New Zealand, having large areas of hilly landscapes, is subject to the risk of soil erosion, and summer and autumn droughts that limit pasture growth, which in turn affects the livestock-based economy. The nitrogen and phosphorus input in fertilisers coupled with livestock excreta and soil disturbance impose a serious threat to downstream water quality. The planting of trees is one option used to decrease soil erosion, increase the quantity of forage and manage runoff. To date, research has mainly focused on wide spaced poplar trees for feed quality and their effects on understorey pasture growth. However, there is increasing interest in the use of densely planted willow and poplar for fodder purpose. The effects of young (< 5 yrs old) willow and poplar planted at close spacing on runoff, soil erosion, growth of understory pasture and nutrient losses have never been studied in New Zealand. Three field trials (two at Crop and Research Unit, Moginie, Manawatu and one at Riverside Farm, Masterton) were conducted between October 2004 and November 2006 that incorporated comparative establishment and growth of densely planted willow and poplar and their effects on soil moisture, runoff, sediment load and nutrient losses from grazed and fertilised farmland. It was concluded that densely planted willow and poplar (3-4 yrs) reduced total nitrogen (TN) and dissolved reactive phosphorus (DRP) by 47 % each and sediment load by 52 %. Young trees reduced surface runoff and soil moisture more as they aged. However, due to their deciduous nature willow and poplar were not effective in reducing surface runoff in winter and early spring. Sheep preferred camping under trees, especially in late spring and summer, and this led to greater deposition of dung and urine under trees than open pasture. Sheep grazing, especially in winter, significantly increased sediment and nutrient loads in runoff water. The N and P fertiliser application increased nutrient load in runoff water well above the threshold level required to initiate algal growth to create eutrophication. Densely planted willow and poplar significantly reduced understorey pasture growth by 23 % and 9 %, respectively, in their second year at Moginie, mainly due to shade, but coupled with soil moisture deficit in summer. The pasture growth in a willow browse block was 52 % of that in open pasture as a result of shade and differences in pasture species composition. Sheep browsing reduced willow leaf area significantly. Willow and poplar survival rates were similar (P > 0.05) after two years of establishment (100 % vs 90.5 %, respectively). However, willow grew faster than poplar in height (1.90 vs 1.35 m), stem diameter (43.5 vs 32.6 mm), canopy diameter (69 vs 34 cm) and number of shoots (8.7 vs 2.3) at the age of two years, respectively. The research clearly demonstrated that densely planted young willow and poplar trees can reduce runoff, sediment load and nutrient losses from farmland to freshwater, but shade and soil moisture can limit pasture growth under trees. It is recommended that willow and poplar should be planted at wide spacing on the whole farm to minimise loss of pasture. Where blocks of trees are necessary, such as willow browse blocks, sheep browsing can be used as a tool to reduce shade to improve pasture growth. Livestock access to riparian strips should be minimal to avoid livestock camping that can have deleterious effects on water quality.

Aboveground and belowground ectomycorrhizal (ECM) communities associated with different age classes of the exotic plantation species Pinus radiata were investigated over the course of two years in the North Island of New Zealand. ECM species were identified with a combined approach of morphological and molecular (restriction fragment length polymorphism (RFLP) and DNA sequencing) analysis. ECM species richness and diversity of a nursery in Rotorua, and stands of different ages (1, 2, 8, 15 and 26 yrs of age at time of final assessment) in Kaingaroa Forest, were assessed above- and belowground; furthermore, the correlation between the above- and belowground ECM communities was assessed. It was found that the overall and stand specific species richness and diversity of ECM fungi associated with the exotic host tree in New Zealand were low compared to similar forests in the Northern Hemisphere but similar to other exotic plantations in the Southern Hemisphere. Over the course of this study, 18 ECM species were observed aboveground and 19 ECM species belowground. With the aid of molecular analysis the identities of Laccaria proxima and Inocybe sindonia were clarified. In the aboveground study, five species were found associated with P. radiata that were previously not reported with this host in New Zealand (Inocybe sindonia, Lactarius rufus, Lycoperdon gunii, Rhizopogon pseudoroseolus and Wilcoxina mikolae). Belowground, the species Psudotomentella sp., P. tristis, R. luteorubescens, Tomentella sp., Wilcoxina mikolae were found as new associates of P. radiata in New Zealand, additionally nine ECM types were found that could not be identified with molecular analysis. There was little correlation between the species fruiting and the species colonising root tips. Only seven species were found in common between the above- and belowground communities, furthermore the dominant species aboveground were not observed in the belowground ECM communities. The influence of host age on the above- and belowground ECM communities of different age classes of P. radiata plantations was investigated. The aboveground species richness increased from the nursery to the oldest age group investigated (26 yrs), while diversity increased to the 15 yr old age group and decreased slightly to the oldest stand. A clear sequence of ECM species changes was observed to be related to stand age with a growing complexity over the chronosequence. The belowground ECM communities showed a different picture and richness and diversity initially decreased from the nursery to the outplanting but increased thereafter. Belowground no change in ECM composition that was directly related to the age of the host was observed, but two distinct groups of ECM species were found – a 'young' and a 'plantation forest' group, with the respective discriminating species being Rhizopogon rubescens and Type unknown Basidiomycete/Amanita muscaria. Another aspect of the study was the fate of the nursery ECM species in the outplanting and the arrival of non-nursery species. The ECM communities of seedlings in the nursery were investigated in 2006 and these seedlings were followed up over eight assessments in the field for one year, furthermore data from the 1-, 2 and 8 yr old plantation stands was analysed. It was found that the nursery species do survive the first year of outplanting and are dominant in the first year. The first non-nursery species occurred six months after outplanting but was only in minor abundance. Nursery ECM were dominant for two years after the seedlings were planted, and were completely replaced after seven years. Rhizopogon rubescens was found to be the most persistent and dominant species in the outplanting, facilitating the successful establishment of the seedlings in the plantation forest.

This thesis investigates the social and economic factors affecting small-scale forestry development in Leyte Province, the Philippines, and in particular, the potential to use typologies of rural households to aid the description and interpretation of the diversity of households in relation to forestry development. Data for the analysis of the relationships between socioeconomic factors and tree management behaviour and intentions and the construction of a typology of rural households in Leyte was gathered from four case study communities on the Island. Following focus group discussions in each of the participating communities to gather background data and populate the structured interview schedule, representative samples of 50 households were selected and interviewed in each of the four communities. Analysis of the present tree management activities of households in the four communities revealed that most households surveyed (approximately 80%) indicated that they are presently managing at least a few trees, the primary purpose of most tree management activities being to supply timber for the households' own needs. Only 10% of respondents indicated that they intend to sell trees they are presently managing, and 25% stated that they intend to plant and manage trees for the production of timber for sale in the future. Approximately 60% of responding households indicated an interest in developing commercial tree farming on the land they manage. Thus it is concluded that small-scale commercial tree growing is uncommon in the communities involved in the survey, and that many households are interested in developing their tree planting and management activities but feel constrained from participation by various factors. The first level of exploration of the socioeconomic factors affecting rural households' tree management behaviour involved univariate analyses of the relationships between households' tree management behaviour and intentions, their socioeconomic characteristics and their attitudes to forestry. The level of resources controlled by the household, in terms of the area of the land managed by the household, their tenurial security and their cash income, are correlated with higher levels of participation in forestry activities, and greater intentions to plant higher numbers of trees in the future. Some farming system variables are also related to higher levels of tree planting and management activity, including the management of livestock and of farm plots distant from their house. While control over higher than average levels of productive resources are, in general, positively correlated to the active management of trees on their land, there are patterns of exceptions to this trend. The exploration of the interrelationships between socioeconomic factors and attitudes affecting households' tree management behaviour was undertaken through the definition of a typology of rural households in relation to forestry. Five types were defined, each having different attitudes to forestry activities. The interpretation of the types was undertaken by describing and comparing the socioeconomic and behavioural characteristics of the types in the typology. The types were characterised by differences in their control of productive resources, differences in their present and intended levels and types of forestry activity, and by differences in their participation in training activities run by development programs. The characteristics of the types were found to correspond highly with descriptions of the socioeconomic factors affecting forestry activities of smallholder households reported by previous studies into and theories about the socioeconomic factors affecting smallholder forestry development. The typology of rural households does help to describe and interpret the variation within each of the four communities in terms of households' attitudes to forestry development and their socioeconomic characteristics. It is concluded that these variations between households mean that the various types of households will be affected in different ways by forestry development programs. It is also concluded that the present state of forestry policies and the market for timber products is such that substantial increase in the level of forestry activity by smallholders is unlikely without comprehensive land use planning, policy reform in regards to tree registration and transport permits, and market development. Recommendations for further research and policy development arising from the thesis focuses on the need to create enabling conditions in which forestry activities can occur and on ways to address the differing needs of the various types defined in the typology.

This thesis investigates the social and economic factors affecting small-scale forestry development in Leyte Province, the Philippines, and in particular, the potential to use typologies of rural households to aid the description and interpretation of the diversity of households in relation to forestry development. Data for the analysis of the relationships between socioeconomic factors and tree management behaviour and intentions and the construction of a typology of rural households in Leyte was gathered from four case study communities on the Island. Following focus group discussions in each of the participating communities to gather background data and populate the structured interview schedule, representative samples of 50 households were selected and interviewed in each of the four communities. Analysis of the present tree management activities of households in the four communities revealed that most households surveyed (approximately 80%) indicated that they are presently managing at least a few trees, the primary purpose of most tree management activities being to supply timber for the households' own needs. Only 10% of respondents indicated that they intend to sell trees they are presently managing, and 25% stated that they intend to plant and manage trees for the production of timber for sale in the future. Approximately 60% of responding households indicated an interest in developing commercial tree farming on the land they manage. Thus it is concluded that small-scale commercial tree growing is uncommon in the communities involved in the survey, and that many households are interested in developing their tree planting and management activities but feel constrained from participation by various factors. The first level of exploration of the socioeconomic factors affecting rural households' tree management behaviour involved univariate analyses of the relationships between households' tree management behaviour and intentions, their socioeconomic characteristics and their attitudes to forestry. The level of resources controlled by the household, in terms of the area of the land managed by the household, their tenurial security and their cash income, are correlated with higher levels of participation in forestry activities, and greater intentions to plant higher numbers of trees in the future. Some farming system variables are also related to higher levels of tree planting and management activity, including the management of livestock and of farm plots distant from their house. While control over higher than average levels of productive resources are, in general, positively correlated to the active management of trees on their land, there are patterns of exceptions to this trend. The exploration of the interrelationships between socioeconomic factors and attitudes affecting households' tree management behaviour was undertaken through the definition of a typology of rural households in relation to forestry. Five types were defined, each having different attitudes to forestry activities. The interpretation of the types was undertaken by describing and comparing the socioeconomic and behavioural characteristics of the types in the typology. The types were characterised by differences in their control of productive resources, differences in their present and intended levels and types of forestry activity, and by differences in their participation in training activities run by development programs. The characteristics of the types were found to correspond highly with descriptions of the socioeconomic factors affecting forestry activities of smallholder households reported by previous studies into and theories about the socioeconomic factors affecting smallholder forestry development. The typology of rural households does help to describe and interpret the variation within each of the four communities in terms of households' attitudes to forestry development and their socioeconomic characteristics. It is concluded that these variations between households mean that the various types of households will be affected in different ways by forestry development programs. It is also concluded that the present state of forestry policies and the market for timber products is such that substantial increase in the level of forestry activity by smallholders is unlikely without comprehensive land use planning, policy reform in regards to tree registration and transport permits, and market development. Recommendations for further research and policy development arising from the thesis focuses on the need to create enabling conditions in which forestry activities can occur and on ways to address the differing needs of the various types defined in the typology.

Aerial control using 1080 (sodium monofluoroacetate) baits is widely used in New Zealand for the control of introduced brushtail possums (Trichosurus vulpecula), with the aim of protecting national conservation and agricultural values from these damaging pests. This thesis integrates research, completed over 25 years, that was motivated by growing recognition in the 1970s of the extent of possum impacts and the need to improve the effectiveness and efficiency of the control operation. Field research assessed the palatability of three types of cereal-based pellet baits and carrot baits in different regions, habitat types and seasons. Palatability was assessed by the consumption of the different bait types presented independently of each other on 15-30 plots, with rotation of bait types at plots on successive nights to provide equal exposure to each bait type. There was regional variation in possums' bait preferences, possibly reflecting genotypic differences, whereas seasonal variation was less evident. Carrot bait was preferred or equally preferred to cereal bait in 14 out of 20 field trials. The proportion of possums eating baits was then investigated by, firstly, developing a technique for tracing bait acceptance using rhodamine B, a UV-fluorescent dye. In four field trials, more than 95% of possums accepted three types of dye-marked bait, eliminating bait refusal as a major reason for low kills in winter control operations. In a fifth trial, conducted in summer, only 68% of possums accepted bait suggesting that seasonal availability of favoured foods may influence bait acceptance. Since possums must encounter baits before deciding whether to eat them, field studies were undertaken to assess the coverage achieved in normal aerial baiting operations. Large gaps, up to 400 m in width, were often found between baiting swaths; these could allow some possums to survive. A controlled field experiment, using acceptance of rhodamine-dyed bait as a measure of effectiveness, showed that bait distribution was least accurate where flight paths were not marked. Where gaps of 100 m between flight paths were deliberately created, bait acceptance was slower and less than where coverage was complete. Sowing baits at 3 kg/ha was as effective as at 10 kg/ha, indicating the potential for substantially reducing operational costs by using machinery capable of faultlessly distributing baits at low rates. Navigational guidance systems were evaluated and found to improve the accuracy of bait distribution. During 1993-1997, when a lower sowing rate of 5 kg/ha was adopted operationally by regional managers, control effectiveness was unchanged but annual savings of around $9 million accrued. Because of the lack of suitable sowing machinery, a bucket was developed to permit faultless distribution of baits at lower rates, demonstrating the possibility of yet further cost-savings. The possibility of seasonal food availability affecting bait acceptance was investigated in three different forest habitats. Dyed baits were aerially distributed on 100 ha at each site in each season over two years. In each trial, fat-based condition indices of possums were calculated and the abundance of possum-preferred plant foods described. Bait acceptance was consistently high (85-100%) in the 24 trials, and was not influenced by either condition or availability of preferred foods. It seems likely that seasonal variation in operational effectiveness is caused by either the availability of sharply seasonal, scarce foods that possums may feed on intensively for brief periods, or by warmer temperatures that render 1080 less effective. The influence of 1080 on acceptance of (rhodamine-dyed) baits was investigated in a field trial. Examination of possums for dye-marking showed that 25% of possums refused to eat either a lethal quantity of bait or any bait at all, compared with 98% of possums eating non-toxic bait. This indicated that 1080 is aversive to possums, which is a potential major reason for their surviving control operations. Pen trials were therefore conducted to further examine the problem and to seek solutions. Toxic carrot baits were rejected by 27.5% of possums, equally by smell and taste aversion, whereas toxic cereal pellets were rejected by 34%, mainly by taste aversion. Orange and cinnamon were shown to be among the most preferred of 42 flavours tested and, when applied to toxic baits, 1080 was effectively masked. Bait refusal was reduced to ≤7%, the same as that recorded for possums presented with flavoured non-toxic baits. For long-term control of possum populations, aerial 1080 baiting can be used sequentially with other poisoning methods. However, the compatibility of these methods is dependent on the likelihood of possums developing bait shyness if sublethally dosed. Studies were therefore conducted to characterise and compare the four main toxicants used (1080, cyanide, cholecalciferol and brodifacoum) for induction and mitigation of bait shyness. Shyness was induced in approximately 80% of possums sublethally dosed with cyanide, 60% with 1080, 20% with cholecalciferol, and 0% with brodifacoum. Cyanide and 1080 shyness were found to persist in many possums for at least 12 and 24 months, respectively. Use of alternative bait types, and of baits containing an alternative slow-acting toxin (brodifacoum) were shown to be effective ways of overcoming shyness. This, and other related research, is reviewed to provide operational specifications that maximise the likelihood that all targeted possums will (i) encounter bait, (ii) eat it, and (iii) die. The likely future use of aerial 1080 baiting is described and the technological, economic, environmental and social constraints on its sustainability are discussed. Finally, the uptake of the research by possum managers is considered, and areas identified in the thesis where information is incomplete are summarised as prioritised topics for further research.

An established host-parasitoid-hyperparasitoid system was used to investigate how the physiological and behavioural characteristics of parasitoids influence the outcomes of laboratory-based host specificity tests. The characteristics of the two pteromalid egg parasitoids, Enoggera nassaui (Girault) and Neopolycystus insectifurax Girault, were assessed and interpreted in regard to the particular host specificity testing methods used and the control of the eucalypt defoliating beetle Paropsis charybdis Stål (Chrysomelidae) in New Zealand. The physiology of N. insectifurax was examined to determine how to increase production of female parasitoids that were physiologically capable and motivated to parasitise P. charybdis eggs in laboratory trials. Neopolycystus insectifurax were found to be more synovigenic than E. nassaui. Provisioning them with honey and host stimuli for three days, and allowing females to parasitise hosts in isolation (i.e. in the absence of competition) was an effective means of achieving these goals. No-choice tests were conducted in Petri dish arenas with the four paropsine beetles established in New Zealand. All four were found to be within the physiological host ranges of E. nassaui and N. insectifurax, but their quality as hosts, as indicated by the percent parasitised and offspring sex ratios, varied. The results of paired choice tests between three of the four species agreed with those of no-choice tests in most instances. However, the host Trachymela catenata (Chapuis), which was parasitised at very low levels by E. nassaui in no-choice tests, was not accepted by that species in paired choice tests. A much stronger preference by N. insectifurax for P. charybdis over T. catenata was recorded in the paired choice test than expected considering the latter was parasitised at a high level in the no-choice test. The presence of the target host in paired choice tests reduced acceptance of lower ranked hosts. Both no-choice and choice tests failed to predict that eggs of the acacia feeding beetle Dicranosterna semipunctata (Chapuis) would not be within the ecological host range of E. nassaui and N. insectifurax. Behavioural observations were made of interspecific competition between E. nassaui and N. insectifurax for access to P. charybdis eggs. Two very different oviposition strategies were identified. Neopolycystus insectifurax were characterised by taking possession of, and aggressively guarding host eggs during and after oviposition. They also appeared to selectively oviposit into host eggs already parasitised by E. nassaui, but did not emerge from significantly more multi-parasitised hosts than E. nassaui. Enoggera nassaui did not engage in contests and fled when approached by N. insectifurax. Although often prohibited from ovipositing by N. insectifurax, E. nassaui were able to locate and begin ovipositing more quickly, and did not remain to guard eggs after oviposition. It is hypothesised that although N. insectifurax have a competitive advantage in a Petri dish arena, E. nassaui may be able to locate and parasitise more host eggs in the field in New Zealand, where competition for hosts in is relatively low. The biology of the newly established encyrtid Baeoanusia albifunicle Girault was assessed. It was confirmed to be a direct obligate hyperparasitoid able to exploit E. nassaui but not N. insectifurax. Field and database surveys found that all three parasitoids have become established in many climatically different parts of New Zealand. Physiological characteristics were identified that may allow B. albifunicle to reduced effective parasitism of P. charybdis by E. nassaui to below 10%. However, the fact that hyperparasitism still prevents P. charybdis larvae from emerging, and that B. albifunicle does not attack N. insectifurax, may preclude any significant impact on the biological control of P. charybdis. Overall, parasitoid ovigeny and behavioural interactions with other parasitoids were recognised as key characteristics having the potential to influence host acceptance in the laboratory and the successful biological control of P. charybdis in the field. It is recommended that such characteristics be considered in the design and implementation of host specificity tests and might best be assessed by conducting behavioural observations during parasitoid colony maintenance and the earliest stages of host specificity testing.

1. The resource management issue under consideration is the conservation and ecological restoration of natural lowland forest and its associated plant and animal communities (referred to here as native bush), which have considerable human-use and intrinsic values. 2. Within the issue of native bush conservation in New Zealand, the topic of this research is the phenomenon of citizens’ involvement as pro-active volunteers, tackling ‘bush restoration’ on public or trust-owned sites. This phenomenon is not rare but is little researched. Some key initial questions are: How extensive is citizen-led bush restoration as a means of environmental management? What motivates the volunteers and what factors advance or hinder their efforts? Why is this phenomenon not more widespread in New Zealand society, and how could it be? 3. The research strategy adopted to address the selected topic is firstly a scoping study to define the issue in context, including multi-disciplinary literature review, followed by a structured qualitative telephone survey of a representative sample of restoration project spokespeople. An analysis of these information sources provides insights and understanding related to the research questions, and raises additional questions. 4. Conclusions offer a potential contribution to the national and local-scale policy agenda in support of citizen-led bush restoration. Recommendations seek to recognise the value of voluntary bush restoration, better enable the committed volunteers, promote new voluntary action and improve the skills of volunteers with a view to enhanced quality of work.

A series of glasshouse and laboratory experiments was carried out to enable comparison of two woody perennial legumes, gorse and broom, with other legumes, nodulated non-legumes and other biological nitrogen fixing systems. Both species had distinct juvenile phases in which broom closely resembled herbaceous species in appearance, but adult plants of both species bore little resemblance to each other or to other legume species. Nodule development was similar to that of other legumes, but mature nodules exhibited structural adaptations to longevity - meristematic activity, a well developed vascular system and numerous cytoplasmic granules in cortical cells. Acetylene reduction and ¹⁵N₂ fixation continued for much longer following excision than has been observed in other legumes. In all experiments, broom nodules exhibited higher rates of acetylene reduction and nitrogen fixation than did gorse nodules. The first detectable product of nitrogen fixation in excised nodules - ammonia - was rapidly incorporated into amide and ∝ amino groups and another unidentified fraction. The principle free amino acid in nodules and sap was asparagine. Its preponderance increased as plants aged. Whole nodulated plants and excised nodules of both species exhibited a relatively low temperature optimum for growth and nitrogen fixation (22°C). They were very sensitive to elevated temperatures. Results indicated that gorse and broom have relatively low light requirements. When aeration was sufficient, combined nitrogen had little effect on growth of nodulated plants. Nodulation in both species was reduced by increasing amounts of combined nitrogen. High levels (100 mg/1) of nitrate and ammonia caused considerable inhibition of nitrogen function. Both species showed large responses to phosphate, but were able to grow and fix nitrogen when supplied with low amounts of phosphate. Boron deficiency reduced nitrogen fixation. Nodulation was increased to compensate for this. Considerable amounts of nitrogen can be contributed to the ecosystem in gorse end broom litter. Direct transfer between gorse or broom and Pinus radiate is likely to be small and may be masked by competition for other nutrients. These findings are discussed with respect to the use of gorse and broom to overcome nitrogen deficiency in reafforestation on the Moutere Gravels, in Nelson, N.Z.

Although species maintenance in small forest fragments relies on successful regeneration and recruitment, few studies have examined the effects of fragmentation on regeneration processes. New Zealand's podocarp species rely on large disturbance openings operating across a vegetated landscape to stimulate regeneration. Clearance of vegetation that results in small fragments of forest removes regeneration opportunities for podocarps by destroying the intact vegetation mosaic, and as a result may exclude disturbances of the scale necessary for podocarp regeneration. Fragmentation alters the disturbance regime of the landscape, with important implications for the regeneration of podocarps on Banks Peninsula. The four remaining lowland podocarp-hardwood fragments on Banks Peninsula were sampled to determine the structure and regeneration patterns of podocarps and to assess their long term viability. Density, basal area, and size and age class distributions were used to examine current composition, and in conjunction with spatial analysis, to identify past regeneration patterns and infer likely future changes in composition and population structure. Podocarp size and age class structures for three of the four fragments were characteristically even-sized and relatively even-aged (eg; Prumnopitys taxifolia c. 350 to 600 years), with little or no regeneration for approximately the last 200 years (old-growth fragments). Regeneration of the current podocarp canopy in the old-growth fragments may have been stimulated by flooding. The fourth younger fragment showed much more recent regeneration with Prumnopitys taxifolia, Podocarpus totara and Dacrycarpus dacrydioides mostly 80-160 years old, and substantial populations of seedlings and saplings, probably as a result of anthropogenic fire. In the absence of major disturbance the podocarp component in forest fragments on Banks Peninsula is likely to decline with composition shifting towards dominance by hardwood species. There is some evidence to suggest that canopy collapse will stimulate some podocarp regeneration within the fragments, however it appears to be unlikely that podocarps will persist on Banks Peninsula indefinitely within the fragments studied. There is an urgent need for more quantitative research in New Zealand fragmentation literature, and a need for more emphasis on processes. Banks Peninsula offers potential for a more landscape scale approach in forest management, and the maintenance of regenerating scrub in pockets about the Peninsula may offer the regeneration opportunities for podocarps that are lacking within protected fragments. My study took a quantitative approach in examining the effects of forest fragmentation on the demographics of podocarps and compositional change in forest fragments on Banks Peninsula.

Saltwater Forest is a Dacrydium cupressinum-dominated lowland forest covering 9000 ha in south Westland, South Island, New Zealand. Four thousand hectares is managed for sustainable production of indigenous timber. The aim of this study was to provide an integrated analysis of soils, soil-landform relationships, and soil-vegetation relationships at broad and detailed scales. The broad scale understandings provide a framework in which existing or future studies can be placed and the detailed studies elucidate sources of soil and forest variability. Glacial landforms dominate. They include late Pleistocene lateral, terminal and ablation moraines, and outwash aggradation and degradation terraces. Deposits and landforms from six glacial advances have been recognised ranging from latest Last (Otira) Glaciation to Penultimate (Waimea) Glaciation. The absolute ages of landforms were established by analysis of the thickness and soil stratigraphy of loess coverbeds, augmented with radiocarbon dating and phytolith and pollen analysis. In the prevailing high rainfall of Westland soil formation is rapid. The rate of loess accretion in Saltwater Forest (ca. 30 mm ka⁻¹) has been low enough that soil formation and loess accretion took place contemporaneously. Soils formed in this manner are known as upbuilding soils. The significant difference between upbuilding pedogenesis and pedogenesis in a topdown sense into an existing sediment body is that each subsoil increment of an upbuilding soil has experienced processes of all horizons above. In Saltwater Forest subsoils of upbuilding soils are strongly altered because they have experienced the extremely acid environment of the soil surface at some earlier time. Some soil chronosequence studies in Westland have included upbuilding soils formed in loess as the older members of the sequence. Rates and types of processes inferred from these soils should be reviewed because upbuilding is a different pedogenic pathway to topdown pedogenesis. Landform age and morphology were used as a primary stratification for a study of the soil pattern and nature of soil variability in the 4000 ha production area of Saltwater Forest. The age of landforms (> 14 ka) and rapid soil formation mean that soils are uniformly strongly weathered and leached. Soils include Humic Organic Soils, Perch-gley Podzols, Acid Gley Soils, Allophanic Brown Soils, and Orthic or Pan Podzols. The major influence on the nature of soils is site hydrology which is determined by macroscale features of landforms (slope, relief, drainage density), mesoscale effects related to position on landforms, and microscale influences determined by microtopography and individual tree effects. Much of the soil variability arises at microscales so that it is not possible to map areas of uniform soils at practical map scales. The distribution of soil variability across spatial scales, in relation to the intensity of forest management, dictates that it is most appropriate to map soil complexes with boundaries coinciding with landforms. Disturbance of canopy trees is an important agent in forest dynamics. The frequency of forest disturbance in the production area of Saltwater Forest varies in a systematic way among landforms in accord with changes in abundance of different soils. The frequency of forest turnover is highest on landforms with the greatest abundance of extremely poorly-drained Organic Soils. As the abundance of better-drained soils increases the frequency of forest turnover declines. Changes in turnover frequency are reflected in the mean size and density of canopy trees (Dacrydium cupressinum) among landforms. Terrace and ablation moraine landforms with the greatest abundance of extremely poorly-drained soils have on average the smallest trees growing most densely. The steep lateral moraines, characterised by well drained soils, have fewer, larger trees. The changes manifested at the landform scale are an integration of processes operating over much shorter range as a result of short-range soil variability. The systematic changes in forest structure and turnover frequency among landforms and soils have important implications for sustainable forest management.