Weed life history: identifying vulnerabilities - Ecological Management of Agricultural Weeds - page 74

per plant, probably due to a short growth period. Moreover, when growing
with sweet corn, A. retroflexus averaged only 28 seeds per plant. Thus, cultural
practices and competition from the crop act as important regulators of weed
seed production (Zanin & Sattin,1988; Mohler & Callaway,1995; Blackshaw &
Harker, 1997).
Several models have shown that including the effects of seed production on
future crops lowers economic weed density thresholds by a factor of 3 to 8 rel-
ative to the effect of competition on the current crop alone (Cousens et al .,
1986; Doyle, Cousens & Moss, 1986; Bauer & Mortensen, 1992). Some authors
have argued that the damage inflicted on future crops by seed production is so
great that certain weeds should not be allowed to reproduce at all ( Abutilon
theophrasti - Zanin & Sattin, 1988; Echinochloa crus-galli - Norris, 1992).
Although extreme efforts to prevent spread of new, localized populations are
often justified, the economic utility of a zero tolerance policy for long-
established populations remains to be demonstrated.
In any case, measures should be taken to reduce seed production.
Depending on the phenology of the weed relative to the crop, a substantial
proportion of potential seed production can sometimes be prevented by
prompt post-harvest weed control measures (Young, 1986; Kegode, Forcella &
Durgan, 1999).This is particularly true for cereals and early season vegetables
where harvest of the crop releases the weeds from competition at a time in the
season when temperature and day length allow rapid growth and maturation
of previously suppressed weeds.For example,Webster,Cardina & Loux (1998)
found that killing weeds in July or August following wheat harvest controlled
70% to 95% of various weed species in maize the following spring relative to
control plots in which weeds were allowed to mature.
In some grain crops, a large portion of the weed seed produced passes
through the combine. For example, Ballaré et al . (1987 a ) found that

2% of
Datura ferox in soybean were shed prior to harvest, and that all three of the
combines tested took up nearly all capsules. In such cases, if equipment were
added to the combine to capture or destroy weed seeds rather than dispersing
them with the chaff, substantial reductions in the annual addition of viable
seeds to the seed bank could be achieved. Slagell Gossen et al . (1998) proposed
attaching hammer mills or roller mills to grain combines to destroy weed
seeds before they were returned to the field.They found that both types of mill
killed a high percentage of Bromus secalinus seeds.In many crop-weed systems,
however, the benefit of capturing or killing weed seeds in the combine would
be small because most of the seeds disperse prior to harvest (Moss, 1983).
Although seed collection and post-harvest weed control usually will not
provide effective control by themselves, they can contribute substantially to