Grantee Research Project Results
Final Report: Characterization of Ecological Integrity of Commercially Grazed Rangelands Using Remote Sensing-based Ecological Indicators
EPA Grant Number: R826112Title: Characterization of Ecological Integrity of Commercially Grazed Rangelands Using Remote Sensing-based Ecological Indicators
Investigators: West, Neil E. , Washington-Allen, Robert , Ramsey, R. Douglas
Institution: Utah State University
EPA Project Officer: Packard, Benjamin H
Project Period: November 1, 1997 through September 30, 1999
Project Amount: $340,617
RFA: Ecosystem Indicators (1997) RFA Text | Recipients Lists
Research Category: Ecological Indicators/Assessment/Restoration , Aquatic Ecosystems
Objective:
The purpose of this research project was to assess the ecological integrity of a semi-arid landscape grazed by domestic and wild animals using modern views of land degradation, succession, landscape ecology theory, and retrospective analyses of remotely-sensed LANDSAT images and geographic information systems (GIS). The objectives of our study were to: (1) develop software to process and standardize large multi-temporal satellite datasets; (2) describe the behavior and trend of remote sensing-based ecological metrics; (3) retrodict the relationship of these metrics to land management (grazing, presented fire, etc.), wild fire, and climatic constraints; and (4) develop software to analyze the effects of grazing along grazing gradients.
Summary/Accomplishments (Outputs/Outcomes):
If the goal for managing sustainable rangelands is to simultaneously achieve a balance between positive financial returns to the grazier, while maintaining the capacity of that land to continue to yield income as well as maintain its native flora and fauna, high-quality water, and other ecosystem services, then monitoring is essential to detect changes. Corrective actions must be applied when necessary. Traditional point-based rangeland assessments are usually focused on plant species composition and surface soil characteristics. While providing useful information, these traditional approaches have generally failed because obtaining timely, repeatable, and statistically adequate time series was too demanding of technical expertise and correct timing of data collection and budgets to complete adequate analyses and reporting. Because these prerequisites are diminishing, we need to explore alternatives to monitoring rangelands. Furthermore, we need methods that will provide answers to the conditions and trends that exist on larger areas that are of concern to other sectors of the public (e.g., regulatory agencies, environmental organizations).
Our study site was the 54,000 hectares (ha), lower elevation, drier eastern half of the Deseret Land and Livestock Ranch (DLL) in northeastern Utah where livestock have been raised over the past 109 years. We first developed a GIS with physical, biological, and administrative characteristics, including historical and recent records of the actions taken by the ranch managers. These records are particularly thorough for the past 17 years when the ranch assumed control from the current owners. Access to private records on kinds and numbers of livestock and where they were at particular times, prescribed fires, herbicide and reseeding activities are essential for dependable interpretations of rangeland trends and conditions, though exceedingly hard to acquire. Because big game animals are a growing component of the income to DLL, wildlife biologists are in residence at the ranch, and data on kinds, number, timing, and places of use by these animals also were made available to us, an even more unusual occurrence than the livestock data. We were able to lay the animal use data over the locations of waterpoints, ecological sites, fence lines (delineating pastures), as well as all the other physical, biological, and cultural features (57 layers total) that we incorporated into the GIS for DLL.
The four major ecological indicators that we measured from 1972 to 1998 (27 years) were: soil-adjusted vegetation index (SAVI); metrics of landscape composition and patterns, such as patchiness using FRAGSTATS, evidence of soil erosion using a soil stability index (SSI), and resilience of plant production in relation to grazing gradients (piospheres). Our basic approach was to describe the behavior and trend of the time series of the ecological indicators, and secondly, to retrodict their relationships to historical land management, fire, and climate. Dynamical systems analysis, the study of time-evolving processes was used to characterize the historical responses of these indicators. Linear regression, non-linear curve-fitting, time series analyses, and catastrophe theory were used to relate indicator responses to constraints.
We found that: (1) there was too little collective impact by fire to be detected at the coarsest scales of observation; (2) the sagebrush steppe dominated portion of DLL was constrained linearly by climate and non-linearly by herbivory; (3) shrub cover increased from 1972 to 1973 and then remained near that level through 1997; (4) grassland cover was greater than shrubland cover initially, but has been declining since 1972; (5) the landscape abruptly showed lower dispersion and low contagion after 1990, conditions indicative of fragmentation relative to earlier larger, clustered patches; (6) patches of bare soil were stable from 1972 to 1997; (7) the historical decline in grass cover was coincident with an increase in cattle number (neither stocking rate or climate had a significant effect on the observed changes in metrics, but the number of days grazed had the most significant fit with landscape metrics); although not significant, the increase in shrub cover was coincident with increasing aridity and cattle number; (8) the trend in soil erosion, as reflected in the SSI, agreed with the bare soil patch thematic results, reflecting stable conditions; (9) a threshold of coincident high stocking rates and drought in the early 1970s was detected, but an El Niño event and probably changes in grazing management averted permanent degradation; (10) we identified a decreasing trend in vegetation response in riparian corridors; and (11) we developed piosphere or grazing gradient analyses software, which allowed analysis of the impact of grazing at individual waterpoints aid separation of the effects of grazing from that of climate and fire. The piospheric analyses applied to DLL, however, did not yield consistent results, most likely because of the high density and increasing number of watering points and progressively smaller pastures erected by current managers.
We have shown that using LANDSAT imagery and GIS makes it possible to synoptically reconstruct the history of changes in amounts of plant cover and balances between plant growth forms since 1972 on the portion of a ranch dominated by sagebrush steppe or its management derivatives. Employment of a dynamical systems approach allowed the interactions of drought, grazing, and fire to be interpreted. The ranch provided information on the kinds and numbers of animals and where they were at particular times, as well as where and when cultivation, burning, pond and fence building took place. This information was pivotal in interpreting the vegetation, soil and landscape patchiness responses.
From the intense analysis that this study provided, we can now identify ways to make monitoring shortcuts to reduce the expense of such monitoring efforts in the future. For instance, most of the change in vegetation came during El Niño and La Niña events. Because these can be predicted up to about six months prior to occurrence, the manager can decrease his/her risks (e.g., by selling off excess animals before a drought connected with a La Niña period or seeding degraded stands during an El Niño event). Similarly, monitoring should be intensified during these periods. This would decrease the mounts of both imagery and field time required to detect most of the changes in these kind of systems.
Journal Articles:
No journal articles submitted with this report: View all 28 publications for this projectSupplemental Keywords:
ecological health, condition, trend, landscape metrics, geographic information systems, monitoring, satellite imagery, sagebrush steppe., RFA, Scientific Discipline, Water, Geographic Area, Ecosystem Protection/Environmental Exposure & Risk, Water & Watershed, Ecosystem/Assessment/Indicators, Ecosystem Protection, State, Ecological Effects - Environmental Exposure & Risk, Environmental Monitoring, Ecology and Ecosystems, Watersheds, Ecological Indicators, agricultural watershed, ecosystem integrity, Midwestern Watersheds, Utah (UT), ecosystem indicators, terrestrial, commercial rangeland, aquatic ecosystems, GIS, synoptic monitoring, water quality, multiscale assessment, agriculture ecosystemsRelevant Websites:
http://www.esd.ornl.gov/people/washington-allen/washington_allen.html Exit
Progress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.