USGS Home
Contact USGS
Search USGS

The study was conducted at the University of Nebraska, Agricultural Research and Development Center, near Mead, Nebraska (41 10°N, 96 25°W; 344 m above sea level). The study area at Mead was an 8-ha field planted in 1971 to big bluestem (Andropogon gerardii Vitman), little bluestem (Schizachyrium scoparium [Michx.] Nash), sideoats grama (Bouteloua curtipendula [Michx.] Torr.), Indiangrass (Sorghastrum nutans [L.] Nash), and switchgrass (Panicum virgatum L.). In 1981, the study site was divided into forty-two 30 x 30-m study areas as part of a long-term investigation of seasonal fire effects (T. Bragg, professor of biology, University of Nebraska, Omaha, pers. com.). Two areas (#13 and #30) were not used in the long-term study and, except for occasional mechanical removal of woody species, were neither burned nor mowed for the 7 years prior to this study. Smooth brome invaded these two areas and became a codominant with big bluestem. During the 3-year study period, precipitation measured at the Mead Agronomy Laboratory 6 km from the study site, varied considerably. Late winter and spring precipitation was near normal in 1990, 47% below normal in 1989, and 35% above normal in 1991 (National Oceanic and Atmospheric Administration 1989, 1990, 1991).

Sampling Methods

In October 1988, study area #30 was divided into sixteen 6 x 6-m plots, each surrounded by a 1.25-m mowed lane. Fall density and biomass of smooth brome tillers were determined for each plot by counting and hand-clipping at the soil surface all current-year tillers within ten 0.1- m² randomly placed quadrats. Biomass samples were dried at 55^oC until they reached a constant weight. In spring 1989, one of four treatments was applied to each plot in a completely randomized design with four replications per treatment. Treatments included burning plots at smooth brome tiller emergence, at elongation, and at heading, and no burning. The three growth stages were essentially the same as the VI (first leaf collared), EI (first node palpable/visible), and RI (inflorescence emergence/first spikelet visible) growth stages of perennial forage grasses described and quantified by Moore et al. (1991). In fall 1989, smooth brome density and biomass were determined by randomly resampling the plots using ten 0.1-m² quadrats.

Burn dates were timed according to the morphological stage of growth of smooth brome and did not vary by more than 10 days over the 3-year period of study (Table 1). After 50% of randomly sampled tillers within the plots had reached a particular morphological stage, burns were conducted as soon as conditions met the following prescription criteria: temperature between 10^oC and 27^oC, wind from any direction between 16 and 32 km ^h-1, and relative humidity between 30% and 70%. Generally, there was only a 2- to 3-day delay between detection of a morphological stage and application of the burn treatment. A back fire was used to burn each plot.

Table 1. Smooth brome morphological growth stages and burn dates for 1989-91 at Mead, Nebraska.

In fall 1989, study area #13 was divided into sixteen 6 x 6-m plots, and density and biomass of smooth brome tillers were determined as in 1988. The experiment was repeated in spring 1990 with the addition of a fourth burn treatment at smooth brome flowering (the R4 [anther emergence/anthesis] growth stage [Moore et al. 1991]). Treatments were applied to each 6 x 6-m plot in a completely randomized design with three replications per treatment. Also, in spring 1990, each of the 16 plots in study area #30 was divided in half along a north/south axis. Either the west or the east half randomly received the same treatment that had been applied in 1989, and the other half received no burn treatment. This provided data for repeated burn treatments (burn/burn) as well as recovery from a single burn (burn/unburned). In spring 1991, the 16 plots in study area #13 were divided in half, as had been those plots in study area #30 in 1990. Half of each plot was then reburned, while the other half was not burned. This provided a second year of data for plots burned once in 2 years and twice in 2 years.

Data Analysis

An analysis of covariance with a means comparison test (F-protected LSD) was used to test for differences in smooth brome tiller density and biomass among treatments within years (α = 0.05) (Montgomery 1984). Fall 1988 and 1989 smooth brome tiller density and biomass were the covariates. An analysis of variance with an F-protected LSD test was used to determine smooth brome tiller density and biomass differences by treatment among years (α = 0.05). Analyses were performed using SOLO statistical software (BMDP Statistical Software 1991).