Role of brown bears (Ursus arctos) in the flow of marine nitrogen into a terrestrial ecosystem

We quantified the amount, spatial distribution, and importance of salmon (Oncorhynchus spp.)-derived nitrogen (N) by brown bears (Ursus arctos) on the Kenai Peninsula, Alaska. We tested and confirmed the hypothesis that the stable isotope signature (δ¹⁵N) of N in foliage of white spruce (Picea glauca) was inversely proportional to the distance from salmon-spawning streams (r=–0.99 and P<0.05 in two separate watersheds). Locations of radio-collared brown bears, relative to their distance from a stream, were highly correlated with δ¹⁵N depletion of foliage across the same gradient (r=–0.98 and –0.96 and P<0.05 in the same two separate watersheds). Mean rates of redistribution of salmon-derived N by adult female brown bears were 37.2±2.9 kg/year per bear (range 23.1–56.3), of which 96% (35.7±2.7 kg/year per bear) was excreted in urine, 3% (1.1±0.1 kg/year per bear) was excreted in feces, and <1% (0.3± 0.1 kg/year per bear) was retained in the body. On an area basis, salmon-N redistribution rates were as high as 5.1±0.7 mg/m² per year per bear within 500 m of the stream but dropped off greatly with increasing distance. We estimated that 15.5–17.8% of the total N in spruce foliage within 500 m of the stream was derived from salmon. Of that, bears had distributed 83–84%. Thus, brown bears can be an important vector of salmon-derived N into riparian ecosystems, but their effects are highly variable spatially and a function of bear density. Received: 11 February 1999 / Accepted: 7 July 1999     

Return of Salmon-Derived Nutrients from the Riparian Zone to the Stream during a Storm in Southeastern Alaska

Spawning salmon deliver nutrients (salmon-derived nutrients, SDN) to natal watersheds that can be incorporated into terrestrial and aquatic food webs, potentially increasing ecosystem productivity. Peterson Creek, a coastal watershed in southeast Alaska that supports several species of anadromous fish, was sampled over the course of a storm during September 2006 to test the hypothesis that stormflows re-introduce stored SDN into the stream. We used stable isotopes and PARAFAC modeling of fluorescence excitation–emission spectroscopy to detect flushing of DOM from salmon carcasses in the riparian zone back into a spawning stream. During the early storm hydrograph, streamwater concentrations of NH₄–N and total dissolved phosphorus (TDP), the fluorescent protein tyrosine and the δ¹⁵N content of DOM peaked, followed by a rapid decrease during maximum stormflow. Although δ¹⁵N has previously been used to track SDN in riparian zones, the use of fluorescence spectroscopy provides an independent indicator that SDN are being returned from the riparian zone to the stream after a period of intermediate storage outside the stream channel. Our findings further demonstrate the utility of using both δ¹⁵N of streamwater DOM and fluorescence spectroscopy with PARAFAC modeling to monitor how the pool of streamwater DOM changes in spawning salmon streams.     

Marine-derived nitrogen and carbon in freshwater-riparian food webs of the Copper River Delta,southcentral Alaska

After rearing to adulthood at sea, coho salmon (Oncorhynchus kisutch) return to freshwater to spawn once and then die on or near their spawning grounds. We tested the hypothesis that spawning coho salmon return marine N and C to beaver (Castor canadensis) ponds of the Copper River Delta (CRD), Cordova, southcentral Alaska, thereby enhancing productivity of the aquatic food webs that support juvenile coho salmon. We sampled three types of pond treatment: (1) natural enrichment by spawning salmon, (2) artificial enrichment via addition of salmon carcasses and eggs, and (3) ponds with no salmon enrichment. All ponds supported juvenile coho salmon. Seasonal samples of stable isotopes revealed that juvenile coho salmon, threespine sticklebacks (Gasterosteus aculeatus), caddisfly larvae, leeches, and chironomid midge larvae were enriched with marine N and C. The aquatic vascular plants bur reed (Sparganium hyperboreum), pondweed (Potamogeton gramineus), and mare’s tail (Hippuris vulgaris) were enriched with marine N only. Riparian vegetation (Sitka alder Alnus viridis ssp. sinuata and willow Salix spp.) did not show enrichment. Artificial additions of adult carcasses and eggs of coho salmon increased the δ¹⁵N and δ¹³C values of juvenile coho salmon. In this dynamic and hydrologically complex coastal environment, spawning coho salmon contributed marine N and C comprising 10–50% of the dietary needs of juvenile coho salmon through direct consumption of eggs and carcass material. Invertebrates that have assimilated marine N and C yield a further indirect contribution. This perennial subsidy maintains the productivity of the ecosystem of the coho salmon on the CRD.     

Effects of Pacific salmon spawning and carcass availability on the caddisfly Ecclisomyia conspersa (Trichoptera: Limnephilidae)

1. The effects of spawning coho (Oncorhynchus kisutch) and chum salmon (Oncorhynchus keta) on the limnephilid caddisfly Ecclisomyia conspersa were evaluated by experimentally excluding salmon from the upper 14‐m stretch of a spawning channel by a wire‐meshed fence. Density, and development and growth rates, of larvae upstream of the fence (without salmon) were compared with those downstream (with salmon). 2. Larval density in the stretch with salmon declined during spawning, but increased again after spawning subsided and the carcasses of dead fish became available. In the stretch with salmon, larval density on salmon carcasses was seven to 37 times greater than on the adjacent channel substratum. The rate of larval development in the stretch with salmon was greater than that in the stretch without salmon. Two months after carcasses became available, 98% of larvae sampled from the stretch with salmon were in the fifth instar, compared to only 23% from the stretch without salmon. Body weight of E. conspersa in the stretches with and without salmon increased by an average of 3.04 and 2.38 mg, respectively, over a 6‐month period. 3. ¹⁵N values of larvae from the stretch with salmon increased following the arrival of the fish, suggesting that the larvae were feeding on salmon‐derived material, such as eggs and carcasses, which contain a high proportion of the heavier stable isotope. In contrast, ¹⁵N values of larvae from the stretch without salmon remained relatively constant throughout the experiment. The availability of salmon carcasses as a high‐quality food source late in larval development may increase survival and fecundity of E. conspersa. 4. These substantial differences were consistent with the view that they were due to the experimental exclusion of salmon and salmon carcasses from the upstream stretch, though the study was un‐replicated and thus precludes ascribing causation more definitely.     

Stable isotope evidence indicates the incorporation into Japanese catchments of marine‐derived nutrients transported by spawning Pacific Salmon

相似文献   

Ecosystem N Distribution and δ<Superscript>15</Superscript>N during a Century of Forest Regrowth after Agricultural Abandonment

Abstract Stable isotope ratios of terrestrial ecosystem nitrogen (N) pools reflect internal processes and input–output balances. Disturbance generally increases N cycling and loss, yet few studies have examined ecosystem δ¹⁵N over a disturbance-recovery sequence. We used a chronosequence approach to examine N distribution and δ¹⁵N during forest regrowth after agricultural abandonment. Site ages ranged from 10 to 115 years, with similar soils, climate, land-use history, and overstory vegetation (white pine Pinus strobus). Foliar N and δ¹⁵N decreased as stands aged, consistent with a progressive tightening of the N cycle during forest regrowth on agricultural lands. Over time, foliar δ¹⁵N became more negative, indicating increased fractionation along the mineralization–mycorrhizal–plant uptake pathway. Total ecosystem N was constant across the chronosequence, but substantial internal N redistribution occurred from the mineral soil to plants and litter over 115 years (>25% of ecosystem N or 1,610 kg ha⁻¹). Temporal trends in soil δ¹⁵N generally reflected a redistribution of depleted N from the mineral soil to the developing O horizon. Although plants and soil δ¹⁵N are coupled over millennial time scales of ecosystem development, our observed divergence between plants and soil suggests that they can be uncoupled during the disturbance-regrowth sequence. The approximate 2‰ decrease in ecosystem δ¹⁵N over the century scale suggests significant incorporation of atmospheric N, which was not detected by traditional ecosystem N accounting. Consideration of temporal trends and disturbance legacies can improve our understanding of the influence of broader factors such as climate or N deposition on ecosystem N balances and δ¹⁵N. Electronic supplementary material The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Effects of land use on 15N natural abundance of soils in Ethiopian highlands

We used the natural abundance of ¹⁵N in soils in forests, pastures and cultivated lands in the Menagesha and Wendo-Genet areas of Ethiopia to make inferences about the N cycles in these ecosystems. Since we have described the history of these sites based on variations in ¹³C natural abundance, patterns of δ¹⁵N and δ¹³C values were compared to determine if shifts of ¹⁵N correlate with shifts of vegetation. At Menagesha, a > 500-yr-old planted forest, we found δ¹⁵N values from −8.8 to +3.5‰ in litter, from −3.5 to +4.5‰ in 0–10 cm soil layer, and from −1.5 to +6.8‰ at >20 cm soil depth. The low δ¹⁵N in litter and surface mineral soils suggests that a closed N cycle has operated for a long time. At this site, the low δ¹³C of the surface horizon and the high δ¹³C of the lower soil horizons is clear evidence of a long phase of C₄ grass dominance or cultivation of C₄ crops before the establishment of the forest >500 years ago. In contrast, at Wendo-Genet, high δ¹³C of soils reveals that most of the land has been uncovered by forests until recently. Soil δ¹⁵N was high throughout (3.4–9.8‰), and there were no major differences between forested, cultivated and pasture soils in δ¹⁵N values of surface mineral soils. The high δ¹⁵N values suggest that open N cycles operate in the Wendo-Genet area. From the points of view of soil fertility management, it is interesting that tall forest ecosystems with relatively closed N cycling could be established on the fairly steep slopes at Menagesha after a long period of grass vegetation cover or cultivation. This revised version was published online in June 2006 with corrections to the Cover Date.     

Effects of spawning Pacific salmon on terrestrial invertebrates: Insects near spawning habitat are isotopically enriched with nitrogen‐15 but display no differences in body size

When Pacific salmon (Oncorhynchus spp.) spawn and die, they deliver marine‐derived nutrient subsidies to freshwater and riparian ecosystems. These subsidies can alter the behavior, productivity, and abundance of recipient species and their habitats. Isotopes, such as nitrogen‐15 (¹⁵N), are often used to trace the destination of marine‐derived nutrients in riparian habitats. However, few studies have tested for correlations between stable isotopes and physiological responses of riparian organisms. We examined whether increases in δ ¹⁵N in terrestrial insect bodies adjacent to salmon spawning habitat translate to changes in percent nitrogen content and body size. This involved comparisons between distance from a salmon‐bearing river, marine‐derived nutrients in soils and insects, soil moisture content, and body size and nitrogen content in two common beetle families (Coleoptera: Curculionidae, Carabidae). As predicted, δ¹⁵N in riparian soils attenuated with distance from the river but was unaffected by soil moisture. This gradient was mirrored by δ¹⁵N in the herbivorous curculionid beetles, whereas carabid beetles, which feed at a higher trophic level and are more mobile, did not show discernable patterns in their δ¹⁵N content. Additionally, neither distance from the river nor body δ¹⁵N content was related to beetle body size. We also found that nitrogen‐15 was not correlated with total percent nitrogen in insect bodies, meaning that the presence of spawning salmon did not increase the percent nitrogen content of these insects. We conclude that while salmon‐derived nutrients had entered terrestrial food webs, the presence of δ¹⁵N alone did not indicate meaningful physiological changes in these insects in terms of percent nitrogen nor body size. While stable isotopes may be useful tracers of marine‐derived nutrients, they cannot necessarily be used as a proxy for physiologically important response variables.     

Effects of sediment removal on nitrogen uptake by riparian plants in the higher floodplain of the Chikuma River, Japan

Riparian plants can use nitrogen (N) from soil and river water, but the use of river water N might be limited in higher floodplain environments of the Chikuma River. The purpose of this study is to reveal the relationship between N uptake by riparian plants and the floodplain topography (relative height and distance from a river channel). We examined the hypothesis that surface sediment removal from the higher floodplain increases river water N uptake by riparian plants by using a stable isotope analysis. The δ¹⁵N value of river water samples (ca. 8‰) were significantly higher than those of the soil extracts (ca. 3‰) in the study area. The δ¹⁵N value of riparian plants increased from +3.0‰ (standard deviation, SD ±2.1‰) before sediment removal to +9.6‰ (±2.1‰) after sediment removal, although there was no significant change in the δ¹⁵N value in N sources of soil and river water. The sediment removal enhanced frequency of flood disturbance, relative ground water level, and river water N uptake by riparian plants on the floodplain.     

Large predators and biogeochemical hotspots: brown bear (Ursus arctos) predation on salmon alters nitrogen cycling in riparian soils

Two important themes in ecology include the understanding of how interactions among species control ecosystem processes, and how habitats can be connected through transfers of nutrients and energy by mobile organisms. An impressive example of both is the large influx of nutrients and organic matter that anadromous salmon supply to inland aquatic and terrestrial ecosystems and the role of predation by brown bears (Ursus arctos) in transferring these marine-derived nutrients (MDN) from freshwater to riparian habitats. In spite of the recognition that salmon-bear interactions likely play an important role in controlling the flux of MDN from aquatic to riparian habitats, few studies have linked bear predation on salmon to processes such as nitrogen (N) or carbon (C) cycling. We combine landscape-level survey data and a replicated bear-exclosure experiment to test how bear foraging on salmon affects nitrous oxide (N₂O) flux, carbon dioxide (CO₂) flux, and nutrient concentrations of riparian soils. Our results show that bears feeding on salmon increased soil ammonium (NH₄ ⁺) concentrations three-fold and N₂O flux by 32-fold. Soil CO₂ flux, nitrate (NO₃ ⁻), and N transformation differences were negligible in areas where bears fed on salmon. Reference areas without concentrated bear activity showed no detectable change in soil N cycling after the arrival of salmon to streams. Exclosure experiments showed that bear effects on soil nutrient cycles were transient, and soil N processing returned to background conditions within 1 year after bears were removed from the system. These results suggest that recipient ecosystems do not show uniform responses to MDN inputs and highlight the importance of large mobile consumers in generating landscape heterogeneity in nutrient cycles.     

Plant and soil natural abundance <Emphasis Type="Italic">δ</Emphasis><Superscript>15</Superscript>N: indicators of relative rates of nitrogen cycling in temperate forest ecosystems

Watersheds within the Catskill Mountains, New York, receive among the highest rates of nitrogen (N) deposition in the northeastern United States and are beginning to show signs of N saturation. Despite similar amounts of N deposition across watersheds within the Catskill Mountains, rates of soil N cycling and N retention vary significantly among stands of different tree species. We examined the potential use of δ ¹⁵N of plants and soils as an indicator of relative forest soil N cycling rates. We analyzed the δ ¹⁵N of foliage, litterfall, bole wood, surface litter layer, fine roots and organic soil from single-species stands of American beech (Fagus grandifolia), eastern hemlock (Tsuga canadensis), red oak (Quercus rubra), and sugar maple (Acer saccharum). Fine root and organic soil δ ¹⁵N values were highest within sugar maple stands, which correlated significantly with higher rates of net mineralization and nitrification. Results from this study suggest that fine root and organic soil δ ¹⁵N can be used as an indicator of relative rates of soil N cycling. Although not statistically significant, δ ¹⁵N was highest within foliage, wood and litterfall of beech stands, a tree species associated with intermediate levels of soil N cycling rates and forest N retention. Our results show that belowground δ ¹⁵N values are a better indicator of relative rates of soil N cycling than are aboveground δ ¹⁵N values.     

Spatial variation of the stable nitrogen isotope ratio of woody plants along a topoedaphic gradient in a subtropical savanna

Variation in the stable N isotope ratio (δ¹⁵N) of plants and soils often reflects the influence of environment on the N cycle. We measured leaf δ¹⁵N and N concentration ([N]) on all individuals of Prosopis glandulosa (deciduous tree legume), Condalia hookeri (evergreen shrub), and Zanthoxylum fagara (evergreen shrub) present within a belt transect 308 m long × 12 m wide in a subtropical savanna ecosystem in southern Texas, USA in April and August 2005. Soil texture, gravimetric water content (GWC), total N and δ¹⁵N were also measured along the transect. At the landscape scale, leaf δ¹⁵N was negatively related to elevation for all the three species along this topoedaphic sequence. Changes in soil δ¹⁵N, total N, and GWC appeared to contribute to this spatial pattern of leaf δ¹⁵N. In lower portions of the landscape, greater soil N availability and GWC are associated with relatively high rates of both N mineralization and nitrification. Both soil δ¹⁵N and leaf [N] were positively correlated with leaf δ¹⁵N of non-N₂ fixing plants. Leaf δ¹⁵N of P. glandulosa, an N₂-fixing legume, did not correlate with leaf [N]; the δ¹⁵N of P. glandulosa’s leaves were closer to atmospheric N₂ and significantly lower than those of C. hookeri and Z. fagara. Additionally, at smaller spatial scales, a proximity index (which reflected the density and distance of surrounding P. glandulosa trees) was negatively correlated with leaf δ¹⁵N of C. hookeri and Z. fagara, indicating the N₂-fixing P. glandulosa may be important to the N nutrition of nearby non-N₂-fixing species. Our results indicate plant ¹⁵N natural abundance can reflect the extent of N retention and help us better understand N dynamics and plant-soil interactions at ecosystem and landscape scales.     

Carbon isotopic fractionation in subtropical brazilian grassland soils. Comparison with tropical forest soils

The natural relationship¹³C/¹²C determined in three soil profiles under grass vegetation indicated a depletion in organic¹³C at depth: theδ ¹³C was between −18‰ and −15‰ in the A horizons and ranged from −18 to −22‰ at depth. Previous work showed that in forest soils, whereδ ¹³C was near −28‰ in the upper horizon, there was, on the contrary, a relative enrichment of the lower strata. This meant thatδ ¹³C, initially different in the various topsoils, became more equal at depth. Comparison between dark, deep horizons (sombric horizons), which are certainly of illuvial origine, would confirm this:δ ¹³C of grassland and a forest sombric horizon were almost equal at around −22‰. These results might mean that, in natural ecosystems, the isotopic carbon composition of the soil underlying humus would be independent of the vegetation type. This would have practical implications for the use of¹³C as a tracer for soil organic matter studies.     

Increased soil stable nitrogen isotopic ratio following phosphorus enrichment: historical patterns and tests of two hypotheses in a phosphorus-limited wetland

We used a P enrichment gradient in the Everglades to investigate patterns of the stable N isotopic ratio (δ¹⁵N) in peat profiles as an indicator of historic eutrophication of this wetland. We also tested two hypotheses to explain the effects of P on increased δ¹⁵N of organic matter including: (1) increased N mineralization/N loss, and (2) reduced isotopic discrimination during macrophyte N uptake. Spatial patterns of δ¹⁵N in surface litter and soil (0–10 cm) mimic those of the aboveground macrophytes (Typha domingensis Pers. and Cladium jamaicense Crantz). Peat profiles also show increased δ¹⁵N in the peat accumulated in areas near the historic P discharges since the early 1960s. The increased δ¹⁵N of bulk peat correlated well with both measured increases in soil total P and the historical beginning of nutrient discharges into this wetland. In 15-day bottle incubations of soil, added P had no effect on the δ¹⁵N of NH₄⁺ and significantly increased the δ¹⁵N of water-extractable organic N. Measurements of surface soils collected during a field mesocosm experiment also revealed no significant effect of P on δ¹⁵N even after 5 years of P addition. In contrast, δ¹⁵N of leaf and root tissues of hydroponically grown Typha and Cladium were shown to increase up to 12‰ when grown at elevated levels of P and fixed levels of N (as NH₄⁺). The magnitude of changes in δ¹⁵N resulting from altered discrimination during N uptake is significant compared with other mechanisms affecting plant δ¹⁵N, and suggests that this may be the dominant mechanism affecting δ¹⁵N of organic matter following P enrichment. The results of this study have implications for the interpretation of δ¹⁵N as an indicator of shifts in relative N limitation in wetland ecosystems, and also stress the importance of experimental validation in interpreting δ¹⁵N patterns.     

Symbiotic nitrogen fixation in eight Acacia senegal provenances in dryland clays of the Blue Nile Sudan estimated by the 15N natural abundance method

The symbiotic biological N₂fixation by Acacia senegal was estimated using the ¹⁵N natural abundance (δ ¹⁵N) procedure on eight provenances collected from different environments and soil types grown in a clay soil in the Blue Nile region, Sudan. Balanites aegyptiaca (a non-legume) was used as a non-N₂-fixing reference plant to allow ¹⁵N-based estimates of the proportion of the Acacia N derived from atmospheric N₂ (N_dfa) to be calculated. Results show variation in leaf δ ¹⁵N between A. senegal and the reference plant and among years. The relative δ ¹⁵N values (‰) were higher in B. aegyptiaca than in the N₂-fixing acacia provenances. Provenances originally collected from clay soils fixed little N in the first year, but the amount fixed increased as the trees aged. All provenances showed a decrease in δ ¹⁵N with age. The N_dfa varied between 24% (Mazmoom provenance) and 61% (Rahad provenance) 4 years after planting. There was no significant difference in δ ¹⁵N between provenance groups based on soil type or rainfall at original growing site. The amount of N_dfa increased significantly with age in all provenances. The above-ground contribution of fixed N to foliage growth in a 4-year-old A. senegal was highest in the Rahad sand–soil provenance (46.7 kg N ha⁻¹) and lowest in the Mazmoom clay-soil provenance (28.7 kg N ha⁻¹). Our study represents the first use of the δ ¹⁵N method for estimating the N input by A. senegal to the clay plain soils of the gum belt in the Sudan.     

Estimation of symbiotically fixed nitrogen in field grown soybeans: An application of natural15N/14N abundance and a low level15N-tracer technique

Summary Plants from agricultural and natural upland ecosystem were investigated for¹⁵N content to evaluate the role of symbiotic N₂-fixation in the nitrogen nutrition of soybean. Increased yields and lower δ¹⁵N values of nodulating soybeansvs, non-nodulating isolines gave semi-quantitative estimates of N₂ fixation. A fairly large discrepancy was found between estimations by δ¹⁵N and by N yield at 0 kg N/ha of fertilizer. More precise estimates were made by following changes in plant δ¹⁵N when fertilizer δ¹⁵N was varied near¹⁵N natural abundance level. Clearcut linear relationships between δ¹⁵N values of whole plants and of fertilizer were obtained at 30 kg N/ha of fertilizer for three kinds of soils. In experimental field plots, nodulating soybeans obtained 13±1% of their nitrogen from fertilizer, 66±8% from N₂ fixation and 21±10% from soil nitrogen in Andosol brown soil; 30%, 16% and 54% in Andosol black soil; 7%, 77% and 16% in Alluvial soil, respectively. These values for N₂ fixation coincided with each corresponding estimation by N yield method. Other results include: 1)¹⁵N content in upland soils and plants was variable, and may reflect differences in the mode of mineralization of soil organics, and 2) nitrogen isotopic discrimination during fertilizer uptake (δ¹⁵N of plant minus fertilizer) ranged from −2.2 to +4.9‰ at 0–30 kg N/ha of fertilizer, depending on soil type and plant species. The proposed method can accurately and relatively simply establish the importance of symbiotic nitrogen fixation for soybeans growing in agricultural settings.     

Trophic ecology of Pacific salmon (<Emphasis Type="Italic">Oncorhynchus</Emphasis> spp.) in the ocean: a synthesis of stable isotope research

Increasing interest in the marine trophic dynamics of Pacific salmon has been motivated by the recognition of their sensitivity to changing climate and to the competitive effects of hatchery fish on wild stocks. It has become more common to use stable isotopes to supplement traditional diet studies of salmon in the ocean; however, there have been no integrated syntheses of these data to determine whether stable isotope analyses support the existing conventional wisdom of feeding strategies of the Pacific salmon. We performed a meta-analysis of stable isotope data to examine the extent of trophic partitioning among five species of Pacific salmon during their marine lives. Pink, sockeye, and chum salmon showed very high overlap in resource use and there was no consistent evidence for chum relying on alternative food webs dominated by gelatinous zooplankton. δ¹⁵N showed that Chinook and coho salmon fed at trophic levels higher than the other three species. In addition, these two species were distinctly enriched in ¹³C, suggesting more extensive use of coastal food webs compared to the more depleted (pelagic) signatures of pink, sockeye, and chum salmon. This paper presents the first synthesis of stable isotope work on Pacific salmon and provides δ¹⁵N and δ¹³C values applicable to research on the fate of the marine derived nutrients these organisms transport to freshwater and riparian ecosystems.     

Estimating the uptake of traffic-derived NO2 from 15N abundance in Norway spruce needles

The ¹⁵N ratio of nitrogen oxides (NO_x) emitted from vehicles, measured in the air adjacent to a highway in the Swiss Middle Land, was very high [δ¹⁵N(NO₂) = +5.7‰]. This high ¹⁵N abundance was used to estimate long-term NO₂ dry deposition into a forest ecosystem by measuring δ¹⁵N in the needles and the soil of potted and autochthonous spruce trees [Picea abies (L.) Karst] exposed to NO₂ in a transect orthogonal to the highway. δ¹⁵N in the current-year needles of potted trees was 2.0‰ higher than that of the control after 4 months of exposure close to the highway, suggesting a 25% contribution to the N-nutrition of these needles. Needle fall into the pots was prevented by grids placed above the soil, while the continuous decomposition of needle litter below the autochthonous trees over previous years has increased δ¹⁵N values in the soil, resulting in parallel gradients of δ¹⁵N in soil and needles with distance from the highway. Estimates of NO₂ uptake into needles obtained from the δ¹⁵N data were significantly correlated with the inputs calculated with a shoot gas exchange model based on a parameterisation widely used in deposition modelling. Therefore, we provide an indication of estimated N inputs to forest ecosystems via dry deposition of NO₂ at the receptor level under field conditions. Received: 7 November 1997 / Accepted: 16 September 1998     

δ15N of forest soil and understorey vegetation reflect the former agricultural land use

Since the middle of the 19th century, the area covered by forests in France has doubled. These new forests grow on previous agricultural lands. We have studied the influence of this agricultural history on the ¹⁵N abundance of present-day forests planted on farmlands in the Vosges mountains (north-eastern France) between 1898 and 1930. Different types of land use were identified from old cadastres (1814–1836) of 16 farms. Ancient forests adjacent to farmlands were used as controls. Former pastures, meadows, croplands, gardens and ancient forests were compared for soil δ¹⁵N (fraction <50 μm and total soil), C/N, P and N content and fern (Dryopteris carthusiana) δ¹⁵N. The mean δ¹⁵N of soil increased in the order ancient forests (+0.0‰)<pastures (+1.4‰)<croplands (+1.6‰)<meadows (+2.5‰)<gardens (+3.8‰). This increase in soil δ¹⁵N with the intensity of former land use was related to the former input of ¹⁵N-enriched manure, and to an activation of soil nitrification leading to ¹⁵N-depleted nitrate export on previously manured parcels. Fern δ¹⁵N increased in the same order as soil δ¹⁵N in relation to past land use. The mean δ¹⁵N of fern in ancient forests (–4.4‰) and former pastures (–3.4‰) was 5‰ lower than soil δ¹⁵N and the two variables were strongly correlated. The δ¹⁵N of fern in formerly manured parcels varied little (cropland: –2.7‰, meadows: –2.6‰ and gardens: –2.2‰) and independently of soil δ¹⁵N, suggesting that the soil sources of fern N differed between unmanured and manured parcels. Understorey plant δ¹⁵N and soil δ¹⁵N appear to be excellent tracers of previous land use in forests, and could be used in historical studies. The persistence of high isotopic ratios in previously manured parcels, almost a century after afforestation, suggests a long-term influence of former land use on the N cycle in forest soils. Received: 22 January 1999 / Accepted: 22 July 1999     

Water sources accessed by arid zone riparian trees in highly saline environments,Australia

The flow regimes of arid zone rivers are often highly variable, and shallow groundwater in the alluvial aquifers can be very saline, thus constraining the availability and quality of the major water sources available to riparian trees—soil water, shallow groundwater and stream water. We have identified water sources and strategies used by riparian trees in more highly saline and arid conditions than previously studied for riparian trees of arid zone rivers. Our research focused on the riparian species Eucalyptus coolabah, one of the major riparian trees of ephemeral arid zone rivers in Australia. The water sources available to this riparian tree were examined using δ¹⁸O isotope data from xylem, soil water, groundwater and surface water. Additionally, soil chloride and matric potential data were used to infer zones of water availability for root uptake. Despite the saline conditions, the trees used a mixture of soil water and groundwater sources, but they did not use surface water directly. The study identified three strategies used to cope with typically high groundwater and soil water salinities. Firstly, the trees preferentially grow in zones of most frequent flushing by infiltrating streamflow, such as the bank-tops of channels. Secondly, the trees limit water use by having low transpiration rates. Thirdly, the trees are able to extract water at very low osmotic potentials, with water uptake continuing at chloride concentrations of at least 20,000–30,000 mg L⁻¹.