Location of Sampling Stations: Plant composition is determined on upland topographic locations (Florence soils), lowland topographic locations (Tully soils), and slope locations. Maps of sampling locations are available as following PVC Map.

Sampling History: Collection of plant composition data began in 1983 in watersheds 1c, 1d, 4b, 20b, N1b, N4d, and N20b. Since then, numerous other watersheds have been added and an annual plant census is currently conducted in:1d, 2c, 2d, 4a, 4b, 20b, N1a, N1b, N2a, N2b, N4a, N4d, N20a, N20b, Fa, Fb, Wa, Wb, Spa, Spb, Sua, Sub, R1a, R1b, R20a, R20b. Sampling in the eight seasonally burned watersheds began in 1994. Four reversal watersheds were established in 2001 in which 1a became R20a, 20a became R1a, 1c became R20b, and 20d became R20b. Only Tully soils (t) are sampled in R1a, R20a, N2a, and N2B. Sampling on slope sites (s) in 1d, 1c, 4a, 4b, 20b, 20d, N4a, N4d, N20a, and N20b were conducted from 1991-2001, but are currently sampled only in the bison-grazed watersheds. A map of sampling locations can be found here [26].

Frequency of Sampling: Initially, plots were surveyed 3 times each year. Beginning in 1991, however, sampling frequency was changed to twice a year (late-May to Mid-June for the spring census, and August to early-September for the summer census). Generally, cool-season species are sampled in the spring census whereas warm-season species are sampled in the summer census when they are more developed.

Variable Measured: Canopy covers of all vascular plant species in each plot are estimated.

Methods: In each watershed, four 50-m long transects (A, B, C, and D) were established on each topographic position. In each transect, five evenly spaced, permanently marked plots are located (n = 20 plots for each topographic position in a watershed). A surveyor’s pin with a 1.78 m long chain is placed in the conduit marking each plot. Canopy cover of all vascular plant species in a 10-m2 circular area within each conduit are estimated using a modified Daubenmire cover scale (Bailey and Poulton, 1968. Ecology 49:1-13). Seven cover classes were used to estimate species canopy coverage. 1 - 0-1% cover; 2 - 2-5% cover; 3 - 5-25% cover; 4 - 25-50% cover; 5 - 50-75%; 6 - 75-95% cover; 7 - 95-100% cover. Codes used in the data set: t = tully (lowland) soil; f = florence (upland) soil; s = slopes.

Form of Data Output: Raw data contains the cover class value for each species detected in the plot. For species that are sampled on both census dates, the highest cover class of each plot is used for analyses. Percentage cover for each species is computed by averaging the mid-points of the cover classes for the 20 plots.The presence or absence of a species in the 20 plots can be used to estimate frequency of occurrence.

Summary of All Changes: Transect E only occurred on watershed N20B florence from 1983 to 1987. This transect was established before bison reintroduction and was abandoned after fence construction left it outside of the treatment area. The current N20b florence transect D was established in 1986 to replaced the old transect E. Many watersheds have experienced name changed over the duration of data collection, so one physical sampling location may have difference watershed designations in the dataset. Use the following file to reconcile watershed names: Watershed Name Matrix [27].

Notes: For watershed r20b, no data were collected in the transect A and B in the summer of 2011 due to wildfire. Summer sampling for watersheds R1a and R20a did not occur in 2014 due to personel changes. Plots in the bison area were established prior to bison reintroduction. Grazing began for the 1988 growing season for Phase I watersheds (N1a, N2a, N4a, N4c, N20a) and 1992 for Phase II watersheds (N1b, N2b, N4b, N4d, N20b).