Exotic Brome-Grasses in Arid and Semiarid Ecosystems of the Western US

(ff) #1

352


cover and annual biomass production data for B. tectorum. We also limited our
inquiry to the following four B. tectorum treatment methods: burning (i.e., wildfi re
and/or prescribed fi re), defoliation (i.e., grazing and/or mowing), herbicide applica-
tion, and revegetation (i.e., seeding).
For the 45 articles that met our criteria, we extracted data from tables or esti-
mated values from digital scans of fi gures using DataThief III Version 1.5 com-
puter software (Myers and Harms 2009 ). In most cases, individual articles offered
numerous control-treatment contrasts when they included more than one treatment
and if experiments were conducted at different, independent study locations. When
data were collected at the same location over time, we acquired both the fi rst (1 year
posttreatment application) and fi nal census (>2 years posttreatment) to determine
potential differences due to timeframe after treatment.


12.4.2 Analysis of B. tectorum Treatment Outcomes

To assess the effect of various treatment methods on B. tectorum , we used effect size
estimates, an approach for synthesizing responses measured across a set of contrast-
ing studies (Osenberg et al. 1999 ). Effect sizes were calculated with the following


equation: lnRR = ln( X (^) E / X (^) C ), where lnRR is the natural log response ratio, X (^) E = treat-
ment mean, and X (^) C = control mean (Goldberg et al. 1999 ). We report mean lnRR ± 95 %
confi dence intervals without considering within-study variance estimates (Hedges
et al. 1999 ). Mean lnRR’s were considered signifi cant if 95 % confi dence intervals
did not overlap zero. While we recognize that the impact of Bromus control treat-
ments are site specifi c and outcomes vary across precipitation and temperature
regimes (Chambers et al. 2014a ), our goal was to assess general, treatment out-
comes, identify research gaps, and inform adaptive management planning
(Fig. 12.1 ).
Table 12.1 (continued)
Study
Control
method a
Response
variables b Timeframe c
Location
(USA state)
Turner ( 1971 ) D C L CO
West and Yorks ( 2002 ) B, D C L UT
Wicks et al. ( 1965 ) H B E, L NE
Young et al. ( 1972a ) B B, C E CA
a B, burn; D, defoliate; H, herbicide; R, revegetation
b C, B. tectorum cover; B, B. tectorum shoot biomass production
c Timeframe posttreatment application; E, early (≤1 year); L, late (≥2 year)
T.A. Monaco et al.

Free download pdf