A method is described for obtaining pasture growth rates using a mower to cut samples at regular grazed intervals from small movable cages. This method was used in a standard series of experiments on grass-clover pastures to ascertain the seasonal growth pattern at various sites throughout New Zealand. Growth pattern calculations and variability are described, and the uses and limitations of the yield data are discussed. Details are given of supplementary climatological data.

Rainfall, soil moisture, and soil and air temperatures were measured over five years on north and south aspects (about 25°slope) of a grazed pasture and on the exposed intervening ridge crest at Coopers Creek, North Canterbury (43°IS'S) at about 380 m a.s.l. Solar radiation and wind were measured only on the ridge, and radiation received on north and south slopes was derived from ridge measurements. Potential evapotranspiration on all aspects was calculated from solar radiation and air temperature according to the Priestley and Taylor formula. Daily mean solar radiation on the ridge averaged 3321angleys per day with a maximum of 5911angleys per day in December and a minimum of 1181angleysper day inJune. Over a year the north aspect received 80% more radiation than the south. Northwest winds prevailed; windrun at 2.1 m above ground was usually over 200 km.d" and occasionally was over 2000 krn.d". Total potential evapotranspiration on the ridge averaged 1093 mm.a" with a maximum of 5.9 mm.d" in December and a minimum of 1.0 mm.d" in June. Over a year, potential evapotranspiration on the north aspect was almost double that on the south aspect. Average annual rainfall was about 1400 mm and was distributed evenly over all aspects, although the catch from individual rain periods varied by up to 80% among aspects. The north aspect soil was always drier than the south aspect soil through the growing season and sometimes approached wilting point between December and March. Moisture content in the ridge soil was intermediate between north and south aspects. Mean annual Stevenson screen air temperatures (1.2 m above ground) were measured as 1O.5°c on the ridge aspect and calculated as 13.7°c and 9.9°c on the north and south aspects respectively. These are equivalent to a latitudinal displacement of about 9°within New Zealand. Soil temperatures on the north aspect were always warmer than on south or ridge aspects under both clear and overcast skies. Ridge temperatures were similar to those on the south aspect. Mean annual temperatures at 10ern depth were 7.7°c on the ridge aspect and 13.8°c and 8.2°c on north and south aspects respectively. On the north aspect, soil maximum temperatures under short grass (2-4 ern) were higher than those under tussock (30-40 em) and soil minima were lower. On the ridge, soil maxima and minima under short grass were lower than those beneath tussock.

Properties in the· top 7.5 cm of soil on steep hillsides under pasture were studied at four sites nn Banks Peninsula, Canterbury, New Zealand. At each site a comparison was made of physical and chemical soil conditions in terracettes (or sheep tracks), which follow hillside contours. and in the slopes between adjacent tracks. Studies were conducted on north and south aspects of tracked hillsides on Takahe and StewartSummit soil types.On all sites the soils of the tracks were wetter than those of the slopes. They also contained more organic C and total N, and slightly more quick test P and K. Within each site the track and slope soils had similar porosities and bulk densities. Hillside ledges or benches made by sheep and used by them as resting places contained considerably more organic C and quick test P and K than tracks located below the benches. Bare parts of benches and tracks had similar total N values.On each soil type the soils on south aspects were colder, wetter, and more porous than soils on north aspects. They also contained more organic C and total N, and less quick test P and K.

Pasture growth rates are presented for two sites, Mona Bush and Winton, on the Southland Plains. Yields were obtained from grazing trials cut by a standardised method at regular intervals using a movable cage technique. The mean yield and standard error of individual samples over all years of measurement are given for standard dates at 14-dav intervals.

ABSTRACT