bush in September on the control plot ranged from 14501b. per acre in 1946 to 4701b. 

 per acre in 194S, with a mean of 9701b. per acre over j2 seasons. The seasonal fluc- 

 tuation in the amount of all other forage was much greater; its green weight on the 

 ungrazed plot ranged from 23001b. per acre in 1946 to 151b. per acre in 1948. 



It has been found possible by correlation and regression to eliminate the direct 

 effect of season on the weight of bluebush forage in the grazed plots, and to convert 

 these weights to a 'constant season' basis, represented by the mean of the control 

 plot. (Fig. 5) 



After accounting for seasonal variation, the amount of bluebush on the lightly 

 and moderately grazed plots remained practically constant over the period 1940- 1945, 

 whereas there was a downward trend in the weight of bluebush on the heavily and 

 very heavily grazed plots. This trend was associated with heavy grazing of bluebush 

 during three seasons of low influential rainfall, including two drought years, 1943, 

 1944, followed by only moderate rainfall prior to September, 1945. All plots responded 

 to the favourable rains of 1946, but after eliminating the response due to season alone, 

 there remained an extraordinary increase in the weight of bluebush on the moderately 

 to very heavily grazed plots, which can only be ascribed to the more intensive graz- 

 ing with sheep at the higher livestock concentrations. There was a small but less 

 marked residual response on the lightly grazed plots. Since 1946, the seasons have 

 been unusually favourable, with the exception of 1948, and on the three plots sub- 

 jected to moderate to very heavy stocking the weights of bluebush forage, after some 

 recession from the peak of 1946, have fluctuated about levels considerably in excess 

 of their initial values. The high level has been maintained on the most heavily grazed 

 plot despite an increase in the mean stocking rate from 90 to 168 sheep per square 

 mile. The amount of bluebush on the lightly grazed plot has shown a gradual upward 

 trend over this period. 



Part of the increase in production of forage by bluebush on the grazed plots, rela- 

 tive to the ungrazed plot, is undoubtedly due to the substantial increase in the num- 

 bers of bushes on the former plots from 1944 to 1946. (Table 3.) A heavy seeding of 

 bluebush occurred early in 1945, and large numbers of seedlings germinated on all 

 plots, including the control. The greater vigour of the seedlings on the grazed plots, 

 and particularly where grazing was heaviest, was most noticeable and accounted for 

 the larger number of young plants which became established on these plots. (Fig. 6) 

 flfithin the grazed plots, a greater increase in numbers of young plants occurred 

 towards the southern end of the plots. (Table 4.) Sheep graze into the prevailing 

 southerly wind, with the result that this part of a paddock is inevitably most heavily 

 grazed. 



The increase in the weight of bluebush per unit area was not entirely due to in- 

 crease of numbers, and indeed, on all the stocked plots except those most heavily 

 grazed, the numbers have since fallen to levels comparable with the ungrazed plot. 

 The increased vigour and leafiness of the older bluebush on the heavily grazed plots, 

 compared with the ungrazed control, have been outstanding. Moreover, the larger 

 numbers of stock on the heavily grazed plots have been successfully carried in prime 

 condition, and with only a slight reduction in wool production per head. 



138 



