The work was conducted at the Rafael Fernandes Experimental Farm, from August 2011 to January 2012, with the objective of determining the amount of roostertree incorporated in the soil, and the divisions of this amount that should be used for the agronomic performance of the carrot. The experiment used a completely randomized block design with three replications, in a 4 x 3 factorial scheme (4 amounts and 3 divisions). Treatments consisted of four amounts of roostertree (6; 19; 32 and 45 t ha-1 in a dry base), in three divisions (30% 15 days before sowing (DBS) + 70% at 30 days after sowing (DAS), 40% at 15 DBS + 60% for DAS and 50% at 15 DBS + 50% at 30 DAS). The carrot cultivar planted was 'Brasilia'. The characteristics evaluated were: plant height; number of stems per plant; shoot dry weight; and the commercial, total and classified productivity of the roots. The greatest agronomic performance of the carrot was obtained from the amount of 45 t ha-1 roostertree incorporated in the soil, divided into 30% 15 days before sowing + 70% 30 days after sowing.
Daucus carota; Spontaneous-species fertilisation; Agronomic efficiency
