Rangeland Health Workshop - Breakout Session 1

Updated 02/27/2009

Points for discussion:

In terms of the two presentations on scaling from fine-scale ground data to coarse scale remotely sensed data (see presentations by Washington-Allen and Jensen)?

1. What are the limits?
2. What are the opportunities?
3. What other options exist?
4. Are the basic principles sound?

Key points presented by group 1:

Several factors currently limit the viability of using remote sensing to monitor rangeland health. Major factors are the cost of imagery, where the greater resolution equals greater cost, and associated with that the availability of imagery and the frequency of data collection. Once data are collected, data processing can also be problematic. The spectral and spatial resolution of available data are often lacking, and there is a paucity of available ground based data with which to make correlations -- for example, weather stations are often few and far between in areas of interest. We are also often left having to use site specific data to make extrapolations to other sites.

Often, indicators reflect the current status of a site, a Ïpost mortemÓ of a site that has already been degraded. In order to use indicators in a predictive fashion we need a better understanding of ecosystem processes and the relationship of indicators to these processes. Decisions about whether an indicator is reflecting a ÏgoodÓ or ÏbadÓ change in the system ultimately depends on the values placed on different states.

Despite the limitations above, opportunities exist for using remote sensing to monitor rangeland health. The hardware and software needed to obtain imagery and process data exist or can be developed. Geomorphometry yields DEMÌs, and radar and laser based technologies (LIDAR, SAC) are currently in use. Coupled with advances in ecological theory and work on ecotones (B. Milne, Gosz), scaling (Gosz, Jornada), and ground based reference sites such as the Long Term Ecological Research (LTER) sites and appropriate experimental designs, the core requirements for distributing finer scale data to coarser scales now exist.

Some programs such as the FEWS program, USAID, and the Drought Alertness program already utilize some of these technologies. However, we need to make data sharing easier through standardization to meet sub regional and regional needs. This will also enhance the availability of information to judge the success of a program. We also need to increase our efforts in educating producers and consumers of this information.

Summary: While there are many limits to scaling up from fine to coarse level, there are also many opportunities that can yield information for rangeland health assessment. The choice of indicators of ecosystem health are value driven. However, the indicators themselves must be based on basic scientific principles expressed in the fields of radiation physics, soil physics, ecological succession and plant and soil physiology.

Key points addressed by group 2:

The group discussion began by considering two key issues brought up by the earlier presentations: Can point (ground based, site specific) issues be related to large areas? And, in terms of remotely sensed data, can the available spatial data be substituted for the relatively unavailable temporal data?

Soil loss is the most critical factor that affects the long term stability of any ecosystem. Most ecosystems have evolved resistance to any substantial resource loss during normal climatic cycles. However, infrequent and unpredictable catastrophic events (climatic, biotic, or other) will cause substantial resource loss from most ecosystems. Soil loss, for example, is often driven by hydrologic processes during large volume, intense rain events.

In order to accurately scale up from fine-scale ground based data to coarse-scale remotely sensed data we must take into consideration several factors including:

1. The accuracy of the data (the remotely sensed imagery). There are many factors that can influence the accuracy of remote sensing data, including the timing of ÏgreeningÓ of vegetation and atmospheric conditions.
2. The sensitivity of the information collected. Indicators of ecosystem health may have different sensitivities to the various components that make up Ïecosystem healthÓ. Thus the choice of indicators must match the questions being asked, and the data interpreted in the context of the sensitivity of the indicator(s) being used.
3. Interactions between management and climate. The suggestion that management effects be separated from climatic effects is untenable. Management decisions must take into account climatic probabilities and be able to respond to significant changes in climate.

Any methodology must include components for risk assessment, error analysis and take into account socio-economic factors which often tend to be neglected when ecosystem health is evaluated. A very necessary component to this approach is the definition of threshold values which serve as markers that activate various management options. Whatever methodology is finally put into place to monitor ecosystem health, be it at the landscape, regional or global levels, the primary aim must be to maximize management options.

Summary: Soil loss is the most critical factor that affects the long-term stability of any ecosystem, and most soil loss is driven by catastrophic events. Limits to using remote sensing imagery includes the accuracy of the data, the sensitivity of the information collected, and interactions between management and climate. Any methodology must also include components for risk assessment, error analysis, socio-economic factors and the reliability of thresholds used for assessment.

Key points addressed by group 3:

A primary problem that needs to be addressed, even before beginning to tackle the problems of scaling from fine resolution to coarse resolution data is the question of how to define rangeland condition or health. There are too few boundaries to define the term rangeland condition, and there are even fewer boundaries with which to define rangeland health, even from a narrow biological standpoint. A possible solution to this dilema is to translate [define] how a good condition community functions. However, lands of intermediate condition, those that are at risk but not yet severely degraded, may be misclassified. Evaluations of the health of ecosystems are often not linked to overall management decisions and perceptions. Additionally, thresholds of undesirable change can not be applied similarly to all ecosystems. Nonetheless, despite numerous problems with the concept of ecosystem health, there are basic facets of site integrity that can be measured. Therefore, can the integrity of a site, i.e. the ability of a site to retain soil and water resources and maintain a stable biotic component, be used as a way to measure ecosystem health? An alternative approach would be to consider degrees of integrity rather than attempt to define the meaning of ecosystem ÏhealthÓ. This would enable us to focus on those parameters that can be measured and aggregated to yield an understanding of the stability and sustainability of a site.

The identification of areas at risk would allow us to invest the limited resources on triage for ecosystems that can be recovered, not on those that are so heavily degraded that recovery is cost-prohibitive. For example, in a small watershed the focus could be on erosion control and the development of a few criteria about the landscape, to decide what the land needs (e.g. changes in land management?), and what can be reasonably done. Thus the focus needs to be on site integrity and not on a utilitarian demand. Overall, we need organized information which can be analyzed to yield estimates of ecosystem integrity at the local level.

Summary: Rangeland health may be defined on the basis of how a community, in good condition, functions. While this raises its own problems of how to define thresholds and the risk of misclassifying sites of intermediate health, there are basic facets of site integrity that can be measured. A preferred method may be to classify on the basis of degrees of site integrity, rather than an overall health rating. Identification of areas at risk would allow investment of limited resources on triage rather than the cost prohibitive revival of severely degraded ecosystems.

< Back to Agenda