Climatic and aspect influences on pasture production in New Zealand

Abstract

In view of the scarcity of published information on pasture production and climate relationships in New Zealand and especially such relationships in hill country these factors were studied at 23 lowland sites throughout the country and on three contrasting hill aspects in North Canterbury.

At 23 lowland sites, perennial ryegrass-white clover pastures, cut at 2 or 3 weekly intervals, were examined for their seasonal and annual yields, the variability of these yields, their seasonal pattern of growth and the contribution of clovers to seasonal and annual herbage production. Regional climates were also examined.

On the hill sites, microclimate studies over five years showed that north, south and ridge aspects differed substantially in temperature, radiation and potential evapotranspiration under comparable monthly precipitation. Differences between north and south aspects in temperature and radiation were equivalent to a latitudinal displacement of at least 6 degrees. Herbage production of these browntop-dominant swards was measured at 4 weekly intervals on a north and south aspect over four years within clearly defined soil strata. The variation in yields between strata within an aspect was usually greater than the mean difference in yields between aspects. Annual yields averaged over 4 years, ranged from 5.0 to 8.6 t DM/ha on the north aspect and from 5.5 to 7.5 t DM/ha on the south aspect. Within each aspect, herbage production, botanical composition, the percentage of digestible dry matter on offer to stock, and soil nutrients varied between soil strata.

When pasture production from each stratum was weighted according to the area occupied by that stratum within an aspect, annual production from the south aspect averaged 6 t DM/ha and that from the north aspect averaged 5.3 t DM/ha. Differences in annual production between aspects varied from near zero to 28%. The seasonal pattern of pasture growth was similar on north and south aspects usually with high growth rates in late spring and autumn, low growth rates in summer and negligible production in winter. Clovers contributed about 11% of total herbage production. Dead plant material contributed up to 25% of total herbage yield in some seasons, despite a preharvest trim cut, and was negatively correlated with the percentage of digestible dry matter in the available herbage on offer to stock.

Pasture growth-climate relationships were examined by means of forward stepwise regressions in which high statistical significance and/or early selection in the stepwise procedure were interpreted as indicating influential causal relationships. At the 23 lowland sites, the relationships of 3-monthly seasonal climatic parameters (winter=June to August) to annual yield were examined. Relationships of previous and current climatic parameters and previous yield to seasonal yield were also examined. Over all sites, the most influential covariates, expressed as a percentage of all occasions when the covariate, was available for selection from the covariate pool were as follows. Annual yields were influenced by rainfall (52%), minimum air temperature (48%), solar radiation (33%) and soil moisture deficit (20%). Seasonal yields were influenced by rainfall (31%), previous seasonal yield (23%), days of soil moisture deficit (21%), solar radiation (19%), soil moisture deficit (17%) and minimum air temperature (17%). Mean or maximum air temperature and temperature degree days above 5°C and 10°C were less often selected in regressions. At each site, the climate or yield in the previous season influenced yield in at least one season.

Similar relationships were examined for nine site groupings using regression analyses after adjustment for the effect of individual sites within the group. In many of these analyses, spring, summer and autumn seasonal yields were significantly influenced by either the previous season's rainfall and/or the current season's rainfall usually in association with the previous season's yield.

More detailed studies at six lowland sites and on three hill aspects examined relationships over harvest periods of two to four weeks. At the lowland sites, growth rates over all harvests were usually significantly influenced by soil temperatures, rates of pan evaporation, potential evapotranspiration, hours of sunshine and growth rates in the previous harvest period. Rainfall, days of rain, mean air temperatures, air temperature degree days >5°C or 100C and soil degree days >6°C were seldom significantly correlated with growth rates. Similar studies were made for particular phases in the seasonal growth pattern.

On the hill sites, relationships over all harvests differed according to aspect. On the north aspect and also at lowland sites which had seasonal soil moisture deficits, growth rates were most influenced by either a moisture or evaporative parameter together with growth rates in the previous harvest period. On the south aspect and also at lowland sites which had less serious soil moisture deficits, growth rates were most influenced by either rates of evapotranspiration or potential evapotranspiration.

The hill and lowland environments, pasture production and growth-climate relationships are compared and discussed.