Ac. Grundy et al., Modelling the effect of cultivation on seed movement with application to the prediction of weed seedling emergence, J APPL ECOL, 36(5), 1999, pp. 663-678
1. Effective weed control is essential in field vegetables. However, the ra
nge of available herbicides has been continually reduced for commercial and
toxicological reasons over the last decade. In order to predict the optimu
m weeding period and to apply alternative control strategies successfully,
a realistic estimate is needed of the size, timing and duration of a flush
of weed emergence in a crop. The soil weed seed bank is the primary source
of future weed populations, and therefore provides a unique resource for pr
edictive management purposes.
2. Models have previously been developed to predict the emergence of weed s
pecies from different burial depths and to simulate the vertical movement o
f weed seeds following seed bed preparation.
3. In this investigation a vertical movement model was extended to include
the effects of four cultivation implements on the horizontal displacement o
f weed seeds. These implements were a rotavator, a spring tine, a spader an
d a power harrow.
4. The rotavator caused a backward movement of seeds; neither the spring ti
ne nor spader had a significant effect on the horizontal displacement of se
eds; whilst the power harrow had the greatest capacity to move seeds forwar
d > 0.5 m in the soil.
5. This investigation combined depth of burial and vertical movement models
to simulate the likely outcome of different sequences of spring tine, spad
er, rotavator and power hat-row on subsequent weed seedling emergence. For
example, sequences including multiple passes of a spader increased the numb
ers of emerged seedlings, whilst for those where the rotavator dominated th
e sequence, a marked reduction in seedling numbers was predicted. The findi
ngs of a series of simulations are viewed in the light of existing methods
of weed control based on soil cultivation, for example the stale seed bed t
echnique.
6. The combined model provides the basis for a decision support system to a
id the control of weeds. Additionally, it provides a research tool to impro
ve understanding of the dynamics of the weed seed bank and the implications
of seed bed preparations for future populations. The combined model has he
lped to identify areas of weed seed ecology requiring further study, essent
ial for the development of true dynamic models.