Understory plant community responses to hazardous fuels reduction treatments in pinyon-juniper woodlands of Arizona, USA.
Although hazardous fuels reduction projects are being implemented widely in dry forests of the western United States, information concerning ecological responses of pinyon-juniper woodlands to fuels treatments is minimal. In this study we used a randomized block experimental design to investigate effects of fuels treatments on understory responses at a pinyon-juniper woodland site in northern Arizona. Treatment alternatives replicated (n=12) on 1-ha units across the 760-ha site were thinning from below (Thin), prescribed burning, (Burn), thinning followed by prescribed burning (Thin+Burn), and untreated controls (Control). We measured cover of vascular plant species, biological soil crust, and forest floor substrate prior to treatment and one year, 2 years, and five years post-treatment. After five years, Thin and Thin+Burn plots showed statistically similar total plant cover but both showed significantly greater cover than Burn and Control plots. Cover increases on Thin+Burn plots were driven mainly by annual forb and nonnative species groups. Graminoid species with the C4 photosynthetic pathway showed significantly greater cover in all treatment groups compared with Control plots. Burning alone tended to reduce shrub cover whereas thinning alone tended to increase shrub cover. We found no significant effects of treatment on cover of biological soil crust or bare soil. Regression analysis showed significant negative relationships between total understory plant cover and overstory canopy cover on burned and unburned plots. We found no significant effect of treatment on species richness over the measurement period. We observed 124 taxa in total, and mean richness ranged from 31.8 to 36.8 species per 500 m2 in 2011. Similar to results for plant cover, we found significantly greater numbers of annual forb species in Thin+Burn compared with Control plots and significantly greater numbers of nonnative species in Thin+Burn compared with all other treatment groups. Permutational multivariate analysis of variance showed no significant differences in overall plant community composition among treatments. These results indicate that fuel reduction treatments on woodland sites may not dramatically affect understory abundance or composition within five years. Although treatments did not significantly differ in cover of biological soil crust or bare soil - indicators of ecosystem function - our results showed that thinning treatments followed by prescribed fire led to increases in cover of nonnative plant species.