The experiment was carried out at the Universidade Federal de Mato Grosso do Sul, in Dourados (Brazil), from September 24th, 1999, to May 21st, 2000, in a Dystrorthox clay texture soil. Fresh matter yield of 'Cem/Um' and 'Macaquinho' taros cultivated under three populations (100,000; 125,000 and 150,000 plants.ha-1) was evaluated. A complete randomized blocks experimental design with four replications and treatments distributed in a 2x3 factorial scheme was employed. Plants were harvested at 217 and 240 days after planting date (DAP). Leaf blades, petioles, corms (RM) and cormels (RF) and fresh mass were evaluated. Petiole yield, harvested at 217 DAP, presented a significant increase of 2.2 t.ha-1 for 'Macaquinho' in relation to 'Cem/Um', and of 2.0 t.ha-1 under a population of 125,000 plants.ha-1, in relation to 100,000 plants.ha-1; when harvested at 240 DAP presented an interaction effect. RM fresh mass of plants under populations of 125,000 and 150,000 plants.ha-1, when harvested at 217 DAP, were significantly greater in 3.0 and 1.9 t.ha-1, respectively, in relation to those under 100,000 plants.ha-1. RF fresh mass yield at 217 DAP from 'Cem/Um' was significantly higher in 6.4 t.ha-1 than 'Macaquinho'. Harvesting at 240 DAP presented the best yields for 'Cem/Um', both RM (11.4 t.ha-1) and RF (56.7 t.ha-1), under populations of 125,000 plants.ha-1 and for 'Macaquinho' under populations of 150,000 plants.ha-1 (10.2 t.ha-1 of RM and 58.0 t.ha-1 of RF).
Colocasia esculenta; clones; populations; yield
