In an experimental area of a dense Phalaris aquatica L. sward, seeded at th
e rate of 8,5 Kg/ha in drill at 30cm apart, two subareas were thinned out t
o every other row, one removing the plants mechanically and another applyin
g glyphosate. The experimetal units were distributed in an experimental des
ign of complete randomized plots, with three replicates. The seed was harve
sted in three consecutive years. Seed production/ha was not affected (P con
gruent to 0,05) either by the thinning out or the way it was done. The resu
lts are discussed in relation to the yield components: density and length o
f panicles.