Get reliable science

The effect of distance from a water point on range condition was examined for two rangeland types in central Australia; an open woodland community and a mulga annual community. Range condition was measured on a systematic grid pattern around the bore using the STARC method. In open woodland, range condition improved at about 2 km from water, but in mulga annual no change was detected due partly to a low number of sites close to the bore. The location of sampling sites adjacent to station roads did introduce some bias vrhen compared to the mean score over the whole area. The amount of bias acceptable must be seen in relation to factors such as time available for survey and the ease with which sample sites can be relocated. Two to four sampling sites located at a distance greater than 2 km, in rangeland preferred by cattle gave a mean score within 30% of the true mean for the whole area with the influence of 'sacrifice area' excluded. A reasonable accuracy of range condition sampling can be achieved in extensive central Australian rangelands. by stratifying the areas sampled with respect to country preferred by cattle and distance grazed from water.