Improvement of depleted grasslands can be restricted by the lack of re
liable, cost effective and ecologically sound establishment techniques
, This paper presents results from an experimental strip seeder drill
designed to mechanically remove competing vegetation, Lucerne (Medicag
o sativa L.), birdsfoot trefoil (Lotus corniculatus L.), cocksfoot (Da
ctylis glomerata L.) and tall fescue (Festuca arundinace Schreb.) were
drilled into undeveloped natural grassland with either a triple disc
drill or a strip seeder drill in the absence or presence of herbicide.
There were small differences between the drills in initial seedling e
stablishment 7 weeks after September sowing but seedling survival over
summer varied with drill and herbicide treatment, With the strip seed
er drill, 80% and 85% of seedlings survived in the absence and presenc
e of herbicide, respectively, while with the triple disc drill 74% sur
vived in the presence of herbicide but only 40% survived in the absenc
e of herbicide, Five months after sowing, 56% and 63% of the strip see
der furrow was covered by sown species in the absence and presence of
herbicide compared with only 6% and 35% of the triple disc furrow, The
se differences in cover were also reflected in herbage dry matter of s
own species in the 3 years following drilling, The results demonstrate
the agronomic superiority of the strip seeder drill for establishment
of improved pasture species.