Search

Aim: Global studies show strong climatic control of leaf area index (LAI) and related ecosystem properties. However, alternative states stabilized by positive feedbacks have the potential to produce appreciable differences in LAI in a given environment. On certain sites in New Zealand, ectomycorrhizal (EM) southern beech forest (Nothofagaceae) appear to form a stable alternative to more species‐rich conifer‐broadleaved forest dominated by arbuscular mycorrhizal (AM) trees. We asked which climate variables best predict LAI of New Zealand's remaining old‐growth forests, and if stands dominated by Nothofagaceae have lower LAI and better‐lit understories than conifer‐broadleaved forests, when climate and other site factors are held constant. Location: New Zealand. Taxon: Seventy‐two tree species belonging to 37 angiosperm, conifer and pteridophyte families, the most frequent occurrences being Cunoniaceae, Cyatheaceae, Dicksoniaceae, Lauraceae, Nothofagaceae, Podocarpaceae and Violaceae. Methods: We quantified overstorey composition of 45 old‐growth stands, and used the LAI‐2200C Plant Canopy Analyzer to measure LAI and understorey light. We used generalized linear mixed models to determine how these variables were influenced by environment and by dominance of Nothofagaceae. Results: LAI ranged from 3.5 to 6.9, and increased linearly with mean annual temperature (MAT); percentage light transmission to the understorey spanned an order of magnitude and decreased exponentially with increasing MAT. However, at a given MAT, LAI of stands dominated by Nothofagaceae averaged 0.8–0.9 units lower than conifer‐broadleaf stands, resulting in twofold higher understorey light levels in the former. Soil water deficits also significantly depressed LAI, and increased understorey light; atmospheric water deficits had the opposite effect. Main conclusions: Temperature is the best climatic predictor of LAI and understorey light in New Zealand's remaining old‐growth forests. However, the lower LAI and canopy light interception in southern beech stands presumably reflects the light‐demanding nature of Nothofagaceae, and their apparent ability to inhibit invasion by more shade‐tolerant AM trees with heavier crowns. This inhibitory effect may reflect reduced N supply to AM plants because of suppression of N mineralization by EM fungi, as observed in some EM systems elsewhere. [ABSTRACT FROM AUTHOR]

Aim To determine the soil characteristics of Nothofagus-dominated rain forests in an ultramafic region (i.e. soils having high concentrations of metals including Mg, Fe and Ni), and whether soil characteristics may explain the location of monodominant rain forest in relation to adjacent mixed rain forest and maquis (shrub-dominated vegetation). Location New Caledonia. Methods Soil characteristics were compared among six Nothofagus-dominated rain forests from a range of altitudes and topographic positions. At four of these sites, comparisons were made with soils of adjacent mixed rain forest and maquis. Results Soil characteristics varied among the monodominant Nothofagus forests, largely due to differences between ultramafic soils and soils influenced by non-ultramafic intrusions. The soils of all vegetation types had low concentrations of nutrients, particularly P, K and Ca (both total and extractable/exchangeable), and high total concentrations of Ni, Fe, Cr and Mn. There were significant differences between the rain forests and adjacent maquis in soil concentrations of several elements (N, P, Ca, Mg and Mn), more so in surface soils than at depth, but much of this pattern may be caused by effects of vegetation on the soil, rather than of soil on the vegetation. However, there were no significant differences in soil concentrations of any mineral elements between Nothofagus forest and adjacent mixed rain forest. Main conclusions We found no evidence for soil mediation of boundaries of Nothofagus rain forest with mixed rain forest, and little evidence for the boundaries of either forest type with maquis. We suggest that the local abrupt boundaries of these monodominant Nothofagus forests are directly related to temporal factors, such as time since the last wildfire and frequency of wildfire, and that disturbance is therefore a major causal factor in the occurrence of these forests. [ABSTRACT FROM AUTHOR]

To identify the edaphic, environmental and historical factors influencing the patchy distribution of the semelparous bamboo Bambusa arnhemica F. Muell. at global, catchment and streambank scales. The entire range of B. arnhemica, a highly fire-prone savanna matrix with generally infertile soils in the north-west of the Northern Territory of Australia above the 1200 mm mean annual rainfall isohyet. Distribution surveys were conducted by air, boat and on the ground. Plot data were collected throughout the range of the species. Bambusa arnhemica occurred predominantly in gallery forests on flood-prone but nevertheless well-drained and deep alluvial soils on sloping stream banks. It ranged widely along lentic watercourses from ephemeral headwater streams to the banks of major rivers and levees on the coastal floodplain. The species did not occur in savannas; savannas adjacent to B. arnhemica gallery forests were also flood-prone and on deep alluvial soils, but were upslope on level ground. Bambusa arnhemica's infrequent non-riparian occurrences were on a wide variety of substrates but generally on soils of moderate fertility and in coastal and/or rocky areas where at least partial topographic protection from fire is likely. Within and between catchments, the distribution of B. arnhemica was idiosyncratic, occurrence being almost always continuous downstream from highly variable ‘starting’ points to the poorly drained coastal floodplain. At local scales, B. arnhemica appears constrained by poor drainage and high fire-frequencies. Enhanced soil fertility may increase its capacity to cope with fire. At the catchment and global scales, we propose that the distribution of B. arnhemica is the product of infrequent and as yet incomplete dispersal across and away from watercourses by seed that lacks specialized dispersal mechanisms, combined with passive dispersal along streams. From this we infer that B. arnhemica is neither a very recent, nor very ancient colonist from Asia. Bambusa arnhemica's circumscribed global distribution has no parallel amongst co-occurring rain forest plants and may be the product of poor dispersal capacity and a substantial rock and floodplain barrier to the east. Limited dispersal capacity may be inextricably linked to local domination of space and the subsequent creation of regeneration space by parental death. [ABSTRACT FROM AUTHOR]

Abstract Nothofagus spp. dominate the upper canopy of some rainforests on ultramafic soils in New Caledonia. These monodominant forests typically occur within, or contiguous with, larger areas of mixed-canopy rainforest. In this study the structure, diversity and composition of six Nothofagus-dominated plots were investigated, and comparisons were made with three adjacent mixed rainforest plots. Stand density and basal area (all stems ≥ 1.3 m high) in the Nothofagus plots were in the range 16,056–27,550 stems/ha and 43.1–69.9 m2/ha, respectively. There was no significant difference (P ≥ 0.05) in total stand density or basal area between the paired Nothofagus and mixed rainforests, but there were consistently fewer trees and less basal area of trees ≥ 40 cm d.b.h. in the Nothofagus forests. Species richness, species diversity (Shannon-Wiener, based on basal area) and equitability (based on basal area) of trees ≥ 20 cm d.b.h. on 0.1 ha Nothofagus plots were in the range 4–17, 0.96–3.76 and 0.45–0.87, respectively. No significant differences (P ≥ 0.05) were recorded in these three parameters between the paired Nothofagus and mixed rainforests, although species diversity was consistently lower in the paired Nothofagus forests. Comparison of dominance by density and basal area indicated that although the uppermost canopy of the Nothofagus forests was dominated by Nothofagus (70–95%), the basal area and density contribution was ≤ 55% except at Col de Yaté (≈ 85%). Analysis of similarity indicated no significant difference in stand composition of trees ≥ 20 cm d.b.h. (following removal of Nothofagus from the data set) between Nothofagus and mixed rainforests using basal area, density or presence-absence data. It is concluded that the Nothofagus-dominated forests differ from the adjacent mixed rainforests mainly by (1) dominance of the uppermost canopy, without necessarily dominance of the stand by basal area or density, and (2) the smaller basal area contributed by large trees (all species). [ABSTRACT FROM AUTHOR]