Weed evolution and community structure - Ecological Management of Agricultural Weeds - page 458

and Vismia guianensis have become pasture weeds as permanent pasture has
replaced forest in Amazonia (Dias Filho, 1994, pp.7-8).
Movement of species into agriculture is also facilitated by introduction of a
new cropping system into a region. For example, the advent of the native
Californian marsh grass Echinochloa microstachya as a weed presumably post-
dates the initiation of rice cultivation in that state during the period 1912-15
(Barrett & Seaman, 1980; Barrett, 1988).This species has subsequently spread
to rice-growing areas in Australia (Barrett, 1988). Similarly, the native wild
rice Oryza punctata was first noted in rice fields in Swaziland in the mid 1950s.
By the 1970s the species had forced abandonment of mechanized rice cultiva-
tion over large areas (Barrett, 1983). Throughout the eastern USA, the advent
of no-till planting has resulted in colonization of fields by robust native per-
ennial species like Solidago altissima and Rhus typhina that were previously not
associated with row crop agriculture.
Evolution of weed races from wild populations
Although the initial movement of early successional species from
natural habitats into human disturbances involves simply the exploitation of
somewhat similar habitats by preadapted genotypes,subsequent selection can
be expected to modify weedy populations substantially. Tilled fields differ
from the ancestral habitats of weeds in several important respects. First, the
temporal pattern of soil disturbance by tillage is typically much more predict-
able than natural disturbance by animals, shifting dunes, or erosional/deposi-
tional events along streams and cliffs. Second, the increase in competition as
the crop matures is also more predictable than in a natural succession. Third,
the selective pressures from weed pulling,cultivation,and,more recently,her-
bicides differ from those exerted by natural mortality factors in wild habitats.
Finally, with the exception of some animal-caused disturbances, the nitrogen
fertility of most naturally open habitats is low; cliffs, dunes, beaches, and
gravel bars lack the organic matter necessary for generation of a pulse of
nitrate following soil disturbance. Thus, human disturbance selects from the
wild not only preadapted species but particular preadapted genotypes. These
then become the ancestors of weedy races, and ultimately, species.
Because only certain genotypes of a potential agrestal weed species may be
capable of founding a weed race, the invasion of cultivated fields by wild
species is probably an ongoing process partially determined by chance events.
Due to the diversity of relatively innocuous weeds that are often present at low
abundance in a given field, the new arrival of a weedy strain of a wild species
usually goes unnoticed until it spreads. By then it is usually too late to docu-
ment the time and place of the weed's origin. Consequently, few studies