Summary: | <p>Impact of weed control methods on the persistence of <em>Parthenium hysterophorus L.</em>, and tomato (<em>Solanum lycopersicum L.</em>) yield was studied in two farmer fields i.e. a site heavily infested with<em> P. hysterophorus</em> (Parthenium-site) and Parthenium free site, in the Jaffna peninsula of the Northern Province of Sri Lanka. The split-plot design consisted of two main plots i.e. manual weeding and ploughing (both followed by harrowing), and three sub-plots i.e. use of a pre-emergent herbicide (Oxyfluorfen; 240 g/L EC at 2 days before planting), mulching [<em>Gliricidia sepium (Jacq.) </em>Kunth ex Walp. leaves at 12 kg per plot on fresh weight basis], and un-weeded plot after crop establishment (control), in three replicates. The soil seed bank of the Parthenium-site was dominated by <em>P. hysterophorus</em> as expected, and the Parthenium-free site was dominated by the perennial sedge <em>Cyperus rotundus L.</em> At tomato harvest, the un-weeded sub-plots recorded the highest weed densities (p<0.05) when compared to the rest, The Parthenium-site showed a higher weed dry weight (p<0.05) when compared to that of the Parthenium-free site, due to higher emergence of the Parthenium weed in the former. Mulching was the best sub-plot treatment for Parthenium control in combination with the main plot treatments. The average fruit weight of tomato at the Parthenium-free site was 7.8 % higher than that of the Partheniumsite (47.8 g per fruit). In the Parthenium-site, mulching resulted in a 6.4% higher fruit weight and 58% higher total yield (p<0.05) when compared to the rest of the sub-plot treatments. The weed competition negatively affected the tomato yield (Y=-0.7551X+7.88; R2=0.58; p<0.05), with Parthenium weed playing a dominant role. Mulching with <em>G. sepium</em> coupled with manual weeding during land preparation or ploughing would suppress growth and development of weeds including<em> P. hysterophorus </em>and enhance yield of tomato.</p><p> </p><p>Tropical Agricultural Research Vol. 25 (1): 56-68 (2013)</p>
|