from Overstory-understory relationships for broom snakeweed-blue grama grasslands
Kirk C. McDaniel, L. Allen Torell, and John W. Bain
Abstract
Data collected over a 11-year period at 2 study areas near Vaugn and Roswell, N.M. were used to define equations that relate grass biomass to the amount of broom snakeweed (Gutierrezia sarothrae [Pursh] Britt. & Rusby) occupying blue grama (Bouteloua gracilis [H.B.K. Lag]) rangeland over time. A 5 parameter sigmoidal growth equation and a negative exponential equation best expressed the relationship between understory grass biomass and overstory broom snakeweed biomass. Explanatory variables included realized precipitation during the second (April to June) and third (July to September) quarters, which coincides primarily with warm-season grass growth. Minimum suppression of grass biomass occurred with complete elimination of broom snakeweed, suggesting control strategies with high overstory mortality will likely be most beneficial to understory production.
Key Words: herbage production, Gutierrezia sarothrae, Bouteloua gracilis weed control, range improvement
Overstory-understory relationships have been defined for many woody and herbaceous plants common on western rangelands (Bartlett and Betters 1983, Ffolliot and Clary 1972). Published overstory-understory equations include linear, logarithmic, quadratic, cubic, and various nonlinear, exponential functional forms (Ffolliot 1983, McPherson 1992, Scanlan 1992). In general, the relationship between herbaceous production and woody cover has been found to be a downward-sloping curve that is either convex to the origin or S-shaped over the relevant range (Jameson 1967). The convex shape, reported to be appropriate for numerous overstory species (Jameson 1967, Hull and Klmop 1974, Ffolliot 1983, Pieper 1990) suggests marginal suppression of herbaceous biomass declines as overstory cover increases, and implies that the first woody plants to invade an area suppress herbaceous production the most. Similarly, a sigmoid shaped curve implies a light or scattered stand of overstory species results in little if any suppression of understory production until a threshold level is eventually reached, and grass yield then rapidly declines (Scifres et al. 1982, Jameson 1967).
Several studies have investigated competitive relationshiops between broom snakeweed (Gutierrezia sarothrae [Pursh] Britt. & Rusby), an undesirable woody weed, and associated grasses growing on rangeland (Campbell and Bomberger 1934, Gardner 1951, McDaniel and Sosebee 1988). Ueckert (1979) and McDaniel et al. (1982) found that thinning a heavy stand of broom snakeweed by 25%, 50%, or 75% did not greatly increase grass yield, whereas complete removal of the weed resulted in a grass production increase from 200% to 400%, suggesting some type of nonlinear relationshiop between overstory broom snakeweed cover and understory grass production.
To define overstory-understory relationships for an economic analysis of broom snakeweed control, Carpenter et al. (1991) used a linear equation to relate expected grass production to broom snakeweed canopy cover. Variables reported to influence grass production included snakeweed cover, summer rainfall, and soil type (site shifters). For a similar economic analysis, Torell et al. (1990) used a logarithmic model to define overstory-understory relationships with explanatory variables defined to be broom snakeweed yield, climatic conditions, and site characteristics. A shortcoming noted by Torell et al. was that the logarithmic model overestimated grass production when broom snakeweed biomass was near zero. A problem with the linear estimation is other research (McDaniel et al. 1982 and Ueckert 1979) suggests the correct broom snakeweed-understory relationship is curvilinear.
The purpose of this study was to define equations expressing the overstory-understory relationshiop for broom snakeweed growing on blue grama (Bouteloua gracilis [H.B.K. Lag])-dominated grasslands in central New Mexico. Model estimation is provided from broom snakeweed and grass biomass data collected over an 11-year period at 2 permanent study sites near Vaughn and Roswell, N.M.
Table 2. Estimated equations defining overstory-understory relationships for the Vaughn and Roswell study sites.
===================================================================================
Equation n R^2 RMSE Estimated Equation
-----------------------------------------------------------------------------------
Sigmoid
Vaughn 104 .68 151.2 Y=278-569(1-e^-.0015X)^.601+17.8R2+16.0R3
Roswell 100 .59 231.9 Y=367-698(1-e^-.0023X)^1.027+22.7R2+10.9R3
Combined 204 .61 195.0 Y=258-629(1-e^-.0020X)^.881+20.5R2+16.6R3
Exponential
Vaughn 104 .68 151.3 Y=242-503e^-.0031X+17.9R2+16.2R3
Roswell 100 .59 230.7 Y=333-701e^-.0022X+22.8R2+11.0R3
Combined 204 .61 194.5 Y=363-615e^-.0022X+20.6R2+16.6R3
-----------------------------------------------------------------------------------
where
Y=grass biomass (kg/ha)
X=broom snakeweed biomass (kg/ha)
R2=2nd quarter (April, May, June) precipitation (cm)
R3=3rd quarter (July, August, September) precipitation (cm)
Using an allowance of 780 pounds of forage per Animal Unit Month (9360 pounds per AUY), and using the combined sigmoid equation, we get
forage=(258-629*(1-e^-.0020X)^.881+20.5R2+16.6R3)/1.1227 (lb/acre)
and
capacity=forage*640/9360 (AUY/section)