In semiarid regions such as the Great Basin, riparian areas function as oases of cooler and more stable microclimates, greater relative humidity, greater structural complexity, and a steady flow of water and nutrients relative to upland areas. These qualities make riparian areaʼs attractive not only to resident and migratory wildlife, but also to visitors in recreation areas such as Great Basin National Park in the Snake Range, east-central Nevada. To expand upon the system of ten permanent plots sampled in 1992 (Smith et al. 1994) and 2001 (Beever et al. in press), we established a collection of 31 cross-sectional transects of 50-m width across the mainstems of Strawberry, Lehman, Baker, and Snake creeks. Our aims in this research were threefold: a) map riparian vegetative communities in greater detail than had been done by past efforts; b) provide a monitoring baseline of hydrogeomorphology; structure, composition, and function of upland- and riparianassociated vegetation; and edaphic properties potentially sensitive to management; and c) test whether instream conditions or physiographic variables predicted vegetation patterns across the four target streams.

In each of the four watersheds, we performed walking transects from the lower-elevation boundary of the park along creek mainstems to a point well above the point at which vehicle access stopped. In these transects, we ranked, by cover, the riparian and upland woody species on each side of the creek, in 0.32-km segments. These walking transects also facilitated selection of a suite of cross-sectional transects that might serve as an early-warning signal of change for natural (e.g., aggradative) and anthropogenic changes (e.g., due to visitor impacts or climate change). At each cross-sectional transect, we used several methods: a) measurement of the number, approximate volume, and total length of instream logs greater than 10 cm in diameter that were within 5 m up- or downstream of the transect; b) counts of pebbles by size class, following Wolman (1954); c) line-point intercepts, which provided various measures of percent cover; d) gap-intercept transects, following Herrick et al. (in press), to measure susceptibility of uplands to erosion by wind or water; e) 1-m^₂ quadrats, to obtain frequency of woody species; f) nested-frequency plots, to measure frequency of all plant species in quadrats of varying size; g) a field-based soil aggregate stability test following Herrick et al. (2001); and h) an impact penetrometer, to measure penetration resistance of soil horizons.

We used species-accumulation curves to assess the ability of our methods to detect the majority of plant species at sites, using the most species-rich and species-poor sites as illustrations. We compared characteristics of hydrogeomorphic valley types (designated by Frissell and Liss 1993), vegetation types, and creeks individually and, using multivariate analyses for the first two ʻtypes,ʼ simultaneously. For the latter, using both the nested-frequency and 1-m² frequency data, we first used nonmetric multidimensional scaling (NMS) to assess relationships of plant communities among sites. Secondly, we used multi-response permutation procedures (MRPP) to test whether plant-community differences existed among either hydrogeomophic valley types or vegetation types. To increase the value of these comparisons for management, we used indicator species analyses to quantify the indicator value of each individual plant species for separating groups.

In contrast to the more incised riparian channels of central Nevada, we observed knickzones, downcutting, and incision only rarely and usually with limited extent in the walking surveys. Downcutting occurred most frequently and extensively in Strawberry and Snake creeks, due in part to their more erodible soils. According to a hydrogeomorphologist with extensive experience in Great Basin riparian systems, the sediment-delivery and hydrologic systems appeared relatively undisturbed in most reaches, with respect to grazing animals and other types of anthropogenic alteration. Site elevation of the 31 transects ranged from 1,950-2,987 m, and stream slope (i.e., gradient) was relatively steep (mean = 9.3%, range 3-16%). Strawberry Creek averaged the lowest maximum water depth, and correspondingly had greatest width/depth ratios. Baker Creek sites averaged the smallest amount of tree-canopy gaps, whereas Snake Creek sites on average had the largest proportion of gaps in understory vegetation. Sites in terrace-bound valley types averaged the lowest slope in the channel as well as the least cover of trees, litter, and vegetation overall, whereas alluviated, boulder-bed canyon sites averaged the greatest widths of the active channel. Sites in Lehman Creek averaged nearly twice as much coarse woody debris as sites from any other creek, whereas Baker Creek sites averaged greatest tree cover (mean = 67%, range 40 – 96%) and species richness (mean = 17.3 species). Multivariate ordinations suggested that sites in leveed outwash valleys and alluvial-fan-influenced valleys had the greatest inter-site heterogeneity in plant composition, whereas sites in incised moraine-filled valleys appeared most homogeneous. Differences among homogeneity of sites within vegetation types were less pronounced, but sites dominated by either aspen and Woodsʼ rose or narrow-leaved cottonwood had the most similar plant communities among sites of the same vegetation type. A number of species were faithful indicators of various valley and vegetation types, using either set of plant-frequency data. We estimate that all 31 sites could be subsequently re-sampled in 14-18 field days by individuals possessing familiarity of the riparian flora of the southern Snake Range. As with any research, monitoring-focused investigations must balance the concerns for number of ecosystem attributes measured, extensiveness in time and space of sampling periods and locations, and the time and cost of sampling.