Proceedings brand creation for a prescribed fire culture – utilizing key social media parameters. Lars Coleman*1, J. Kelly Hoffman1, Thomas McDaniel1, R. Patrick Bixler2, Urs P. Kreuter1, Morgan Russell3

Huisache (Acacia farnesiana [L.] Willd.) is expanding in range and density and displacing more desirable forage species in South Texas. We characterized optimal timing and environmental conditions for effective herbicide control by analyzing root Total Nonstructural Carbohydrate (TNC) content and mortality of sprayed trees over two years. Each month between April 2012 and November 2014, five shrubs in each of four different study sites were excavated, and the root crowns were collected and analyzed for TNC. In addition, two herbicide formulations were foliar-applied to five to ten shrubs every month at the four sites from July 2012 to November 2014, and mortality was evaluated following two growing seasons. Data were analyzed using a randomized complete block design ANOVA, and Akaike’s Information Criterion (AIC) was used to determine the best-fit model for mortality. We found significant TNC increases (compared to the prior month) during May, August, and December. Herbicide-induced mortality was greatest during the months of May, October, November, and September. The best fit model for mortality was a sixth-order polynomial function of mortality vs. month; when month was removed as a parameter, the best fit model was a quadratic function of mortality vs. soil temperature, combined with a quadratic function of mortality vs. phenology. These results indicate two windows in which to chemically treat huisache to achieve high mortality in the coastal plains of South Texas:  spring (especially May) and autumn (especially October and November). Besides month (which can have variable environmental conditions), high huisache mortality rates can also be expected when soil temperature at 0.3 m is at or near a peak of 24.5°C, during the full canopy stage. 

AMINOPYRALID IN COMBINATION WITH PICLORAM AND FLUROXPYR FOR PICKLYPEAR CONTROL IN TEXAS. James R. Jackson*¹, Morgan Russell², Charles R. Hart³; ¹Texas A & M AgriLife Extension Service, Stephenville, TX, ²Texas A&M Agrilife Extension, San Angelo, TX, ³Dow AgroSciences, Stephenville, TX

Pricklypear cactus (Opuntia spp.) is a common invading species on central, south and west Texas rangelands that limits forage production and decreases grazeable acres. Traditionally pricklypear has been controlled by applications of picloram (Tordon 22K) or picloram plus fluroxypyr (Surmount). In the fall of 2013 research was started to analyze the effects of aminopyralid when mixed with Tordon 22K or Surmount. The objectives were to evaluate the speed of desiccation with the addition of aminopyralid and if equivalent mortality could be achieved with lower rates of Tordon or Surmount herbicides when combined with aminopyralid. Trials were conducted in 2013 and 2014 using various tank mix combinations of picloram, fluroxypyr and aminopyralid on pricklypear applied by individual plant treatment, ground broadcast and aerial applications. Based on results obtained from 2013 and 2014 trials, a proprietary formulation (GF-2969, Dow AgroSciences) containing a mixture of aminopyralid, picloram and fluroxypyr was tested.  Trials were established in 2015 and 2016 to test the speed and efficacy of pricklypear control with GF-2969.  Treatments were applied by Individual plant treatment, ground broadcast, and aerial application methods. All trials were evaluated for percent desiccation at 4, 8, 12 and 24 months after treatment.  Results for the tank mix trials in 2013 and 2014 trials indicate that the addition of aminopyralid with Tordon 22K and Surmount resulted in reduced usage rates while maintaining a high rate of efficacy. Results of the 2015 and 2016 trials indicate that a higher level of efficacy is achieved with lower use rates of GF-2969 as compared to Surmount.  GF-2969 also outperformed Tordon 22K at equivalent rates with both greater efficacy and speed.

ESTABLISHED ALIEN INVADER TREES AND PROBABLE INVADER TREES IN ARIZONA. John H. Brock*; Arizona State University Polytechnic, Mesa, AZ

John. H. Brock, Professor Emeritus, College of Integrative Sciences and Arts, Arizona State University Polytechnic, Mesa, Arizona, USA.

DEVELOPMENT OF AN AUTOMATED METHOD TO QUANTIFY BEEF CATTLE DRINKING ACTIVITIES IN EXTENSIVE GRAZING SYSTEMS. Lauren R. OConnor*¹, Greg J. Bishop-Hurley², Dave L. Swain³; ¹CQUniversity Australia, Rockhampton, Queensland, Australia, ²CSIRO, St Lucia, Australia, ³CQUniversity Australia, Rockhampton, Australia

Development of an Automated Method to Quantify Beef Cattle Drinking Activities in Extensive Grazing Systems

Williams, L., Fox, D., Bishop‐Hurley, G., Swain, D. (in press) A pilot study to record drinking behaviour of beef cattle using accelerometers. Animal Production Science.

USE OF AN UNMANNED AERIAL VEHICLE (UAV) TO EVALUATE GRAZING STRATEGIES IN THE NEBRASKA SANDHILLS. Amanda E. Shine Sanford*; University of Nebraska, Lincoln, NE

Nutrient inputs are commonly modeled at the pasture level as if cattle dung and urine deposition were spatially uniform.  However, nutrient return by cattle on grazinglands is patchy and is influenced by a variety of factors, including variation across a pasture in vegetation quality and species composition, location of water and mineral sources, shade availability, and pasture topography.  Stocking rate and grazing strategy also affect dung and urine distribution, leading to significant variation both temporally and spatially in nutrient return to rangelands, but these effects are not yet well known or described in the scientific literature.  In order to gain more insight into the spatial dynamics of nutrient return on grazinglands, an unmanned aerial vehicle (UAV) with a 4-band multispectral sensor was used to monitor the effects of stocking density on the spatial and temporal changes in dung distribution on pastures grazed by yearling cattle on a sub-irrigated meadow located in the Nebraska Sandhills.  Different stocking densities were created by implementing two different grazing strategies in a 60-day grazing season:  a four-pasture grazing rotation with one 15-day occupation per pasture and a 120-pasture mob grazing system with one 0.5-day occupation per pasture.  Stocking densities were 7,000 kg/ha and 225,000 kg/ha, respectively, for the treatments.  Dung was identified through image analysis techniques and then mapped using a geographic information system (GIS).  The resulting distribution maps were analyzed using spatial statistics to identify clustering patterns and then evaluate how patterns differed between strategies, over time within the same groups, and over the course of the grazing season.  Results of this study are being used to develop and inform a nutrient cycling model which accounts for patterns of dung distribution tied to different grazing strategies when estimating the pulses of nitrogen, phosphorus and carbon being returned to the grazed ecosystem.  

ESTIMATING FORAGE BIOMASS AND UTILIZATION IN A DESERT GRASSLAND WITH SMALL UNMANNED AERIAL SYSTEM IMAGERY. Jeffrey K. Gillan*¹, Mitch McClaran¹, Tyson Swetnam¹, Phil Heilman²; ¹University of Arizona, Tucson, AZ, ²USDA-Agricultural Research Service, Tucson, AZ

ABSTRACT

Forage biomass and utilization are important indicators for evaluating livestock and range management in dryland ecosystems. Traditional field methods are typically obtained from few locations within a management unit because of large investment in travel and field time. This small spatial coverage and few samples can limit the accuracy of representing these indicators in a large management unit. To address this challenge of efficiently covering large areas without diminishing the quality of information, we deployed a small unmanned aerial system equipped with high resolution true color camera, operated with autonomous mission planning, and processed the data with advanced image analysis capable of estimating indicator values.

The high-resolution 3D models represented approximately 40% of the grass heights which contains 80% of the biomass. Preliminary results show 1) realistic estimates of biomass based on image-based extrapolations of volume based on field estimates and 2) estimates of utilization that are closely related to field methods at both the transect and plot scales. Our preliminary results are consistent with the promise of more efficiently obtaining high-resolution information for larger area than traditional field methods that under-sample the large of extent of rangelands. 

ANNUAL, HIGH RESOLUTION, PERCENT VEGETATION COVER MAPS OF US RANGELANDS FROM 1984-2015. Matthew O. Jones*, Brady Allred, Dave Naugle; University of Montana, Missoula, MT

Land cover maps are essential tools for tracking land surface changes, informing land management, monitoring environmental conditions, and assessing conservation efforts.  Current products, however, lack essential spatial or temporal resolutions required for such efforts.  The categorical classes in a majority of these maps also fail to capture the inherent heterogeneity of the land surface.  To address the need for high spatial and temporal resolution land cover maps that capture natural landscape variation, we produced annual, 30 meter, continuous land cover maps for U.S. rangelands from years 1984 to 2015.  We used over 23,000 vegetation plots from the NRCS National Resources Inventory and BLM Landscape Monitoring Framework spanning years 2004-2014 to train and validate an assortment of machine learning regression models.   We capitalized on a cloud-based platform for planetary-scale geospatial analysis with massive computational capabilities (Google Earth Engine) to process and store over 250 spatially contiguous gridded climatic, biotic, and abiotic spatiotemporal variables (inclusive of satellite remote sensing data) to drive the model.  The resulting yearly maps provide percent cover of annual and perennial herbaceous vegetation, shrubs, bare ground, and litter for U.S. rangelands, inclusive of the Great Plains.  The modeling framework allows for annual generation of land cover maps into the future.  These maps provide essential information for rangeland management and conservation, and will be used to create a 30 meter spatial resolution primary productivity model optimized for rangeland vegetation.

TESTING THE ECOLOGICAL SITE GROUP CONCEPT
. Shawn W. Salley*¹, Jonathan J. Maynard¹, Travis W. Nauman², Curtis J. Talbot³, Joel R. Brown³; ¹USDA-ARS, Las Cruces, NM, ²US Geological Survey, Moab, UT, ³USDA-NRCS, Las Cruces, NM

ABSTRACT

The 2016 “Ecological Sites for Landscape Management” special issue of Rangelands recommended an update to our thinking of Ecological Sites, suggesting that in our desire to make Ecological Sites more quantitative, we abandoned consideration of Ecological Sites’ spatial context. In response, Ecological Site Groups (ESGs) and associated general state-and-transition modelwere proposed as a framework for describing landscape-level processes occurring across multiple ecological sites, and thus integrating multiple ecological sites from similar landscapes into common behavioral units. We hypothesized that the spatial distribution of ESGs could be predicted using readily available geospatial data due to the theoretical association between ESGs and landscape biophysical properties. Here we test ESG concepts with a spatial modeling framework using machine learning algorithms, a SSURGO modified NASIS point dataset, and a suite of remote sensing-based spatial covariates (e.g., hyper-temporal remote sensing, terrain attributes, climate data, land-cover, and lithology). Our modeling approach was tested on two Major Land Resource Area (MLRA) study areas within the western U.S., representing 6.1 million ha within MLRA 35 and 7.5 million ha within MLRA 42. Results show our approach was effective in mapping ESGs, with a 64% correct classification based on 1,406 point observations across 8 expertly-defined ESG classes in MLRA 35 and a 75% correct classification based on 2,626 point observations across 9 expertly-defined ESG classes in MLRA 42. National coverage of the training and covariate data used in this pilot study provides opportunities for a consistent national-scale mapping effort of ESGs.

CHARACTERIZATION OF DIFFERENT RANGELAND SITES IN SEMI-ARID AREAS OF SUDAN USING REMOTE SENSING TECHNIQUES
. Nancy I. Abdalla*, Abdelaziz K. Gaiballa; Sudan University of Science and Technology, Khartoum, Sudan

ABSTRACT

This study was conducted at North Kordofan State in the semi-arid areas of Sudan, to provide information for identifying characteristics of different range sites using remote sensing data based on understanding the interactive relationships that include topographic, soil feature and vegetation cover. Three sites representing the main range types in the study area were selected according to variations in soil type (flat sandy sites, sand dune site and hard clay site). Vegetation measurements mainly cover, biomass and trees density were measured. Remote Sensing data in form of MODIS/TERRA surface reflectance at 250 m spatial resolution was used to study seasonal vegetation variability between different rangeland sites using the Normalized Difference Vegetation Index (NDVI) for five years (2010-2014). The study results showed variations among different rangeland sites with different NDVI values and pattern of monthly changes, where hard clay site showed high NDVI values early and late in the rainy season for the five years. Late in the season sand dune sites showed the highest NDVI values along the five years. The sand sheet site tends to reflect lower NDVI values during September and October. Variations of plant cover, biomass and tree density during the growing season and as dictated by different soil types at each site are the main factors behind variations in values of NDVI. On other hand, soil characteristics influence plant cover quantitatively and qualitatively. Values of Normalized Difference Vegetation Index (NDVI) and their pattern along months of year can be used for characterization of rangelands sites as part of range classification and for understanding spatial change in rangeland types.  

HISTORICAL TRENDS OF THE DISTRIBUTION OF SAND SHINNERY OAK PRAIRIE IN THE SOUTHERN HIGH PLAINS

We investigated historical trends of the extent and geographical distribution of sand shinnery oak prairies (SSOP) on the Southern High Plains in New Mexico and Texas. Our objective was to create a baseline dataset on land cover change in the historical distribution of SSOP that facilitates future investigations into how vegetation heterogeneity could have shaped lesser prairie-chicken (Tympanuchus pallidicinctus) habitat and population demography in the SSOP. We analyzed historical maps and documentation prior to the 1930s and remotely sensed data (aerial photography and satellite imagery) collected from 1930 to 2015. Land cover and land use maps from 19th and 20th centuries were digitized into a Geographic Information System. The presence of sand shinnery oak (Quercus harvardii), sand sagebrush (Artemisia filifolia) and grasses was identified through photo interpretation of canopy shape and texture information in 120 aerial photomosaics from 1930-1970 that were acquired for 22 sites within the study region. Distribution of the SSOP was analyzed using satellite imagery (Landsat time series) from 1975 to 2015. For New Mexico, results show that sand shinnery oak, sand sagebrush and grass associations have been reduced from an approximate potential extent of ~1,800,000 ha by the late19th century (mostly Lea, Eddy, Chaves and Roosevelt counties) to ~900,000 ha in 1977 and then to ~600,000 by 2015, with decreasing dominance of sand shinnery oak in the last 40 years. In Texas, sand shinnery oak prairies have been reduced to less than 25% of its historical potential extent (from 2,000,000 ha to 430,000 ha). Remnants are found mostly in the Yoakum, Terry, Cochran, Hockley, Andrews and Gaines counties. Results from aerial and satellite image analysis show a decrease (35%) in areas dominated by sand shinnery oak during the last four decades in Texas, with isolated recovery of mixed vegetation communities in Cochran and Terry counties.

CLASSIFYING IMAGES: FROM MILLIMETER PIXELS TO TWENTY MILLION ACRES

Since 2010 Open Range Consulting has been taking overhead pictures of vegetation throughout the Western United States. These images which are up to 20 million pixels a piece now number over 5,000.  These images have a classification accuracy of greater than 90 percent. The classified images are used to create landscape wide maps covering millions of acres. They are highly accurate and depict functional vegetation cover attributes which are vital to land management. These maps we develop are used by agencies, businesses, and ranch personnel to report on vegetative characteristics and drive management decisions.  Accurately classifying each image was taking an hour per image, this is a problem. Streamlining this process could significantly reduce costs and the timeliness of this product. 

APPLICATION OF VIRTUAL REALITY TECHNOLOGIES IN SUPPORT OF REDD RELATED WOODLAND INVENTORIES. Robert A. Washington-Allen*¹, Natasha S. Ribeiro², Paulo A. Raposo³, Connor W. Vermilyea⁴, Robert Friedrichs⁵, Aires Banze⁶, Kyle Landolt⁴; ¹University of Nevada, Reno, Reno, NV, ²Eduardo Mondlane University, Maputo, Mozambique, ³University of Tennessee, Knoxville, TN, ⁴Oak Ridge National Laboratory, Oak Ridge, TN, ⁵Oak Ridge National Laborator, Oak Ridge, TN, ⁶Universidade Nova de Lisboa, Lisbon, Portugal

We have been conducting the monitoring, verification, and reporting (MRV) of carbon dynamics for a Level II Reduction of Emissions from Deforestation & Degradation in Developing Countries (REDD) related project in the 14th largest protected area in the world: the 42,000 km2 Niassa National Reserve’s (NNR) in northern Mozambique. We have conducted this study for an 11-year period,  between 2005-2015, and have found that anthropogenic fires, declining elephant herbivory (due to increased poaching between 2009 to 2015), and water limitations drive the Miombo woodland’s carbon dynamics. We collected data on miombo woodland structure (including fuel extraction and elephant and fire damage), composition, demographics including recruitment and mortality, biomass, and vegetation and soil carbon stock density (CSD) using traditional forest inventory methods in fifty 30-m diameter circular sample plots. In 2015, we tested a new survey technology: a 1550-nm wavelength terrestrial laser scanner (TLS) with a 2.25-mm laser spot size and a ranging accuracy of ± 2 mm over 350-m to collect three-dimensional (3-D) virtual woodland structural data in 35 of the 50 plots at a sampling rate of 976,000 points sec-1 at a spacing of 8-mm 10 m-1 range. The resulting GPS positioned 5 scans per plot were registered and consolidated into a single 3-D ~23 million points cloud for each period with a mean spacing of 1-cm. We compared TLS data for 11 of these plots to field collected tree height and DBH for 10 to 15 trees per plot. We both manually and automatically measured forest parameters in the 3-D virtual environment using commercial and open source software. Manual measures of tree height and DBH were also taken directly from the trees within the 3-D TLS environment projected onto reality by a Microsoft HoloLens. We found that the HoloLens augmented TLS measures of DBH and height were highly related to field measures ( r2 > 0.85). We estimated overall mean tree density was > 550 trees ha-1 distributed across 79 species.Consistent with broadscale remote sensing studies of “greening trends in Africa”, above ground biomass is increasing over the last 11 years suggesting that NNR represents a potential carbon sink.

THE HERBAGE ASSESSED REMOTELY TO PREDICT ENVIRONMENTAL RISK (HARPER) PROCESS IMPROVES EFFECTIVENESS OF RANGELAND MONITORING.