ARS | Publication request: SENSITIVITY TESTING OF INDICATORS OF ECOSYSTEM HEALTH

Submitted to: Ecosystem Health
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: November 1, 1996
Publication Date: N/A

Interpretive Summary: Before we use any indicator of ecosystem health we must first know how sensitive it is to changes in ecosystem health. We developed and evaluated three indicators: 1) the size of bare patches, 2) the proportion of total grass cover that is contributed by long-lived grasses, and 3) soil stability as measured by a slake test. In desert rangelands livestock congregate at watering points scattered throughout the landscape, causing high levels of disturbance at these points and decreasing with distance from the watering points. Indicator values were obtained at four distances from the watering points up to 0.7 miles (= grazing gradients). We then obtained indicator values from fenced exclosures that had been excluded from grazing for 50 years, and adjacent grazed areas. These values were compared to values from the grazing gradients. All three indicators were sensitive to livestock induced disturbance, with ungrazed exclosures having gindicator values in the same range as grazing gradient locations furthest from watering points, and the adjacent grazed areas having indicator values similar to 0.1 to 0.2 miles from watering points. The slake test indicator appeared to be most sensitive at the early stages of degredation in ecosystem health while the size of bare patches and the proportion of long-lived grasses were sensitive over a wider range.

Technical Abstract: The sensitivities of three indicators of ecosystem health were evaluated at several sites in the Jornada Basin of the Chihuahuan Desert. The size of bare patches, proportion of total grass cover contributed by long-lived perennial grasses and soil stability are interdependent indicators of ecosystem functions related to the retention and use of water and nutrients. Sensitivity tests were chosen using data collected along disturbance gradients and then tested using independent, ungrazed exclosures and adjacent grazed pastures. The mean size of bare soil patches was sensitive to anthropogenic disturbance. When bare soil patch data were transformed using natural logarithms, the skewness of the frequency distribution weighted by mean bare patch size could be used to indicate early disturbance to the ecosystem. The proportion of total vegetation that was long-lived also was sensitive to anthropogenic disturbance, and appears to be a good indicator of ecosystem degradation. The slake test for soil surface stability was extremely sensitive to disturbance and may serve as an early-warning indicator of soil degradation for the coarse-textured soils which were evaluated.