REDUCING LOSSES OF SURFACE-SOWN SEED DUE TO HARVESTING ANTS

Four experiments were carried out at Orange, New South Wales, in 1978 to test the effect of coating seeds with insecticides on their removal by seed harvesting ants (Pheidole sp.). Treated or untreated seeds of Phalaris aquatica were placed 10 cm from the entrance to an ant nest and the number of seeds taken by ants noted daily. Further experiments tested the effect of coating seeds with insecticides on the germination of P. aquatica and Medicago sativa and on the viability of rhizobia applied to the seed. Permethrin, at 1.50 and 2.25 g a.i. kg-1 seed and bendiocarb at 0.75, 1.50 and 2.25 g a.i. kg-1 seed significantly reduced the rate of removal of seed by ants when compared with the rate of removal of untreated seed. Ants removed untreated seed at > 150 seedsinest day-1 while treated seed was removed at an average of 5 seedslnest day-1 over a 14 day period. The wettable powder formulation of permethrin was as effective as the miscible oil formulation. The activity of ants from nests that took treated seed was reduced by the higher rates of bendiocarb, but not affected by permethrin and the low rate of bendiocarb when compared with the activity of ants from nests that took untreated seed. The miscible oil formulation of permethrin reduced the rate of germination of P. aquatica, but the wettable powder formulation of both permethrin and bendiocarb had no deleterious effect on germination. Neither permethrin nor bendiocarb had deleterious effects on the survival of Rhizobium meliloti or R. trifolii. © 1979 ASEG.