HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (6) Citing Articles via Google Scholar Google Scholar Articles by Dise, N. B. Articles by MacDonald, J. Search for Related Content PubMed PubMed Citation Articles by Dise, N. B. Articles by MacDonald, J. Agricola Articles by Dise, N. B. Articles by MacDonald, J. Related Collections Surface Water Quality Water Quality Soil Pollution Water Pollution

TECHNICAL REPORT
Surface Water Quality

Aluminum Output Fluxes from Forest Ecosystems in Europe

A Regional Assessment

^a Dep. of Earth Sciences, The Open Univ., Milton Keynes, MK7 6AA, UK
^b Dep. of Soil Ecology, Bayreuth Institute for Terrestrial Ecosystem Research (BITÖK), Univ. of Bayreuth, D-95440 Bayreuth, Germany

* Corresponding author (n.b.dise{at}open.ac.uk)

Received for publication May 4, 2000.

	ABSTRACT

TOP NOTES ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Data from 89 forested catchments and plots across Europe were used to define empirical relationships between aluminum leaching and input fluxes of major ions, output fluxes of major ions, ecosystem parameters such as soil pH, and combinations of these. Forests that release dissolved Al to seepage or surface waters are located primarily in areas receiving the highest loading of acid rain, and the output flux of Al shows the highest correlations to the throughfall flux of inorganic nitrogen, the output fluxes of NO^-₃, H⁺, and SO^2-₄, and the mineral soil pH. If the speciation of Al is taken to be Al³⁺ (an overestimate), Al is released in a nearly 1:1 molar charge ratio with the sum of NO^-₃ and SO^2-₄ in runoff or seepage water over a wide range of base-poor bedrock types and acid deposition across Europe. The empirical data point to a threshold range of N deposition of 80 to 150 mmol_c N m^-2 yr^-1 and a (less clearly defined) range of S deposition of 100 to 200 mmol_c SO^2-₄ m^-2 yr^-1 above which Al released from forests exceeds 100 mmol_c Al m^-2 yr^-1. Within this threshold range, the sites that release little or no dissolved Al are those that continue to assimilate input N and/or have high soil pH (>4.5).

	INTRODUCTION

TOP NOTES ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
ALUMINUM IN SOLUTION is one of the most important ions determining the functioning of terrestrial and aquatic ecosystems. The toxic effects of ionic aluminum on aquatic and terrestrial biota are well documented (e.g., Baker and Schofield, 1982; Verbost et al., 1995; Cronan and Grigal, 1995) and have received increased attention since the recognition of the effects of acidic deposition on surface waters in the 1970s.

In surface water, ionic Al can be toxic to fish, invertebrates, and a wide variety of other freshwater organisms. In forests, Ulrich (1995) and Matzner and Murach (1995) concluded that high levels of aluminum ions reduce the fine root biomass of trees in relation to the aboveground biomass. In addition, the roots in soil with high ionic Al are often concentrated in the upper organic layer (where aluminum is complexed), which may be only a few centimeters thick. Such a shallow root distribution can lead to drought stress and greatly increased susceptibility to wind throw. Because of these effects, aluminum was hypothesized to be a major agent of forest damage in acid deposition–affected forests in central Europe (Ulrich, 1995). Other potential effects of Al in the soil include interfering with ion transport through root hairs and reducing diversity of soil biota.

Elevated soil solution Al is an effect of prolonged soil acidification; the Al concentration of rainfall and snowmelt is nearly always extremely low. Instead, Al is released from the soil if the pH of the soil solution declines below about 4.5. The so-called "Al-buffer-range" (Ulrich, 1981) is characterized by the release of Al ions into the soil solution as an effect of H⁺ buffering.

The Al chemistry of soils and soil solutions and its regulation by solid phase properties have been addressed in several papers (e.g., Dahlgren et al., 1989; Mulder and Stein, 1994; Berggren and Mulder, 1995). Aluminum is most toxic in its reactive ionic forms: Al⁺₂, Al(OH)²⁺, and Al³⁺. In general, the higher the charge, the more toxic the species. Monomeric aluminum, Al³⁺, frequently dominates in low-pH soils containing little organic matter. Although upper soil horizons may have high concentrations of total aluminum, this is often complexed with organic matter so the (toxic) reactive aluminum concentration is low.

Once in the soil solution, aluminum ions are transported with seepage to deeper layers or to ground waters and surface waters. In contact with solutions of pH > 5, Al³⁺ ions can act as cation acids by the precipitation of aluminum hydroxides and the related formation of H⁺ (Dahlgren et al., 1989). Thus, beside the direct toxic effects to biota, the transport of aluminum ions with seepage to deeper soil layers and aquatic systems and their acidification is a major detrimental effect of Al.

The modeling and prediction of Al output from acid soils has received attention in numerous papers. So far, modeling has followed a deterministic approach using equilibrium reactions of H⁺ and soil solids with a given pH–pAl dependence. For mineral soils gibbsite dissolution often is used. More recently it was shown that the release of organically bound Al may regulate the soil solution concentration, especially in very acid soils (Berggren and Mulder, 1995; Matzner et al., 1998).

In general, prediction of the H⁺ and Al concentrations in soil solutions using equilibria approaches is very difficult because of the uncertainty related to the thermodynamic constants used (Schecher and Driscoll, 1987). Furthermore, disequilibria of soil solutions with respect to defined Al-containing solids has often been found (Matzner and Prenzel, 1992; Franken et al., 1995) as well as long-term changes in the Al chemistry (Mulder and Stein, 1994). Nevertheless, deterministic models predicting Al output are used today on both local and regional scales (e.g., MAGIC, SAFE, SMART), including the calculation of critical loads for acidic deposition (deVries et al., 1994; Warfinge and Sverdrup, 1992).

Here we follow a different approach. By using long-term data on fluxes of Al and other mineral ions from a large number of case studies, we define empirical relationships between ecosystem parameters and the input of ions from atmospheric deposition on the one side, and Al output with seepage and runoff on the other. Such relationships have already been developed for NO^-₃ output, where it has been shown that N input, the C to N ratio of the forest floor, and the mineral soil pH were largely able to explain the regional variation of NO^-₃ output (Dise et al., 1998a).

Considering the clear relationship between soil acidity and soil solution Al, and the demonstrated toxic affects of Al on freshwater biota, the challenge is to assess to what extent Al leaching is already initiated across European forests, and to what extent this is related to the pattern of acid (N and S) deposition and ecosystem changes in soil processes. In this paper we analyze a database of 89 forested plots and catchments across Europe to establish the regional pattern of Al output from these forests and relate output patterns to other ecosystem parameters such as the level of acid deposition, the soil chemistry, and the input and output fluxes of other ions. The empirical relations between Al output and rates of atmospheric deposition might also be of use for the evaluation of critical loads. Our hypothesis is that sites that release Al to seepage or runoff waters are those that receive the highest loading of acidifying ions (SO^2-₄, NO^-₃, NH⁺₄), release the largest levels of these ions (SO^2-₄, NO^-₃), and/or show suppressed soil pH and other symptoms of soil acidification.

	METHODS

TOP NOTES ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
A subset of the Indicators of Forest Ecosystem Functioning (IFEF) database (Dise et al., 1998b) of European forested plots and catchments is used for this study. The database was compiled from published reports and contacting individual research groups across Europe (Table 1). The database has average annual fluxes of major mineral elements with throughfall, seepage, and runoff, as well as various ecosystem characteristics (e.g., chemistry of soil and vegetation) and site information (e.g., management history, average precipitation). Not all information is available for all sites.

View larger version (16K): [in this window] [in a new window]   Fig. 4. 3 x Altot flux in seepage water or runoff versus the B-horizon soil pH (in H2O).

In order to test these hypotheses, we need to assume a charge on the Al ions. We do not have information on Al species from any of the sites, and the Al_out in our database is total dissolved Al. In this study we assign the Al ions a charge of 3, bearing in mind that this will be an overestimate of the actual charge for many sites, and may be a major overestimate in sites with highly organic soils. In general, the more acid sites will be expected to have a high proportion of ionic Al as Al³⁺ (Cronan, 1994). Fluxes are thus calculated as 3 x (Al_tot) and expressed either as that or as mmol_c Al m^-2 yr^-1 (again with the recognition that this is an upper estimate).

We consider output fluxes of >100 mmol_c Al m^-2 yr^-1 (9 kg Al ha^-1 yr^-1) as the threshold value for initiation of the risk of soil acidification and potential ecosystem degradation as described in the introduction of this paper. The exact value is arbitrary, but the approximate value of 7 to 15 kg Al ha^-1 yr^-1 in streamwater or leachate differentiated in the ALBIOS study (Aluminum in the Biosphere; Cronan, 1994) between sites in which some effects of ecosystem acidification (e.g., base saturation <10–15%) were apparent and those that were relatively unaffected by acidification. We also in this paper define sites that leach <35 mmol_c NO₃ m^-2 yr^-1 (<5 kg NO₃–N ha^-1 yr^-1) as N retaining, regardless of the N input.

We use regression analyses to identify significant relationships between Al_out and other input, output, or catchment variables, and to identify variables that show a similar relationship to Al_out (e.g., slopes of regression lines statistically identical). This is useful for testing hypotheses about the conditions under which forests begin to release ionic Al. In addition to this, the regression models using input fluxes or ecosystem variables can potentially be used as predictors of Al_out flux; but in this case the assumptions of regression analyses (normality of the data; random scatter of residuals) must be more strictly adhered to. We do not attempt any predictions of Al fluxes in this paper. Potential nonlinear relationships are examined by logarithmic or exponential transformation of variables. In some cases we consider subsets of the data (plots only, conifer sites only, similar bedrock) to determine if they behave differently with respect to Al_out from the sites considered collectively.

Our database contains both plot and catchment sites. It may be expected that the chemical and physical fates of some ions of biogeochemical interest would be different between plots and catchments. In particular, flowpaths in catchments tend to be longer than in plots, allowing more time for biological removal of nitrogen and buffering reactions on acids and Al. Furthermore, Al complexation in peaty layers and processes in riparian zones such as SO^2-₄ reduction and denitrification increase solution alkalinity and reduce Al_out fluxes in catchments as compared with plot studies.

However, in our study, we were not able to draw any inferences about the behavior of plots versus catchments per se. The reason is that plots and catchments tend to be segregated by location and receive different amounts of acid deposition, both of which confound the detection of trends based simply on their morphology. Catchments were dominant in areas with higher relief, especially Scandinavia and southern Europe, and these received on average one-third the inorganic N (see Fig. 3a) and one-half the SO^2-₄ deposition (Fig. 3b) of the plots, which dominated in central Europe. (An exception to this is six catchments located in the Czech Republic, which received intermediate to high N and S deposition, similar to that of many of the plots. Some of these sites also released relatively high amounts of Ca in streamwater, as described earlier.)

View larger version (15K): [in this window] [in a new window]   Fig. 3. (a) 3 x Altot flux in seepage water or runoff versus the combined flux in throughfall. (b) 3 x Altot flux in seepage water or runoff versus SO2-4 flux in throughfall. (c) 3 x Altot flux in seepage water or runoff versus the combined flux in throughfall.

	RESULTS

TOP NOTES ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Overall Distribution of Aluminum Output
Forests that release the largest amounts of Al to seepage or runoff are predominantly located across northwestern Germany (up to 590 mmol_c Al m^-2 yr^-1) and the Netherlands (210–870 mmol_c Al m^-2 yr^-1), with lower Al output in sites located in eastern Germany (up to 480 mmol_c Al m^-2 yr^-1), southwestern Sweden (up to 340 mmol_c Al m^-2 yr^-1), western Denmark (up to 270 mmol_c Al m^-2 yr^-1), and southern Norway (up to 170 mmol_c Al m^-2 yr^-1) (Fig. 1). The lowest levels of Al output (<130 mmol_c Al m^-2 yr^-1) occur in sites located in the rest of Scandinavia, Wales, Ireland, the Czech Republic, and southern Germany.

View larger version (104K): [in this window] [in a new window]   Fig. 1. 3 x Altot flux in seepage water or runoff survey sites; data are the average of at least one year of regular measurement. Note: sites are located at the center of the column.

View larger version (16K): [in this window] [in a new window]   Fig. 2. (a) 3 x Altot flux in seepage water or runoff versus NO-3 flux in seepage water or runoff. (b) 3 x Altot flux in seepage water or runoff versus SO2-4 flux in seepage water or runoff. (c) 3 x Altot flux in seepage water or runoff versus the combined flux in seepage water or runoff.

View this table: [in this window] [in a new window]   Table 3. Single output flux variables significantly correlated to Al output flux. Fluxes expressed as mmolc m-2 yr-1.

Well after NO^-₃, the output ions that correlated next best to the output flux of Al are SO^2-_{4 out} and H⁺_out (r² = 0.42 with N = 84 and r² = 0.46 with N = 44, respectively, Table 3). Transforming the data does not improve the fit. Although many sites release Al in a roughly molar equivalent proportion to SO^2-₄, there is considerable scatter from sites that release relatively high amounts of SO^2-₄ (>100 mmol_c SO^2-₄ m^-2 yr^-1), but retain Al almost completely (Fig. 2b). These sites are sulfate-saturated but either retain nitrogen, leach high amounts of calcium, or both. All three of the largest outliers (leaching >400 mmol_c SO^2-₄ m^-2 yr^-1 and retaining nearly all Al) leach high levels of Ca (370–800 mmol_c Ca²⁺ m^-2 yr^-1), suggesting a local strong source of Ca buffering. One of these sites is also strongly N retaining (N_out = 31 mmol_c N m^-2 yr^-1). Although several other output ions (notably NH⁺_{4 out} and K⁺_out) show significant relationships to Al_out, none other than NO^-_{3 out}, SO^2-_{4 out}, and H⁺_out show a higher regression coefficient than 0.2 (Table 3).

The best-fit two-variable regression of output ions to Al_out combines NO^-_{3 out} and H⁺_out (R² = 0.79, N = 43, p < 0.0001). However, with nearly the same R² value, and almost twice the number of sites, the regression combining NO^-_{3 out} and SO^2-_{4 out} is arguably a stronger relationship:

Considering only the plots (which overall receive higher acid deposition), and those sites not suspected of a local source of calcium, the slope of the relationship between Al_out and the combined fluxes of NO^-_{3 out} and SO^2-_{4 out} is slightly less than 1 (Fig. 2c):

Thus, equivalents of Al are released in a maximum ratio of 0.9:1.0 with equivalents of SO^2-₄ and NO^-₃ in the sites that are most influenced by acid deposition. This ratio is a maximum estimate because we overestimate the Al³⁺ flux by not accounting for other Al species of lower charge. However, for most of these fairly to highly acid sites we expect the ionic aluminum to be dominated by Al³⁺ (Cronan, 1994).

The highest proportion of variability in output Al flux for all sites together (plots + catchments) can be explained by combining the output fluxes of NO^-₃, SO^2-₄, and H⁺:

Throughfall Fluxes Related to Aluminum Output Flux
The best regression fit for Al_out using input fluxes is with the total throughfall flux of inorganic N , hereafter referred to as N_in (Fig. 3a, Table 4). Slightly poorer, but still highly significant relationships occur between Al_out and the input of the two inorganic nitrogen ions (NO^-₃, NH⁺₄) considered separately (Table 4). The relationship between Al_out and throughfall input of nitrogen shows a broadly positive trend with wide scatter (Fig. 3a). Fitting a log relationship does not improve the proportion of variability in Al_out explained. Al_out shows poorer relationships to SO^2-_{4 in} (Fig. 3b) or (N_in + SO^2-_{4 in}) (Fig. 3c) because there are a number of sites that receive high levels of SO^2-₄ (and high combined levels of N + SO^2-₄) but release low levels of Al (<100 mmol_c m^-2 yr^-1). These include the six Czech catchments discussed in the introduction of this paper.

View this table: [in this window] [in a new window]   Table 4. Single input flux variables significantly correlated to Al output flux. Fluxes expressed as mmolc m-2 yr-1.

The most significant two-variable regression of input ions on Al output combined NO^-₃ flux in throughfall and NH⁺₄ flux in throughfall:

The only other combination of two input ions that was more significant than the ions alone was with K⁺_in and SO^2-_{4 in} (R² = 0.39, N = 71, p < 0.004).

Ecosystem Characteristics Related to Aluminum Output Flux
Of the ecosystem characteristics we considered (Table 2), the strongest correlation to Al_out was the B horizon soil (H⁺) (r² = 0.37; Table 5) followed by the C horizon (H⁺) (r² = 0.32) and the A horizon (H⁺) (r² = 0.17). Another way of considering the relationship is to relate the log of Al_out to the soil pH (Table 5). The regional data show a broadly negative relationship between subsoil pH and Al_out (Fig. 4). Sites with relatively high-pH soils (pH > 4.5) in general release less Al than sites with relatively low-pH soils. Sites that release the highest levels of Al (Al_out > 300 mmol_c m^-2 yr^-1) all have soil B horizon pH < 4.5. Considering only plots improves this relationship significantly (Table 5, Fig. 4), in large part because of the removal of several catchments that appear to be influenced by either a ground water Ca²⁺ source or highly organic soils.

View this table: [in this window] [in a new window]   Table 5. Ecosystem variables significantly correlated to Alout flux (concentrations in µmolc L-1). Fluxes expressed as mmolc m-2 yr-1.

Related to soil pH, there is some suggestion that bedrock type modifies the dominant effect of acid input on Al_out, although data are limited. For any given input of acid precursor ions (N_in + SO^2-_{4 in}), the flux of Al from plots dominated by loamy sands and marls is generally (but not always) lower that from poorer sandstones and metamorphic rocks, which is in turn generally (but not always) lower than that from more resistant igneous rocks (Fig. 5a). (We considered plots only in this analysis as we were more confident that the bedrock type described from plots related directly to the seepage chemistry described).

View larger version (20K): [in this window] [in a new window]   Fig. 5. (a) 3 x Altot flux in seepage water versus the combined (NO-3 + NH+4 + SO2-4) flux in throughfall; data divided into major bedrock classes. Plots only. (b) 3 x Altot flux in seepage water versus the combined (NO-3 + SO2-4) flux in seepage water; data divided into major bedrock classes. Plots only.

	DISCUSSION

TOP NOTES ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Analyzing regional trends from data compilations is inevitably subject to high uncertainty. In particular, the data we use come from different studies and time periods, often using different methods for different purposes. The quality of data and local effects such as amounts of dust deposition, site history, ground water contribution to runoff, or deposition history cannot be assessed in detail. Despite this, when a large number of case studies are considered together, patterns emerge that may be due to ecosystem responses to a limited number of dominant processes (e.g., Dise et al., 1998a; Gundersen et al., 1998; deVries et al., 1999). Local or site-specific effects then become part of the unexplained variation in the data.

In the database the overall amounts of Al released to deeper soil layers or to the aquatic environment can be high: in 40% of the sites it exceeds 100 mmol_c m^-2 yr^-1 and in highly polluted areas it can exceed 500 mmol_c m^-2 yr^-1 (Fig. 1 and 3). In addition, there is a clear pattern of enhanced Al output across areas, such as central Germany and the Netherlands (Fig. 1), which have been identified as receiving the highest loads of acid deposition (EMEP, 1994). The areas releasing intermediate amounts of Al (western Denmark, southwestern Sweden, southern Norway) receive intermediate amounts of acid deposition, and those releasing little or no Al receive the lowest levels of acid deposition. This is supported by the regressions, which show that a large part of the variability in the output flux of Al across European forests is explained by the input fluxes of the acid-precursor ions NO^-₃, NH⁺₄, and, to a lesser extent, SO^2-₄ (Table 4, Fig. 3). Thus, the first result from this study is the demonstration that, across Europe, Al is released preferentially from forests receiving acid (especially N) deposition.

Furthermore, the correlations demonstrated between Al_out and the mobile anions NO^-_{3 out}, SO^2-_{4 out}, and H⁺_out (Fig. 2, Table 3), and Al_out and soil pH (Fig. 4, Table 5), indicate that Al is released preferentially from sites that not only receive acid deposition, but are acidified. The high correlations between the output fluxes of these acid ions is not due to autocorrelation with water fluxes, since there is no relationship between output fluxes of Al and discharge, or Al_out and conservative ions such as Cl^- (Table 3). Indeed, some of the sites with the highest discharge (e.g., in Scandinavia) have the lowest Al fluxes. Overall, the strong linear relationship between Al_out and the combined fluxes of SO^2-_{4 out} and NO^-_{3 out} (Fig. 2c), across a variety of bedrock types (Fig. 5b), suggest that buffer mechanisms other than Al release are only of minor importance in the base-poor soils included in our data, especially in sites that are influenced by acid deposition.

Relationships between high NO^-₃ and SO^2-₄ fluxes and high Al fluxes in seepage and runoff have been shown in individual catchment or plot studies over the course of a season (especially during spring runoff or high-flow acid pulses) (e.g., McAvoy, 1989) and for concentrations rather than fluxes (deVries et al., 1999); we now demonstrate this as a regional pattern across European forests. Overall, these findings confirm the basic concepts of the effects of acid deposition on soil chemistry and, especially, on the release of Al (Reuss, 1991; Ulrich and Sumner, 1991).

The relation of Al_out to the input of different chemical ions in throughfall (Table 4, Fig. 3) showed that of all the major ions, the input of inorganic N showed the highest relationship to the output flux of ionic aluminum. The threshold value of N deposition between sites that overwhelmingly retain Al (as defined as leaching <100 mmol_c m^-2 yr^-1) and sites that overwhelmingly leach Al is 80 to 150 mmol_c N m^-2 yr^-1 (11–21 kg N ha^-1 yr^-1) (Fig. 3a). Within that range, sites that show higher N retention, higher S retention, and/or higher soil pH have lower losses of Al.

There is no other input ion, combination of input ions, or site characteristic explored that so clearly differentiates between sites that are and are not at current risk of acidification (as defined in this paper by release of Al > 100 mmol_c m^-2 yr^-1). In particular, total inputs of SO^2-₄, or of the acid precursor ions (Fig. 3b,c) showed much poorer relationships to Al_out. Why nitrogen input in particular appears to be such an important factor for Al leaching is not clear. It may be that N saturation and resulting acidification is more closely related to overall forest "health" than S saturation, which is widespread across forest ecosystems in Europe. Determining the processes underlying these relationships should be a priority of future research in this area.

Mean regional critical loads on a European scale have been calculated from deterministic models as 48 mmol_c m^-2 yr^-1 for N, 77 mmol_c m^-2 yr^-1 for S, and 80 mmol_c m^-2 yr^-1 for acidity (deVries et al., 1994). The empirical data point to a threshold level of N deposition of 80 to 150 mmol_c N m^-2 yr^-1 and a (less clearly defined) threshold level of S deposition of 100 to 200 mmol_c SO^2-₄ m^-2 yr^-1, above which Al release from forests exceeds 100 mmol_c Al m^-2 yr^-1. Considering that critical loads are intended to protect ecosystems over the long term and therefore have a built-in protection aspect, the data from our study match the average given by deVries et al. (1994) quite well. However, our results suggest that N + S retention, rather than base cation buffering, is the most important factor for distinguishing whether forests on base-poor soils subject to acid deposition are acidified and release dissolved aluminum to seepage or runoff.

	CONCLUSIONS

TOP NOTES ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
The analyses of the database of 89 forested plots and catchments across Europe suggest the following:

(i) Forests that leach the highest amount of dissolved ionic aluminum are located in central Europe, primarily Germany, the Netherlands, and the western Czech republic. A scarcity of data from other central European countries, such as Poland, Russia, and the Baltic states, as well as France and southern Europe, does not allow us currently to make inferences about these regions.

(ii) Among input or site variables, the output flux of aluminum is most closely correlated to the input flux in throughfall of inorganic nitrogen (Fig. 3a, Table 4), and then to the mineral soil pH (Fig. 4, Table 5).

(iii) If a forest leaches NO^-₃ at levels >100 mmol_c m^-2 yr^-1 it is also highly likely to leach Al at levels >100 mmol_c m^-2 yr^-1 (Fig. 2a, Table 3). The same does not necessarily hold for sites leaching SO^2-₄ (Fig. 2b).

(iv) Over a broad range of sites and base-poor bedrocks, equivalents of Al are released in a ca. 0.9:1.0 ratio with equivalents of NO^-₃ + SO^2-₄ (Fig. 2c).

(v) There is some suggestion that soil parent material mitigates Al release (Fig. 5a); however, this is because the richer sites have higher nitrogen and sulfur retention (Fig. 5b) rather than significantly higher buffering capacity due to weathering of base cations.

	ACKNOWLEDGMENTS

 
Financial support for this work was provided by the German Ministry for Education, Science, Research and Technology (BMBF, Grant BEO-51-0339476) to N.B. Dise while working as a guest scientist at BITÖK, University of Bayreuth, and by the Open University Research Development Fund to J. MacDonald. We also thank the many researchers across Europe who have contributed to the IFEF database.

	NOTES

TOP NOTES ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
M. Armbruster, current address: Technical Institute of Dresden, Postfach 1117, D-01735 Tharandt, Germany. J. MacDonald, current address: Scottish Environmental Protection Agency (SEPA), Stirling, FK94TR, UK.

	REFERENCES

TOP NOTES ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 

• Baker, J.P., and C.L. Schofield. 1982. Aluminum toxicity to fish in acidic waters. Water Air Soil Pollut. 18:289–309.[ISI]
• Berggren, D., and J. Mulder. 1995. The role of organic matter in controlling Al solubility in acidic mineral soil horizons. Geochim. Cosmochim. Acta 59:4167–4180.
• Bredemeier, M., K. Blank, J.-Y. Xu, A. Tietema, A.W. Boxman, B. Emmet, F. Moldan, P. Gundersen, P. Schleppi, and R.F. Wright. 1998. Input–output balances at the NITREX sites. For. Ecol. Manage. 101:57–64.
• Cronan, C.S. 1994. Aluminum biogeochemistry in the ALBIOS forest ecosystems: The role of acidic deposition in aluminium cycling. p. 51–81. In D.L. Godbold and A. Hütterman (ed.) Effects of acid rain on forest processes. Wiley–Liss, New York.
• Cronan, C.S., and D.F. Grigal. 1995. Use of Ca/Al ratios as indicators of stress in forest ecosystems. J. Environ. Qual. 24:209–226.[Abstract/Free Full Text]
• Dahlgren, R.A., C.T. Driscoll, and D.C. McAvoy. 1989. Aluminium precipitation and dissolution rates in Spodosol B horizons in the northeastern USA. Soil Sci. Soc. Am. J. 53:1045–1052.[Abstract/Free Full Text]
• DeVries, W., G.J. Reinds, H.D. Deelstra, J.M. Klap, and E.M. Vel. 1999. Soil solution chemistry. Chapter 8. In Intensive monitoring of forest ecosystems in Europe. Forest Intensive Monitoring Coordinating Inst. (FIMCI) EC-UNECE, Brussels, Belgium.
• DeVries, W., G.J. Reinds, and M. Posch. 1994. Assessment of critical loads and their exceedance on European forests using a one layer steady state model. Water Air Soil Pollut. 72:357–394.
• Dise, N.B., E. Matzner, and M. Forsius. 1998a. Evaluation of organic horizon C:N ratio as an indicator of nitrate output in conifer forests across Europe. Environ. Pollut. 102:453–456.
• Dise, N.B., E. Matzner, and P. Gundersen. 1998b. Synthesis of nitrogen pools and fluxes from European forest ecosystems. Water Air Soil Pollut. 105:143–154.
• EMEP. 1994. Transboundary acidifying pollution in Europe: Calculated fields and budgets 1985–1993. EMEP/MSC-W Rep. 1/94 [Online]. Available at http://www.emep.int/ladm.html (verified 21 Feb. 2001).
• Forsius, M., S. Kleemola, M. Starr, and T. Ruoho-Airola. 1995. Ion mass budgets for small forested catchments in Finland. Water Air Soil Pollut. 79:19–38.
• Fottova, D., and I. Skorepova. 1998. Changes in mass elements fluxes and their importance for critical loads: Geomon network, Czech Republic. Water Air Soil Pollut. 105:365–376.
• Franken, G., M. Pijpers, and E. Matzner. 1995. Al chemistry of the soil solution in an acid forest soil as influenced by percolation rate and soil structure. Eur. J. Soil Sci. 46:613–619.
• Gundersen, P. 1995. Nitrogen deposition and leaching in European forests—Preliminary results from a data compilation. Water Air Soil Pollut. 85:1179–1184.
• Gundersen, P., B.A. Emmett, O.J. Kjonaas, C.J. Koopmans, and A. Tietema. 1998. Impact of nitrogen deposition on nitrogen cycling in forests: A synthesis of NITREX data. For. Ecol. Manage. 101: 37–55.
• Hauhs, M., K. Rost-Siebert, G. Raben, T. Paces, and B. Vigerust. 1989. Summary of European data. p. 5.1–5.37. In J.L. Malanchuk and J. Nilsson. (ed.) The role of nitrogen in the acidification of soils and surface waters. Miljörapport 89 (10). Council of Nordic Ministers.
• Johnson, D.W. 1995. Temporal patterns in beech forest soil solutions: Field and model results compared. Soil Sci. Soc. Am. J. 59:1732–1740.[Abstract/Free Full Text]
• Kleemola, S., and M. Forsius (ed.) 1996. International cooperative programme on integrated monitoring of air pollutant effects on ecosystems. 5th Annu. Rep. The Finnish Environ. 27. Finnish Environ. Inst., Helsinki, Finland.
• Kleemola, S., and M. Forsius (ed.) 1997. International cooperative programme on integrated monitoring of air pollutant effects on ecosystems. 6th Annu. Rep. The Finnish Environ. 116. Finnish Environ. Inst., Helsinki, Finland.
• Kleemola, S., and M. Forsius (ed.) 1999. International cooperative programme on integrated monitoring of air pollutant effects on ecosystems. 8th Annu. Rep. The Finnish Environment 325. Finnish Environ. Inst., Helsinki, Finland.
• Kölling, C. 1999. Air pollution and its effects on Bavarian forests. Results of the element budget calculations at the Bavarian Forest Climate Station 1991–1998 (In German). Berichte aus der Bayerischen Landesanstalt für Wald und Forstwirtschaft 22:1–46.
• Kreutzer, K., C. Beier, M. Bredemeier, K. Blank, T. Cummins, E.P. Farrell, N. Lammersdorf, L. Rasmussen, A. Rothe, P.H.B. de Visser, W. Weis, T. Weis, and Y.-J. Xu. 1998. Atmospheric deposition and soil acidification in five coniferous forest ecosystems: A comparison of the control plots of the EXMAN sites. For. Ecol. Manage. 101:125–142.
• Matzner, E., and D. Murach. 1995. Soil changes induced by air pollutant deposition and their implication for forests in Central Europe. Water Air Soil Pollut. 85:63–73.
• Matzner, E., M. Pijpers, W. Holland, and B. Manderscheid. 1998. Al in soil solutions of forest soils: Influence of water flow and soil Al-pools. Soil Sci. Soc. Am. J. 62:445–454.[Abstract/Free Full Text]
• Matzner, E., and J. Prenzel. 1992. Acid deposition in the German Solling area: Effects on soil solution chemistry and Al-mobilization. Water Air Soil Pollut. 61:221–234.
• McAvoy, D.C. 1989. Episodic response of aluminium chemistry in an acid sensitive Massachusetts catchment. Water Resour. Res. 25:233–240.
• Mulder, J., and A. Stein. 1994. The solubility of Al in acidic forest soils: Long-term changes due to acid deposition. Geochim. Cosmochim. Acta 58:85–94.
• Paces, T. 1985. Sources of acidification in central Europe estimated from elemental budgets in small basins. Nature 315:31–34.
• Reuss, J.O. 1991. The transfer of acidity from soils to surface waters. p. 203–218. In B. Ulrich and M.E. Sumner (ed.) Soil acidity. Springer–Verlag, Berlin, Germany.
• Schecher, W.D., and C.T. Driscoll. 1987. An evaluation of uncertainty associated with Al equilibrium calculations. Water Resour. Res. 23: 525–534.
• Ulrich, B. 1981. Ecological grouping of soils according to their chemical characteristics. (In German.) Z. Pflanzenernähr. Bodenkd. 144 Band Heft3 S. 289–305.
• Ulrich, B. 1995. The history and possible causes of forest decline in central Europe, with particular attention to the German situation. Environ. Rev. 3:262–276.
• Ulrich, B., and M.E. Sumner. 1991. Soil acidity. Springer–Verlag, Berlin, Germany.
• Verbost, P.M., M.H.G. Berntssen, F. Kroglund, E. Lydersen, H.E. Witters, B.O. Rosseland, B. Salbu, and S.E. Wendelaar Bonga. 1995. The toxic mixing zone of neutral and acidic river water: Acute aluminum toxicity in brown trout (Salmo trutta L.). Water Air Soil Pollut. 85:341–346.
• Warfinge, P., and H. Sverdrup. 1992. Calculating critical loads of acid deposition with PROFILE: A steady-state soil chemistry model. Water Air Soil Pollut. 63:119–143.

This article has been cited by other articles:

				K. A. Aronsson and N. G. A. Ekelund Biological Effects of Wood Ash Application to Forest and Aquatic Ecosystems J. Environ. Qual., September 1, 2004; 33(5): 1595 - 1605. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (6) Citing Articles via Google Scholar Google Scholar Articles by Dise, N. B. Articles by MacDonald, J. Search for Related Content PubMed PubMed Citation Articles by Dise, N. B. Articles by MacDonald, J. Agricola Articles by Dise, N. B. Articles by MacDonald, J. Related Collections Surface Water Quality Water Quality Soil Pollution Water Pollution