Effects of roads,topography, and land use on forest cover dynamics in the Brazilian Atlantic Forest

Roads and topography can determine patterns of land use and distribution of forest cover, particularly in tropical regions. We evaluated how road density, land use, and topography affected forest fragmentation, deforestation and forest regrowth in a Brazilian Atlantic Forest region near the city of São Paulo. We mapped roads and land use/land cover for three years (1962, 1981 and 2000) from historical aerial photographs, and summarized the distribution of roads, land use/land cover and topography within a grid of 94 non-overlapping 100 ha squares. We used generalized least squares regression models for data analysis. Our models showed that forest fragmentation and deforestation depended on topography, land use and road density, whereas forest regrowth depended primarily on land use. However, the relationships between these variables and forest dynamics changed in the two studied periods; land use and slope were the strongest predictors from 1962 to 1981, and past (1962) road density and land use were the strongest predictors for the following period (1981–2000). Roads had the strongest relationship with deforestation and forest fragmentation when the expansions of agriculture and buildings were limited to already deforested areas, and when there was a rapid expansion of development, under influence of São Paulo city. Furthermore, the past (1962) road network was more important than the recent road network (1981) when explaining forest dynamics between 1981 and 2000, suggesting a long-term effect of roads. Roads are permanent scars on the landscape and facilitate deforestation and forest fragmentation due to increased accessibility and land valorization, which control land-use and land-cover dynamics. Topography directly affected deforestation, agriculture and road expansion, mainly between 1962 and 1981. Forest are thus in peril where there are more roads, and long-term conservation strategies should consider ways to mitigate roads as permanent landscape features and drivers facilitators of deforestation and forest fragmentation.     

Past,present and future land-use in Xishuangbanna,China and the implications for carbon dynamics

Land use/land cover change is an important driver of global change and changes in carbon stocks. Estimating the changes in carbon stocks due to tropical deforestation has been difficult, mainly because of uncertainties in estimating deforestation rates and the biomass in the forest that have been cut. In this study, we combined detailed land-use change over a 27-year period based on satellite images and forest inventory data to estimate changes in biomass carbon stocks in the Xishuangbanna prefecture (1.9 million ha) of China. Xishuangbanna is located in southwestern China in the upper watershed of the Mekong River, and the major forest types are tropical seasonal rain forest, mountain rain forest, and subtropical evergreen broadleaf forest. In the past when the region was completely forested the total biomass carbon would have been approximately 212.65 ± 8.75 Tg C. By 1976 forest cover had been reduced to 70%, and in addition many forests had been degraded resulting in a large decrease in the total biomass carbon stocks (86.97 ± 3.70 Tg C). From 1976 to 2003, the mean deforestation rate was 13 722 ha year⁻¹ (1.12%), and this resulted in the loss of 370,494 ha of forest, and by 2003 total biomass carbon stocks had been reduced to 80.85 ± 2.64 Tg C. The annual carbon emissions due to land-use change, mainly forest conversion to agriculture and rubber plantations, were 0.37 ± 0.03 Tg C year⁻¹ between 1976 and 1988 and 0.13 ± 0.04 Tg C year⁻¹ between 1988 and 2003. During the next 20 years, if rubber plantations expand into forests outside of reserves, shrublands, grasslands, and shifting cultivation below 1500 m the total biomass carbon stocks of Xishuangbanna will decrease to 76.45 ± 1.49 Tg C in 2023. This would reflect a loss of 4.13 ± 1.14 Tg C between 2003 and 2023, or an annual loss of 0.21 ± 0.06 Tg C year⁻¹. Alternatively, if rubber plantations only expand into areas of shifting cultivation below 1500 m, and all areas presently in shrublands and grasslands are allowed to recover into secondary forests, total biomass carbon stock of the region would increase to 92.65 ± 3.80 Tg C in 2023. Under this scenario, the growth of existing forests and the expansion of new forests would result in a net sequestration of 0.60 ± 0.06 Tg C year⁻¹. This study demonstrates that the uncertainty of biomass estimates can be greatly reduced if detailed land-use analyses are combined with forest inventory data, and that slight changes in future land-use practices can have large implications for carbon fluxes.     

Using indicators of deforestation and land-use dynamics to support conservation strategies: A case study of central Rondônia,Brazil

In Rondônia State, Brazil, settlement processes have cleared 68,000 km² of tropical forests since the 1970s. The intensity of deforestation has differed by region depending on driving factors like roads and economic activities. Different histories of land-use activities and rates of change have resulted in mosaics of forest patches embedded in an agricultural matrix. Yet, most assessments of deforestation and its effects on vegetation, soil and water typically focus on landscape patterns of current conditions, yet historical deforestation dynamics can influence current conditions strongly. Here, we develop and describe the use of four land-use dynamic indicators to capture historical land-use changes of catchments and to measure the rate of deforestation (annual deforestation rate), forest regeneration level (secondary forest mean proportion), time since disturbance (mean time since deforestation) and deforestation profile (deforestation profile curvature). We used the proposed indices to analyze a watershed located in central Rondônia. Landsat TM and ETM+ images were used to produce historical land-use maps of the last 18 years, each even year from 1984 to 2002 for 20 catchments. We found that the land-use dynamics indicators are able to distinguish catchments with different land-use change profiles. Four categories of historical land-use were identified: old and dominant pasture cover on small properties, recent deforestation and dominance of secondary growth, old extensive pastures and large forest remnants and, recent deforestation, pasture and large forest remnants. Knowing historical deforestation processes is important to develop appropriate conservation strategies and define priorities and actions for conserving forests currently under deforestation.     

Transpiration and hydraulic traits of old and regrowth eucalypt forest in southwestern Australia

We tested the hypothesis that overstorey of eucalypt forest dominated by tall, large diameter trees uses less water than regrowth stands in the high rainfall zone (>1100 mm year⁻¹) of the northern jarrah (Eucalyptus marginata) forest in southwestern Australia. We measured leaf area, cover, sapwood area and sapwood density at three paired old and regrowth stands. We also measured sapflow velocity at one paired stand (Dwellingup) from June 2007 to October 2008. Old stands had more basal area but less foliage cover, less leaf area and slightly thinner sapwood. The ratio of sapwood area to basal area decreased markedly as tree size increased. Sapwood area of the regrowth forest stands (6.6 ± 0.30 m² ha⁻¹) was nearly double that of the old stands (3.4 ± 0.17 m² ha⁻¹), despite larger basal area at the old stands. Leaf area index of the regrowth stands (2.1 ± 0.26) was only one-third larger than that at the old stands (1.5 ± 0.15); hence, the ratio of leaf area to sapwood area was larger in old stands than in regrowth stands (0.45 ± 0.022 m² cm⁻² versus 0.32 ± 0.045 m² cm⁻²). Our results are consistent with theories that trees have evolved to optimize carbon gain rather than maintain stomatal conductance. Neither sapwood density (540–650 kg m⁻³) nor sap velocity differed greatly between regrowth and old stands. At the old forest site, daily transpiration rose from 0.5 mm day⁻¹ in winter to 0.9 mm day⁻¹ in spring–summer, compared to 0.9 mm day⁻¹ and 1.8 mm day⁻¹ at the regrowth site. Annual water use by the overstorey trees was estimated to be ∼230 mm year⁻¹ for the old stand and ∼500 mm year⁻¹ at the regrowth stand, or 20% and 44% of annual rainfall. The overwhelming role of stand sapwood area in determining stand water use, combined with the marked changes in the ratio of sapwood area to basal area with tree age and size, suggest that stand overstorey structure can be managed to alter overstorey water use and catchment water yield. Silviculture to promote old-forest-like attributes may be a viable means of delivering multiple water and conservation benefits.     

Future quantities and spatial distribution of harvesting residue and dead wood from natural disturbances in Canada

Interest in the use of bioenergy is increasing because of the need to mitigate climate change, the increasing costs and finite supply of fossil fuels, and the declining price of lumber and paper. Sound bioenergy policies must be informed by accurate estimates of potential feedstock production, rights to the production, social values and economics. Two of the main sources of bioenergy feedstock from forests are (i) harvesting residue and (ii) dead wood resulting from natural disturbances (i.e. standing dead timber). We modeled the production of bioenergy feedstock from these two sources from 2005 to 2020 for Canada's managed forest south of 60° N so that this information can be used in provincial and national strategic planning. Published estimates of harvesting residue vary widely, and our objective was to provide more precise estimates based on new forest inventory data and regional modeling. Natural disturbances result in very large quantities of dead wood on the landscape, but estimates of future stocks and annual production have not previously been made. Our estimates included a 50% discount factor to net-down theoretically available quantities to a more realistic estimate of potential ecologically sustainable bioenergy feedstock. The total future annual production averaged 51 ± 17 Tg year⁻¹ from natural disturbances and 20 ± 0.6 Tg year⁻¹ from clearcut harvesting residues. Harvesting residue for the area logged varied spatially from a low of 1.0 ± 0.77 kg m⁻² year⁻¹ to a high of 6.7 ± 0.1 kg m⁻² year⁻¹. Dead wood production due to insects was forecast to peak in the Montane Cordillera of British Columbia (BC) at 16.7 Tg year⁻¹ due to the current mountain pine beetle outbreak. Total dead wood production due to fire was highest in the western portion of the boreal forest (3.6 Tg year⁻¹ in the Boreal Shield of Saskatchewan), in part due to the high frequency of fires in these ecosystems and the large area of western boreal forest, but the highest density production was in BC: >9 kg m⁻² year⁻¹ in the burned area. Our results showed that the dead wood stocks of 331 Tg oven-dry matter potentially available for bioenergy in 2020 are much smaller than the 3100 ± 84 Tg of dead wood stocks estimated based on ecosystem dynamics. While bioenergy use will accelerate the release of greenhouse gases compared to on-site decay, the energy is renewable and can be used as a substitute for fossil fuels. The net benefit to the atmosphere of forest bioenergy use is affected by many factors, and future research should further assess which sustainable wood-based bioenergy strategies yield the greatest net greenhouse gas benefits over the different time scales needed for post-disturbance forest recovery.     

Wood density in forests of Brazil's ‘arc of deforestation’: Implications for biomass and flux of carbon from land-use change in Amazonia

Wood density is an important variable in estimates of forest biomass and greenhouse-gas emissions from land-use change. The mean wood density used in estimates of forest biomass in the Brazilian Amazon has heretofore been based on samples from outside the “arc of deforestation”, where most of the carbon flux from land-use change takes place. This paper presents new wood density estimates for the southern and southwest Brazilian Amazon (SSWA) portions of the arc of deforestation, using locally collected species weighted by their volume in large local inventories. Mean wood density was computed for the entire bole, including the bark, and taking into account radial and longitudinal variation. A total of 403 trees were sampled at 6 sites. In the southern Brazilian Amazon (SBA), 225 trees (119 species or morpho-species) were sampled at 4 sites. In eastern Acre state 178 trees (128 species or morpho-species) were sampled at breast height in 2 forest types. Mean basic density in the SBA sites was 0.593 ± 0.113 (mean ± 1 S.D.; n = 225; range 0.265–0.825). For the trees sampled in Acre the mean wood density at breast height was 0.540 ± 0.149 (n = 87) in open bamboo-dominated forest and 0.619 ± 0.149 (n = 91) in dense bamboo-free forest. Mean wood density in the SBA sites was significantly higher than in the bamboo dominated forest but not the dense forest at the Acre site. From commercial wood inventories by the RadamBrasil Project in the SSWA portion of the arc of deforestation, the wood volume and wood density of each species or genus were used to estimate average wood density of all wood volume in each vegetation unit. These units were defined by the intersection of mapped forest types and states. The area of each unit was then used to compute a mean wood density of 0.583 g cm⁻³ for all wood volume in the SSWA. This is 13.6% lower than the value applied to this region in previous estimates of mean wood density. When combined with the new estimates for the SSWA, this gave an average wood density of 0.642 g cm⁻³ for all the wood volume in the entire Brazilian Amazon, which is 7% less than a prior estimate of 0.69 g cm⁻³. These results suggest that current estimates of carbon emissions from land-use change in the Brazilian Amazon are too high. The impact on biomass estimates and carbon emissions is substantial because the downward adjustment is greater in forest types undergoing the most deforestation. For 1990, with 13.8 × 10³ km² of deforestation, emissions for the Brazilian Amazon would be reduced by 23.4–24.4 × 10⁶ Mg CO₂-equivalent C/year (for high- and low-trace gas scenarios), or 9.4–9.5% of the gross emission and 10.7% of the net committed emission, both excluding soils.     

Derivation of a spatially explicit 86-year retrospective carbon budget for a landscape undergoing conversion from old-growth to managed forests on Vancouver Island,BC

To understand the influence of disturbance, age–class structure, and land use on landscape-level carbon (C) budgets during conversion of old-growth forests to managed forests, a spatially explicit, retrospective C budget from 1920 through 2005 was developed for the 2500 ha Oyster River area of Fluxnet-Canada's coastal BC Station. We used the Carbon Budget Model of the Canadian Forest Sector (CBM-CFS3), an inventory-based model, to simulate forest C dynamics. A current (circa 1999) forest inventory for the area was compiled, then overlaid with digitized historic disturbance maps, a 1919 timber cruise map, and a series of historic orthophotographs to generate a GIS coverage of forest cover polygons with unique disturbance histories dating back to 1920. We used the combined data from the historic and current inventory and forest change data to first estimate initial ecosystem C stocks and then to simulate forest dynamics and C budgets for the 86-year period. In 1920, old-growth forest dominated the area and the long-term landscape-level net ecosystem C balance (net biome productivity, NBP) was a small sink (NBP 0.2 Mg C ha⁻¹ year⁻¹). From 1930 to 1945 fires, logging, and slash burning resulted in large losses of biomass C, emissions of C to the atmosphere, and transfers of C from biomass to detritus and wood products (NBP ranged from −3 to −56 Mg C ha⁻¹ year⁻¹). Live biomass C stocks slowly recovered following this period of high disturbance but the area remained a C source until the mid 1950s. From 1960 to 1987 disturbance was minimal and the area was a C sink (NBP ranged from 3 to 6 Mg C ha⁻¹ year⁻¹). As harvest of second-growth forest began in late 1980s, disturbances again dominated the area's C budget, partially offset by ongoing C uptake by biomass in recovering young forests such that the C balance varied from positive to negative depending upon the area disturbed that year (NBP from 6 to −15 Mg C ha⁻¹ year⁻¹). Despite their high productivity, the area's forests are not likely to attain C densities of the landscape prior to industrial logging because the stands will not reach pre-logging ages. Additional work is underway to examine the relative role historic climate variability has had on the landscape-level C budget.     

Influences of forest tending works on carbon distribution and cycling in a Pinus densiflora S. et Z. stand in Korea

This study was conducted to determine carbon (C) dynamics following forest tending works (FTW) which are one of the most important forest management activities conducted by Korean forest police and managers. We measured organic C storage (above- and below-ground biomass C, forest floor C, and soil C at 50 cm depth), soil environmental factors (soil CO₂ efflux, soil temperature, soil water content, soil pH, and soil organic C concentration), and organic C input and output (litterfall and litter decomposition rates) for one year in FTW and non-FTW (control) stands of approximately 40-year-old red pine (Pinus densiflora S. et Z.) forests in the Hwangmaesan Soopkakkugi model forest in Sancheonggun, Gyeongsangnam-do, Korea. This forest was thinned in 2005 as a representative FTW practice. The total C stored in tree biomass was significantly lower (P < 0.05) in the FTW stand (40.17 Mg C ha⁻¹) than in the control stand (64.52 Mg C ha⁻¹). However, C storage of forest floor and soil layers measured at four different depths was not changed by FTW, except for that at the surface soil depth (0–10 cm). The organic C input due to litterfall and output due to needle litter decomposition were both significantly lower in the FTW stand than in the control stand (2.02 Mg C ha⁻¹ year⁻¹ vs. 2.80 Mg C ha⁻¹ year⁻¹ and 308 g C kg⁻¹ year⁻¹ vs. 364 g C kg⁻¹ year⁻¹, respectively, both P < 0.05). Soil environmental factors were significantly affected (P < 0.05) by FTW, except for soil CO₂ efflux rates and organic C concentration at soil depth of 0–20 cm. The mean annual soil CO₂ efflux rates were the same in the FTW (0.24 g CO₂ m⁻² h⁻¹) and control (0.24 g CO₂ m⁻² h⁻¹) stands despite monthly variations of soil CO₂ efflux over the one-year study period. The mean soil organic C concentration at a soil depth of 0–20 cm was lower in the FTW stand (81.3 g kg⁻¹) than in the control stand (86.4 g kg⁻¹) but the difference was not significant (P > 0.05). In contrast, the mean soil temperature was significantly higher, the mean soil water content was significantly lower, and the soil pH was significantly higher in the FTW stand than in the control stand (10.34 °C vs. 8.98 °C, 48.2% vs. 56.4%, and pH 4.83 vs. pH 4.60, respectively, all P < 0.05). These results indicated that FTW can influence tree biomass C dynamics, organic C input and output, and soil environmental factors such as soil temperature, soil water content and soil pH, while soil C dynamics such as soil CO₂ efflux rates and soil organic C concentration were little affected by FTW in a red pine stand.     

Forest regeneration in artificial gaps twelve years after canopy opening in Acre State Western Amazon

The main objectives were to study the effect of gap size and canopy openness on the natural regeneration dynamics considering the parameters of sapling growth, recruitment, mortality, density, species composition and above-ground biomass accumulation. The study was carried out in 32 artificial gaps with sizes varying from 100 to 1200 m² and canopy openness from 10 to 45%, from the second to the twelfth year after gap creation. The gap size was measured using the vertical projection of the tree crowns on the ground (Brokaw's definition), and the canopy openness measurement by hemispherical photography. In the first five years, mean sapling growth (0.54 cm year⁻¹), mortality (3.9% year⁻¹) and AGB (26.2 Mg ha⁻¹ or 8.7 Mg ha⁻¹ year⁻¹) were significantly higher in the gaps than in the forest understorey (0.17 cm year⁻¹, 1.5% year⁻¹ and −0.59 Mg ha⁻¹ year⁻¹ respectively) and positively correlated with gap size and canopy openness. In the same period, recruitment was also significantly higher in the gaps (5.8% year⁻¹) than in the forest understorey (0.4% year⁻¹) but decreased with gap size and negatively correlated with canopy openness. In the first five years, the relative density of pioneer species was higher in the gaps but not significantly correlated with gap size or canopy openness. AGB increased linearly since canopy opening, and twelve years after gap creation it was still higher in larger (121.2 Mg ha⁻¹ or 10.1 Mg ha⁻¹ year⁻¹) rather than smaller (62.5 ha⁻¹ or 5.2 ha⁻¹ year⁻¹) gaps. Twelve years after gap creation there were no significant differences in the parameters of sapling growth, recruitment, and mortality which could be attributed to the original gap size and canopy openness.     

Above-ground biomass dynamics after reduced-impact logging in the Eastern Amazon

Changes in above-ground biomass (AGB) of 17 1 ha logged plots of terra firme rain forest in the eastern Amazon (Brazil, Paragominas) were monitored for four years (2004–2008) after reduced-impact logging. Over the same time period, we also monitored two 0.5 ha plots in adjacent unlogged forest. While AGB in the control plots changed little over the observation period (increased on average 1.4 Mg ha⁻¹), logging resulted in immediate reductions in ABG that averaged 94.5 Mg ha⁻¹ (±42.0), which represented 23% of the 410 Mg ha⁻¹ (±64.9) present just prior to harvesting. Felled trees (dbh > 55 cm) accounted for 73% (±15) of these immediate losses but only 18.9 Mg ha⁻¹ (±8.1) of biomass was removed in the extracted logs. During the first year after logging, the annual AGB balance (annual AGB gain by recruitment and growth − annual AGB loss by mortality) remained negative (−31.1 Mg ha⁻¹ year⁻¹; ±16.7), mainly due to continued high mortality rates of damaged trees. During the following three years (2005–2008), average net AGB accumulation in the logged plots was 2.6 Mg ha⁻¹ year⁻¹ (±4.6). Post-logging biomass recovery was mostly through growth (4.3 ± 1.5 Mg ha⁻¹ year⁻¹ for 2004–2005 and 6.8 ± 0.9 Mg ha⁻¹ year⁻¹ for 2005–2008), particularly of large trees. In contrast, tree recruitment contributed little to the observed increases in AGB (1.1 ± 0.6 Mg ha⁻¹ year⁻¹ for 2004–2005 and 3.1 ± 1.3 Mg ha⁻¹ year⁻¹ for 2005–2008). Plots with the lowest residual basal area after logging generally continued to lose more large trees (dbh ≥70 cm), and consequently showed the greatest AGB losses and the slowest overall AGB gains. If 100% AGB recovery is desired and the 30-year minimum cutting cycle defined by Brazilian law is adhered to, current logging intensities (6 trees ha⁻¹) need to be reduced by 40–50%. Such a reduction in logging intensity will reduce financial incomes to loggers, but might be compensated for by the payment of environmental services through the proposed REDD (reduced emissions from deforestation and forest degradation) mechanism of the United Nations Framework Convention on Climate Change.     

Deforestation and fragmentation of Chaco dry forest in NW Argentina (1972–2007)

The sustained increase in the global food demand has favoured the recent expansion of industrial agriculture into neotropical dry forest ecosystems, with resulting changes in their extent and spatial configuration. Based on Landsat satellite images, we analyzed changes in forest cover and landscape configuration over an area of 600 by 100 km located in NW Argentina (Tucumán and Salta provinces) in four periods between 1972 and 2007. The study area, one of the most active deforestation frontiers of South American dry forest, was divided into six relatively homogeneous sectors in terms of land property structure and biophysical characteristics. During the study period 1.4 millions hectares of dry forest were cleared. Deforestation started in the 1970s as a result of technological changes and increasing rainfall; continued (with spatial and temporal fluctuations) during the 1980s and 1990s in association to the sustained global demand of soybean, and was accelerated (to ca. 100,000 ha year⁻¹) between 2001 and 2007 following the global increase in commodity prices, and the national peso devaluation. We described the landscape structure using eight landscape indices, summarized as one synthesis value: the Euclidian distance across the seven dimensions from a theoretical non-fragmented situation. In areas with soil limitations deforestation resulted in relatively stable, highly fragmented landscapes. In contrast, the sites with regional coarser-scale limitations (rainfall), deforestation produced a less fragmented landscape where agriculture concentrates in sites with high rainfall. Sites with no limitations for agriculture tend to a largely deforested landscape with few small and poorly connected forest patches. The land properties size seems to influence some indices of fragmentation, but the synthetic index of fragmentation suggests an overall convergence of fragmentation patterns towards a similar configuration across different biophysical and land tenure conditions.     

Soil–atmosphere CO2, CH4 and N2O fluxes in boreal forestry-drained peatlands

Greenhouse gas emissions from managed peatlands are annually reported to the UNFCCC. For the estimation of greenhouse gas (GHG) balances on a country-wide basis, it is necessary to know how soil–atmosphere fluxes are associated with variables that are available for spatial upscaling. We measured momentary soil–atmosphere CO₂ (heterotrophic and total soil respiration), CH₄ and N₂O fluxes at 68 forestry-drained peatland sites in Finland over two growing seasons. We estimated annual CO₂ effluxes for the sites using site-specific temperature regressions and simulations in half-hourly time steps. Annual CH₄ and N₂O fluxes were interpolated from the measurements. We then tested how well climate and site variables derived from forest inventory results and weather statistics could be used to explain between-site variation in the annual fluxes. The estimated annual CO₂ effluxes ranged from 1165 to 4437 g m⁻² year⁻¹ (total soil respiration) and from 534 to 2455 g m⁻² year⁻¹ (heterotrophic soil respiration). Means of 95% confidence intervals were ±12% of total and ±22% of heterotrophic soil respiration. Estimated annual CO₂ efflux was strongly correlated with soil respiration at the reference temperature (10 °C) and with summer mean air temperature. Temperature sensitivity had little effect on the estimated annual fluxes. Models with tree stand stem volume, site type and summer mean air temperature as independent variables explained 56% of total and 57% of heterotrophic annual CO₂ effluxes. Adding summer mean water table depth to the models raised the explanatory power to 66% and 64% respectively. Most of the sites were small CH₄ sinks and N₂O sources. The interpolated annual CH₄ flux (range: −0.97 to 12.50 g m⁻² year⁻¹) was best explained by summer mean water table depth (r² = 64%) and rather weakly by tree stand stem volume (r² = 22%) and mire vegetation cover (r² = 15%). N₂O flux (range: −0.03 to 0.92 g m⁻² year⁻¹) was best explained by peat CN ratio (r² = 35%). Site type explained 13% of annual N₂O flux. We suggest that water table depth should be measured in national land-use inventories for improving the estimation of country-level GHG fluxes for peatlands.     

Planting mahogany in canopy gaps created by commercial harvesting

Attempts at natural forest management of mahogany (Swietenia macrophylla King) have so far met with limited success, whilst many plantations are beset by the shoot borer Hypsipyla spp. In this paper we present preliminary results of an approach to enrichment planting that aims to balance economic returns (rapid growth and good silvicultural form) with intervention costs and changes to forest structure. Mahogany seedlings were planted in gaps created by selective timber harvesting and that ranged in vertical projected area from 91 to 542 m² (mean = 257 m²). Seedlings grew within the matrix of gap regrowth, with limited control of competing vegetation. Sixty-one percent of seedlings had survived by 4.4 years (equivalent to an annual mortality rate of 10.5% year⁻¹), and had reached a mean height of 4.5 m. Stocking levels of mahogany were similar to that of naturally regenerated commercial species in unplanted gaps of the same age, but mahogany seedlings were significantly taller. The incidence of shoot borer attack on mahogany stems was relatively low (54.7%), but, more importantly, most damaged stems (58%) responded by producing a single replacement leader. The cost of the proposed methodology (US$ 94 per gap over 4.4 years) was low compared to the high value of mahogany timber relative to other species in the forest. The implications of planting mahogany in gaps for forest management and the potential benefits to conservation of the species are considered.     

Forest dynamics after selective timber harvesting in Papua New Guinea

Forest dynamics after timber harvesting is a major issue for tropical forest managers and communities. Timber harvesting provides income to communities and governments and resources to industry but it has also been identified as a potential contributor to deforestation and degradation of tropical forests. In Papua New Guinea (PNG) harvesting is primarily occurring in accessible primary forests however, the fate of these forests under current harvesting practices is poorly understood.In this study we investigated the impacts of selective harvesting on stand structure, growth and dynamics, recovery and degradation, and species diversity. We also assessed the impacts of forest fire after the 1997-98 El Nino on basal area (BA) growth and mortality rates of natural tropical forests in PNG. For this study we used data from 118 (105 in selectively harvested and 13 in un-harvested forest), one-hectare permanent sample plots distributed across the country and measured for over 15 years by the PNG Forest Research Institute (PNGFRI). We analysed data from 84 of these plots in harvested forest to examine temporal trends in stand condition following harvesting. Mortality rates were investigated in 10 of the 21 plots in harvested forest that were burned during the 1997-98 El Nino drought with sufficient data for analyses. We tested a model developed in Queensland tropical forests to determine whether or not a critical threshold residual BA existed for the recovery of harvested tropical forests in PNG. Results from a logarithmic regression analysis of the relationship between starting BA (BA at first census) and stand BA increment after selective harvesting showed a positive increase in BA growth (r² = 0.74, p < 0.05). However, there was no critical threshold in residual BA that determined whether a harvested forest was likely to degrade or recover BA growth after harvesting. Our analyses suggested that the response to harvesting was variable, with the majority of un-burned plots (75%) showing an increase in BA and remainder a decrease. Average BA of selectively-harvested tropical forests was about 17 m² ha⁻¹ ± 4.17 (SD). Average annual increment in BA across the 84 un-burned plots was 0.17 m² ha⁻¹ year⁻¹ ± 0.62 (SD). Thus these forests generally show capacity to recover after selective harvesting even when the residual BA is low. A proportion of the BA increment is made up of non-commercial pioneer species that originate in significant gaps after harvesting. On burned plots, BA is affected by high mortality rates. The fate of these forests will depend on the degree of future harvesting, potential conversion to agriculture and the impact of fire and other disturbances.     

Water use by Tabor and Kermes oaks growing in their respective habitats in the Lower Galilee region of Israel

We estimated water use by the two main oak species of the Lower Galilee region of Israel—Tabor (Quercus ithaburensis) and Kermes (Quercus calliprinos)—to develop management options for climate-change scenarios. The trees were studied in their typical phytosociological associations on different bedrock formations at two sites with the same climatic conditions. Using the heat-pulse method, sap flow velocity was measured in eight trunks (trees) of each species during a number of periods in 2001, 2002 and 2003. Hourly sap flux was integrated to daily transpiration per tree and up-scaled to transpiration at the forest canopy level. The annual courses of daytime transpiration rate were estimated using fitted functions, and annual totals were calculated. Sap flow velocity was higher in Tabor than in Kermes oak, and it was highest in the youngest xylem, declining with depth into the older xylem. Average daytime transpiration rate was 67.9 ± 4.9 l tree⁻¹ d⁻¹, or 0.95 ± 0.07 mm d⁻¹, for Tabor oak, and 22.0 ± 1.7 l tree⁻¹d⁻¹, or 0.73 ± 0.05 mm d⁻¹, for Kermes oak. Differences between the two oak species in their forest canopy transpiration rates occurred mainly between the end of April and the beginning of October. Annual daytime transpiration was estimated to be 244 mm year⁻¹ for Tabor oak and 213 mm year⁻¹ for Kermes oak. Adding nocturnal water fluxes, estimated to be 20% of the daytime transpiration, resulted in total annual transpiration of 293 and 256 mm year⁻¹ by Tabor and Kermes oaks, respectively. These amounts constituted 51% and 44%, respectively, of the 578 mm year⁻¹ average annual rainfall in the region. The two species differed in their root morphology. Tabor oak roots did not penetrate the bedrock but were concentrated along the soil–rock interface within soil pockets. In contrast, the root system of Kermes oak grew deeper via fissures and crevices in the bedrock system and achieved direct contact with the deeper bedrock layers. Despite differences between the two sites in soil–bedrock lithological properties, and differences in the woody structure, annual water use by the two forest types was fairly similar. Because stocking density of the Tabor oak forests is strongly related to bedrock characteristics, thinning as a management tool will not change partitioning of the rainfall between different soil pockets, and hence soil water availability to the trees. In contrast, thinning of Kermes oak forests is expected to raise water availability to the remaining trees.     

Recovery process of a mountain forest after shifting cultivation in Northwestern Vietnam

The recovery process of fallow stands in the mountainous region of Northwestern Vietnam was studied, based on a chronosequence of 1–26-year-old secondary forests after intensive shifting cultivation. The number of species present in a 26-year-old secondary forest attained 49% of the 72 species present in an old-growth forest. Total stem density decreased gradually from 172,500 ha⁻¹ in a 3-year-old forest to 24,600 ha⁻¹ in the 26-year-old stand, but stem density of larger trees (diameter at breast height (D) ≥ 5 cm) increased from 60 ha⁻¹ in a 7-year-old to 960 ha⁻¹ in the 26-year-old forests, which was similar to that of an old-growth forest. Annual biomass increment of the 26-year-old stand was 4.2 Mg ha⁻¹ year⁻¹. A saturation curve was fitted to biomass accumulation in secondary forests. After an estimated time of 60 years, a secondary forest can achieve 80% of the biomass of old-growth forests (240 Mg ha⁻¹). Species diversity expressed by Shannon Index shows that it takes 60 years for a secondary forest in fallow to achieve a plant species diversity similar to that of old-growth forests.     

Net primary productivity of Bruguiera parviflora (Wight & Arn.) dominated mangrove forest at Kuala Selangor,Malaysia

The net primary productivity of Bruguiera parviflora dominated mangrove forest at Kuala Selangor, Malaysia was estimated from the average yearly biomass increment and litter production. The average yearly biomass increment in saplings and trees was 0.58 and 16.51 t ha⁻¹, respectively, and the annual amount of total litter production was 10.35 t ha⁻¹. The biomass increment in saplings and trees was not significantly different (t-test, p > 0.05) in 2 successive years and the estimated net primary productivity was 27.44 t ha⁻¹ year⁻¹. The ratio (2.65:1) of net primary productivity and litterfall suggests that this mangrove forest is at a juvenile stage.     

Landscape dynamics of Amazonian deforestation between 1984 and 2002 in central Rondônia,Brazil: assessment and future scenarios

Central Rondônia is one of the most deforested regions in the Brazilian Amazon and contains areas at different stages of degradation forming a gradient from mature forest to highly urbanized and built-up areas. Regional data from satellite imagery are available from the 1980s, but apart from studies that quantify deforestation, the broad-scale landscape dynamics of Rondônia have not been examined well. This paper assesses the landscape changes between 1984 and 2002 in a watershed located in the central region of the state of Rondônia, Brazil, which has undergone systematic and rapid deforestation due to introduction of pasture that began in the 1970s. Bi-annual Landsat TM/ETM+ images from 1984 to 2002 were classified into three broad land cover types: mature forest, secondary forest, and pasture, resulting in a time series of land-use/land-cover maps. Landscape changes were evaluated by computing a cross tabulation between years, transition rates, and landscape metrics related to size, density, connectivity, configuration, and deforested patches distribution related to patch size and spatial proximity to roads and old pastures. Transition probability functions were fitted to the temporal series to predict land-use changes for the next 10 years, for three different scenarios: (1) continued land-use change; (2) eliminating clear-cutting and selective logging; and (3) eliminating clear-cutting, selective logging, and secondary vegetation clearing. Current land-use transitions cannot be sustained beyond the next 10–15 years. A more sustainable scenario for the region requires a major reduction of deforestation activities, implementing the “permanent preservation area” policy along riversides, and controlling land-use transitions at balanced levels. We recommend that forest managers in regions facing similar deforestation pressures should strive to maintain at least 35% mature or primary forest, because landscape fragmentation proceeds rapidly below this critical threshold.     

Annual allowable cut for merchantable woody species in a community managed forest in western Kenya

Changes in stand density, basal area, off-take and annual increment were determined from 18 permanent sample plots established in 1997 in Got Ramogi Forest in western Kenya. The plots were assessed in 2003 and 2008. A total of 824 stems ?1.5 m in height were recorded from 43 woody species. Key merchantable woody species comprised 20% of the woody species and 67% of the overall stem density. There was a significant reduction in the overall stand density and in the stem density of key merchantable woody species, but not among other woody species between 1997 and 2008. The basal area decreased significantly among key merchantable woody species, but not for the overall forest. The basal area decreased from 22.6 to 9.7 m² ha⁻¹ for key merchantable woody species. The stand volume of key merchantable woody species decreased from 156 m³ ha⁻¹ in 1997 to 61.7 m³ ha⁻¹ in 2008. The mean annual off-take declined from 10.3 m³ ha⁻¹ year⁻¹ between 1997 and 2003 to 9.1 m³ ha⁻¹ year⁻¹ between 2003 and 2008, while the mean annual increment increased from 2.9 to 3.3 m³ ha⁻¹ year⁻¹. It was predicted that forest recovery would surpass the 1997 stand volume of 156 m³ ha⁻¹ if off-take levels between 10% and 90% of the mean annual increment were adopted. We settled on an annual allowable cut of 80% of the mean annual increment as a compromise between consumptive and conservation interests. We identified over-harvesting as the main cause of the reduction in stem density among key merchantable woody species. A management plan with compartment registers indicating the diversity, abundance and distribution of each woody species was recommended to guide their utilization and monitor their population dynamics.     

Forest stand dynamics and sudden oak death: Mortality in mixed-evergreen forests dominated by coast live oak

Sudden oak death (SOD), caused by the recently discovered non-native invasive pathogen, Phytophthora ramorum, has already killed tens of thousands of native coast live oak and tanoak trees in California. Little is known of potential short and long term impacts of this novel plant–pathogen interaction on forest structure and composition. Coast live oak (Quercus agrifolia) and bay laurel (Umbellularia californica) form mixed-evergreen forests along the northern California coast. This study measured tree mortality over a gradient of disease in three time periods. Direct measurements of current mortality were taken during 2004, representing a point-in-time estimate of present and ongoing mortality. Past stand conditions, c. 1994, were estimated using a stand reconstruction technique. Future stand conditions, c. 2014, were calculated by assuming that, given a lack of host resistance, live trees showing signs of the disease in 2004 would die. Results indicate that coast live oaks died at a rate of 4.4–5.5% year⁻¹ between 1994 and 2004 in highly impacted sites, compared with a background rate of 0.49% year⁻¹, a ten-fold increase in mortality. From 2004 to 2014, mortality rates in the same sites were 0.8–2.6% year⁻¹. Over the entire period, in highly impacted sites, a 59–70% loss of coast live oak basal area was predicted, and coast live oak decreased from 60% to 40% of total stand basal area, while bay laurel increased from 22% to 37%. Future stand structures will likely have greater proportions of bay laurel relative to coast live oak.