Effects of sagebrush fire and fire surrogate treatments on a Great Basin seed bank community

Background/Question/Methods

The influence of sagebrush fuels and restoration treatments on Great Basin vegetation dynamics has been well documented but impacts of treatments on the seed bank community has received little attention. Effects of fuels reduction treatments (prescribed fire, tebuthiuron herbicide, and imazapic herbicide) on seed bank community composition and dominant seed bank species densities were evaluated. In addition, we determined whether the pre-treatment seed bank or the aboveground vegetation was more similar to the post-treatment vegetation. Line-point intercept quantified relative abundance of species aboveground in 30 x 33 m subplots. Seed bank samples were collected before and after treatments at random points along four 30-m transects per subplot. Samples were cold-moist stratified for 60 days to break dormancy. Seed pools were evaluated by direct germination in a greenhouse. Seedlings were identified, counted, and removed. After 115 days, samples were dried, mixed, rewatered and emergence was censused for an additional 21 days. Individuals that were not identified as seedlings were transplanted and grown to maturity. Community structure was assessed with non-metric multidimensional scaling (NMDS). Densities were analyzed with mixed model factorial ANOVAs. Similarities were assessed with Bray-Curtis (BC) distance including both presence and relative abundance.

Results/Conclusions

The seed bank community shifted after treatment in the imazapic and tebuthiuron treatments.  This shift was away from a set of species dominated by annual forbs and towards a more mixed set of lifeforms. Alyssum desertorum, Bromus tectorum, Ceratocephala testiculata, and Poa secunda dominated the seed bank community; the first three are exotics.  Dominant species varied in their seed density responses to treatments: A. desertorum was unaffected by any treatment, but B. tectorum, C. testiculata, and P. secunda all decreased substantially immediately following fire. B. tectorum and P. secunda recovered to levels found in the control by one year post-treatment while densities of C. testiculata increased but did not fully recover relative to the control. In addition, C. testiculata seed bank densities were decreased one year post-treatment by both herbicides. Post-treatment vegetation community was more similar to the pre-treatment vegetation than to the pre-treatment seed bank community. Results suggest seed bank community composition and seed density were temporally and spatially variable.  Furthermore, tebuthiuron and imazapic altered community composition whereas fire selectively reduced seed density.  Results also imply that the pre-existing vegetation may be a better indicator of the vegetation community following fuels/restoration treatments than is the seed bank community.