Dense (Climo and Richardson 1984) and long (Watkin and Clements 1978) pastures at grazing offer some protection to soils from treading damage and reduce sediment loss compared with loss from a very short canopy pasture (Betteridge et al.
(1998) showed sediment loss was greater from slope soils, which were more prone to treading damage as a consequence of lower plastic and liquid limits, than from adjacent soils with easier contours.
To determine the immediate effects of sheep and cattle on the soil and pasture resource of a hill land, a single severe treading was imposed by sheep and cattle on the wet soil above its plastic limit.
Treading subplots comprised two 150-[m.sup.2] plots for cattle [[C.sub.1], [C.sub.2], treading+grazing, the second of which was stocked twice]; one 67-[m.sup.2] subplot for sheep [treading+grazing]; and one 67-[m.sup.2] untrodden plot (except for a light grazing by sheep to control pasture covers).
After the second treading of [C.sub.2], pastures were returned to their respective cattle and sheep rotational managements.
Within each treading subplot, five 2-m contometer transects were established down-slope, crossing banks, slopes, and tracks, to enable representative changes in microtopography during a grazing event to be accurately described.
The contometer was used to quantitatively measure `disturbance' and visually describe soil surface `damage' and pasture characteristics at each pin site, before and after treading. Measurements and scores were made before and after the first treading on all 4 subtreatments and following the second cattle treading on untrodden and [C.sub.2] treatments only.
Surface disturbance, or the difference in height of each pin caused by a treading event, was analysed as net disturbance and absolute disturbance.
This standard deviation value was used as a variable in analyses of variance to describe soil random roughness before and after treading. The effect of animal treading on soil roughness was described as the difference in units of standard deviation between roughness before treading and after treading.
The compactive effort used was reduced to 60% of standard, to simulate more closely the force applied by cattle treading. Bulk density, particle density, and total porosity were determined by the method of Gradwell and Birrell (1979) on 100-mm-diameter by 80-mm-deep intact cores taken randomly from 5 slope and 5 track sites from each of the treatments.
The layout was a split-plot design with 3 replicated blocks within each pasture type, which were each split into the 4 treading treatments.
Volumetric soil water content (vSWC) in both the 0-75 and 75-150 mm horizons at the time of the first treading averaged 55 and 58 [m.sup.3]/100 [m.sup.3], respectively, across tracks, slopes, and pasture types.
Soil on cattle pasture slopes was the least susceptible to compaction under stock treading as indicated by the smallest bulk density, the wettest CWC, and a comparatively flat compaction curve.