Executive Summary

A strategy for monitoring post-fire seedings in the sagebrush steppe of the Intermountain West was developed and used to monitor four example fires in the Vale, Oregon District of the Bureau of Land Management (BLM). We began to develop a potential approach by (1) reviewing previous vegetation monitoring manuals produced by the Federal government to determine what techniques and approaches had been approved for use, and (2) monitoring a set of example fire rehabilitation projects from 2006 through 2008.

We reviewed seven vegetation monitoring manuals approved for use by the Federal government. From these seven manuals, we derived a set of design elements appropriate for monitoring post-fire rehabilitation and stabilization projects. These design elements consisted of objectives, stratification, control plots, random sampling, data quality, and statistical analysis. Additionally, we chose three quantitative vegetation field procedures that were objective and repeatable to be used in conjunction with these six design elements.

During the spring and summer of 2006 to 2008, U.S. Geological Survey personnel monitored vegetation in seven post-fire seeding treatments in four burned areas in the Vale district of the BLM in eastern Oregon. Treatments monitored included a native and non-native seeding in each of the Farewell Bend, Double Mountain, and Keeney Pass fires, and a native seeding at the Cow Hollow fire. All fires occurred in 2005.

There generally was a low level of plant establishment for all seedings by 2008. The quantitative objective established by the BLM was to achieve 5 seeded grass plants/m² by the end of 3 years as a result of the seeding. There was an estimated 3.97 and 6.28 plants/m² in 2006 and 1.06 and 0.85 plants/m² seeded perennial grasses in 2008 from the Keeney Pass non-native and native seeding, respectively. The Cow Hollow seeding resulted in the lowest establishment of perennial seeded grasses of the four project areas with 0.69 plants/m² in 2006 and 0.09 plants/m² in 2008. Density of seeded perennial grasses at the Double Mountain non-native and native seeding were 2.72 and 3.86 plants/m² in 2006 and 0.90 and 1.74 plants/m² in 2008, respectively. The Farewell Bend non-native seeding resulted in 5.62 plants/m² in 2006 and 0.42 plants/m² in 2008 while the native seeding had 2.22 seeded grass plants/m² in 2006 and 0.44 plants/m² by 2008. The primary reason for low level of establishment on most treatments except the Cow Hollow seeding was most likely the unfavorable timing and amount of precipitation in 2007 and 2008.

Measurements of density within the first 3 years provide the best estimate of initial seeding success. Increases in cover due to the seedings were not detectable in the first 3 years following seeding in this monitoring effort. Changes in cover resulting from the treatments may be detectable in cases where the seedings were very successful in the first 3 years following seeding, but in areas with lower annual average precipitation, may not occur consistently. As a result, cover of seeded species may not be a good indication of seeding success in the early years after treatment. However, cover is useful for monitoring initial patterns of abundance of naturally recovering vegetation, exotic annual grasses and forbs, and bare ground. Cover measurements at these four sites revealed patterns common to most of the treatment areas in cover of litter, bare ground, and exotic annuals in response to drill seeding and weather patterns. There was a rapid increase in litter at all treatments after the fire. Additionally, there was less litter in treatment plots than in the control plots in 2006 probably due to the mechanical action of the seed drill. There also was a corresponding decrease in bare ground from 2006 to 2008. Initially, higher bare ground cover at treatment plots appears to be due to the mechanical action of the seed drill.

Cover of annual grasses, primarily Bromus tectorum, increased from 2006 to 2007 and then decreased slightly in 2008. The highest cover and density of exotic annual grasses generally occurred in the second year following fire. Immediately after fire, lower densities of B. tectorum may emerge due to the loss of seed, but B. tectorum plants that do emerge often produce an abundance of seed due to high nutrient availability and reduced competition, resulting in higher densities the following year.

There was a consistent, negative linear relationship between the amount of cover of existing perennial grasses and annual grass cover at treatment areas. This relationship also was apparent in the gap data, and annual grass cover was greatest when basal gaps in the greater-than-200-cm size class were more frequent. The inverse relationship between cover of perennial and annual grasses suggests that post-fire seedings, when successful, can improve rangeland condition where annual invasive grasses are problematic.

Overall, quantitative objectives are a valuable part of monitoring the initial success of post-fire seedings. However, they need to be adapted for specific situations and areas. The potential of a particular area to reach a certain density or cover of desirable plant species (and the condition of the pre-fire plant community, for example, healthy or degraded) can be used to set initial objectives, which could be further modified by conditional statements that depend on environmental conditions after seeding. These conditional objectives may be developed to include a range of values rather than a specific target objective. Eventually, using data from many projects over time, a model could be developed to predict optimum seeding success over a range of conditions. Using such a dynamic approach to setting objectives would minimize the numbers of projects that are deemed failures due to unrealistic objectives or environmental factors that are outside the control of land management. Monitoring at the four areas from the Vale, Oregon District of the Bureau of Land Management that burned in 2005 also demonstrated potential uses and difficulties associated with monitoring ES&R (Emergency Stabilization and Rehabilitation) treatment effectiveness. Overall, the monitoring approach, combined with the quantitative techniques, performed reasonably well in burned areas previously dominated by sagebrush. Future monitoring efforts should take into account the logistical constraints of each design element and quantitative technique to arrive at the most cost-effective yet statistically valid monitoring plan. In the future, procedures that encompass more of the natural variability, either through sampling at more locations or incorporating the use of remote sensing may be able to capture more of the natural variability at the landscape scale. The value of the three quantitative techniques for interpreting success of post-fire seedings depends on the time frame in which they will be used. For the first 3 years following seeding (the period for which monitoring is usually funded), density is the most directly applicable measurement of treatment effect and is emphasized in this report. Changes in plant cover and basal-gap intercept measurements are small during the first 3 years and, when combined with environmental and observer variation, could not be used for determining success. As the seeding ages and plants become larger, however, comparison of cover and gap-intercept data between treatment and control plots can be used to determine long-term effects. Whether initial densities in the first 3 years correlate to later cover and basal-gap intercept measurements is unknown and warrants further investigation. In addition to assessing level of establishment at a variety of different post-fire seedings, the use of similar techniques to monitor Vale fires helped identify patterns common throughout multiple treatments. Consistent patterns of vegetation attributes identified in these four burned sites include the rate of accumulation of litter, the rate of decrease of bare ground, the inverse relationship between annual grass and forb cover, and the relationships between annual grasses with perennial grass cover and basal–gap intercept. Identifying additional patterns at a greater number of projects in a wider geographic area and correlating with site factors (such as soil, elevation, and climate) will aid efforts to improve seeding success through adaptive management.