Tephra deposits from volcanic eruptions may constitute a major disturbance to vegetation, with potentially persistent and complex effects. However, tephra disturbance has received limited study. The 1980 eruption of Mount St. Helens allowed us to address this topic. Here we ask how vegetation change has progressed in forest understory beneath an intact tree canopy over the subsequent 30 years and how vegetation in 2010 compares to that before disturbance. In summer 1980, soon after tephra deposition, we established permanent plots (1 m2) in four old-growth conifer forests differing in tephra depth (4.5 or >12 cm), snow cover at the time of the disturbance, and species composition (herb-rich or herb-poor). At each site, we established 100 natural plots and 50 plots from which we removed tephra; these cleared plots both simulated rapid erosion and indicated the pre-disturbance understory. At the deep-tephra, herb-rich site (DR) we removed tephra from 50 additional plots in 1982 to simulate delayed erosion. We recorded plant species cover and density periodically, 8-10 times from 1980 to 2010.
Results/Conclusions
The tephra deposit almost eliminated bryophytes at all sites. Deep tephra greatly reduced herb cover, whereas shrubs and small trees were substantially reduced wherever tephra fell on snow. After 30 years, some growth forms and species reached pre-disturbance abundance, a few exceeding their pre-eruption importance, while others were still much reduced in abundance. Bryophyte cover remained low in 2010 (<half of pre-disturbance) and was increasing only slowly at all sites. Total herb cover was still reduced at deep-tephra sites, but species differed greatly in response. At DR, a previously dominant herb, Erythronium montanum, still had <1% of its former cover, whereas other species (also strongly reduced initially) exceeded pre-disturbance cover. Delayed clearing of tephra facilitated recovery of some herbs, confirming the importance of post-disturbance events. The shrub layer had reached pre-eruption abundance at all sites except DR. Tree cover had attained or exceeded pre-eruption values; at DR establishment of conifers on the bare tephra produced a tree seedling layer with 44% cover in 2010, compared to the 12% pre-disturbance value. Although many attributes recovered to roughly pre-disturbance abundance, convergence is not apparent in others, indicating that disturbance can deflect vegetation change in new directions.