Thesis (BS Biology) -- University of the Philippines Mindanao, 2007

Seedbanks, counts, dry weights, and diversity indeces of broadleaf weed species were compared under two contrasting weed management in a banana plantation. The weed management treatments were manual weeding and chemical weed control using paraquat and glyphosate. Seedbanks observed at the start of the experiment and six months after treatments yielded nine broadleaf weed species. These were Ageratum conyzoides L., Bidens pilosa L., Cleome rutidosperma DC., and other broadleaf weed species namely, Borreria laevis (Lam.) Griseb., Calopogonium muconoides Desv., Hedyotis corymbosa (L.) Lam., Mimosa Invisa Mart., unidentified 1, and unidentified 2. The same set of broadleaves weed species was found in the manual and chemical plots. The Shannon-Weiner Index shows that species diversity was not significantly affected by weed management treatments. Over the six month sampling period, the manual treatment did not significantly reduce species diversity: but the chemical treatment significantly increased the species diversity in 15% slope. The counts of broadleaf weed species seedling from the seedbanks were not significantly affected by weed control method. In one year (10 sampling months) of field observation, there were significantly lower counts and dry weights of the broadleaf weed species in the chemical plots than in the manual plots in both 15% and 25% slopes. The rainfall patterns in the two experimental sites were very similar. There was a peak in March and a sudden drop in October. There was no significant association between rainfall and counts of broadleaf weed species in the two plots in both 15% and 25% slopes. However, rainfall affected the relative efficacy of the weed control treatments. The use of paraquat and glyphosate for weed control reduced weed management cost compared to manual weeding by 85.56%