Progress report for GW18-024

Project Type: Graduate Student
Funds awarded in 2018: $25,000.00
Projected End Date: 04/30/2021
Grant Recipient: University of Arizona
Region: Western
State: Arizona
Graduate Student:
Major Professor:
Steven Archer
The University of Arizona
Expand All

Project Information

Summary:

Many rangelands have experienced shrub encroachment in concert with the loss of native grasses. Efforts to combat this phenomenon include a variety of ‘brush management’ practices (BM) traditionally aimed at restoring forage production – but are seldom economically viable from that standpoint. However, shrub encroachment also affects numerous other ecosystem services (ESs). A broader evaluation of the impacts of shrub encroachment and BM on ES would enable: (i) more accurate assessments of the utility of BM and (ii) development of guidelines for determining when, where and under what circumstances to use BM to promote desired ESs. Shrub encroachment/BM influences on ES are locally constrained by soils, topography, natural disturbance, and livestock grazing, but no conceptual framework integrating these factors with ES exists. This project aims to take a holistic approach to the shrub encroachment phenomenon, how it has altered ESs and the ability of BM to recover and restore valued lost services.

Outcomes will include:

(1) Improved understanding of producer demands for ESs on shrub-encroached landscapes

(2) Enhanced knowledge of shrub encroachment/BM patterns and processes

(3) Improved targeting of range management planning and practices by characterizing spatiotemporal changes in ES, and

(4) Identification of ES trade-offs/synergies associated with shrub encroachment/BM in rangeland environments.

To achieve these outcomes, I have been (i) quantifying rates/patterns of shrub cover across the Las Cienegas National Conservation Area (LCNCA), a working rangeland, with contrasting soils, topography, and management histories in order to (ii) set baseline levels of ESs and (iii) identify maximum potential shrub cover based on topo-edaphic variables. This information is being used to (iv) model how future shrub encroachment/BM actions will impact capacities of stakeholder-desired ES across. Recent and ongoing outreach activities have included presentations at local workshops, annual meetings of the Society for Range Management, and county-level Cooperative Extension programs. Written documents (e.g. fact sheets, articles for popular outlet and peer-reviewed scientific journals) are if various draft stages.

Project Objectives:

The overarching goal of this research is to take a holistic approach to the shrub encroachment phenomenon, how it has altered ESs s and the ability of brush management to recover and restore valued services valued by stakeholders.

Specific objectives are to:

(1) Identify and weigh producer demands for rangeland ESs through semi-structured interviews and surveys.

(2) Characterize spatial and temporal dynamics of shrub cover on both shrub-encroached and managed sites with contrasting soils, topography, and management histories.

(3) Model spatiotemporal ecosystem service capacity using

  1. Land cover classifications from (2), and assign values for ES based upon field data and peer-reviewed literature, and
  2. Statistical analysis to evaluate relationships between multiple ESs and identify possible trade-offs and synergies.

Research

Materials and methods:

Objective 1: Identify and weigh producers’ demands for rangeland ES.

Pre-Test:

Semi-structured interviews were conducted between November 2019 and July 2020 with relevant stakeholder groups which included producers, governmental employees, non-governmental land managers, and academicians. The interview contained closed- and open-ended questions concerning participants’ knowledge of issues relevant to this study. Focal topics included (i) shrub encroachment and its environmental impacts, and (2) brush management and motives/desires for engaging in this management tool with respect to ecosystem services. An example of this semi-structured interview can be found here: https://projects.sare.org/wp-content/uploads/Appendix-1.png). Images of rangelands with differing shrub cover (Figure 1) were used to ascertain participants’ preferred degrees of shrub cover and the level of shrub cover at which they would consider implementing brush management. Participants were also asked to review a list of ecosystem services relevant to rangelands and rank them based on preference. A total of 18 participants were interviewed. and insights gleaned from this pre-test was used to create an online survey. Based on responses from the semi-structured interviews, images were updated and the list of ecosystem services was pared down to highlight those deemed beneficial.

Figure 1. Photographs presented to participants during semi-structured interviews showing areas with differing shrub cover. 1) No shrubs; 2) Low cover; 3) Medium-low cover; 4) Medium cover; 5) Medium-high cover; 6) High cover; 7) Closed shrub canopy. Participants were asked to order images from most- to least-preferred based on their aesthetic appeal and by the ecosystem services provided. Participants are also asked to identify which image would represent the level of shrub cover at which brush management should be undertaken.

Online Survey:

The online survey, created using the Qualtrics platform, was distributed in to an audience across Southern Arizona and New Mexico on November 29th, 2020 with follow-up distribution on February 8th, 2021. Stakeholder groups targeted for the online survey included producers, governmental employees, non-governmental land managers, academicians, and people who recreate and/or live near rangelands. The survey consisted of three sections. In the first section, an image-based survey approached was used. Participants were shown Image 1 above and asked to score the panels according to six characteristics with respect to: aesthetics, importance to human well-being, cultural heritage, habitat for biodiversity, recreation and ranching opportunities, whether or not brush management should be undertaken (Figure 2). A Post-hoc Tukey pairwise comparison was performed to test for differences for each variable across each image and Spearman’s rank correlation coefficients were computed from participant scores for the aforementioned characteristics to measure the strength of the associations between these quantitative variables.

Figure 2. Example question from section 1 of the online survey. Participants were shown seven photographs of rangelands with varying levels of shrub cover (Figure 1) and asked to rate each for the above characteristics. Each characteristic was rated on a 0-100 point scale.

In the second section, a Best Worst Scaling approach (Figure 3; Flynn et al. 2007) was used to rank participant perceptions of a suite of seven Ecosystem services (water quality, biodiversity, erosion control, aesthetic value, cultural heritage, recreation/tourism and forage potential). Each service was shown to participants four times and compared to each other service a total of two times. Conditional logit model was used to analyze the relative importance of each ecosystem service preference and a Wald p-Test was used to compare these services. The third and final section collected basic demographic information for survey respondents. The full online survey can be explored using the following link:  https://uarizona.co1.qualtrics.com/jfe/form/SV_0OHbRHfRCXpcfFc

Figure 3. Example Best Worst scaling question from Section 2 of the online survey. Survey design used a balanced incomplete box to rate best and worst of seven ecosystem services (water quality, biodiversity, erosion control, aesthetic value, cultural heritage, recreation/tourism and forage potential). Each service was shown a total of four times and was rated against each other service twice

Figure 3. Example Best Worst scaling question from Section 2 of the online survey. Survey design used a balanced incomplete box to rate best and worst of seven ecosystem services (water quality, biodiversity, erosion control, aesthetic value, cultural heritage, recreation/tourism and forage potential). Each service was shown a total of four times and was rated against each other service twice

Objective 2: Characterize spatial and temporal dynamics of shrub cover on both shrub-encroached and managed sites with contrasting soils, topography.

Aerial image dates used to classify P. velutina cover change across LCNCA ranged from 1936-2017 (Tab. 1). Images for years between 1950 and 1960 were not available.

The 1936 aerial photographs were acquired from Arizona State University Map and Geospatial hub and scanned at 1200 dpi. All other images were downloaded in digital format from EarthExplorer (USGS; http://earthexplorer.usgs.gov/). To remedy issues related to differences in photo scale and resolution, all images were resampled to a common spatial resolution (1 m). Images for 1936 and 1975 were georectified to the 2017 orthorectified base image using ArcGIS (ESRI 2016).

An unsupervised image classification approach [Integrative Self-Organizing Data Analysis Technique (ISODATA)] was used. Because shrub canopies often touched or overlapped it was not possible to reliably consistently identify individual plants Accordingly, shrub patches were our classified sample units. One hundred classes were initially produced by the ISODATA classifications for each year. These classes were then manually assigned as “shrub canopy” or “non-shrub” and reclassified. Region grouping and zonal geometry tools in the GIS program were used to aggregate shrub patches into classes based on the number of 1-m pixels they covered. Entities smaller than 2 pixels (2 m) created “salt and pepper” or speckled appearances to the output classified images and were omitted as it was uncertain whether this represented bunch grasses, succulents (Yucca spp., Opuntia spp) or small shrubs. To minimize shadow effects, poorly illuminated areas in the 1936 image were masked and omitted. These same areas were also omitted from analysis for all other years. Classification accuracies for each year were calculated using a point-based assessment (Stehman 2009), wherein random sample points (n =200) were stratified between “shrub” (n=100) and “non-shrub” (n=100) classes.

We next up-scaled shrub cover from 1-m to 1-ha to buffer errors associated with georeferencing and resampling to a common resolution on the various images. Because we aimed to quantify shrub encroachment potential and response to brush management, areas with other known disturbances (e.g. wildfire, human development, etc.) were masked and excluded from analyses. Areas where brush management practices had been conducted in past years were identified and excluded from analyses aimed at quantifying maximum shrub cover potential. Rate changes of shrub cover change post-brush management were then compared to rates of cover change on undisturbed portions of the landscape (Fig. 4).

Figure 4. Shrub cover (%) on the 42,000 acre Las Cienegas National Conservation Area in 1936, 1975, 1996, 2007 and 2017. Shrub cover was aggregated to 1-ha (100-m) to account for slight differences in georeferenced image positions.

Topo-edaphic template

Ecosites

We used an ecological site description (hereafter “ecosites”) map of LCNCA developed by the Natural Resource Conservation Service (NRCS; https://www.nrcs.usda.gov/wps/portal/nrcs/site/national/home/).

Topography

Elevation was obtained from a 1/3 arc-second (~10-m) digital elevation model (DEM) from the U.S. Geological Survey (USGS) National Elevation Dataset (https://nationalmap.gov/elevation.html). Slope inclination and slope aspect were calculated from this DEM using ArcGIS tools.

Soil Texture

We used clay content from the global SoilGrids map (Hengl et al. 2017) to represent soil texture at 0 (surface), 5, 15, 30, 60, 100 and 200 cm depths. To serve as an index for infiltration potential, percent clay for the top two depths (surface and 5-cm) was averaged.

Topographic wetness index

A topographic wetness index (TWI) was used as an indicator of the effect of topography on soil moisture. TWI values for LCNCA were calculated using the TauDEM 5.0 software suite (Tarboton 2010).

Statistical analysis

We considered the 95th percentile of shrub cover as the maximum potential shrub cover across the entire study area and on specific ecosites. To characterize the effects of topo-edaphic constraints (elevation, soil texture, slope inclination, slope aspect and TWI) on the upper limit of shrub cover across sites, we used additive, non-parametric quantile regression analysis (Koenker et al. 1994). Quantile regression was chosen over other regression methods (i.e. linear, least square and mean) due to its ability to buffer the influence of outliers such as those caused by unaccounted for past disturbances. Quantile regressions were calculated using the ‘rqss’ function within the ‘quantreg’ library in R (http://r-project.org/).

Objective 3: Model Spatiotemporal Ecosystem Service Capacity

With the completion of shrub cover layers for created in Objective 2, modeling of spatial temporal ecosystem services has begun. This portion of the project will continue through the end date of this project and be completed in during the academic year of 2021.

Research results and discussion:

Objective 1: Preliminary Results

Image-based analysis

Of the 86 respondents to date, 10 (12%) were residents, 7 (8%) were recreationists, 15 (17%) were non-profit/non-governmental, 11 (13%) were ranchers/producers, 22 (26%) were governmental/land managers, and 21 (24%) were academicians. Correlation analysis found the two strongest-correlated variables were (i) human well-being and cultural heritage (Spearman’s rho = 0.783, p < 0.001) and (ii) aesthetics and cultural heritage (Spearman’s rho = 0.778, p < 0.001). Habitat for biodiversity best correlated with aesthetics (Spearman’s rho = 0.663, p < 0.001). Recreational opportunity was correlated strongest with cultural heritage (Spearman’s rho = 0.701, p < 0.001) and ranching potential with recreational opportunities (Spearman’s rho = 0.686, p < 0.001).  Need for restoration practices was correlated highest with ranching potential (Spearman’s rho = -0.487, p < 0.001) and lowest with human well-being (Spearman’s rho = -0.301, p < 0.001) (Table 1).

Table 1. Spearman correlations’ coefficients (Spearman’s rho) between the seven benefits evaluated in this study: aesthetics, human well-being, habitat for biodiversity, cultural heritage, recreation opportunities, ranching potential, and requiring restoration.

Post-hoc Tukey tests confirmed that respondents identified most positively with images depicting lower shrub cover with mean ranking scores dropping as shrub cover increased (Fig. 5). The exception to this was observed with rankings towards biodiversity which was statistically similar (Tukey’s post-hoc tests) across all images except Image 7 (highest cover). Insights from an option explanation section within the survey suggests that respondents did not view biodiversity as a net-loss with encroachment. Instead, it was perceived that while some grassland specialist species may be negatively impacted, some generalist species would benefit from higher shrub cover.

Figure 5. Respondent perceptions of images with different levels of shrub cover (Image 1 = lowest cover; Image 7 = highest cover; See Fig. 1) for the seven benefits evaluated: aesthetics, human well-being, biodiversity, cultural heritage, recreation opportunities, ranching potential, and requiring restoration. Bars indicate the mean (+1 SE) respondent rating using a visual analog scales. Different letters indicate statistically significant differences (Tukey post-hoc tests, p < 0.05).

Best-worst Scaling

Of the ecosystem services analyzed, habitat for biodiversity was ranked highest by participants (BW score = 0.68) followed by erosion control (BW score = 0.35) and water quality (BW score = 0.23) (Table 2). Aesthetics, cultural value, recreation and tourism, and forage potential all received a negative score meaning that those services were chosen as the “worst” option more often than being chosen as the “best” option. Cultural value ranked lowest (BW score = -0.43).

Table 2. Frequency each service was selected as the Best and Worst reason to conserve or restore desert grasslands.

A conditional logit was used to estimate the “shared preference” for each ecosystem service which is an indication of the probability that a respondent showed a preference for one service over all others (Table 3). Results show that habitat for biodiversity, which was ranked highest amongst respondents, was approximately 16 times ( = 0.49/0.03) more important than the lowest ranked service cultural heritage and 2.5 times ( = 0.49/0.19) more important than the second highest-ranked service erosion control. Furthermore, P-values calculated in the conditional logit model represents values under the null hypothesis, in which coefficients are zero. Since all variables do not have significant values, this indicated only water quality, habitat for biodiversity, erosion control and aesthetics were regarded as more important than Cultural Value, which is the base variable having a coefficient of zero.

Table 3. Conditional logit estimations of a best-worst experiment to rank the relative importance of ecosystem services.

Specific comparisons between services can further be seen in the All-Variable Comparison Report (Table 4). For example, erosion control was found to have a significantly higher preference among respondents than aesthetics, cultural value, recreation & tourism, and forage production but was less preferable than habitat for biodiversity and statistically comparable to water quality.

Table 4. All-Variables Comparison Report between the seven ecosystem services evaluated in this study. Each comparison is the difference in preference between the ecosystem service type in the row and the service in the column. Statistically significant values (p < .05) are colored either blue or red depending on the sign (+/-) of the difference.

Objective 2: Preliminary Results

Shrub cover change across LCNCA

Shrub cover at the 1-ha scale varied greatly across the study site (Fig. 6) with total shrub cover across the entire site steadily increasing over the past 81 years (1.6% in 1936, 2.8% in 1974, 4.1% in 1996, 5.8% in 2007 and 6.1% in 2017).

Figure 6. Percent shrub canopy cover from 1936-2017 on contrasting soil types. Note: a 100 m buffer zone was created around drainages and these were excluded from analysis. Known areas of past brush management were also digitized but are omitted from the data presented here.

Shrub cover on the various soil types (but excluding intermittent drainages and areas subject to brush management) ranged from 0.04% to 5.04% in 1936, 0.11% to 9.91% in 1975, 0.18% to 14.20% in 1996, 0.20% to 18.62% in 2007 and 0.26% to 18.48% in 2017 (Fig. 6).  Sandy washes, sandy loam, and clayey swales had the highest initial coverage and experienced the greatest subsequent increases in shrub cover. Clay loam upland, limy slopes and loamy bottoms had the lowest initial shrub cover and also experienced lowest levels of subsequent encroachment.

Post brush management re-encroachment

Areas in LCNCA that have undergone past brush management have been mapped and analyzed to assess rates of shrubs re-establishment. Because aerial imagery is more readily available beginning in the 2000s, additional National Agricultural Imagery Program (NAIP) years have been included for these analyses.  Two examples, a loamy upland site (Fig. 7) and a sandy wash site (Fig. 8) are shown below). Both were mechanically cleared of shrubs in the 1970s and in 1975 both had < 1% shrub cover. Re-establishment rates on these sites varied greatly as the loamy upland saw shrub canopy cover rise to 2.3% in 2010 at which time re-treatment was implemented and brought cover back down to < 1% by 2017. Re-establishment occurred much more rapidly on the sandy wash sites, reaching 18.4% by 2010 and 22.1% by 2017 (Fig. 9). These results will be valuable for land managers as they provide a basis for understanding on how shrub re-establishment rates are controlled by topo-edaphic factors and estimating when follow-up treatment may be required to meet a given management objective.

Figure 7 Loamy upland site which underwent mechanical brush management in the 1970s with re-treatment occurring in 2010. Fire occurred in this area in 2007 but was not found to have an effect on larger shrubs.
Figure 8 Sandy wash which up to 1975 was an agricultural field and kept clear of shrubs. Fields were abandoned following 1975, thus opening the door for shrub encroachment.
Figure 9. Post-treatment re-establishment of shrub cover on a loamy upland and sandy wash ecological sites.

Shrub encroachment and topo-edaphic constraints

Additive non-parametric quantile regression analysis was used ascertain the potential upper limit of shrub cover (95th percentile) for various topo-edaphic settings. Shrub cover values across all years were aggregated for these analyses with the assumption that the 95th percentile approximates the upper limit of shrub cover in that location.

Ecosites

Pooled across LCNCA ecosites the upper limit of shrub cover was 17.6% (Fig. 10). Upper limits of shrub cover were highest on loamy bottoms (37.4%) and clayey swales (35.8%) and lowest on granitic hills volcanic hills loamy (2.0%) and limestone hills (1.8%). The two most spatially extensive ecosites on the LCNCA, loamy uplands 4919 ha) and loamy uplands limy slopes (6734) had upper limits of shrub cover of 5.4 and 16.6%, respectively. This difference is believed to reflect elevation-effects.

Figure 10. Box and whisker plots of shrub cover (%) by ecosite. The center, bottom, and top of boxes indicate the median, the 25th percentile, and the 75th percentile; diamonds are the 95th percentiles.

Elevation

Within the LCNCA, the upper limits of shrub cover potential was found to decrease with increasing elevation: as elevation increased from 1360-m to 1465-m potential shrub cover decreased from 56% to 9%. From 1465-m to the max elevation of 1572-m potential shrub cover stayed relatively steady with only a slight decrease from 9% to 4% (Fig. 11). Elevation was found to a have similar relationship with shrub cover potential across all of the major ecosites at LCNCA.

Figure 11: Shrub cover (%) by elevation (m). Data points are color coded by ecosites (see above key). Black line represents the 95th quantile regression line. Maximum potential shrub cover was typically highest at lower elevations and declined as elevation is increased. Clayey swales, loamy bottoms, loamy uplands and loamy uplands limy slopes were found to have similar relationship.

Slope inclination

Upper limits of potential shrub cover across LCNCA was shown to have a U-shaped relationship with slope inclination with potential maximum cover decreasing up to ~ 12°, at which point potential cover increased as slope steepness increased (Fig. 9). Unlike elevation, this relationship was not the same across the major ecosites with both loamy bottoms and clayey swales showing increases in cover potential as slope increased. Loamy uplands and limy slopes were found to have an initial drop in potential cover which then leveled off; slope inclination did not influence potential shrub cover on the loamy slope ecosite. 

Figure 12: Shrub cover (%) by slope (°). Data points are color coded by ecosites (see above key). Black line represents the 95th quantile regression line.

Slope aspect

Potential shrub cover was found to be highest on East- (24.7%) and Northeast- (21.1%) facing slopes and lowest on Southwest- (10.5%) and South- (11.6%) facing slopes (Fig. 13). On clayey swales shrub cover potential was highest on Southwest- (53.4%) and lowest on West- (16.9%) facing slopes while potential cover was highest on Northeast- (12.1%) and lowest on Northwest- (5.9%) facing slopes across the limy slope ecosite. For the loamy bottom ecosite, potential shrub cover was highest on Northwest- (41.1%) slopes and lowest on Southern- (33.5%) slopes. On loamy slopes shrub cover potential was highest on Northeast- (12.1%) and lowest on West- (6.2%) facing slopes while potential cover was highest on East- (20.7%) and lowest on Southwest- (12.4%) facing slopes across the loamy uplands ecosite. For the loamy uplands limy slopes ecosite, potential shrub cover was highest on East- (7.4%) facing slopes and lowest on Southern- (3.9%) slopes.

Figure 13. Box and whisker plots of shrub cover (%) by Slope Aspect. The center, bottom, and top of boxes indicate the median, the 25th percentile, and the 75th percentile; diamonds are the 95th percentiles. Top figure represents the shrub cover by slope aspect across entire LCNCA.

Surface Clay (0 – 5cm)

Spatial variation in soil texture is believed to influence the spatial variation in potential shrub cover due to trade-offs with precipitation infiltration/percolation and soil fertility. Soils with high clay content may be more resistant and sandy soils more susceptible to shrub encroachment. These trade-offs were observed based on shrub cover potential and surface clay content across LCNCA (Fig. 14). On soils with surface (0-5cm) clay contents of 19% to 22%, maximum potential shrub cover was relatively low and stable at ~ 9%. As surface clay increased from 22% to 28% potential shrub cover increased to 33%, but then quickly declined with further increases in clay content. Clayey swales, loamy bottom, loamy uplands, and loamy uplands limy slopes ecosites all shared a similar relationship while potential shrub cover on limy slopes and loamy slopes was not found to be influenced by surface clay content.

Figure 14: Shrub cover (%) as a function of surface (0-5-cm) clay content. Data points are color coded by ecosites (see above key). Black line represents the 95th quantile regression line.

Topographic Wetness Index (TWI)

Contrary to studies at the Journada Experimental Range in southern New Mexico, TWI was found to have a slight positive correlation with potential shrub cover on the LCNCA (Fig. 15). Of the six major ecosites loamy uplands, limy slopes and slopes shared a similar relationship. Both loamy bottoms and loamy uplands limy slopes showed an initial decrease in potential shrub cover as TWI increased with cover abruptly leveling out. Potential cover on clayey swales decreased with increasing TWI values.

Figure 15: Shrub cover (%) in relation to the Topographic Wetness Index (TWI). Data points are color coded by ecosites (see above key). Black line represents the 95th quantile regression line.
Participation Summary
25 Farmers participating in research

Educational & Outreach Activities

2 Tours
12 Webinars / talks / presentations
1 Workshop field days

Participation Summary

50 Farmers
200 Ag professionals participated
Education/outreach description:

Participation Summary

To date, 18 participants have been directly involved in the study through semi-structured interviews as well as 86 participants who completed the online survey. Participants include academicians, government employees, non-government personnel, and land users. February, 2021 I sent out an additional 50 surveys to ranchers/landowners in an effort to beef up input from this stakeholder group.

Outreach & Education Description

2019-2021

On April 1st, 2021 I am scheduled to deliver a “lunch and learn” presentation (~1 hour) to members of the Cienega Watershed Partnership. This event are usually well-attended by a broad diversity of people, many of whom will likely have only a superficial knowledge or understanding of shrub encroachment/brush management.

At this year’s virtual Society for Range Management Annual Meeting on February 15th, 2021 I presented a poster covering results from Objective 1. Due to COVID concerns forcing this conference to be held virtually interactions were challenging. With that said, some good discussions did occur – especially with how to increase stakeholder participation and inclusion of diverse groups within projects like this one. I was also contracted directly by five people who wanted to be kept updated on results as they emerge. 

A poster presentation on Objective 2 was given on February 18th, 2020 at the annual meeting of Society for Range Management in Denver, Colorado. This national/international conference draws a diverse set of participants including research scientists, governmental and non-governmental land managers, and producers., I was able to interact with a number of people regarding my research during my poster session. Discussions ranged from GIS/Remote sensing methods for classifying historical imagery, acquisition of historical imagery, and perspectives (professional and anecdotal) on drivers of shrub encroachment.

2018-2019

A 30-minute oral presentation was given at the June 22, 2019 Science on the Sonoita Plain Symposium in Elgin, Arizona (http://www.cienega.org/event/sosp-10th-annual-symposium/). The conference was attended by ~50 participants and included scientists, governmental and non-governmental land managers, producers, and interested local citizens. The presentation generated good discussion and interest as many people saw connections of how this work can be applied to ongoing and planned activities in the region. Furthermore, a large brush management project is slated to begin late 2019 early 2020 at LCNCA. Participants were interested to see how this research can help better plan best places to treat and when to plan for retreatments. From this, I was able to provide shrub cover maps for the allotted treatment areas spanning from 1936 to present. These maps are being used by the lead brush treatment manager to target which sites are best suited for treatment (i.e. has underdone significant encroachment vs. high shrub cover always present) as well as logistics of the treatment themselves. 

A poster presentation of Objective 2 results was given on December 4, 2018 at the University of Arizona’s Fall Student Showcase. The poster was presented on behalf of the Arid Lands Resource Sciences program. The showcase had roughly 100 attendees which included undergraduate and graduate students, post-docs, professors and university technical staff. Many of the attendees were unaware of the shrub encroachment phenomenon and were interested to learn about the topic.

2017-2018

A poster presentation of preliminary results for Objective 2 was given on June 2, 2018 at the Science on the Sonoita Plain Symposium in Elgin, Arizona. The conference was attended by ~75 participants and included scientists, governmental and non-governmental land managers, producers, and interested local citizens. The poster was well received and a number of attendees exchanged information in order to be updated on final results as they become available.

Aspects of this research was also presented in Spring of 2018 to roughly 150 undergraduate students enrolled in a general science course (Global Change 170) at the University of Arizona during a section on global change and its impacts on the Southwest. Students enrolled in this course were mainly non-science majors focusing on business, communication and psychology. A number of students expressed interest in this project and if given again in the future, a site visit may be included to those interested.

From 2017-2019, I attended five conferences and five workshops in Arizona (listed below). These events provided great networking opportunities in which I was able to explain my research and methodology and interact with relevant stakeholders who I am hoping to interview for Objective 1.

Conferences:

  • Research Insights in Semiarid Ecosystems (RISE) October 21st, 2017 Tucson, Arizona. Organized by Agricultural Research Services and College of Agriculture and Life Sciences University of Arizona. Approximately 100 participants.
  • Madrean Conference May 14th-18th, 2018 Tucson, Arizona. Organized by Sky Island Alliance. Approximately 500 participants.
  • Science on the Sonoita Plain June 2nd, 2018, Elgin, Arizona. Organized by Audubon Society. Approximately 75 participants
  • Research Insights in Semiarid Ecosystems (RISE) October 20th, 2018 Tucson, Arizona. Organized by Agricultural Research Services and College of Agriculture and Life Sciences University of Arizona. Approximately 100 participants.
  • Science on the Sonoita Plain June 22nd, 2019, Elgin, Arizona. Organized by Cienega Watershed Partnership. Approximately 50 participants
  • Research Insights in Semiarid Ecosystems (RISE) October 26st, 2019 Tucson, Arizona. Organized by Agricultural Research Services and College of Agriculture and Life Sciences University of Arizona. Approximately 100 participants.
  • Society of Range Management Annual Meeting, February 15th-18th, 2020 Denver, Colorado. < 1000 participants.
  • Society of Range Management Annual Meeting, February 15th-18th, 2020 Denver, Colorado. < 1000 participants.
  • Society of Range Management Annual Meeting, February 16th-19th, 2020 Virtual. < 1000 participants

Workshops:

  • State of the Cienega Watershed March 6th, 2018 Tucson, Arizona. Organized by Cienega Watershed Partnership. Approximately 50 participants.
  • Altar Valley Brush Management Workshop April 19th, 2018 Arivaca, Arizona. Organized by Altar Valley Conservation Alliance. Approximately 75 participants.
  • Altar Valley Watershed Workshop March 28th, 2019. Organized by Altar Valley Conservation Alliance. Approximately 35 participants.
  • State of the Cienega Watershed May 21tst, 2019 Tucson, Arizona. Organized by Cienega Watershed Partnership. Approximately 50 participants.
  • Altar Valley Conservation Alliance Community Meeting May 29th, 2019. Organized by Altar Valley Conservation Alliance. Approximately 50 participants.

A preliminary website has been created for this project: https://cals.arizona.edu/research/archer/es_changes.html. As results become available the site will be updated.

Projected Outcomes

Did the project contribute to a larger project?: No

Project Outcomes: Results from the activities to date are currently being analyzed and outreach activities are ongoing. Final results for both Objectives 1 & 2 and anticipated to be available in the Spring 2021. Results from Objective 2 are currently being drafted for publication in a refereed scientific journal. Data related to Objective 1 are still coming in, as I am trying to increase rancher/producer responses to the online survey. I expect analysis for this objective to be finalized in mid-April 2021 with preparation for publication to follow immediately.

Knowledge Gained:

Working on this project has increased both my knowledge of extension activities as well as the importance of good outreach practices which target a variety of stakeholders. Over the past three years, I have had the opportunity to participate in the planning process of a broad-scale brush management project focused in Southern Arizona. I was invited to participate with the expectation that my project will shed light on where and when to implement brush management to achieve desired conservation objectives and when re-treatment might be needed. Spatial layers created from this project have been shared with those planning said treatments in order to help inform them the past history on areas targeted for treatment. Seeing the number of stakeholders who are participating in this planning (i.e. both governmental and non-governmental land managers, producers, hunters and community members) has shed light on the complexity of sustainability issues and why a variety of stakeholders are vital for the success and longevity of conservation actions. Due to my involvement with said issues and because of the relationships I have fostered through outreach efforts, in 2019 I was invited to join the Science Advisory Board for the Altar Valley Conservation Alliance.

Over the course of this project I have learned a number of skills and tools that will help me in my future conservation endeavors. Through this project, I now have an advanced working knowledge of Geographic Information Systems and other remote sensing applications. I also learned how to program and analyze data using R and how to develop and deploy online surveys. Additionally, I have also gained knowledge on the subject matter itself which is being shared with local groups and land managers.

Project Outcomes

Did this project contribute to a larger project?:
No
Project outcomes:

The results and final outreach for this project are still in preparation. Final results for both Objectives 1 & 2 and anticipated to be available in the near future and this annual report will be updated when available.

Knowledge Gained:

Working on this project has increased both my knowledge of extension activities as well as the importance of good outreach practices which target a variety of stakeholders. Over the past two years, I have had the opportunity to participate in the planning process of a broad-scale brush management project focused in Southern Arizona. I was asked to participate because it is anticipated that my project can help shed light on where to treat to achieve desired conservation objectives and when re-treatment can be expected. Spatial layers created from this project have been shared with those planning said treatments in order to help inform them the past history on areas targeted for treatment. Seeing the number of stakeholders who are participating in this planning (i.e. both governmental and non-governmental land managers, producers, hunters and community members) has shed light on the complexity of sustainability issues and why a variety of stakeholders are vital for the success and longevity of conservation actions.

Once analysis for this project is completed, we will have a better understanding of the rates/dynamics of the shrub encroachment phenomenon, how it has impacted various ecosystem services, stakeholder demand for those services, and when, where and how often to re-treat to meet desired conservation goals. This will not only benefit the above mentioned planning process but other producers and land managers across the west dealing with encroachment.

Knowledge Gained:

Working on this project has increased both my knowledge of extension activities as well as the importance of good outreach practices which target a variety of stakeholders. Over the past year, I have had the opportunity to participate in the planning process of a large scale brush management project focused in Southern Arizona. I was asked to participate because it is anticipated that my project can help shed light on where to treat to achieve desired conservation objectives and when re-treatment can be expected. Seeing the number of stakeholders who are participating in this planning (i.e. both governmental and non-governmental land managers, producers, hunters and community members) has shed light on the complexity of sustainability issues and why a variety of stakeholders are vital for the success and longevity of conservation actions.

Once analysis for this project is completed, we will have a better understanding of the rates/dynamics of the shrub encroachment phenomenon, how it has impacted services, and when, where and how often to treat to meet desired conservation goals. This will not only benefit the above mentioned planning process but other producers and land managers across the west dealing with encroachment.   

Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.