin, New Zealand: University of Otago, Retrieved from. http://www. srnr.arizona.edu/wgimblett/NZ_RBSIM_Final2.html [13.07.12]. Kliskey, A. D. (2000). Recreation terrain suitability mapping: a spatially explicit methodology for determining recreation potential for resource use assessment. Landscape and Urban Planning, 52, 33e43. Kliskey, A. D., Lofroth, E. C., Thompson, W. A., Brown, S., & Schreier, H. (1999). Simulating and evaluating alternative resource-use strategies using GIS-based habitat suitability indices. Landscape and Urban Planning, 45, 163e175.

J.A. Beeco, G. Brown / Applied Geography 38 (2013) 76e85 Lew, A., & McKercher, B. (2006). Modeling tourist movements: a local destination analysis. Annals of Tourism Research, 33(2), 403e423. Lyle, J., & Stutz, F. (1983). Computerised land use suitability mapping. The Cartographic Journal, 20(1), 39e49. Manning, R. E. (1979). Impacts of recreation on riparian soils and vegetation. Water Resource Bulletin, 15(1), 30e43. Manning, R. E. (2007). Parks and carry capacity: Commons without tragedy. Washington: Island Press. Manning, R. E. (2011). Studies in outdoor recreation: Search and research for satisfaction (3rd ed.). Corvallis: Oregon State University Press. Manning, R. E., Jacobi, C., & Marion, J. L. (2006). Acadia National Park visitor use and impact monitoring programs for carrying capacity decision-making. George Wright Forum, 23(2), 59e72. Marion, J. L. (1995). Capabilities and management utility of recreation impact monitoring programs. Environmental Management, 19(5), 763e771. Marion, J. L., & Wimpey, J. (2011). Informal trail monitoring protocols: Denali National Park and Preserve. Final Report. Blacksburg, Virginia, USA: USDI, U.S. Geological Survey, Virginia Tech Field Station, Retrieved from. https://profile.usgs.gov/ myscience/upload_folder/ci2012Feb2415041636429DENA%20Trails%20Final% 20Rpt.pdf. Marion, J. L., Wimpey, J., & Park, L. (2011). The science of trail surveys: recreation ecology provides new tools for managing wilderness trails. Park Science, 28(3), 60e65. McIntyre, N., Moore, J., & Yuan, M. (2008). A place-based, value-centered approach to managing recreation on Canadian Crown lands. Society of Natural Resources, 21, 657e670. McIntyre, N., Yuan, M., Payne, R. J., & Moore, J. (2004). Development of a valuesbased approach to managing recreation on Canadian Crown lands. Proceedings of the second International Conference on Monitoring and Management of Visitor Flows in Recreational and Protected Areas, Rovaniemi: Finland. National Park Service. (1997). Visitor Experience and Resource Protection (VERP) Framework: A handbook for planners and managers. Ormsby, T., Napoleon, E., Burke, R., Groessl, C., & Bowden, L. (2008). Getting to know ArcGISÒ desktop. Redlands: CA: ESRI Press. Páez, A., & Scott, D. (2004). Spatial statistics for urban analysis: a review of techniques with examples. GeoJournal, 61, 53e67. Park, S., Tussyadiah, I. P., Mazanec, J. A., & Fesenmaier, D. R. (2010). Travel personae of American pleasure travelers: a network analysis. Journal of Travel and Tourism Marketing, 27(8), 797e811. Pfueller, S., Xuan, Z., Whitelaw, P., & Winter, C. (2009). Spatial mapping of community values for tourism planning and conservation in the Murray River Reserves. Gold Coast, Queensland: CRC for Sustainable Tourism.

85

Plog, S. C. (1972). Why destination areas rise and fall in popularity. Cornell Hotel and Restaurant Administration Quarterly, 14(4), 55e58. Plog, S. C. (1991). Leisure travel: Making it a growth market.again. New York: Wiley. Raymond, C., & Brown, G. (2006). A method for assessing protected area allocations using a typology of landscape values. Journal of Environmental Planning and Management, 49(6), 797e812. Reed, P., & Brown, G. (2003). Values suitability analysis: a methodology for identifying and integrating public perceptions of forest ecosystem values in national forest planning. Journal of Environmental Planning and Management, 46(5), 643e658. Shelby, B., & Heberlein, T. A. (1984). A conceptual framework for carrying capacity determination. Leisure Sciences, 6, 433e451. Shoval, N., & Isaacson, M. (2010). Tourist mobility and advanced tracking technologies. New York: Taylor and Francis. Sieber, R. (2006). Public participation geographic information systems: a literature review and framework. Annuals of the Association of American Geographers, 93(3), 491e507. Stankey, G., Cole, D., Lucas, R., Petersen, M., & Frissell, S. (1985). The Limits of Acceptable Change (LAC) system for wilderness planning. USDA Forest Service General Technical Report INT-176. Steiniz, C., Rojo, H., Bassett, S., Flaxman, M., Goode, T., Maddock, T., et al. (2003). Alternative futures for changing landscapes: The upper San Pedro river Basin in Arizona and Sonora. Washington: Island Press. Strickland-Munro, J. K., Allison, H. E., & Moore, S. A. (2010). Using resilience concepts to investigate the impacts of protected area tourism on communities. Annals of Tourism Research, 37(2), 499e519. Tchetchik, A., Fleischer, A., & Shoval, N. (2009). Segmentation of visitors to a heritage site using high-resolution time-space data. Journal of Travel Research, 48(2), 216e229. Vries, S., & Goossen, M. (2002). Modeling recreational visits to forests and nature areas. Urban Forestry & Urban Greening, 1, 5e14. Wagar, J. A. (1974). Recreational carrying capacity reconsidered. Journal of Forestry, 72(5), 274e278. Weber, P. (2007). Principles of sustainable trails. In P. Weber (Ed.), Managing mountain biking: IMBA’s guide to providing great riding. Boulder, CO: International Mountain Biking Association. Whittaker, D., Shelby, B., Manning, R., Cole, D., & Haas, G. (2011). Capacity reconsidered: finding consensus and clarifying differences. Journal of Park and Recreation Administration, 29(1), 1e20. Wong, D., & Lee, J. (2005). Statistical analysis of geographic information with ArcView GISÒ and ArcGISÒ. Hoboken, New Jersey: John Wiley & Sons, Inc. Zanon, D. (n.d.). Parks visitor market segmentation. Retrieved from http://www. weplan.parks.vic.gov.au/sites/default/files/ParkVisitorSegmentationSummary_ 0.pdf [July 11.07.12].