Priority river metrics for residents of an urbanized arid watershed

Landscape and Urban Planning 133 (2015) 37–52

Contents lists available at ScienceDirect

Landscape and Urban Planning journal homepage: www.elsevier.com/locate/landurbplan

Research paper

Priority river metrics for residents of an urbanized arid watershed Matthew A. Weber ∗ , Paul L. Ringold 1 United States Environmental Protection Agency, Office of Research and Development, Western Ecology Division, 200 SW 35th St., Corvallis, OR 97333, USA

h i g h l i g h t s • • • •

The general public provided extensive and thoughtful input on river monitoring goals. Native flora and fauna were usually preferred, despite not knowing specific species. Visible defacement such as garbage rivaled the importance of ecological features. Findings are reported as metrics accessible to natural and social scientists.

a r t i c l e

i n f o

Article history: Received 17 October 2013 Received in revised form 9 September 2014 Accepted 11 September 2014 Keywords: Qualitative research Focus groups Depth interviews Final ecosystem services Urban ecosystem services River monitoring

a b s t r a c t In river and stream assessment and management, a persistent question is isolating appropriate indicators of resource condition. We employ qualitative research techniques to identify features of rivers and streams important to the general public in an urbanized arid watershed of the southwestern US, based on interview and focus group data. After detailed analysis of transcriptions, findings were member-checked with new study participants and further revised. Theme frequencies are reported to provide an indication of participants’ informational priorities. Recurrent ecological themes were Water, Vegetation, and Fish and Wildlife; recurrent human themes were Garbage and Graffiti, Odor, Infrastructure, Other People, and Noise. Themes are further described along with illustrative quotes from participants. We interpret participant input into actionable metrics which could serve to track resource condition. Results are compared to previous research and current monitoring practice. The findings are particularly relevant for scientists and managers interested in the perspectives on rivers and streams held by residents of urbanized watersheds in arid landscapes. Published by Elsevier B.V.

1. Introduction In arid regions, water resources are a frequent focus of environmental management. In the US, perhaps nowhere else is water scarcity more acutely felt than in the Southwest. Furthermore, the consensus is that the Southwest will become even more arid within a period of years to decades (Seager et al., 2007). The instream flows and associated riparian ecosystems that remain disproportionately contribute to biodiversity as compared with other land cover in the area (Naiman, Décamps, & Pollock, 1993). Rivers support a variety of recreational activities, and provide a water supply for municipal, agricultural, and industrial uses. A systematic way of assessing the condition of arid river resources would assist with the difficult

∗ Corresponding author. Tel.: +1 541 754 4315; fax: +1 541 754 4799. E-mail addresses: [email protected] (M.A. Weber), [email protected] (P.L. Ringold). 1 Tel.: +1 541 754 4565; fax: +1 541 754 4799. http://dx.doi.org/10.1016/j.landurbplan.2014.09.006 0169-2046/Published by Elsevier B.V.

tradeoffs inherent in management, for example, balancing instream and extractive water uses. A consistent approach to assessing the condition of river resources requires a focus on specific, measureable features. Traditionally, the details of selecting these features have been left to biophysical scientists such as ecologists. Yet acknowledging a need for public input in river management (e.g., Kondolf & Yang, 2008: Ch. 4) implies a necessity that the metrics of river condition be publicly relevant. Some biophysical scientists engaged in monitoring recognize that what is measured should be things people value (e.g. Jackson, Kurtz, & Fisher, 2000; National Research Council, 2000) but-perhaps due to the specialized and technical nature of deriving actionable river metrics-social scientists have not frequently addressed this issue. In this study we utilize social science techniques to systematically identify important river features, relying directly on input from residents of an arid case study location. We purposefully solicit feedback from the general public rather than high profile river stakeholders who may not represent general public interests.

38

M.A. Weber, P.L. Ringold / Landscape and Urban Planning 133 (2015) 37–52

The features emerging from this input provide a foundation for our interpretation of specific river metrics. These metrics represent information relevant to the research participants while providing a detailed reference for scientific monitoring, modeling, mapping, and communication. Our goal is to inform at least some of the difficult decisions involved in monitoring program execution. Challenges such as the subjectivity of what is ultimately measured given the wide range of possibilities and constraints have been detailed by biophysical scientists (Hughes, 1993; Hughes & Peck, 2008), but may not be obvious to those who have not had to address them. We believe additional gains in thoughtful monitoring design can be made by examining strategic decisions with the benefit of social science research. Most of our participants resided in an urban setting. While their comments did not exclusively pertain to the urban environment, much discussion did focus on urban river conditions. The importance of urban ecosystems is receiving increased recognition (e.g., a special issue of Landscape and Urban Planning, 2013, 109(1)). In the US, over 80% of people live in urban locations; in the Southwest the value exceeds 90% (US Census, 2010a). Human modifications are a given, but rivers and streams still provide some of the last dynamic natural areas in the urban landscape. Problems of urban environmental management can be considered distinct from rural or wilderness locations since urban areas are proximate to large human populations, thus careful management has a potential for high impact. The critical question we address is: what is the breadth and depth of river features biophysical scientists should measure and/or model that is of direct relevance to people? Choices that impact rivers are continuously being made. Identifying publicly relevant features is crucial in order to represent social values in river status and assessment. Although various sources contribute partial insights, documentation of the range of features people value is virtually nonexistent. Some river attributes, such as water itself, are traded in markets. However there are also a broad range of so-called “nonmarket” river attributes such as ecological amenities and scenic views relevant for quality-of-life. Nonmarket attributes of rivers do not have easily observed units or price tags. In favorable data circumstances these attributes can be explored through tactics such as nonmarket valuation (e.g. Freeman, 2003). Several valuation studies have addressed river-related attributes in the Southwest (e.g., Bark, Osgood, Colby, Katz, & Stromberg, 2009; Berrens, Bohara, Silva, Brookshire, & McKee, 2000; Colby & Wishart, 2002; Larson & Perrings, 2013; Weber & Stewart, 2009; Weber, Mozumder, & Berrens, 2012). However research designs have only treated a limited set of variables per study, and results are sometimes limited to specific groups. A separate branch of literature elucidates categories of importance for rivers, such as naturalness, access, and aesthetics (see Asakawa, Yoshida, & Yabe, 2004; Gobster & Westphal, 2004; Smith & Moore, 2011). Understanding categorical motivations for river values adds insight, but is an imperfect guide to identifying key features themselves. What is considered more or less aesthetic? What metrics need to be included in a description of “naturalness”? We know of only two prior studies involving social science research that consider an array of publicly important river features: Schiller et al. (2001), and Ringold, Boyd, Landers, and Weber (2009). In this paper we continue research on this topic, comparing and contrasting with these prior studies in the discussion. We utilize qualitative research techniques to identify recurrent participant themes. From these we interpret specific metrics to represent the status of the resource. These metrics guide specific measurements that could be made at a given river location. Our findings have application to a range of river monitoring and management questions for the study region and may reflect perspectives held by residents of other urbanized watersheds in the arid Western US.

2. Case study location All study participants were residents of southern Arizona, with the vast majority residing within the Santa Cruz River watershed, and Tucson in particular. Tucson is the largest population base in the Santa Cruz watershed, situated at an elevation of about 800 m above sea level, with just over 1 million inhabitants in the metropolitan area. Tucson is enriched in hispanic/latino culture, with 40% of persons from that racial group as compared with 15% nationally (US Census, 2010b). Less than 30 cm of rain fall in Tucson in an average year, concentrated in summer monsoon and winter rainy seasons. The geography is Sonoran Desert basin and range. Although surface water is scarce in the region, flowing rivers do exist. The Santa Cruz River itself is perennial in its uppermost reaches in the San Rafael valley of Southeastern Arizona, and in Northern Sonora, Mexico. The river was perennial within recorded history downstream in downtown Tucson (Logan, 2002). The Tucson reach was heavily impacted by downcutting induced by natural floods, human-assisted channelization to mitigate flooding, and groundwater development, removing the historic riparian area (Webb, Leake, & Turner, 2007: Ch. 21, esp. p. 254). Most hydrological accounts indicate the 1940s as when groundwater-dependent perennial flow ceased in the Tucson reach of Santa Cruz River, as a result of centrifugal pumps for agricultural irrigation, a continuing irrigation method in the region. However, two reaches north of Mexico have become perennial in modern times due to the discharge of treated wastewater. Three major treatment plants discharge into the channel downstream of Mexico, one just north of the border, and two in northwest Tucson. These reaches replace some of the vegetative communities and other habitat that previously existed. For additional watershed background see Norman et al. (2010).

3. Methods 3.1. Conceptual approach In engaging research participants our overarching approach was to ask them to focus on physical, measureable features of rivers and streams of direct importance to them. If a participant were to mention a topic such as water quality, we would probe to learn more about why water quality was important to that person, such as important manifestations of high or low water quality. Furthermore, we would verify that these manifestations were of direct relevance. For example, dissolved oxygen is a technical water quality term which is unlikely to have direct meaning to a layperson, but which was occasionally mentioned. Through follow-up and probe questions the reason why dissolved oxygen was considered important could be determined, e.g. as a potential indicator for whether or not the water would support fish. Thus we endeavored to isolate features of direct importance to participants as opposed to underlying or intermediate ecological features or processes. Our research is an empirical application of the “Final Ecosystem Services” conceptual framework described by previous authors (Boyd & Banzhaf, 2007; Boyd & Krupnick, 2013; Ringold, Boyd, Landers, & Weber, 2013). Encouraging participants to think about and express final outcomes rather than intermediate factors promotes clarity on what it ultimately relevant to them, and eliminates double counting of potentially interdependent features, such as dissolved oxygen and fish. Collecting public input to inform river monitoring required both a breadth and depth of public input. This directed us toward qualitative methods of focus groups and interviews since they allow wide-ranging commentary as well as in-depth discussion with participants. In a focus group, multiple perspectives can be

M.A. Weber, P.L. Ringold / Landscape and Urban Planning 133 (2015) 37–52

represented, and there is an opportunity to witness discourse between participants. In an interview, concentrated feedback from an individual can be collected, free of influence from other respondents. Interviews are regarded as a safer forum for controversial topics (Kaplowitz, 2000). For both approaches, open-ended questions invite broad exploration of the research topic, while back and forth dialogue using probe and follow-up questions can be employed when detail is needed. For study design, fieldwork, and analysis methods we relied mainly on Morgan and Krueger (1998) for focus groups, and Rubin and Rubin (2005) for depth interviews.

39

A four-part script was used for both stages 1 and 2. In the first part, an overview of the study purpose was given. Then, the confidentiality of participant input was described. Participants were then asked if they had any questions before the session began, after which informed consent was obtained. In focus groups, to help each attendee feel comfortable speaking, the script began with an icebreaking question, asking how each person felt about the heat in southern Arizona. The second part of the script was oriented to gathering feedback regarding important, measureable features of rivers and streams in the entire region, using the following three questions:

3.2. Data collection stages Data collection followed a phased progression of three stages. Iterative, adaptive data collection is common in qualitative inquiry to allow refinement of research strategies (: p. 43–46; Patton, 2002). For all stages, residents of southern Arizona were recruited as research participants. Stage 1 was a pilot phase of the study designed to refine ideas with participants recruited from a contact list of Tucson neighborhood association presidents. These persons were representing officers of specific bounded areas registered with the city as official neighborhoods. From the full list of 140 registered neighborhoods, those adjacent to the Santa Cruz River were selected. Neighborhood presidents for these areas were contacted directly by the first author, and told that it was a study on rivers and streams including the Santa Cruz River. Ultimately 72% of those contacted agreed to participate, resulting in 13 interviews taking place in the fall of 2010. A critical insight of this stage was noting the difficulty participants had orienting their input toward specific, measureable river and stream features; this observation was used to develop more effective probe and follow-up questions for stage 2. Stage 2 was the bulk of the data collection, entailing ten focus groups involving seventy persons, occurring in the spring of 2011. In this stage we wished to limit self-selection of people particularly interested in rivers and streams. For example, when making cold calls the recruiter (a marketing firm) used a broadly worded script: Hello, this is calling from [recruiting firm] to invite you to join a paid focus group in early May. You will receive $75 for your participation. This is a public opinion study about an environmental topic and involves nothing of a personal nature. The focus group will last about 2 hours. To see if you qualify or to get more information, please call our toll free # [toll free # here with extension]. Most recruits (65%) were secured via cold calls to a randomly generated list of households sorted by zip code, using a phone list purchased from a survey research firm. Telephone recruiting was supplemented with internet notices and flyers placed in local library branches, with a given focus group typically containing a mix of recruiting sources. After recruits were scheduled, in the interest of transparency they were then told it was a government sponsored study on public opinion regarding rivers and streams in southern Arizona. A purposive sample approach was taken (Patton, 2002) in targeting five zip code clusters in Tucson having diverse sociodemographics. Within stage 2 two additional focus groups were held further south in rural areas but still within the Santa Cruz River watershed. These allowed exploration of potentially contrasting views outside of Tucson (population ∼1 million in the metropolitan area). One group was held in the small town of Rio Rico (population ∼19 thousand), and another was held in the community of Tubac (population ∼1 thousand). Since a relatively large percentage of residents speak a language other than English at home in South Tucson and in Rio Rico (US Census, 2010b), these focus groups were conducted in Spanish with the assistance of a Spanish speaking comoderator. The recruiter gathered sociodemographic background on all stage 2 participants including their outdoor water-related recreation habits.

1. ‘What is your familiarity with rivers and streams?’ 2. ‘What are the first things that come to mind for you for rivers and streams in southern Arizona?’ 3. ‘Is there anything important to you about rivers and streams in southern Arizona?’ The third part of the script turned to the Santa Cruz River in particular, although many participants had by then already mentioned the Santa Cruz River. Focusing on the Santa Cruz River allowed the moderator to provide background, a map, and photographs for this specific river likely to be familiar to participants, and gauge their reaction to that information. After question 4, the moderator distributed Santa Cruz River background and photos, attached as Fig. 1A and B. Four questions total were asked in relationship to the Santa Cruz River, with two final wrap-up questions, all designed to help participants isolate measureable features important to them: 4. ‘What are the first things that come to mind for you for the Santa Cruz River?’ 5. ‘Is there anything important to you about the Santa Cruz River?’ 6. ‘Please rank the Photos 1–4 in terms of what is most important to you’ (emphasis on why). 7. ‘If you were in charge of the Santa Cruz River what would you do?’ 8. ‘If you had to name one thing you want me to pay attention to, what is it?’ 9. ‘Is there anything else you’d like to add?’ Before questions 2, 3, 4, 5, and 6 focus group participants were asked to jot down notes on a provided piece of paper, to reinforce individual thinking. Follow-up and probe questions were utilized particularly for questions 3 and 5 to isolate measureable features ultimately important to participants. Stage 3 involved reviewing a draft set of research findings with fifteen additional persons in the spring of 2013 to ‘member-check’ validity of results (Creswell & Miller, 2000). The first part of the script was the same as that used in stages 1 and 2. However, in contrast to the more open-ended questioning previously used, a table of important river and stream features as identified by previous participants was now furnished. Member-checkers were asked to react to the features listed, endorse features important to them, point out what might be missing, and describe what could be dropped. Similar to stage 2, a marketing firm was used to recruit, screen, and compensate participants, with a script designed to limit selfselection. Member-checking sessions occurred in Tucson as well as Phoenix. Phoenix was the only recruitment source outside of the Santa Cruz River watershed, allowing some investigation of whether earlier study findings would apply for participants from another large metropolitan area in southern Arizona. One focus group of four persons was held in both Tucson and Phoenix. There were five depth interviews in Tucson, and two depth interviews in Phoenix. Input during this study stage largely confirmed findings in previous sessions but also resulted in new insights, described in the relevant section of the results and discussion.

40

M.A. Weber, P.L. Ringold / Landscape and Urban Planning 133 (2015) 37–52

Fig. 1. (A) English version of map and background text. (B) English version of photos accompanying map.

The first author moderated all focus group and interview sessions. The moderator had resided in Tucson from 1995 to 2007 and had extensive knowledge of rivers and streams throughout southern Arizona. This helped build rapport with participants

during casual conversation held before a session began, and helped with understanding regional references made during the meeting, allowing more efficient construction of follow-up and probe questions.

M.A. Weber, P.L. Ringold / Landscape and Urban Planning 133 (2015) 37–52

3.3. Analysis All human subject interactions were tape recorded and transcribed, becoming the primary data for analysis. Each data collection session comprised a separate transcript. The data were reviewed by the first author through an iterative process of reading transcripts and attaching codewords to reoccurring types of participant input, referred to as themes (Morgan & Krueger, 1998; Rubin & Rubin, 2005). To offer direct guidance for river and stream assessment metrics the analysis focused on coding themes where interest in distinct features could be identified. Interest in ecological features was recorded in a separate category of themes from interest in obvious human modifications of rivers and streams. This was done to distinguish features of rivers and streams important due to their ecological characteristics (such as various flora and fauna) versus features important due to human inputs (such as recreational infrastructure). Provisional themes were refined with additional reviews of the data, ultimately defining strict rules to finalize the codebook and guide when to count an occurrence of a particular code. Themes were then organized into subcategories for ease of presenting results. We emphasize that only references to measureable quantities are represented in code frequencies: for example the words “wildlife” and “pollution” were ubiquitous, but only when more specific types of wildlife could be elicited, or specific negative manifestations of pollution stated, could the frequency count for a corresponding theme be impacted. To systematically count code frequencies, continuous blocks of text associated with different speakers were used to break each transcription into units (: p. 105; Wutich, Lant, White, Larson, & Gartin, 2010). A block of text for a speaker that was briefly interrupted but then continued speaking was treated as a single unit for that speaker. A given transcript contains multiple units associated with each speaker. A given theme could be coded only once within each unit; however, if a speaker returned to the same theme it could be coded again later. It was verified that none of the major themes were represented by a small number of participants. However there are several themes with relatively low incidence. The minimum criterion for a theme was that it be supported by at least two participants in different sessions. In the results and discussion, themes referred to as “minor” occur less frequently than the median theme code count of 18. Software (The Ethnograph, v6) was used to assist in tracking codes and codebook organization. Participants were encouraged to think in terms of measureable quantities, yet researcher judgment was needed to translate themes into the more technical language of metrics. In most cases metrics are a straightforward interpretation of data, worded so as to give clear direction for actual measurements that could be taken at a stream site. To economize space, we tabulate interpreted metrics but do not report them within the text. In some cases a difficult choice had to be made in developing the metric; these instances are documented in the relevant section of the results and discussion. The wording of metrics was led by the second author who has experience developing stream and riparian indicators for stream monitoring programs. Descriptions of metrics include units and temporal character. There are a number of challenges associated with focus group and interview research, some of which we highlight here. Interviewer-induced-bias via the moderator is a concern. We believe moderator introduced information and topics influenced the frequency of a few of our codes. These cases are discussed within the relevant section of the results. The phenomenon of ‘groupthink’ due to dominant participants is a known source of bias in focus groups (Janis, 1972). Undoubtedly this affects our results; however we safeguard against it with multiple groups. Focus group and interview sample sizes tend to be relatively small, and insufficient to statistically represent the population. Instead of

41

representativeness, we sought to collect input from participants of diverse backgrounds, and to achieve a ‘saturation’ of input (Glaser & Strauss, 1967) with minimal new information collected by the end of the study. However, new minor codes might appear if the sample size were expanded, and relative code frequencies might shift. Criteria for judging the validity of qualitative research are debated, but it has been suggested that validity be based on how well the researcher applies the method to fit the study objective (: p. 95; Fern, 2001). Our goal was to document detailed input on important stream characteristics from a diverse sample of the general public in the case study location. We describe possibilities for further study in the conclusions.

4. Results and discussion Sociodemographic data for participants of all three stages are presented in Table 1. Focus group summary information such as group number, location, and number of attendees is summarized in Table 2. A diverse sample of ages, incomes, occupations, and durations of residence in the region were included in the study. Participants were roughly split between those self-identifying as engaging in significant river and stream related recreation, and those who did not. A handful of people expressed little to no recreational interest in rivers and streams; also rare were the especially avid. We found the main theme categories for ecological features to be Water, Vegetation, and Fish and Wildlife. Theme categories for human modifications were Garbage and Graffiti, Odor, Infrastructure, Other People, and Noise. Themes and interpreted metrics are summarized in Table 3. The frequency of coding for a given theme, including the breakdown across data collection stages, and the number of focus groups the theme appeared in, is reported to offer a sense of the relative importance of each theme. Frequencies are listed in Table 3 in descending order within a theme category. In the text we describe the few cases for which rural group code frequencies were remarkably different from urban groups, rather than further disaggregating Table 3. Feedback from the relatively few residents of Phoenix included within stage 3 did not appear to be remarkably different from residents of the Santa Cruz River watershed. Thus, results for Phoenix participants alone are not discussed in the text. Throughout results and discussion, quotations are included to illustrate main findings. Quotations are italicized and are presented with gender-specific pseudonyms.

4.1. Water 4.1.1. Presence of water The mere presence of water was the standout theme voiced by respondents. Surface water is exceedingly scarce in the region, thus just seeing water in a riverbed is noteworthy. Jacob was one of many who recounted their surprise upon moving to the area. Jacob (group 5): I came out from California. And they said we got a river! And I was like shoot, cool! You know? And it was like a bed of sand. What is that? That’s the river. You just gotta wait some months, and then the water will come. And it only lasts like, a week or something? That’s the river. A frequent adjective focus group participants used to describe regional rivers and streams was “dry”, in contrast with expectations or desires. Many comments regarding the lack of “real” rivers were imbued with a sense of loss—people believed that things used to be better. Participants commonly spoke about how the Santa Cruz River used to have water. Gloria was a 70 year-old woman, born in Tucson, who witnessed changes.

42

M.A. Weber, P.L. Ringold / Landscape and Urban Planning 133 (2015) 37–52

Table 1 Descriptive statistics for participants of all three study stages. Information demonstrates the diverse range of sociodemographics of research participants; we do not claim that the sample was statistically representative of the study area population. Stage 1* n = 13

Stage 2 n = 70

Stage 3 n = 15

– – – 13 (100%)

46 (66%) 10 (14%) 14 (20%) –

7 (47%) 8 (53%) – –

7 (54%) 6 (46%)

39 (56%) 31 (44%)

9 (60%) 6 (40%)

– 1 (8%) 3 (23%) 7 (54%) 2 (15%)

13 (19%) 10 (14%) 7 (10%) 21 (30%) 19 (27%)

1 (7%) 3 (20%) 7 (47%) 2 (13%) 2 (13%)

Household income <20k 20 to 40k 40 to 60k 60 to 100k >100k Refused

18 (26%) 21 (30%) 7 (10%) 5 (7%) 4 (6%) 15 (21%)

2 (13%) 3 (20%) 3 (20%) 4 (27%) 2 (13%) 1 (7%)

Employment Homemaker Student Other employment Unemployed Retired Disabled

3 (4%) 5 (7%) 45 (64%) 2 (3%) 13 (19%) 2 (3%)

1 (7%) – 13 (87%) 1 (7%) – –

Years in present city 0 to 9 10 to 19 20 to 29 30 to 39 40 to 49 50 to 59 60 or more

18 (26%) 10 (14%) 14 (20%) 14 (20%) 6 (9%) 6 (9%) 2 (3%)

6 (40%) 4 (27%) 2 (13%) 3 (20%) – – –

Years in Southwestern US 0 to 9 10 to 19 20 to 29 30 to 39 40 to 49 50 to 59 60 or more

9 (13%) 10 (14%) 17 (24%) 14 (20%) 8 (11%) 8 (11%) 4 (6%)

4 (27%) 4 (27%) 2 (13%) 3 (20%) 1 (7%) 0 (0%) 1 (7%)

Variable Source Cold call random sample Internet posting Library flyer Neighborhood president contact list Gender M F Age 18 to 29 30 to 39 40 to 49 50 to 59 60 and over

User type (self-identified during screening) Significant recreation No significant recreation

8 (62%) 5 (38%)

37 (53%) 33 (47%)

7 (47%) 8 (53%)

Residence (or closest city) Tucson Rio Rico Tubac Phoenix

13 (100%) – – –

56 (80%) 9 (13%) 5 (7%) -

9 (60%) – – 6 (40%)

* Stage 1 sociodemographic data is based on observation or participant comments. “” = No data. “–” = No participants in this category for this stage.

Gloria (group 7): . . .the past when things were good, and we were young and in the river, and all that, and I don’t see that anymore. I mean, basically because there’s not enough water, and we can’t just say okay, we want water and it’s going to rain. It won’t happen. It won’t happen. It should be noted that the reaches of the Santa Cruz River pictured in photos A and D (Fig. 1B) flow due to releases of treated effluent. Release of municipal effluent into waterways is common worldwide, and a primary water source for numerous waterways in Arizona (: Ch. 2). However, ironically, we found that the vast majority of participants were completely unaware of these ‘new’

perennial reaches of the Santa Cruz River despite their expressed interest in visible surface water. The source of Santa Cruz River surface flow was read out loud to stage 1 and 2 participants along with other background material provided between questions 4 and 5 to ensure they understood it was effluent (see Fig. 1A and 1B). Tucson residents were so used to seeing a dry Santa Cruz River, expressions of disbelief were common upon seeing the photos. So accustomed were respondents to surface water scarcity, every drop literally mattered. When looking at the photos, people peered at the dry streambeds in photos B and C for evidence of recent moisture, and strained at the photo margins to glimpse

M.A. Weber, P.L. Ringold / Landscape and Urban Planning 133 (2015) 37–52

43

Table 2 Focus group information. Group no.

Stage

Location

Language

Character

Number of attendees

1 2 3 4 5 6 7 8 9 10 11 12

2 2 2 2 2 2 2 2 2 2 3 3

Northwest Tucson Northwest Tucson Central Tucson Central Tucson Midtown Tucson North Foothills Tucson South Tucson South Tucson Rio Rico Tubac Central Tucson Central Phoenix

English English English English English English English Spanish Spanish English English English

Urban Urban Urban Urban Urban Urban Urban Urban Rural Rural Urban Urban

7 7 7 4 9 9 7 6 7* 7* 4 4

* The number of attendees in the Rio Rico and Tubac groups was both 7, but 2 of the attendees of the Tubac group were actually from nearby Rio Rico. Thus Table 1 reflects 9 residents of Rio Rico and 5 residents of Tubac.

vegetation just out of view, that might signal an oasis. It was typical for Q6 photo ranking to be based on the presence of water, even the possibility of water (note that photo ranking was an exercise to elicit deeper discussion of preferences for specific river features, we do not present photo rankings themselves). Dave (group 10): [Photo] A number one. It looks like a natural setting. Uh, D number two because there’s some water in it [laughs]. I picked C number three because it looks like there might be water in it, you know, at some point. And then B just, it looks, looks like a dirt road. Contrary to what might be assumed, affinity for visible water wasn’t limited to recreational reasons, the motivations were much broader. Louise was a non-recreationalist who said “I stay inside as much as possible, to avoid the sun”. Her reason for interest in surface water reflects comments many participants made. Louise (group 2): . . .it would mean we are respecting nature, we are making the best of it, getting things back to where they were before we had so many people in the area, in that regard we want to take a step back. About one-sixth of the references to the presence of surface water mentioned flowing water in particular. A minor theme counterpoint to the presence of water was several negative references to stagnant water and vector-borne diseases such as West Nile virus, indicating that surface water was not always viewed positively.

by capturing episodic stream runoff, for either direct use or to promote recharge to the aquifer. Aside from quantity, the quality of water supply as based on riverine sources was the third strongest water theme. Associated comments were overwhelmingly focused on human health concerns, with a minor, separate theme on disagreeable tap water taste. Similar results were found in nearby Phoenix by Gartin, Crona, Wutich, and Westerhoff (2010), who documented doubt among urban residents that authorities are adequately investing to protect water quality. Delilah (group 6): Because we used to have well water which we used to drink which was really good here and now we have CAP water which nobody drinks. [Crosstalk] Moderator: I’m sorry? Leroy: I said it is horrible water. I’m used to drinking out the faucet back east like everybody else and I came down here and stopped. It just messes with me. John: I heard Tucson was one of the worst, for calcium. I didn’t find that out until I found out I had kidney stones. They’re not fun. The Rio Rico group was the most concerned with water supply quality, accounting for one-third of the stage 2 count for this theme. Rio Rico participants cited an array of issues: polluted groundwater from surface water discharges from factories; poorly constructed septic systems; and outdated asbestos water piping.

4.1.2. Water supply Just as the presence of water in the riverbed wasn’t taken for granted, neither was it assumed there would be enough water for one’s household or the larger community; this was the second most frequent water theme. Dependence on groundwater in Tucson has been partially mitigated by the Central Arizona Project (CAP) canal bringing water from the Colorado River since the early 1990s, but water supply appears to remain a top concern within the regional consciousness. There is a cognitive association that wet rivers signal adequate water supply. Rose responds to Q2:

4.1.3. Flooding Ironically, there was also a theme of too much water. Although the region is marked by low rainfall overall, flooding does occur. The most important flooding theme was a concern for property damage. Several people remembered the flood of 1983 that damaged numerous structures. Some people expressed frustration with houses built on the floodplain as disasters waiting to happen.

Rose (group 5): I put down: they’re dry; too few; need to be renewed; a major issue. I mean how long can we continue to support the number of people who are continuing to move in here with the water supply that we know is, has a finite amount to it.

Pedro (group 8): Like I said . . . around that new neighborhood, the houses are right along the edge of the river. There is no protection. Those people used to grow crops, and they sold the land to builders. They made their money, but the people that are buying there don’t know about the consequences.

Residential water harvesting practices were commonly described. The effort some participants made (e.g. retaining rooftop rain runoff into cisterns) suggested a significant water conservation ethic since local water rates are not especially high. Many respondents expressed confusion and concern as to the lack of concerted water harvesting efforts on a regional scale, such as

Certain roads, river crossings, and underpasses are commonly flooded and closed in Tucson during seasonal “monsoon” runoff. There were concerns for human safety, as when people or cars crossing an arroyo are caught in a “flashflood”. The inconvenience of driving during periods of high water was also mentioned as a minor theme.

44

M.A. Weber, P.L. Ringold / Landscape and Urban Planning 133 (2015) 37–52

Table 3 Summary of respondent themes, interpreted metrics, and code counts. Code counts are included to convey a sense of relative levels of interest among research participants, but should not be interpreted as strictly representative of the study area population. Stage 2 data includes the number of groups the theme appeared in. Category

Theme

Metric

Stage 1 count

Stage 2 count (number of groups)

Stage 3 count

Total count

Appreciation of the presence of surface water Concern about water supply quantity being enough for human demand Concern about water supply safety, health risks of drinking water associated with river water Concern about the safety of recreational contact with river water Concern about property damage from floods Interest in swimming

Percentage of days of surface water per year Minimum surface water flow, minimum surface water volume, and minimum aquifer volume, per year Probability of water-borne illness from drinking tap water associated with river water Probability of water-borne illness from partial body contact and full body contact with river water Annual probability of flooding inundating sensitive property Whether minimum thalweg depth allows for swimming Minimum main channel depth and width, class of rapids, and presence of navigation hazards such as downed trees Annual probability of flooding at or above bankfull Annual probability of sudden increase in flow volume and velocity making wading or driving through established crossings dangerous Probability of illness due to nearby surface water via vectors such as mosquitoes Total volume of flow per year and minimum surface water flow per year Clarity of water, depth of visibility Presence of the sound of flowing water

34

170 (10)

6

210

37

57 (9)

3

97

15

52 (10)

6

73

4

35 (10)

4

43

6

33 (8)

3

42

6

27 (9)

3

36

5

26 (7)

3

34

9

13 (7)

1

23

0

14 (6)

4

18

0

7 (3)

3

10

0

9 (3)

1

10

1 2

9 (5) 4 (3)

0 1

10 7

Presence and abundance of algae Local tap water taste rating Annual probability of flooding inundating nearby roads Presence and abundance of trees large enough to provide shade for people Presence and abundance of lush green vegetation including shrubs, grass and reeds (other than large trees) Presence and abundance of invasive plants

1 0 0

2 (1) 2 (1) 5 (4)

4 3 0

7 5 5

10

136

Ecological Water

Interest in any form of navigation on surface water Interest in large flow events as natural phenomena Concern of drowning risk due to flashflooding

Vegetation

Concern of disease risk associated with standing water Interest in flow being increased or concern that flow will be reduced Appreciation of water clarity Appreciation of the sound of flowing water Dislike of algae Dislike of tap water taste Inconvenience of flooding interfering with driving Appreciation of large trees Appreciation of greenery or lush vegetation (other than trees) Concern of non-native plants overtaking native species Appreciation of there being different kinds of plants Appreciation of wildflowers Dislike of non-native nuisance plants

Fish and wildlife

Concern for plants in danger of going extinct Appreciation of harvesting opportunities for medicinal or edible plants Appreciation of birds in general or a specific bird species Appreciation of wildlife fitting into the mammal class Appreciation of fishing opportunities or reference to common gamefish Appreciation of the presence of fish in general or of any specific kind of fish (other than gamefish) Concern of nuisance fish and wildlife damaging property, harming people, or that are commonly feared Appreciation of wildlife fitting into the reptile class

40

86 (10)

11

34 (9)

5

50

6

14 (4)

6

26

3

5 (4)

0

8

2 0

4 (3) 7 (3)

1 0

7 7

0

0

3

3

1

0

1

2

12

40 (10)

2

54

6

34 (10)

9

49

6

27 (9)

5

38

Presence and abundance of all fish species

9

24 (9)

4

37

Presence and abundance of wildlife known to harm humans, damage property, or that are commonly feared

0

22 (9)

5

27

Presence and abundance of reptile species

6

13 (7)

0

19

Total number of different types of appreciated plant species (see other Vegetation themes) Presence and abundance of wildflowers Presence and abundance of non-native plants high in allergens or high in water use List of plant species present that are in danger of extinction List of plant species present that are edible or medicinal Presence and abundance of bird species Presence and abundance of mammal species, especially larger mammals, including predators Presence and abundance of game fish species

M.A. Weber, P.L. Ringold / Landscape and Urban Planning 133 (2015) 37–52

45

Table 3 (Continued) Category

Theme

Metric

Concern about fish and wildlife that overtake native species Concern for fish and wildlife in danger of going extinct Appreciation of there being different kinds of fish and wildlife

Presence and abundance of invasive wildlife List of fish and wildlife species present that are in danger of extinction Total number of different types of appreciated fish and wildlife species (see other Fish and Wildlife themes) Presence and abundance of abnormalities in wildlife susceptible to pollution-induced mutations, such as amphibians Presence and abundance of frog species Presence and abundance of butterfly species

Concern for pollution-induced mutations in fish and wildlife Appreciation of frogs Appreciation of butterflies Human Garbage and Graffiti

Dislike of garbage or graffiti

Odor

Dislike of human-caused odor

Infrastructure

Positive reference to an existing paved trail or appreciation of trails for biking along waterways Dislike of obvious human modifications to the channel detracting from site aesthetics

Other people

Noise

Dislike of obvious human modifications on the bank detracting from site aesthetics Appreciation of trails for walking or hiking along waterways Appreciation of basic recreational facilities along waterways other than trails Interest in others recreationally using waterways Concern for personal safety or discomfort regarding potential interactions with other people Dislike of crowds or seeing others while recreating along waterways Dislike of human-caused noise

Stage 1 count

Stage 2 count (number of groups)

Stage 3 count

0

10 (3)

8

18

4

4 (2)

6

14

3

9 (6)

0

12

0

1

3

4

0 1

3 (3) 2 (2)

1 0

4 3

83 (10)

9

105

8

33 (8)

3

44

17

23 (7)

0

40

6

15 (5)

1

22

3

9 (5)

4

16

3

11 (7)

1

15

Presence of picnic tables, bathrooms, drinking water, and shade structures

3

6 (5)

1

10

Number of recreational users present

13

16 (6)

0

29

Presence of homeless persons or camps, and presence of crime

7

12 (5)

1

20

Number of recreational users present

0

2 (2)

5

7

Presence and description of sound of human origin

0

14 (6)

5

19

Presence and abundance of garbage and/or graffiti along waterways Presence and description of odor of human origin Presence and extent of paved trails

Description of infrastructure to the channel visually impeding otherwise “natural” viewscapes such as bank protection, powerline pole placement, and bridges Description of infrastructure on the bank visually impeding otherwise “natural” viewscapes such as railings or buildings Presence and extent of unpaved trails

13

Total count

The second most frequent flooding theme was positive comments and/or comments of awe. Many people described the strong, episodic flows associated with the rainy seasons as a powerful spectacle worth experiencing.

concerns for all waterways, but heightened when discussion turned to the Santa Cruz River since it is prone to urban pollution and was also described as carrying treated wastewater in the provided background. Typical was Henry’s comment:

Ella (stage 1 interviewee): . . .the kids and everybody goes to see it, the Santa Cruz I‘m talking, just here, so we have the monsoons and it just feeds us, it’s like natural beauty.

Henry (group 1): If I walk through that water barefoot, am I going to have to worry about being ill from it, getting cancer twenty years down the road from constantly walking in there.

4.1.4. Water-related recreation Despite it being a desert environment, or perhaps because of this, water-related recreation was frequently mentioned. Numerous recreational activities were cited throughout the sessions. Most of the locations mentioned for water contact recreation were in mountainous areas outside of the city. Jessie (group 6): A lot of people will ask if there’s water–‘Is there water up there? Okay I‘ll go.’ A lot of people won’t go if there’s no water. Kind of goes along with gettin’ wet, you know the kids want to get wet everybody wants to get wet it’s the thing here in Arizona, well especially in southern Arizona, people flock to any kind of water. The most important water recreation theme was a concern of the safety of contacting potentially polluted water; people were concerned for themselves, their children, their dogs, or other people. The health effects of contacting river water were background

The second most important water recreation theme was interest in navigation. Navigation references were overwhelmingly in regards to small craft, such as tubing, kayaking, and canoeing. Another water recreation theme was numerous references to swimming, for which a metric was developed to indicate whether the water was deep enough to allow swimming. References to wading were also made, however this recreational activity does not appear in a separate theme since it did not lead to new guidance for monitoring. References to wading are instead represented in the code counts for appreciating the presence of surface water, and concern about the safety of recreational contact with water. There were a number of comments appreciative of waterrelated recreation of a non-contact form. Just being near water can be enough. Connie (group 7): . . .it’s one thing to like live in the Southwest and be aware of like you know monsoon season and rains and streams and stuff like that and actually BE right next to a river.

46

M.A. Weber, P.L. Ringold / Landscape and Urban Planning 133 (2015) 37–52

There’s just something about it that just feels like you are home. It’s like deep inside you bones, where you feel like you are supposed to be near water [agreement from others]. You know what I mean? It just feels very natural [emphasis in the original]. There were several minor aesthetic themes associated with noncontact water recreation. Water clarity was prized, as opposed to water that was muddy or contained algae. Some people spoke of the sound of running water, an aesthetic preference also found by Hetherington, Daniel, and Brown (1993). The quantity of flow was another theme mentioned in association with aesthetics, but also in regards to water supply, recreation, and flow-dependent ecology. For additional context we note that there were also references to recreation in dry washes. These were sometimes seen as places to walk, explore, or drive off-highway vehicles. Two people described dry washes as safer places for kids to play, rather than flowing streams with potentially contaminated water. 4.2. Vegetation Lush vegetation is sparse in the region and concentrated in places with water. Our participants were strongly aware of this association, mentioning vegetation associated with rivers almost as often as the presence of water, even noting that increased vegetation is actually a more reliably present feature of riverbeds than water. A human preference for riparian vegetation in arid lands was speculated by Barmil, Daniel, and Hetherington (1999), who noted that few studies had directly addressed it. Bark et al. (2009) established that Tucson homebuyers are willing to pay more to be near woody plants and tall trees along waterways. The preferred type of vegetation was trees, the key characteristic being their height. The shady refuge trees provide was seen as a welcome contrast to the typical desert landscape. The photos were an effective device for seeding discussion on vegetative differences. Blake (group 1): I think the thing, the big difference [with D as compared with A], is there’s no trees so that kind of like turns me off to going there [agreement from others] . . .honestly I wouldn’t go that far out of my way just to enjoy a river, unless it was like a trip I planned. But yeah I mean just like in town, more than anything, there’s not really shade. You’re just sitting out in the sun. So as green as you want to make this [photo D] if there’s no trees it kind of takes away from it. The Tubac group had the most references to trees, accounting for one-sixth of the stage 2 count for this theme. This small community is concentrated along the Santa Cruz River channel. In contrast to Tucson groups, the reach of the effluent-dominated Santa Cruz River nearest them was well-known. In the Tubac area tree roots can still reach the groundwater table and tree growth along the river is more extensive than in Tucson, something participants appreciated. Photo A was far and away the winning landscape. This is partially evident in what people said about photo A, but also evident in what they said about photo B, which drew surprisingly strong negative reactions from several people. One woman pictured herself “dying in the dirt, it’s so dry”. It was also described as “pretty worthless soil”, “hideous”, “sad”, and “pretty ugly”. Penny (group 4): And then B looks very barren to me. Nothing even a bush that looks more than a foot high, two feet [thirty centimeters, sixty centimeters]. So if you were walking along there you couldn’t get cooled off by the water and you couldn’t even hide under a bush for shade [group laughter]. Interestingly there were relatively few references to specific vegetation, despite moderator probes. The emphasis was simply a desire to have more of it, especially trees. The second most

frequent type of vegetation desired was references to green or lush vegetation. The third most important vegetation theme was dislike of invasive plants. The issue of invasive species typically arose spontaneously. However the research team directly queried participant perspectives on this issue in two groups that had extra discussion time remaining. Thus there is some impact from moderation on related code frequencies. Discussion showed an overall preference for native vegetation, with some voices expressing neutrality. However particular native species were not often named and the existence of non-native species by itself was not of broad concern. Thus particular measurements could not be derived from these perspectives. Instead, the emphasis was on invasiveness. The concern was for plants that “overgrow” or “take over” other native plant species. There was the sense that non-native species would throw things “out of balance”, with unknown and likely negative consequences. A relatively minor vegetation theme was appreciation of a diversity and/or variety of plants. Interpreting this into a metric is challenging since there is a rich debate on how to measure biodiversity, but only minimal participant input on this theme was collected. We did not have the time to introduce terminology and investigate issues such as species diversity vs. genetic diversity, or species richness vs. species evenness (Stirling & Wilsey, 2001; United Nations Environment Program, 1995). In the absence of more in-depth participant input we propose the number of different species (species richness) as the simplest possible metric consistent with the few statements that were collected. One such statement follows: Susanna (group 2): In [photo] A it looks like there is life there, there is diversity in the different types of species. . . To constrain what would be counted, we narrow the species richness metric to being the sum of species of appreciated types of vegetation represented in other themes. The richness metric is also limited to native vegetation due to the documented concern of non-native plants overtaking native species. A few additional minor vegetation themes were also found. Discussions regarding native vs. non-native plants occasionally turned in the direction of disliking certain traits of non-native species (aside from invasiveness), such as their potential to introduce new allergens or high water needs. The minor theme of non-native nuisance plants was thus created. During member-checking, concerns regarding plants in danger of extinction were raised three times. We note that our metric “List of plant species present that are in danger of extinction” should be interpreted as biological criteria rather than legal criteria; that is, something could be in significant danger of extinction and not be on an official threatened or endangered species list. Finally, wildflowers and edible and medicinal plants were each mentioned a small number of times as appreciated vegetation types, with corresponding metrics appearing in Table 3. 4.3. Fish and wildlife The word “wildlife” was omnipresent, but typically unspecific. This was a crucial insight of respondent input: probes and followups were repeatedly asked for preferred forms of wildlife and respondents generally resisted. Moderator: What’s important about wildlife, in your words, what sort of information on wildlife is important to you. . . Daisy (stage 1 interviewee): Just maintaining habitat for the wildlife. Moderator: Is there certain types of wildlife you’re concerned about?

M.A. Weber, P.L. Ringold / Landscape and Urban Planning 133 (2015) 37–52

Daisy: I‘m concerned about ‘em all [laughs]. We offer two explanations for ‘general’ wildlife preferences. For some people, it may not matter what kind of wildlife it is; it is just important that the river support something. This point of view may stem in part from an unfamiliarity of the possibilities. Rather than naming preferred wildlife such as something that could be hunted, many participants instead expressed curiosity to learn more about what was present, as well as how human impacts had affected them. Moderator: . . .what about wildlife, what could we tell you. Audrey (group 3): What kinds of wildlife. There’s a lot of stuff we don’t see, snakes, frogs. So it would be nice to know what there is and if any of them are endangered. On the other hand, people with outdoor recreational pastimes did seem more able to name particular species. However, these listings had a rote character. The word “wildlife” was used by participants in such a way that it seemed to represent a package deal, rather than a la carte. The second explanation for general wildlife preferences is the pervasive sentiment across respondents that “everything is interconnected”. A commonly held belief in the “balance of nature” has been described by other authors (Ladle & Gillson, 2009). Participants frequently wondered about the unpredictable and negative effects that might come from tinkering with nature’s complex interrelationships. To endorse a specific form of wildlife would be contrary to this holistic view. It is possible that aggregated metrics of wildlife would better represent status of the resource to the layperson as opposed to the disaggregated metrics we provide. For example, an index of biotic integrity (e.g. Stoddard, Larsen, Hawkins, Johnson, & Norris, 2006) is a possible approach. Here, we provide metrics as finely resolved as possible, and must leave development of aggregated metrics as a topic for future research. Although there was usually a lack of specificity to direct queries, ‘high profile’ wildlife can be discerned. Some things were mentioned more than others. There were slightly more references to land wildlife rather than aquatic species. This reflects public awareness of desert riparian areas supporting numerous forms of life. The most commonly mentioned wildlife class was birds. Birds were typically referred to in the aggregate sense; only rarely was more specific language such as species used. The vagueness of bird comments is notable, given that southern Arizona is a world-renowned birding location due to geography of migratory paths, the mix of ecoregions, and the relative scarcity of surface water that concentrates birding sites. Another important category of wildlife was mammals, although this was a title that we applied; participant references to “mammals” were rare. A variety of animals fitting into the mammal class was mentioned, such as deer, coyote, bobcat, and mountain lion, as well as desert-specific species such as javelina. The most such references came from the Rio Rico group, accounting for one-fourth of the stage 2 count for this theme. Rio Rico residents mentioned seeing wildlife near their homes, as might be expected to be more common in a rural area than in the city. For all groups the emphasis was on larger mammals, perhaps because these are the most visible. During stage 3, another potential explanation was raised by two persons who made the distinction of predator mammals, predators being an indicator for them of ecosystem health. Another land wildlife theme was reptiles, especially snakes, turtles, and lizards. Butterflies were a minor theme. For aquatic wildlife, the most frequent class mentioned was fish. Similar to birds, only rarely were specific kinds of fish mentioned. Rather, it seemed to be of great significance if fish of any kind could be found. References to fishing were among the more prevalent water-related pastimes mentioned (usually in association with stocked lakes, rather than streams). Comments about fishing trips

47

usually did not mention target species, although trout was a game fish a few people noted in member-checking. Only a few people mentioned an interest in consuming caught fish. In fact, instead of discussing the fish, more frequently other reasons for fishing were mentioned, such as relaxation: Pearl (group 6): I love to fish. Being that I’m from the Midwest, and I rarely catch anything. But, it’s just the unstressfulness of it. So I enjoy that. There were two minor aquatic life themes, one was for frogs. The second was for what we call sensitive species; there were two references in stage 2 regarding concern about mutations of aquatic life. This theme was also supported by two people during memberchecking who noted that presence and abundance of listed types of fish and wildlife would not represent the full picture of wildlife health. While typically there was appreciation for fish and wildlife, some people raised exceptions for nuisance wildlife. Mosquitoes were maligned. Coyotes, javelina, and snakes, to the extent they threaten people or pets or are a nuisance to property were also sometimes mentioned negatively. Mathilda exemplifies how a simultaneous preference can be held for wildlife in general, even while subconsciously excepting certain ‘problem’ species, in this case coral snakes. Coral snakes, among the most venomous of North American snakes, resemble nonvenomous kingsnakes. Both are native to the area. Mathilda (group 4): I‘m concerned about the animals [agreement from male voice]. I don’t think that most of the people in Tucson are educated about snakes, the good ones, the bad ones, I don’t know if that could help. Because I know of people that have killed kingsnakes because they thought they were dangerous, and this type of thing. As described in the vegetation section, the topic of native vs. non-native species often arose spontaneously but was directly introduced in two sessions with remaining discussion time. Almost all comments were in favor of native animals, with a somewhat stronger sentiment than was shown for favoring native plants. As with plants, an important threshold was when non-native animals were invasive and threatened to exclude native animals. Concern was not necessarily warranted just because something was nonnative. The theme of extinction received more references in association with fish and wildlife than for vegetation. As with vegetation our metric should be interpreted as biological criteria rather than legal criteria since something need not be officially ‘listed’ as threatened or endangered to actually be in danger of extinction. In most cases there was not a sense that species would be individually missed, but rather that it would disrupt the balance of the ecosystem. Particular species of concern were mentioned by name only a handful of times. The abstract aversion to extinction appeared to be stronger than affinity for specific species. Fish and wildlife diversity also received a higher theme frequency than vegetation diversity. For context, there appeared to be less of an affinity for biodiversity per se, than a desire to keep things as they are “supposed to be”. For wildlife diversity we follow a similar strategy as with plants, focusing on species richness. We limit the sum of wildlife species to native wildlife, and sum the species of appreciated types of wildlife already appearing in other Fish and Wildlife themes.

4.4. Garbage and graffiti Trash, litter, garbage, and graffiti; for many people these words were the first thing that came to mind when discussing waterways.

48

M.A. Weber, P.L. Ringold / Landscape and Urban Planning 133 (2015) 37–52

This was the strongest theme within the human modifications category. Several dry washes, including the Santa Cruz, are visible from major thoroughfares in town. Leroy was one of the few people who engaged in almost no river-related recreation, yet like others was highly aware of the issue. Leroy (group 6): . . .have you ever seen the washes when they’re dry I think that’s pretty gross but maybe that’s just me, I walk a lot, and I walk by a lot of washes, and there’s a lot of stuff down there, and I‘m like, I can’t believe we even have this, it just grosses me out, I‘m not used to seeing stuff like that, so, maybe possibly someone could go down there and pick it up you know, back to her, her kids, they go there and explore, you can’t explore if there’s things like that, shirts and bottles. Delilah: It’s the pollution, it’s in the water and in the washes. Leroy: Like somebody really just goes there and dumps . . . you know how streams carry trash, doesn’t even look like that. Looks like somebody just picked a spot and emptied a garbage bag [agreement]. The implication is that even if other wishes could not be met, such as visible water, vegetation, and fish and wildlife, at least the area could be kept clean. Cleanliness would seem to be a significant step to enhancing the value of a stream site, or even reversing it from being an eyesore to an amenity. Garbage and graffiti was a strong urban theme, and even stronger for the rural groups. Rio Rico and Tubac had the highest occurrences, together accounting for one-third of the code count for stage 2. 4.5. Odor The introduction of Santa Cruz River background halfway through the sessions described the association between the river and the wastewater treatment plant, which was generally unknown. There were then numerous comments, influenced by this moderator-introduced information, associated with past experience of being near the wastewater treatment plant, such as driving by on the interstate. Very few people had visited or knew of the effluent-dominated river pictured in photo D. To the extent the river itself has an unpleasant odor, or even if there is an ambient odor stemming from a nearby source, people were concerned. 4.6. Infrastructure Recreational amenities were almost always seen as positive, although it was possible to overdo it. A balance between facilitating human use while retaining as many natural elements as possible was the typical goal. Most of the comments were in reference to paved trails, with the existing paved trail along the Santa Cruz River in Tucson cited most often. The concrete path pictured in photo C was seen as a positive feature for the majority, even though photo C was typically rated third overall due to lack of water or greenery. Some people wanted additional amenities for this location, for example extensions to the path so that it would go further and connect to more places. The existence of the path was even seen as a comfort. It guides the way, it indicates that authorities thought enough of the resource to provide a path, and it protects nature by separating people from the river channel. Minor recreational amenity themes were appreciation of unpaved trails for hiking, and facilities such as water fountains and bathrooms. There is some evidence of preferences for ‘zoning’ recreational infrastructure depending on distance from urban centers. Nobody suggested removing the concrete path in photo C; if anything they wanted more amenities there. However only a few people suggested recreational developments for photo A. Photo A typically ranked first, people felt that it had an “untouched” and “pristine”

look. A minority were cautious regarding recreational developments anywhere. This was the especially strict version of keeping it “natural”. Whereas most people supported recreational amenities, there were strong negative associations with other infrastructure. Such infrastructure was criticized as visually or structurally deterring enjoyment of the river. In discussing her photo ranking, Sadie noted: Sadie (group 7): I would much rather look at the water and the riparian area surrounding it than I would the manmade structures. We separated codes referring to modifications to the channel from codes referring to modifications on the bank. Prevalent examples were the soil cement channel and the metal railing, respectively, apparent in photos C and D. These two themes were barely discussed before photos were distributed, thus we believe that moderator-introduced information significantly influenced these code counts. Urban development in general was also often discussed, and associated impacts on listed themes such as overextending the local water supply, reducing wildlife habitat, and worsening flood damages to human property.

4.7. Other people In addition to human modifications to waterways, people themselves were also important. There was a surprising amount of interest in waterways being recreationally useful for others. Many of these comments were in relation to children in particular. Vivian (stage 1 interviewee): I would like to see where kids could actually enjoy the water like in the old days. There are areas like this where really it isn’t accessible to people, where they could put in the reclaimed [water] and try to build up the vegetation and have wildlife come back and stuff like that. Conversely, there were also conflicts with other users. A relatively minor theme was other people interfering with solitude; the most cited conflict was a concern regarding personal safety from perceived threats of other people in the area. Concerns related to homeless people living in the dry washes were mentioned numerous times. Madrona (group 8): In the area where I live. . . there are no walking paths, and if you go down to the river all the homeless people have their homes. . .There needs to be some sort of security, because people complain and they go kick them out, but they come back the next day. They also have dogs that are running loose and I think there needs to be more stable security in the area.

4.8. Noise A surprise was a significant number of comments valuing rivers for offering a quiet, peaceful setting. Potential for noise pollution was a factor for how some rated photographs. As noted earlier, photo B typically rated last. However, people who rated it higher did so because it looked peaceful and showed an overall lack of human influence. Susanna (group 2): I would rather be somewhere where I’m going to have to bring my own bottled water, where I’m going to have to make my own shade, than be in an area where I know we’re being disrespectful to nature, where there’s noise pollution where there’s all these other things. . .

M.A. Weber, P.L. Ringold / Landscape and Urban Planning 133 (2015) 37–52

4.9. Comparison with prior research and transferability As stated in the introduction, while biophysical scientists have described the need to measure river features people value, there are few social science investigations of river monitoring. We know of only two prior studies involving social science research that consider a breadth of publicly important river and stream features; Schiller et al. (2001), and Ringold et al. (2009). The three studies differ markedly in scope, methods, and results. Schiller et al. used focus groups and interviews with non-scientist residents of the southern Appalachians to make lay translations of preexisting USEPA scientific indicators for streams, forests, and landscapes. Ringold et al. used an expert workshop to hypothesize indicators for rivers that would apply across the US, tailored for predefined archetypal human-interest categories such as recreation, industry, and agriculture. In the present study, like Ringold et al., we take the approach of buildings indicators “from the ground up” rather than translating preexisting indicators. Like Ringold et al. we have the relatively narrow scope of rivers and streams, but are narrower still with a focus on layperson residents of an urbanized, arid watershed in southern Arizona. Similar to Schiller et al. we employ empirical qualitative research techniques to engage layperson participants. Our study is the only one to include not only publicly important ecological features, but also human modifications to rivers and streams. In comparing results across the three studies, we limit it to ecological stream features. Overall we see many parallels between our themes and Ringold et al. results, although our findings are more specific than broadly worded Ringold et al. indicators. For example, under plants, for the recreational use of “viewing”, Ringold et al. list “Presence and abundance of visible plant taxae at all points at all times”. We are able to report that for this case study participants appreciated large trees and lush green vegetation in particular. Similarly, our themes under Fish and Wildlife are more specific than “charismatic fauna” cited by Ringold et al. Furthermore, we are able to offer a sense of relative priorities with code frequency results. In comparing with Schiller et al., it is important to keep in mind that the structure and scope of their results was constrained by the USEPA indicators they were translating. Thus their lay themes show how people interpret technical metrics, which is only indirect evidence that the metrics associated with those translations were of high importance. They list three lay interpretations of USEPA stream indicators: the chemical characteristics of stream water that help determine how water can be used by plants, animals, and people; the kind and number of living things, other than fish, in a stream; and the kind, number, and edibility of fish present in the stream. These have some similarity with our results. We also found much interest in human, plant, and animal usage of streams, but are able to list more specific types of human uses, plants, and animals of interest. We did find an interest in fish, but did not find strong interest in aquatic life otherwise. Regarding fish edibility, we received just one comment and thus did not include it as a theme. We have thirty-five ecological themes to their three, but this is partially explained by our participants thinking beyond the wetted stream. Major findings of ours such as flooding, vegetation, and wildlife appear in Schiller et al. results for forests and landscapes. Unconstrained by USEPA indicators we also have the possibility of finding, and indeed do find, additional themes emerging from public input, such as a deep concern for water supply quantity. At the same time there are also instances in which Schiller et al. results range beyond our limited rivers and streams scope, such as their indicator “Woodland productivity for forest products.” In general our results are worded in more detail, but again it was part of our purpose to give new monitoring insight while Schiller et al. already had USEPA indicators and were not seeking to modify them,

49

just investigating whether the public could understand them and valued them. Our study has strongest applicability as an in-depth investigation of a diverse sample of urban residents in an arid region of the US. We hypothesize that there would be interest in these themes and associated metrics particularly where sociodemographic and geographic conditions are similar. The entire southwestern US is characterized by the overarching factor of water scarcity, and urban population centers are common in the region. Many of the most frequently cited themes may be unique to the arid context, such as interest in the presence of surface water, the relationship between rivers and groundwater supply quantity, and appreciation of trees and other lush vegetation. There are also likely to be at least a few idiosyncratic findings. This case study is in the Sonoran Desert, which is characterized by strong episodic precipitation events. Flooding themes may not be as important in other southwestern locales. Another notable factor is that all perennial flow in the Santa Cruz River in Tucson is due to effluent, and introducing this background information to participants significantly affected the code count for human-caused odor concerns. Each case study will have unique circumstances affecting participant themes. We cannot disregard the possibility that urban residents elsewhere may have fundamentally different river and stream preferences, but we would expect the main themes to remain the same with some thematic variation stemming from the influence of changing environmental factors.

4.10. Comparison with existing monitoring practice Our research pertains to what information residents are interested in knowing about rivers. While this is a critical criterion, many more questions need to be answered to make the strategic decisions necessary to develop a monitoring program, such as how to select river sites for which to report data, the cost of different metrics, which features are vulnerable to change, and the frequency of data collection. While keeping in mind numerous factors that bear upon what data are ultimately collected, we offer a high-level comparison of our results with existing monitoring practice, focusing on themes occurring 18 or more times (the median code count). The agency with the most direct responsibility for stream monitoring in the state is the Arizona Department of Environmental Quality (ADEQ). ADEQ monitoring protocols are designed to protect specific “designated uses” within the three broad use categories of Aquatic and Wildlife, Human Health, and Agricultural (ADEQ, 2010: R1811-104). The first two use categories have strong accordance with many of our ecological themes. However ADEQ testing for water quality characteristics (such as dissolved chromium) that support the Aquatic and Wildlife use category does not mean that aquatic life and wildlife are actually found at the site, nor does ADEQ collect data on these final biological outcomes. ADEQ designated use standards for Human Health have direct relevance with metrics such as safe partial and full body contact. However since designated uses change from one river reach to the next, data is not readily available for every reach for every designated use. For example, the effluentdominated portions of the Santa Cruz River are not held to the full body contact standard. The vast majority of our participants lived in the city of Tucson, which is served by utility water providers, as is common in US urban areas. Such utilities have federally mandated water quality standards and rigorous testing of water quality. Adequate monitoring effort appears to be in place, however it is not unilaterally convincing to residents, evident in the prevalence of drinking water quality concerns. There are notices supplied by utilities which list each species and technical test results in a series of numbers (e.g. parts per million); these might be overly technical or difficult for

50

M.A. Weber, P.L. Ringold / Landscape and Urban Planning 133 (2015) 37–52

consumers to translate into actual health risks. Furthermore, there was evidence of continuing mistrust in connection to the transition to Central Arizona Project water. We also found several significant ecological themes outside the monitoring scope of ADEQ or public water utilities. At least some data for these are provided by additional entities. For example, the Arizona Department of Water Resources has the responsibility of tracking available water quantity in critical areas; FEMA does flood mapping in some areas, which identifies vulnerable property and roads; the USGS tracks flow heights and volumes where gages exist, and American Whitewater provides navigability information for a surprising number of river reaches. However potentially important data gaps remain. The presence of surface water was the most frequent theme of all. While maps do exist of springs and perennial waterways, we know of only limited efforts to track where surface flows are actively changing (Turner & Richter, 2011). Navigability details for the Santa Cruz River are lacking, despite the two effluent-dominated reaches being designated “navigable” by the US Army Corps of Engineers. For Vegetation, there are numerous monitoring efforts, many of which rely on remote sensing. For Fish and Wildlife, we do not know of reasonably comprehensive monitoring programs to match our metrics, although data do exist for some things in some locations. The Arizona Game and Fish Department, National Audubon Society, Friends of the Santa Cruz River, Sonoran Institute, and many other entities have important and relevant efforts. Yet when and where data are available, it is more common to assess habitat quality than the more difficult task of monitoring mobile organisms. For human themes, again we know of at least some monitoring for many of the metrics. For our study area there is information from either Pima County or the City of Tucson on recreational amenities, crime statistics by location, even odor near the wastewater treatment plants. However, the most frequent human theme of all, garbage and graffiti, does not appear to be systematically monitored in the study area (but it should be noted that there are periodic river clean-up efforts, often led by volunteers). Other data gaps are locations of high recreational use, locations of unsightly infrastructure, and locations of human-caused noise. For data that do exist, because there is public interest in so many varied river features, from so many different sources, there would appear to be much value in consolidating available information. It would be challenging for a layperson to gather existing data for all the metrics appearing in Table 3 for a given reach of river. The concept of “river report cards” is useful in this regard. We note that existing examples of river report cards often treat only a few metrics, and some of these metrics tend to be difficult for the public to directly interpret, such as pH or dissolved oxygen (Anacostia Watershed Society, n.d.; Milwaukie Riverkeeper, 2013; Sonoran Institute, 2011).

5. Conclusions Although there is a vast literature dealing with rivers and streams, there is relatively little investigation of specific features important to the public. In this paper we empirically document such features, and furthermore develop monitoring metrics easily translatable to field protocols. The results are based on in-depth interactions with 98 residents of southern Arizona. Participants were recruited from the general public to represent a wide range of sociodemographics. This research provides broad input at the foundation of resource management; defining what might be included as important to the public when assessing resource status and trends. Our main contribution is documenting themes of public interest related to specific river and stream features, as well as

a sense of relative priorities evident in code frequencies. Without such documented input, it would be difficult to judge whether river features considered in a given river management context are overly broad or overly narrow, and communication of management goals may be deficient. The measureable features society chooses to focus on have enormous weight when it comes time to consider the tradeoffs inevitable in resource management. Deciding which features to consider, given the numerous possibilities, is a significant accomplishment. There were several surprise insights from this research. We hadn’t expected to collect human themes, but these were recurrent and intertwined with ecological themes. The negative sentiment toward trash, constructed banks, and other visible defacement of rivers rivaled the importance of ecological factors; e.g., garbage and graffiti were mentioned almost as frequently as appreciation of large trees. Less obvious human features such as odor and noise were also significant themes. The presence of water was a foremost theme for our participants, whereas agencies tend to focus on measuring surface water quality or quantity. A sense of loss associated with rivers and streams in the case study area where there used to be more surface water was omnipresent. The strong connection people drew between river and stream conditions and their own water supply was remarkable, showing understanding of hydrological processes. Water security concerns of both quantity and quality were widespread, despite most participants being connected to centralized, closely regulated water utilities. Ecological themes went far beyond water, or even the active channel, reflecting public appreciation of the role riparian areas have for regional vegetation and wildlife. Interestingly, there was a generality to these preferences; affinity for particular species of flora or fauna was rare. Large trees were the favored vegetation. Participants did not seem to overly focus on specific fish and wildlife species (e.g. charismatic megafauna) despite repeated probes. The importance of avoiding extinctions was raised, but with a context of preserving ecological interconnections rather than to keep a certain species for its own sake. In fact, people often spoke of concern for “the environment” as a monolithic concept; it was challenging to disaggregate it into specific features. The stated desire for “healthy” and “natural” ecosystems was ubiquitous. People wanted to preserve inherently natural characteristics, even when they didn’t know or couldn’t express what those features were. We were surprised at the thoughtfulness participants expressed on river and stream issues. Only a handful of participants displayed minimal interest. Part of this attentiveness is accounted for by water supply being a salient issue in the region, however there was also significant attention to river and stream ecology. This is noteworthy given the highly impacted state of the most frequently discussed river, the urbanized reach of the Santa Cruz. Rather than being dismissive, people consistently expressed high interest, and described their desire to see improvements. At the same time, lack of basic river knowledge was sometimes apparent. Most of our participants were completely unaware of the nearby, perennial, publicly accessible, effluent dominated reach of the Santa Cruz River, even while often discussing ephemeral reaches just upstream. Our goal in this paper is to present a broad overview of the perspectives of residents of an urbanized, arid watershed on priority features of rivers and streams. With this scope, there are related residual topics. Input we received on the ecological theme categories of Water, Vegetation, and Fish and Wildlife is rich enough for a stand-alone exposition of each. Some individual themes beg follow-up research, such as preferences for native species and diversity, since these are prevalent issues in the ecological literature. Public engagement could be expanded to include more rural

M.A. Weber, P.L. Ringold / Landscape and Urban Planning 133 (2015) 37–52

participants, as well as distinct stakeholders such as industrial, agricultural, or environmentalist interests. In addition, public input on priority locations to collect river and stream data is needed, to complement our research question of what to measure at a given site. We speculate that locations nearest participant residences may be a priority, but there may be additional considerations such as more intensive monitoring at high quality sites. This qualitative research provides a foundation for quantitative research possibilities. We provide a detailed hypothesis of what residents of an urbanized arid watershed value that could be tested in larger sample survey research. Knowledge of what measureable features people value constrains river monitoring decisions at a broad, strategic level. With this “wish-list” in hand, additional considerations can be layered on, such as where to collect data, the cost of gathering different types of information, and an assessment of features vulnerable to near-term threats. The wish-list also supplies a guide for river information biophysical scientists could focus on when seeking to provide quantitative data to support river management choices. Partnering social science could then be anchored to the same river metrics, to assess public preferences for management scenarios through quantitative techniques such as environmental valuation.

Acknowledgments The research was subject to US Office of Management and Budget review, and US EPA Human Subjects review. Claire Zugmeyer (Sonoran Institute) provided photo C. David Morgan (Portland State University) and Melissa Cheyney (Oregon State University) contributed to the research design. Participants of the 2012 International Congress for Qualitative Inquiry Sustainability Panel assisted with early analysis. Leslie Richards (Oregon State University) provided a helpful review of the draft manuscript. Three anonymous reviewers improved the final paper. Any faults are due to the authors alone.

References Anacostia Watershed Society. (n.d.). 2011 State of the Anacostia River. Retrieved from http://www.anacostiaws.org/programs/publicaffairs/state-of-the-river Arizona Department of Environmental Quality. (2010). Title 18: Environmental quality. Arizona Department of Environmental Quality (Chapter 11: Department of environmental quality, Article 1: Water quality standards). Effective April 13, 2010. Retrieved from: http://water.epa.gov/scitech/swguidance/ standards/wqslibrary/upload/az wqs 1.pdf. Asakawa, S., Yoshida, K., & Yabe, K. (2004). Perceptions of urban stream corridors within the greenway system of Sapporo, Japan. Landscape and Urban Planning, 68, 167–182. Bark, R. H., Osgood, D. E., Colby, B. G., Katz, G., & Stromberg, J. (2009). Habitat preservation and restoration: Do homebuyers have preferences for quality habitat? Ecological Economics, 68, 1465–1475. Barmil, S., Daniel, T. C., & Hetherington, J. D. (1999). Human values and perceptions of water in arid landscapes. Landscape and Urban Planning, 44, 99–109. Berrens, R. P., Bohara, A. K., Silva, C. L., Brookshire, D., & McKee, M. (2000). Contingent values for New Mexico instream flows with test of scope, group-size reminder and temporal reliability. Journal of Environmental Management, 58, 73–90. Boyd, J., & Banzhaf, S. (2007). What are ecosystem services? The need for standardized environmental accounting units. Ecological Economics, 63, 616–626. Boyd, J. W., & Krupnick, A. J. (2013). Using ecological production theory to define and select environmental commodities for nonmarket valuation. Agricultural and Resource Economics Review, 42, 1–32. Colby, B. G., & Wishart, S. (2002). Quantifying the influence of desert riparian areas on residential property values. The Appraisal Journal, LXX(3), 304–308. Creswell, J., & Miller, D. (2000). Determining validity in qualitative inquiry. Theory Into Practice, 39(3), 124–130. Fern, E. F. (2001). Advanced focus group research. Thousand Oaks, CA: Sage. Freeman, A. M., III. (2003). The measurement of environmental and resource values (2nd ed.). Washington, D.C.: Resources for the Future. Gartin, M., Crona, B., Wutich, A., & Westerhoff, P. (2010). Urban ethnohydrology: Cultural knowledge of water quality and water management in a Desert City. Ecology and Society, 15(4), 36.

51

Glaser, B., & Strauss, A. (1967). The discovery of grounded theory. New York, NY: Aldine de Gruyter. Gobster, P. H., & Westphal, L. M. (2004). The human dimensions of urban greenways: Planning for recreation and related experiences. Landscape and Urban Planning, 68, 147–165. Hetherington, J., Daniel, T. C., & Brown, T. C. (1993). Is motion more important than it sounds? The medium of presentation in environment perception research. Journal of Environmental Psychology, 13, 283–291. Hughes, R. M. (1993). Stream indicator and design workshop. In Report no. EPA/600/R93/138. Environmental Protection Agency. Hughes, R. M., & Peck, D. V. (2008). Acquiring data for large aquatic resource surveys: The art of compromise among science, logistics, and reality. Journal of the North American Benthological Society, 27(4), 837–859. Jackson, L. E., Kurtz, J. C., & Fisher, W. S. (Eds.). (2000). Evaluation guidelines for ecological indicators. In EPA/620/R-99/005. Research Triangle Park, NC: Environmental Protection Agency, Office of Research and Development. Janis, I. L. (1972). Victims of groupthink: A psychological study of foreign-policy decisions and fiascoes. Boston, MA: Houghton Mifflin. Kaplowitz, M. D. (2000). Statistical analysis of sensitive topics in group and individual interviews. Quality & Quantity, 34, 419–431. Kondolf, G. M., & Yang, C.-N. (2008). Planning river restoration projects: Social and cultural dimensions. In S. Darby, & D. Sear (Eds.), River restoration: Managing the uncertainty in restoring physical habitat (p. 160). Chichester, West Sussex, England: John Wiley and Sons Ltd (Chapter 4). Krippendorff, K. (2013). Content analysis: An introduction to its methodology. Thousand Oaks, CA: Sage Publications. Ladle, R. J., & Gillson, L. (2009). The (im)balance of nature: A public perception timelag? Public Understanding of Science, 18, 229–242. Larson, E. K., & Perrings, C. (2013). The value of water-related amenities in an arid city: The case of the Phoenix metropolitan area. Landscape and Urban Planning, 109(1), 45–55. Logan, M. F. (2002). The lessening stream: An environmental history of the Santa Cruz River. Tucson, AZ: University of Arizona Press. Milwaukie Riverkeeper. (2013). 2012 river report card. Milwaukie Riverkeeper. Retrieved from http://www.mkeriverkeeper.org/content/2012-river-reportcard. Morgan, D. L., & Krueger, R. A. (1998). The focus group kit (6 volumes). Thousand Oaks, CA: Sage Publications. Naiman, R. J., Décamps, H., & Pollock, M. (1993). The role of riparian corridors in maintaining regional biodiversity. Ecological Applications, 3, 209–212. National Research Council. (2000). Ecological indicators for the nation. Washington, DC: National Academy Press. Norman, L., Tallent-Halsell, N., Labiosa, W., Weber, M., McCoy, A., Hirschboeck, K., et al. (2010). Developing an ecosystem services online decision support tool to assess the impacts of climate change and urban growth in the Santa Cruz watershed; where we live, work, and play. Sustainability, 2, 2044–2069. Patton, M. Q. (2002). Qualitative research and evaluation methods (3rd ed.). Thousand Oaks, CA: Sage Publications. Ringold, P. L., Boyd, J., Landers, D., & Weber, M. A. (2009). Report from the workshop on indicators of final ecosystem services for streams. In EPA/600/R-09/137. (Meeting date: July 13 to 16, 2009). Retrieved from http://www.epa.gov/ nheerl/arm/streameco/index.html. Ringold, P. L., Boyd, J., Landers, D., & Weber, M. A. (2013). What data should we collect? A framework for identifying indicators of ecosystem contributions to human well-being. Frontiers in Ecology and the Environment, 11, 98–105. Rubin, H. J., & Rubin, I. S. (2005). Qualitative interviewing (2nd ed.). Thousand Oaks, CA: Sage Publications. Schiller, A., Hunsaker, C. T., Kane, M. A., Wolfe, A. K., Dale, V. H., Suter, G. W., et al. (2001). Communicating ecological indicators to decision-makers and the public. Conservation Ecology, 5(1), 19. Seager, R., Ting, M., Held, I., Kushnir, Y., Lu, J., Vecchi, G., et al. (2007). Model projections on an inminent transition to a more arid climate in southwestern North America. Science, 316, 1181–1184. Smith, J. W., & Moore, R. L. (2011). Perceptions of community benefits from two wild and scenic rivers. Environmental Management, 47, 814–827. Sonoran Institute. (2011). A living river: Charting the health of the upper Santa Cruz river: 2010 water year. Sonoran Institute. Retrieved from http://sonoraninstitute.org/where-we-work/southwest/santa-cruz-river/311charting-the-health-of-the-upper-santa-cruz-river.html. Stirling, G., & Wilsey, B. (2001). Empirical relationships between species richness, evenness, and proportional diversity. The American Naturalist, 158(3), 286–299. Stoddard, J. L., Larsen, D. P., Hawkins, C. P., Johnson, R. K., & Norris, R. H. (2006). Setting expectations for the ecological condition of streams: The concept of reference condition. Ecological Applications, 16, 1267–1276. Turner, D. S., & Richter, H. E. (2011). Wet/dry mapping: Using citizen scientists to monitor the extent of perennial surface flow in dryland regions. Environmental Management, 47, 497–505. United Nations Environment Program. (1995). In V. H. Heywood (Ed.), Global biodiversity assessment (p. 1152). Cambridge, England: Cambridge University Press. United States Census. (2010a). 2010 Census urban and rural classification and urban area criteria. Retrieved from: (https://www.census.gov/geo/reference/ ua/urban-rural-2010.html) United States Census. (2010b). State and county quick facts. Retrieved from: (http://quickfacts.census.gov/qfd/index.html)

52

M.A. Weber, P.L. Ringold / Landscape and Urban Planning 133 (2015) 37–52

Webb, R. H., Leake, S. A., & Turner, R. M. (2007). The ribbon of green: Change in riparian vegetation in the southwestern United States. Tucson, AZ: The University of Arizona Press. Weber, M., & Stewart, S. (2009). Public valuation of river restoration options on the Middle Rio Grande. Restoration Ecology, 17(6), 762–771. Weber, M., Mozumder, P., & Berrens, R. P. (2012). Accounting for unobserved time-varying quality in recreation demand: An application to a Sonoran desert wilderness. Water Resources Research, 48(5), W05515.

White, M. S. (2011). Effluent-dominated waterways in the southwestern United States: Advancing water policy through ecological analysis. Arizona: Arizona State University (Ph.D. dissertation). Wutich, A., Lant, T., White, D. D., Larson, K. L., & Gartin, M. (2010). Comparing focus group and individual responses on sensitive topics: A study of water decision makers in a Desert City? Field Methods, 22(1), 88–110.