Assessment of standing dry matter on hill country by cutting to ground level—sources of error

Five trials were carried out to determine important sources of error associated with the estimation of standing dry matter (DM) in hill country. Pasture samples were taken by cutting to ground level with portable shearing handpieces. Cutting within quadrats of area 0.125 m 2 gave estimates of pasture yield which had a similar standard error to estimates obtained from 0.25 m 2 quadrats. Components of variation resulting from the cutting machine, placement of quadrat, and person collecting the cut sample were all negligible, whereas the component resulting from the person doing the cutting (cutter) was highly significant—this was so in the green and dead fractions of the sward as well as for the total DM. It is not practicable to estimate cutter variation precisely nor does it seem possible to reduce it significantly by staff training. Subjective methods of site selection were compared with semi‐random site selection — the estimates given for pasture DM were not noticeably different. The associated standard errors were smaller by subjective selection.