Initial responses of woody vegetation, water quality, and soils to harvesting intensity in a Texas bottomland hardwood ecosystem

Sustainable management of bottomland hardwood forest ecosystems requires a knowledge of responses to management impacts, including timber harvesting. The effects of clearcutting and partial cutting on woody vegetation regeneration dynamics, surface and groundwater quality, soil physical properties, and soil respiration were tested in a bottomland hardwood ecosystem in southeastern Texas, USA, through comparison with non-cut control areas. Overstory removal only slightly affected composition of woody vegetation regeneration 1 year after harvesting compared with pre-harvest composition. Initial composition in both cutting treatments appeared to be the strongest determinant of post-harvest composition, at least for the first year after harvesting. There were few significant differences in groundwater properties when harvesting treatments were compared with control areas during a 17-month period following harvest. Turbidity, temperature, electrical conductivity, dissolved O₂, NH₄-N, NO₃-N, and PO₄-P of streamwater did not vary significantly among treatments. Slight decreases in total and macroporosity were observed in association with higher bulk densities at 0–5 cm depth in the clearcut and partial cut treatments. Saturated hydraulic conductivity values did not decline significantly with treatment intensity. No significant differences among treatments in measured soil physical properties were observed at 5–10 cm depth. Although in situ soil respiration increased with harvest intensity, treatment had no significant effect on mineral soil respiration. In summary, most variables showed only slight response to harvesting, thereby indicating that harvesting practices can be conducted with minimal initial impacts on measured response variables.     

Thinning, burning, and thin-burn fuel treatment effects on soil properties in a Sierra Nevada mixed-conifer forest

More than a century of fire exclusion and past timber management practices in many Sierra Nevada mixed-conifer forests have led to increased stand densities and fuel accumulation, with a corresponding risk of large, high severity wildfires. To reduce hazardous fuel accumulations and restore the health and natural processes of forest ecosystems, fuel management programs often employ thinning and prescribed fire treatments, both alone and in combination. We evaluated forest floor and mineral soil chemical and physical characteristics following these treatments in a managed Sierra Nevada mixed-conifer forest using a fully replicated study design with four separate treatments: THIN, BURN, THIN + BURN, and an untreated CONTROL. Compared to the CONTROL, the BURN and THIN + BURN treatments consumed a large amount of the forest floor, reducing the mass and depth by more than 80%. These treatments reduced the forest floor C and N pools by more than 85%, resulting in reductions of 25 Mg C ha⁻¹ and more than 700 kg N ha⁻¹ from the forest floor. Despite these large losses from the organic horizons, no significant differences in mineral soil total C and N pools were detected among treatments. Compared with the CONTROL and THIN treatments, the BURN and THIN + BURN significantly increased the mineral soil NO₃-N concentration, pool of inorganic N, pH, and exposed bare soil. The THIN + BURN treatment significantly increased the concentrations of NH₄-N and exchangeable Ca relative to the CONTROL. No significant differences in the net rates of nitrification, N mineralization, or bulk density were detected among the four treatments. The BURN treatment reduced mineral soil C concentration and CEC, while the THIN + BURN treatment had the greatest increase in inorganic N. Fire effects on soil pH and inorganic N were moderated in skid trails due to reduced fuel continuity and consumption. In light of the current management emphasis on hazardous fuels reduction, we recommend that researchers investigating fire effects in harvested stands include skid trail influences in their study design.     

The use of ground vegetation and humus type as indicators of soil nutrient regime for an ecological site classification of British forests

In many countries ground vegetation and humus type are used as indicators of forest soil quality, especially nutrient regime. This paper reports the development of such methods for use in British forests, within a new Ecological Site Classification combining climate, soil moisture regime and soil nutrient regime.

To develop a field assessment method for soil nutrient regime, a study was made of soil chemistry, humus type and ground vegetation in British forests. Sites were selected in both mature plantations and semi-natural woodlands. Soil and humus profiles were described and the soil was sampled volumetrically for later chemical analysis. Vascular ground vegetation was recorded in quadrats by species cover fraction, and classified according to the existing British National Vegetation Classification. Soils were analyzed for a number of chemical variables. Vegetation data were treated by application of the species indicator values for soil reaction (R) and soil nitrogen (N), as proposed by Ellenberg [Vegetation Ecology of Central Europe, 4th Edition. Cambridge University Press, Cambridge]. Site mean indicator values mR and mN (weighted by species cover fraction) were then calculated. Multivariate statistical analysis techniques were applied to both the soil chemical and the vegetation sample data.

Soil nutrient regime was shown to be a composite gradient of several soil chemical variables, of which the pH value and the availability of mineral (especially nitrate) nitrogen, and of calcium, were of particular importance. The species composition of the ground vegetation was related to position on this soil nutrient gradient. The vegetation: soil nutrient correlation using the site mean Ellenberg values was satisfactory (r=0.89), but was improved by using indicator values generated from within the present data. The occurrence of the major humus types (mor, moder and mull) is broadly related to soil nutrient regime defined in this way. Both ground vegetation and humus type can therefore be used as soil nutrient indicators in British forests.

A division of the soil nutrient gradient into five classes (Very Poor, Poor, Medium, Rich and Very Rich) is proposed. Future sampling work may lead to the definition of an additional class of soils with carbonate nutrient regimes. The Ecological Site Classification will provide forest managers in Great Britain with an improved basis for the selection of tree species for planting, and the adoption of silvicultural methods best suited to the site.     

Nitrate dynamics after clear felling monitored by in vivo nitrate reductase activity (NRA) and natural N abundance of Deschampsia flexuosa (L.) Trin.

The N dynamics following clear felling, focusing on NO₃⁻ turnover, were studied at four forested sites in southern Sweden. Two different methods were used to study N availability: (i) an in vivo nitrate reductase activity (NRA) bioassay and (ii) measurements of natural abundance of stable N isotopes in leaves of the grass species Deschampsia flexuosa, and in organic soil horizons. At each of the four sites, six plots were established and each year, for 5 consecutive years (1989–1993), one plot per site was felled. Thus, in 1993 there were five plots with different ages since clear felling and one control (closed forest) plot at each site. NRA was analyzed three times annually during the years 1989–1993. Samples for grass and soil analysis of δ¹⁵N, total N and soil pH were taken in 1993 only. NRA rapidly increased after the felling and remained high throughout the studied period. This suggests that there was an increased pool of plant-available soil NO₃⁻ more than 5 years after clear felling. Despite differences in site productivity and N deposition between the four sites, no significant differences in NRA were found between the sites. There were also rapid changes in δ¹⁵N in leaves of D. flexuosa, coinciding with the increases in NRA, during the first 3 years after felling. In contrast to NRA, shoot δ¹⁵N decreased 3–4 years after the felling at three out of four sites. Variations in the δ¹⁵N figures between sites may have been largely due to between-site differences in field-layer retention of N. At two of the sites, where NO₃⁻ leaching was also measured, a correlation was found between the NO₃⁻ concentration in the water and the difference in δ¹⁵N between D. flexuosa leaves from felled and closed forest plots. The data presented here suggest that NO₃⁻ leakage after clear felling is a rapid process, which is influenced by the development of field-layer biomass after the felling. Furthermore, losses of NO₃⁻ through leaching rapidly change the natural abundance of the plant available N pools in the soil.     

Forest soils and carbon sequestration

Soils in equilibrium with a natural forest ecosystem have high carbon (C) density. The ratio of soil:vegetation C density increases with latitude. Land use change, particularly conversion to agricultural ecosystems, depletes the soil C stock. Thus, degraded agricultural soils have lower soil organic carbon (SOC) stock than their potential capacity. Consequently, afforestation of agricultural soils and management of forest plantations can enhance SOC stock through C sequestration. The rate of SOC sequestration, and the magnitude and quality of soil C stock depend on the complex interaction between climate, soils, tree species and management, and chemical composition of the litter as determined by the dominant tree species. Increasing production of forest biomass per se may not necessarily increase the SOC stocks. Fire, natural or managed, is an important perturbation that can affect soil C stock for a long period after the event. The soil C stock can be greatly enhanced by a careful site preparation, adequate soil drainage, growing species with a high NPP, applying N and micronutrients (Fe) as fertilizers or biosolids, and conserving soil and water resources. Climate change may also stimulate forest growth by enhancing availability of mineral N and through the CO₂ fertilization effect, which may partly compensate release of soil C in response to warming. There are significant advances in measurement of soil C stock and fluxes, and scaling of C stock from pedon/plot scale to regional and national scales. Soil C sequestration in boreal and temperate forests may be an important strategy to ameliorate changes in atmospheric chemistry.     

Nitrogen-fertilization impacts on carbon sequestration and flux in managed coastal Douglas-fir stands of the Pacific Northwest

We examined whether N-fertilization and soil origin of Douglas-fir [Psuedotsuga menziesii (Mirb.) Franco] stands in western Washington state could affect C sequestration in both the tree biomass and in soils, as well as the flux of dissolved organic carbon (DOC) through the soil profile. This study utilized four forest sites that were initially established between 1972 and 1980 as part of Regional Forest Nutrition Research Project (RFNRP). Two of the soils were derived from coarse-textured glacial outwash and two from finer-textured volcanic-source material, primarily tephra, both common soil types for forestry in the region. Between 1972 and 1996 fertilized sites received either three or four additions of 224 kg N ha⁻¹ as urea (672–896 kg N ha⁻¹ total). Due to enhanced tree growth, the N-fertilized sites (161 Mg C ha⁻¹) had an average of 20% more C in the tree biomass compared to unfertilized sites (135 Mg C ha⁻¹). Overall, N-fertilized soils (260 Mg C ha⁻¹) had 48% more soil C compared to unfertilized soils (175 Mg C ha⁻¹). The finer-textured volcanic-origin soils (348 Mg C ha⁻¹) had 299% more C than glacial outwash soils (87.2 Mg C ha⁻¹), independent of N-fertilization. Soil-solution DOC collected by lysimeters also appeared to be higher in N-fertilized, upper soil horizons compared to unfertilized controls but it was unclear what fraction of the difference was lost from decomposition or contributed to deep-profile soil C by leaching and adsorption. When soil, understory vegetation and live-tree C compartments are pooled and compared by treatment, N-fertilized plots had an average of 110 Mg C ha⁻¹ more than unfertilized controls. These results indicate these sites generally responded to N-fertilization with increased C sequestration, but differences in stand and soil response to N-fertilization might be partially explained by soil origin and texture.     

Effects of strip wise tillage in combination with liming on chemical and physical properties of acidic spruce forest soils after clear cutting

We investigated the initial effects of strip wise soil loosening (0–35 cm depth) on soil chemical and physical parameters by using a deeply working rotary cultivator in combination with liming and mixing of the dolomite with the soil material of acidic forests. The investigations took place 8 months after the treatment. pH values and contents of exchangeable Ca and Mg increased significantly at the tilled depth whereas the content of exchangeable Al and easily soluble P decreased. The rate of mineralisation increased at this depth which was shown by a loss of Corg, Ntot and short-term loss of NO₃-N. The treatment led to a mobilisation of Mn at the tilled depth. However, the content of exchangeable Pb decreased due to an increased pH value. Below the tillage depth of 35 cm only partly significant changes of exchangeable Mn and NO₃-N were found. The total porosity and bulk density at 10–15 and 40–45 cm depths were not significantly different from those in the control plot, but the rate of infiltration increased significantly.     

Nitrogen deposition in Casuarina equisetifolia (Forst.) plantation stands in the dry tropics of Sonbhadra, India

Periodic variations in the concentration, deposition and canopy impact of different forms of N on annual N deposition through rainfall, throughfall and stemflow in 5 and 8 year old stands of Casuarina equisetifolia were studied. Throughfall and stemflow ranged from 70 to 76% and 5–6% of annual precipitation respectively. The total N deposition by rainfall was 11.1 kg ha⁻¹ year⁻¹, and by throughfall was 13.6 kg ha⁻¹ year⁻¹ and 16.5 kg ha⁻¹ year⁻¹ in 5-year-old and 8-year old plantations, respectively. The quantities of N deposited through stemflow in the two plantations were nearly identical, accounting for 1.6 kg ha⁻¹ year⁻¹. Observations of the monthly deposition of NH₄,N, NO₃-N, Kjeldahl-N and organic-N revealed that maximum deposition occurred in July and the minimum in September. Organic-N deposition was 17% less (5-year) than the rainwater content. Net deposition of N, as an effect of canopy, was 7–8.7 kg ha⁻¹ year⁻¹, which was added directly to the available nutrient pool of soil.     

Effects of forest management on soil C and N storage: meta analysis

The effects of forest management on soil carbon (C) and nitrogen (N) are important to understand not only because these are often master variables determining soil fertility but also because of the role of soils as a source or sink for C on a global scale. This paper reviews the literature on forest management effects on soil C and N and reports the results of a meta analysis of these data. The meta analysis showed that forest harvesting, on average, had little or no effect on soil C and N. Significant effects of harvest type and species were noted, with sawlog harvesting causing increases (+18%) in soil C and N and whole-tree harvesting causing decreases (−6%). The positive effect of sawlog harvesting appeared to be restricted to coniferous species. Fire resulted in no significant overall effects of fire on either C or N (when categories were combined); but there was a significant effect of time since fire, with an increase in both soil C and N after 10 years (compared to controls). Significant differences among fire treatments were found, with the counterintuitive result of lower soil C following prescribed fire and higher soil C following wildfire. The latter is attributed to the sequestration of charcoal and recalcitrant, hydrophobic organic matter and to the effects of naturally invading, post-fire, N-fixing vegetation. Both fertilization and N-fixing vegetation caused marked overall increases in soil C and N.     

The role of fire and nutrient loss in the genesis of the forest soils of Tasmania and southern New Zealand

The dominant soil patterns in forested or previously forested landscapes in southern New Zealand and Tasmania are described. Soil properties on adjacent sunny and shady aspects in hill country of the South Island of New Zealand are compared to soil properties under adjacent ‘dry’ and ‘wet’ eucalypt forest in Tasmania.

A soil contrast index or SCI is defined for comparing soil contrasts on parent materials of different absolute nutrient contents. Three soil groups are defined using the SCI. Group 1 soil pairs are stable New Zealand soils in which exchangeable Ca + Mg + K values are higher on drier sunny aspects than on moister shady aspects. Group 2 soil pairs are New Zealand soils in which soils on sunny aspects display evidence of topsoil erosion by wind; consequently some soil pairs on dry (sunny) aspects have lower levels of exchangeable Ca + Mg + K than soils on moister (shady) aspects. Group 3 soil pairs are Tasmanian. Soils on drier sites (under dry eucalypt forest) invariably have lower exchangeable Ca + Mg + K values than soils on moister sites (under wet eucalypt forest), which is the reverse of the pattern in SCI Group 1 soils in New Zealand.

Except on clay-rich parent materials, Tasmanian soils under dry forest generally have texture-contrast profiles and a mean C/N ratio in topsoils (A1 horizons) of 29. Soils under wet forest generally have uniform or gradational texture profiles and a mean topsoil C/N ratio of 15. The texture-contrast soils show strong clay eluviation with sand or sandy loam textures in upper horizons and clayey textures in lower horizons. However, in New Zealand texture-contrast soils are all but absent, and do not occur in the previously forested areas described in this paper. Topsoils (Ah horizons and soils sampled to 7.5 cm depth) in New Zealand areas sampled in this study have a mean C/N ratio of 15, regardless of whether they occur on sunny or shady aspects.

We propose that the frequency and spatial occurrence of fire are the dominant processes causing: (1) the marked difference in levels of nutrients and different topsoil C/N ratios in soils of Tasmania; (2) the development of texture-contrast soils under dry forests in Tasmania; and (3) the difference between soil patterns in New Zealand and Tasmania. Fire depletes nutrients in forests by causing losses to the atmosphere, losses by runoff, and losses by leaching. Nutrient loss by fire encourages fire-tolerant vegetation adapted to lower soil nutrient status, so frequent fire is a feedback mechanism that causes progressive soil nutrient depletion. By destroying organic matter and diminishing organic matter supply to the soil surface fire inhibits clay–organic matter linkages and soil faunal mixing and promotes clay eluviation. Fire frequency is likely to have increased markedly with the arrival of humans at ca. 34 000 years B.P. in Tasmania and ca. 800 years B.P. in New Zealand. We argue that texture-contrast soils have not formed in New Zealand because of the short history of frequent fires in that country. A corollary of this conclusion is that texture-contrast soils in Tasmania are, at least in part, anthropogenic in origin.     

Changes in soil acidity and organic matter following the establishment of conifers on former grassland in New Zealand

Effects of a land use change from grassland to coniferous plantation forestry (Pseudotsuga menzieii [Douglas fir]; Pinus radiata [radiata pine]) on soil acidity and organic matter were assessed at two sites in New Zealand. The sites differed with respect to soils, climate, vegetation cover and type, relative maturity and management of the forest stands. Results obtained at the different sites were, therefore, not directly comparable, although they represented a comparison of a similar change in land use and some overall trends were evident. The change from grassland to conifers decreased levels of organic carbon, total nitrogen and exchangeable cations and increased exchangeable acidity in the upper 20–30 cm of soil. Exchangeable aluminium and exchangeable acidity were more sensitive measures of the effects of afforestation on soil acidity than pH.     

拖拉机集材对林地土壤的影响

本文从环境保护的观点出发.分析了拖拉机集材造成的林地土壤压实、水土流失。并对集材前后迹地土壤营养成份流换情况进行对比。结果表明,拖拉机集材对森林生态环境破坏较为严重。不符合生态效益对林业生产的要求。     

Sitka spruce site index in response to varying soil moisture and nutrients in three different climate regions in Ireland

We examined the relationships between Sitka spruce (Picea sitchensis (Bong.) Carr.) site index, windspeed, former land use, soil moisture and soil nutrients with a view to identifying factors limiting the growth potential of the species in three climate regions in Ireland. We selected plantations covering three climate regions (delineated on the basis of ‘growing season’ balance of precipitation and potential evapotranspiration; representing dry, moist and wet climate regions) located on sites representing the range of soil moisture regimes (SMR), soil nutrient regimes (SNR), and land use types found throughout Ireland. Site index of Sitka spruce varied among climate region, with significantly lower site index associated with the wet climate region mainly due to the deterioration in edaphic conditions and adverse climatic conditions. The effect of edaphic variables (SMR, SNR) on site index was consistent across climate regions, site index increasing with increasing SNR, and decreasing with excess moisture or moisture deficit. Site index reached a maximum on fresh/very rich sites in the dry and moist climate regions and on moist/rich sites in the wet climate region. In the dry climate region, water supply (SMR) was the most important variable regulating growth, in wetter windier climates nutrient supply (SNR) was the most important factor, accounting for 69% of the variation in site index. The study has allowed region-specific recommendations to be made for successful plantation establishment in Ireland and for countries with similar climatic regions.     

Effect of clear-cutting and site preparation on the level and quality of groundwater in some headwater catchments in eastern Finland

The impact of forest management (clear-cutting and site preparation) on stream hydrology has been studied in five small catchments in eastern Finland from 1991 and on groundwater levels and quality from 1994. The period 1992–1996 was a calibration period and in the autumn of 1996, 10% and 30% of the area of two of these catchments were clear-cut according to the forest management plan. Regeneration was carried out by disc-plowing in the autumn of 1998 and planting with Scots pine (Pinus sylvestris) and Norway spruce (Picea abies) seedlings in the spring of 1999. The depth and quality of groundwater was monitored with four to nine groundwater wells installed in each catchment. There were 32 wells in all, 16 on upland mineral soils and 16 on peatlands (the perforated part of the pipe was totally put into the underlying mineral soil at eight sites). Sampling was made monthly during spring (March–May) and autumn (November–December) and bimonthly during summer (June–October). The samples were chemically analyzed for pH, electrical conductivity, and concentrations of total P and Fe (before filtration), and total N, NO₃⁻-N, NH₄⁺-N, total P, PO₄³⁻-P, SO₄²⁻-S, Ca²⁺, Mg²⁺, K⁺, Na⁺, Mn, Zn, Fe, Al³⁺, Cl⁻ (after filtration through 0.45 μm membrane filter). Data collected until the end of 2001 are reported. Groundwater was found in the down-slope wells in lower-lying areas, but not in those installed on the slopes with a thin (1–2 m) soil layer. Clear-cutting did not significantly affect groundwater levels in the wells. Nitrate N concentrations increased from 0.03 mg L⁻¹ level after clear-cutting and again after site preparation in the wells on upland soils and peatlands receiving water from the managed parts of the catchment. In one well at the lower edge of the managed area NO₃⁻-N concentrations reached 1–1.4 mg L⁻¹ in 2001 (fifth year after clear-cutting, third after disc-plowing), but mean concentrations remained <0.3 mg L⁻¹. Chloride concentrations also increased (50–100%) after treatments but the concentrations of other solutes showed no significant effect of treatment. It was concluded that changes in groundwater quality and quantity related to the clear-cutting were small and did not represent a danger to water quality or quantity.     

Soil denitrification rates in a subalpine watershed of the eastern Sierra Nevada

Denitrification rates in soils of six subalpine plant communities in an eastern Sierra Nevada watershed were determined by the acetylene inhibition method. Soil atmosphere samples were collected monthly from June 1986 through May 1987 in a riparian, wet meadow, dry meadow, north-facing forest, south-facing forest and barren site and analyzed for N₂O content using gas chromatography. Soil temperature, moisture, organic matter, C,N,C:N ratio, NO₃N and pH were examined to assess their effects on denitrification rates. Mean denitrification rates for the year varied from 103.3 μg m⁻²h⁻¹ in the north-facing forest to 120.2 μg m⁻²h⁻¹ at the barren site, but did not differ significantly among any of the six plant communities. However, comparisons among months within individual communities revealed that the denitrification rates in each community varied significantly over the year, and in three of the six sites significant correlations between denitrification rates and other soil parameters were detected. Soil acidity was positively correlated with denitrification rate in the riparian and wet meadow communities, and in the dry meadow, soil moisture was positively correlated while soil temperature and organic matter were negatively correlated with denitrification rate. Comparisons among sites within individual months revealed significant differences in denitrification rates in June, September, October and January, but no single site consistently exhibited the highest or lowest rate in all 4 months, and only in October, when denitrification rates were positively correlated with soil temperature and moisture, was variation in denitrification rates among sites explained by other soil parameters. For the six plant communities overall, soil denitrification rates were highly variable from June to October, increased sharply from October to December, and then declined from December to May.     

四川盆地西缘山地典型地段不同林分对土壤质量的影响

以四川盆地西缘山地典型地段的桦木林、柳杉林、杉木林、竹林和天然常绿阔叶为研究对象,从土壤剖面特征、土壤机械组成、酸碱性和土壤养分含量等方面对比分析了林地土壤质量。结果显示:牛尾竹林土壤各项指标处于较低水平,土壤质量很差;人工栽植的针叶林(柳杉林),其人工林土壤表现出很好的发育和熟化特征,其有机质含量和速效N、P、K总量也仅次于天然林,且团聚体数量较多,但水稳性能较差,表土的养分淋溶特征明显;而盐基交换量和盐基饱和度以桦木幼林最大,其次为天然常绿阔叶林、竹林、杉木林,最小为柳杉林。表明以木材为主要经营目标的人工针叶林(柳杉林和杉木林)加速了盐基离子的流失,最终导致了土壤的酸化,因此合理的植被构成及林分经营模式是维持土壤质量的关键技术。     

Floristic and ecological differences between recent and ancient forests growing on non-acidic soils

The aim of this work was to investigate differences in soil chemistry and understory composition between recent forests (sites afforested in the last 170 years) and ancient forests growing on non-acidic soils. The study was carried out on hardwood forests at moderate elevation (400–600 m asl) in the Jura Mountains (N.E. France) on four main pedological substrates with different characteristics. The floristic composition of 127 stands from recent forests (n = 65) or ancient forests (n = 62) was surveyed. Some functional traits and the Ellenberg indicator values of the surveyed species were recorded. In addition, the topsoil from 30 stands was analysed. The composition of the flora was analysed by Detrended Correspondence Analysis and the species which were typical of one class of forest age were identified using a chi-square (χ²) test. The difference between forest classes for plant traits, their indicator values, or soil chemistry was tested using the generalized linear model and Bonferroni t-tests (or Kruskall–Wallis tests). The floristic composition of the ancient forests was significantly different from that of the recent forests and was characterized by a high occurrence of shrub species in recent forests. These differences were associated with higher specific leaf area, low-range seeds dispersal, and some life forms like geophytes. There was no clear difference in soil chemistry between the two classes of forests, except for δ¹⁵N values. The weakness of the difference in the soil between ancient and recent forests suggested that changes in soil chemistry caused by a former agricultural land use were not responsible for the differences in understory composition recorded. The differences in functional traits between the two forest classes supported this conclusion. We finally concluded that (i) past land use modifies the vegetation composition of current forests, even on neutral soils and that (ii) in our context, biological filters were probably responsible for these changes.     

Soil-solution chemistry in black locust, pine/mixed-hardwoods and oak/hickory forest stands in the southern Appalachians, U.S.A.

Soil-solution chemistry was measured over a 15-month period in three forest stands of contrasting nitrogen mineralization and nitrification rates in the southern Appalachians of North Carolina, U.S.A., using porous-cup lysimeters. In a black-locust-dominated stand, soil solution NO₃---N was 3.73 and 5.04 mg l⁻¹ at 30- and 60-cm depth respectively, and dissolved organic N ( ) was 0.718 and 0.582 mg l⁻¹ respectively. Values at 30 and 60 cm for a pine/mixed-hardwood stand were 0.032 and 0.058 mg l⁻¹ NO₃---N, and 0.201 and 0.168 mg l⁻¹ (values are means over the whole duration of the study). At both depths, soil solution conductivity, pH, Ca, Mg, K and PO₄---P were higher in black locust than in pine/mixed-hardwoods, and there were no differences in soil solution Na. In an oak/hickory stand, soil solution NO₃---N at 30-cm depth was 0.008 mg l⁻¹, and was 0.357 mg l⁻¹. At 30-cm depth, soil-solution conductivity, Ca, Mg and PO₄---P were higher in black locust than in oak-hickory, with no differences in pH, K and Na; , pH and K were higher in oak/hickory than in pine/mixed-hardwoods. In the oak/hickory and pine/mixed-hardwoods forest stands, with relatively lower soil N turnover rates, was a major portion of soil solution N.     

Change in soil macrofauna and vegetation when fast-growing trees are planted on savanna soils

Many forest species can be found in understory vegetation of old plantation plots, despite the fact that the native vegetation was a poor savanna growing on highly nonfertile sandy soils. The aim of the present paper is to describe the changes that occur in the environmental conditions when savanna is planted with fast-growing trees, and is particularly concerned with vegetation and soil macrofauna. The study was carried out in industrial eucalyptus plantations, and in experimental Acacia and pine plantations. Most plots were located on sandy soil, but some measurements were also carried out on clay soil planted with the same species in order to assess the influence of soil type.

A strong correlation was shown between the age of the eucalyptus trees and the percentage of forest species in undergrowth, emphasizing the progressive change from savanna vegetation towards forest vegetation.

Biomass and density of macrofauna were very low in both sandy and clayey savanna soils, total biomass being 3.3 and 5.8 g/m² respectively. Soil macrofauna became more important as the age of plantations increased, and biomass reached 29 g/m² in the 20-year-old eucalyptus plot on sandy soil, and 74 g/m² in 26-year-old eucalyptus plantation on clay soil, compared to 33 g/m² in the natural forest plot on sandy soil; however, frequency of occurrence and number of taxa were lower in old eucalyptus plot as compared to forest. Large differences in the abundance of macrofauna were observed in relation to planted species. Acacia was most favourable to soil macrofauna, with a total biomass of 60 g/m² on sandy soil and many taxa present. Pine plantations had a poor macrofauna and several taxa were lacking, particularly in the sandy soil.

Total macrofauna frequency was significantly correlated with the percentage of forest species in understory vegetation. Both were correlated with soil pH and soil organic-matter content. The results suggest that soil organic matter and litter quality are of main importance in changing the above- and below-ground habitat in plantations.     

Changes in soil properties and vegetation characteristics along a forest-savanna gradient in southern Venezuela

Vegetation cover in the Gran Sabana highlands (southern Venezuela) appears as a complex mosaic of tall to low forests, bush vegetation and savannas. In this study we described the changes in structure and floristic composition along a forest-savanna gradient consisted of tall forest (TF), medium forest (MF), low forest (LF) and open savanna (S), and analyse the possible reasons for the observed changes. The results showed no obvious differences in the soils properties along the vegetation gradient. All sites presented shallow soils (<50 cm depth) with high percentage of sand and with dominance of quartz, kaolinite and oxides of iron and aluminum. The soil chemical characteristics were unfavorable and similar along the vegetation gradient. The major soil difference was related with the presence of an organic layer on the soil surface of TF and MF and their absence on the soils of LF and S. Abundant residues of large trees were found on the forest floor of TF, MF and LF. These residues presented no signs of burning in TF, while in MF and particularly in LF were charred. This observation joined to the presence of charcoal within the mineral soil of S and the absence of the organic surface layer in LF and S indicated that fire has affected with different intensity or frequency the studied vegetation gradient. Large differences in the structure and floristic composition were found between TF, MF, LF and S. These differences could not be explained by changes in the mineralogical and chemical characteristics of the soil but only by fire which triggers the conversion. We concluded that the studied vegetation gradient represents stages in a temporal change from forest to savanna caused by fire, and this change has implied an impoverishment of tree species, a drastic reduction of biomass in terms of basal area, a drastic change of the floristic composition and the loss of the organic surface layer, which play an important role to maintain the fertility of these soils.