ANIMAL PRODUCTION UNDER A SERIES OF PINUS RADIATA-PASTURE AGROFORESTRY SYSTEMS IN SOUTH-WEST VICTORIA, AUSTRALIA

Liveweight and greasy wool production data from young wethers were col lected for the period autumn to early summer for the years 1986 to 199 4 from an on-going Pinus radiata-pasture project at Carngham, Victoria . A plantation established in 1981 on a grazing property was thinned i n 1983 to give trees spaced at: (i) no trees (a grazing control), (ii) 8 m x 12 m, (iii) 4 m x 9 m, (iv) 4 m x 3 m in a 5-row belt then a 10 -row pasture gap, and (v) 2 m x 3 m (original spacing). By 1989 the de nsity of trees in these systems was slightly reduced by windthrow, ani mal damage and thinning to (ii) 60 trees/ha, (iii) 200 trees/ha wide-s paced, (iv) 200 trees/ha B-row belt and (v) 1360 trees/ha. All trees i n systems (ii), (iii) and (iv) had been pruned to 6 m by late 1992. Th e sheep stocking policy was initially based on a set 10 sheep ha(-1). Since 1991 the stocking rate has been varied yearly in an attempt to m aintain similar production per sheep. The pastures were re-sown in 198 9 with perennial rye-grass and subterranean clover cultivars. Since 19 91, wool production per ha from the 200 trees/ha wide-spaced system ha s been consistently less than in systems (i), (ii) and (iv). Wool prod uction among systems differed significantly since 1992. Expressed as a percentage of wool production achieved on open pasture, the relative values for the agroforestry systems from 1990 to 1994 as follows. [GRA PHICS] The results also show a consistent advantage of the belt arrang ement of trees rather than a wide-spaced layout. There are many practi cal advantages of the belt layout, although timber production may poss ibly be smaller and offset other gains.