To cite this article: P. T. P. Clifford & S. J. M. McCartin (1985) Effects of pre-harvest treatment and mower and header types on seed loss and hard seed content at mowing, recovery, and separation when harvesting a white clover seed crop, New Zealand Journal of Experimental Agriculture, 13:4, 307-316, DOI: 10.1080/03015521.1985 Abstract A preliminary on-farm survey of white clover (Trifolium repens L.) crop seed losses at pickup for threshing and separation indicated that operator skills and/or machinery were responsible for losses rather than crop management skills. Losses ranged from 12 to 39%, the mean being 200 kg/ha of which 130 kg/ha was hard seed. In subsequent seasons at Lincoln, losses resulting from crop condition at mowing (crops were either cut green or desiccated before cutting morning or afternoon) were measured after cutting by a standard sickle bar (17 fingers/m), a lespedeza bar (25 fingers/m), a double reciprocating knife, or a twindrum rotary mower. The latter 2 mowers were also assessed in an out-of-condition crop. Losses from mowing ranged from 17 to 59% of the total harvestable seed yield (200-540 kg/ha). Use ofa rotating-finger type pickup attachment for swath recovery for threshing reduced mowing losses by 10% to range from 60 to 340 kg/ha. Lowest losses (5%) were recorded on a desiccated crop cut in the morning with a double reciprocating knife mower. The rotary mower gave the highest losses under good harvesting conditions for all treatments, but performed better than the double reciprocating knife on out-of-condition crops. Threshing and separation losses range from 20 to 200 kg/ha. Increase in pickup width tended to further concentrate losses in the header offal. The overall range of offal trail + pickup losses (460-720 kg/ha) indicated that some form of recovery would be practical. Hard seed content of seed lost both at mowing and at crop pickup for threshing ranged from 60 to 71 % -a hard seed return to the soil for the crop pickup process of 40-210 kg/ha. The drastic reduction in hard seed content which occurred in the threshing and separation processes was more than offset when this loss was restricted to the limited distribution of the offal trail and its associated recovery loss. The overall hard seed return in this area ranged from 160 to 420 kg/ha. These results indicated an urgent need for improved harvesting technology (a) to reduce seed loss and (b) to reduce hard seed content of the fraction remaining.