Livestock grazing for weed management - Ecological Management of Agricultural Weeds - page 436

and yield animal products profitably.Essential to these goals are grazing man-
agers who regularly monitor the state of their animals, vegetation, and
natural resources (Chapter 3) (Bingham & Savory, 1990, pp.87-120; Watters,
1990; Sheath & Clark, 1996; Stafford Smith, 1996).
Apractical challenge for the application of improved grazing management
for weed control at the landscape scale is the development of simple sampling
methods to categorize the state of a given paddock or range and to detect
incipient transitions. A comparison of four sampling methods by Stohlgren,
Bull & Otsuki (1998) illustrated difficulties associated with the evaluation of
range vegetation.Whereas common transect methods satisfactorily estimated
forage availability and soil cover of the major plant species, rare species were
generally underestimated. They proposed multiscale sampling instead of
transect methods. Small quantitative sample plots serve to monitor the abun-
dance of common species,while a search of a larger surrounding area for addi-
tional species would detect rare plant species or recent invaders. The difficult
task of early detection of noxious species is key to their successful manage-
ment (Zamora & Thill, 1999). Temperate pasture sampling has focused on
biomass or height to estimate available forage (Sheath & Clark, 1996) with
little attention to weed presence.Multiscale sampling in these pastures would
permit accurate estimation of available forage, while monitoring the status of
less-frequent weed invaders.
The task of detecting transitions from sustainable vegetation states to dete-
rioration (Figure 9.3) is especially challenging. Stockwell et al . (1994) found
that farmers readily verified the states proposed by scientists for Chrysopogon
fallax grasslands. However, identification of the factors related to transitions
was more difficult, attributed by the authors to the lack of clear documenta-
tion of the nature of transitions,highly variable experiences among observers,
and the circumstantial, non-systematic nature of the information. Detection
or prediction of transitions requires not only the sampling of species and
biomass, but also the recording of information that aids in understanding
why and how the vegetation may be changing. Such information includes
rainfall events, soil moisture status, grazing levels, fire, and seed production.
For temperate humid grass farms, Watters (1990) and Webby & Sheath (1991)
suggested daily monitoring of rainfall and soil temperature and bi-weekly
estimates of pasture cover, growth rates, and grazing levels.Animal body con-
dition should be assessed at least monthly.In their proposal,other animal per-
formance data and soil fertility levels were monitored less frequently. For the
monitoring to be used widely by scientists, technicians, and farmers, Watters
(1990) recommended that pasture sampling methods be standardized and a
practical notebook developed. Although not mentioned by Watters (1990),
standardized sampling methods for weeds are also needed.