Predicting nitrogen deficiency in Eucalyptus nitens plantations using soil analysis and budgeting methods

Areas of Eucalyptus nitens plantations in Tasmania are increasing and are expected to enhance Australia's production of wood products. Standard silvicultural practises involve N fertilisation of E. nitens plantations at planting and later ages, however not all plantations respond to N fertiliser applied either at planting or at a later age. A method to predict N fertiliser responses is required to prevent wastage and obtain maximum productivity of Tasmanian E. nitens plantations. In this study nitrogen budgeting and soil analysis methods were examined as predictors of the timing of N fertiliser responses. Fourteen established research localities within Tasmania were used for the study, covering a wide range of sites planted to E. nitens. Of the 14 sites, 11 were on basalt, with single representatives on siltstone, granite, and alluvium. Rainfall ranged from 1039-1913 mm per annum and elevation ranged from 170 m to 650 m. Sites were variously fertilised, some at planting, others at growth stages up to ten years. Net nitrogen mineralisation (NNM) was estimated in situ at five sites encompassing a wide range in N fertility. NNM in these sites ranged from 13 to 188 kg N ha\\(^{-1}\\) year\\(^{-1}\\). Soil analyses for total N, total P, total C, hot KCl extractable N, soil solution and cold KCl extractable N were examined as indices of NNM. Total N, total P, total C, and hot KCl extractable N did not show large temporal variation and the values attributed to these indicators separated the five sites into two groups, being sites with NNM greater or less than 40 kg N ha\\(^{-1}\\) year\\(^{-1}\\). Sites of NNM >40 kg N ha\\(^{-1}\\) year\\(^{-1}\\) had total N greater than 0.4%, total P greater than 0.2%, total C greater than 8% and hot KCl extractable N greater than 100 µg N g soil\\(^{-1}\\). Sites with lower NNM had concomitantly lower values of these soil analyses. The biomass of tree components was estimated from pre-determined regressions with tree size. Measurements or estimates of N concentrations led to estimates of N content in tree components at 14 sites. Nitrogen content of litterfall was estimated at two sites with high soil N analysis values, one of known high in situ NNM rates. Maximum estimated N uptake in the combined above-ground biomass, the below-ground biomass and the litterfall was 162 kg N ha\\(^{-1}\\) year\\(^{-1}\\). Fertiliser responses were deemed to be significant (P<0.05) when increments in stem diameter at breast height (1.3 m) over bark of fertilised trees was significantly greater than diameter increments of unfertilised trees. For sites fertilised at planting the initial year of significant response was recorded, while for sites fertilised at a later age (age 3-10 years) relative responses (diameter increment of fertilised trees / diameter increment of unfertilised trees) were recorded. Of sites fertilised at planting, two had responded by age two years, one by age three years and three had not responded by age three years. Relative responses of sites fertilised at a later age ranged from 99% to 171%. The formation of a simpler partial budget, where N supply was only in situ NNM (0-10 cm depth; 5 sites) and N demand was only the N increment into the above-ground biomass, was always able to predict a significant response to fertiliser. However, when NNM was estimated with soil analyses, fertiliser responses were accurately predicted in only five of 14 sites. The six sites fertilised from planting could be separated, on the basis of total soil N, into those that responded [significant (P<0.05) increase in stem diameter at breast height in N fertilised trees compared to unfertilised trees] before age three years (n=2), at age three years (n=1) and after age three years (n=3) with total N of <0.28%, 0.28-0.51% and >0.51% respectively. All sites that responded to N fertiliser had soil solution and cold KCl extractable nitrate below 0.1 mM and 1 µg N g\\(^{-1}\\) soil respectively. The combination of total N and soil-solution or cold KCI-extractable nitrate allowed prediction of N fertiliser responses at all of the 14 study sites. Both soil analysis and one budgeting method were successful in predicting N fertiliser responses of E. nitens plantations. The budgeting method was most successful when NNM was estimated in situ, which together with the tree measurements required, is labour- and time-intensive. Unless estimation of NNM and N uptake can be simplified, the budgeting technique is unlikely to become part of standard silvicultural practices. A soil analysis method using total N and soil solution or cold KCl extractable N is simpler and more likely to be used by forest managers.