Corrigendum to NZJFS 25(3) 318-327: Development of a composite taper equation to predict over- and under-bark diameter and volume of Eucalyptus saligna in New Zealand.

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

Work study of pruning of plantation Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) in Kaingaroa Forest showed that production rates in first and second lifts (to 2 m and 3.8 m) were approximately 20 and 30 trees per productive hour, and varied with walk distance, hindrance rating, number of branches removed, and tree diameter. The physiological workloads of both lifts were classified as "heavy".

Three nozzle types were compared in a trial investigating the potential for herbicide drift during aerial spraying. Water containing a colorimetric tracer, a fluorimetric tracer, and a foaming agent was sprayed from a helicopter boom along a single flight line. Deposition on steel plates located on the ground was measured to a distance of 300 m downwind. The airborne flux was measured to 225 m downwind, using "Rotorod" samplers. Conventional D8-45 nozzles were found to have greater drift potential than foaming nozzles (both pointing straight down). Lowest airborne drift occurred during use of D8 nozzles, pointing straight back. Deposition measurements using steel plates on the ground showed that peak deposition was greatest and closest to the flight line with D8 nozzles, followed by foaming and D8-45 nozzles. Steel plates on the ground were found to be unsuitable for estimation of airborne flux.

In a trial designed to quantify the reduction of Pinus radiata D.Don growth caused by a range of weed species, tree seedlings were grown weed-free or with herbaceous broadleaves (a volunteer mixture of species from which grasses were excluded), Cytisus scoparius L. (broom), Ulex europaeus L. (gorse), Buddleja davidii Franchet (buddleia), Holcus lanatus L. (Yorkshire fog) plus Lolium multiflorum L. (Italian ryegrass), Lotus uliginosus Schk. (lotus), or Cortaderia selloana (Schult) Asch. et Graeb. (pampas). Water and nutrient levels were varied by factorial irrigation and fertiliser treatment. After 3 years tree stem volume was greatest in weed-free, lotus, gorse, and grass treatments and least with buddleia and pampas. The effect of herbaceous broadleaves and broom was intermediate. It was concluded that tall, fast-growing, weed species reduced P. radiata growth by restricting light availability to tree crowns. No convincing evidence was found to link the large growth losses with interference in water or nutrient supply.

An objective index describing the radial "evenness" of lateral tree roots has been developed. The index may be applied on the basis of root length or mass or image analysis of root quantity. Unevenness, defined as deviation from a perfectly uniform root distribution, is weighted according to the scale at which it occurs, with unevenness between halves or quarters of the root system receiving more weight than unevenness between sixteenths or thirty-seconds. A sliding frame of reference eliminates the problem of rotational dependence in the initial positioning of the "grid" used to count roots. For practical application the index requires a computer-readable root map, either from a field root-mapping exercise, or from digitised pictures of the exposed root system. Agreement between evenness rankings from visual inspection and from the index is generally excellent.

Relationships between cone penetration resistance (PR), soil moisture (SM), and bulk density (BD) were derived for: (i) cultivated (ripped) and uncultivated, hardsetting and non-hardsetting, field soils; and (ii) repacked cores of the uncultivated soils. Each of the soils supports commercial Pinus plantations in the coastal lowlands of south-east Queensland, Australia.

Penetration resistance was positively correlated with bulk density and negatively correlated with soil moisture for all soils. In the uncultivated soils, penetration resistance was less sensitive to bulk density than typically reported in the literature, or than observed in the cultivated soils where a wider range of bulk density values was studied. In both the cultivated and the repacked soils, penetration resistance was more sensitive to soil moisture at higher bulk density, and more sensitive to bulk density at lower soil moisture.

It was not possible to fit the same models to uncultivated, repacked, and cultivated soils, and therefore not possible to compare relationships for each statistically. Relationships between penetration resistance, bulk density, and soil moisture were best described by additive models in the uncultivated soils and multiplicative models in the cultivated soils. For the repacked soils, models had to be developed relating penetration resistance to bulk density for each soil moisture class separately.

The study demonstrated that: (i) relationships between penetration resistance, bulk density, and soil moisture were insufficiently sensitive to predict responses in the penetration resistance of field soils tochanges in soil moisture, as might occur temporally, or bulk density, as might occur with compaction or reconsolidation after cultivation; and (ii) repacked soils could not be used to simulate the relationships between penetration resistance, bulk density, and soil moisture for cultivated field soils. Therefore, penetration resistances measured at different times in studies in which either bulk density or soil moisture are expected to change cannot be easily compared. In these situations, which include compaction and consolidation studies, both penetration resistance and bulk density, or bulk density alone, should be used to monitor change.

Relationships between penetration resistance, soil moisture, and bulk density, together with moisture characteristic drying curves for individual soils, were used to define relationships between penetration resistance and matric suction. These relationships define a soil characteristic that may be useful for: (i) explaining varying responses of different soils to drying; (ii) explaining various Pinus seedling growth responses to cultivation and compaction; and (iii) delineating soils which are functionally hardsetting upon drying.

Seeding trials with Pinus species and Pseudotsuga menziesii (Mirb.) Franco were undertaken to investigate known constraints to low-cost afforestation of depleted short grasslands of the South Island high country.

In field trials mycorrhizal development of Ps. menziesii seedlings was enhanced by drilling seed with soil collected under an existing stand. Seed inoculation with mycorrhizal fungal spores appeared to be more successful than soil inoculation in promoting pine mycotrophy. Seed inoculation was less effective for Pinus nigra Arn. than for P. radiata D.Don, possibly because the method used resulted in attachment of more spores to the larger P. radiata seed. Inoculation of pine seed with spores of Rhizopogon rubescens Tul. was more effective than inoculation with R. luteolus Fr. or Suillus luteus (L. ex Fr.) S.F.Gray.

Exclosure trials at two Mackenzie Basin sites showed that browsing of seedlings was due to rabbits. Mechanical removal of residual herbaceous species resulted in reduced tree seedling numbers as rabbits were attracted to the bare soil or to seedlings exposed by turf removal. Control of rabbits for several years may be necessary to allow successful establishment by seeding. In the absence of rabbits, turf removal may improve seedling establishment in drier areas. Fertiliser application decreased tree seedling numbers through increased competition from the herbaceous sward.

Establishment of pines by seeding on the drought-prone outwash gravel soils of the Mackenzie Basin is not reliable. On less drought-prone soils forest establishment by seeding is possible if rabbits are controlled. Further experimentation with mycorrhizal inoculation is required.

An establishment trial was planted on a dry-land hill forest site in the Okuku area of Ashley Forest, Canterbury, to compare the effects of seedling quality, handling, cultivation (ripping), and post-planting weed control on subsequent survival and growth of Pinus radiata D.Don up to 13 years after planting, i.e., at half-rotation. Weed control was the most significant factor, followed by seedling quality and cultivation (ripping). Lack of weed control, use of conventional seedling quality, and conventional handling resulted in a 43% loss of volume growth compared with the best treatment combination of weed control, "best practice" seedling quality, and "best practice" handling. Ripping was less important, but gave a 7% increase in volume.

A Corrigendum to this paper is available here: Corrigendum for Balneaves, J. M. et al. 1996: Establishment practices can improve longer-term growth of Pinus radiata on a dry-land hill forest”

A site which had previously carried a high quality Eucalyptus regnans F.Mueller forest, but which had marginal soil nutrient reserves, was selected to trial four fertiliser options for establishment of Eucalyptus nitens (Dean et Maiden) Maiden and E. regnans sawlog plantations. Apart from the fertiliser options, current operational plantation establishment practices were used.

Application of a nitrogen-phosphorus fertiliser shortly after planting produced increased growth for E. nitens amounting to 30 m³/ha by age 7 years. Eucalyptus regnans also responded to fertiliser application, but overall growth was less than half that of the E. nitens. This slower growth was due to early severe and repeated insect attacks. Insect attack also caused increased double and multiple leadering in the E. regnans.

Normal rates of nitrogen-phosphorus fertiliser application (25 g N and 11 g P/tree) resulted in an increase in branch size compared with trees without fertiliser, and the mean largest branch diameter was increased by 6 mm for E. nitens. Multiple leadering also increased, from 2% to 10%. Very high rates of fertiliser addition over the first 6 months after planting caused substantially greater deterioration in branching habit in both species; the mean largest branch diameter was 10 mm larger than with normal fertiliser additions, and 27% of trees had multiple leaders. While most stems in stands treated with normal rates of fertiliser were suitable for pruning for sawlog production, very few stems in stands treated with high rates of fertiliser were suitable.

Despite substantial increased growth from fertiliser application, at age 7 years foliar-phosphorus levels were low and foliar-nitrogen levels were marginal in all treatments. Increased growth rates could be expected from further application of fertiliser. Results indicated that soil nutrient profiles together with foliar analysis might be useful in developing fertiliser regimes for eucalypt plantations.

A site which had previously carried a high quality Eucalyptus regnans F.Mueller forest, but which had marginal soil nutrient reserves, was selected to trial four fertiliser options for establishment of Eucalyptus nitens (Dean et Maiden) Maiden and E. regnans sawlog plantations. Apart from the fertiliser options, current operational plantation establishment practices were used.

Application of a nitrogen-phosphorus fertiliser shortly after planting produced increased growth for E. nitens amounting to 30 m³/ha by age 7 years. Eucalyptus regnans also responded to fertiliser application, but overall growth was less than half that of the E. nitens. This slower growth was due to early severe and repeated insect attacks. Insect attack also caused increased double and multiple leadering in the E. regnans.

Normal rates of nitrogen-phosphorus fertiliser application (25 g N and 11 g P/tree) resulted in an increase in branch size compared with trees without fertiliser, and the mean largest branch diameter was increased by 6 mm for E. nitens. Multiple leadering also increased, from 2% to 10%. Very high rates of fertiliser addition over the first 6 months after planting caused substantially greater deterioration in branching habit in both species; the mean largest branch diameter was 10 mm larger than with normal fertiliser additions, and 27% of trees had multiple leaders. While most stems in stands treated with normal rates of fertiliser were suitable for pruning for sawlog production, very few stems in stands treated with high rates of fertiliser were suitable.

Despite substantial increased growth from fertiliser application, at age 7 years foliar-phosphorus levels were low and foliar-nitrogen levels were marginal in all treatments. Increased growth rates could be expected from further application of fertiliser. Results indicated that soil nutrient profiles together with foliar analysis might be useful in developing fertiliser regimes for eucalypt plantations.

Timing of GA_4/7 application was investigated for the 1993 year at the Amberley and Waikuku seed orchards in Canterbury. February was confirmed as being a more favourable month for application than March. The response "window" was considerably wider when GA_4/7 was applied by stem injection than when it was applied externally to buds. At the dosages used, stem injection was also confirmed as being a much more effective way to increase female strobilus yield than bud application. The optimal application time for different clones varied little: in the three clones studied the mean range was only 7 days. Also, response to the timing of application did not differ appreciably between ramets of different size and age.

The area of potentially productive land occupied by skid sites (logging landings) created during logging operations was measured at three sites in New Zealand-Kaingaroa Forest (Bay of Plenty), Golden Downs Forest (Nelson), and Berwick Forest (Otago). Several skid sites in each forest were subjected to rehabilitation treatments, either ripping or ripping plus return and spread of topsoil and woody debris. These treatments markedly improved soil physical properties related to tree growth. When topsoil was returned the soil nutrient status also improved. Rehabilitation treatments improved the growth of young trees on the landings.

Plantation forestry on the widespread Imperata cylindrica (L.) Beauv. secondary grasslands in Indonesia is technically feasible and financially profitable, provided that fast early growth and high yield per hectare are achieved. Fast early growth and canopy closure ensure rapid suppression of the grass, thereby reducing competition and fire susceptibility. Considerable differences in growth between species and among provenances indicate the need for careful species and provenance selection and species/site matching. Production of high-quality seedlings is a prerequisite for survival over the critical early phase when competition by grass is fierce. Total cultivation is better than strip cultivation, and a stand density of at least 1100 seedlings/ha is required. Plantation establishment costs have an insignificant effect on the financial profitability of the scheme, but a high yield per hectare has a highly significant effect. Plantation establishment has considerable environmental effects which make it profitable from a wider economical perspective. These include sequestration of atmospheric carbon and maintenance of biological diversity.

Three herbicides (glyphosate, hexazinone, and metsulfuron methyl) at three rates were tested on indigenous herbaceous shrubs (macchia) in a 1-year-old virgin Pinus radiata D.Don stand. Mechanical ring weeding of 1 m radius was included as a control treatment. Herbicides were applied over-the-top to uncovered P. radiata. These treatments were tested for each of four seasonal applications. Tree height and diameter growth were measured to determine efficacy of the treatments. Two floristic surveys indicated satisfactory weed suppression.

Summer was the best season to apply herbicides. Glyphosate and hexazinone were significantly better than metsulfuron methyl. Hexazinone had no detrimental effect on P. radiata and improved height growth, whereas glyphosate scorched the trees but improved their diameter growth. The rates at which metsulfuron methyl was applied were too low for the control of macchia vegetation.

Optimum application rates were 2000 g and 1500 g active ingredient (a.i.)hexazinone/ha and 1500g a.i. glyphosate/ha. Metsulfuron methyl at 27 g a.i./ha successfully controlled Rubus spp.

A modification of the GLEAMS model was used to determine application windows which would optimise efficacy and environmental safety for herbicide application to a forest site. Herbicide/soil partition coefficients were determined using soil samples collected from the study site for two herbicides (imazapyr, K_oc=46; triclopyr ester, K_oc= 1038) and published values for two other herbicides (hexazinone, K_oc=54; triclopyr amine, K_oc=20) were used in the model. Other site-specific characteristics were taken from catchment topographic maps and soil data. Long-term climatic records for the region were then used to provide meteorological data for use in the hydrology component of the model. The model was run with herbicide application for each day of the manufacturers' recommended growing-season application windows. Average surface run-off losses, expressed as a percentage of applied, were low for all herbicides modelled (hexazinone, 0.37%; imazapyr, 0.34%; triclopyr amine, 0.21 %; triclopyr ester, 1.85%). Model predictions of herbicide loss for each application day were then summarised and the application days with the lowest predicted loss within the manufacturers' application windows were identified as the environmentally safest days (environmental window) for application. This application of the GLEAMS model predicts an environmental window, for each herbicide for the site under consideration, during which the probability of adverse environmental impacts is at the lowest level achievable based upon long-term climatic records. The environmental window for pesticides other than herbicides can also be determined through modelling in a similar fashion. GLEAMS is the model chosen for this example, but other pesticide fate models may be equally applicable.

A low-volume laboratory research spray system and the spinning disk atomisers Flak, Herbi, and Micromax were used to study the effect of droplet size and spray volume on glyphosate efficacy on trembling aspen (Populus tremuloides Michx). Five droplet sizes, each with a narrow droplet spectrum, ranging from 177 to 1589 _mm VMD were used. Spray volumes from 15 to 120 l/ha were tested. The phytotoxicity of glyphosate increased as carrier volume was reduced (thereby increasing herbicide concentration), but was not affected by changes in droplet size. There was no significant interaction between droplet size, spray volume, and active ingredient (a.i.) rate per hectare. It is possible that the concentration gradient between the droplet and leaf, rather than the droplet coverage, is important for glyphosate phytotoxicity.

Spot spraying, where only the area around individual trees is treated, is becoming an increasingly important method of herbicide application during establishment of Pinus radiata D. Don plantations in New Zealand. Minimising the spot size reduces costs and has perceived environmental benefits from reduced herbicide use. Trials were undertaken at two sites to determine the effect of area and duration of spot weed control on P. radiata growth. One year after planting, crop growth benefits from weed control were proportionally greater on the more productive site. Here, crop diameter growth continued to increase, albeit at a declining rate, as spot size was increased to the point of complete weed control. A similar trend was apparent with height growth up to spot diameters of about 1.5 m. With larger spot sizes, further gains in height growth were minimal. On the less productive site, significant growth benefits were apparent only from the smallest spot size treatment. At both sites, diameter growth was more sensitive to weed control than height growth.

World-wide, sediment is the major water quality problem. The use of herbicides for controlling competing vegetation during stand establishment can be beneficial to forest ecosystem sustainability and water quality by minimising off-site soil loss, reducing on-site soil and organic matter displacement, and preventing deterioration of soil physical properties. Sediment losses from sites where competing vegetation is controlled by mechanical methods can be 1 to 2 orders of magnitude greater than natural losses from undisturbed watersheds. On a watershed basis, vegetation management techniques in general increase annual erosion by <7%. Herbicides do not increase natural erosion rates. Organic matter and nutrients that are critical to long-term site productivity can be removed off-site by mechanical vegetation-management techniques and fire, or redistributed on-site in a manner that reduces availability to the next stand.

For several decades, research has been conducted on the fate of forestry-use herbicides in various watersheds throughout the southern and western United States, Canada, and Australia. This research has evaluated chemicals such as 2,4-D, glyphosate, hexazinone, imazapyr, metsulfuron methyl, picloram, sulfometuron methyl, tebuthiuron, and triclopyr. Losses in streamflow, and leaching to groundwater have been evaluated. Field study data indicate that residue concentrations tend to be low, except where direct applications are made to ephemeral channels or streams, and do not persist for extended periods of time. Regional environmental impact statements in the United States demonstrate that forestry herbicide presence in surface and groundwater is not a significant risk to water quality or human health. They also clearly indicate that herbicides can greatly reduce water quality deterioration that is produced by erosion and sedimentation.

Public perception of environmental quality is primarily visual, and forest operations, particularly logging, can have a significant negative impact on a visual amenity. In order to visualise landscape change and evaluate reaction to such change, graphic simulations can be produced with a range of technologies. Objective and dynamic, though abstract, models of the landscape can be produced using digital terrain modelling. Photorealistic, though static and subjective, impressions of landscape change can be produced employing digital image-editing software. Image-rendering technologies are available for reducing the abstract appearance of digital terrain models and for increasing the quality of viewer/ image interaction. Spatial analysis and image processing software facilitate quantitative measurement of visual impacts.

Existing components of a forest establishment decision-support system (DSS), including a model of the initial growth of Pinus radiata D.Don sensitive to site preparation, a knowledge-based module for designing establishment strategies, and a knowledge-based system for selecting herbicides, were combined with the STANDPAK stand modelling system (which integrates growth and yield models with models of stem defect and log valuation) and a spreadsheet in order to design a cost-effective establishment regime for P. radiata on a specific site. The exercise showed that there were some important deficiencies in models, and software components need to be better integrated.

A modelling approach was used to evaluate the effects of weed control on the growth and survival of Picea mariana (Mill.) B.S.P. (black spruce) up to 11 years after planting. The data were generated from a split-plot experimental design with a completely randomised arrangement of whole-plot treatments (2 herbicides x 3 replicates). There were six split-plot treatments distinguished by stock type (0.4-, 0.6-, and 1.5-gpaperpots, and 1.5/1.5 bareroot transplants) and planting season (spring and summer). The two weed-control treatments were an untreated control and glyphosate applied at 70 l/ha with a spinning disc applicator at 2.14 kg a.e./ha. The experiment was located in north-eastern Ontario, Canada, on an upland mixed-wood herb-rich site in the boreal forest region.

Of the models tested, the exponential and linear-exponential, respectively, provided the best fits to the seedling growth and survival data. An interpolation procedure was developed to augment the diameter observations with additional estimates for years when only seedling height (and survival) were recorded. Because of the serial correlation over time in the data, model parameter estimates were used as primary data in multivariate analyses to test for treatment effects.

Reduction of weed competition almost always accelerated the growth of the black spruce outplants. Eight growing seasons after weeding, the trees on the weeded plots were up to almost three growing seasons ahead of their counterparts on the non-weeded plots. By the end of the experiment, the growth advantage for trees in weeded plots relative to those on non-weeded plots was increasing with respect to volume at about 1.5 times its rate with respect to height. Tree survival was not significantly affected by weed control, planting season, or stock type. Planting season and stock type did affect tree growth, however. The relative rates of volume growth of the spring-planted stock exceeded that of the corresponding summer-planted stock by 10-14%. The bareroot stock was initially taller and increased in volume at a relative rate which was 4-22% faster than stock in the 0.4-g paperpots in the same weed control and planting season regimes. This superiority of the bareroot stock over the paperpot stock was 3-5% greater when planted in the summer than in the spring.

A physically based model was developed to simulate changes in plant water use and soil water distribution over time. It was built from two main submodels which dealt with soil water flow and plant water uptake. The soil water flow submodel employed the Richards equation combined with vapour flow procedures to predict soil water distribution, deep drainage, and soil evaporation. The plant water uptake submodel used a potential driven approach in that it depended on the water potential gradient between the roots and soil matrix. The model was used to analyse the effects of rooting patterns on deep drainage, which is one of the main differences in plant properties between trees and annual crops or grasses. Root distribution affects plant water uptake, soil water distribution, and deep drainage. The deep drainage rate under Armidale (NSW, Australia) weather conditions is episodic in character and root distribution influences deep drainage during wet periods. From discussion of the optimal root distribution for an agroforestry system it was concluded that to prevent too much deep drainage, it may not be necessary to have only deep-rooted species present. Simulation can indicate the appropriate ratio of deep-rooted trees to shallow-rooted crops or grasses in an agroforestry system where the aim is to minimise deep drainage while maximising the percentage of productive shallow-rooted crops or grasses.

Although New Zealand tree plantations have traditionally been established by planting bare-root seedlings, there is now increased interest in the use of container-grown stock. Optimum container design is currently seen as one that incorporates lateral root-pruning. Size specifications for stock for planting out in New Zealand vary according to species.

Two experiments were established in the central North Island of New Zealand to examine survival and growth of Pinus radiata D.Don in response to weed control and methods of soil cultivation. Fifth- and sixth-year tree height, diameter and survival were examined in relation to (a) initial tree size expressed in various ways, (b) intensity of weed control, and (c) method of soil cultivation. Of four measures of initial seedling size tested, seedling ground line diameter (OLD) was best correlated with tree performance at one site while initial GLD squared Î height was most significant at the other. Control of weeds improved tree growth at both sites, and markedly improved survival of trees at the higher altitude site. Analysis of residuals of an initial growth model constructed with data from 27 experiments suggested that stocks of 1/0 seedlings with mean initial root collar diameters of less than 5 mm performed poorly compared with larger 1/0 stocks.

Survival and early growth of two grades of bare-root Pinus radiata D. Don seedlings were studied in response to soil cultivation (augering, disking, pitting, or ripping) combined with standard or intensive weed control. Soil tillage after clearfelling did not improve volume yields but did increase establishment costs. Overall, intensive weed control improved fourth-year volume per hectare by 108%. Planting Grade A seedlings (4.1 mm average root-collar diameter) instead of Grade B seedlings (2.8 mm average root-collar diameter) brought about a 36% increase in volume per hectare. For the pitting treatment, the unit cost of the additional volume produced by planting Grade A stock was approximately US$7/m³ compared to $10/m³ for the additional volume resulting from intensive weed control. This study suggests that planting larger diameter seedlings and allowing 1500 kg competing biomass/ha (1 year after planting) may be ecologically more advantageous and economically equivalent to planting small grade seedlings and total eradication of vegetation cover.

New Zealand has built a very successful, intensively managed, plantation forest industry. The key for its future success will be to ensure all aspects of forestry have low impacts on both the environment and the community. Comprehensive planning prior to carrying out operations is considered pivotal for ensuring that environmental standards are not compromised.

The New Zealand Forest Code of Practice provides a planning process that will allow the practical requirements of sound and effective environmental management to be met. Key components of the planning procedure include the impact appraisal procedure, which systematically evaluates the potential impact of proposed operations on identified site values, and the operations database which aids selection of suitable techniques in addition to providing the basis for an operational monitoring programme.

Making good decisions for natural resource management has become increasingly difficult. Forest managers have too much information, too many decisions to make, and too little time to do any of it. In addition, the objectives are changing and the diversity of needs that must be met is growing. Computer decision aids can help sort and process information and expand our ability to make good decisions in the face of these constraints. Artificial intelligence (AI) technology allows inclusion of knowledge processing in the decision support environment. Management of forest ecosystems involves a mixture of quantitative and qualitative elements that can be captured effectively with multi-component decision support systems (DSS) including expert systems, databases, models, geographic information systems, and user interface components such as hypertext and graphics. AI technology and expert systems are less visible as separate entities and more often occur as part of broader decision support environments. Decision support systems have been developed fora diversity of resource management applications, including management of wetlands, grazing lands, landscapes, community forests, and biodiversity, to name just a few. Two areas of current research are development of more effective integrating paradigms and transfer of the products of AI/DSS technology to the end user.

Factorial combinations of insect, nutrient, and weed control treatments were applied repetitively to eight Pinus ponderosa Dougl. ex Laws. var. ponderosa (ponderosa pine) plantations across a broad span of site qualities in California. Findings from the first 6 years showed that plantation growth potential was several times greater than previously realised. On the average, tree volumes from combined treatments were 4 times greater than for untreated controls. Sites of poor quality were limited primarily by soil drought, and fertiliser offered no further advantage over weed control alone. However, droughty sites often are infertile. Weeding not only improved moisture availability, but improved nutrient availability as well, and the joint effects were not easily separated. Controlling vegetation repeatedly with herbicides tripled volume growth and produced significant increases in foliar nutrient concentrations. Sites of medium quality also were limited by soil drought as well as by soil infertility. However, fertiliser application without vegetation control boosted weed growth which in turn blocked trees from a fertiliser response. Plantations on the best sites responded positively to both herbicides and fertilisers and effects were additive when treatments were combined. Despite an extreme range of site and plant stress conditions, insects had little influence on plantation development. Enhanced soil fertility and elevated foliar nutrient concentrations did not render trees more susceptible to insects. Thus, insecticides had no effect on tree growth. Findings suggest that weed control is essential for plantation development on poor sites under the Mediterranean climatic conditions, but not on better sites. Fertiliser responses are expected to continue for several years.

Planted Cryptomeria japonica D.Don and Chamaecyparis obtusa (Sieb, et Zucc.) Endl. seedlings were weeded to leave weeds around a seedling and growth was studied. The more weeds left around the seedling, the smaller was the exposure ratio of the seedling. Seedling growth decreased as the exposure ratio decreased. The higher the site quality, the greater was the advantage that weeding gave seedlings. These results implied that the value of mechanisation for labour-saving in weeding work depended on the site quality.

A field experiment was established between 1989 and 1993 on 4 sites in southern Sweden to study the effects of the clearcut age on damage by pine weevils and competing vegetation. On each site, Norway spruce (Picea abies) seedlings were planted on 0- to 4-yr-old clear felled areas. In the study reported here, the effects of clearcut age, mounding, herbicide, mowing, and removal of slash on growth of ground vegetation were investigated. The study was restricted to analysis of second-year growth of planted Norway spruce seedlings, i.e. established seedlings. Ground vegetation was sparse on fresh and 1-yr-old clearcuts. On older clearcuts the dry weight of ground vegetation ranged between 1 and 4 Mg/ha, with a considerable variation between years and sites. Seedling growth was negatively influenced by ground vegetation. Mounding and herbicide treatments reduced ground vegetation and increased seedling growth, especially on older clearcuts, while mowing had no effect. Slash removal had no significant effect on the amount of ground vegetation and did not affect seedling growth. Carbon isotope analysis (¹³C abundance), predawn water potential, and needle conductance did not reveal a consistent difference in water stress between seedlings in undisturbed and vegetation-controlled plots, even during dry periods. There was no evidence that competitor effects on radiation were related to seedling growth response. Therefore, it was concluded that competition between ground vegetation and planted seedlings for water and light alone could not explain the observed differences in growth. Some data are also presented on nitrogen (nitrate and ammonium) in soil water, and on nitrogen isotope analysis (15N abundance) in seedlings, and it is suggested that the results indicate that growth was restricted by nitrogen availability.

Growth during the first two growing seasons in response to different levels of aboveand below-ground competition was evaluated for Norway spruce (Picea abies L. Karst) seedlings on a clay moraine in southern Sweden. Spruce seedlings were planted in plots and left undisturbed, or mowed, or mowed and shaded, or treated with herbicides, or treated with herbicides and shaded. The aim of the shading treatment was to simulate the light levels below the canopy of the ground vegetation in the undisturbed plots. Spruce seedling growth was reduced by the presence of competing vegetation regardless of whether the vegetation was mown or not. Shading only reduced growth to a minor extent. Soil water potentials were reduced in all treatments during periods of drought. However, undisturbed plots showed the lowest levels of soil moisture and the periods of drought extended over longer periods of time than in plots where the vegetation was either mown or removed with herbicides. Carbon isotope analysis and measurement of needle lengths indicated no differences in water stress for seedlings in undisturbed plots compared to seedlings in herbicide-treated plots. Seedlings in undisturbed and mown plots had lower needle nitrogen concentrations than seedlings in herbicide-treated plots. Therefore, it was concluded that competition between weeds and seedlings planted on clay moraine was mainly below ground and that lower availability of nitrogen in untreated and mown plots may have been an important limitation for growth.

A loblolly pine (Pinus taeda L.) plantation study was installed across the southeastern United States by members of the Auburn University Silvicultural Herbicide Cooperative. This study was designed to quantify and model the effects of varying levels of competing vegetation on long-term pine growth. A matrix of initial stand conditions was established encompassing three site index classes, six plantation ages (0-5 years), four hardwood rootstock density classes (class limits vary by age), and absence or presence of herbaceous weed control.

Five 0.13-ha treatment plots, with 0.06-ha measurement plots, were established at each location. Variation among plots was controlled by sampling and matching plots based on pre-treatment pine and hardwood tree and stand attributes. Each plot was selected for no treatment (check), total hardwood control for 1 year, or specific levels of intermediate hardwood control (one-time treatment by basal spray of herbicide). Herbaceous weed control was combined with hardwood control for certain treatments at selected locations to evaluate the impact of herbicides that control both plant components. Establishment of plots began in 1987, and 56 locations were active in 1995.

The first model presented predicted response in pine basal area as a function of age of treatment and hardwood basal area response at age 8. This model expressed the tradeoff between hardwood and pine basal area and clearly showed larger pine responses per unit of hardwood control at younger ages of release (1.61 m² of pine basal area response per square metre of hardwood control at treatment age O, v. 0.01 m² pine basal area response at treatment age 5). The second model predicted age 8 hardwood basal area as a function of sum of hardwood rootstock heights per hectare and number of hardwood rootstocks per hectare from early stand evaluations (ages 1-5). Prediction of age 8 hardwood basal area allowed stands of different ages to be ranked for need of release through projection of long-term yield loss. Pine yield reduction at rotation due to hardwood competition can be estimated for young stands by using this hardwood basal area prediction model in growth and yield models which use an estimate of hardwood basal area at or past age 8.

The use of low-cost, large-scale aerial photographs to measure woody plant attributes was assessed for a variety of early successional forest ecosystems. Two conventional 35-mm cameras were mounted on a boom and suspended from a balloon or tripod platform to obtain nominally vertical stereo photographs (contact scales ranging from 1:250 to 1:1000). Photo measurements of individual plant total height were generally unbiased and precise, with regression standard errors ranging from 3.4 to 10.6 cm for plants up to 4 m tall. Standard errors for individual crown diameter measurements ranged from 7.1 to 18.4 cm for crowns up to 3.1 m in width. Direct photo estimates of crown area were unbiased and consistent (standard error=0.107 m², plants up to 6 m²). With measurements being confined to the inner 70% of the overlap portion of each photo pair, relief displacement had no discernible effect on the accuracy of crown estimates. For all measurement variables examined, the relationships between ground- and photo-measured factors were generally unaffected by species but did shift in response to high levels of crown closure and/or discrepancies in ground- and photo-measurement protocols. The most precise photo estimates originated from the larger photo scales tested. Results suggest that evaluations made from large-scale aerial photographs may be used to augment field evaluations in surveys of early successional woody plant communities situated on level terrain.

Optimum timing and duration of herbaceous vegetation control during the early development of forest plantations can be assessed using critical-period analysis. Critical periods are being developed for black spruce (Picea mariana (Mill.) B.S.P.), jack pine (Pinus banksiana Lamb.), eastern white pine (Pinus strobus L.), and red pine (Pinus resinosa Ait.) seedlings in the Great Lakes/St Lawrence forest type of Ontario, Canada. Six patterns of herbaceous vegetation control were examined: 3 consecutive years, firstyear only, first 2 years, second and third years, third-year only, and no control. Third-year survival and height were not affected by herbaceous vegetation for any conifer species. Stem diameter, however, decreased substantially without vegetation control. Third-year diameters for white pine, jack pine, black spruce, and red pine without vegetation control were 55, 56, 61, and 64%, respectively, of that observed for trees under 3 consecutive years of vegetation control. Stem volume of all species without vegetation control was reduced to between 27 and 36% of that observed under 3 consecutive years of control. Height/stem diameter ratios decreased as the degree of vegetation removal increased. Critical-period analysis indicated that herbaceous vegetation control is important immediately after planting for both tolerant and intolerant conifer species. Stem diameter gains also were proportional to the number of years of herbaceous vegetation control.

In a simple conceptual model of competition for resources the net interaction between plants is broken down into two distinct components: competitive effect on resources, or the rate at which resources are depleted by neighbouring plants, and competitive response to resources, or the degree to which a target plant is limited by resource availability. This mechanistic description of the process of competition suggests a number of ways in which predicting the impact of competition on individual plants could be simplified and made more general. All these possible simplifications have important assumptions that have rarely been tested. Therefore, they are currently best regarded as null hypotheses rather than firmly-established guides to forestry practice. Firstly, when the number of species is large, there are many fewer possible plant-resource interactions to be quantified than plant-plant interactions. This assumes that the net interaction between plants is indeed simply the composite of their competitive effects and responses. Secondly, not all possible plant-resource interactions must be studied at all stages. Specifically, the most important phase in forest vegetation management is seedling establishment of trees. In this situation, it should be necessary to quantify only the effect on resources of non-commercial vegetation and the response to resources of commercial tree seedlings to make predictions relevant to decisions about forest vegetation management. Thirdly, it should be possible to ignore species identity of competing vegetation around target tree seedlings if species of neighbours have equivalent effects on resources. This would greatly simplify the quantification of competitive interactions at any particular site. Equivalent competitive effects are most likely if sizes of species of competing plants are incorporated into measures of competitive effect. Finally, perhaps the most important factor currently complicating general predictions about the outcome of competition is variation among sites. If both effect and response curves for particular resources are known, along with the magnitude of change in abiotic resource supply, it should be possible to predict competitive effects across sites without repeating the detailed studies of effect and response.

There is debate amongst ecologists concerning the validity of particular experimental designs for the study of interference (competition) and the quest for a "best" design. The debate has become increasingly dogmatic and confusing to outsiders. Most of the criticisms relate not to the validity of the designs themselves, as is usually claimed, but to the interpretability of indices calculated from them. It is argued that it is illogical to condemn agroup of experimental treatments for all purposes simply because of the ways in which some researchers choose to interpret the results. Assertions that one particular design, the Replacement Series, is totally unusable are argued to be unwarranted and driven by an assumption that every researcher has the same objective. Each experimental design has both its merits and its restrictions; it is unreasonable to expect that any design will be universally applicable. A researcher must match the aim of the experiment with an appropriate design and method of analysis.