Fjölrit RALA - 05.12.1999, Qupperneq 137

Kevin F. O’Connor etal.
135
Weights of total soil N and P and amounts of exchangeable K and Ca in 0-200 mm
soil layers are also shown in Table 2, demonstrating also that above-ground live nutri-
ent pools and surface soil pools are not closely related over all nine localities. Total
soil N and total soil P in this 0-200 mm layer show less variation than any other pa-
rameter examined. Williams et al. (1978ab) examined the altitudinal distribution of
values for topsoil indicators of available nutrients in tall tussock grassland soils and
the influence of some such indicators, especially for P, on nutrient weights in live
Chionochloa shoots. Their larger survey did not reveal any marked relationships, ex-
cept for the tendency of total shoot P in different species to be more closely related to
particular soil P tests or different soil P fractions. It is not surprising, therefore, that
soil and live biomass nutrient contents vary independently in the present study over
several species. Similarly, there is no clear evidence of influence of live biomass pools
on soil pools.
Carbon and nutrient comparisons between stages of ecological degradation
Data for biomass comparisons of degradation stages are at present available ífom four
localities, but nutrient data are incomplete for all of these stages. Table 3 reports C
levels (calculated as 50 percent of biomass, inclusive of ash), N, P, K and Ca levels
for degradation stages at three of these localities.
Table 3 indicates substantial and consistent reductions in C in above-ground live
biomass, from tall tussock grasslands to shorter grasslands and weed communities.
These reductions are small in comparison with that in the single example of transfor-
mation from forest to tall grassland. Litter C is much lower in forest than is above-
ground dead C (including litter) in tall tussock grassland. Reductions in C in above-
ground dead associated with transformation of tall tussock grasslands to short grass-
lands are generally substantial, without being closely parallel to changes in above-
ground live material.
When the three localities are considered as replicates and tall tussock grassland
compared with weedy short grassland, above-ground C means are highly significantly
different (p=0.001). Apart from the substantial reduction in C in roots with the change
from forest, reductions in root biomass are not consistent. Nevertheless, tall tussock
grassland remains significantly greater in total biomass C than weedy short grasslands
(p=0.01). Differences in weight of soil C in 0-200 mm layer or in combined 0-400
mm layer are not clearly discemible from the three comparable sequences. When the
total ecosystem is considered to 400 mm depth, weedy short grassland appears to have
less C than tall tussock grassland, but the difference does not reach significance
(p=0.185).
The Old Man Range three-stage degradation sequence at 1300 metres measured by
Meurk (1978) demonstrated similar declines in above-ground live C, an overall de-
cline in above-ground C and total biomass C. It has not been included in the statistical
analyses here because the induced vegetation is botanically different from the weedy
short grassland of the other three comparisons, and because lack of herbage nutrient
data and of soil C and nutrient data for the degradation stages prevents valid system
comparisons.