Product positioning in industrial markets is an important consideration when new materials or products are introduced into a mature market. An investigation was carried out into the position of established timber species as perceived by UK end-users of timber. Multi-dimensional scaling techniques were employed in obtaining and explaining the perceptual patterns of four market segments (ie, general furniture, furniture frames, general joiner, and mouldings). The potential position of P. radiata D.Don was determined by overimposing objective measures of its physical properties. The findings indicateds that although there were differences in the perceptions of the end-use segments, all four segments shared a common element in their differentiation of species as hardwoods or softwoods. It was proposed that P. radiata should be positioned amongst the premium softwoods. Therefore, promotional efforts should emphasise its superior finishing properties and should be supported by a unique selling proposition centred around the sustainable availability of long and wide clear lengths. Efforts should be aimed at end-users who exhibit diversity in their utilisation of species eg manufacturers of mouldings.

A computer program, FLGRADE, has been developed that can grade random-width factory lumber to the US Western Lumber Grading Rules. FLGRADE uses data obtained from digitized boards. Board profile (including wane) is represented using a series of line segments which combine to form a polygon. All other defect data are represented using diagonal coordinates representing the smallest encompassing rectangle. The algorithm creates a list of cuttings by considering areas of placement. Using dynamic programming, an initial 2-stage solution is generated based on the rip-first cutting procedure. Alternative solutions which allow more flexible cutting sequences (i.e. either rip first or cross-cut first) and which do not restrict the number of stages are then generated to determine whether a higher percentage of cuttings, and hence a higher grade, can be obtained.

Grades generated by FLGRADE for a set of boards for which defect data had previously been digitized, resulted in 80% receiving the same grade as had been manually assigned. Of the remaining boards, approximately 6% received a higher grade and 14% a lower grade. All computer assignments agreed with manual assignments within one factory lumber grade.

A study of spiral grain in Canterbury-grown Pinus radiata D.Don revealed that spiral grain varies significantly both within the tree and between trees; bending strength and stiffness decrease with an increase in the angle of spiral grain; corewood stiffness is less sensitive to spiral grain than outerwood stiffness; and spiral grain downgrades a significant proportion of the sawmill output from Canterbury-grown P. radiata.

Preliminary tests to measure free shrinkage strain on Pinus radiata at increased temperatures (dry-bulb 70 deg C, wet-bulb 43-68 deg C) were conducted on discs taken from an internodal portion of the stem of an 8-yr-old tree from a plantation at Rolleston, SE of Christchurch, New Zealand. Radial, tangential, and sectional shrinkage strains were determined against the moisture content. These tests showed unexpected scatter due to large and variable amounts of compression wood in the disks. There was little correlation between the measured strains and the percentage of latewood or the density of the samples.

The tests were therefore repeated using wood slats 10 mm thick from a tree grown in an environment which reduced the amount of compression wood (in Southland which is less windy than the area in the Canterbury region where the test discs originated). These tests indicated that a very small strain appeared at high volume-averaged moisture contents (greater than the fibre saturation point). This may have been due to contraction of the surface layers which had dried below the fibre saturation point. However, significant strains began at a volume-averaged moisture content of 33% (dry basis). Of particular significance is the development of a new technique to measure strain continuously at higher temperatures which should be useful for kiln-drying studies.

Alternative strategies for aggregating stands into croptyes were evaluated. Strategies tested included traditional grid methods, clustering methods, and a variable resolution approach. A benchmark stand-level model was developed within the FOLPI forest estate modelling system for a 231 stand estate. Croptying strategies were evaluated for their ability to match the results from this stand-level model. Comparisons were made in terms of the variation within croptypes, objective function value, forecast cash flows and woodflow volumes, stands specified for harvest over the short term, and problem size. The best croptyping strategy was the variable resolution approach in which the unique identity of the 41 stands within 6 yr of harvesting was preserved while the 190 younger stands were aggregated into croptypes. This hierarchical approach represents a compromise between the need for short-term detail and the desirability of modelling long-term consequences within the same model.

Data from 240 trees sampled throughout the climatic range of Eucalyptus saligna Sm. in New Zealand were used to develop and validate equations from which taper and barkthickness could be estimated. Tree breast-height diameter over bark (dbh) and tree height were used as the predictor variables. Conditioning guaranteed that the taper equation would predict dbh as over-bark diameter at breast height. Combination of the equations produced a composite under-bark taper equation which can be integrated to derive underbark volume estimates for any stem section, ensuring compatibility of taper and volume.

A Corrigendum to this paper is available here: Corrigendum - Development of a composite taper equation to predict over- and under-bark diameter and volume of Eucalyptus saligna in New Zealand

Understorey plant species composition of second- and third-rotation Pinus radiata D. Don stands was investigated in relation to site variation within three Kinleith Forest compartments varying in time since planting (1,13,29 years). Detrended correspondence analysis and detrended canonical correspondence analysis were used to analyse indirect and direct compositional gradients respectively. Within compartments, compositional variation related to topography and soil chemistry, and supported a view that these management units can be inherently variable over short distances. For example, in the 1-year-old compartment Microlaena stipoides (Lab.) R.Br, characterised exposed sites low in exchangeable soil cations, Buddleja davidii Franch. sites low in soil cations, with Pneumatopteris pennigera G.Forst, on sheltered sites, and Brachyglottis repanda J.R. et G.Forst, var. repanda on sites with high soil cations. Most species recorded in each compartment were indigenous, the proportion increasing with time since planting. The 1-year-old compartment showed some compositional affinities to the pre-planting shrubland, whereas the 29-year-old compartment had stronger affinities with indigenous forest, although it was not a close analogue of a nearby indigenous forest. Species richness was greatest in the 1-year-old compartment and least in the 13-year-old compartment, with the 29-year-old compartment of similar richness to comparatively diverse indigenous forests. This high species richness could be a consequence of repeated disturbance by forestry operations. This study does not support the view that Pinus radiata plantations are necessarily "biological deserts".

Pinus radiata D. Don cuttings from 3-yr-old donor trees were compared with seedlings in a spacing trial planted on farmland near Rotorua and grown on a direct sawlog regime. At age 7 yr, the cuttings were taller but slimmer and had less taper than the seedlings. Closer spacing at planting enhanced height but not diameter growth for both stock types. The most pronounced contrast between the two stock types was the better resistance to toppling and superior form, particularly straightness, of the cuttings. This trend had a marked effect on crop selection raitos, resulting in the need to plant at least double the number of seedlings to approach the crop quality of the cuttings. PC-STANDPAK was used to model results at plantation age 28 yr. It was predicted that the superior shape and form of the butt log of the cuttings would result in a substantially higher recoverable volume of clear grade timber than the seedlings. Consequently, the value of pruned logs per hectare for the cuttings was predicted to be 20% higher than that for the seedlings. This trial was planted on a fertile site, and early differences in form between seedlings and cuttings were more pronounced than observed at less fertile sites. However, the results observed here could be expected on other highly fertile sites.

Topsoils under planted conifer forests in the Canterbury high country have high levels of 0.5 M H₂S0₄-P (inorganic phosphorus) compared to published records from topsoils of grasslands on soils of the same high country soil groups. Group means under conifers were from 16% to 140% higher than under grassland but such differences did not appear to be consistently related to moisture class or natural fertility. Concentrations of 0.5 h Olsen-phosphorus in conifer topsoils were also compared with topsoil records for unimproved grassland, semi-improved grassland, and improved pasture for the same soil sets and soil groups. For forest topsoil samples, inorganic phosphorus and Olsenphosphorus values were correlated, especially within the naturally more fertile groups of drier and younger soils. Both forests and grasslands exhibited a similar but small decline in Olsen-phosphorus values with increasing precipitation. For most soil taxa, variability in Olsen-phosphorus was high under any vegetation cover, but Olsenphosphorus was clearly greater under conifer plantation than under grassland. Soils from the three development classes of grassland had similar mean Olsen-phosphorus levels whereas the average forest Olsen-phosphorus levels were generally 2 to 4 times higher than the average grassland level for the corresponding soil.

There was a significant enhancement of "plant available" topsoil phosphorus by conifer plantations in the montane zone across a wide precipitation range. Despite the high variability, this enhancement appeared to be of similar magnitude along the whole precipitation range sampled, a situation apparently different from that outlined for "inorganic phosphorus".

Analysis of wood properties in 257 pieces of dried, dressed, 90 x 45-mm lumber, randomly selected from a sawing study on Pinus radiata D. Don logs from a 25-year-old stand in Kaingaroa Forest, indicated that grain deviation in the corewood zone could be severe (up to 18°). An assessment of kiln drying degrade showed a severe problem resulting from twist. Cross grain has also been implicated in strength reduction, and so a study was designed to provide preliminary data on the relative effects of wood density, knot area ratio, and spiral grain (cross grain in sawn lumber) on tensile strength and stiffness in structural lumber.

The most influential of the measured properties were knot area ratio and wood density, but cross grain also had a significant impact on tension strength of "in-grade" lumber. On the other hand, grain deviation led to only a small, though significant, reduction in stiffness (E plank). It is therefore unlikely that spiral grain in standing trees contributes strongly to the low stiffness of structural lumber from young P. radiata plantations.

A computer sawing system, which simulated a multiple saw edger, was used to investigate the effect of log rotation on the timber conversions of swept logs of three sizeclasses. A sample of 100 logs was stratified with respect to both log size and sweep, and four sawing methods (half-taper live sawing, half-taper cant sawing, offset live sawing, and offset cant sawing) were applied with the computer sawing simulation system AUTOS AW to each log at 37 positions representing rotation at 10° intervals from 0° (equivalent to the "sweep up" position) to 360° inclusive. All 14 800 simulations used the same carriage configuration, saw kerfs, and target dimensions, and the timber conversion percentage was determined for each.

The results suggest that, under the conditions selected for the simulations, any log rotation can be used with half-taper cant sawing without sacrifice of substantial volume, but rotation is more important with offset cant sawing. Increased conversions were obtained when the log rotation was approximately 145°, i.e., midway between the "sweep in" and "sweep down" positions. For every angle, half-taper sawing obtained conversions that were at least equal to, if not greater than, those obtained with offset sawing. Results were consistent with the general rule of thumb that each 0.1 increase in the sweep:diameter ratio (deviation : s.e.d.) leads to a 5% decrease in timber recovery.

Trials in three sawmills directly compared conversion of pruned, plantation-grown (in New Zealand), Pinus radiata logs by split-taper and full-taper sawing systems. No decided advantage from either system was found in recovery of defect-free clears, either by volume or by lengths. Full-taper sawing reduced total conversion by 3.4-4.7%, conversion to combined clears grades by 2.2-6.5%, and gross log values by 4.9-10.9%. Full-taper sawing was more difficult to implement, required additional cutting to square cants and blocks, and was estimated to lower mill production rates by 5-20%. Combined results proved full-taper sawing inappropriate for conversion of pruned plantation-grown logs.

Increment cores (5 mm diameter) and wood discs (50 mm thick) were used to study wood basic density patterns in Pinus radiata D. Don in relation to tree age, ring number from pith, position in the stem, and site in about 1000 stems from the North Island of New Zealand. Based on the relationships determined within trees, between trees, and between sites, a model was constructed which can satisfactorily predict wood basic density according to tree age, ring number from pith, and height in the stem at any location, and can explain the main contributions to variations in the wood density of individual trees. Furthermore, the model predicts the average density of given log height classes from data collected from breast height (1.3 m) outerwood increment cores.

The detailed variation of wood density within the tree revealed by the simulation software can be directly applied by industry, e.g., in "growing forward" density predictions for stands assessed prior to harvest, and allocation of stands and log types to alternative processes.

Spiral grain in Pinus radiata D. Don is a well-known feature of the corewood region of the stem. The pattern of variation in a 25-year-old plantation was recently documented and detailed measurements in another two crops indicated that the general pattern established remained valid. The results were used to construct a model of spiral grain in P. radiata as part of a series of Single Tree Wood Property models.

Effects of physiological ageing on physical wood properties of Pinus radiata D.Don trees were assessed. In the first trial, seedlings and cuttings propagated from 1 - to 5-year-old parent trees grown in field trials on two contrasting forest sites (Kaingaroa Forest— pumice, Tairua Forest—clay) were sampled at 11 years. The second trial involved 25-year-old seedlings and cuttings, physiologically aged 12 to 16 years, which were sampled in Kaingaroa Forest. Wood characteristics assessed were stem volume, wood density, heartwood development, spiral grain, compression wood, tracheid length, and corewood percentage.

In the trial with trees aged 11 years (physiological age _£5 years), there were no significant differences or trends in wood properties associated with physiological age. However, the impact of site was significant for all wood properties measured. In the trial with trees aged 25 years (physiological age 12—16 years), aged cuttings had significantly lower wood density, longer tracheids in rings 2 and 5, and higher spiral grain angles in the first 14 rings when compared to seedlings.

Wood basic density was assessed on cuttings from 11 clones in a 16-year-old clonal trial of Pinus radiata D. Don, using breast-height increment cores, and wood disc samples from the butt, breast height (1.4 m), 6, 11, 16, and 20 m height. The cuttings came from 5-year-old trees, and were thus physiologically "aged". The sample clones were chosen from the 120 clones in the trial, to cover the available range of wood density and tracheid length. Tree average wood density values were calculated from gravimetric measurements on the discs and within-tree trends were assessed using a scanning X-ray densitometer.

There were significant differences among clones, among trees within clones, among rings, and in interaction between rings and clones. The effects of clones and rings were dominant. There were also significant differences in pith-to-bark trends among clones, and in height up the stem at a given ring number from the pith in half of the clones. There were significant differences in the breast-height ring density profile between ring groups 1-5 and 11-15 due to increases in overall density, and in the amount and density of latewood, but there were no apparent differences in ring profile among clones. Mean ring density for rings outside ring 7 was highly correlated with weighted mean density at breast height and with weighted tree means. Estimates of clonal repeatability indicate a high degree of genetic inheritance for basic density, in agreement with results from earlier studies. All clones, however, were of below-average density compared to typical seedling stock at the same age, suggesting a physiological age effect in these clones, derived from 5-year-old trees. Using a corewood definition of <400 kg/m³, only one of the clones reached this density level by ring 10 and most failed to reach it by ring 15.

Microfibril angle was measured at breast height on growth rings 1,5,10, and 15 from the pith, for two trees from each of 11 clones of Pinus radiata D. Don. Average breast height values for both trees and clones were calculated by weighting individual ring values by relative growth ring area. Significant variation was observed both among clones, and within and between trees for each clone. The clonal repeatability (an estimate of broad-sense heritability) of unweighted microfibril angle was 0.7. The corresponding single ring values were high for rings 1 and 15 but were much lower for rings 5 and 10. Differences between trees within clones could be attributed to the presence of compression wood in some cases. Compression wood was associated with larger angles on average, although this was a variable effect for individual ring comparisons. Microfibril angle in ring 5 was the best predictor of average weighted breast-height microfibril angle for individual trees, indicating a potential for screening of young trees if required. The poor correlation of ring 15 values with weighted tree means suggests that outerwood microfibril angle is independent of corewood microfibril angle, or that the relationship varies among clones.

Some evaluations of solid wood samples (including internal checking assessment, shrinkage, collapse and tension tests) were carried out on 15-year-old Eucalyptus nitens trees grown in the Kaingaroa Forest, New Zealand. There were significant differences between the stand density classes in basic density and moisture content, with the low density class having significantly lower earlywood density than the medium and high density classes. Significant differences were observed between all three classes in average ring density and latewood percentage. The extent of the internal checking observed will present a definite problem to many solid wood products. Collapse occurred readily and could be severe, but steaming effectively restored the collapse. The scanner used gave an accurate and reliable estimate of radial and tangential shrinkage combined. The tension test results were positively correlated with density and moisture content.

Significant areas of Pseudotsuga menziesii var. menziesii (Mirb.) Franco (Douglas-fir) have been planted in New Zealand using a relatively small number of provenances. The resource is regarded as suitable for a range of structural products, with little information available on provenances in terms of wood quality. Two provenance trials were sampled—34-year-old trees of 39 provenances across six sites, and 19-year-old trees of six provenances across three sites—and wood density patterns were examined for 10 provenances on four sites with 30 trees per provenance. There was a strong site effect on properties measured and correlations between growth rate and density were negative on all sites and at most five-ring group positions. Juvenile wood extended to between 10 and 15 rings. In the 34-year-old trial, significant differences were found both within and between sites for density, stem diameter, and level of heartwood. Differences between provenances in resin contents were also significant. Two Californian provenances, Stinson Beach and Mad River, were found to be of consistently high density and diameter growth on all sites. In the 19-year-old trial there were significant differences within and between sites for density but diameter differences were significant only between sites. The results indicate that the Ashley Forest provenance is superior, having good diameter growth and higher wood density.

Stratified sampling has been applied to the estimation of volume in export consignments of pulplogs from New Zealand. Minimal bias and greatly improved precision were demonstrated when this type of sampling was compared with the alternative cluster sampling procedure. Adoption of stratified sampling techniques has reduced the cost and increased the precision of average piece volume estimates used to derive overall volumes from a total piece count.

Stump removal and windrowing were used as a site-preparation technique to control infection by Armillaria spp. in two second-rotation Pinus radiata D. Don stands. Significant reductions in mortality due to Armillaria spp. from 10% to <1%, and from 22% to 5% at 5 years from planting were achieved. Parasitic sub-lethal infection at age 8 years in the same stands was reduced from 85% to 10% and 67% to 31%. Economic analysis predicted returns of $2761 to $8826 (mean $6249) for a cost of $3623 (at 8% interest rate).

A new species of Vizella, V. tunicata, distinguished from other Vizella species by the gelatinous coat surrounding each ascospore, is described. The fungus is parasitic on leaves of Pittosporum tenuifolium Solander ex Gaertner and Myrsine australis (A. Richard) Allan.

Coniothyrium ovatum Swart occurring on leaves of Eucalyptus leucoxylon F.J.Mueller is reported as a new host record and a new fungus record for New Zealand.

Results from four trials indicated that tree selection which takes place during the normal pruning phase of Pinus radiata D.Don (ages 3 to 10 years) is inefficacious. At these early ages it is not possible to identify many of those trees that are likely to be of superior form or size at the age of clearfelling. Accuracy of prediction improves with age, and by the time of production thinning (age 12-14) a manager can place considerable reliance in the choice of crop trees. Change of dominance appears to be more pronounced in lower-stocked stands.

Results from 29 semi-mature trials of New Zealand Pinus radiata D.Don indicate that height growth is positively correlated with final-crop stocking in certain circumstances, and where stockings are less than 800 stems/ha (higher stockings were not analysed). A subsidiary dataset, with ages from planting to 7 years, gave results that confirmed these. Twelve of the trials demonstrated a significant (p<0.05) decrease in annual height increment with a reduction in stocking. For these trials there was an average height loss of 0.13 m/year after thinning to final stockings, or approximately 2 m over one typical rotation, for every halving in stocking. One possibility is that this effect is due to wind: lower stockings incur greater wind turbulence, which reduces height growth.

In order to standardise descriptions of stand height, a new equation was calculated to predict Mean Top Height from Predominant Mean Height.

Changes in soil boron fractions, dry weights of biomass components (needles, branches, stem bark, stem wood) and uptake of nitrogen, phosphorus, potassium, calcium, magnesium, boron, copper, and zinc were studied in a 4-year-old Pinus radiata D.Don stand in Ashley Forest, North Canterbury, 1 year after application of urea at 0 and 400 kg N/ha and/or ulexite at 0 and 7.4 kg B/ha.

Significant responses to both the applied nitrogen and boron were measured in total above-ground tree biomass. The needle, branch, and stem components were significantly heavier in the trees fertilised with boron alone, but only the needle component was significantly heavier where nitrogen fertiliser had been added alone. Increased tree growth was associated with increased total uptake of all nutrients except phosphorus and potassium which increased only in trees treated with ulexite. Increased nutrient uptake was measured mainly in needles of trees treated with urea, but in needles and branches of those to which ulexite was applied. Nutrient concentrations of non-fertiliser elements remained similar or declined in trees to which fertiliser was applied. This was attributed to dilution effects and/or internal translocation between biomass components.

Boron application significantly increased boron concentrations in the current and 1-year-old needles whereas nitrogen application had no effect on nitrogen concentrations. Some of the added boron was retained in plant-available fractions in the top 20 cm of the soil 1 year after application and this, together with the potential retranslocation of boron within the ulexite-treated trees, will provide a future supply of boron for tree growth.

The results confirmed that ulexite is a suitable long-term supplier of fertiliser boron to P. radiata growing in low rainfall areas.

A pot trial was used to investigate the effect of the form of nitrogen (ammonium or nitrate) and the rate of nitrogen addition, at 1, 2.5, or 5 mM (14, 35, and 70 _ug/ml respectively), on Pinus radiata D.Don seedlings grown in sand culture.

Ammonium-fed seedlings were taller and had heavier shoots and roots than nitratefed seedlings, particularly at the highest rate of nitrogen addition. The ammonium-fed seedlings contained greater total amounts of copper and boron and had a higher percentage of copper, boron, and iron in the shoots than nitrate-fed seedlings. The nitratefed seedlings contained significantly more iron.

Increasing the rate of nitrogen addition from 14 to 35 _ug/ml increased shoot weights in the ammonium-fed seedlings, increased shoot height in the nitrate-fed seedlings, and increased shoot:root ratios for both forms of nitrogen. Shoot heights of ammonium-fed seedlings increased when the rate of nitrogen increased from 14 to 70 _ug/ml. Within each nitrogen form, the rate of nitrogen addition did not generally affect micronutrient concentrations in the shoots or roots.

The concentrations of boron and iron (at all nitrogen rates) and copper (at 70 _ug N/ml) were higher in the nitrate-fed roots than in the ammonium-fed roots but the concentrations of iron and copper (at all nitrogen rates) in the nitrate-fed shoots were significantly less than ammonium-fed shoots. The nitrate-fed roots were darker in colour and were more branched and finer than ammonium-fed roots. Organic acids, produced during nitrate reductase activity, may have a role in these differences.

Pinus radiata D. Don seedlings grown in a range of magnesium solution concentrations showed differences in root: shoot ratios, with those exhibiting magnesium deficiency symptoms allocating proportionately less resources to the roots than healthy seedlings. A foliar spray of 2% magnesium solution with 0.2% Pulse™ in water alleviated the deficiency symptoms and improved dry matter allocation to the roots. In a 7-year-old P. radiata fertiliser trial, magnesium fertiliser treatments caused no improvement in basal area or height after 6 years but foliar magnesium concentrations had been raised above the critical level. Trees with adequate foliar magnesium had nearly double the fine root biomass of those with inadequate concentrations. This suggested that below-ground dry matter allocation was decreased in deficient trees, and that the noted slow growth response of P. radiata to magnesium fertiliser may be due to the need to rebuild the root system before an above-ground response occurs. However, while fine root (<1 mm) biomass was increased in 3-year-old trees treated 18 months previously with magnesium fertiliser, no relationship between root:shoot ratio and magnesium application was found. It was suggested that the changes in root: shoot ratio may develop over a period longer than 3 years.

The Resource Management Act 1991 emphasises the importance of assessing the potential environmental effects of land-use change. Forestry is a potential land-use in the Mackenzie/Waitaki Basin but its effects are not acceptable to everyone. The preferences of stakeholders were investigated for different land-use options involving forestry. Using a "Q sort" technique, stakeholders expressed preferences for cards that presented environmental effects of a range of technically feasible land-use options, including forestry, agriculture, and conservation. Several indicators of the environmental effects for each land-use option were modelled, such as the visual effects of land-use regimes and consequential impacts such as wilding spread, and the non-visual environmental effects such as local income and employment and soil status. Seventy-seven respondents rated a total of 36 cards covering four landforms. Analysis of the results identified several clear "themes", or sets of preferences, characterised by a distinctive set of preferred landuse options, and a distinctive combination of effects. Each theme has specific criteria for judging acceptability. These preferences have possible implications for planning involving forestry under the Resource Management Act 1991.

The effect of pasture reversion to tea tree communities (comprising manuka (Leptospermum scoparium J.R. et G.Forst.) and/or kanuka (Kunzea ericoides var. ericoides (A.Rich.) J.Thompson) of varying ages) on landslide damage resulting from Cyclone Bola was studied in hill country on the East Coast of the North Island, New Zealand. Eighteen sites containing areas of both pasture and regenerating forest aged 6-79 years were selected from areas of steep, colluvial-covered slopes prone to shallow landslipping. Vegetation composition and stand dynamics were ascertained from sample plots, and aerial photographs were used to measure landslide damage. Manuka dominated younger stands but within 20-30 years of establishment, kanuka had become dominant. Mean canopy height stabilised at 13 m by age 40, basal area reached 40 m²/ha by age 30 years, and stand density reduced from 20 000 stems/ha at 10 years to 3000 stems/ha at 40 years. Landslide damage showed a rapid and highly significant reduction against increasing age of tea tree stands and was estimated to be 65% less than pasture at 10 years and 90% less at 20 years. These findings have implications for land-use options, including clearfelling of indigenous vegetation for planting of Pinus radiata D.Don.