Spatial heterogeneity and other sources of variance in sward height as measured by the sonic and HFRO sward sticks

AbstractMeasurements of undisturbed sward surface height (sward height) were made using the HFRO and sonic sward sticks at 20 cm intervals along 100 m transects laid across paddocks grazed by either sheep, cattle or goats. Measurements of sward height were also made at 5 cm intervals along 10 m transects laid across the same paddocks using the HFRO sward stick alone. The variance of the measurements from each transect was partitioned across the range of spatial scales of roughness present using spectral analysis. The variance of the sward surface height appeared to be dominated by variation with spatial scales of less than 1 to 2 m. The variances of measurements made with the two sward sticks were similar despite the greater sampling area of the sonic sward stick. This was due to the presence of variation at spatial scales much larger than the sampling area of either instrument and to a greater contribution from non‐spatially related sources to the variance of the sonic sward stick measurements. The use of weighted disc meters for obtaining the mean height of continuously grazed sward deserves re‐examination, while automation of the sonic sward stick may be worthwhile if the spatial distribution of herbage is to be described.