PIs: Chris Heckman (Lead PI) and Nikolaus Correll (co-PI), Computer Science, University of Colorado; Nichole Barger, EBIO, University of Colorado.
Overview: Degraded ecosystems make up a significant amount of desert land in the western United States as well as globally. These ecosystems require active intervention, often through artificial seeding, in order to recover from this state and support plant diversity and productivity. We propose the active revegetation of desert landscapes through the deployment of autonomous robotic platforms with mobile manipulators that will learn to identify favorable seeding locations, navigate to these locations, and discern small variations in terrain that might affect later site effectiveness. To do this, we will develop new techniques in robotic navigation, environmental perception, and visual-tactile feedback. The end result of this project will be new ecosystems that can support vegetation and that are re-introduced into productivity through active intervention alongside ecologists.
Intellectual Merit: The proposed research will develop new techniques in degraded ecosystem recovery through visual-tactile perception, multimodal learning, and exotic prior integration through currently-available data. The latter focus will involve the integration of low-resolution mapping data into high-resolution in situ data, such as that collected by high-resolution 3D lidar on an autonomous platform. Multimodal learning will be performed by learning correlating sensor features for effective revegetation via a variety of onboard sensors which will collect data through a rigorous series of field deployments on a multifunctional platform that has in large part already been developed by the proposal team. Visual-tactile perception serves as the crowning achievement of this proposal, wherein techniques for terrain compatibility determination will be combined in active interrogation of the environment and silty, sandy, and other environments will be classified. All of these features will be combined into a global planner that will maximize the likelihood of ecosystem recovery by determining where scarce resources (seeds and other active elements) will be deployed.
Broader Impacts: Beyond education, the proposed research will have significant impact on two major factors of environmental recovery. The first is in the deployment of autonomous cooperative agents that can support land use and management through autonomous revegetation, thereby increasing economic capability of federal lands that have previously been considered a loss for grazing capabilities. The second is the education and training of ecologists in these autonomous techniques, thereby providing future capability to researchers and field practitioners in scaling environmental recovery efforts into the future, with the potential impact to other environments. These elements will be included into K12, college, and graduate training, especially in the propagation of machine learning techniques to accomplish these goals, even if they are not included in an embodied autonomy deployment.
PIs: Nichole Barger, University of Colorado; Mike Duniway, USGS; Akasha Faist, New Mexico State University; Dave Hoover, USDA-ARS
Livestock grazing is a critically important livelihood that supports people and communities across the globe. Inappropriate grazing land management, however, has led to degradation of about 20 percent of the world’s pastures and rangelands resulting in significant declines in a broad range of ecosystem services. The prevalence of such undesired ecosystem change, coupled with the more recent increased importance of other ecosystem services such as recreation and tourism has resulted in new interest in effectively restoring the broad range of ecosystem services these lands provide. The many examples of unsuccessful restoration projects focused on degraded grazing lands highlights the challenge of restoring native plant communities and soil resources--challenges often driven by unstable and highly eroding soils and the lack of predictable rainfall required for native plant establishment. Recent advances in soil stabilization products and field application methods and newly developed approaches to enhance seed germination and establishment have the potential to increase restoration success on degraded grazing lands. Following this, our overarching goal in this project is to employ the latest technologies and implementation strategies in soil stabilization and native plant establishment to restore important ecosystem services while maintaining agricultural services provided by rangelands. Our primary research objectives in this project are to: 1) quantify changes in key regulating (air quality maintenance, erosion control, water regulation), and supporting (net primary production, soil fertility, soil formation) ecosystem services with a combination of innovative native plant seeding and soil stabilization strategies and 2) evaluate the ecosystem service trade-offs of maintaining agricultural activities by returning cattle to the landscape after variable rest periods. In this project we will implement two soil stabilization strategies that were recently shown to be highly effective in stabilizing eroding soil surfaces (biological soil crust inoculum with polyacrylamides and connectivity modifiers) in combination with native plant seeding. Livestock grazing will then be introduced at the typical 2-year rest period and a longer rest period of > 2 year period. To compliment the research component, we have developed an extension component to deliver our science-based knowledge of grazing land restoration not only to allow for more informed management decisions but to develop restoration practices that are feasible and cost effective. Our extension objectives are to: 1) identify the best candidate restoration practices that not only increase delivery of key ecosystem services but are logistically feasible for land managers and private landowners and 2) share knowledge of best practices in increasing ecosystems services while maintaining agricultural activities with a range of federal and state land managers and private land owners on the Colorado Plateau. We will create an advisory committee with members from the broad community of federal land managers, private landowners, and rangeland professionals. We will seek ongoing guidance from the advisory committee on the feasibility of our restoration strategies and the most effective outlets in which to disseminate our findings across different stakeholder groups in the region.
PI: Nichole Barger, University of Colorado
Increasing human population pressures and land use intensification across dryland ecosystems has resulted in extensive land degradation or “desertification” across many of these regions. Once degraded, recovery times in dryland ecosystems range from decades to centuries, which presents a significant challenge to restoring the ecological structure, function, and diversity of these ecosystems. In undisturbed desert environments, biological soil crusts (‘biocrusts’) are an important functional component of these environments influencing soil stability, hydrology and nutrient availability. Biocrusts are diverse communities of cyanobacteria, mosses, lichens, and fungi that colonize the top few mm of soil surfaces and make up a high percent of the living ground cover in dryland systems where vascular plants are sparse. Soil surface disturbance due to human activities results in dramatic losses in biocrust communities altering soil erosion and nutrient and water cycles. Biocrusts recovery after soil surface disturbance is slow, thus there may be limited capacity for natural recovery. In response to these slow recovery times for biocrust communities, recent advances have been made in developing biocrust restoration strategies for degraded dryland soils. One issue in restoring biocrust communities in dryland ecosystems is monitoring biocrust recovery at the landscape sale. Current methods involve time consuming and lab intensive ground based measures that are limited in their spatial extent. Thus there is a critical need to develop methods to monitor biocrust restoration and recovery at the landscape scale. Our aim in this proposal is to develop high resolution remote sensing capabilities to monitor recovery of biocrust communities on degraded dryland soils. Working in collaboration with the Integrated Remote and In-Situ Sensing (IRISS) group at University of Colorado at Boulder (UCB) and federal scientists from U.S. Department of Agriculture and the U.S. Geological Survey, we will deploy unmanned aerial vehicles (UAVs) equipped with multispectral sensors to measure the reflectance spectra of early successional biocrust communities that were generated from lab grown organisms at two western U.S. desert research sites. The collaboration with IRISS, a partner with the UCB Grand Challenge, will take my lab in a new and innovative research direction and expand the tools we have available to monitor the recovery of degraded landscapes.
PIs: Nichole Barger, University of Colorado
Akasha Faist, University of Colorado
BLM Canyon Country Fire Zone Collaborators:
Gabe Bissonettte Jason Kirks Funding Source: State of Utah, Department of Natural Resources,
Division of Wildlife Resources Time Period: 5/16 – 5/17
PI: Nichole Barger, University of Colorado
Co-PI: Ferran Garcia-Pichel (ASU), Matthew Bowker (NAU), Jayne Belnap (USGS)
Co-PI: Mike Duniway (USGS), and Sasha Reed (USGS)
Funding Source: SERDP, Strategic Environmental Research and Development
Time Period: 3/1/13 - 2/29/18
Biological soil crusts ('biocrusts') are communities of microorganisms that develop on soil surfaces and are a critically important functional component of dryland systems of the globe. They are often associated with increased soil nutrient and water retention—resources that are highly limiting to plant productivity in these ecosystems. But most importantly, biocrusts stabilize soil surfaces against wind and water erosion. While resilient to wind and water erosion, biocrusts are highly susceptible to compressional forces, such as those generated from foot and vehicle traffic associated with ground-based military training activities. Due to the functional importance of biocrust communities to the ecological functioning of dryland ecosystems there is keen interest in restoring these communities. Thus our overarching research objective in this project is to facilitate the recovery of degraded arid and semi-arid Department of Defense (DoD) lands by restoring biocrust communities. In this project we will: 1) establish a biocrust nursery as an inoculum testing and supply center for biocrust restoration 2) identify successful field application methods of biocrust inoculum in a series offield trials 3) evaluate soil and plant responses to biocrust restoration in multi-factorial field experiments and 4) share knowledge of biocrust restoration success and challenges with DoD and federal land managers.
PIs: Jason Neff, Nichole Barger, Lisa Dilling, Jana Milford
Funding source: Funded by USDA in a joint NASA/USDA call in Carbon Cycle Science, Research Opportunities in Space and Earth Sciences
Time Period: 5/2011 - 4/2015
In this project we take an integrated multi-scale approach to the evaluation of carbon stocks and fate under different management on public lands in the Intermountain West. We developed this project in close collaboration with the US Department of Agriculture (USDA) Forest Service and the Department of Interior (DOI) Bureau of Land Management. There are four primary objectives in our work and these are 1) the Analysis of forest and woodland carbon stocks across elevational gradients in Southwest Colorado and Eastern Utah, 2) the Evaluation of changes in forest and woodland carbon associated with land management activities including fire mitigation, and forestry, 3) the Projection of management impacts on carbon balance using ecosystem carbon models, and 4) the Integration of carbon into land management decisions including evaluation of potential implications of different federal approaches to carbon management. The work uses remote sensing and field biomass measurements to develop regional carbon maps for federal land management centers in Colorado and Utah. At these sites, we also monitor changes in carbon stocks with management using ground based, data-collection techniques and use existing USFS models of forest biomass change to project carbon accumulation through time. In all of our analysis, monitoring, and modeling work, we deal directly and explicitly with error and uncertainty and all spatial or temporal projections of carbon will include uncertainty bounds. Finally this study includes the development of new decision support tools for carbon management that will reside in federal land management databases at our study sites. We recognize that federal carbon management policy and interaction with existing land management policy is uncertain and likely to remain so for some time to come. Accordingly, our study includes not just the development of decision support tools but also an explicit study of decision-making. This project will help to identify the barriers and complexities involved in integration of carbon management into broader federal land management goals.
United States PIs: Dr. Jason Neff, Geosciences Department, CU Boulder Dr. Nichole Barger, Ecology and Evolutionary Biology, CU Boulder South Africa
Collaborating Scientist: Dr. John Stockton, Department of Geological Sciences, University of Cape Town
Funding Source: Mellon Foundation Time Period: 8/2008 - 10/2013
In this project we outline a study of nitrogen (N) cycling across the diverse flora of the Western Cape province of South Africa. This region of South Africa has a remarkable diversity of floral assemblages that are arrayed across (and associated with) an equally diverse collection of geologic settings. We propose a study of nitrogen inputs and cycling that will be closely tied to an ongoing NRF funded study led by John Stockton of the University of Cape Town that is focused on understanding the role of geologic and geochemical variation in the control of floral composition in this region. This current project is oriented around the study of the inputs and cycling of macro and micronutrients (excluding N) in ecosystems that range from the Strandveld coastal ecosystems to the mountain Fynbos ecosystems of Table Mountain National Park. This existing set of sites and studies offers a remarkable opportunity to not only examine N cycling across a range of settings, but also to better understand the interaction between the N cycle and the cycling of other P and the micronutrients. In this project we will specifically focus on the response of N fixation to variation in soil nutrient status and marine aerosol N input into Western Cape Ecosystems. We expect to find increasing reliance on N fixation derived N in the Fynbos ecosystems compared to the coastal settings where marine aerosol inputs are higher. However, the extraordinary diversity of geochemical settings in this area (and the corresponding variation in vegetation cover) suggests the possibility that micronutrient and P availability may interact with, and potentially control, the biological fixation of N to these ecosystems. The combination of the proposed N studies and the ongoing geochemical and floristic opportunities offer a unique opportunity to examine the interactions between the major biogeochemical cycles and the role that these interactions play in structural biological communities.
PI: Nichole Barger, University of Colorado
Co-PI: Mark Miller, US Geological Survey Southwest Biological Science Center, Kanab, UT
Co-PI: Jeff Herrick, USDA ARS. Las Cruces, NM
Funding Source: USDA Cooperative State Research Extension and Education Services (CSREES)-Managed Ecosystems program
Time Period: 6/1/08-5/31/2012
The focus of this integrated research program is the development of approaches and tools for restoration and long-term sustainable management of pinyon-juniper (P-J) ecosystems that are based on principles of adaptive ecosystem management. Our research objectives are 1) to identify the P-J treatment strategies that are most effective in overcoming constraints to understory restoration and 2) evaluate the impacts of P-J treatment strategies on important ecosystem attributes and functions. By understanding the important constraints to restoring understory vegetation and the potential risks associated with different treatment strategies, we expect that the long-term outcome of this research will be more effective restoration of understory vegetation communities in combination with improved ecosystem conditions. Building on the research program, our extension objectives are 1) to use the results of our research to develop a series of science-based decision making models and 2) to provide training to land managers and other stakeholders in using these models to plan and implement future projects. Training land managers to use science-based decision making models that are grounded in fundamental principles of ecology should lead to increased awareness of ecological processes and how these processes function across managed landscapes. Following this, application of such models to managing P-J woodlands should lead to more scientifically defensible management plans and protection from litigation, which is always a management concern when working with a range of stakeholders that differ in how they believe these landscapes should be managed. Our educational objectives are to provide both classroom and field-based training to undergraduate and graduate students in methods of adaptive ecosystem management. This will be accomplished by developing an Ecosystem Management course for undergraduates, directing a summer undergraduate research and management internship program and training graduate students to direct their research to address the needs of land management goals. Our goal in providing these educational opportunities to students is to forge stronger ties between future ecologists and land managers with the outcome of bridging the knowledge gap between academic science and applied management issues.
PI: Greg Asner, Department of Global Ecology, Carnegie Institution of Washington, Stanford University
Co-PI: Jason Neff, Department of Geological Sciences and Environmental Studies, University of Colorado
Co-PI: Nichole Barger, University of Colorado Funding Source: NASA North American Carbon Program
Time Period: 2/2004-5/2008
Arid and semi-arid ecosystems cover about 3.4 million square kilometers of North America. The spatial patterns and abundance of herbaceous and woody plants throughout these regions are determined by bio-climatic conditions, topography, soil properties, and disturbance regimes. During the past century, the balance between woody and herbaceous plants has shifted in many U.S. dryland ecosystems to favor trees and shrubs. Recent syntheses suggest that woody plant encroachment contributes significantly to a North American carbon sink. However, current estimates of gross or net rates of woody cover change in the western U.S. are crude and have not been linked to changes in C storage, thus the potential contribution of woody encroachment to the U.S. carbon budget remains elusive. While our regional knowledge of current woody cover distributions, woody vegetation changes over time, and ecosystem C responses is very crude, what we do know has largely come from studies in “lowland” arid and semi-arid regions. Pinyon-juniper (P-J) woodlands are among the least understood systems in terms of woody encroachment and thickening. We do not know: (1) the current distribution, cover and carbon stocks of P-J ecosystems in a ~ 500,000 km2 portion of the Southwest; (2) regional rates of P-J cover and carbon change in relation to soils and grazing history; (3) soil organic carbon responses to changes in P-J cover; and (4) how to model current and future distributions of carbon stores in the P-J region.The broad goal of this project is to quantify and understand regional effects woody encroachment on carbon storage in pinyon-juniper ecosystems of the Southwest U.S. We will quantify the contribution of woody encroachment in these systems to a proposed U.S. carbon sink and the interaction of aboveground changes in carbon with direct grazing impacts on soil carbon. We will combine multi-platform remote sensing, field biogeochemical and dendrochronology studies, and spatio-temporal