Abstract Minerals, nitrogen, structural carbohydrates, and lignin analyses from field collections of four species of high country tall-tussocks, species of Chionochloa, are reported. Differences among species were found for cellulose and hemicellulose, and values for these were higher than is common in pasture grasses. In C. macra and C. rubra cellulose: hemicellulose ratios were equal to 1.0, a condition not common to herbaceous plants. C. rubra was highest in estimated lignin. Substantial intraspecific ranges for N, P, K, Ca, and Mg occurred. Nitrogen rarely exceeded \% and was lowest in C. rubra. There were significant interspecific differences for P, K, Ca, and Mg; P was higher in C. macra and C. rigida than in C. flavescens and C. rubra; K and Mg were lower in C. rubra than in other species; for Ca there was a clear differentiation C. flavescens>C. rigida~> C. macra> C. rubra. In the field survey C. macra was significantly higher in total ash than other species. Silicon reached almost 2% in a field sample of C. macra, a value twice that of other species, but was not conspicuously higher in garden conditions. C. rubra contained less Al than other species both in the field and in the uniform environment gardens. Variation in the frequency with which tall-tussocks are grazed by stock may be related to differences in mineral composition within species. Differences between species in acceptability to stock appear to be related to differences in mineral and lignin composition.
Bannister & Ward (1981) recently examined the changes in non-structural carbohydrate contents of Chionochloa rigida and macra during winter . We agree that their observations help to provide a physiological explanation for the success of rigida at lower and the restriction of macra to higher altitudes (p. 239). However, we cannot agree that their observations in any way support the interpretation that mixed stands of macra and rigida result from e. rigida extending its range upslope in response to a 'post-Little Ice Age' Warming (Meurk 1978, p. 47). For Canterbury mountains Connor (1965) showed that irrespective of aspect, at lower in the Rakaia catchment, macra (then known for that area as rigida) dominated on fine-textured soils, whereas C. flavescens (eastern South Island form) dominated on coarse-textured rubbly soils. Nearby in the Ashburton catchment, a similar relationship was found between C. macra (by then known as Ozionochloa Q) and rigida Connor & MacRae 1969). For Central Otago mountains, where land surfaces are more stable and altitudinal zonation is strong, macra was not recognised as a distinct taxon but merely as a high altitude ecotype of rigida influenced by climatic gradients (Mark 1965). Numerous experiments and plant tissue analyses have since shown that macra has very distinctive nutrient requirements and a mineral ecology that has adapted it to grow on older soils than those of rigida or flavescens (eastern South Island form) (Connor et a1. 1970, Molloy & O'Connor 1970, O'Connor et al. 1972, Williams et a1. 1978). These observations and experiments provided an explanation for the patterns described by Connor (1965) and Connor & MacRae (1969). They also enable us to suggest that land surfaces with pure macra, or rigida with macra, will have some underlying soil differences compared with pure stands of rigida on the same slope and with the same climate - a process of logic first described by Jenny (1941). It was such a situation in the Ashburton catchment that Williams et al. (1977) referred to. Here there are no major differences in parent rock, but the soils beneath the mixed stands are finer textured than those beneath pure rigida. We interpret these patterns as resulting from the influence of the state factors-parent material and time--on the whole ecosystem, in the sense of Jenny (1958, 1980), without having to introduce the more complex and tenuous notion of climatic change. We have abundant unpublished evidence to further support our argument, but have based this reply solely with reference to published field examples and experimental or analytical data.
Summary There are three major tussock grassland communities in the Mackenzie Country: (i) fescue-tussock grassland where Festuca novae-zeladiae is abundant, (ii) red- tussock grassland where Chionochloa rubra is the physiognomic dominant, and (iii) snow-tussock grassland where Chionochloa rigida is the physiognomic dominant. Socioligical analyses of these communities in which severa phases are recognised are given in tables. The history of the grasslands is traced from the destruction of the forests through tall-tussock stages to the current extensive short-tussock grassland of recent origin.