Nature–Human Relational Models in a Riverine Social–Ecological System: San Marcos River, TX, USA

: A social–ecological system is a highly connected organization of biophysical and social actors that interact across multiple scales, share resources, and adapt to the actors’ changes. The ways in which humans and nature interact have traditionally been characterized and inﬂuenced by competing intrinsic and utilitarian values. However, recently, relational values and relational models have been used to unpack the myriad of values society assigns to nature and create general typologies of nature–human relationships. Here, we investigate the spectrum of environmental values that exist in the San Marcos River (SMR)—a social–ecological system (SES) in which a spring-fed river ﬂows through an urban environment in central Texas (USA) including a university campus that attracts regional and international tourists. Recognizing that scholars have struggled to identify a nuanced understanding of environmental values and how these values shape nature–human relationships in SES, we use the SMR case study to capture the nature–human relational models that exist among social and user groups of the blue space. Analyzing different groups of visitors and stakeholders of the SMR (n = 3145), this study serves as a pilot to apply relational models using a variety of metrics to build a framework for understanding models of nature–human relationships, beyond ecosystem services and dualistic valuations. In our sample, most respondents were classiﬁed under the stewardship model (59%). The utilization model (34%) was the second most common, followed by wardship (6%). We found that patterns of place identity emerged to support the development of relational models beyond utilization. Despite the differences among perceptions, values, and some variation in relational models, one commonality was the innate, ubiquitous preference to protect natural habitat, water quality, and the river’s aquifer water source. Our study contributes to the growing literature around relational values and is a pathway to integrate ecosystem services, environmental values, and human–environment interactions into a more holistic approach to environmental valuation.


Introduction
Conservation biologists, practitioners, economists, and the environmental ethics community have long been divided over the appropriate ways to value nature [1,2]. Over the past two decades, the ecosystem services (ES) framework has been broadly used to quantify the utilitarian benefits nature provides to people and the ways those benefits are impacted by human actors [3,4]. The ES framework, as presented in the Millennium Ecosystems Assessment [5], made direct connections between healthy ecosystems and healthy humans, such as air and water quality [6]. Subsequent works expanded on the role ecosystem services-specifically knowledge systems, inspiration, cultural heritage, sense of place, and social relations-play in fostering a sense of responsibility and connectedness to nature [7,8]. Quantifying ecosystem services can lead to the prioritization of monetized, utilitarian values [3,7,9], which fails to capture the complex nonmonetary, intrinsic, cultural, The headwaters of the SMR are in the subhumid subtropical savanna-like Edwards Plateau ecoregion of central Texas, better known as Texas' scenic Hill Country. The City of San Marcos is located in Hays County between two major metropolitan areas, Austin and San Antonio, two of the nation's top 10 cities in terms of population [35]. San Marcos was named the nation's fastest-growing city from 2012 to 2014 [36]. From 2010 to 2019, the city's population increased by about 50% [35]. From 2010 to 2020, Hays County was the fastest-growing county (100,000 minimum population) in the nation by percentage [36]. San Marcos is also a major tourist destination, with over 14 million people per year as of 2018, and home to Texas State University, which has a student enrollment of over 38,000 (2020). The recent population boom combined with intense droughts has led to water shortages and restrictions, and water demand is expected to exceed local supply in the near future [37].
In addition to a rapidly growing population, San Marcos is also home to diverse and valuable river ecosystems. The San Marcos River is spring-fed from the karst Edwards Aquifer, a highly protected aquifer because of the presence of seven endangered species The headwaters of the SMR are in the subhumid subtropical savanna-like Edwards Plateau ecoregion of central Texas, better known as Texas' scenic Hill Country. The City of San Marcos is located in Hays County between two major metropolitan areas, Austin and San Antonio, two of the nation's top 10 cities in terms of population [35]. San Marcos was named the nation's fastest-growing city from 2012 to 2014 [36]. From 2010 to 2019, the city's population increased by about 50% [35]. From 2010 to 2020, Hays County was the fastest-growing county (100,000 minimum population) in the nation by percentage [36]. San Marcos is also a major tourist destination, with over 14 million people per year as of 2018, and home to Texas State University, which has a student enrollment of over 38,000 (2020). The recent population boom combined with intense droughts has led to water shortages and restrictions, and water demand is expected to exceed local supply in the near future [37].
In addition to a rapidly growing population, San Marcos is also home to diverse and valuable river ecosystems. The San Marcos River is spring-fed from the karst Edwards Aquifer, a highly protected aquifer because of the presence of seven endangered species [38]. Water emerges through the San Marcos Spring at the impounded Spring Lake and continues to form the San Marcos River with several park access points throughout the university and city (Figure 1). In recorded history, the springs have never run dry, and evidence shows the watershed to be one of the longest inhabited places in North America [24]. The ecosystem boasts excellent water quality, pristine habitats, high biodiversity, and the presence of endangered species that ensure the ecosystem's legal protection through a Habitat Conservation Plan [23]. The demand for biological and hydrological services is well-documented at the river, but social demand is less so.
The SMR is a social-ecological system (SES), and as such, "the delineation between social and natural systems is artificial and arbitrary" [39,40]. Because of the multitude of benefits provided by the river, there are often competitions between the functions of the ecosystem and the overwhelming social demand for the recreational and aesthetic opportunities the river provides. Within this SES, it is important to consider the impact of social demand for ecosystem services on those benefits and the biodiversity supported by the river through a more complex method of nature-human relational models.

Data and Methods
To measure social demand of the SMR SES and ultimately construct nature-human relational models, we conducted a questionnaire survey (Appendix A) in spring/summer 2015 across the city of San Marcos, Texas, USA. No risks were anticipated as a result of participation in this study (IRB EXP2015Y951777). All subjects gave their informed consent for inclusion before they participated in the survey. Our survey collected mostly closed responses; however, there were many additional comments and explanations of responses. As detailed in Table 1, the survey focused on five measurements: a ranked measurement of ecosystem services; statements that reflected utilitarian, intrinsic, or relational values; a measurement of the perception of the river (if the river is well-managed and protected, if the river is clean, and if the river is sensitive to urban growth); and behavior (activities) and environmental management priorities (Table 1). Table 1. Survey questionnaire and categories of analysis and associated grammars.

Ecosystem Services
Rank the importance of water in the San Marcos River for the following: •

Survey Questions Grammar (s)
The San Marcos River and its environment are sensitive to rapid urban growth.

Preferences
Usually the water in the San Marcos River is clean and clear. If the river became dirty or cloudy, would you still use and enjoy it the way you do now?
• Still enjoy the river, but less than I do now.

•
The cleanliness and clarity of the river has no effect on how much I use or enjoy the river.

•
If the river became dirty or cloudy, it would greatly reduce my ability to use or enjoy the river. • I would avoid the river if it was a dirty or cloudy river. • I do not currently use or enjoy the river.
• Goal orientation What is your personal preference of the amount of people in the river and parks when you visit?
• Other questions on the survey asked for sociodemographic information to help categorize participants into user groups and social groups. Social groups were categorized based on sociodemographic information that was independent of their use of the river. User groups were categorized based on frequency of use and residential status, both of which can influence or be determined by relationships with use of the river [22].
We used descriptive statistics and a series of nonparametric statistical tests (chisquared, Kruskal-Wallis, Mann-Whitney-depending on the number of groups-and Dunn's multiple comparison) to understand if significant differences existed among the user and social groups and the ranking of ecosystems services, environmental values, perceptions, and preferences. All descriptive statistics and nonparametric tests were conducted using JMPro. Cronbach's α was 0.7218 for the entire set of responses.

Framing Conventional Metrics to Relational Models
For the present study, we use conventional metrics to then follow Muradian and Pascual's (2018) argument for the adoption of discrete relational models, each characterized by an interaction of social conventions that provide a holistic model of nature-human relationships. We propose two ways to include and expand on the concept of relational values toward relational models. The first is the adoption of relational models, each with its own "grammar" of associated goals, emotions, and perceptions. The second is to investigate cognitive frameworks that shape a given relationship with nature [7]. This second approach requires integrating theories from cognitive and social psychology, particularly social representation theory, or how individuals assign meaning to the world based on the social constructs to which they are exposed. In brief, social representation theory is the content and production of common sense, or how individuals make sense of the world by adhering to social codes and constructs [41]. Because fully unpacking social representation theory is beyond the scope of this study, we aim to categorize users of the SMR into relational models based on patterns of shared values and preferences. Muradian and Pascual (2018) use seven parameters of reported "grammars" to distinguish between relational models: ontology including a clear society-nature distinction, whether nature is an entity with agency, how nature is positioned vis-à-vis humans, goal orientation, emotional drivers, practices, and main mode of interaction [7]. Out of these dimensions forms a typology of seven nature-human relational models. We propose that using this relational framework will uncover the ways in which blue spaces impact human well-being, as well as how relationships with blue spaces affect environmental management.
Though assigning relational models requires rich qualitative data, in this pilot study, we offer a means by which to transition from traditional metrics toward a relational model by attempting to identify or determine relational models through a survey of concepts related to ecosystem services and utilitarian, intrinsic, and relational values. In other words, our method uses traditional categories to uncover the "grammar" that defines relational models (Table 1).
Our study utilizes measures of ecosystem services and social demand (uses, perceptions, preferences, and environmental values) at the San Marcos River to categorize user groups and social groups and determine whether there are any significant relationships between groups and various perceptions, behaviors, values, or preferred services. Table 1 shows the survey questions related to understanding these grammars. The complete survey instrument is in Appendix A.
Using a subset of our survey questions and logical arguments, we classified our respondent sample into relational models (Table 2) based on the criteria of Muradian and Pascual (2018) [7]. Specifically, our method used a question hierarchy, which moved from detachment to wardship; our survey questions did not capture responses that could be classified as ritualized exchange or devotion. We first identified those with a detachment model and then sorted out respondents based on signals from key questions. For example, to examine what, if any, respondents may view "nature as inexistent" (detachment), we set a criterion of three salient survey questions with corresponding answers: (1) river provides benefits to fish and wildlife; corresponding answer as disagree or neutral, i.e., not agreeing that the river provides benefits; (2) river provides benefits to human well-being; corresponding answer as disagree or neutral; (3) importance of water; not environmental health (cultural use or human consumption). Here, the river (and therefore "water") is nature, and respondents, in this model, are unable to see the value or existence of nature, i.e., nature not important ( Table 2).
We continued with sets of three questions, some of which overlap and carry over to the next model, though using different corresponding answers based on the relational model, e.g., for the stewardship model, the corresponding answer for importance of water is environmental health, rather than cultural uses or human consumption. Finally, with wardship comes a preference for pristine conditions. As such, we selected those that identify habitat as the primary river benefit and prefer no people in the river. The question hierarchy was designed to be mutually exclusive where no one could be assigned to more than one relational model.

Social Actors and Social Demand
The questionnaire survey was completed by 3145 participants. The sampling method was mixed-modal: surveys along the river parks captured river users, an online version was distributed to the student population at Texas State University, and mail-out surveys captured rural residents who may not visit the river frequently. We also sampled at neighborhood markets, the city's activity center, and outlet malls for diversity/inclusion and to ensure our survey sample corresponded with the 2015 American Community Survey (ACS) demographic data. Overall, our survey participants were younger, had lower income, and were more educated than the census statistics for the San Marcos population (Table 3). This result is reflective of the disproportional number of college students that completed our survey. While this could be viewed as a limitation, students are indeed the largest subpopulation in San Marcos and an influential user of the San Marcos River (Julian et al., 2018). Another notable difference was the over-representation of women in our sample (+16 percentage points relative to the population), but this is a common pattern in questionnaire surveys [42]. Racial/ethnic diversity in our survey was similar to census statistics; however, we over-sampled the White-only population by 10 percentage points and under-sampled the Hispanic/Latino population by 12 percentage points.
Our survey sample included a spectrum of social actors. There were long-time residents (50+ years) who visit the river daily on one end of the spectrum, and international tourists who only visited the river this one time on the other end of the spectrum. We had 226 respondents who said they have never visited the San Marcos River, including 185 students, 38 tourists, and 3 non-student residents. However, the vast majority of our sample visited the river multiple times. Thirty-one percent of river users visited between one and five times per year, which we classified as low frequency for later statistical analyses. medium-frequency (6-15 visits per year) river users made up 26% of our sample. The remaining river users (43%) were classified as high frequency (>15 visits per year). We found that river use was partly explained by residency and student status (Table 4, Appendix B). In general, non-student residents were the most frequent visitors, with a mean of 72 annual visits. Median annual visitation for this group (30) was three times higher than university students and regional tourists. River usage among students did not vary with residency status; both resident and non-resident students visited the river weekly on average during school sessions. Approximately 70% of the university students visited the river in groups of three or more people, but it was non-regional tourists who visited the river in the largest groups-5+ for half of them. Some of these respondents came to the river on large tour buses. Compared to other user groups, non-student residents were twice as likely to visit the river alone.
In addition to quantifying river usage (Table 4), we also measured social demand by asking (in a free-response format) what activities they participated in when visiting the San Marcos River (Figure 2). The most common categories of activities were swimming, floating/tubing, relaxing/stress relief/meditation/sunbathing, socializing/community event/picnic/drinking, park exercise (walking, running, jogging), park sports/recreation, water sports/recreation, dog activities, work/school/research, wildlife/nature viewing, river clean-up/volunteering, and reading (Appendix C). Swimming was the most popular activity, being mentioned by 41% of participants who listed at least one activity. Most river users (66%) listed 2-3 activities and as many as 19 activities. While people enjoyed the river in multiple ways, there were different preferences among the user groups. Non-student residents used the river more for physical exercise, while the other user groups (students and tourists) participated more in leisure activities (Table 4).  Figure 2. Activities reported by river users, where size of word is proportional to frequency. Verbs with multiple tenses were changed to the same tense. Categorical data are reported in Appendix C. Word cloud was generated using Word it Out, a product of Enideo.

Perceptions and Preferences
Other measures of social demand for a social-ecological ecosystem are the perceptions and preferences of the social actors (Tables 5 and 6; Appendices B and D). Regarding perceptions, two-thirds (67%) of respondents agreed or strongly agreed that "the environmental health of the San Marcos River (SMR) is well managed and well protected", while 9% disagreed or strongly disagreed with this statement. The remaining 24% were neutral (neither agree nor disagree). Similarly, 9% of the respondents perceived the river as slightly dirty or extremely dirty, with the vast majority citing trash as the reason. Four out Figure 2. Activities reported by river users, where size of word is proportional to frequency. Verbs with multiple tenses were changed to the same tense. Categorical data are reported in Appendix C. Word cloud was generated using Word it Out, a product of Enideo.

Perceptions and Preferences
Other measures of social demand for a social-ecological ecosystem are the perceptions and preferences of the social actors (Tables 5 and 6; Appendices B and D). Regarding perceptions, two-thirds (67%) of respondents agreed or strongly agreed that "the environmental health of the San Marcos River (SMR) is well managed and well protected", while 9% disagreed or strongly disagreed with this statement. The remaining 24% were neutral (neither agree nor disagree). Similarly, 9% of the respondents perceived the river as slightly dirty or extremely dirty, with the vast majority citing trash as the reason. Four out of every five respondents (80%) perceived the river as mostly clean or very clean; the remaining 11% were neutral. Newer residents (less than 4 years) were more likely to perceive the SMR as well-managed, well-protected, and clean (Appendix D). Frequency of Visits to SMR Low frequency visitors had more positive perceptions Low frequency visitors had more positive perceptions Table 6. Preferences of river use by social actors. Level of significance: * (0.05), ** (0.01), *** (0.001). Another perception we measured was the sensitivity of the river to rapid urban growth. Only 13 respondents (less than 1%) disagreed with the statement "the San Marcos River and its environment are sensitive to rapid urban growth". The vast majority (87%) believed the river is sensitive to rapid urban growth; older users, those with higher incomes, and those with higher levels of education were more likely to have this perception (Table 5). To expand on the perception of sensitivity and, thus, social demand, we asked two preference questions.

Loss of Water Quality Crowding Preferences Interpretation
When asked "if the river became dirty or cloudy, would you still use and enjoy it the way you do now?" the most popular response for those who use the river (46%) was "it would greatly reduce my ability to use or enjoy the river" ( Figure 3A). Almost a quarter (24%) said they "would still enjoy the river, but less than I do now", and more than a fifth of the respondents (22%) said they "would avoid the river" if it became dirty or cloudy. These preferences did not vary among user groups, but those with higher income were more likely to avoid the river if water quality degraded (Table 6). In addition to wanting a clean and clear river, most respondents who use the river (58%) preferred only a few people in the river ( Figure 3B). Some preferred no people in the river (14%). A quarter (26%) said they preferred many people in the river and 2% preferred very many people in the river. Compared to water quality preference, there was much more variation in crowd preferences in the river among social and user groups ( Table 6). Those reporting longer-term residencies and higher education levels preferred fewer people in the river when visiting compared to other groups. White river users also preferred fewer people when compared to non-White river users. Regional tourists differed from resident and student user groups by preferring more of a crowd and enjoying the river the same regardless of water quality (Table 6).

Ecosystem Services and Environmental Values
Survey participants overwhelmingly (90%) agreed (or strongly agreed) that "the San Marcos River provides benefits to fish and wildlife." Even more (96%) agreed (or strongly agreed) that "the San Marcos River provides benefits to human well-being." With acknowledgment of the benefits provided by the river, we explored ecosystem services in the traditional categories of provisioning (food and water source), regulating (habitat and clean water), and cultural services (recreation and culture). Results showed that people generally assigned the highest importance to regulating services: 39% chose habitat as the most important benefit and 23% chose clean water first (Appendix E). Cultural services followed in importance: 20% chose culture first and 13% chose recreation as the most important benefit. Provisioning services ranked last: only 4% chose water source as most important and less than 1% chose food source. To assess environmental values, we asked participants to rank the relative importance of water and fish (separately) in the San Marcos River for human consumption (utilitarian, U), environmental health (intrinsic, I), and nonmaterial (cultural) uses (relational, R) (Appendix E). For water, most respondents (70%) prioritized the intrinsic value of environmental health, with utilitarian and relational values each ranked highest by 15% of the sample. Among user groups, non-student residents were the only one significantly different than the others, with only 6% prioritizing human consumption (U). Almost four out of every five non-student residents (79%) prioritized environmental health (I), ten percentage points higher than the other user groups. Fish in the San Marcos River were overwhelmingly prioritized for their intrinsic value (84%), followed by relational value (10%) and utilitarian value (6%). Non-regional tourists were significantly different than other user groups: 10% prioritized human consumption (U), 16% prioritized cultural uses (R), and 75% prioritized environmental health (I), the lowest among all user groups.
We then assessed environmental values across social and user groups using multiple divisions of intrinsic (I), utilitarian (U), and relational (R) values (Tables 7 and 8; Appendix E). Intrinsic values, such as habitat and clean water, were more prevalent among respondents with higher education levels and long-term residents of San Marcos. Respondents with lower education levels (e.g., less than a bachelor's degree) and tourists prioritized utilitarian values, such as food and water source. Relational values-spanning culture, inspiration, education, identity, and spirituality-varied across residential status and years of residency. Newer residents, student residents, and tourists placed higher value on culture and recreation, whereas long-term residents (non-students) reported stronger relational values of spirituality and identity.
Overall, the user group of residential status exhibited more differences across environmental values than the user group frequency. For example, residential status affected how respondents were likely to value variables in each of the three environmental values (Table 8). Users who were residents (non-students) were more likely to value clean water (I) and spirituality (R), while more transient visitors such as tourists and students valued recreation (R), culture (R), and water source (I). Frequency groups showcased differences only in relational values with those who visit less frequently valuing recreation (R) and aesthetics (R) higher on average ( Table 8).
The final method we used to measure environmental values was asking participants to distribute money (percentage-wise, totaling 100%) from a hypothetical annual fund dedicated to improving the San Marcos River. The funding distributions were relatively consistent across all user groups ( Figure 4), with water quality protection, water quantity protection, and habitat protection being the top three funding priorities. Riverfront development for housing, dining, and shopping was by far the least important-less than 5% for all user groups. However, there were some notable differences among user groups. Students (both resident and non-resident) placed greater importance on habitat protection for fish and wildlife. Non-student residents placed relatively high importance on increasing riverfront park acreage. Regional tourists placed the greatest importance on protection of water quality and water quantity, almost equally.

Relational Models
Using a hierarchical subset of our survey questions and logical arguments (Table 2), we classified each participant into a relational model that ranged from detachment (nature as inexistent) to wardship (nature to be protected above all else). Only 1% of our surveyed sample fell into the detachment category, and only seven participants (<1%) were labeled as domination (human control over nature). Almost a third of our sample (34%) fell into the utilization relational model, where their priority was to benefit from nature rather than protect it. The most common relational model was stewardship (59%), where participants prioritized environmental health over human benefits. Finally, 6% of our sample were labeled as wards with a preference for a healthy ecosystem absent of people.
Once the respondents were sorted into relational models, we then looked at how the different user groups fell into the five relational models (Table 9). Detachment, although uncommon overall, was highest among regional tourists (4%). Domination was rare, only representing seven respondents overall, most of whom were resident students. Utilization was most common with regional students (37%) and nonregional tourists (36%), followed by resident students (34%) and regional tourists (31%). Twenty-seven percent of residents (non-students) were categorized as having a utilization relational model with the SMR, the lowest among all groups. Stewardship was the predominant relational model for all user groups, but highest for non-student residents (67%). Wardship models comprised a small but significant percentage of student and resident respondents (5-6%), but only accounted for 1% of tourist user groups.

Intrinsic
Habitat for plants and animals Clean Water Clean and reliable water from the aquifer groundwater Resident (Non-Student) valued higher than Student Resident and Regional Student Utilitarian Food A source of fish for your meals Least valued Regional Tourist valued higher than Regional Student and Resident (Student) Water Source A source for municipal, industrial, or agricultural water uses Nonregional Tourist valued higher than Student Resident, Regional Student, and Resident (non-student) Regional Tourist valued higher than Student Resident and Regional Student Relational Culture A place for relaxing or enjoying the scenery and local culture Resident (non-student) valued lower than Student Residents and Regional Students Residents (students) valued higher than Nonregional Tourist and Regional Tourist Regional Students valued higher than Nonregional Tourist and Regional Tourist

Physical activities in nature
Resident (Non-Student) valued lower than Resident (Student), Regional Student, Regional Tourist, and Nonregional Tourist Visitors in the "Low" group valued more than visitors in the "High" group Inspiration Artistic, cultural, or work-related activities Aesthetics Relaxation, scenery, or sentimental value Visitors in the "Low" group valued more than visitors in the "Medium" and "High" groups Education Opportunity to experience, learn about, or appreciate nature Identity Cultural heritage, local pride, sense of place, symbol of San Marcos Regional Tourist valued higher than Student Resident; Resident (Non-student), and Regional Student Spirituality Sacred, religious, or mental health activities Resident (non-student) valued higher than Student Resident and Regional Student

Relational Models
Using a hierarchical subset of our survey questions and logical arguments (Table 2), we classified each participant into a relational model that ranged from detachment (nature as inexistent) to wardship (nature to be protected above all else). Only 1% of our surveyed sample fell into the detachment category, and only seven participants (<1%) were labeled as domination (human control over nature). Almost a third of our sample (34%) fell into the utilization relational model, where their priority was to benefit from nature rather than protect it. The most common relational model was stewardship (59%), where participants prioritized environmental health over human benefits. Finally, 6% of our sample were labeled as wards with a preference for a healthy ecosystem absent of people.
Once the respondents were sorted into relational models, we then looked at how the different user groups fell into the five relational models (Table 9). Detachment, although uncommon overall, was highest among regional tourists (4%). Domination was rare, only representing seven respondents overall, most of whom were resident students. Utilization was most common with regional students (37%) and nonregional tourists (36%), followed by resident students (34%) and regional tourists (31%). Twenty-seven percent of residents (non-students) were categorized as having a utilization relational model with the SMR, the lowest among all groups. Stewardship was the predominant relational model for all user groups, but highest for non-student residents (67%). Wardship models comprised a small but significant percentage of student and resident respondents (5-6%), but only accounted for 1% of tourist user groups. Using open-ended voluntary comments provided by respondents, we found supportive evidence for different relational models (Table 10). For example, to showcase "nature as nonexistent", a student resident suggested that the river "needs more publicity", which can be interpreted as more people should know about it and use it. The SMR is a tourist destination and was cited as being overcrowded in other user comments. In another example, those categorized as having a utilitarian relationship provided comments such as "more parking in general" and "more consideration for the humans who recreate" (Table 10). Stewardship model respondents stated that the SMR should be "treasured and preserved" and that it is "best to honor and protect it." Wardship respondents would like "more strict actions of people within the river..." and show a willingness to "pay more or limit [...] recreational use for its sake" (Table 10).
Another finding from the open-ended comments is that there is overlap in social demand and environmental values between some of the relational models, particularly between stewardship and wardship. We expand on this overlap in the Discussion. One river feature that occurred frequently in the comments was Texas wild rice (Zizania texana), the rare and endangered plant species (endemic to our study area) that is protected and influences management of the river. This aquatic species-referred to as rice, grass, plant, and weed-was mentioned at least 50 times in the open-ended comments, sometimes in a positive context and sometimes in a negative context. We explore this controversial river feature further in the Discussion as well. Table 10. Example comments from survey respondents (user group in parentheses) paired with their relational model.

Detachment
Is urban growth sensitive to the San Marcos River? (Resident, Student) Have they ever thought of lifeguards? We would love it. (Tourist, Regional) It needs more publicity. I suggest there has to be policies that increment strict habits to protect San Marcos river. (Student, Regional) Use as much money and man power it takes to keep the river clean and habitable for the rare species of wildlife that live in the river. (Resident, Student) I would be willing to pay more or limit my recreational use of the river for its sake, and I think that through education other people may begin to agree with that. (Resident, Non-student) I feel the river needs to be more protected. As far as the ecosystem and all the animals that use it as a resource. (Student, Regional)

Relational Models Built from Social Demand
The San Marcos River is a social-ecological system (SES) where different social actors (non-student residents, student residents, regional/commuter students, regional tourists, and extra-regional tourists) interact with one another and a natural environment that provides many benefits to society. Using this riverine SES as a natural experiment, we first assessed social demand-the uses, behaviors, preferences, perceptions, and values of society (sensu [32,33]). We found that the San Marcos River is heavily used (Table 3) for a multitude of activities ( Figure 2, Appendix C). It is also a highly valued resource, particularly for its clean water, natural habitat, and tranquil milieu (Appendices C and D). Indeed, most river users preferred only a few people (or no people) when they visit and said that their use/enjoyment of the river would be greatly reduced (or they would avoid the river altogether) if the water became dirty or cloudy ( Figure 3). The vast majority of survey participants prioritized environmental health over human use of the river. These findings accord with other SES studies from around the world [15,[43][44][45][46].
Our study was also analogous to these international SES studies in that different social actors interacted with their environment in different modes (Table 3). In general, non-student residents visited the river weekly with a relatively small group and used the river parks for swimming, exercising, and socializing. Students visited the river much less (monthly or semi-monthly) and also used the river parks primarily for swimming, but floating (in an inner tube) and relaxing were their second and third most popular activities, respectively. Tourists mostly used the river for floating (often with an alcoholic beverage cooler) a few times a year depending on how close they live to the river. While students and tourists used the river for alike activities and with similar frequency, a notable difference between the two social actors was that students placed significantly higher value on the cultural aspects of the river (Table 8; Appendix E). The overall sum of social demand for the San Marcos River showed that people have preference for an ecologically healthy river that is swimmable and clean enough to float. A decline in any of those attributes could result in negative (and possibly unsustainable) outcomes [40,46].
We used the ecosystem services framework to collect some of our social demand data because of its convenience and common language, but as recent scholars have pointed out, this framework fails to capture the complex ways humans and nature interact, particularly the cultural and moral dimensions (i.e., relational values) [3,7,12]. Thus, we expanded our SES analysis by exploring the relational models outlined in Muradian and Pascual (2018) [7]. We identified three prominent relational models within the San Marcos River SES: stewardship, utilization, and wardship (Table 10; Figure 5). Those that reflect the stewardship model showed a respect for nature by prioritizing environmental health over human use. Stewards also acknowledged the role of functioning aquatic ecosystems, showing knowledge and respect for the maintenance of SES. This system mindset and respect demonstrate a mostly relational value [13]. Those who fall into the utilization model agreed that water has benefits to human well-being but did not prioritize environmental health. They value nature's agency, but not its rights; its benefit is as a service to society, i.e., its primary value is utilitarian. Wards demonstrate mostly intrinsic values. They prioritize environmental health above all else; they view the river as habitat and prefer no people in that habitat. These perceptions and preferences reveal adherence to the relational model of wardship, which views nature as an entity with rights that should be protected.
Geographies 2023, 3, FOR PEER REVIEW 20 prefer no people in that habitat. These perceptions and preferences reveal adherence to the relational model of wardship, which views nature as an entity with rights that should be protected. Our relational model's methodology placed individuals into specific categories (Table 10; Figure 5); however, value expressions often overlapped. For example, a resident student classified as a steward stated: "[the] SM river is an amazing place that should be preserved, and if that means that I (and others) should not use it, I wouldn't." This comment is perhaps more reflective of a ward, who believes nature needs to be protected, in this case, from humans. In the same vein, a student resident also classified as a steward posed the question: "It's a great river and homes a lot of living creatures, why do [we] as humans view that we are more important than them?" In this example, humans should Our relational model's methodology placed individuals into specific categories (Table 10; Figure 5); however, value expressions often overlapped. For example, a resident student classified as a steward stated: "[the] SM river is an amazing place that should be preserved, and if that means that I (and others) should not use it, I wouldn't." This comment is perhaps more reflective of a ward, who believes nature needs to be protected, in this case, from humans. In the same vein, a student resident also classified as a steward posed the question: "It's a great river and homes a lot of living creatures, why do [we] as humans view that we are more important than them?" In this example, humans should not interfere with nature, i.e., nature as a separate entity with distinct rights (wardship model). Finally, the stewardship model boundary also blurred with the utilitarian model boundary. For example, a student resident said, "I think there should be more places to drink at the river and more trash bins along the river so people can throw away their trash with more ease," which leans toward using the river as a place to consume alcohol while having the amenity of trash receptacles. This blurring of boundaries, or value heterogeneity, is quite common in social-ecological systems with valuable natural resources [46].
The San Marcos River is teeming with valuable natural resources: clear, flowing water rich in biodiversity and surrounded by scenic, amenity-rich parks. This natural wonderland is extensively managed and largely protected for the endangered species that rely on a healthy ecosystem [23]. One of these protected endangered and endemic species is Texas wild rice (Zizania texana). From the more than 50 comments that specifically mentioned this protected species, we found both shared (Table 10) and conflicting values. Respondents classified as stewards stated the following regarding the management of the endangered species: "I think it's pointless to keep weeds in the water. All to save some rice?? Sorry but I put humans needs over rice...." (Resident, student; punctuation original) "Don't restrict access because of wild rice." (Resident, non-student) "The rice grass is growing too much it makes the river look really gross and tacky." (Resident, non-student) These attitudes may be attributed to differences in social demand, and the dislike of the wild rice may simply reflect a more utilization relational model. To be sure, respondents who were sorted into a utilization model felt similar, as supported in the following example comments: "Sometimes we spend too much time and money protecting the endangered species in preference to making it a nicer recreational area. Cut the weeds, they grow back and I think we pay far too much time and money to protect the wild rice. It grows back." (Resident, student) "Stop making barricades in the river to save the over abundance of wild rice grass." (Resident, student) "Trim some of the grass so it's not sticking out of the water." (Resident, student) One difference between the two sets of comments is that the stewardship model included non-student residents, while those who held negative perceptions of Texas wild rice and fell into a utilization model were all students. This result could be a consequence of a lack of education about the importance of the species and its proper management, i.e., "cutting the grass". Nonetheless, in this pilot approach to classifying users of a blue space into relational models, we find that-similar to relational values-there are patterns, yet the edges and transitions are blurry.
Relational values are dependent on the desire for harmony between social and ecological actions, meaning intrinsic values and utilitarian values are both important but are interwoven to promote a relational value with place [13]. If someone enjoys time spent in blue spaces, is it because they receive a direct utilitarian benefit (stress relief) or because of their relationship with the place? The individual making the value attribution may be unable to articulate a distinction. It is important to acknowledge that bundles of services and intertwined values create a more holistic picture of individuals' experiences [33]. Indeed, there are differences between the ways in which practitioners, academics, and the public conceptualize ecosystem services [11]. For example, Lehnen and colleagues have used the IPBES and nature's contributions to people framing to "zoom into connections between people and nature" with their "individual relationships with entities of nature framework", which acknowledges issues of justice and equity when assessing environmental values [47]. That is, their multi-dimensional examination of nature-human relationships helps "detect inequalities in the benefits and detriments individual people receive" ( [47] p. 596).

Our Relationship with Nature Is a Shared Value
Despite differences in social demand among groups, several key values and preferences were shared. Across social groups, intrinsic values (habitat) and relational values (aesthetics and sense of place) were highly regarded. Additionally, inspiration, education, and culture were similarly valued among groups. In this way, relational values blur the lines between utilitarian and intrinsic values because they prioritize healthy ecosystems to promote deeper connections. Further blurring the lines of relational values is the fact that across groups, most of the relational values we measured held few differences-and the differences that did exist were mostly from long-term residents having deeper levels of place meaning and connection. These shared valuations and preferences for ecosystem services show a general societal preference for the maintenance of ecosystems for relational values over utilitarian and intrinsic values [13,15,22].
An important finding from our research is that social actors have idealized expectations of their social-ecological system. The majority of social actors (across all user groups) prefer clear water and would "greatly reduce" their use and enjoyment of the river if it became dirty or cloudy. More than a fifth (22%) of the respondents would avoid the river altogether if it was in this degraded condition. This general demand for clean water has been shown for a spectrum of blue spaces around the world [48][49][50][51][52][53][54]. These studies, along with our study, have shown that people are willing to travel far and incur higher costs to enjoy clean lakes and rivers. Indeed, the visual characteristics of an environment, e.g., water clarity, are often used by the public to gauge environmental quality [55]. While the previous studies attribute this increased demand for aesthetics and recreational preferences, our multi-dimensional data reveal deeper connections-connections between environmental health and our overall well-being. While the interpretation of "aesthetics" is highly variable and dependent on personal experiences, as well as frames of beauty [56], we can reasonably assume that respondents did not just prioritize clean water to meet their subjective view of aesthetics but prefer clean water because it supports ecosystem functioning as 70% of the social actors prioritize environmental health over any human uses supports this stewardship relational model. Furthermore, social actors want pristine conditions with few or no people in the river; yet, they use the river a lot and for many activities. Non-student residents best exemplify this hypocrisy on account that 75% of them wanted either a few or no people in the river; yet, they were the most frequent users of the river (Table 4; Figure 3). Regional students had similar expectations but were the most likely to avoid the river if it became dirty or cloudy (26% of them). We attribute this "delight or flight" attitude to their greater mobility. These regional students come to the San Marcos River specifically for its natural benefits, but if these benefits are degraded, it is easy for them as non-residents to visit a more desirable blue space. Residents are less mobile and will continue to use their river even if it is degraded, although their enjoyment will be less. Because residents are more place-invested, they are more likely to be stewards and wards of their precious resource.
In addition to place-investment, one possible explanation for residents reflecting behaviors of stewardship-as shown in other studies-is that the activities themselves allow the "relation" to form [54,57,58]. According to Tuan's theory of "body ballet", it is the habitual movements in and across a space that create place [59]. In previous studies, users of blue places with higher levels of place dimensions (sense of place, place dependence, and place identity) were found to have a stronger response to the place, i.e., "a willingness to take action to preserve" ( [59], p. 675) [46]. We found evidence of the desire to protect a place-based, communal asset and identity-expressive place meanings [60,61] in some of the additional, optional comments from respondents (Table 11). Table 11. Example comments reflective of place-based meanings.

Place-Based Asset Identity-Expressive
"The river is a holy place, a blessing to mankind."-Resident, non-student "[I] see myself as a keeper of the river."-Resident, non-student "San Marcos residents and government need to realize the importance of the river in our town, as well as how to protect it, to the fullest."-Resident, non-student "I've noticed in the past 43 years that as people become acquainted with our river, they develop an intimate desire to call it their own. Perhaps for its beauty and clarity, or ability to change your overall mood during the hot summers, but there is certainly something special about this river for each of us."-Resident, non-student "Our river truly is a gem, and an important resource to enrich the environment, and the lives of many. It needs to be protected."-Resident, student "SM river is an amazing place that should be preserved, and if that means that I (and others) should not use it, I wouldn't."-Resident, student We see this phenomenon reflected here through the additional comments left by residents, including student residents, possibly showcasing identity-expressive meanings [61] with the SMR. We explore place identity further below.

Social Connectivity and Place Identity
The manifold social demand for the San Marcos riverscape reflects multiple scales and dimensions of connectivity [34,62,63]. The preference for readily available clean water (Figure 3, Appendix D) and the high value placed on protecting the aquifer, water quality, and fish/wildlife (Figure 4), combined with the consensus that the river is sensitive to rapid urban growth (Table 5), establishes the watershed-scale connections of the river. That is, most river users recognize that the health of their beloved river reach is affected by land use and management at the watershed-scale.
In addition to being a biodiversity hotspot, the San Marcos River is a cultural hotspot [64]. Kondolf and Pinto suggest the social connectivity-or the "communication and movement of people, goods, ideas, and culture along and across rivers", namely through the recognition of "longitudinal lateral, and vertical connectivity" ( [62], p. 182)-plays a role in "river culture" wherein the "intersection of hydrologic, biological, and cultural uses and values of rivers as a basis for preserving ecological and cultural diversity along rivers" [57] ( [62], p. 182). Our study site, the upper San Marcos River, offers exceptional social connectivity. There is a high degree of lateral connectivity on account of the river being narrow enough (~20 m) to observe human activities on the other side of the river and shallow enough (mostly waist-deep) to cross the river and interact with both riverbanks. The multiple pedestrian and vehicle bridges enhance this lateral connectivity. Vertical and longitudinal connectivity are maintained through the various parks and public access points along the river, with the right side of the river being continuous in this regard. Additional longitudinal connectivity is provided by a~2 km sidewalk/trail system that connects the downstream extent of our study area all the way to the headwaters, with paths underneath the vehicle bridges.
Furthermore, the small river with excellent water quality and social connectivity enables activities in the stream, along the bank, and throughout riverside parks. The opportunity for various activities-swimming, tubing, snorkeling, kayaking, nature-watching, sunbathing, and music, just to name a few (Figure 2)-plays a vital role in creating and reinforcing relational values in the City of San Marcos and abroad. That is to say, diverse populations have the opportunity to form a relationship with the (clear flowing, not too crowded) blue space through their preferred method. It is this relationship that leads to "place identity"-a deep form of place attachment wherein the self develops in relation to the physical environment by means of preferences, beliefs, values, and goals [65]. This place identity fosters a stewardship ethic, especially in natural areas [66,67]. These stewards feel a responsibility to take care of their environment and, thus, visit it more frequently, intensifying relational values in the process. This positive feedback phenomenon most likely manifests in long-term residents, as length of time in residency is the strongest predictor of place identity [68]. Indeed, the residents subgroup in our study gravitated toward relational models of stewardship and wardship. Their relations with the SMR seem to be a part of their overall identity (as shown in the comments in Table 11), whereas more transient visitors such as tourists have not had the extended interaction needed to create deep place meanings.
In the middle of this spectrum-between long-term resident stewards and casual tourists-are students, some short-term residents and some commuters. While most of these students classified as stewards (with a heightened appreciation of the cultural benefits of the river), they were more likely to have stronger utilitarian values. Their utilitarian activities (e.g., tubing and sunbathing) are substitutable, meaning one can tube or sunbathe at a different blue space, one that best meets their social demands. Commuter students were particularly acute to this scenario, where more than a quarter of them said they would avoid the river if its water clarity did not meet their preference. Earlier, we related this "delight or flight" attitude to the commuter students' greater mobility (ease of finding a new blue space), but this attitude is also related to a lack of place identity [58].
The importance of place-identity and environmental stewardship goes well-beyond social connectivity and relational values. A persistent problem is how to meet the increasing social demands placed on the environment and maintain healthy ecosystems [69]. Recent scholars have suggested and shown that environmental stewardship (voluntary action on behalf of the environment [70]) can fill these multi-scale gaps in natural resource management [71,72]. Motivation for volunteerism/stewardship has been widely investigated [72][73][74][75][76], and researchers have found that participation in stewardship activities-from individual litter pick-up to organized group efforts such as water-quality monitoring-can catalyze deepening the human-environment relationship, thereby crafting a stewardship relational model that may transcend geographic boundaries [71]. Such a relational model could serve as a leverage point for guiding sustainable human-environment behavior [47], as broader society should seek to implement "stewardship at all levels to maintain and improve ecosystem services" [77].
Understanding and enacting sustainable human-environment behavior is a critical component of social-ecological systems (SES) research, as it is often "problem-oriented" by seeking to inform environmental management policy and practice [40]. SES governance is centered on knowledge from multiple actors, using socio-ecological relationships to make decisions that prompt system sustainability [78], thereby adapting management plans based in part on stakeholder knowledge. User surveys, such as the one in this study, are a useful mechanism to capture potential relational models and individual dimensions of SES, which lead to adaptive plans that reflect the local context [79].
Our study, situated in an SES framework, establishes the social demand (uses, preferences, perceptions, and values) of stakeholders. Relational values, coupled with multidimensional data and relational models, are a path toward incorporating social dynamics into environmental planning, rather than relying on market value preferences or preferences for ecosystem services, including willingness-to-pay studies [80,81]. Preferences for ubiquitous values of ecosystems (clean water and fewer people) can be leveraged into heuristics to guide decision-making in complex conditions with multiple stakeholder opinions (see [82] for an education tool). For example, our case study quantifies social dynamics within an SES to adjust local environmental planning. Since 2013, the San Marcos River has been managed under the Edwards Aquifer Habitat Conservation Plan (EAHCP) [23]. The EAHCP works to ensure suitable habitat for threatened and endangered species within the Edwards Aquifer system, which includes the San Marcos River. The HCP is implemented through a "stakeholder driven process," which can be informed by this study [23]. Environmental education, recreation, and user behavior change are crucial to the EAHCP's success, which balances the sensitive habitat with social demand.
Limitations include applying relational models to a single social-ecological system, i.e., in a limited context and geographical scope. Had the survey been administered to various users of multiple SES, the results may be different. In addition, we did not include ecosystem disservices, or the harm nature can cause humans, in our questionnaire and analysis [83]. Future research could use a questionnaire crafted to address all grammars of relational models, i.e., to include statements about domination, detachment, etc., as related to ecosystem disservices. Furthermore, our questionnaire and analysis did not address the multifaceted aspects of aesthetics as it relates to the overall experience of the SES. We acknowledge that aesthetics is examined in a variety of disciplines (philosophy, urban design, and environmental psychology to name a few) but that unpacking the breadth of the term was outside of the scope of our study. We suggest that future research works to ameliorate discrepancies between users' lived experiences and references points for aesthetics [56] and applies the framework of relational models to users of multiple and varied blue spaces.

Conclusions
This study is one of the largest survey samples to date of a social-ecological system (SES) wherein ecosystem services, environmental values, social demand, and nature-human relational models were assessed. Our SES is also noteworthy because it is one of the most intensively managed aquatic ecosystems in the fastest growing region of the United States. We collected a wealth of data on ecosystem services and social demand to gain a deeper understanding of this exceptional SES, and then applied the framework of Muradian and Pascual (2018) to develop empirical relational models [7]. These empirical models provide a real-world assessment (beyond just theory) of how social actors engage with and value a blue space. Broader implications of applying relational models suggest that stewardship and wardship models could be leveraged to encourage policies and practices encouraging environmental stewardship. An environmental stewardship ethic, once introduced, is likely to create a positive feedback loop where frequent relational experiences in nature lead to human behaviors that are aware, attached, informed, engaged, and nurturing.
Looking forward, we found that relational values as a third category (beyond intrinsic and utilitarian) does offer additional insights; however, the nuances found (such as the blurring of boundaries and plural relational values) are a signal that three categories may be insufficient. Multiple relational models and system archetypes may offer multi-dimensional solutions, but nature-human relationships that transcend epistemological boundaries are on the horizon.

Acknowledgments:
We are grateful to Graham Daly and Emily McBroom for helping administer the survey. Graham Daly also helped design the survey and with initial data organization. The City of San Marcos and the San Marcos Convention and Visitor Bureau, respectively, provided access to facilities and resources to assist in the commissioning of the survey.

Conflicts of Interest:
The authors declare no conflict of interest. The funder had no role in the design of the study, in the collection, analyses, or interpretation of data, in the writing of the manuscript, and in the decision to publish the results.

TEXAS STATE UNIVERSITY: Department of Geography Consent Form:
This survey is designed to measure how people use and value the San Marcos River. Even if you do not visit the San Marcos River, your input is still valuable for a better understanding of the relationship between people and the river. This survey should take about 15 min to complete. All survey responses are confidential and remain anonymous and can in no way be linked to your identity. Your participation is voluntary and very much appreciated, but you may opt out at any moment. However, it is preferred that you answer all of the questions so that the data we collect will be accurate and representative of the people surveyed. It is our intention to conduct a thorough study of the relationship between people and the San Marcos River to inform a better understanding of this important relationship.
This ___ Non-material human uses: associated with recreation, aesthetics, education, inspiration, spirituality, and identity. ___ Human consumption: use associated with municipal, agriculture, and industry water supply. (22) Rank the importance of fish in the San Marcos River.
[1 being the most important and 3 being the least important.] ___ Environmental health: one part of the ecosystem and food web which also includes birds, mammals, insects, and plants. ___ Non-material human uses: recreational fishing, aesthetics of viewing, and education of the environment. ___ Human consumption: a food high in protein, low in fat, and a source of fatty acids. (23) The environmental health of the San Marcos River is well-managed and well-protected. ___ Strongly agree ___ Agree ___ Neither agree nor disagree ___ Disagree ___ Strongly disagree (24) If you were in charge of an annual fund dedicated to improvement projects for the San Marcos River, how would you distribute the money? [The 100% is representative of all of the money in the fund. Total must add up to 100%, whole numbers only.] [If you chose Slightly dirty or Extremely dirty.] (28) Usually the water in the San Marcos River is clean and clear. If the river became dirty or cloudy, would you still use and enjoy it the way you do now? ___ I would continue to use or enjoy the river the way I do now. ___ I would still enjoy the river, but less than I do now. ___ The cleanliness and clarity of the river has no effect on how much I use or enjoy the river. ___ If the river became dirty or cloudy, it would greatly reduce my ability to use or enjoy the river. ___ I would avoid the river if it was a dirty or cloudy river. ___ I do not currently use or enjoy the river.  (27) Please list the reason(s) you chose to describe the river as dirty.
[If you chose Slightly dirty or Extremely dirty.] (28) Usually the water in the San Marcos River is clean and clear. If the river became dirty or cloudy, would you still use and enjoy it the way you do now?
___ I would continue to use or enjoy the river the way I do now. ___ I would still enjoy the river, but less than I do now. ___ The cleanliness and clarity of the river has no effect on how much I use or enjoy the river. ___ If the river became dirty or cloudy, it would greatly reduce my ability to use or enjoy the river. ___ I would avoid the river if it was a dirty or cloudy river. ___ I do not currently use or enjoy the river.  (37) Please explain your answer to the above question concerning the sensitivity of the San Marcos River to rapid urban growth. (38) Describe the amount of time you spent enjoying outdoor activities during childhood and adolescence. ___ Regularly ___ Occasionally ___ Rarely ___ Never (39) List the outdoor activities you enjoyed doing during childhood and adolescence and then rank them accordingly with 1 being the most enjoyed.
[Enter "None" if your answer to the previous question was "Never".]  Thank you for completing this survey. We appreciate your participation and are happy to discuss any questions or comments you may have.     Table A26. Environmental values of user groups, as tested using Kruskal-Wallis and Dunn's multiple comparison.

User Group
Chi-Square

Resident
(Non-Student) Resident (Student) Regional Student

Non-Regional Tourist
Post Hoc Summary

Med Mean SD
Habitat For plants and animals

Resident (Non-Student)
differs from Resident (Student), Regional Student, Regional Tourist, and Nonregional Tourist. Table A27. Environmental values of user groups, as tested using Kruskal-Wallis and Dunn's multiple comparison.