Scholarly article on topic 'Informal Rangeland Monitoring and Its Importance to Conservation in a U.S. Ranching Community'

Informal Rangeland Monitoring and Its Importance to Conservation in a U.S. Ranching Community Academic research paper on "Agriculture, forestry, and fisheries"

CC BY-NC-ND
0
0
Share paper
Academic journal
Rangeland Ecology & Management
OECD Field of science
Keywords
adaptive management / ethnoecology / knowledge integration / local knowledge / social–ecological system / traditional environmental knowledge

Abstract of research paper on Agriculture, forestry, and fisheries, author of scientific article — Steven R. Woods, George B. Ruyle

Abstract Effective natural resource management relies on accurate and timely information on the natural environment, which may be obtained by formal (“scientific”) or informal (“local” or “traditional”) methods. Formal monitoring methods are well documented and widely accepted among the rangeland science community, yet adoption by U.S. ranchers is inconsistent. In contrast, informal monitoring appears to be widely used by ranchers, but its practice and importance have rarely been documented or assessed. By interviewing ranchers and government agency personnel, we evaluated informal monitoring in and around the Altar Valley, Arizona, United States. Informal monitoring techniques included qualitative visual appraisals of forage quantity, indicator species and erosion, and incorporated local environmental history. The environmental knowledge embedded in informal monitoring was generally compatible with natural science. Informal monitoring was conducted continuously throughout the year and provided near real-time assessments that integrated observations of most land in individual pastures and ranches. In contrast, formal monitoring was generally performed only once per year, in a limited number of areas and with a delay of a few months between observation and completion of analysis. Thus informal monitoring had higher spatial coverage and temporal resolution and provided assessments faster than formal monitoring. Consequently, ranchers generally considered informal monitoring to be more relevant than formal monitoring to formulating yearly grazing plans and responding rapidly to unpredictable changes in the natural environment. Ranchers incorporated informal monitoring into assessments of rangeland trends and outcomes of conservation measures and thereby into choices of grazing system and planning of brush management and erosion control. Thus informal monitoring was foundational to long-term conservation, annual rangeland management planning, and adaptive natural resource management on subyearly timescales. If informal monitoring is of comparable utility in other rural communities, it would appear advantageous to document and evaluate informal approaches and to incorporate them into formal conservation planning.

Academic research paper on topic "Informal Rangeland Monitoring and Its Importance to Conservation in a U.S. Ranching Community"

ELSEVIER

Contents lists available at ScienceDirect

Rangeland Ecology & Management

journal homepage: http://www.elsevier.com/locate/rama

Informal Rangeland Monitoring and Its Importance to Conservation in a U.S. Ranching Community^

Steven R. Woods a'*, George B. Ruyle b

a Adjunct Assistant Professor School of Natural Resources and the Environment, University of Arizona, Tucson, AZ 85721, USA b Professor, School of Natural Resources and the Environment, University of Arizona, Tucson, AZ 85721, USA

ARTICLE INFO

ABSTRACT

Article history: Received 25 July 2014 Accepted 1 July 2015

Key words:

adaptive management

ethnoecology

knowledge integration

local knowledge

social-ecological system

traditional environmental knowledge

Effective natural resource management relies on accurate and timely information on the natural environment, which may be obtained by formal ("scientific") or informal ("local" or "traditional") methods. Formal monitoring methods are well documented and widely accepted among the rangeland science community, yet adoption by U.S. ranchers is inconsistent. In contrast, informal monitoring appears to be widely used by ranchers, but its practice and importance have rarely been documented or assessed. By interviewing ranchers and government agency personnel, we evaluated informal monitoring in and around the Altar Valley, Arizona, United States. Informal monitoring techniques included qualitative visual appraisals of forage quantity, indicator species and erosion, and incorporated local environmental history. The environmental knowledge embedded in informal monitoring was generally compatible with natural science. Informal monitoring was conducted continuously throughout the year and provided near real-time assessments that integrated observations of most land in individual pastures and ranches. In contrast, formal monitoring was generally performed only once per year, in a limited number of areas and with a delay of a few months between observation and completion of analysis. Thus informal monitoring had higher spatial coverage and temporal resolution and provided assessments faster than formal monitoring. Consequently, ranchers generally considered informal monitoring to be more relevant than formal monitoring to formulating yearly grazing plans and responding rapidly to unpredictable changes in the natural environment. Ranchers incorporated informal monitoring into assessments of rangeland trends and outcomes of conservation measures and thereby into choices of grazing system and planning of brush management and erosion control. Thus informal monitoring was foundational to long-term conservation, annual rangeland management planning, and adaptive natural resource management on subyearly timescales. If informal monitoring is of comparable utility in other rural communities, it would appear advantageous to document and evaluate informal approaches and to incorporate them into formal conservation planning.

© 2015 The Authors. Published by Elsevier Inc. on behalf of Society for Range Management. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Introduction

Rangeland monitoring is foundational to informed, deliberate management of rangelands (Elzinga et al., 1998; Holechek et al., 2004). Monitoring enables pasture and livestock management decisions to account for the condition of land, its plants, animals and soils, and their responses to human activity and the wider natural environment. Of the various techniques available, formal ecological monitoring is well understood (Coulloudon et al., 1999a; Holechek et al., 2004; Lindenmayer and Likens, 2010), and informal or traditional methods have often been studied in nonindustrialized or indigenous societies (Berkes et al., 2000; Thornton and Scheer, 2012). Informal environmental knowledge and monitoring in industrialized societies are less commonly studied

☆ This research was funded by grant GW10-034 from the USDA Western Sustainable Agriculture Research and Education Program.

* Correspondence: Dr. Steven Woods, School of Natural Resources and the Environment, University of Arizona, Tucson, AZ 85721, USA.

E-mail address: srwoods@email.arizona.edu (S.R. Woods).

but can play an important role in natural resource management (Ballard et al., 2008; Meuret and Provenza, 2015; Millar and Curtis, 1999). Informal rangeland monitoring appears widespread among U.S. ranchers, but its practice, uses, and value have rarely been documented or assessed (Knapp and Fernandez-Gimenez, 2008, 2009; Sayre, 2004).

For our purposes, formal monitoring is defined as standardized procedures based in the scientific method and widely accepted and used among natural resource management professionals in academia and government agencies (Raymond et al., 2010). Procedures are well documented, consistently repeatable and, usually, quantitative and amenable to statistical analyses, thus minimizing bias and dependence on place or practitioner (Ruggiero, 2009). Formal rangeland monitoring methods are developed, practiced, and promoted by, among others, the academic community and by U.S. federal agencies within the Department of Agriculture and the Department of the Interior (Coulloudon et al., 1999a; Holechek et al., 2004; Lindenmayer and Likens, 2010; USDA-NRCS, 2003; USFWS, 1980). Formal monitoring can be effective in assessing and improving natural resource management, though efficacy is not guaranteed in all circumstances and cost can be prohibitive

http://dx.doi.org/10.1016/j.rama.2015.07.005

1550-7424/© 2015 The Authors. Published by Elsevier Inc. on behalf of Society for Range Management. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

(Elzinga et al., 1998; Lindenmayer and Likens, 2010). Published data on adoption by U.S. ranchers are scarce, but available data suggest it is used on approximately half the livestock ranches in Arizona (Fernandez-Gimenez et al., 2005; Peterson, 2010).

In contrast, informal monitoring is nonstandardized, relies on personal practice and experience, is typically embedded in local cultural and natural environments (Raymond et al., 2010), and is practiced on approximately 95% of ranches in Arizona (Peterson, 2010). Informal knowledge of the natural environment may be localized and may or may not be compatible with natural science (Ellis, 2005; Sillitoe etal., 2004), and informal monitoring procedures typically do not conform to the scientific method (Raymond et al., 2010). The degree of compatibility between informal knowledge and natural science is variable and should not be assumed (Raymond et al., 2010; Tibby et al., 2008). Thus informal monitoring is vulnerable to charges of practitioner bias and unreliability, and is typically not officially sanctioned by government agencies (Ruggiero, 2009).

Informal or traditional monitoring can, however, have advantages over formal monitoring, including greater effective sample sizes (in a broad sense, in terms of numbers of plants, animals, and areas observed), longer duration and greater frequency of observation, integration of greater variety of observations, and lower cost (Moller et al., 2004), attributes that can increase the effectiveness of monitoring as a tool for understanding ecological change and its causes (Elzinga et al., 1998; Herrick et al., 2006; Lindenmayer and Likens, 2009; Vaughan et al., 2001). Thus informal and formal monitoring may be to some extent complementary, and incorporation of both systems may improve the management of natural resources (Reed et al., 2013).

If informal and formal methods are to be integrated systematically, informal methods must first be clearly identified and their validity or compatibility with formal methods assessed (Raymond et al., 2010; Sillitoe et al., 2004). It is common to equate the validity of traditional environmental knowledge with its degree of conformance to formal science (Ellis, 2005; German, 2010). However, this approach is controversial because disagreements between the two knowledge systems can be due to inadequacies of formal science (Ellis, 2005; Fairhead and Scoones, 2005). There can also be disagreements within natural science. For instance, formal field monitoring can correspond more closely to informal monitoring than to remote sensing assessments of rangelands (Herrmann et al., 2014; Kong et al., 2015). Therefore bias in favor of either informal or formal monitoring should be minimized, for example, by assessing compatibility between the two systems rather than treating one system as a benchmark or standard reference (Ellis, 2005). Then, apparent contradictions between informal and formal knowledge would represent opportunities to re-evaluate and refine both sets of observations and conclusions, and thereby improve or correct either or both of them.

Formal rangeland monitoring methods are typically evaluated and selected with reference to their purpose or application, whether in ecological research or natural resource management (Elzinga et al., 1998; Lindenmayer et al., 2011). We suggest that informal monitoring should be similarly evaluated in the context of its uses. Informal monitoring typically varies between practitioners and regions (Raymond et al., 2010). In this study our objective was to document the informal rangeland monitoring practiced in one ranching community and compare it with formal monitoring and natural science. We used qualitative methods (Patton, 2002; Sayre, 2004) to gain detailed, in-depth understanding of informal rangeland monitoring practices, the application of informal and formal monitoring to rangeland management, and the perspectives of participants on their utility. We compare informal and formal monitoring in the study area, and compare informal monitoring with published literature on formal monitoring and natural science. Thereby, we assess the compatibility and complementarity of the two monitoring systems and their utility to local rangeland management and conservation. We submit that such description and analysis are necessary if we are to determine whether it is feasible and meritorious to integrate the two methodologies or their outcomes (Ellis, 2005; Huntington, 1998).

Methods

Biophysical Setting

The study area comprises the Altar Valley and adjacent rangelands in the Santa Cruz Valley in Pima and Santa Cruz counties, Arizona, United States. The area lies west of 111.1°W and south of 32.0°N; borders the Schuk Toak, Baboquivari, and Chukut Kuk districts of the Tohono O'Odham Nation to the west and Mexico to the south; and totals approximately 2 300 km2 (900 square miles). Landforms include mountains of up to 2 350 m elevation, pediments, alluvial fans, and a floodplain down to 750 m elevation (Andrews, 1937; Sayre, 2007). Mean annual precipitation varies with elevation and ranges from 300-650 mm (NOAA-NCDC, 2012). Peak precipitation occurs between July and September in the monsoon season, with a smaller peak in winter and a pronounced spring dry season. Mean daily temperature ranges from 4-10°C in January and 21-32°C in July (NOAA-NCDC, 2002).

Vegetation communities vary from Quercus-Pinus (oak-pine) and oak savanna at higher elevations to herbaceous and wooded riparian areas along principal channels, with desert scrub and semidesert grassland and savanna being the predominant rangeland types (Meyer, 2000; Strittholt et al., 2012). Grasslands in the region have changed considerably since the early 20th century. Many are now dominated by the nonnative Lehmann lovegrass (Eragrostis lehmanniana Nees.), while others have undergone considerable encroachment by native woody plants, particularly velvet mesquite (Prosopis velutina Woot.). Other common woody plants include the shrubs catclaw acacia (Acacia greggii A. Gray.), paloverde (Parkinsonia spp.), ocotillo (Fouquieria splendens Engelm.), and prickly pear cactus (Opuntia spp.), and the subshrubs huajillo (or fairyduster, Calliandra eriophylla Benth.), burroweed (Isocoma tenuisecta Greene), and snakeweed (Gutierrezia spp.). Common native grasses include threeawns (Aristida spp.), sideoats grama (Bouteloua curtipendula [Michx.] Torr.), tanglehead (Heteropogon contortus [L.] P. Beauv. ex Roem. & Schult.), and sacaton grass (Sporobolus wrightii).

Social Context

The study area is predominantly rural and includes the small town of Arivaca. Cattle ranching has been a major industry since the early 19th century (Sheridan, 1995) and is currently the most extensive land use. The majority of rangelands in the area are Arizona State Trust lands, administered by Arizona State Land Department (ASLD) and leased to ranchers for livestock grazing (USDA-NRCS et al., 2008). Private ranches and the Buenos Aires National Wildlife Refuge (BANWR) account for most of the remaining area, with smaller holdings under the jurisdiction of the U.S. Department of Agriculture (USDA) Forest Service (USFS), Pima County and the U.S. Department of Interior Bureau of Land Management (BLM). Most livestock ranches utilize a combination of State Trust land and private land, with some also grazing USFS, BLM, or county lands. The 12 largest ranches in the Altar Valley cover approximately 125 km2 on average, including both private and public land (Sayre, 2007). BANWR is administered by the US Fish and Wildlife Service (USFWS) and is closed to livestock grazing. Other government agencies involved in local rangeland management include the USDA Natural Resources Conservation Service (NRCS) and Arizona Game and Fish Department (AGFD). The Altar Valley Conservation Alliance (AVCA) provides a forum for members of the local ranching community, NGOs, representatives of government agencies and others who aim to cooperate in rangeland conservation.

Data Collection

We conducted 28 semistructured, conversational interviews (Wilson and Sapsford, 2006) with 27 participants between February 2010 and January 2011. Interviews were semidirective, allowing both participant and interviewer to cooperatively direct the interview into

the areas they felt were important (Huntington, 1998). Each participant was personally involved in monitoring and management of rangelands in the study area. The participants included 14 ranchers from nine ranches, representing approximately half of the ranches in the study area, and 13 government personnel representing the principal agencies involved in rangeland management in the area (AGFD, ASLD, BLM, NRCS, Pima County, USFWS, and USFS), one of whom was concurrently a rancher. These sample sizes are typically sufficient to determine common issues or practices through semistructured interviews (Guest, 2014). The ranches included one guest ranch and one principally growing fruit. The remainder raised cattle, in common with most of the other ranches in the study area, and all ranches had horses. All ranchers had been involved in ranching for at least 7 years and the majority for > 40 years. Five had family ranching histories in the study area going back two, three, or four generations. Another six were third-, fourth-, or fifth-generation ranchers with family ranching histories both in the study area and elsewhere in southeast Arizona and, in two cases, also elsewhere in the western United States.

We used multiple methods to identify potential participants in order to minimize the influence of any sampling bias that any one method may introduce. Potential participants were identified by a combination of networking through local community meetings held by the AVCA and University of Arizona contacts, snowball sampling (Patton, 2002), and the use of maps and Internet searches to identify ranches in the study area. Participants were contacted initially either in person at community meetings or by phone. Where we could not find a telephone number for a ranch we did not cold call them in person, without an introduction or appointment, in case this jeopardized the rapport that is essential between interviewer and interviewee (Berg and Lune, 1995). Of those ranchers and agency personnel we contacted, none refused to be interviewed.

Three of the interviews had two participants, each pair being ranchers who were husband and wife. The remaining 25 interviews were with one participant each. Multiple interviews were conducted with two ranchers, and these included five go-along interviews (Evans and Jones, 2011; Jones et al., 2008) conducted while traveling through and observing rangelands. The go-along interviews constituted "place-based" discussion of the natural environment within the ranches under consideration and enabled participants to use features of the vegetation and soils to prompt and illustrate their explanations. Three of the go-along interviews also included observation of ranch and livestock management discussions between participating ranchers. The numbers of participants during these observation periods were 2, 3, and 4. All interviews were conducted in a place of the interviewee's choosing (Mack et al., 2005). We gave each participant the choice of where to be interviewed to minimize the inconvenience to them and ensure they were comfortable with the circumstances of their interviews. Ranchers were interviewed in their homes, elsewhere on their ranch, or in their offices. Agency personnel were interviewed in their offices or, in one case, in a restaurant.

Each interview lasted between 40 min and 3 h. Primary topics of the interviews were the same for ranchers and agency personnel. They included informal and formal rangeland monitoring, as well as rangeland management practices and decisions. Participants' concerns and aims regarding the natural environment were elicited to enable interviews to begin with areas participants considered most relevant. Three preliminary focus groups had been conducted independently of this study with a total of 27 experienced Arizona ranchers in 2009, and had identified brush encroachment and soil erosion as areas of common concern. These topics were discussed with each participant in this study to ensure discussions were grounded in practical conservation issues of widespread interest.

Precise questions and the ordering of interview topics varied between interviews. This allowed the interviewer to follow up on comments made by participants, either to understand the current topic in more detail or to lead into another topic. It also enabled

participants to offer information they felt relevant to the discussion even if the interviewer had not prepared a precise, focused question on the matter. These techniques were intended to generate a greater depth of understanding than would have been attained through fully structured interviews and to reveal relevant aspects of the interview topics of which the interviewer had hitherto been unaware (Huntington, 1998; Wilson and Sapsford, 2006). The variation in the precise questions used in different interviews meant that not all participants addressed every detail presented in the results. In these cases we have avoided giving the proportion of participants who held a view or used a particular technique, to avoid implying that the remainder did not.

Notes were taken during all interviews. Digital audio recordings of interviews were also made except when impracticable during go-along interviews. This study was approved by the University of Arizona Institutional Review Board (Human Subjects Protection Program, project number 09-1135-02).

Analysis

Audio recordings were transcribed, and S. R. Woods coded the transcriptions and notes in NVivo 8 (Berg and Lune, 1995; Fisher, 1997; QSR, 2008). Initial codes were based on the a priori themes encapsulated in the preprepared interview topics. Emergent subthemes and relationships between themes were added to the code book during preliminary coding of the first eight interview transcripts and the notes from two go-along interviews. The resulting code book was then used to code all transcripts and notes. Coded instances of themes were summarized, tabulated, and compared to determine and characterize common methods of monitoring, their uses, and participants' views of their utility.

Compatibility between a) ranchers' informal rangeland monitoring, its uses, and its embedded environmental knowledge and b) formal monitoring, its uses, and natural science was assessed by a search of published literature based on formal natural scientific methods (Reed et al., 2008). This literature included articles in refereed ecology and earth science journals, books and book chapters explicitly based on such articles, and formal monitoring manuals published by government agencies. Informal monitoring was also compared with formal monitoring techniques commonly used for similar rangeland management purposes in the study area.

Results

informal Monitoring by Ranchers

All livestock ranchers made informal, ocular estimates of forage abundance and condition and of precipitation and its effects on vegetation. Most (93%) also reported monitoring signs of soil erosion and deposition, and most (93%) said they made judgments of overall rangeland condition. Livestock ranchers generally considered forage abundance the most critical feature of the natural environment to assess. One rancher said:

1 look at forage plants. That's the first thing that gets your attention, but you also look at the condition of the mesquite trees and whether you're getting any beans, and you look at the health of the general landscape. I mean when you're looking at it all the time it doesn't take much to recognize that, yeah, it's getting tougher and tougher.

Descriptors were generally qualitative rather than numeric (Table 1). Spatial scales of monitoring varied according to the sizes of features being monitored and the degree of patchiness encountered. Observations were contextualized within the characteristics of individual areas of a ranch, including topography, historical rangeland conditions and management practices, and infrastructure. Thus an individual gully may be described as showing signs of recent erosion; a small portion of a hillside as being in poor condition; or a whole pasture as recovering from wildfire with good, fresh grass growth but still a lot of bare ground.

Comparison of informal and formal rangeland monitoring, based on interviews with 20 ranchers and government agency personnel who described using both methodologies. Views of the advantages and disadvantages of each methodology were similar among this group of participants and were not contradicted by any of the 17 participants who used only one monitoring methodology. All participants were involved in rangeland management in the Altar Valley or neighboring ranches in southeast Arizona, USA.

Property Informal monitoring Formal monitoring

Rangeland observations Descriptors/variables Qualitative Usually numeric

Time needed for observation Short Long (e.g., 2 transects per day)

Frequency of observations Daily or near-daily Yearly

History of similar, prior observations in situ From a few years to over a century From 1 yr to a few decades

Areal extent of observations Whole pastures and whole ranches 1 or 2 small areas per pasture, in some or all pastures in a ranch

Level of detail Variable High

Dependence on observer High Low

Standardization Low High

Documentation Usually little or none Comprehensive

Time to availability Typically the same day as observation 2-3 mo after observation

Results and Analyses Documentation Occasional; nonstandardized Always; standardized

Repeatability and reliability Uncertain, due to dependence on observer High, due to reliance on recorded numbers and standardized methods

Comparability between years Uncertain, due to dependence on memory and oral history High, due to reliance on recorded numbers and standardized methods

Uses Long-term planning of pasture management1 Yes Yes

Yearly pasture and livestock plans Yes Rarely2

Adaptive rangeland management on subyearly timescales Yes No

Planning erosion control and brush suppression treatments Yes Yes

Managing wildlife and endangered species populations No Yes

Evidence of compliance with environmental laws No Yes

1 For example, choice of grazing system or rest-rotation pattern.

2 In all but one ranch, analyses of the most recent formal monitoring were not available before yearly management plans were made, as formal monitoring and most ranch's yearly planning was conducted at the same time of year, and formal analysis would take 2-3 mo before availability.

The areas directly monitored informally constituted most or all pasture land within each ranch.

All ranchers compared current rangeland conditions with personal recollection of conditions and land management practices in previous years, and most ranchers also incorporated orally transmitted local environmental history from previous generations. For instance, some ranchers used the highest forage production they had seen in a given pasture as a benchmark from which to gauge the current condition of that pasture, and most ranchers observed change over multiple years to assess long-term trends and outcomes of rangeland conservation practices such as brush suppression and erosion control. Often, observations of change in one location over multiple years would be integrated with comparisons between that location and others nearby. Most ranchers had informal photographs of their ranch from 20 or more years previously, which they could compare with current conditions, and one rancher occasionally made written notes of his informal observations.

Most ranchers (64%) incorporated observations of indicator plant species into rangeland assessments. Indicator species have long been used by ecologists to help assess the condition of natural environments (Niemi and McDonald, 2004). Generally, ranchers' use of indicator species appeared compatible with ecological literature (Table 2). For example, rangeland trends can be signaled by changes in abundance of unpalatable herbaceous species, which in some areas include tanglehead (H. contortus) and threeawns (Aristida spp.) (Cable and Martin, 1975; Canfield, 1948). Ranchers interpreted the abundance or condition of indicator species in the context of other environmental conditions. As one rancher said:

It depends on which end of the utilization you are at. If you're in a pasture that you've been kind of resting after heavy grazing, if you've got some good rainfall and you're seeing an expansion of tanglehead, that would be a good sign. Not that you're going for tanglehead as your climax, but it means that your range is going in the right direction. When you get in bad shape, your least desirable species are always the ones that grow fastest, tanglehead and the threeawns. So that's your first indicator that things are getting better, the country's getting better.

Those ranchers who personally managed livestock and maintained pasture infrastructure observed the natural environment while

performing such work. Thus they monitored informally on a near-daily basis (Table 1). The ranchers who employed ranch hands to do most of the day-to-day work in the field used the observations reported to them by their employees and accompanied them to jointly observe their rangelands when a notable change was reported. In all cases, ranchers informally monitored their rangelands personally on at least a weekly basis.

The remainder of this section details the main components of ranchers' informal monitoring. Each description is followed by a comparison with literature based on natural science.

Forage

All livestock ranchers assessed forage quantity visually in terms of bulk, mass, or more generally quantity of vegetation. The primary aims for ranchers were to assess the amount of forage at the time of monitoring and use that as a basis to anticipate the amount that would likely be available to livestock during ensuing seasons, particularly to the end of the dry season when forage availability is typically at its annual minimum.

Informal methods of estimating forage amount had similarities with the formal method of double sampling, wherein biomass estimates are based on ocular estimates (Elzinga et al., 1998). However, in formal double sampling, ocular estimates of standing crop are numeric and are calibrated by weighing dry vegetation mass of subsamples. In contrast, informal estimates were qualitative and were calibrated against prior informal assessments of forage abundance and livestock carrying capacity.

Composition. Informal assessments of forage quantity accounted for vegetation composition in terms of quantities or proportions of native versus non-native grasses, perennial versus annual grasses, perceived nutritional value and palatability of different species, and plant vigor and growth stage. Different categories or species of plant were assessed or implicitly weighted according to perceived forage value.

The principal vegetation category assessed for overall forage quantity was grasses for 13 ranchers and woody plant leaves for one cattle rancher. Perennial grasses were considered to retain more nutrition longer into the dry season than annuals. Thus annual grasses were included in

Table 2

Indicator species used by ranchers in rangeland monitoring in the Altar Valley and neighboring areas of southeastern Arizona, USA.

Common name (Species)

Life form

Observation

Inference

Rancher explanations

Level of agreement with natural science literature

Velvet mesquite and huajillo Shrub and Curled up leaves

(Prosopis velutina and Calliandra eriophylla)

Ocotillo (Fouquieria splendens)

Burroweed and snakeweed (Isocoma tenuisecta and Gutierrezia spp.)

Prickly pear (Opuntia Mill.)

Tanglehead and threeawn (Haemonchus contortus and Aristida spp.)

subshrub, respectively

Green leaves

Yellow leaves Absence of leaves

Little or no rainfall over the previous few weeks

Very recent rainfall

Rainfall has recently ceased No recent rainfall

Subshrubs High abundance in spring High winter rainfall

Cactus Plump, green, pads

Thin, yellow or purple pads Grasses High utilization

Numerous healthy stands over a wide area, after drought or heavy grazing Numerous stands over a wide area, after high rainfall

Recent rainfall

Lack of recent rainfall Overall forage utilization has been high, and more palatable species have been grazed heavily. Rangeland health is improving, and other native perennial grasses are likely to increase. Rangeland health is deteriorating.

Deep roots can delay and reduce effects of low rainfall on mesquite and huajillo.

Ocotillo produces and loses leaves rapidly in response to moisture levels, and ocotillo leaves can be assessed from greater distance than grasses due to ocotillo's greater height.

Wet winters benefit deeply rooted plants.

Color depends on prickly pear variety and individual plant health, as well as on water status.

Tanglehead and threeawns and are less palatable to livestock than most other locally common native grasses, increase in abundance rapidly in response to favorable growing conditions, and are persistent under unfavorable conditions.

High1 with mesquite (Pasiecznik etal., 2001; Phillips, 1963; Ryel et al., 2008), uncertain with huajillo2 (Burgess, 1995) High1 (Kozlowski, 1976; Nobel and Zutta, 2005; White et al., 2006)

High13 (Burgess, 1995; Cable, 1967,1969; Ralphs and McDaniel, 2010; Weaver, 1958) Moderate4 (Knipling, 1970; Stintzing et al., 2001)

Moderate5 (Burgess, 1995; Cable and Martin, 1975; Canfield, 1948; USDA-NRCS, 2003)

Note: Most ranchers reported assessing mesquite and/or huajillo leaves. Use of each other indicator species was reported by a minority of the ranchers interviewed.

1 Highly similar relationships are found in natural scientific literature.

2 Insufficient information was found on huajillo to evaluate the level of agreement with natural science.

3 Snakeweed is not deeply rooted but, like burrowed, responds to winter precipitation as described by ranchers.

4 The natural science literature describes similar relationships for vascular plants in general and for prickly pear fruit, but we found no formal studies on prickly pear pad (i.e., cladode) color changes due to water stress.

5 Effects of heavy grazing on the relative abundance of grass species can vary greatly between sites and regions, but in southeast Arizona, tanglehead and threeawns can show the patterns described by ranchers. In general, herbaceous species of low palatability often respond to improved conditions more rapidly than palatable species.

assessments of current forage but tended to be discounted when assessing the amount of forage that would be available through the dry season. Large stands of unpalatable plants were excluded from forage quantity assessments.

Native grasses were generally considered to be of higher nutritional value than non-natives. For instance, assessments of forage quantity generally accounted for the lower nutritional value for cattle of Lehmann lovegrass compared with most native grass species, and for the relatively high volume but low mass of mature Lehmann lovegrass plants due to their open, diffuse growth form. However, young, green Lehmann lovegrass shoots were considered palatable to cattle and would often appear before native grass shoots when forage availability is typically at a yearly minimum and in response to winter rains when native warm season perennial grasses can be relatively unpalatable. Thus early in the monsoon and in winter, young, green Lehmann lovegrass shoots could form a significant proportion of forage assessments for cattle. In addition, some ranchers considered mature Lehmann lovegrass to be palatable to horses and incorporated abundance of all growth stages of Lehmann lovegrass into forage assessments throughout the year.

Most livestock ranchers (92%) assessed amounts of leguminous shrub leaves and seed pods, particularly during the dry season when herbaceous vegetation is scarce and of low quality. Five ranchers considered leaves of the subshrub huajillo to be highly palatable to cattle and an important component of forage. Three ranchers considered mesquite leaves to be an important and valuable component of cattle forage during all seasons. Two of these ranchers estimated that cattle consumed approximately 10% of their diet from mesquite leaves even in the presence of abundant native grass feed.

Abundance of palatable and unpalatable plants was used to assess general range condition. For example, the more palatable native grass

there was in a pasture, the better its condition was considered to be. For some ranchers, unpalatable species were also used as indicators of trends in general rangeland condition (Table 2).

Informal categories of vegetation and ranchers' views on their palat-ability and nutritional content were generally consistent with those described in the formal literature and used in formal assessments of rangeland condition (Coulloudon et al., 1999a; Holechek et al., 2004; Ruyle, 2003). Vegetation composition, in terms of numeric proportions of species and plant types by weight, density, or cover, has been used extensively to formally describe ecological sites and evaluate rangeland condition (Coulloudon et al., 1999a). Ranchers' categorization of nutritional value to livestock largely corresponded to formal classifications. For example, grazers such as cattle generally prefer grasses to forbs and shrub leaves, but woody plant leaves can constitute a high proportion of their food intake on some ranges; and the nutritional value of shoots and leaves peaks early in the growing season and declines markedly after the growing season, with this decline being more pronounced in annual than perennial grasses (Holechek et al., 2004; Ruyle, 2003). Diets of cattle vary between ranches and seasons in southeast Arizona (Ogden, 2003), with mesquite consumption typically less than 10% in March and April and often exceeding 20% in other months.

Vigor and Color. Individual plant vigor was based on overall size of herbaceous plants, color of leaves (and also culms of grasses), and in one case also by grass leaf width. Growth stage would be assessed by size, color, and developmental features such as presence of inflorescences or seeds. Very young shoots were considered more nutritious and greener than mature shoots. Grass shoots were considered to become less green, more yellow, and less nutritious and palatable during dry periods within or following the growing season. After subsequent rainfall, shoots would become more green and more palatable and nutritious.

Color was also used to help assess grass species composition without close anatomic study. One rancher stated:

They all have their own color. You can look across a field or across a

hillside or a pasture, and say, 'over there is threeawn', because it has

a certain look to it, or 'over there, that's sideoats' or 'that's sacaton'.

Sideoats is bluer, when it's healthy and growing.

Thus ranchers recognized that color varied according to plant species, vigor, water status, and phenological stage and that color can be an indicator of palatability and nutritional value.

Although color can aid formal plant species identification (Elzinga et al., 1998), it is typically absent from formal field-based monitoring protocols (Coulloudon et al., 1999a; Habich, 2001). Objective ocular assessments of vegetation color are typically considered problematic in the formal literature (Kent, 2011). However, stress such as drought can alter plant color at visible wavelengths, generally decreasing reflectance of green light (Jackson et al., 1983; Knipling, 1970). Consequently, remote and in-situ measurement of visible light spectra can be used to assess primary production and growth stage, although near infrared wavelengths are typically also incorporated in remote sensing assessments of vegetation (Coppin et al., 2004; Kurc and Benton, 2010; Tucker, 1979; Tucker and Sellers, 1986).

Utilization. More than half of the ranchers reported making ocular estimates of forage utilization. Most described this as how "hard" livestock had grazed grasses and other forage plants. This appeared to be a combination of assessing changes in height, overall volume, and typical individual size of grass in a pasture. Two ranchers compared grazed areas with livestock exclosures set up for formal monitoring.

Utilization is often assessed in formal rangeland monitoring (USDA-NRCS, 2003). Visually assessing height of forage is comparable with the formal stubble height method (Coulloudon et al., 1999b). However, ranchers appeared to be comparing current grass height with recollections of height either earlier in the growing season or in previous years, whereas the stubble height method requires numeric heights to be set for specific plant communities. Visually assessing the general appearance of utilization is similar to the formal landscape appearance method (Coulloudon et al., 1999b). However, this method is considered prone to high variation between observers. Visually comparing forage utilization within and outside an exclosure is highly similar to the formal paired plot method. As applied by two of the ranchers in this study, the method of observation may be considered largely formal and their mental, nonsta-tistical method of analysis informal.

Recommended maximum utilization levels for key forage species vary between plant communities but in semiarid regions can be 40% to 50% (Holechek, 1988). Ranchers generally aimed to limit forage utilization to a similar extent. Two took a "take half, leave half" approach. One of them said, "It's a visual thing. You just get a feel for, you've taken about half, it's time to move." Another rancher considered 50% utilization a maximum. In all cases, high utilization was considered a strong reason to move livestock to a different area.

Precipitation

Ranchers assessed rainfall levels qualitatively by combinations of observations of its effects on vegetation, soil surfaces, water channels and stock ponds, and by direct observation during storms. Most ranches had at least one rain gauge, and three had 10 or more distributed throughout the ranch. Three ranchers reported that they would inventory their ranches immediately after storms to determine where rain had fallen, observe its effects, and check rain gauge levels.

Most ranchers used informal, visual observations of herbaceous vegetation to judge recent levels of precipitation and its spatial patchiness. As one rancher stated, "In the summer it can be even as obvious as just a green patch in a bunch of dry, and you'll see where a shower came through. It's very localized." This is in accord with published literature.

In semiarid regions, vegetation can green up rapidly in response to relatively large precipitation pulses following drought, and green-up can be spatially patchy, corresponding to rainfall patterns (Kurc and Benton, 2010; Pennington and Collins, 2007).

Most ranchers reported using one or more plant species to help assess precipitation or moisture levels (Table 2). One rancher said:

Ocotillo is a main indicator of rainfall. In the spring through the summer, I watch the ocotillo leaves like a hawk. It's my prime indicator. [After rainfall,] in 3 to 5 days, they can put on new leaves. If it gets dry, they start dropping their leaves. The leaves can start turning yellow, and then it rains, and they green back up. So it's a prime indicator of rainfall. In August, if you go into an area and the ocotillo leaves have turned yellow, you know you've got a problem out there with not as much rain.

Observed, direct effects of precipitation levels on soil surfaces included encrusting during prolonged dry periods, splash marks from individual rain drops after light rains, puddles, and channel flow. Levels of water in stock ponds and riparian areas were used to infer precipitation levels over wider areas and longer timescales than direct effects on the soil surface. One rancher stated:

If you go along and you never see the washes run, you just get little showers here and there, you're not going to have enough feed to make it, usually. You need the washes to run here on this ranch at least twice a year. It would be better if they run three or four times during the monsoon and run big. That makes a difference of a good year or a not so good year. The year that the washes never run is usually the year thatyou're going to have water problems and you're going to have forage problems, too. Usually.

Rainsplash causes micro-erosion, and in the absence of overland flow, these erosion marks can be visible to the naked eye (Jyotsna and Haff, 1997). Relationships between precipitation and water flows in arid and semiarid regions are highly variable, being influenced by catchment size, topography, and climate among other factors (Langbein et al., 1951; Osborn and Renard, 1970; Thornes, 2009). Surface runoff, channel flow, and groundwater recharge tend to be highly irregular, infrequent, and dependent on relatively infrequent, high-intensity precipitation over short periods of the order of several days' duration (Thornes, 2009). Stockponds may be recharged by surface, subsurface, and/or channel flow. In Arizona, stockponds are typically recharged in fewer than 3 months per year and, below 1 500m elevation, predominantly in the peak rainfall months of the monsoon (Langbein et al., 1951). Thus the natural science literature is in accord with ranchers' observations that water levels in riparian areas and stockponds are indicative of recent rainfall, although the details of such linkages are likely to be highly site specific.

Erosion

Most ranchers (93%) observed change over multiple years in riparian channels such as head cutting, bank erosion, and bed scouring. Most also noted effects of erosion on upland soils and vegetation. Ranchers typically assessed the proportions of land that were covered by bare ground and by herbaceous vegetation. Reductions in herbaceous cover and increases in bare ground were considered indicative of susceptibility to soil erosion, increasing soil erosion, and/or low rainfall. The exact interpretation appeared to account for other observations indicative of rangeland trend, such as brush encroachment, or of rainfall levels.

In like manner, formal monitoring can include observations of signs of active erosion in gullies, such as head cutting and incised sides, as well as signs of recovery from erosion, such as increased vegetation growing on gully sides and beds (USDA-NRCS, 2003). In both uplands and riparian areas, herbaceous cover can protect against soil erosion, and soil erosion and drought can reduce herbaceous cover (George et al., 2011; Puigdefabregas, 2005; Snyder and Tartowski, 2006).

Consequently, herbaceous and bare ground cover can be used in conjunction with other indicators to help assess rangeland trend and susceptibility to erosion (Coulloudon et al., 1999a; USDA-NRCS, 2003).

Ranchers also compared depths of water channels to recollections of conditions in previous years and, in most cases, over multiple decades. Two ranchers incorporated oral history from previous generations into long-term erosion assessments, noting that down-cutting of washes had increased markedly since the early 20th century, at which time surface water was more commonly available in the dry season than at present. Long-term formal assessments of erosion are considered valuable in natural science-based rangeland and watershed management, although comprehensive studies over multiple decades are not common in the United States (George et al., 2011; Moran et al., 2008).

Ranchers visually monitored effects of attempts to reduce erosion such as rocks placed in water channels and raised sections of ranch roads. Such observations included increasing in-channel sedimentation and vegetation close to erosion control structures and, in uplands, greater vegetation density, height, and greenness near erosion control structures compared with surrounding areas. One rancher said:

From a visual standpoint, these [raised road sections] are old enough that you can see just taller, more vigorous grass plants in the areas where that runoff comes off the road. It can be pretty stark sometimes, especially in a dry summer. You can see the difference. In the one-rock dams, the soil accumulation above them, and the grass production in those areas, is just absolutely amazing. They kind of explain themselves.

Similar effects of in-channel rock emplacements have been shown in formal studies, although there appears to be considerable variation of effect according to the details of implementation (Gellis et al., 1995; Nichols et al., 2012).

Brush Encroachment and Suppression

Ranchers typically assessed brush encroachment into grasslands from visual observations over periods ranging from several years to several decades. Four ranchers compared current abundance of velvet mesquite with conditions in the early 20th and late 19th centuries, as told by now-deceased former generations of ranchers.

Velvet mesquite was considered to have proliferated in grasslands in the study area much more than any other shrub or tree species, a view consistent with formal studies in the region (McClaran, 2003). However, some small areas (on the order of 1 ha) were dominated by catclaw acacia and were considered by some ranchers to be more impenetrable to livestock than areas dominated by velvet mesquite. Most ranchers were of the opinion that brush encroachment reduced grass forage availability and therefore viewed it as undesirable, in accord with formal studies (Owens et al., 1991). However, one rancher viewed velvet mesquite leaves as having sufficiently high forage value and year-round availability that mesquite encroachment into grasslands on their ranch did not significantly diminish livestock production. Thus the perceived impacts of woody plant encroachment on livestock production integrated observations of changes in plant community, accessibility to livestock, and perceived forage value of grasses and woody species in all seasons.

Changes in herbaceous vegetation and soil erosion were also observed after brush suppression. One rancher noted increased grass growth, lower soil erosion, and mesquite sapling recruitment up to 20 years after brush suppression by herbicide in one area, in comparison with both that area's previous condition and an adjacent, untreated area. Three ranchers noted that combinations of fire, chemical, and mechanical treatments a few years apart appeared more successful than single treatments in reducing woody plant abundance. This is consistent with formal assessments of brush suppression strategies (Archer et al., 2011).

Informal Monitoring by Agency Personnel

Six of the agency personnel described using informal rangeland monitoring for rangeland management. Another agency interviewee said he used only formal monitoring to assess and manage rangelands but nevertheless described making informal observations of rangeland condition. In response to further probing, this interviewee said that useful recommendations on rangeland management could not be made solely on the basis of formal monitoring data but also required time spent on the rangelands in question. The following description of informal monitoring is drawn from the interviews with these seven agency personnel.

Six of these participants made ocular assessments of rangeland condition and forage quantity and utilization. Of these, five also made informal assessments of erosion, and two of recent precipitation and fire. Five agency participants made informal assessments of brush encroachment and effects of brush suppression treatments, and considered informal monitoring to be a normal and important component of rangeland management. Four agency personnel said they took note of ranchers' oral environmental histories and local environmental knowledge, and another three indicated that they often respected the conclusions drawn by experienced ranchers from informal monitoring.

When asked about informal, qualitative assessments of rangeland condition, one participant said, "Every time you go out you're looking, of course." According to another:

Most of what we do is probably ocular reconnaissance. Because when you've got a transect on this side of the pasture that's mostly grass and the other side of the ridge is all prickly pear, you notice it. I don't know how you wouldn't notice that.

A third participant described the importance of informal monitoring's ability to account for diverse aspects of rangeland ecology, thus:

I think to be a good and effective range manager you need to have your eyes open at all times. You've got to look around and see what's going on in general. Is it the climate? Was there a fire up on the mountain that caused a lot of runoff down here? Just a bunch of different factors that can play in. To be a good manager of anything, you need to be aware of everything around you. There are so many variables involved you can't just base it on numbers.

Agency participants' descriptions of informal monitoring were similar to those of ranchers but were typically less comprehensive, forming a subset of the types of observations described by ranchers. For instance, agency personnel made ocular assessments of forage quantity, species composition, and utilization but, unlike ranchers, did not describe using color to assess plant vigor.

The two agency personnel who reported informally monitoring precipitation did so in similar manner to ranchers, by assessing vegetation greenness, water flow in channels, and water depth in stock ponds. In contrast to ranchers, they did not describe assessing effects of precipitation on soil surfaces or using direct visual observations of storms.

Agency personnel informally monitored erosion by assessing head cutting and down cutting over the periods they had been involved in the study area. One also noted effects of erosion control treatments, with increased vegetation close to rocks placed in water channels.

Agency participants also observed changes in woody plant abundance over multiple years, including effects of brush suppression treatments. One participant said, "The sprays that they did, the chemical treatment on the brush, you can really see where that's really helped." Another described assessing chemical brush treatments both formally and informally:

There we have numbers. Our number count on brush has gone down. And the grass production—herbaceous cover—has really increased, as an effect of that. So we've got it in the data. And then also visually, we can see that it's improving, too. You can see the benefit, visually and in the data.

More generally, most agency personnel who described using informal monitoring considered it to be complementary to formal monitoring (Table 1). One agency participant said:

If you get too focused on numbers or data, you're going to miss the big picture. The benefits [of informal monitoring] would be that you'd be looking at the ranch as a whole, or the pasture as a whole. You could see all the aspects that are going on. Versus if you just looked at numbers you just have the numbers, and maybe not see what else is going on. You might miss something. And I guess the disadvantages are, you might miss the signals the data are showing you. When you run the transect it makes you look, really look, and you end up with a plant list that you had no idea you would have gotten, just at first glance. And numbers give you something. When you look, you say, 'I think it looks better than last year,' whereas the numbers will show you, this grass is increasing or decreasing, or grasses in general are producing more this year, or not. The numbers do keep you focused. And they would give you a more definite baseline to work with.

Formal Monitoring

Formal monitoring was performed annually at all but one of the ranches represented by participants in this study and overlapped in purpose with ranchers' informal monitoring. Rangeland condition, long-term trend, and ability to support livestock grazing were assessed in the autumn, at the end of the main growing season. Unlike informal monitoring, this was also intended to facilitate compliance with federal laws such as the National Environmental Policy Act (Coggins et al., 2007). NRCS personnel collaboratively monitored with ranchers on private, county and state land, while USFS and BLM personnel monitored the land under their respective jurisdictions. Agency personnel conducted statistical analyses, which would usually be complete 2-3 months after field assessments. Each pasture's analyses were shared with ranchers grazing that pasture.

Most formal vegetation monitoring was conducted in permanent, representative key areas. Some large pastures had more than one key area, but most had one or none. Key areas were typically smaller than 16 ha, a minority of the area in a pasture. Utilization was estimated in key areas by techniques such as percentage of grazed versus ungrazed plants and double sampling (Elzinga et al., 1998; USDA-NRCS, 2003). Each key area typically had one or a few permanent transects, used to assess rangeland condition and trend based on plant species frequency, dry weight rank, and cover. Two people could collect data from approximately two transects per day. A new transect would usually be monitored each year for the first few years, and thereafter either alternate years or in a minority of years according to a planned schedule. In addition, small fenced grazing exclosures were used as reference sites on some of the ranches in this study (Coulloudon et al., 1999a).

Government agencies also conducted formal monitoring to comply with legislation protecting wildlife, endangered species, and archaeological sites (Coggins et al., 2007; Fish, 1980). Methods included surveying populations and mapping habitats of wildlife and endangered species, as well as archaeological surveys. The objectives were to locate and quantify endangered species and wildlife populations, to identify potentially important habitat locations, and to determine whether proposed rangeland treatments may harm archeological sites. These objectives and monitoring techniques had no direct comparators in the informal monitoring described by any participants. Agency personnel and ranchers considered formal monitoring to be more acceptable than informal monitoring as evidence of compliance with legislation and agency policies, due to its perceived objectivity and lack of bias.

¡ncorporation of Informal Monitoring into Formal Assessments

Formal monitoring reports sometimes included qualitative observations made by agency personnel. An illustrative example given by an

agency rangeland specialist was, "I passed a particular exclosure in a particular riparian area, and the fence was up, and there was no livestock use noted inside the exclosure." Some reports would also include ranchers' informal monitoring observations.

NRCS personnel sometimes included their informal observations on data sheets. One NRCS participant said:

We'll say, 'Last year we saw in this pasture there was a lot of green sprangletop. This year, just by looking at it, there's tons of plains lovegrass,' or something like that. Or, 'I noticed this ditch that wasn't here last year.'

NRCS has also encouraged ranchers to write short histories of their ranches, usually about half a page to a page long, and most ranchers involved with NRCS in the study area had done so. These are filed confidentially by NRCS. An NRCS participant said, "[These] environmental histories give an idea of why a ranch is like it is."

Informal rangeland assessments were sometimes used to adapt formal monitoring plans to changing conditions. If informal observations indicated there was a concern in a pasture, that pasture's transect may be monitored even if it was not in the schedule for that year. One agency participant said, "If there's an obvious concern in a pasture, you will read its transect." Another said:

There are a couple of instances where ranchers have told us that this transect doesn't really represent the entire pasture. Originally, maybe 20 years ago, it did, before that cross fence was put in, a hundred yards away. Maybe at one time it represented the pasture, but they're like, 'No, the rest of my pasture looks a lot better than this part right here, off the side of the road.' So in some instances they want to add new transects that are maybe more representative. Keep the old transect because we have so much data, and maybe that would be an area of concern that we want to improve, so let's keep monitoring it, but add a new one to reflect the rest of the pasture. Or there are other instances where the transect's on a really good little site, but it doesn't represent the entire pasture. It goes both ways.

Ranchers' Use of Monitoring Information

Informal monitoring provided the principal input of information on rangeland condition for most ranchers (86%) when making rangeland management decisions. This included long-term planning, yearly pasture and livestock management plans, and within-year adaptions of yearly plans in response to fluctuating environmental conditions (Table 1). According to one rancher, "the key to any pasture management is observation."

Informal assessments of vegetation could influence the type of business ranchers operate. One rancher stated, "We haven't been in the cow-calf business for quite a few years now, but we go into steers when the feed situation makes that beneficial."

Decisions on what brush suppression and erosion control measures to take and where to implement them were often based on several years of cumulative informal assessments of forage availability and brush abundance or erosion. Planning for rangeland improvements also took into account informal assessments of the effects of past brush or erosion treatments. Such planning often incorporated input from agency personnel and/or rangeland consultants on the basis of formal academic training, research, and monitoring. This was always the case on government land and often the case on private land, where NRCS was the most common source of formal advice.

Two ranchers described changing grazing systems on the basis of informal assessments of forage levels made over several successive years. In both cases, rapid rotation systems were seen to have been insufficiently flexible to respond to severe droughts. After several years of use, rapid rotations were replaced by simpler, more flexible rotation

systems that ranchers felt could be more readily adapted to accommodate variations in precipitation and forage availability.

All livestock ranchers made a yearly management plan, which specified livestock numbers and, for all but one ranch, the timing of pasture usage for the coming 12 months. One ranch produced its yearly plan in January each year, taking into account analyses of formal monitoring conducted the previous autumn. All other ranches produced their plans in the autumn, after the monsoon growing season but before receiving analyses of any formal monitoring conducted that year. All yearly plans covered the upcoming spring dry season and summer monsoon season on the basis of the most current estimates of the forage levels that would be available. Thus yearly plans for all but one ranch were based solely on informal assessments of the amount of forage that would be available until the end of the dry season.

All ranchers varied their yearly plans to some degree according to the condition of vegetation and its responses to precipitation, drought, or wildfire. Water availability also influenced livestock and pasture management directly because water and feed availability were not always spatially and temporally coincident. Subyearly responses to unpredictable conditions included altering the timing of livestock movement between pastures, resting a pasture that had undergone wildfire, purchasing supplemental feed, and selling livestock. Such adaptive responses were aimed at maintaining the condition and productivity of livestock and vegetation, in particular through minimizing overgrazing. Adaptive responses were, of necessity, based solely on informal monitoring of the natural environment, as annual formal monitoring could not provide timely data on or help predict subyearly variation in precipitation levels or wildfire, or their effects on vegetation.

Most ranchers used formal monitoring to assess trends in rangeland condition and plant community composition. Formal monitoring provided greater detail on plant species than was generally attained with informal monitoring. In particular, formal analyses sometimes detected changes in the abundance of minor species, which may be overlooked by informal monitoring. Although these minor species may contribute little to forage production, knowledge of changes in their abundance may be important indicators of trends in rangeland condition.

Ranchers compared and combined formal analyses with informal observations. Combining formal and informal assessments appeared to generate more comprehensive views of rangeland trends than would be possible with either methodology alone. For instance, one rancher compared formal and informal observations of Rothrock grama (Bouteloua rothrockii Vasey). Formal data showed a declining abundance in a pasture over several years, followed by marked increase in 1 yr. The rancher compared this with his own similar, informal observations and with a conversation with another rancher who spoke of rare spikes in Rothrock grama abundance.

Formal monitoring did not generally provide sufficiently timely analyses to inform yearly plans or subyearly adaptations to unpredictable conditions. However, formal monitoring appeared to help inform ranchers' thinking about how to manage rangelands a year or more into the future. Formal monitoring also helped ranchers verify that their forage utilization levels were within the limits set by government agencies, which can be essential to maintaining permission to graze public land. More generally, most ranchers considered both formal and informal monitoring to increase their understanding of the rangelands they managed and therefore to be valuable.

Discussion

Ranchers' informal rangeland monitoring in the study area generally appears compatible with natural science and with formal monitoring practices. This is not to say that either system is incapable of improvement or that they will always agree in practice. Indeed, several ranchers said they did not claim to always "get it right." We did not attempt to quantify the level of agreement between informal and formal assessments of field conditions, and there was variation among ranchers'

views on some issues such as the forage value of velvet mesquite leaves and the impact of woody plant proliferation on livestock production. However, there do not appear to be fundamental conceptual or practical impediments to harnessing both monitoring systems in the cause of rangeland conservation.

Approximately half the government agency participants said they did no informal monitoring. The remainder described using informal methods similar to those employed by ranchers, with some considering it a normal, if not inevitable, activity during field work. Our data do not indicate clear reasons for this divergence of view. Some natural resource managers may use informal or local knowledge implicitly without a clear sense of their precise methods or role (Raymond et al., 2010). Therefore we would advocate further research to determine whether there are major, widespread differences in the degree to which agency personnel conduct informal rangeland assessments, and the extent to which their use and integration with formal assessments may be implicit or tacit.

Each monitoring system has advantages over the other (Table 1). Informal monitoring can more effectively "sample" whole pastures and so account for resource patchiness better than formal methods that are often limited to small proportions of each pasture. In addition, informal monitoring can be, and is, conducted much more frequently than formal monitoring. It is legitimate to question the extent to which adaptive rangeland management could be effectively implemented only using annual field-based formal methods, which cannot detect influential changes that occur on shorter timescales. This is particularly important to rangeland conservation in arid and semiarid lands, where vegetation is dependent on precipitation that is highly variable in both space and time (Snyder and Tartowski, 2006). Further, informal environmental management in drylands has often been refined over decades or centuries to accommodate longer-term climate fluctuations, such as droughts, which are expected to become more severe or commonplace (Knapp and Fernandez-Gimenez, 2008; van Ginkel et al., 2013). Thus informal methods may facilitate proactive adaptation to some of the likely adverse effects of climate change, and hence partially mitigate their adverse effects on ecosystem services and livelihoods.

In our study, ranchers and agency personnel typically considered formal monitoring to be more verifiably objective and accurate than informal monitoring. Thus formal analyses were considered valuable in official or legal disputes concerning rangeland management and conservation, but informal assessments were not. This can be of particular importance regarding compliance with environmental legislation and forage utilization limits set by government agencies. One rancher said they were considering adding formal monitoring transects to their ranch so that its ecological condition could be verified should litigation over grazing permission ever arise.

Both formal and informal monitoring can account for long-term ecological trends. We did not seek to evaluate the objectivity of long-term informal observations. Rather, we suggest that comparisons of formal and informal assessments of trends may give clearer, more comprehensive pictures than either methodology alone. In particular, both methodologies may give insights into rangeland management issues such as livestock grazing sustainability, brush encroachment and soil erosion, and the outcomes of measures used to address them.

Each of these key conservation issues has been assessed informally over multiple decades in our study area. Worldwide, there have been numerous long-term, formal studies of the effects of grazing practices on vegetation communities and forage production (Briske et al., 2011). However, long-term formal studies on the efficacy of brush control treatments are scarce, forcing rangeland scientists to rely primarily on qualitative or informal assessments (Archer et al., 2011). There are also significant gaps in formal evidence on outcomes of erosion control measures, with experienced rangeland and watershed scientists often relying partially on trial and error (George et al., 2011). Thus previous research and the current study indicate that some rangeland scientists concur with our conclusion that informal and formal monitoring can be

mutually compatible and with the predominant view among ranchers in this study that integrating both systems can enhance understanding and conservation of the natural environment.

Mention of informal monitoring is generally absent from formal rangeland monitoring manuals and the range science literature (Coulloudon et al., 1999a; Elzinga etal., 1998; Habich, 2001; Holechek et al., 2004; USDA-NRCS, 2003; Vaughan et al., 2001). This is understandable when considering the need for demonstrably unbiased, re-peatable, and verifiable techniques (Ruggiero, 2009). However, this omission discourages open discussion of the potential value of informal monitoring and stifles creative thinking about how it might be improved, standardized, and integrated into formal monitoring schemes. Thus the formal range science literature largely neglects practical needs that, in some cases, can be at least partially met by informal monitoring.

If the advantages of integrating informal and formal monitoring are to be realized more fully, informal methods must be assessed more widely and deeply than has been possible in this study. Informal techniques used by ranchers and rangeland scientists should be observed in practice, and variations among regions, communities, and individual practitioners should be documented (Kawulich, 2005; Raymond et al., 2010). Informal and formal monitoring systems should be assessed jointly and on an equal basis, along with land managers' existing methods of integrating the two systems (Stringer et al., 2014; van Ginkel et al., 2013). The formal monitoring methods may include remote sensing estimates of above-ground plant biomass, particularly if spatial resolutions of remote sensing analyses match those of informal field assessments (Herrmann et al., 2014). Where possible, the levels of agreement between informal and formal assessments in the field should be quantified, and areas of disagreement investigated further without a predisposition toward either methodology being inherently more valid than the other (Ellis, 2005; Reed et al., 2008). Knowledge should be shared in a two-way dialogue between natural scientists and ranchers working in partnership, and a participatory research approach should be considered (Dyer et al., 2014; Oba, 2012; van Ginkel et al., 2013). Such a process should help develop the mutual respect, trust, and communication necessary to assess and improve monitoring methods, as well as integrate informal with formal monitoring more explicitly and effectively than has so far been common (Geeson et al., 2015; Meuret and Provenza, 2015; Reed et al., 2014).

Implications

This study suggests that informal monitoring by U.S. ranchers can have advantages and disadvantages compared with formal monitoring; that the two systems can be compatible and highly complementary and can be integrated in rangeland conservation; and that ranchers can possess considerable practical knowledge of local natural history. The study did not attempt to represent all U.S. ranching communities, and more work is necessary to determine how widely our conclusions apply. Where they do apply, documenting informal monitoring, acknowledging its value, and integrating it more closely with formal monitoring is likely to improve communication and cooperation among rangeland scientists, land management agencies and ranchers, and lead to more ecologically sustainable land use (Brunson and Huntsinger, 2008; Sayre, 2004; Tanakaet al., 2005).

Integrating informal and formal environmental knowledge is likely to become increasingly important in managing U.S. rangelands. Knowledge of natural history is vital to natural resource conservation, but its teaching in U.S. universities has declined markedly over recent decades (Tewksbury et al., 2014). On retiring, experienced rangeland scientists will often be replaced by ecologists who are highly educated and professional but must make up a shortfall in field experience and training in natural history (Noss, 1996). We suggest that this shortfall could be significantly alleviated if early career rangeland conservationists take advantage of experienced ranchers' informal environmental knowledge. Sharing of knowledge may be achieved by a variety of

methods, including interviews, participatory research, and group workshops oriented to practical rangeland management issues (Ballard etal., 2008; Millar and Curtis, 1999). Whatever methods are used, the critical first steps would appear to be the recognition of the value of informal environmental monitoring and knowledge, and of shared interests in rangeland conservation.

Acknowledgments

We are grateful to all the participants in this study. Diane Austin provided valuable advice on methodology. She and Steven Archer reviewed draft versions of this paper.

References

Andrews, D.A., 1937. Ground water in the Avra-Altar Valley, Arizona. Contributions to the hydrology of the United States. Government Printing, Office, Washington DC, USA, pp. 163-180.

Archer, S.R., Davies, K.W., Fulbright, T.E., McDaniel, K.C., Wilcox, B.P., Predick, K.I., 2011. Brush management as a rangeland conservation strategy: a critical evaluation. In: Briske, D.D. (Ed.), Conservation benefits of rangeland practices: assessment, recommendations, and knowledge gaps. U.S. Department of Agriculture, Natural Resources Conservation Service, Washington DC, USA, pp. 105-170. Ballard, H.L., Fernandez-Gimenez, M.E., Sturtevant, V.E., 2008. Integration of local ecological knowledge and conventional science: a study of seven community-based forestry organizations in the USA. Ecol. Soc. 13 (2), 37 Available at: http://www.ecologyandsociety. org/vol13/iss2/art37/.

Berg, B.L., Lune, H., 1995. Qualitative research methods for the social sciences. Allyn &

Bacon, Boston, MA, USA 432 pp. Berkes, F., Colding, J., Folke, C., 2000. Rediscovery of traditional ecological knowledge as

adaptive management. Ecol. Appl. 10,1251-1262. Briske, D.D., Derner, J.D., Milchunas, D.G., Tate, K.W., 2011. An evidence-based assessment of prescribed grazing practices. In: Briske, D.D. (Ed.), Conservation benefits of rangeland practices: assessment, recommendations, and knowledge gaps. U.S. Department of Agriculture, Natural Resources Conservation Service, pp. 21-74. Brunson, M.W., Huntsinger, L., 2008. Ranching as a conservation strategy: can old

ranchers save the new West? Rangel. Ecol. Manag. 61,137-147. Burgess, T.L., 1995. Desert grassland, mixed shrub savanna, shrub steppe, or semidesert scrub? In: McClaran, M.P., Van Devender, T.R (Eds.), The desert grassland. University of Arizona Press, Tucson, AZ, USA, pp. 31-67 Cable, D.R., 1967. Fire effects on semidesert grasses and shrubs. J. Range Manag. 20, 170-176.

Cable, D.R., 1969. Competition in the semidesert grass-shrub type as influenced by root

systems, growth habits, and soil moisture extraction. Ecology 50, 27-38. Cable, D.R., Martin, S.C., 1975. Vegetation responses to grazing, rainfall, site condition, and mesquite control on semidesert range. Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO, USA 25 pp. Canfield, R.H., 1948. Perennial grass composition as an indicator of condition of southwestern mixed grass ranges. Ecology 29,190-204. Coggins, G.C., Wilkinson, C.F., Leshy, J.D., Fischman, R.L., 2007. Federal public land and resources law. Foundation Press, Westbury, NY, USA 1141 pp. Coppin, P., Jonckheere, I., Nackaerts, K., Muys, B., Lambin, E., 2004. Digital change detection methods in ecosystem monitoring: a review. Int. J. Remote Sens. 25,1565-1596. Coulloudon, B., Eshelman, K., Gianola, J., Habich, N., Hughes, L., Johnson, C., Pellant, M., Podborny, P., Rasmussen, A., Robles, B., Shaver, P., Spehar, J., Willoughby, J., 1999a. Sampling vegetation attributes. US Department of the Interior, Bureau of Land Management, Denver, CO, USA 163 pp. Coulloudon, B., Eshelman, K., Gianola, J., Habich, N., Hughes, L., Johnson, C., Pellant, M., Podborny, P., Rasmussen, A., Robles, B., Shaver, P., Spehar, J., Willoughby, J., 1999b. Utilization studies and residual measurements. US Department of the Interior, Bureau of Land Management, Denver, CO, USA 165 pp. Dyer, J., Stringer, L.C., Dougill, A.J., Leventon, J., Nshimbi, M., Chama, F., Kafwifwi, A., Muledi, J.I., Kaumbu, J.M.K., Falcao, M., Muhorro, S., Munyemba, F., Kalaba, G.M., Syampungani, S., 2014. Assessing participatory practices in community-based natural resource management: experiences in community engagement from southern Africa. J. Environ. Manag. 137,137-145. Ellis, S.C., 2005. Meaningful consideration? A review of traditional knowledge in environmental decision making. Arctic 58, 66-77. Elzinga, C.L., Salzer, D.W., Willoughby, J.W., 1998. Measuring and monitoring plant populations. US Department of the Interior, Bureau of Land Management, Denver, CO, USA 477 pp.

Evans, J., Jones, P., 2011. The walking interview: methodology, mobility and place. Appl. Geogr. 31, 849-858.

Fairhead, J., Scoones, I., 2005. Local knowledge and the social shaping of soil investments: critical perspectives on the assessment of soil degradation in Africa. Land Use Policy 22, 33-41.

Fernandez-Gimenez, M.E., Ruyle, G., McClaran, S.J., 2005. An evaluation of Arizona Cooperative Extension's rangeland monitoring program. Rangel. Ecol. Manag. 58,89-98. Fish, P.R., 1980. Federal policy and legislation for archaeological conservation. Ariz. Law Rev. 22, 681-699.

Fisher, M., 1997. Qualitative computing: using software for qualitative data analysis. Ashgate Publishing Ltd., Aldershot, UK 146 pp.

Geeson, N., Quaranta, G., Salvia, R., Brandt, J., 2015. Long-term involvement of stakeholders in research projects on desertification and land degradation: how has their perception of the issues changed and what strategies have emerged for combating desertification? J. Arid Environ. 114,124-133.

Gellis, A.C., Cheama, A., Laahty, V., Lalio, S., 1995. Assessment of gully-control structures in the Rio Nutria watershed, Zuni reservation, New Mexico. J. Am. Water Resour. Assoc. 31, 633-646.

George, M.P., Jackson, R.D., Boyd, C.S., Tate, K.W., 2011. A scientific assessment of the effectiveness of riparian management practices. In: Briske, D.D. (Ed.), Conservation benefits of rangeland practices: assessment, recommendations, and knowledge gaps. US Department of Agriculture, Natural Resources Conservation Service, Washington, DC, USA, pp. 213-252.

German, L.A., 2010. Local knowledge and scientific perceptions: questions of validity in environmental knowledge. In: German, L.A., Ramisch, J.J., Verma, R (Eds.), Beyond the biophysical: knowledge, culture, and power in agriculture and natural resource management. Springer, Dordrecht, Netherlands, pp. 99-125.

Guest, G., 2014. Sampling and selecting participants in field research. In: Bernard, H.R., Gravlee, C.C. (Eds.), Handbook of methods in cultural anthropology. Rowman & Littlefield, Lanham, MD, USA, pp. 215-249.

Habich, E.F., 2001. Ecological site inventory, technical reference 1734-7. US Department of the Interior, Bureau of Land Management, Denver, CO, USA 112 pp.

Herrick, J.E., Schuman, G.E., Rango, A., 2006. Monitoring ecological processes for restoration projects. J. Nat. Conserv. 14,161-171.

Herrmann, S.M., Sall, I., Sy, O., 2014. People and pixels in the Sahel: a study linking coarse-resolution remote sensing observations to land users' perceptions of their changing environment in Senegal. Ecol. Soc. 19 (3), 29 Available at http://www.ecologyandsociety. org/vol19/iss3/art29/.

Holechek J.L., 1988. An approach for setting the stocking rate. Rangelands 10,10-14.

Holechek, J.L., Pieper, R.D., Herbel, C.H., 2004. Range management: principles and practices. Prentice Hall Inc., Upper Saddle River, NJ, USA 607 pp.

Huntington, H.P., 1998. Observations on the utility of the semi-directive interview for documenting traditional ecological knowledge. Arctic 51, 237-242.

Jackson, R.D., Slater, P.N., Pinter Jr., P.J., 1983. Discrimination of growth and water stress in wheat by various vegetation indices through clear and turbid atmospheres. Remote Sens. Environ. 13,187-208.

Jones, P., Bunce, G., Evans, J., Gibbs, H., Ricketts Hein, J., 2008. Exploring space and place with walking interviews. J. Res. Pract. 4 (2), D2 Available at: http://www.jrp.icaap. org/index.php/jrp/article/view/150/161.

Jyotsna, R., Haff, P.K., 1997. Microtopography as an indicator of modern hillslope diffusivity in arid terrain. Geology 25, 695-698.

Kawulich, B.B., 2005. Participant observation as a data collection method. Forum Qual. Soc. Res. 6, 43 Available at: http://www.qualitative-research.net/index.php/fqs/ article/view/466/996.

Kent, M., 2011. Vegetation description and data analysis: a practical approach. Wiley-Blackwell, Chichester, West Sussex, UK 428 pp.

Knapp, C.N., Fernandez-Gimenez, M., 2008. Knowing the land: a review of local knowledge revealed in ranch memoirs. Rangel. Ecol. Manag. 61, 148-155.

Knapp, C.N., Fernandez-Gimenez, M.E., 2009. Knowledge in practice: documenting rancher local knowledge in northwest Colorado. Rangel. Ecol. Manag. 62, 500-509.

Knipling, E.B., 1970. Physical and physiological basis for the reflectance of visible and near-infrared radiation from vegetation. Remote Sens. Environ. 1,155-159.

Kong, T.M., Marsh, S.E., van Rooyen, A.F., Kellner, K., Orr, B.J., 2015. Assessing rangeland condition in the Kalahari Duneveld through local ecological knowledge of livestock farmers and remotely sensed data. J. Arid Environ. 113, 77-86.

Kozlowski, T.T., 1976. Water supply and leaf shedding. In: Kozlowski, T.T. (Ed.), Water deficits and plant growth. Academic Press, New York NY, USA, pp. 191-231.

Kurc, S.A., Benton, L.M., 2010. Digital image-derived greenness links deep soil moisture to carbon uptake in a creosotebush-dominated shrubland. J. Arid Environ. 74, 585-594.

Langbein, W.B., Hains, C.B., Culler, R.C., 1951. Hydrology of stock-water reservoirs in Arizona. US Geological Survey, Washington D.C., USA 18 pp.

Lindenmayer, D.B., Likens, G.E., 2009. Adaptive monitoring: a new paradigm for long-term research and monitoring. Trends Ecol. Evol. 24, 482-486.

Lindenmayer, D.B., Likens, G.E., 2010. The science and application ofecological monitoring. Biol. Conserv. 143,1317-1328.

Lindenmayer, D.B., Likens, G.E., Haywood, A., Miezis, L., 2011. Adaptive monitoring in the real world: proof of concept. Trends Ecol. Evol. 26, 641 -646.

Mack, N., Woodsong, C., MacQueen, K.M., Guest, G., Namey, E., 2005. Qualitative research methods: a data collector's field guide. Family Health International, Research Triangle Park NC, USA 120 pp.

McClaran, M.P., 2003. A century of vegetation change on the Santa Rita Experimental Range. In: McClaran, M.P., Ffolliott, P.F., Edminster, C.B. (Eds.), Proceedings: Santa Rita Experimental Range: 100 years (1903-2003) of accomplishments and contributions. Us Department of Agriculture, Ogden, UT, USA, pp. 16-33 Forest Service, RMRS-P-30.

Meuret, M., Provenza, F.D., 2015. When art and science meet: integrating knowledge of French herders with science of foraging behavior. Rangel. Ecol. Manag. 68,1-17.

Meyer, W.W., 2000. Altar Valley watershed resource assessment. Unpublished report prepared on behalf of the Altar Valley Conservation Alliance 49 pp.

Millar, J., Curtis, A., 1999. The nature and role of farmer knowledge in temperate pasture management in the Murray-Darling Basin. Rural. Soc. 9,301-312.

Moller, H., Berkes, F., Lyver, P.O.B., Kislalioglu, M., 2004. Combining science and traditional ecological knowledge: monitoring populations for co-management. Ecol. Soc. 9 (3), 2 Available at: http://www.ecologyandsociety.org/vol9/iss3/art2/.

Moran, M.S., Peters, D.P.C., McClaran, M.P., Nichols, M.H., Adams, M.B., 2008. Long-term data collection at USDA experimental sites for studies of ecohydrology. Ecohydrology 1,377-393.

Nichols, M.H., McReynolds, K., Reed, C., 2012. Short-term soil moisture response to low-tech erosion control structures in a semiarid rangeland. Catena 98,104-109.

Niemi, G.J., McDonald, M.E., 2004. Application of ecological indicators. Annu. Rev. Ecol. Evol. Syst 35, 89-111.

NOAA-NCDC, 2002. Climate atlas of the United States: Version 2.0. US Department of Commerce, National Oceanic and Atmospheric Administration, Asheville, NC, USA.

NOAA-NCDC, 2012. Annual observational data. Available at:, http://gis.ncdc.noaa.gov/ map/cdo/?thm=themeAnnual Accessed 10 July 2012.

Nobel, P.S., Zutta, B.R., 2005. Morphology, ecophysiology, and seedling establishment for Fouquieria splendens in the northwestern Sonoran Desert. J. Arid Environ. 62, 251-265.

Noss, R.F., 1996. The naturalists are dying off. Conserv. Biol. 10,1-3.

Oba, G., 2012. Harnessing pastoralists' indigenous knowledge for rangeland management: three African case studies. Pastoralism Res. Policy Pract. 2 (1), 1 Available at http:// www.pastoralismjournal.com/content/2/1/1.

Ogden, P.R., 2003. Seasonal diets of cattle grazing rangelands in Pima and Pinal counties, Arizona. Report for University of Arizona Cooperative Extension. University of Arizona, School of Renewable Natural Resources, Tucson, AZ, USA 22 pp.

Osborn, H.B., Renard, K.G., 1970. Thunderstorm runoff on the Walnut Gulch experimental watershed, Arizona, USA. Proceedings of IASH-UNESCO symposium on results of research on representative and experimental basins; Victoria University of Wellington, New Zealand. International Association of Scientific Hydrology publication 96, Gentbrugge, Belgium, pp. 455-464.

Owens, M.K., Launchbaugh, K.L., Holloway, J.W., 1991. Pasture characteristics affecting spatial distribution of utilization by cattle in mixed brush communities. J. Range Manag. 44,118-123.

Pasiecznik, N.M., Felker, P., Harris, P.J.C., Harsh, L.N., Cruz, G., Tewari, J.C., Cadoret, K., Maldonado, L.J., 2001. The Prosopis juliflora-Prosopis pallida complex: a monograph. HDRA, Coventry, UK 172 pp.

Patton, M.Q., 2002. Qualitative research and evaluation methods. Sage Publications, Inc., Thousand Oaks, CA, USA 688 pp.

Pennington, D.D., Collins, S.L., 2007. Response of an aridland ecosystem to interannual climate variability and prolonged drought. Landsc. Ecol. 22, 897-910.

Peterson, D.J., 2010. Rangeland monitoring survey for ranchers: a report of findings. University of Arizona, School of Renewable Natural Resources. Report for University of Arizona Cooperative Extension, Tucson, AZ, USA 32 pp.

Phillips, W.S., 1963. Depth of roots in soil. Ecology 44,424.

Puigdefábregas, J., 2005. The role of vegetation patterns in structuring runoff and sediment fluxes in drylands. Earth Surf. Process. Landf. 30,133-147.

QSR 2008. NVivo version 8 getting started. Melbourne. QSR International Pty Ltd, Australia, p. 48.

Ralphs, M.H., McDaniel, K.C., 2010. Broom snakeweed (Gutierrezia sarothrae): toxicology, ecology, control, and management. Invasive Plant Sci. Manag. 4,125-132.

Raymond, C.M., Fazey, I., Reed, M.S., Stringer, L.C., Robinson, G.M., Evely, A.C., 2010. Integrating localand scientific knowledge for environmental management. J. Environ. Manag. 91,1766-1777.

Reed, M.S., Dougill, AJ., Baker, T.R., 2008. Participatory indicator development: what can ecologists and local communities learn from each other. Ecol. Appl. 18,1253-1269.

Reed, M.S., Fazey, I., Stringer, L.C., Raymond, C.M., Akhtar-Schuster, M., Begni, G., Bigas, H., Brehm, S., Briggs, J., Bryce, R., Buckmaster, S., Chanda, R., Davies, J., Diez, E., Essahli, W., Evely, A., Geeson, N., Hartmann, I., Holden, J., Hubacek, K., Loris, A.A.R., Kruger, B., Laureano, P., Phillipson, J., Prell, C., Quinn, C.H., Reeves, A.D., Seely, M., Thomas, R., van der Werff Ten Bosch, M.J., Vergunst, P., Wagner, L., 2013. Knowledge management for land degradation monitoring and assessment: an analysis of contemporary thinking. Land Degrad. Dev. 24,307-322.

Reed, M.S., Stringer, L.C., Fazey, I., Evely, A.C., Kruijsen, J.H.J., 2014. Five principles for the practice of knowledge exchange in environmental management. J. Environ. Manag. 146, 337-345.

Ruggiero, L.F., 2009. The value of opinion in science and the Forest Service Research organization. J. Wildl. Manag. 73,811-813.

Ruyle, G.B., 2003. Rangeland livestock production: developing the concept of sustainability on the Santa Rita Experimental Range. In: McClaran, M.P., Ffolliott, P.F., Edminster, C.B. (Eds.), Proceedings: Santa Rita Experimental Range: 100 years (1903-2003) of accomplishments and contributions. US Department of Agriculture, Ogden, UT, USA, pp. 34-47 Forest Service, RMRS-P-30.

Ryel, R.J., Ivans, C.Y., Peek M.S., Leffler, AJ., 2008. Functional differences in soil water pools: a new perspective on plant water use in water-limited ecosystems. Progr. Bot. 69, 397-422.

Sayre, N.F., 2004. Viewpoint: the need for qualitative research to understand ranch management. Rangel. Ecol. Manag. 57,668-674.

Sayre, N.F., 2007. A history of working landscapes: the Altar Valley, Arizona, USA. Rangelands 29,41-45.

Sheridan, T.E., 1995. Arizona: the political ecology of a desert state. J. Polit Ecol. 2,41-57.

Sillitoe, P., Barr, J., Alam, M., 2004. Sandy-clay or clayey-sand? Mapping indigenous and scientific soil knowledge on the Bangladesh floodplains. In: Bicker, A., Sillitoe, P., Pottier, J. (Eds.), Development and local knowledge: new approaches to issues in natural resources management, conservation and agriculture. Routledge, London, UK, pp. 174-201.

Snyder, K., Tartowski, S., 2006. Multi-scale temporal variation in water availability: implications for vegetation dynamics in arid and semi-arid ecosystems. J. Arid Environ. 65, 219-234.

Stintzing, F.C., Schieber, A., Carle, R., 2001. Phytochemical and nutritional significance of cactus pear. Eur. Food Res. Technol. 212, 396-407.

Stringer, L.C., Fleskens, L., Reed, M.S., de Vente, J., Zengin, M., 2014. Participatory evaluation of monitoring and modeling of sustainable land management technologies in areas prone to land degradation. Environ. Manag. 54,1022-1042.

Strittholt, J.R., Bryce, S.A., Ward, B.C., Bachelet, D.M., 2012. Sonoran Desert rapid ecoregional assessment report. US Department of the Interior, Bureau of Land Management, Denver, CO, USA 33 pp.

Tanaka, J.A., Rimbey, N., Torell, LA, 2005. Rangeland economics, ecology, and sustainability: implications for policy and economic research. West. Econ. Forum 4, 1 -6.

Tewksbury, J.J., Anderson, J.G.T., Bakker, J.D., Billo, T.J., Dunwiddie, P.W., Groom, M.J., Hampton, S.E., Herman, S.G., Levey, D.J., Machniclki, N.J., del Rio, C.M., Power, M.E., Rowell, K., Salomon, A.K., Stacey, L., Trombulak, S.C., Wheeler, T.A., 2014. Natural history's place in science and society. Bioscience 64, 300-310.

Thornes, J.B., 2009. Catchment and channel hydrology. In: Parsons, A.J., Abrahams, A.D. (Eds.), Geomorphology of desert environments. Springer, Dordrecht, Netherlands, pp. 303-332.

Thornton, T.F., Scheer, AM., 2012. Collaborative engagement of local and traditional knowledge and science in marine environments: a review. Ecol. Soc. 17 (3), 8 Available at: http://www.ecologyandsociety.org/vol17/iss3/art8/.

Tibby, J., Lane, M.B., Gell, P.A., 2008. Local knowledge and environmental management: a cautionary tale from Lake Ainsworth, New South Wales, Australia. Environ. Conserv. 34,334-341.

Tucker, C.J., 1979. Red and photographic infrared linear combinations for monitoring vegetation. Remote Sens. Environ. 8,127-150.

Tucker, C.J., Sellers, P.J., 1986. Satellite remote sensing of primary production. Int. J. Remote Sens. 7,1395-1416.

USDA-NRCS, 2003. National range and pasture handbook. US Department of Agriculture, Natural Resources Conservation Service, Washington D.C., USA 84 pp.

USDA-NRCS, AVCA, USFWS, AGFD, ASLD, USFS, P. C. N. Resources, The Nature Conservancy, 2008. Altar Valley fire management plan. US Department of Agriculture, Natural Resources Conservation Service, Tucson, AZ, USA 83 pp.

USFWS, 1980. Habitat evaluation procedures. US Department of the Interior, US Fish and Wildlife Service, Springfield, VA, USA 130 pp.

Van Ginkel, M., Sayer, J., Sinclair, F., Aw-Hassan, A., Bossio, D., Craufurd, P., El Mourid, M., Haddad, N., Hoisington, D., Johnson, N., Velarde, C., Mares, V., Mude, A., Nefzaoui, A., Noble, A., Rao, K.P.C., Serraj, R., Tarawali, S., Vodouhe, R., Ortiz, R., 2013. An integrated agro-ecosystem and livelihood systems approach for the poor and vulnerable in dry areas. Food Secur. 5, 751-767.

Vaughan, H., Brydges, T., Fenech, A., Lumb, A., 2001. Monitoring long-term ecological changes through the ecological monitoring and assessment network: science-based and policy relevant. Environ. Monit. Assess. 67, 3-28.

Weaver, J.E., 1958. Classification of root systems of forbs of grassland and a consideration of their significance. Ecology 39, 393-401.

White, J.D., Gutzwiller, K.J., Barro, W.C., Randall, L.J., Swint, P., 2006. Modeling mechanisms of vegetation change in a semi-arid ecosystem. Ecol. Model. 214, 181-200.

Wilson, M., Sapsford, R., 2006. Asking questions. In: Sapsford, R., Jupp, V. (Eds.), Data collection and analysis. Sage Publications Ltd, London, pp. 93-123.