Fire rehabilitation effectiveness: a chronosequence approach for the Great Basin

Abstract

Federal land management agencies have invested heavily in seeding vegetation for emergency stabilization and rehabilitation (ES&R) of non-forested lands. ES&R projects are implemented to reduce post-fire dominance of non-native annual grasses, minimize probability of recurrent fire, quickly recover lost habitat for sensitive species, and ultimately result in plant communities with desirable characteristics including resistance to invasive species and resilience or ability to recover following disturbance. Land managers lack scientific evidence to verify whether seeding non-forested lands achieves their desired long-term ES&R objectives. The overall objective of our investigation is to determine if ES&R projects increase perennial plant cover, improve community composition, decrease invasive annual plant cover and result in a more desirable fuel structure relative to no treatment following fires while potentially providing habitat for Greater Sage-Grouse, a species of management concern. In addition, we provide the locations and baseline vegetation data for further studies relating to ES&R project impacts. We examined effects of seeding treatments (drill and broadcast) vs. no seeding on biotic and abiotic (bare ground and litter) variables for the dominant climate regimes and ecological types within the Great Basin. We attempted to determine seeding effectiveness to provide desired plant species cover while restricting non-native annual grass cover relative to post-treatment precipitation, post-treatment grazing level and time-since-seeding. Seedings were randomly sampled from all known post-fire seedings that occurred in the four-state area of Idaho, Nevada, Oregon and Utah. Sampling locations were stratified by major land resource area, precipitation, and loam-dominated soils to ensure an adequate spread of locations to provide inference of our findings to similar lands throughout the Great Basin. Nearly 100 sites were located that contained an ES&R project. Of these sites, 61 were seeded by using a drill, 27 were broadcast aerially, and 12 had a combination of both. We randomly sampled three burned and seeded, burned and unseeded, and unburned and unseeded locations in the vicinity of the fire, each within the same ecological site. We measured foliar cover of all plant functional groups (perennial or annual, shrub, grass, forb, native or introduced), biological soil crusts, and abiotic (bare soil and litter) variables using the line-point intercept protocol. Fuel loads and horizontal fuel continuity were measured. We applied linear mixed models to response variables (cover and density of plant groups) relative to the dependent variables (seeding treatments and precipitation/temperature relationships. Post-fire strengths with native perennial grasses or shrubs in mixes did not increase density or cover of these groups significantly relative to unseeded, burned areas. Seeded non-native perennial grasses and the shrub Bassia prostrata were effective in providing more cover in aerial and drill seedings. Seeded non-native perennial grass cover increased with increased annual precipitation regardless of seeding type. Seeding native shrubs, particularly Artemisia tridentata, did not significantly increase shrub cover in burned areas. Cover of undesirable non-native annual grasses was lower in drill seedings relative to unseeded areas but only at higher elevations. Seeding effectiveness after wildfire is unpredictable in drier, low elevation environments, and our findings indicate management objectives are more likely met when focusing efforts on higher elevation or higher precipitation locations where establishment of perennial grasses is more likely. On sites where potential for invasion and dominance of non-native annuals is high, such as lower and drier sites, intensive methods of restoration that include invasive plant control before seeding may be required. Where establishment of native perennial plants is the goal, managers might consider using native-only seed mixtures, because we found that the non-native perennials typically used in Great Basin restoration efforts are selected for their competitive nature and may reduce establishment of less competitive native species. Although we attempted to include information on livestock grazing history after seedings, we were unable to extract sufficient data from files to address this topic that may play an additional role in understanding native plant abundance post-fire seeding. Evaluation of drill and aerial seeding effects on fuel characteristics focused on two metrics that are standard inputs for fire behavior models, fuel load and fuel continuity. Fuel loads were evaluated separately for total fuel load biomass, and the individual components that sum to total biomass, namely herbaceous, shrub, shrub:herbaceous ratio, litter, 10-hour, and 100-hour fuel biomasses. Fuel continuity was evaluated using the following cover categories, total, annual grass, annual forb, perennial forb perennial grass, shrub, litter, vegetative interspace, and perennial interspace. Drill seeding did not affect fuel loads, except to reduce 10-hour fuels, probably due to mechanical destruction of dead and down fuels by the drill seeding equipment. Drill seeding did affect fuel continuity, specifically decreasing total plant cover by increasing perennial grass cover which suppressed annual grass and litter production resulting in a net decrease in continuity, but only at the elevations above approximately 1500m. Aerial seeding had no effect on any fuel load or fuel continuity category. For the Greater Sage-Grouse habitat study, we developed multi-scale empirical models of sage-grouse occupancy in 211 randomly located plots within a 40 million ha portion of the species’ range. We then used these models to predict sage-grouse habitat quality at 101 ES&R seeding projects. We compared conditions at restoration sites to published habitat guidelines. Sage-grouse occupancy was positively related to plot- and landscape-level dwarf sagebrush (Artemisia arbuscula, A. nova, A. tripartita) and big sagebrush steppe, and negatively associated with non-native grass and human development. The predicted probability of sage-grouse occupancy at treated plots was low on average (0.07–0.09) and was not significantly different from burned areas that had not been treated. Restoration was more often successful at higher elevation sites with low annual temperatures, high spring precipitation, and high plant diversity. No plots seeded after fire (n=313) met all overstory guidelines for breeding habitats, but approximately 50% met understory guidelines, particularly for perennial grasses. This trend was similar for summer habitat. Ninety-eight percent of treated plots did not meet winter habitat guidelines. Restoration actions in burned areas did not increase the probability of meeting most guideline criteria. The probability of meeting guidelines was influenced by a latitudinal gradient, local climate, and topography. Post-fire seeding treatments in Great Basin sagebrush shrublands generally have not created high quality habitat for sage-grouse. Understory conditions are more likely to be adequate than those of overstory, but in unfavorable climates, establishing forbs and reducing cheatgrass dominance is unlikely. Reestablishing sagebrush cover will require more than 20 years using the restoration methods of the past two decades. Given current fire frequencies and restoration capabilities, protection of landscapes containing a mix of dwarf sagebrush and big sagebrush steppe, minimal human development, and low non-native plant cover may provide the best opportunity for conservation of sage-grouse habitats. Our database of ES&R locations has used the Land Treatment Digital Library to archive data and location information regarding our study (see Pilliod and Welty 2013). This has contributed to two additional studies. One examined the potential spread of Bassia prostrata (aka Kochia prostrata; forage kochia) from ES&R project locations (Gray and Muir 2013). The second used remote sensing to determine the phenology of vegetation green-up on post-fire seeded sites (Sankey et al. 2013).