USGS - science for a changing world

Northern Prairie Wildlife Research Center

  Home About NPWRC Our Science Staff Employment Contacts Common Questions About the Site

Plant Community Patterns on Upland Prairie
in the Eastern Nebraska Sandhills

Materials and Methods

The study was conducted on the University of Nebraska's Barta Brothers Ranch, a 2,350 ha ranch in the eastern Nebraska Sandhills. The ranch is located 30 km south of Long Pine in Rock and Brown counties (42° 14'N, 99° 39'W). Prairie of this area is typical of the eastern Nebraska Sandhills in terms of topography, soils, and vegetation. Over 90% of the ranch is classified as upland prairie with the remainder being wet meadows concentrated in the southcentral part of the ranch.

Upland prairie is dominated by sands range sites. Dunes are oriented west-by-northwest to east-by-southeast and slope (5 to 15%) down to dry valley bottoms. Steep dunes (>20% slope), characterized by the choppy sands range site, are uncommon. Relief from dune top to adjacent valley floor is less than 40 m. Soils of the uplands are in the Valentine series (mixed, mesic Typic Ustipsamments). The climate is semiarid with a mean annual precipitation of 553 mm (64-year mean) at Ainsworth in northern Brown County (NOAA 1998). As much as 80% of the precipitation falls between April and September with May and June the wettest months. The 1990's were particularly wet with mean annual precipitation at 625 mm (NOAA 1998) which resulted in a relatively high water table. Vegetation of the study area is representative of the Nebraska Sandhills with warm-season tall grasses, i.e., prairie sandreed, sand bluestem, little bluestem, and switchgrass (Panicum virgatum L.), dominating the landscape. Native cool-season grasses, i.e., needle-and-thread, porcupine-grass (Stipa spartea Trin.), and Junegrass [Koeleria pyramidata (Lam.) Beauv.], and sedges also are common. A wide variety of forbs and woody plants is ubiquitous.

The major portion (1,620 ha) of the ranch has been a contiguous set of pastures that have been grazed continuously by cow-calf pairs during the growing season (May 15 to October 15) since the mid-1950's at the recommended stocking rate of 2.0 animal unit months per ha. An animal unit month is the amount of forage (304 kg) consumed by a 450-kg mature cow. Most of the pastures were 260 ha, or smaller, in size and were in good to excellent range condition in the summer of 1998 as determined by field staff of the Natural Resources Conservation Service (Mike Hanna, pers. comm.). Four upland topographic positions were recognized: interdunal valleys, south-facing slopes, dune tops, and north-facing slopes. Dune tops were relatively level areas at the apex of dune slopes and composed about 20 to 25% of the upland area. North- and south-facing slopes were the most abundant positions covering jointly about 60 to 70% of the area (Mike Hanna, pers. comm.). The interdunal valleys typically were narrow and composed only 10 to 20% of the upland.

In May 1998, 87 sampling sites were established with a spacing of about one site per 20 ha. Sites were sampled for plant species frequency over a 30-day period in June and July 1998 when the cool-season plants were mature but still intact and the warm-season plants were growing actively. Rainfall and soil moisture in spring and summer 1998 were relatively high and cool-season plants remained green throughout June. At each sampling site, a transect location was selected randomly and marked permanently on each of the four topographic positions. Transects ran parallel with the contour of the land at the top of dunes, at the midpoint between dune tops and valleys along the north- and south-facing slopes, and through the center of the narrow valleys. Transects on the slopes and dune tops were 100 m long whereas the valley transects were 50 m long. Plant species frequency was estimated along each transect by placing a 0.1-m² quadrat at 4-m intervals for a total of 25 quadrats per transect for the slopes and dune tops and 13 quadrats per transect for the valleys. At each quadrat placement, each plant species was identified and recorded. Length of transect (experimental unit) was shorter for the valleys because the valleys frequently were not extensive and species diversity in the valleys was relatively low. The 4-m interval was selected as a result of a preliminary process of developing a species-area curve relating interval (same size) to number of species (Mueller-Dombois and Ellenberg 1974). The 0.1-m² quadrat was selected because it was the largest frame size that resulted in a frequency of less than 100% for each species encountered in the sampling process (Daubenmire 1968).

Soil samples were collected in June 1999 from each of the topographic positions at two randomly-selected sampling sites. Twelve vertical, undisturbed soil cores, 1.75 cm in diameter and 30 cm in length, were taken at regular intervals along the entire length of each transect. The cores were taken with a step-down probe and divided into two depth increments: 0 to 15 cm and 15 to 45 cm. The subsample cores along each transect were composited and oven dried at 100° C until mass was constant. Samples from each composited increment were used to prepare a 1:1 soil-water mixture with 10 g of soil and 10 g of water. Soil reaction of the mixture was determined by using a pH meter. Soil organic matter content of each depth interval also was determined by using the Walkely-Black Method (Schulte 1988). Dried samples of each depth interval were sieved through a No. 10 standard sieve and texture analysis on 50 g subsamples was conducted by the hydrometer method (Day 1965).

Statistical Analysis

Transect was the experimental unit for analysis of species richness and frequency of occurrence. Species richness by topographic position was calculated as the cumulative number of plant species encountered in the quadrats of a transect. Mean frequency of occurrence of each plant species was calculated as its average frequency over all quadrats within a transect. An arcsine transformation of the frequency data was conducted to make the data normally distributed. The transformed data were analyzed by using analysis of variance (SAS 1996) and separated with Fisher's protected least significant difference at α = 0.05 (Steel and Torrie 1980). Canonical discriminant analysis (PROC CANDISC) was used to derive canonical variables. Canonical variates in our study were linear combinations of species frequency of occurrence and were used to ordinate topographic positions. Species used in the CANDISC procedure were determined by PROC STEPDISC, which selected a subset of the species to produce a discriminant model using a stepwise selection (SAS 1996).

Previous Section -- Introduction
Return to Contents
Next Section -- Results and Discussion

Accessibility FOIA Privacy Policies and Notices

Take Pride in America logo logo U.S. Department of the Interior | U.S. Geological Survey
Page Contact Information: Webmaster
Page Last Modified: Saturday, 02-Feb-2013 06:33:38 EST
Sioux Falls, SD [sdww54]