A seed-source trial was established in 1996 using open-pollinated seed of Douglas-fir (Pseudotsuga menziesii var. menziesii [Mirb.] Franco). The trial involved seed from 11 native populations in coastal California, three Californian seed orchards, one Oregon seed orchard and two Washington seed orchards. Open-pollinated progeny seedlots were bulked by native population. Seed from seven New Zealand seed stands of Washington, Oregon or Californian origin was also included. This trial was conducted at five sites with a range of climatic features.

The purpose of this trial was to test the performance of bulked Californian seedlots on four cold, high-elevation sites prone to snow and exposed to strong winds compared to a mild, low-elevation site. Overall, there were significant differences in height growth of trees from different seedlots but little difference was observed in the growth performance of trees from individual Californian provenances of varying latitude at the four trial sites with harsh climates. Also, the three New Zealand seedlots of California origin grew at about the same rate as the three seed-orchard seedlots of southern Oregon and Washington origin. At the four harsh-climate sites, the severe effect of environment on stem straightness, malformation and acceptability appear to have largely obliterated any provenance differences at the assessment age (eight – nine years).

At the sheltered site, trees from northern Californian seedlots clearly grew best and trees from the New Zealand seedlots originating from California substantially outgrew trees from the New Zealand seedlots of Oregon and Washington origin. This result paralleled that at three progeny-trial sites featuring many of the same seed sources, which all had mild climates. The southernmost provenance from Los Padres at latitude 35° 49´ behaved differently from all other provenances. It had slower growth, but better form that may have resulted from shelter from faster-growing trees.

Douglas-fir (Pseudotsuga menziesii var. menziesii [Mirb.] Franco) is a native of the Pacific North-west of the United States of America. In New Zealand, Douglas-fir is the prominent alternative species to radiata pine (Pinus radiata D.Don) and is planted widely. Original provenances were obtained from Washington USA, but after the introduction of Swiss needle cast (Phaeocryptopus guaemannii [T.Rohde] Petr.), provenances needed to be re-evaluated. Trials planted in 1957 and 1959 demonstrated that the best provenances for New Zealand sites were from coastal locations in California, USA and Oregon, USA. While selections were made from these provenance trials, the need for more genotypes was also identified. Seed, comprising of 222 open-pollinated families from 19 coastal and one inland provenance from north-western USA, was collected in 1993.

This seed was planted in 1996 along with eight New Zealand land-race seedlots as controls. The latitude of the provenances ranged from 35° 07´N in California to latitude 44° 10´N in Oregon. The trial layout was a sets-in-replicate design of seven sets of progenies with 30 replicates of single-tree plots, planted on three sites

Totara (Podocarpus totara D.Don) is an important indigenous timber tree in New Zealand. Old-growth stands have been depleted, but totara regeneration is a prominent feature of the rural pastoral landscape in many parts of the country. These young stands have not been well described or quantified. A method for determining their age is essential if their development and timber potential is to be understood. The reliability of a commonly used procedure for estimating tree age from growth-ring counts was investigated by sampling in totara stands of known age. Increment cores were taken from 11 planted stands ranging in age from 9 to 91 years. Of the 147 cores sampled, 122 (83%) were considered to be suitable for ring counting with the aid of a stereo-microscope. Estimates of age based on latewood band counts showed least bias in stands with diameter growth rates of approximately 5 mm yr^-1. There was a tendency towards underestimation in slower-growing stands because latewood bands were not formed every year. The age of seven well-stocked stands with mean diameter increment ranging between 2.8 and 4.8 mm yr^-1 was underestimated by 7 – 19%. Cores obtained from faster-growing stands did not always exhibit distinct latewood bands and sometimes formed more than one growth ring per year. This led to overestimation of age. In three open-growing stands with diameter increment averaging 7.1 mm yr^-1, age was overestimated by an average of 10%. The latewood-band counting method is likely to underestimate age in regenerating totara stands where intense competition and slow growth rates can be expected. Avoidance of sampling suppressed trees and rejection of core samples with indistinct latewood bands should decrease error and allow estimates of stand age to lie within 10 – 15% of actual age.

Trees are complex living organisms that are continuously laying down new cells with a structure appropriate for survival (including mechanical reliability) in their current environment. Trees need to adapt to new environmental conditions following silviculture operations and one adaptive response is alteration of their wood properties. This study examines how wood density and microfibril angle in individual radiata pine (Pinus radiata D.Don) trees changed in response to seven different pruning and thinning regimes.

Pith-to-bark wood samples were taken from just above stump height (approx. 0.3 m above the base of the tree) in 28-year-old trees that were part of a silvicultural experiment. Ring-average values of wood density and microfibril angle were obtained using SilviScan-3.

A comparison of these radial profiles indicated that severe pruning, together with moderate to heavy thinning resulted in an increase in microfibril angle around the time of thinning which interrupted the commonly observed downward trend in microfibril angle with increasing ring number from the pith. Most trees showed an increase in ring-average wood density in the year following the final thinning, however this may be related to environmental conditions rather than thinning as the increase also occurred in the untreated control.

Near the end of the rotation, average microfibril angle and average wood density for rings 21 to 26 were not influenced by the historic silvicultural treatments.

These results support our understanding that the developmental history of a tree must be taken into account when estimating the overall wood structure and properties and that tree DBH alone is insufficient.

Kauri (Agathis australis (D.Don.) Lindl.) is a coniferous forest species endemic to New Zealand. This unique resource is currently under threat from Phytophthora taxon Agathis infection that causes kauri die-back. This situation highlights the need not only for reliable clonal propagation methodologies to amplify genotypes exhibiting disease resistance but also the development of protocols that could be used for mature material for ex situ conservation of important genotypes. Understanding the viability of previously stored seed for culture initiations is also critical if trees subsequently die.

Six agar-based culture media were compared for their effects on shoot production from in vitro germinated mature zygotic kauri embryos. All root and some hypocotyl tissue was removed from the germinated embryos prior to initiation onto the culture medium. The number of shoots produced was highest (10 per embryo) in a full-strength, modified Quiorin and Lepoivre medium containing 3.5 g L^-1 activated charcoal.

Four treatments incorporating use of rooting hormones were compared for their effects on root development from shoots produced in vitro. The dipping of stem ends in rooting powder containing talc plus 2% indole-3-butyric acid before transfer to potting mix stimulated root development in 68% of the shoots. Only 5 – 14% of shoots maintained in agar-based cultures with added rooting hormones before transfer to potting mix produced roots. Rooted plants continued to grow vigorously when transferred to standard nursery conditions. Variability among seed sources was high and there was no evidence that genotype influenced in vitro production of either shoots or roots.

“Risk intelligence” is a necessary adjunct to management and planning in land and water resource protection and productivity. Following a brief review of how risk is quantified, packaged and sold in the financial sector, risk management in land and water resource management in New Zealand is assessed in the light of the 2007–2009 global financial meltdown. Conventional planning approaches for scenario planning do not accommodate the wide complexities of risk management because of their inability to adequately quantify and forecast risk for value creation. A case study using a farm in Rotorua, New Zealand, is employed to explore how farming risks in terms of nitrate leaching, phosphorus loss, sedimentation and biodiversity loss, may be minimised in exchange for value creation. Historical variance and trends of simulated variables is estimated through random sampling for future trends using Beta, Triangular and Two-sided Power distributions and Monte Carlo simulations. Despite all three statistical distributions performing relatively well, the choice of the most appropriate one will ultimately be determined by expert judgement.

Data from a nationwide set of permanent sample plots and interpolated climate and nutrition surfaces were used to develop multiple regression models describing Sequoia sempervirens (D.Don) mean top height at age forty (referred to as Site Index) and volume mean annual increment at age forty for a stocking of 400 stems ha-1 (referred to as 400 Index). The final Site Index model explained 82% of the variance in the data using mean annual daily air temperature and mean summer vapour pressure deficit, with the variables accounting for 71 and 11% of the variance, respectively. The final 400 Index model accounted for 76% of the variance in the data. Independent model variables for the 400 Index include mean spring air temperature, subsoil acid soluble phosphorus, and mean summer vapour pressure deficit, with these variables respectively accounting for 55, 16 and 5% of the variance in the data. A one-at-a-time validation procedure indicated both final models were relatively unbiased and accurate.

For Site Index, partial response curves show a positive linear relationship with mean annual daily air temperature and a downward facing parabolic relationship with summer vapour pressure deficit that reached an optimum Site Index at 0.53 kPa. Partial response curves show a positive relationship between 400 Index and mean spring air temperature, and acid soluble phosphorus, and a negative linear relationship with summer vapour pressure deficit. Maps illustrating the spatial variation in 400 Index and Site Index for S. sempervirens across New Zealand are provided.

Computer-based decision support systems are an important tool for ensuring good practice in the aerial application of pesticides. By providing access to herbicide-plant dose-response curves through such systems, users are able to predict the effects of herbicides on sensitive plants outside of the spray area and weeds within the target area. However, difficulties arise when the operational scenario requires input of a herbicide/plant combination that is not found within the dose-response database. The purpose of the study described here was to evaluate whether the existing dose-response database could be used to derive generic dose-response curves for use in agriculture, horticulture or forestry.

Each dose-response curve was characterised using a term labelled “s”, the ratio of herbicide doses required to give a 5% and a 95% yield reduction of the test plant species, and an index dose (in this case the dose giving a 50% yield reduction). The value of s differed considerably between trials with a range from 4.8 to 2800 (e.g. the upper extreme indicates that the dose required to give 5% yield reduction must be increased by a factor of 2800 for a 95% yield reduction).

Three generic dose-response curves were derived based on the mean value and the upper and lower extremes of s. With this information plus an estimate of the index dose (i.e. the dose required to reduce plant yield to 50% or some other selected value) an average, or extreme dose-response curve can be generated.

Cyclaneusma needle cast is a very damaging foliar disease of Pinus species. It is particularly widespread and detrimental to the growth of planted forests in New Zealand. The influence of climate change on the spatial distribution of disease severity (S_sev) would provide forest managers with insight into where to deploy disease resistance planting stock in the future. Here we use an existing model of Cyclaneusma needle cast severity, developed from an extensive dataset, to spatially predict disease severity under current and future climate to 2040 and 2090.

Spatial predictions of S_sev under current climate varied widely throughout New Zealand. Values of S_sev were highest in moderately warm, wet and humid high elevation environments located in the central North Island. In contrast, relatively low values of S_sev were predicted in drier eastern and southern regions of New Zealand.

Projections within the North Island show relatively little change in S_sev from current climate over the short term and low to moderate reductions in S_sev within most areas over the long term. In contrast, within the South Island, S_sev was predicted to markedly increase over both projection periods, with more pronounced increases in S_sev projected by 2090. Surfaces presented here are a critical element of decision support systems that describe how climate change is likely to influence plantation productivity and vulnerability to abiotic and biotic risk factors.

Terrestrial laser scanning (TLS) technologies are now being used to provide detailed forest inventory information at the tree or plot scale. A major problem to overcome when using TLS is occlusion by surrounding trees, lower branches and understorey. An occlusion adjustment model was evaluated in 24 stands in Oregon, USA and Australia. The model can be used to predict stand-level tree count densities with minimal errors, especially if an appropriate plot radius is selected. The optimal plot radius may be dependent on the stand-type in which the TLS is being undertaken. Other approaches are likely to be more appropriate if accurate stem counts are required at the individual plot level.

Pseudotsuga menziesii (Mirbel) Franco (Douglas-fir) is a preferred timber species for sites in the South Island of New Zealand above about 500 m elevation where precipitation exceeds 800 mm. However young stands often exhibit leader tip death, leader break and stem distortion that result in poor tree form. These symptoms may be caused by chemical factors such as boron deficiency or by environmental factors such as frost, snow, wind, insects or disease. Trials were installed in two young Douglas-fir stands at Lake Hill in the Rakaia Valley (570 m elevation) and Balmoral Station in the Mackenzie Basin (870 m elevation) to determine if tree form could be improved by applying boron in the form of hydrated sodium calcium borate hydroxide (ulexite). Stem malformation (stem forking or multi-leadering) was 48% and 88% at Lake Hill and Balmoral respectively. Stem malformation was not reduced by boron application at Balmoral, and only reduced at Lake Hill by the highest boron application rate (32 kg B ha^-1), which also greatly reduced tree stem volume growth. These results indicate that stem malformation was not due to boron deficiency. Boron applied at 4 kg ha^-1 significantly increased tree stem growth at Lake Hill, but reduced stem growth at Balmoral, and higher rates reduced stem growth at both sites. Boron rates of 1 - 2 kg ha^-1 may be more appropriate for young Douglas-fir. This study indicates that a foliar boron concentration of 12 mg kg^-1 may be adequate for good form of Douglas-fir, but may not be adequate for optimum growth. It is concluded that further studies are needed to better understand the link between climate and Douglas-fir tree form in montane environments, as well as to better understand relationships between rate of boron application, foliar boron concentration and growth and form of Douglas-fir in different environments.

Resin pockets are a significant source of wood-quality degrade in Pinus radiata D.Don (radiata pine) logs for many forests around New Zealand. Low rainfall and windy conditions coupled with stony soils and poor soil water-holding capacity have been implicated in their occurrence. Resinous defects have little impact on the structural properties of timber and consequently their occurrence may be underestimated in plantations grown for purposes other than appearance grade timber. It has been common in industry trials to describe resin pockets according to three distinct categories; Types 1, 2 and 3. This paper builds on a series of existing studies directed at understanding the physiological causes of resin pocket occurrence and suggests that current distinctions drawn between different types of resin pockets represent a developmental continuum. Specifically, it argues that a morphological continuum exists between what is generally categorised as Type 1, and Types 2 and 3 resin pockets. It is not intended that the proposed gradation replace the current classification of resin pockets. Rather it is suggested that there is a common physiological cause to better focus research directed at managing their occurrence.

The Land-Use, Land-Use Change and Forestry (LULUCF) sector of the Kyoto Protocol requires New Zealand to monitor changes in the country’s carbon stocks, including those within indigenous forests. Podocarpus cunninghamii Colenso was, in pre-human times, a dominant tree species within the forests of the South Island high country. Anthropogenic disturbance, primarily fire, has converted most of these forests to grassland. Despite this mass deforestation, remnant stands of Podocarpus cunninghamii still exist, and may represent important point sinks of carbon. This study provides first estimates of existing above- and below-ground carbon stocks in high country stands of Podocarpus cunninghamii and presents a preliminary model of Podocarpus cunninghamii carbon sequestration rate. Carbon stocks within high country stands of Podocarpus cunninghamii range from 7.3 t ha^-1 in the drylands to 130.1 t ha^-1 in the wetter areas. Estimates based on tree ring widths indicate a high country-wide Podocarpus cunninghamii carbon sequestration rate of 0.1 - 0.5 t ha^-1 yr^-1 for 250 - 1000 stems ha^-1, respectively.

A tree volume equation for New Zealand’s Nothofagus species was validated using sectional measurements from Nothofagus menziesii (Hook f.) Oerst trees at Alton Valley in Southland. Upper stem diameters of 60 trees were measured using a Tele-Relaskop within three plots that had been thinned to stockings of 150, 1500 and 8228 stems ha^-1. Bark thickness measurements were obtained from a range of heights up tree stems. Merchantable under-bark volumes of stems calculated from sectional measurement were compared with merchantable volumes predicted by a model developed by Ellis (1979). The model was found to predict volumes with minimal bias even though errors were kurtotic and a little outside the range predicted. The correlation between predicted and estimated volumes was 0.97.

It can be difficult to provide large numbers of fresh forest insects for use in fumigation experiments. This paper reports on studies aimed at providing large numbers of fresh burnt pine longhorn beetles (Arhopalus ferus Mulsant) (Coleoptera: Cerambycidae). Burnt pine longhorn beetle is one of the most important pests of quarantine concern associated with export of New Zealand radiata (Pinus radiata D.Don) logs, particularly to India, China and other Asian countries.

The timing of the collection of hundreds of adults, obtaining eggs from them and then storing the eggs for use in fumigation trials must be synchronised with the timing of the trial itself. Three separate experiments were conducted to improve rearing and storage of burnt pine longhorn beetles. In the first experiment, burnt pine longhorn beetle eggs were laid by field-collected adults. Larvae were successfully reared on an artificial (huhu) diet at 20 °C ± 1.5 °C and pupated after 216 days (males) and 227 days (females). Adults emerged after a further 2.5 weeks. These adults stayed alive in the laboratory for up to 46 days at 20 °C. In a second experiment field-collected burnt pine longhorn beetle adults of mixed age were stored at 4, 6, 8, 10, 12 and 14 °C for up to 10 weeks. The longest survival time was at 6 °C (50% still alive after 28 days). In a third study we found 12 - 14 °C was the best temperature for storing eggs. The total life times, times to pupation and adult survival times were not significantly different between males and females. Males tended to be more variable than females particularly in the time spent as larvae and as adults.

Spatial maps of density and stiffness patterns within individual trees were developed using two methods: (1) measured wood properties of veneer sheets; and (2) mixed effects models, to test the hypothesis that within-tree patterns could be predicted from easily measurable tree variables (height, taper, breast-height diameter, and acoustic velocity).

Sample trees comprised an assortment of 25 coastal Pacific Northwest Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) trees grown on three sites under a range of thinning regimes. At harvest, trees were 36 years old on one site, and 45 and 51 on the other two. After felling and crosscutting, bolts were peeled into veneer, labelled, dried, weighed, measured, and non-destructively tested using a Metriguard 2600^TM veneer tester. The labels allowed each sheet to be tracked back to the peeler bolt and consequently to the position within the parent tree from which it came, and the measurements allowed calculation of veneer density and, after application of the fundamental equation for propagation of sound, veneer stiffness.

Maps of each parent tree created from the veneer data clearly demonstrated regions of higher density and stiffness. Furthermore, within each tree, density was approximately normally distributed, while stiffness tended to have moderate negative skew. Maps developed using mixed effects models showed very good correspondence between measured and predicted patterns, particularly for density. Despite differences in age, site and silviculture, results from this study suggest that it is possible to predict within-tree wood properties using easily measurable tree variables.