Disturbances such as fire, land clearing, and road building remove vegetation and can have major influences on public health through effects on air quality, aesthetics, recreational opportunities, natural resource availability, and economics. Plant recovery and succession following disturbance are poorly understood in arid lands relative to more temperate regions. This study quantitatively reviewed vegetation reestablishment following a variety of disturbances in the Mojave and Sonoran Deserts of southwestern North America. A total of 47 studies met inclusion criteria for the review. The time estimated by 29 individual studies for full reestablishment of total perennial plant cover was 76 years. Although long, this time was shorter than an estimated 215 years (among 31 individual studies) required for the recovery of species composition typical of undisturbed areas, assuming that recovery remains linear following the longest time since disturbance measurement made by the studies.

Restoring biological soil crusts (biocrusts) in degraded drylands can contribute to recovery of ecosystem functions that have global implications, including erosion resistance and nutrient cycling. To examine techniques for restoring biocrusts, we conducted a replicated, factorial experiment on recently abandoned road surfaces by applying biocrust inoculation (salvaged and stored dry for two years), salvaged topsoil, an abiotic soil amendment (wood shavings), and planting of a dominant perennial shrub (Ambrosia dumosa). Eighteen months after treatments, we measured biocrust abundance and species composition, soil chlorophyll a content and fertility, and soil resistance to erosion. Biocrust addition significantly accelerated biocrust recovery on disturbed soils, including increasing lichen and moss cover and cyanobacteria colonization. Compared to undisturbed controls, inoculated plots had similar lichen and moss composition, recovered 43% of total cyanobacteria density, had similar soil chlorophyll content, and exhibited recovery of soil fertility and soil stability. Inoculation was the only treatment that generated lichen and moss cover. Topsoil application resulted in partial recovery of the cyanobacteria community and soil properties. Compared to untreated disturbed plots, topsoil application without inoculum increased cyanobacteria density by 186% and moderately improved soil chlorophyll and ammonium content and soil stability. Topsoil application produced 22% and 51% of the cyanobacteria density g⁻¹ soil compared to undisturbed and inoculated plots, respectively. Plots not treated with either topsoil or inoculum had significantly lower cyanobacteria density, soil chlorophyll and ammonium concentrations, and significantly higher soil nitrate concentration. Wood shavings and Ambrosia had no influence on biocrust lichen and moss species recovery but did affect cyanobacteria composition and soil fertility. Inoculation of severely disturbed soil with native biocrusts rapidly restored biocrust communities and soil stability such that restored areas were similar to undisturbed desert within three years. Using salvaged biocrust as inoculum can be an effective tool in ecological restoration because of its efficacy and simple implementation. Although salvaging biocrust material can be technically difficult and potentially costly, utilizing opportunities to salvage material in planned future disturbance can provide additional land management tools.

The reference conditions of historical tree density and pattern underpin ecological restoration and management of Pinus ponderosa Douglas ex Lawson & C.Lawson forests in western North America, yet the potential spatial variation in these variables across the landscape remains unclear. We reconstructed historical (1880) tree density and spatial pattern on 1 ha plots at 53 sites within a 110 000 ha P. ponderosa landscape in northern Arizona, compared these variables among US Forest Service ecosystem classification units, and modeled spatial variation with environmental variables. Mean tree density differed 19-fold among nine ecosystem types, and regression trees using four soil or climatic variables explained 62%–74% of the variation in density. Although density was more sensitive to environmental variation than was pattern, we did not find the clumped pattern widely described for P. ponderosa forests to be universal across ecosystems. Results suggest that (i) multivariate combinations of soil and climatic properties influenced historical forest structure, (ii) as much variation exists in reference conditions within the study landscape as between P. ponderosa regions, (iii) ecosystem classification is a useful framework for quantifying spatial variation in reference conditions, and (iv) determining spatial variation in reference conditions can assist resource managers in prioritizing areas for management and in developing ecosystem-specific management strategies within landscapes.

a b s t r a c tIncreasing wildfire activity is one of the most pressing management concerns in arid lands of the American West. To examine post-fire recovery of perennial vegetation in the Mojave and Sonoran Deserts, I analyzed data systematically synthesized from the literature. Post-fire sprouting by desert perennials is generally limited but varies among species. For example, only 3-37% of Larrea tridentata sprouted compared to 64-86% of Yucca schidigera. Four of five studies measuring recovery of perennial cover reported close relationships (r 2 ¼ 0.67-0.99) between time since fire (TSF) and cover. In fact, three studies measuring the longest TSF (!37 years) found that cover had returned to within 10% cover of unburned areas within approximately 40 years. Conversely, post-fire species composition exhibited little convergence with unburned composition in five of six studies even 47 years after fire. Sphaeralcea ambigua, Gutierrezia spp., Achnatherum speciosum, Encelia spp., Hymenoclea salsola, and Baileya multiradiata had the highest burned:unburned abundance ratios, although overall post-fire community composition differed between the Mojave and Sonoran Deserts. Analyzing the literature as a whole suggested some generalities (e.g., that perennial cover reestablishes faster than composition), but more work is required for improving specific knowledge about plant recovery among fires, sites, species, and climates.