Public Participatory Geographic Information Systems (PPGIS) on U.S. Public Lands and Waters

Globally, the use of PPGIS has been widespread in research evaluating visitor activities and public perceptions of values associated with public lands and waters (Brown, Weber, and Bie, 2014; Brown and Fagerholm, 2015; Hansen and others, 2021; Olafsson and others, 2022). In addition to the use of PPGIS for Browns Canyon National Monument, PPGIS has also been used in multiple other U.S. public lands contexts, mostly in national forests. For example, one such study evaluated public values for six national forests in Colorado and Wyoming and found that biocentric values (for example, future, biodiversity, and aesthetic values) made up the majority of responses (Bagstad and others, 2017). This finding is a common result among many PPGIS surveys across the world (Brown and Raymond, 2007; Nielsen-Pincus, 2011; Brown and Weber, 2012). Another study used PPGIS in two national forests in Alaska to understand how the public’s values for wilderness areas differ from nonwilderness areas, finding that life-sustaining, intrinsic, and spiritual values were more common in wilderness areas than nonwilderness areas (Brown and Alessa, 2005). One of those PPGIS studies within Chugach National Forest in Alaska resulted in a forest plan that was consistent with public preferences and values, rather than the alternative plan that was based only on biophysical characteristics (Reed and Brown, 2003). Lastly, Brown and Reed (2012) used a PPGIS mapping tool to support a values-compatibility analysis to determine the compatibility of proposed off-highway-vehicle routes in the Mt. Hood National Forest in Oregon. When accounting for the landscape values of the public, land managers would be able to select the placement of the off-highway-vehicle routes optimally (Brown and Reed, 2012). The main implication from PPGIS studies on public lands is that land managers can incorporate public attitudes and preferences when making decisions for public resource management and development projects on public lands and waters.

Despite the demonstrated use and value PPGIS brings to planning and decision-making processes, it has not yet been widely used across Federal agencies or locations, partially due to constraints with the OMB approval process and the length of time required to comply with the PRA (Brown, 2012). The studies reviewed in the paragraph above all took place in national forests and collected data over a decade ago. This study represents the first known PPGIS application that has been approved by the OMB for use by a DOI bureau. Additionally, researchers that previously used PPGIS noted that it is particularly useful for multiple-use landscapes, like national forests and BLM lands, because management decisions could affect land use (Brown, Kelly, and Whitall, 2014).

Study Site

Mojave Trails National Monument, located in southern California, was designated in 2016 and encompasses 1.6 million acres of Federal land managed by the BLM (BLM, undated; fig. 1). The national monument contains diverse landscape features, including sand dunes, lava flows, and mountain ranges (refer to recreation points of interest on fig. 1). The historic Route 66 (U.S. Highway 66) runs through the national monument, which is a highway that represents historic places and cultural resources from California to Lake Michigan (National Park Service, 2022). Along with recreational activities such as hiking, camping, and wildlife viewing, the Marble Mountains within Mojave Trails National Monument contain well-known sites for rockhounding and invertebrate fossil collecting (BLM, undated). The area is also known for the cultural and historic landmarks, such as the Mojave Indian Trails, which are historic Native American trading routes, and the Desert Training Center, which was a World War II training facility. There were an estimated 63,000 visits by recreators in 2022 (BLM, 2023).

Questionnaire Development

The questionnaire was developed by reviewing prior PPGIS studies and the associated lists of values and definitions used in those studies (for example, Brown and Reed, 2000; Brown, 2004; Allen and others, 2009; Brown, 2012; Brown and Raymond, 2014; Brown and Fagerholm, 2015). We then consulted with BLM collaborators from the Socioeconomics team and the National Operations Center to add and refine questions and values to better fit a BLM context, given the multiple-use landscape. For example, although not common in the prior literature, we added the values “personal income opportunity” and “community economic opportunity” because BLM lands are used for multiple purposes and can offer income opportunities. We also sent the survey to the Interagency Visitor Use Management Council to solicit additional feedback on questions and phrasing. After refining the survey instrument with these groups, we were asked to further edit values and definitions by OMB staff.

There were four sections on the survey: (1) experience use history, (2) GIS application to map uses of and values for the landscape, (3) change in use of the landscape due to land use change or management changes, and (4) demographics. We asked three questions related to experience use history because a person’s past experience and familiarity with the landscape can influence how they use and value that landscape (Moore and Graefe, 1994; Halpenny, 2010). Similarly, demographic data are useful to better understand the characteristics of the sample, which may also affect uses of and values for a landscape.

Data were collected online, and the mapping application asked participants to place points on the map representing their uses of and values for different locations, with Mojave Trails National Monument outlined on the map (refer to fig. 2). We used points rather than polygons because prior research compared the two methods and found that points were more favorable due to greater validity and lower cognitive complexity for participants (Brown and Pullar, 2012). Participants could place up to 30 points on the map (due to the software and how the application was developed). They were asked to select uses and values from a list of 16 different options, all of which had an associated definition. The list of values and their respective definitions can be found in table 1.

Following the mapping portion of the survey, participants were asked to rank their top 3 most important location-value pairs, and the following survey questions related only to these top 3 location-value pairs. This approach was used to understand more information about certain locations and values, but to only ask additional questions up to three times (rather than for every point they placed) to keep the survey response time and burden on participants lower. Participants were given a hypothetical scenario and asked how their use of the area would change if the scenario were to happen. For Mojave Trails National Monument, the scenario was recreational facility development and read: “In this hypothetical scenario, increased visitor use requires the development of new facilities to support camping, information dissemination, restrooms, and other needs. As a result, while you are benefiting from [insert value] you will occasionally see utilities, access roads and vehicles, and new development that are associated with this work.”

Based on this scenario, participants were asked if they would change the location they visit, the number of visits, the timing of visits, or the activities participated in, or if they would stop visiting altogether. Participants were then asked what factors would be important in deciding to substitute locations (for example, similar cost, similar travel time, similar use levels). Finally, we asked to what degree this scenario would affect the overall quality of their experience. These questions were each asked up to three times, once for each of their top three location-value pairs selected. These questions were used to better understand how uses of and values for the landscape would change under the recreational facility development hypothetic scenario. The full survey instrument can be found in appendix 1.

Application Development

All data were collected online through a website application developed by the USGS Fort Collins Science Center. The application was supported on all major browsers on computers, tablets, and mobile devices. The mapping interface uses Mapbox (Mapbox, 2024), and users were able to pan and zoom in on the map to locate specific features. Figure 2 shows a screenshot of one page of the application that has the mapping interface showing the map zoomed out on with points placed.

The development of VaMPIRE was a multiyear effort based on an original proof of concept that has been iteratively updated according to feedback from project partners. An open dialog between project scientists, BLM partners, and developers facilitated the creation of the first VaMPIRE prototype, which was gradually refined to accommodate the survey content for the pilot of Mojave Trails National Monument. The application code was written to remain flexible regarding survey location and survey question generation, such that it is possible to use VaMPIRE surveys for multiple locations simultaneously (for example, using different web addresses for each location). VaMPIRE is the culmination of interdisciplinary collaboration, incorporation of user feedback toward application improvement, and constraints to application content and appearance.

The VaMPIRE application has three main components: a front-end program to display the mapping interface and survey, a database server to store survey responses, and an application programming interface (API) program to facilitate communication between the front end and the database. Industry-standard technologies (for example, GraphQL and Docker) are used to securely and efficiently deliver a streamlined survey experience to users. The front end was designed for intuitive operation and extensively uses the U.S. Web Design System for the appearance of the application. The API and database support the display, filtering, and exportation of survey response data such that the data can be accessed via a graphical user interface, filtered using structured query language (SQL)-like commands, and exported in JavaScript Object Notation (JSON) format.

Data Collection

The sample of people contacted for survey participations resulted from an interested party list of Mojave Trails National Monument and included names and emails previously collected by the monument staff. The email list had 246 email addresses, of which 39 were returned as no longer valid. Therefore, the total potential sample size was 207. Surveys were conducted in January and February 2024. Emails were first sent out with an introduction, a description of the purpose of the survey, and a personalized link to the survey. The first reminder email was sent out a week after the initial email, and the last reminder email was sent out 3 weeks after the initial email (following guidance from Dillman and others, 2014). Each potential participant was sent a unique survey link so we could track if they completed the survey, forwarded the survey, or completed it multiple times. People only received the reminder emails if they had not completed the survey at the time the reminders were sent. The spreadsheet with names, email addresses, and unique links to participate in the survey was never connected to the database of survey responses, and all survey responses are anonymous.

Additionally, one participant who was emailed the survey posted the link to the survey in an off-roading-focused Facebook group, which garnered additional responses. Although soliciting responses through social media was not the original intent of this study, due to the large response from the group, we included these data and analyzed them separately from the email-list group when appropriate. The full dataset is available online (Lindley and Wilkins, 2025).

Data Analysis

All data were analyzed using R (version 4.3.1; R Core Team, 2022), and maps were created in R using the ggplot2 package (Kahle and Wickham, 2013). Given that this is the first pilot of the VaMPIRE tool and the research questions are more exploratory in nature, the results focus on presenting summary maps and descriptive statistics (for example, frequencies) for each survey question; we do not aim to provide formal statistical analyses or comparisons. We first used descriptive statistics to show demographics of respondents and their experience with Mojave Trails National Monument. To answer our first research question, we created a table with frequencies of each value that participants placed on the map. For our second research question, we plotted all locations placed within the boundaries of Mojave Trails National Monument to discern the most popular areas in the monument, regardless of associated value category. Locations were then separated by value category and represented in a map for each individual value category. These maps could aid in explaining how certain areas are valued, and the information could be later used in a compatibility analysis with specific management objectives to better understand the optimal location of certain projects, such as trails or restrooms.

Next, we depicted locations placed by respondents from the email-list group and those placed by respondents from the off-roading group. We mapped the locations separately for respondents from the email-list group and the off-roading group, then conducted a hotspot analysis to highlight the most used areas in Mojave Trails National Monument by each group. Hotspot analysis was conducted in R using sfhotspot (Ashby, 2023) using a cell size of 2,000 feet. We end by showing the frequencies of different responses regarding the hypothetical scenario and how it would affect use of the area and user experiences.

Response Rates and Characteristics of the Sample

There were 74 total responses from the email-list group out of 207 deliverable email recipients, which is a response rate of 35.7 percent. We obtained an additional 47 responses from the survey being posted in an off-roading group on a social media outlet. Combined, this represents a total of 121 responses.

Table 2 shows characteristics of respondents divided into these two groups: respondents from the email-list group and respondents from the off-roading group. Based on the respondent’s year of birth, respondents were grouped in six age bands, as can be seen in table 2. A majority (64.7 percent) of respondents from the email-list group were 60 years or older, while 46.7 percent of respondents from the off-roading group were 60 years or older. Very few people under the age of 40 from either group participated in this survey. Additionally, the samples for both groups had more male participants, with the off-roading group being particularly skewed toward men. For both samples, most respondents have a bachelor’s degree or higher, are white/Caucasian and not of Hispanic, Latino, or Spanish origin, and lived in California at the time of the survey.

Table 3 consists of summary statistics of respondents’ experience use history of Mojave Trails National Monument separated by respondents from the email-list group and respondents from the off-roading group. The two most common responses for the number of years the respondent has been visiting Mojave Trails National Monument were 6–7 years and 20 or more years for both samples. The most common responses for the number of visits to the monument in an average year were 1–2 times and 3–4 times a year for both groups, with 62.2 percent of the email-list group visiting 1–4 times a year on average, and 59.6 percent of the off-roading group visiting 1–4 times a year. Respondents from both groups visit most often in the fall and spring and visit least often in the summer. Additionally, most respondents indicated they visited in more than one season.

Values for Mojave Trails National Monument

Respondents placed a total of 1,069 values locations around the Mojave Trails National Monument area, 1,005 of which were specifically within the boundaries of Mojave Trails National Monument. The following results and figures use only those points placed within the monument boundaries. The number of points placed for each value category is shown in table 4. On average, respondents from the email-list group placed 8.77 points in the national monument per person, while respondents from the off-roading group placed 7.57 points in the national monument per person.

Across both groups, recreation accounted for over half of all points placed by respondents, with 52 percent of locations placed by the email-list group representing recreation, and 74 percent of locations placed by the off-roading group representing recreation (table 4). For the email-list group, all 16 of the value options were selected at least twice, while 2 value options were not selected by any participants in the off-roading group (biodiversity and life sustaining). Apart from recreation, the most common value choices from the email-list group were wilderness, inspirational (intellectual), education or learning, and biodiversity. For the off-roading group, aside from recreation, the most common values placed were therapeutic (mental health) and aesthetic.

Of the top three values ranked by each email-list respondent (185 total responses), 45.41 percent were recreation, 9.73 percent were inspirational (intellectual), and 8.11 percent were wilderness. As for the off-roading group, 107 total locations were placed in the top three values for each person. These responses were skewed more heavily toward recreation, with 78.50 percent recreation, 5.61 percent aesthetic, and 3.74 percent therapeutic (mental health). Of these points, the most common primary recreation activity for the off-roading group was off-roading (65.48 percent), whereas the most common primary activity for recreation for the email-list group was rockhounding (39.23 percent). The follow-up question asking about specific recreational activities was only asked for each individual’s top three locations.

Spatial Patterns of Values for Mojave Trails National Monument

Figure 3 depicts the specific locations respondents use or value (combining responses from both the email and off-roading groups). Areas with the highest concentration of points are in the northwest portion of the national monument, near the Afton Canyon Natural Area and the Pisgah Crater and lava flow area. While there is a higher concentration of locations in the northern area of the national monument, locations valued by respondents exist throughout the entire monument.

Figure 4 shows the locations placed by respondents, separated by value type. While every value category was used, recreation was the most common value associated with a location, whereas life sustaining and personal-income opportunity were the least common. Overall, the locations for each value were placed throughout the national monument, and there are no clear visual differences in how the different values cluster on the landscape.

Respondents from the email-list group placed a total of 649 locations, and respondents from the off-roading group placed a total of 356 locations (as described in table 4). The spatial distribution of locations respondents value comparing the email-list group to the off-roading group can be seen in figure 5. Respondents from the email-list group placed a higher proportion of locations in the northwest section of Mojave Trails National Monument, whereas the respondents from the off-roading group placed more locations in the northern and northeastern section of Mojave Trails National Monument more generally. This can further be seen in the hotspot analysis in figure 6. For both groups, a lower proportion of points were placed in the southern part of the national monument.

Perceptions of the Hypothetical Scenario

Lastly, participants were asked about how they would respond to the hypothetical recreational facility development scenario. Specifically, this question was asked up to three times per respondent for each of their top three location-value pairs (refer to table 5). A majority of respondents from both the email-list group and the off-roading group stated that their ability to receive the value they selected from that location would not change as a result of recreational facility development. Of all the substitution options, respondents from both groups were least likely to substitute a different value in the same location or a different location. Respondents were more likely to substitute the timing of their visits, how often they visit, or the location they visit to receive that benefit. However, 10.49 percent of respondents from the email-list group and 6.67 percent from the off-roading group stated they would stop receiving benefits from their selected value altogether due to recreational facility development.

While 48.96 percent of respondents from the email-list group and 56.70 percent of respondents from the off-roading group stated they would not be willing to substitute locations, the most important attribute reported from both groups to consider when substituting locations was stated to be a location with a similar setting, whereas the least important for both groups was that the substitute location have similar costs. The actual text for this question depended on the answer to the previous question (refer to table 5). Respondents were fairly split on how large of an effect the hypothetical scenario would have on their experience.

Limitations

Important to note from these findings is the bias of our sample. Due to the sampling method (which involved using a list of interested parties to collect data), all respondents were highly interested in the Mojave Trails National Monument in some capacity, which likely influenced the results. Collecting data using an interested-party email list is a low-cost, time-saving method, but will likely provide a sample that is biased if the intent is to capture the general public’s values. However, there may be instances where public land and water managers are more interested in the uses and values of people who are particularly interested in the area, in which case an interested party email list would be appropriate and the target audience. The survey link was also shared with an off-roading group on social media, which biased the sample to have more representation from people who off-road and people who use the particular social media outlet. In any case, we advise land managers who use this tool in the future to think about whose opinions and input are of interest and who may be excluded. For example, if data are collected exclusively using an internet survey, people without reliable internet access may be unintentionally excluded.

The demographics of participants in this study also show that respondents were considerably different from the U.S. population in general, and likely different from typical visitors to the monument. For example, over 80 percent of respondents were 50 years old or older (with less than 10 percent of respondents younger than 40 years old), and about 65 percent were men. Although we do not have data on the demographic profile of all visitors to Mojave Trails National Monument, it is unlikely that this sample is representative of all visitors to the monument. Respondents from the off-roading group were even more highly biased in terms of demographics and their interests, given the off-roading group was centered on a specific recreational activity. All results should be interpreted keeping this bias in mind and recognizing that these findings are not representative of all Mojave Trails National Monument visitors. Prior studies in Canada, Australia, and on national forests in Arizona and Oregon have found that participants in PPGIS studies tend to be disproportionately male and highly educated, so this bias is not unique to this study (Beverly and others, 2008; Brown and Reed, 2009; Raymond and Brown, 2011; Brown and Fagerholm, 2015). Prior research has also found that the types of values mapped are influenced by participant demographics (including gender and level of formal education), indicating that having a representative sample is important (Brown and Reed, 2009). Additional research could help clarify why certain demographics may be more likely to respond to PPGIS surveys and how to mitigate these respondent biases.

Lessons Learned from the First Pilot and Future Considerations

Given that this was the first use of the VaMPIRE tool, we aimed to evaluate challenges, considerations, and possible areas for improvement for future use of the tool. Regarding technical use of the online application, we only received one email indicating confusion with how to use the online application, with the participant noting that they could not figure out how to zoom in on the map (and there was another respondent who put in the open-ended comment box that they also could not figure out how to zoom the map). At the time of this pilot, we did not have a plus or minus button on the application for zooming in or out on the map; participants had to zoom using the scroll wheel on a mouse (or fingers on a touch screen). After receiving this feedback, we added a plus and minus button to the online map to facilitate ease of zooming in and out for future applications. We also received feedback about technical difficulties with being allowed to place only 30 points, and not knowing when 30 points had been reached. This was a software limitation, and for future uses of VaMPIRE we hope to increase this point limit so these challenges are mitigated.

Additionally, a couple of respondents wrote in the open-ended comment box at the end that some places were hard to identify and that it would be useful to include more place names. Future iterations of VaMPIRE may benefit from including key locations within the area that are shown on the map even when it is completely zoomed out (showing the entire monument or area of interest). We also identified one area where the questions could be improved. When analyzing results, it became apparent that the question on how the quality of their experience would be affected by the hypothetical scenario could be perceived as either positive or negative, but we did not ask if the effect would be positive or negative. Future uses of VaMPIRE would benefit from adding a question asking participants if the scenario would have a positive, negative, or neutral effect on their uses and values for the landscape (or updating the response options to clarify).

The VaMPIRE tool was designed to be used in any BLM-managed public lands and resources. Therefore, this pilot survey was not tailored to the specific context of Mojave Trails National Monument, which had the potential to cause confusion. For example, someone choosing recreation as one of their top three values would then see a question about the types of recreation they engaged in at that location. However, the provided list of recreation activities contains 24 options that, although applicable at a national scale, may not be as applicable at the local scale and could be missing certain activities that are exceptionally popular at the survey location. In the case of Mojave Trails National Monument, we did not include rockhounding in the list despite it being very popular at this specific location. There was an “other” option for participants to write in activities, yet we did receive feedback (both through email and through the open-ended comment box) indicating that because rockhounding is such a prevalent activity, not having it appear in the list was a source of confusion or disappointment for some respondents. In the future, it may be useful to tailor activity lists to fit each individual location if possible.

This pilot also demonstrated how easy and likely it is that the survey link could be shared with others outside of the original intended sample. In some cases, this may be beneficial because it can expand the pool of participants to new groups, but in other cases it may not be warranted or may bias the sample. Specifically, it is undesirable for people to complete the survey if their data will not be used or would bias the sample (this would unnecessarily increase the burden on the public). For future iterations, it may be important to consider only allowing one submission per unique link to limit the survey distribution beyond the intended participants. Additionally, there are limits on the total number of participants allowed for these surveys based on the burden estimates submitted to the OMB for approval.

Lastly, both follow-up emails for those who had not yet responded to the survey increased response rates substantially, so it is important that future studies also use multiple email contacts to increase the response rate (as recommended by Dillman and others, 2014). This is one potential issue with allowing the link to be shared outside of the target sample—there is not a way to send follow-up reminders to potential participants (for example, people who may have been forwarded the link or contacted through social media).