Larix sibirica Ledeb. (Siberian larch) pulplogs were assessed for their approximate age, and their methanol and aqueous extractives content. Kraft pulping properties were assessed using whole-log and slabwood chip samples. The slabwood chips gave pulp in somewhat higher yield than pulp from whole-log material, but the papermaking properties of their pulps were generally similar. The L. sibirica samples examined had rather low density for the species, and their papermaking properties reflected this. Kraft pulps were obtained from the L. sibirica samples in about 2% lower yield than from Pinus radiata D. Don, and had papermaking properties similar to those of P. radiata corewood (top log) pulps. They would be suitable for packaging rather than printing grades.

Early performance of planted totara in comparison with other indigenous conifersWhole-tree basic density and internal checking were assessed in Eucalyptus nitens (Deane et Maiden) Maiden at six New Zealand sites, four in the North Island and two in the South Island, by sampling 15 trees/site, each involving one seedlot of Victorian provenance at a stocking of 1111 stems/ha. Sites ranged in altitude from 40 to 540 m and in latitude from 35Ý52?S (Dargaville) to 45Ý55?S (Drumfern). Site mean whole-tree density ranged from 428 kg/m3 at Raweka (Whakatane) to 476 kg/m3 at Mangakahia (Dargaville). Density at Kinleith, Wainui (both central North Island), and Millers Flat and Drumfern (southern South Island) varied little, from 445 kg/m3 to 459 kg/m3. From these and previous results, there was some indication that very high rainfall and high levels of foliar nitrogen, phosphorus, and magnesium led to extremely low wood density. Whole-tree density increases with age and, after an initial decrease in the lower bole, increases with height up the stem by 50 kg/m3 and more as age increases, resulting in lower density in the butt log than in upper logs. Internal checking, assessed in one breast-height disc per tree, was prevalent at all sites, especially in outer heartwood in both slowly kiln-dried and air-dried samples. More checks were found in air-dried discs than in kiln-dried. Many more checks were found at the North Island sites Mangakahia, Raweka, and Kinleith than at the high-altitude central plateau site Wainui or the South Island sites. Higher numbers of checked rings and total checks were associated with higher mean annual temperatures, short green crowns, and poor crown health. Far fewer checks were found at cooler sites where trees had much better crown health and longer green crowns. Checking data were reanalysed from an earlier study of twenty 15-year-old trees from Kaingaroa Forest (altitude 230 m) from four kiln-dried discs per tree at heights of 0, 1.4, 6.4, and 11.4 m, and from a 1-m air-dried diametral board sawn from the base of each tree. Numbers of checks varied enormously among trees and fell to zero above height 11.4 m. Numbers of checks in the board cross-section correlated moderately with those in the breast-height disc. It is considered that excessive checking in E. nitens may seriously reduce its potential for utilisation for appearance-grade lumber, particularly on warmer New Zealand sites where crown health is poor.

In species trials of Eucalyptus nitens (Deane et Maiden) Maiden, E. globulus Labill, and E. maidenii Labill aged 8 years and 11 years, 10 trees of each species/age were selected to compare wood properties and some lumber properties, ancillary to a previously-reported kraft pulping study. The trees were selected across the range of wood density for destructive sampling to approximate the species mean. Wood properties were measured on breast-height increment cores, on discs removed at 5-m intervals, and on a billet from height 5-6 m, and clearwood mechanical properties were measured from six test sticks/tree, cut from the billet. Eucalyptus maidenii and E. globulus had much higher density than E. nitens. Whole-tree densities averaged, respectively, 574, 540, and 451 kg/m3. There was no pith-to-bark gradient in density in the 6-m-height discs of E. nitens, but E. globulus and E. maidenii showed appreciable commensurate increases in density. In E. nitens, density decreased initially from the base to a height of 6 m, then increased rapidly upwards. Density increased steadily from the base in E. globulus, but decreased in E. maidenii. Bark was thickest in E. maidenii, especially at the tree base, and E. nitens had a higher proportion of heartwood than the other species. Tangential shrinkage, measured in the 6-m disc, was 17% for E. nitens, 12% for E. globulus, and 10% for E. maidenii. Tangential and radial collapse, measured as the reduction in shrinkage by steam reconditioning, was much higher in E. nitens than in E. globulus, which in turn was higher than in E. maidenii. Internal checking (assessed on the 6-m disc) was general and often severe in E. nitens, very occasional in E. globulus, and completely absent in E. maidenii. Average spiral grain angles for all species were less than 2.5Ý, and unlikely to cause drying distortion. Mechanical properties of the three species paralleled their wood densities. MoE values (GPa) for the outer rings 7 to 9 at height 5-6 m averaged 14.4 for E. maidenii, 13.7 for E. globulus, and 9.6 for E. nitens (at about 14% mc). Corresponding MoR values were 131, 122, and 88 MPa. The continued good growth and health of E. maidenii in Northland, combined with its lack of checking, low spiral grain angle, low shrinkage, low collapse, and excellent strength and stiffness, indicate its promise for solid wood products. However, growth-stress-related characteristics in sawn timber, and drying distortion remain to be evaluated.

Few growth models have been published relating to native tree species in Chile. The objective of this study was to develop site index curves for two of the most valuable native species in Chile - Roble (Nothofagus obliqua (Moerb.) Oerst.) and Rauli? (Nothofagus alpina Oerst. et Endl.). The study was based on stem analysis data. The method used for the reconstruction of each stem profile corresponded to one proposed previously. Site index models were generated using the difference equation method. The resulting site index systems revealed different growth patterns for both species. Among the two light-demanding species, Rauli? is characterised by better initial growth

We have devised a new field methodology to collect log data necessary to develop a graphical system for 3-D spatial reconstruction of the internal and external shape of the log, based on stem and cross-sectional analysis. The methodology provides a detailed internal and external description of the stem, including the full growth history of the tree as recorded in the rings. The technique may be used to analyse both straight and swept logs, and is specially suited for studying the evolution of stem sweep when a tree has lost its verticality, due to environmental factors such as snow and wind, or through destruction of the leader. We have developed the software required to process the data and to reconstruct the development inside the tree

Shoot pruning, root trimming, and exposure to sunny field conditions reduced root growth potential (RGP) of 1-0 Chinese fir (Cunninghamia lanceolata Hook) and Masson pine (Pinus massoniana Lamb.) seedlings. The total number of new roots(TNR) and number of new roots greater than 1 cm (TNR > 1 cm) of Masson pine decreased by 53% and 56% respectively after removal of half the foliage and decreased by 70% and 82% respectively after pruning of all the foliage; both the TNR and TNR > 1 cm of Chinese fir decreased by 11% after pruning of half the foliage and 19% and 45% respectively after pruning of all the foliage. The TNR and TNR > 1 cm of Chinese fir decreased by 29% and 33% respectively after trimming of 40% of the lateral roots; those of Masson pine decreased by 69% and 78% respectively. RGP of both Chinese fir and Masson pine seedlings decreased rapidly after total exposure to only a few hours of bright sun. After the seedlings were exposed for more than 2 hours, the TNR of Chinese fir fell below 10. The TNR of Masson pine appeared to be less susceptible to exposure and TNR dropped below 20 after 4 h

Conventional sampling schemes using bounded or point plots in a woodlot or stand of small area require a high proportion of the trees to be measured when recoverable (merchantable) volume is being estimated to the degree of precision normally required. Where the cost is high relative to the value of the woodlot, too few plots may be established to obtain useful confidence intervals and the trees within the plots themselves may be assessed too quickly for the predictions to be accurate. Individual tree sampling is often preferred over plot-based schema. PhotoMARVL, a method based on photographic image analysis of the stem, was developed to improve the accuracy of measurement over existing visual systems, but significantly adds to the cost of conventional pre-harvest inventory. Sampling with probability proportional to prediction (3P sampling) can efficiently utilize a more accurate though more costly measurement method such as PhotoMARVL. Additionally, it is useful where the stocked area of a woodlot is uncertain, as all the trees are visited and counted. In this study, 3P sampling was evaluated using pre-harvest inventory data of rotation-age Pinus radiata in New Zealand to test under what circumstances this system could be cost effective. Where individual trees were visually assessed for diameter at breast height 1.4 m to obtain a quick estimate of recoverable volume and a subsample was more accurately measured for product volumes using the conventional MARVL cruising technique, 3P sampling was the most cost-effective technique for woodlots of 5 ha or less. In less than a day, a two-person crew could estimate the potential recoverable volume by broad log-product classes to within 10% of the mean. As the variability of a stand increased, or as the requirements for precision increased, the size of the area below which 3P sampling was competitive increased to between 20 and 40 ha. The coefficient of variation of the ratio of the quickly estimated volume to the MARVL measured recoverable volume is likely to be between 15 and 20%, across all tree sizes, independent of the variability of tree size. Utilizing PhotoMARVL on a small subsample of trees within the woodlot would improve the accuracy of any value estimate and add the cost of approximately 4 person-hours. For larger areas, Point-3P sampling could be applied, but the combination with PhotoMARVL would increase costs by 60-70% over conventional double sampling with MARVL alone

Trials of 49 eucalypt species were established in 1979 in the Wairarapa district at Kahuiti and Pakaraka, New Zealand, originally to test species for their potential to stabilize erodable land for pastoral use. Trials were planted in a randomized complete block design with five replications of four-tree row plots of each seedlot (paired rows of four trees of species with only a single seedlot). The species included Corymbia maculata [E. maculata], E. cladocalyx, four stringybarks (including E. muellerana and E. globoidea), nine ashes (including E. fastigata, E. regnans, and E. obliqua), seven peppermints, and 18 gums (including E. nitens). Because of heavy thinning at Pakaraka, the Kahuiti trial only was assessed at age 22 years on production forestry criteria: diameter at breast height (dbh), stem straightness, malformation, crown health, and number of potential 5-m sawlogs per tree. The 12 best-grown species for mean tree dbh at Kahuiti, were ranked: E. globoidea, E. muellerana (stringybarks), E. obliqua, E. fraxinoides, E. regnans (ashes), E. cordata (gum), E. delegatensis, E. fastigata, E. sieberi (ashes), E. cinerea, E. kartzoffiana, and E. nitens (gums). The ashes, with addition of E. nitens, showed a combination of best diameter growth (apart from the two stringybarks), straightest stems, least malformation, good crown health, and largest number of sawlogs per tree of all groups. The peppermints were generally slower-growing and more sinuous than the ashes. Some of the gums grew well and most survived better on this adverse, eroded site than the other groups. Superiority of E. fastigata and E. obliqua was confirmed by other trials in Hawke's Bay and the Wairarapa region. The apparent good growth and health of E. globoidea and E. muellerana in the Wairarapa district, also reported in trials in Northland, hint at the potential of these known good sawtimber species

Premature defoliation is common in young plantations of Eucalyptus nitens in parts of New Zealand where stands are infected by the leaf spot fungus Phaeophleospora eucalypti, often accompanied by infection by species of Mycosphaerella. Four young eucalypt species and provenance trials aged between 1 and 3 years were evaluated for infection and juvenile leaf retention, to help identify alternative species or provenances that might be a suitable substitute for new plantings on disease-prone sites. Families and provenances of E. nitens from states in Australia, i.e., in New South Wales were less infected by both fungi, and had better foliage retention, than those originating from Victoria. Variation was not apparent between provenances within each State, but there were significant differences between families within provenances. Although common on E. nitens, infection by P. eucalypti was found only at a very low incidence, or not at all, on other species tested from Section Maidenaria. However, infection by Mycosphaerella species and premature defoliation were both heavy on subspecies of E. globulus, particularly subspecies globulus, subspecies pseudoglobulus, and subspecies bicostata. Infection and defoliation by Mycosphaerella species were moderate on E. nitens, comparatively low on E. globulus ssp. maidenii, and very low on two seedlots of E. benthamii. There was significant variation between provenances within several E. globulus subspecies in infection by Mycosphaerella species and in foliage retention. Infection by Mycosphaerella species was inversely correlated with foliage retention for most eucalypt species and subspecies, as was that of P. eucalypti on E. nitens. Tolerance to frost injury, as measured by percentage shoot survival, was greatest in E. nitens and E. globulus ssp. maidenii, and least in E. globulus ssp. globulus. Final selections will depend on further evaluation of health, wood production, and pulping quality, as trees mature and produce only adult foliage later in the rotation. However, initial results based on health alone confirm that there is alternative material available for reducing disease levels in young pulpwood plantations.

Specimens of Pseudocercospora on eucalypt foliage collected over the past 22 years and held in the New Zealand Forest Research Institute Mycological Herbarium were examined. In addition to the previously recorded Pseudocercospora eucalyptorum, three new species of Pseudocercospora: P. acerosa sp. nov., P. crousii sp. nov., and P. pseudobasitruncata sp. nov. - have now been described. These descriptions can be used to identify the species found in New Zealand, and contribute to the worldwide key to Pseudocercospora spp. on eucalypts

A simple water balance model was used to explain the difference in basal area growth over 3 years for 9- to 12-year-old Pinus radiata D. Don growing on a dryland site in the presence of a broom (Cytisus scoparius L.) understorey and where the understorey was removed. With an annual average rainfall of 600 mm, seasonal water deficits are characteristic of the well-drained, stony silt-loam soil. During the summer months, root-zone water content fell to approximately 90 mm, about 50% of the winter maximum. Modelled estimates of daily root-zone water storage in the two treatments were in close agreement with neutron probe measurements. Pinus radiata basal area increment over the experiment was approximately 9% higher in the treatment where the broom understorey was removed. Annual basal area increments for both treatments were highly correlated (r=0.97) with an annual integration of root-zone water deficits. The approach used in this study highlights the importance of root-zone water storage in regulating productivity at dry sites and confirms the success of water balance modelling to explain the impacts of understorey control on tree growth

A database consisting of 299 Pinus radiata D. Don sample plot periodic growth data, soil measurements (mainly from the A horizon profile) gathered from within each plot, as well as relevant climatic data, was updated and revisited. The plots were in various forests in the North Island of New Zealand. The soils relevant to the plots were grouped according to the New Zealand Soils Classification from which 11 classes can be recognised. These classes were examined by a multivariate discrimination analysis utilising A horizon soil variables as predictors. The various classes separated into two major groupings, with virtually all the classes being correctly classified for at least 75% of the data. The major discriminators were silt and clay percentage, the depth of the A horizon, organic carbon (%), and total nitrogen (%). Phosphorus (Bray or Olsen), Bray-extractable cations (calcium, potassium, and magnesium), average resistance, and soil pH had no significant effect on the separation of the classes. These data were augmented by climatic data and attempts were made to build a predictive system for mean top height (average height of the 100 largest trees by diameter at breast height). At first, separate equations were assayed for each soil class, but insufficient replication for some soils necessitated a pooled model. A nonlinear equation was constructed that was largely unbiased over the soil classes with an approximate R2 value equal to 0.77. The residual root of the error mean square was around 3 m. The dominant predictor variables were the age of stand when sampled, depth of the A horizon, average wind velocity, mean annual rainfall, and altitude. The model should be used cautiously as it is very difficult to produce a predictive system of stand growth without partially confounding the effects of soils, location, and climate

A comparison of planted and natural second-growth stands of kauri (Agathis australis Salisb.) in the North Island of New Zealand identified stem diameter as the principal factor for predicting heartwood presence and quantity. Age was a secondary factor, with older smaller stems from natural stands having disproportionately more heartwood than would be predicted from their stem diameter. Heartwood was found in kauri with diameters as small as 10.3 cm in natural trees and 15.8 cm in planted trees, although it represented less than 4% of total stem volume, and it was present in all trees above 35 cm diameter at breast height (dbh). Usable quantities of heartwood did not appear until breast height diameters reached 60 cm when mean heartwood volumes exceeded 0.5 m3 in the bottom 6-m log. Kauri with diameters in excess of 90 cm are predicted to contain a minimum of 1.9 m3 of heartwood per tree in the bottom 6-m log. Diameter increments in the best-performing planted stands in this study suggested that an average diameter over bark of 90 cm would be achieved in 120-150 years. Maximum sapwood width occurred at 40-50 cm dbh for planted kauri and 80-90 cm dbh in natural stands, and stabilised at an average value of 13 cm. In this study sapwood width was less than 15 cm in 93.7% of the kauri. The results of this study will facilitate the prediction of heartwood quantity in planted or second-growth stands using easily obtained diameter over bark measurements

The sand-binding plant Ammophila arenaria (L.) Link, commonly known as marram grass, has many characteristics which contribute to its usefulness in the initial stages of stabilisation of coastal sand and account for its selection for this purpose in most temperate countries. Research and experience in New Zealand have demonstrated the success of revegetation programmes based on its use in a planned vegetation succession, culminating in the permanent stabilisation of extensive sand drifts. Except on the seaward face of the current foredune, native sand-binding plants grow less vigorously than A. arenaria. For sites behind the foredune crest, use of A. arenaria offers the most reliable basis for prevention and arrest of sand mobility and the commencement of an effective vegetation succession involving native or exotic plants. Where rabbits have been eliminated from these sites and some degree of sand stability can be established and maintained, native sand dune species would be an appropriate choice for revegetation programmes. Use of A. arenaria as a temporary nurse for other dune species is currently under investigation

Armillaria root disease is widespread in plantations of Pinus radiata D. Don in many parts of New Zealand. Data from research trials in three central North Island forests were used to estimate the impact of infection, as assessed by the extent of root collar girdling, on stand growth approximately halfway through the rotation period, assuming a causative relationship between infection severity and growth. Volume loss was determined as a little over 2% at age 13 years, in a second-rotation stand with ca. 20% of trees infected prior to thinning, on a site not previously covered in indigenous forest. This value was calculated from the average stem volume for each of four infection severity classes and the relative numbers of trees in each class. Indices to compensate for competition were included in the model, based on stem diameters and local stocking densities influencing the growth of each tree. Mortality was not a significant factor in this stand. By contrast, mortality induced by Armillaria root disease was typically high in first-rotation stands planted on ex-indigenous forest sites, causing gaps in stocking resulting in substantially greater volume loss. In one such stand a comparison of mid-rotation crop tree volumes with early mortality indicated that a crop volume reduction of 21% had occurred. An additional loss of 4% in this stand was attributable to growth reduction in infected but still living trees. Although comparatively low, growth loss in infected trees may have significant economic impact if the disease becomes generally dispersed in second-rotation plantations

A remote sensing method is being developed to improve the cost effectiveness of assessing incidence and distribution of Armillaria root disease in young Pinus radiata D. Don plantations. Two multiple-tree row transects comprising 673 trees were located in a 4-year-old plantation of P. radiata (average height 3.5m) and the health of the crown foliage on each tree was assessed according to a four-point colour classification. Root collars on all trees were examined for infection by Armillaria species as indicated by resinosis, and sometimes by characteristic mycelial fans, accompanied by rhizomorphs. The study trees were then located on high resolution, colour, stereo aerial photographs and the individual canopies were re-assessed using the same colour classification. A total of 199 trees were identified as infected or colonised by Armillaria species (30% of all study trees). Of these, 131 bore healthy green crowns (66%) and 43 had died before attaining a height of 2 m (22%), and so they were not reliably detectable as diseased in colour photographs because of their green crowns or small image size. Of the other 25 diseased trees (13% of all diseased trees) which were taller than 2 m when assessed and either dead or with discoloured crowns, 24 (96%) were correctly classified on the 1:3,000 photographs and 19 (76%) on the 1:6,000 photographs. Most dead trees ?2 m tall were easily recognised at both scales of photography. The aerial photographs therefore enabled an accurate estimate of the incidence and spatial distribution of dying trees taller than 2 m that had Armillaria root disease. If a relationship can be established between such trees and the incidence of infection in a total stand, remote sensing is likely to prove a practical tool for assessing Armillaria root disease in P. radiata plantations

Levels of infestation by Armillaria species are less readily ascertained in many newer plantations of Pinus radiata D. Don in New Zealand, because early mortality is generally lower than when stands were planted directly on ex-indigenous forest sites. Two methods of assessing infection were therefore explored using data from a second-rotation trial for which infection was known for every tree. Theoretical transects of different sizes placed randomly in this stand determined that single-tree transects were the most efficient, requiring examination of the least number of root collars for the same level of accuracy. The numbers of trees were determined that should be sampled to achieve required levels of precision for different intensities of infection. Incidence and severity of infection at age 6 years were found to be greater nearer to trees killed early in the rotation period, suggesting an alternative assessment approach. Computer-generated contour maps of the incidence of killed and living infected trees were used to demonstrate that field maps of the distribution of visible, low-intensity, pre-thinning mortality may have potential for identifying sites with greater overall stand infection. This method could simplify the operational evaluation of stand infection, but more field work is required to establish its feasibility

Armillaria root disease persists in many seemingly healthy stands of Pinus radiata D. Don throughout New Zealand. Options for control or management remain limited, but one possibility may be to manipulate silviculture in a way that minimises chronic stand infection. A 3.1-ha trial in a second-rotation stand planted in 1985 was used to compare the effects of different thinning treatments approximately halfway through the rotation on a site not formerly covered in native forest. Treatments that created stumps at stand age 7 years appeared to increase the incidence of infection 5.5 years later (to 44% and 46% trees infected, for stocking levels of 500 and 250 stems/ha, respectively), compared with unthinned controls (30% infection, 810 stems/ha). Incidence was significantly greater (49%) among trees within 5 mofstumps derivedfrom trees infected by Armillaria species before thinning, than among trees with no neighbourhood stumps (29%). Not all stumps could be found after 5.5 years, but of the 76% that were, at least 58% were colonised by species of Armillaria. Incidence and severity of infection appeared unaffected by pruning to 30% of treeheight as measured at age 13 years. Infection was distributed unevenly across the trial area at ages 6 and 12.5 years. However, the over all spatial infection pattern did not alter over this period. Cultural pairing between isolates of Armillaria species made from 23% of all infected or colonised trees prior to thinning identified a minimum number of 68 genets (at least 22 per hectare). Further cultures isolated 5.5 years after thinning from trees or stumps in six treatment plots belonged to 41 genets (44 per hectare, minimum), of which at least 20% were present prior to treatment. These high colony densities strongly imply a role for basidiospores in the dissemination of Armillaria species, whether or not colonies were established in first-crop stumps or earlier. All colonies were of A. novae-zelandiae (Stevenson) Herink expect for one of A. limonea (Stevenson) Boesewinkel, which was present before thinning and subsequently re-isolated from the same location. Infection will be re-evaluated later in the rotation to determine the effect of larger stumps generated by a second thinning at age 13.5 years. In the meantime, caution appears warranted when thinning on heavily infested sites. Research is under way for an alternative control method for use in association with low-intensity thinning

The root systems of thirteen 13-year-old Pinus radiata D. Don trees with basal resinosis caused by Armillaria novae-zelandiae (Stevenson) Herink in Kaingaroa Forest were excavated and the boles were sectioned. The extent of infection in bark and wood of boles and roots was determined. Armillaria novae-zelandiae appeared to be able to survive and spread in bark, leaving the vascular cambium and the inner living phloem largely intact. Infected bark exhibited special characteristics including wide, light-coloured, resin-soaked layers of dead phloem separated by thick periderms often containing stone phellem, small mycelial fans, and layers of old resin. Penetrations to and death of sections of the cambium occurred on all but one of the trees, but the areas involved were small (mostly <40 mm in tangential and 100 to 150 mm in longitudinal extent). Most were located between major lateral roots, and in none did the dead area extend higher than 300 mm above the root collar. Penetrations generally occurred during the dormant season and a strong host reaction the following growing season resulted in well-developed calluses, which healed xylem penetrations within a few years. Invasion of the cambium by A. novae-zelandiae occurred at various dates from 1990 to the current growing season, and could not be simply related to stress caused by silvicultural operations or climatic conditions. Armillaria novae-zelandiae is likely to survive on infected trees to harvest, causing minor scarring and, in a few trees, resin soaking of xylem, but only at the very base of the tree. A small increment loss is also likely. After harvest and planting, the pathogen may again cause mortality in young plantations

As part of an investigation into the manner in which Armillaria root disease develops in Pinus radiata D. Don plantations in New Zealand, studies were undertaken to examine the ability of spores to colonise freshly cut bark-encased billets of wood in which moisture content was maintained by partial burial in soil. Billets were enclosed in plastic cylinders to protect them from colonisation by soil rhizomorphs, and in different studies were located as spore traps within second-growth indigenous podocarp/broadleaved forest, sometimes adjacent to fruitbody clusters of Armillaria species, or in the open. In two studies, natural airborne inoculum of Armillaria species was supplemented by treatment with spores or aqueous spore suspensions. After periods of 7 to 44 months, studies were terminated by exhuming billets and removing bark in order to determine which were colonised by species of Armillaria, as indicated by the presence of characteristic mycelial fans or rhizomorphs. Many billets of P. radiata (28% of 83 traps) and willow (Salix sp.; 83% of 12 traps with billets pre-treated using 2,4-D herbicide) were colonised by Armillaria species. Cultures isolated from three pine and six willow billets were identified as A. novae-zelandiae (Stevenson) Herink. In one study 65% of 20 pine traps were colonised instead by Phlebiopsis gigantea (Fries) Ju?hlich, a possible candidate as a biological control agent for Armillaria species in pine stumps. Three (9%) of 35 traps placed in the open became colonised by Armillaria species. If typical, this incidence may be sufficient to account for many of the colonies of these pathogens observed in pine plantations, and suggests that new infection centres may become established by means of airborne basidiospores. However, further work is under way to compare the receptiveness of billets and pine stumps to spore colonisation. In separate microscope studies, germination of basidiospores of two Armillaria species, followed by hyphal penetration of the xylem, was demonstrated on the surfaces of discs ofP. radiata and Beilschmiedia tawa (A. Cunn.) Kirk incubated in a moist environment.

Computer modelling of the behaviour of Pinus radiata D. Don wood during and after drying is dependent on the availability of data and this requires rapid and accurate measurement of wood samples. The testing of a prototype device designed to facilitate measurement of dimensions and distortion in standard sample sticks has indicated that crook and twist movement can be expected after the equilibration of moisture content

Altogether, 1862 Pinus radiata D.Don structural boards were tested in tension to examine five selected parameters of tensile strength, and to verify the code-assigned characteristic tensile strength values. Knot area ratio (KAR) alone does not predict tensile strength at the weakest point of the board; local bending stiffness at the failure point EP, fail shows a stronger correlation with tensile strength, and this presents a possibility for non-destructive estimation of tensile strength in P. radiata structural timber. The mean, 5th percentile, and characteristic strength in tension were studied for the test boards and the results indicated that the code values for current visual stress-grading systems of P. radiata need to be revised because the characteristic tensile strengths obtained from the tests were significantly lower than the values assigned in the standards

A 27-year-old Pinus radiata D. Don clonal trial was screened to identify clones with a wide range of branch size, internode length, and mature wood density (outer five rings at breast height). Four trees (ramets) from each of 10 such clones were subsequently chosen for use in wood property studies. Gravimetric wood property measurements were made on discs taken from each end of 5-m logs up to the merchantable limit (200 mm small-end diameter) and samples for densitometry obtained from specific positions in the stems. Structural lumber was recovered from four logs each of two ramets/clone, assessed for drying distortion, and graded mechanically. The discs were used to derive unextracted density for the juvenile (growth rings 1-10) and mature (rings 11 +) portions of the stems at all levels, and samples for X-ray densitometry at 5 m and 20 m. For each annual ring, a number of wood density variables were recorded, including earlywood and latewood measures. Within each of the 10 clones, the overall patterns of density variation were quite consistent from pith to bark and between stem levels, and demonstrated distinct clonal differences. Indicative estimates of heritabilities, age-age correlations, and predictability of final average tree and log densities were derived using linear mixed-effects models from the densitometric data. Both average wood density and intra-ring density component patterns were highly consistent within clones, and heritability estimates were high for most properties assessed. Latewood characteristics were particularly strongly related to growth ring density. For P. radiata, juvenile wood is often conveniently described as the first 10 growth rings from the pith. On this basis, clonal means for juvenile wood volume varied from 39% to 61%. The actual wood properties for this zone, in terms of wood density, differed markedly between clones. Using an alternative technical definition for juvenile wood (proportion of the stem with annual average basic density < 400 kg/m3), the percentage of juvenile wood varied from 15 to 64%, or from 5 to 13 growth rings. This suggests that clonal selection may be an effective way of controlling the impact of juvenile wood properties.

In a preliminary study, the cargo, packaging, the inside, and the outside of 102 air cargo containers were inspected for the presence of contaminants. Most of the contaminants were lying loose inside the containers, and the packaging and cargo were almost wholly free of contaminants. Based on these findings, a larger sample of 991 containers was examined, with attention focused on contaminants in and on the container, without recording the details of the cargo carried within but inspecting the wooden packing. The sample was randomly selected from containers landed at Auckland, Wellington, and Christchurch airports in the period from April to December 1999. The containers were examined in the unpacking sheds at the airports and all soil, plant, animal, and inorganic contaminants found on the outside of and inside the container during and after unpacking were collected. Isolations were made for fungi from all soil samples (from 51 containers) collected. All plant material was examined microscopically for fungi. Insects, spiders, and other invertebrates were collected. A container was classified as potentially quarantinable" if any of the contaminants found in or on it included viable pests or viable fungi belonging to genera which include plant pathogens, as the presence of such organisms indicated the potential risk posed by this particular pathway. Of the 991 containers examined, 750 (75.7%) carried no contaminants, 110 (11.1%) carried only non-quarantinable contaminants, and 131 (13.2%) carried potentially quarantinable contaminants. The quarantinable contamination rate of containers originating from different parts of the world varied from region to region; it was 18.2% for Australia, 16.4% for Europe, 9.4% for North Asia, 7.8% for North America, 5.9% for South-east Asia, and 5.1% for the Pacific. There were few regional differences in the proportion of quarantinable contaminants to the total number of contaminants. Very few contaminants (3.3%), none quarantinable, were associated with packaging. The quarantinable contamination rate varied from 0.0% for baggage containers to 19.6% for open-sided containers. Foliage, twigs, fruit, seed, and woody material made up 62% of all sources of contaminants and soil was the next most common (23%) source. Most of the contaminants were found inside the containers; only 0.8% of the contaminants were found solely on the outside. The finding that fresh plant material carrying plant pests and pathogens is common inside air cargo containers suggests that these containers are a pathway by which such harmful organisms could enter New Zealand. The fact that pest and pathogen incursions similar to those found during the air container survey have been detected in the vicinity of airport cargo sheds provides a strong link to this pathway. The training of facility operators working with the air cargo containers in managing the biosecurity risks from this pathway and the monitoring of their performance are important. The removal of residues and the safe disposal of all material collected from empty containers as soon as possible after the containers are unpacked is essential to minimise the risk of introduction of undersirable organisms"

Although fine roots are important in mineral nutrient and carbon cycling, they have often been ignored in forest ecosystems. The work reported here was carried out to study the growth and spatial distribution of roots of coppicing eucalypts. The study used a 7-year-old eucalypt stand planted at 3 x 2-m spacing at the Patago?nia farm in the Savanna region of Joa?o Pinheiro county in north-east Minas Gerais State, Brazil. The eucalypt plantation belonged to the V & M Forest Company. All the trees in the area were felled, and root biomass was determined in trees whose diameter at breast height corresponded to the population mean. Measurements were carried out at 0, 60, 120, 180, 240, and 330 days after harvesting in order to evaluate the distribution of root biomass. Sampling was performed in 50 x 50-cm units on the beds and at the 0-10, 10-20, 20-40, and 40-60 cm depths, distributed in half the area occupied by the selected trees, and the roots were separated into three diameter classes (fine roots < 1 mm, medium roots 1-3 mm, and coarse roots > 3 mm). Fine- and medium-sized root biomass increased with time after harvesting, particularly to a depth of 20 cm. However, there was little alteration in the biomass of coarse roots. Root biomass decreased with depth and, on average, about 73%, 54%, and 68% of the fine, medium, and coarse roots, respectively, were concentrated in the surface 20 cm of soil. Analyses of the horizontal root distribution indicated that, in general, most of the roots, mainly medium and coarse, were located close to the stumps and that the root distribution was less uniform as root diameter increased

Initiation and early development of microsporangiate strobili (male cones) in Pinus radiata D. Don were studied during two growing seasons (1992/93 and 1996/97) at the Amberley Seed Orchard (43Ý10?S) in the South Island of New Zealand. Further collections were made at the Seddon Seed Orchard (41Ý42?S) and in a breeding archive located at Rotorua (30Ý10?S). Long-shoot buds from putative pollen-producing branches were collected from December to March from different clones and different-aged ramets. On one clone male cones were found on bicyclic shoots. In both years axillary buds, which would develop into male cones, had formed by mid-December. Generally, development was earlier on older ramets, but it also varied between clones, years, and locations. Development was earlier in 1992/93 than in 1996/97. In 1996/97, material collected in the Seddon Seed Orchard was about 1 month ahead of material from the Amberley Seed Orchard. Primordia developing into male cones appeared to start swelling very early. These shoots had far fewer basal cataphylls than did short shoots. Anatomically, male-cone, vegetative long-shoot, and female-cone primordia all go through a similar intermediate stage of development whence their ultimate fate might be altered by environmental factors or hormone applications.