N deposition, N transformation and N leaching in acid forest soils

Nitrogen deposition, mineralisation, uptake and leaching were measured on a monthly basis in the field during 2 years in six forested stands on acidic soils under mountainous climate. Studies were conducted in three Douglas-fir [Pseudotsuga menziesii (Mirb.) Franco] plantations (D20: 20 year; D40: 40 yr; D60: 60 yr) on abandoned croplands in the Beaujolais Mounts; and two spruce (Picea abies Karst.) plantations (S45: 45 yr; S90: 90 yr) and an old beech (Fagus sylvatica L.) stand (B150: 150 yr) on ancient forest soils in a small catchment in the Vosges Mountains. N deposition in throughfall varied between 7–8 kg ha^–1 year^–1 (D20, B150, S45) and 15–21 kg ha^–1 yr^–1 (S90, D40, D60). N in annual litterfall varied between 20–29 kg ha^–1 (D40, D60, S90), and 36–43 kg ha^–1 (D20, S45, B150). N leaching below root depth varied among stands within a much larger range, between 1–9 kg ha^–1 yr^–1 (B150, S45, D60) and 28–66 kg ha^–1 yr^–1 (D40, S90, D20), with no simple relationship with N deposition, or N deposition minus N storage in stand biomass. N mineralisation was between 57–121 kg ha^–1 yr^–1 (S45, D40, S90) and between 176–209 kg ha^–1 yr^–1 in (B150, D60 and D20). The amounts of nitrogen annually mineralised and nitrified were positively related. Neither general soil parameters, such as pH, soil type, base saturation and C:N ratio, nor deposition in throughfall or litterfall were simply related to the intensity of mineralisation and/or nitrification. When root uptake was not allowed, nitrate leaching increased by 11 kg ha^–1 yr^–1 at S45, 36 kg ha^–1 yr^–1 at S90 and between 69 and 91 kg ha^–1 yr^–1 at D20, D40, B150 and D60, in relation to the nitrification rates of each plot. From this data set and recent data from the literature, we suggest that: high nitrification and nitrate leaching in Douglas-fir soils was likely related to the former agricultural land use. High nitrification rate but very low nitrate leaching in the old beech soil was related to intense recycling of mineralised N by beech roots. Medium nitrification and nitrate leaching in the old spruce stand was related to the average level of N deposition and to the deposition and declining health of the stand. Very low nitrification and N leaching in the young spruce stand were considered representative of fast growing spruce plantations receiving low N deposition on acidic soils of ancient coniferous forests. Consequently, we suggest that past land use and fine root cycling (which is dependent on to tree species and health) should be taken into account to explain the variability in the relation between N deposition and leaching in forests.     

Litterfall and nutrient dynamics in a montane moist evergreen broad-leaved forest in Ailao Mountains,SW China

Montane moist evergreen broad-leaved forest, dominated byLithocarpus and Castanopsis species,is the most extensive stand of subtropical mountain in Yunnan Province, SWChina. Litter production, standing crop of litter on forest floor and nutrientreturn patterns were studied over nine years (1991–1999) in a stand ofprimary evergreen broad-leaved forest in northern crest of the Ailao MountainRange. There were significant yearly variations in litter production, which ismainly related with the masting year of canopy species, and exceptionalphysicalevents (strong winds and snow) in the natural forest. The mean annual smalllitterfall is 7.12 t ha^–1 yr^–1ofwhich leaf litter account for 65% of the total litterfall. The seasonality ofsmall litterfall was bia-modal, with the main one in the late dry season(April–May) and a lesser one in early winter (October–November).Decomposition quotient value was relatively low with 0.58 for total smalllitterfall. Mean large-wood ( 2.5 cm in diameter) ranged from0.21 to 1.41 t ha^–1 yr^–1 with amean of 0.52 t ha^–1 yr^–1.Concentrations of most elements in leaf and twig were slightly greater in wetmonths than dry months, except for C and K. Woody litter had low N and Pconcentrations compared with the leaf and reproductive parts. Nutrient returntothe soil through small litterfall decrease in the orderC>N>Ca>K>Mg>Mn>Al>P>Fe, while nutrient reserve inlitteron the forest floor was in the declining sequenceC>N>Ca>K>Mg>P>Fe>Al>Mn.     

Root dynamics in bottomland hardwood forests of the Southeastern United States Coastal Plain

Effects of flooding on root dynamics appear nonlinear and therefore difficult to predict, leading to disparate and often contradictory reports of flooding impacts on production in bottomland hardwood forests. We explored root dynamics in two adjacent wetland habitats by comparing results obtained from several methods of estimating root processes. Also, we tested the influence of flooding on root dynamics of cherrybark, overcup, water and swamp chestnut oaks. Fine root biomass in the laurel oak habitat was greater (< 0.05) than in the swamp tupelo habitat (5.7 vs. 2.4 Mg ha^–1), as was fine root necromass (2.4 vs. 1.3 Mg ha^–1), productivity (2.3 vs. 0.3 Mg ha^–1 yr^–1 when the sum of significant increments method was used, 5.6 vs. 2.5 Mg ha ^–1 yr^–1 when the maximum minus minimum method was used, and 1.2 vs. 1.0 Mg ha^–1 yr ^–1, when the root screen method was used), and turnover (40% and 12% per year). Mortality estimates were lower in the laurel oak habitat (1.3 and 1.2 Mg ha^–1 yr^–1) than in the swamp tupelo community (2.8 and 2.1 Mg ha^–1 yr^–1) when significant increment and maximum minus minimum methods were used, respectively. This apparent contradiction between estimates of production and mortality may be due to more rapid decomposition rates in the more aerated soil of the laurel oak than in the swamp tupelo forest type. Roots in the swamp tupelo habitat appeared to be longer-lived than in the laurel oak habitat. We concluded that there was greater investment in roots in the laurel oak habitat, where a shallow rooting zone and episodes of flooding and drought required drastic changes in root structure and physiology. In contrast, the swamp tupelo habitat had a deeper rooting zone and more consistently moist to flooded hydroperiod, allowing flood adapted roots to persist. The four oak species varied in their phenology of root production and response to flooding, from no difference among treatments for overcup oak to dramatic reductions in root growth during and after flooding for cherrybark oak. Flooding enhanced or at least did not negatively influence root growth in overcup oak, but seriously impacted root growth and survival of cherrybark oak and swamp chestnut oak. Different responses were attributed to the timing of root production: root growth began early for cherrybark oak so spring flooding severely affected this species. Growth in overcup oak began later and ended earlier than the other species tested, allowing the species a means of avoiding flood stress.     

Retention of soluble organic nutrients by a forested ecosystem

We document an example of a forested watershed at the Coweeta HydrologicLaboratory with an extraordinary tendency to retain dissolved organic matter(DOM) generated in large quantities within the ecosystem. Our objectives weretodetermine fluxes of dissolved organic C, N, and P (DOC, DON, DOP,respectively),in water draining through each stratum of the ecosystem and synthesizeinformation on the physicochemical, biological and hydrologic factors leadingtoretention of dissolved organic nutrients in this ecosystem. The ecosystemretained 99.3, 97.3, and 99.0% of water soluble organic C, N and P,respectively, produced in litterfall, throughfall, and root exudates. Exportsinstreamwater were 4.1 kg ha^–1yr^–1of DOC, 0.191 kg ha^–1 yr^–1 ofDON, and 0.011 kg ha^–1 yr^–1 ofDOP. Fluxes of DON were greater than those of inorganic N in all strata. MostDOC, DON, and DOP was removed from solution in the A and B horizons, with DOCbeing rapidly adsorbed to Fe and Al oxyhydroxides, most likely by ligandexchange. DON and DOC were released gradually from the forest floor over theyear. Water soluble organic C produced in litterfall and throughfall had adisjoint distribution of half-decay times with very labile and veryrefractory fractions so that most labile DOC was decomposed before beingleachedinto the mineral soil and refractory fractions dominated the DOC transportedthrough the ecosystem. We hypothesize that this watershed retained solubleorganic nutrients to an extraordinary degree because the soils have very highcontents of Fe and Al oxyhydroxides with high adsorption capacities and becausethe predominant hydrologic pathway is downwards as unsaturated flow through astrongly adsorbing A and B horizon. The well recognized retention mechanismsforinorganic nutrients combine with adsorption of DOM and hydrologic pathway toefficiently prevent leaching of both soluble inorganic andorganic nutrients in this watershed.     

Nitrogen and phosphorus budgets for the sub-tropical Richmond River catchment, Australia

Nitrogen and phosphorus budgets were developed forfour sub-catchments in the Richmond River catchmentfor two study years. The catchment is used for avariety of farming pursuits including dairying, beef,cropping, fruit, nuts, forestry, and sugar cane. Eachsub-catchment varies in hydrology, the proportion ofeach land use, and the population density whichenabled a unique opportunity to study fluxes andstorage associated with a variety of environmentalfactors. Total loadings entering each sub-catchmentvaried from 12 to 57 kg ha^–1yr^–1 fornitrogen and 0.25 to 6.6 kg ha^–1yr^–1 forphosphorus with little inter-annual variation.Averaged across the whole catchment, nitrogen fixation(47%) dominated the inputs; fertiliser (26%) andrainfall (21%) made up the next largest inputs.Fertiliser inputs dominated the phosphorus budget(65.5%); rainfall and manures making up 13% and 12%respectively. Produce dominated the outputs of bothnitrogen and phosphorus from the four sub-catchmentsbeing greater than the riverine export. The deliveryof nitrogen to catchment streams ranged from <1 to24% of the total inputs and the delivery of phosphorus to catchment streams ranged from <1 to 39%. Storage of phosphorus in catchment soils varied between –0.32 and 4.46 kg ha^–1yr^–1. Whendenitrification and volatilisation were estimated using data from other studies, storage of nitrogen ranged from 1 to 24 kg ha^–1yr^–1. Despite theepisodic nature of runoff in the sub-tropical RichmondRiver catchment, the magnitude of nutrient fluxes andstorage appear similar to other catchments of theworld which have mixed land use and relatively lowcatchment nutrient loadings.     

Cation distribution,cycling, and removal from mineral soil in Douglas-fir and red alder forests

Overstory species influence the distribution and dynamics of nutrients in forest ecosystems. Ecosystem-level estimates of Ca, Mg, and K pools and cycles in 50-year old Douglas-fir and red alder stands were used to determine the effect of overstory composition on net cation removal from the mineral soil, i.e. cation export from the soil in excess of additions. Net cation removal from Douglas-fir soil was 8 kg Ca ha^–1 yr^–1, 1 kg Mg ha^–1 yr^–1, and 0.3 kg K ha^–1 yr^–1. Annual cation export from soil by uptake and accumulation in live woody tissue and O horizon was of similar magnitude to leaching in soil solution. Atmospheric deposition partially off-set export by adding cations equivalent to 28–88% of cation export. Net cation removal from red alder soil was 58 kg Ca ha^–1 yr^–1, 9 kg Mg ha^–1 yr^–1, and 11 kg K ha^–1 yr^–1. Annual cation accumulation in live woody tissue and O horizon was three times greater than in Douglas-fir, while cation leaching in soil solution was five to eight times greater. The lack of excessive depletion of exchangeable cations in the red alder soil suggests that mineral weathering, rather than exchangeable cations, was the source of most of the removed cations. Nitric acid generated during nitrification in red alder soil led to high rates of weathering and NO₃-driven cation leaching.     

Nutrient Limitation in Three Lowland Tropical Forests in Southern China Receiving High Nitrogen Deposition: Insights from Fine Root Responses to Nutrient Additions

Elevated nitrogen (N) deposition to tropical forests may accelerate ecosystem phosphorus (P) limitation. This study examined responses of fine root biomass, nutrient concentrations, and acid phosphatase activity (APA) of bulk soil to five years of N and P additions in one old-growth and two younger lowland tropical forests in southern China. The old-growth forest had higher N capital than the two younger forests from long-term N accumulation. From February 2007 to July 2012, four experimental treatments were established at the following levels: Control, N-addition (150 kg N ha^–1 yr^–1), P-addition (150 kg P ha^–1 yr^–1) and N+P-addition (150 kg N ha^–1 yr^–1 plus 150 kg P ha^–1 yr^–1). We hypothesized that fine root growth in the N-rich old-growth forest would be limited by P availability, and in the two younger forests would primarily respond to N additions due to large plant N demand. Results showed that five years of N addition significantly decreased live fine root biomass only in the old-growth forest (by 31%), but significantly elevated dead fine root biomass in all the three forests (by 64% to 101%), causing decreased live fine root proportion in the old-growth and the pine forests. P addition significantly increased live fine root biomass in all three forests (by 20% to 76%). The combined N and P treatment significantly increased live fine root biomass in the two younger forests but not in the old-growth forest. These results suggest that fine root growth in all three study forests appeared to be P-limited. This was further confirmed by current status of fine root N:P ratios, APA in bulk soil, and their responses to N and P treatments. Moreover, N addition significantly increased APA only in the old-growth forest, consistent with the conclusion that the old-growth forest was more P-limited than the younger forests.     

Exchange of N-gases at the Höglwald Forest – A summary

During 4 years continuous measurements of N-trace gas exchange were carried out at the forest floor-atmosphere interface at the Höglwald Forest that is highly affected by atmospheric N-deposition. The measurements included spruce control, spruce limed and beech sites. Based on these field measurements and on intensive laboratory measurements of N₂-emissions from the soils of the beech and spruce control sites, a total balance of N-gas emissions was calculated. NO₂-deposition was in a range of –1.6 –2.9 kg N ha^–1 yr^–1 and no huge differences between the different sites could be demonstrated. In contrast to NO₂-deposition, NO- and N₂O-emissions showed a huge variability among the different sites. NO emissions were highest at the spruce control site (6.4–9.1 kg N ha^–1 yr^–1), lowest at the beech site (2.3–3.5 kg N ha^–1 yr^–1) and intermediate at the limed spruce site (3.4–5.4 kg N ha^–1 yr^–1). With regard to N₂O-emissions, the following ranking between the sites was found: beech (1.6–6.6 kg N ha^–1 yr^–1) >> spruce limed (0.7–4.0 kg N ha^–1 yr^–1) > spruce control (0.4–3.1 kg N ha^–1 yr^–1). Average N-trace gas emissions (NO, NO₂, N₂O) for the years 1994–1997 were 6.8 kg N ha^–1 yr^–1 at the spruce control site, 3.6 kg N ha^–1 yr^–1 at the limed spruce site and 4.5 kg N ha^–1 yr^–1 at the beech site. Considering N₂-losses, which were significantly higher at the beech (12.4 kg N ha^–1 yr^–1) than at the spruce control site (7.2 kg N ha^–1 yr^–1), the magnitude of total gaseous N losses, i.e. N₂-N + NO-N + NO₂-N + N₂O-N, could be calculated for the first time for a forest ecosystem. Total gaseous N-losses were 14.0 kg N ha^–1 yr^–1 at the spruce control site and 15.5 kg N ha^–1 yr^–1 at the beech site, respectively. In view of the huge interannual variability of N-trace gas fluxes and the pronounced site differences in N-gas emissions it is concluded that more research is needed in order to fully understand patterns of microbial N-cycling and N-gas production/emission in forest ecosystems and mechanisms of reactions of forest ecosystems to the ecological stress factor of atmospheric N-input.     

Allocation of carbon in a mature eucalypt forest and some effects of soil phosphorus availability

Pools and annual fluxes of carbon (C) were estimated for a mature Eucalyptus pauciflora (snowgum) forest with and without phosphorus (P) fertilizer addition to determine the effect of soil P availability on allocation of C in the stand. Aboveground biomass was estimated from allometric equations relating stem and branch diameters of individual trees to their biomass. Biomass production was calculated from annual increments in tree diameters and measurements of litterfall. Maintenance and construction respiration were calculated for each component using equations given by Ryan (1991a). Total belowground C flux was estimated from measurements of annual soil CO₂ efflux less the C content of annual litterfall (assuming forest floor and soil C were at approximate steady state for the year that soil CO₂ efflux was measured). The total C content of the standing biomass of the unfertilized stand was 138 t ha^-1, with approximately 80% aboveground and 20% belowground. Forest floor C was 8.5 t ha^-1. Soil C content (0–1 m) was 369 t ha^-1 representing 70% of the total C pool in the ecosystem. Total gross annual C flux aboveground (biomass increment plus litterfall plus respiration) was 11.9 t ha^-1 and gross flux belowground (coarse root increment plus fine root production plus root respiration) was 5.1 t ha^-1. Total annual soil efflux was 7.1 t ha^-1, of which 2.5 t ha^-1 (35%) was contributed by litter decomposition.The short-term effect of changing the availability of P compared with C on allocation to aboveground versus belowground processes was estimated by comparing fertilized and unfertilized stands during the year after treatment. In the P-fertilized stand annual wood biomass increment increased by 30%, there was no evidence of change in canopy biomass, and belowground C allocation decreased by 19% relative to the unfertilized stand. Total annual C flux was 16.97 and 16.75 t ha^-1 yr^-1 and the ratio of below- to aboveground C allocation was 0.43 and 0.35 in the unfertilized and P-fertilized stands, respectively. Therefore, the major response of the forest stand to increased soil P availability appeared to be a shift in C allocation; with little change in total productivity. These results emphasise that both growth rate and allocation need to be estimated to predict changes in fluxes and storage of C in forests that may occur in response to disturbance or climate change.     

Regional analysis of inorganic nitrogen yield and retention in high-elevation ecosystems of the Sierra Nevada and Rocky Mountains

Yields and retention of dissolved inorganic nitrogen (DIN: NO₃ ^– + NH₄ ⁺) and nitrate concentrations in surface runoff are summarized for 28 high elevation watersheds in the Sierra Nevada of California and Rocky Mountains of Wyoming and Colorado. Catchments ranged in elevation from 2475 to 3603 m and from 15 to 1908 ha in area. Soil cover varied from 5% to nearly 97% of total catchment area. Runoff from these snow-dominated catchments ranged from 315 to 1265 mm per year. In the Sierra Nevada, annual volume-weighted mean (AVWM) nitrate concentrations ranged from 0.5 to 13 M (overall average 5.4 M), and peak concentrations measured during snowmelt ranged from 1.0 to 38 M. Nitrate levels in the Rocky Mountain watersheds were about twice those in the Sierra Nevada; average AVWM NO₃ ^– was 9.4 M and snowmelt peaks ranged from 15 to 50 M. Mean DIN loading to Rocky Mountain watersheds, 3.6 kg ha^–1 yr^–1, was double the average measured for Sierra Nevada watersheds, 1.8 kg ha^–1 yr^–1. DIN yield in the Sierra Nevada, 0.69 kg ha^–1 yr^–1, was about 60% that measured in the Rocky Mountains, 1.1 kg ha^–1 yr^–1. Net inorganic N retention in Sierra Nevada catchments was 1.2 kg ha^–1 yr^–1 and represented about 55% of annual DIN loading. DIN retention in the Rocky Mountain catchments was greater in absolute terms, 2.5 kg ha^–1 yr^–1, and as a percentage of DIN loading, 72%.A correlation analysis using DIN yield, DIN retention and surface water nitrate concentrations as dependent variables and eight environmental features (catchment elevation, slope, aspect, roughness, area, runoff, soil cover and DIN loading) as independent variables was conducted. For the Sierra Nevada, elevation and soil cover had significant (p > 0.1) Pearson product moment correlations with catchment DIN yield, AVWM and peak snowmelt nitrate concentrations and DIN retention rates. Log-linear regression models using soil cover as the independent variable explained 82% of the variation in catchment DIN retention, 92% of the variability in AVWM nitrate and 85% of snowmelt peak NO₃ ^–. In the Rocky Mountains, soil cover was significantly (p < 0.05) correlated with DIN yield, AVWM NO₃ ^– and DIN retention expressed as a percentage of DIN loading (%DIN retention). Catchment mean slope and terrain roughness were positively correlated with steam nitrate concentrations and negatively related to %DIN retention. About 91% of the variation in DIN yield and 79% of the variability in AVWM NO₃ ^– were explained by log-linear models based on soil cover. A log-linear regression based on soil cover explained 90% of the variation of %DIN retention in the Rocky Mountains.