Letters

Acid Rain and Soil Chemistry

In their article "Acid rain on acid soil:A new perspective" (5 Aug. 1983, p. 520)Edward C. Krug and Charles R. Frinkhighlight the natural processes that gov-ern the chemistry of soil. We agree withtheir thesis that natural soil formation isa major acidifying process in some envi-ronments, including those affected bycontemporary acid rain. Yet we disagreesharply with their position that acid rainis not really significant. They argue thatsome environments that have been af-fected by acid rain would be acidic natu-rally, and therefore no chemical changescan be induced by acid rain. These con-clusions are misleading or misdirected.The evidence and arguments used by

Krug and Frink revolve around two ba-sic themes: (i) the concept of "totalacidity" versus pH, and (ii) the role ofnatural acids, notably organic and car-bonic, in acidifying landscapes. Krugand Frink invoke the concept of "totalacidity" to purportedly show that acidrain is not much different from naturalrain. Total acidity may be a useful pa-rameter in soil chemistry, but is irrele-vant for characterizing acid rain (1). Be-cause carbon dioxide can be continuous-ly absorbed from the atmosphere, theacid capacity (total acidity) of naturalrain is enormous. However, water satu-rated with atmospheric CO2 has a rela-

Table 1. Stream water acidification by naturalmeans (Jamieson Creek) (5) and by acid rain(Falls Brook) (6). Results of the analysis aregiven in milligrams per liter. DOC, dissolvedorganic carbon.

Jamieson Creek Falls BrookItem 23 July 1982, 6 March 1979,

pH, 4.71 pH 4.75

Ca2+ .29 .72Na+ .40 .43K+ .13 .11Mg2+ .05 .19NH4+ .05 .01SO42- 1.0 5.3N03- .03 1.56Cl- .42 .62Al (total) .30 .61SiO2 3.9 2.2DOC 4.8 - 2.5

1424

tively low acid intensity. For example,natural rain at apH of 5.6 has a hydrogenion activity (acid intensity) of 2.5 micro-equivalents per liter. In contrast, acidrain with apH of 4.6 has an acid intensityten times greater, that is, a hydrogen ionactivity of 25 microequivalents per liter.It is true that, if either acid rain ornatural rain is titrated with strong bases,their acid capacity would not be greatlydifferent. This phenomenon, however, isnot germane to the impact that acid rainmay have on a landscape. Total acidity isonly a measure of the potentially avail-able hydrogen ions that would be avail-able if certain carefully defined condi-tions were met. Total acidity is not relat-ed at all to pH, which is the crucialparameter in deciding whether a givenreaction is possible. The heart of the acidrain problem is that rain at pH 4.6 can dothings chemically that rain at pH 5.6cannot. For example, alumina mineralsare essentially insoluble at pH 5.6, butare quite soluble at pH 4.6 (2). Hydrogenion activity, that is, pH, is well known asthe "master variable" in solution chem-istry (3) precisely because it can exertsuch sensitive control over chemical re-actions. The most important aspect ofacid rain then is not its "total acidity"but rather its pH, that is, its acid intensi-ty. In the thermodynamic considerationof chemical reactions, only pH matters;total acidity is of no concern. To contendthat somehow acid rain, which may be asacid as pH 2.4 (4), is as benign as naturalrain at pH 5.6 is not consistent withchemical theory and ecological under-standing.Krug and Frink show how the soil-

forming process produces organic acidsor carbonic acids, or both. Clearly or-ganic compounds exert a powerful effecton the pH and chemistry of podzolicsoils. Yet Krug and Frink project theviewpoint that it is not possible for acidrain to affect the chemistry of a podzolsoil because ostensibly it is already acid.This view is categorically wrong. Thechemical impact of acid rain in a givensituation involves not only the amount ofacid present but also the kind of acidanions present. The kind of acid has a

profound effect on certain chemical reac-tions taking place in the soil zone and inthe waters draining the soil.

This effect can be illustrated by com-paring the chemistry of the waters drain-ing from a natural, unpolluted watershedand a similar one affected by acid rain(Fig. 1). Although both streams are com-parably acidified, the cause and conse-quences of the acidity are entirely differ-ent in the two systems. The JamiesonCreek watershed near Vancouver, Brit-ish Columbia, is a close analog to theHubbard Brook Experimental Forest inNew Hampshire. Both have podzolicsoil regimes imposed on a hilly, glaciat-ed, granitic terrane (5). A conspicuousdifference between the two areas is thegeneral absence of acid rain over theJamieson Creek catchment and the per-vasive presence of acid rain over Hub-bard Brook. The headwater streams atJamieson Creek are acidified by organicacids, as Krug and Frink would predict.A short distance downstream, however,these organic acids are replaced by car-bonate alkalinity, again as Krug and

6

I

.5 5

4

_-- 0 -

'd

- Downstream

100 200 300 400

Drop in stream elevation (m)

Fig. 1. Chemical differences between natural,podzolic drainage unaffected by acid rain [Ja-mieson Creek (0) near Vancouver, BritishColumbia (5)] and a system affected by acidrain [Falls Brook (0) within the HubbardBrook Experimental Forest, New Hampshire(6)]. Chemistry is plotted as a function ofdownstream pathlength, expressed in units ofstream drop. In the upper panel ion balance isdefined as the sum of the strong bases(B+ = Ca2+ + Na+ + Mg2+ + K+) minusthe sum of the strong mineral acids(A- = S042- + NO3- + Cl-). Both streamsare acidic in their initial stages and becomeless so as downstream pathlength increases(lower panel). The pH in the stream affectedby acid rain (Falls Brook) is clearly controlledby strong mineral acids, while in the naturalsystem pH is not controlled by strong mineralacids (upper panel).

SCIENCE, VOL. 225

m

on

Nov

embe

r 25

, 201

4w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

o

n N

ovem

ber

25, 2

014

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

Frink would predict. Dissolved alumi-num in the Jamieson Creek system ispresent almost entirely in the form oforgano-aluminum complexes. A compa-rable stream at Hubbard Brook (6), anarea affected by acid rain, shows thatthroughout its extent pH is controlled bysulfuric and nitric acids, the same acidsthat characterize acid rain. There is nocarbonate alkalinity in the stream sys-tem, and dissolved aluminum is largelyin the form of inorganic complexes,mostly fluorides and hydroxides (6). Inthe Jamieson Creek watershed, hydro-gen ion is supplied to the system entirelyfrom organic and carbonic acids. Bycomparison, in the Hubbard Brook areaat least 50 percent of the hydrogen ionused by the system comes from acid rain(7).Although the concentration of hydro-

gen ion (pH) of these streams is similar,the dominant kinds of acid present aredifferent (Table 1). The ecological conse-quences of this shift in acid type isexemplified by the behavior of aluminumin both systems. At Jamieson Creek,where pH is controlled by weak naturalacids, dissolved aluminum in stream wa-ter is predominantly all organically com-plexed. Organically complexed alumi-num in solution is essentially nontoxic,and fish can live in such waters (8). Inmarked contrast, inorganic aluminumcomplexes, which characterize much ofthe anthropogenically acidified waters ofthe northeastern United States, are le-thal to populations of young fish (8). Theextent of aluminum toxicity to fish isvery much dependent on the speciationof the aluminum, not just the amount (8).To infer then that aluminum mobilizedby acid rain is the same in every respectas aluminum mobilized by organic acidsis misleading.

Finally, Krug and Frink make asser-tions about land use and its long-termeffects on water acidity. The essence oftheir arguments concerns the accumula-tion or depletion of organic acids in thesoil zone as a function of land use. It isnecessary, however, to distinguish be-tween an acid soil and the acidity ofwaters that drain from it. A watershedwith an acid soil is not always associatedwith acid stream water (Fig. 1). Further-more, our data from deforestation ex-periments at Hubbard Brook show thatacidity produced by clear-cutting the for-est is dominantly nitric acid, not organicacid (9). Most important, we have foundthat the chemical effects of deforestationlast only a few years (10), not decades,as inferred by Krug and Frink.

In summary, Krug and Frink ignorethe substance of published chemical data28 SEPTEMBER 1984

from Scandinavia, Canada, and the Unit-ed States which show that stream andlake acidification over large regions aredue to sulfuric and nitric acids, not or-

ganic acids (11). Instead, they concen-

trate on selected, anecdotal evidence or

hypothetical scenarios to make a case

that acid rain hardly matters, when infact the evidence is overwhelmingly tothe contrary (12). Discussions about a

largely peripheral concept such as "totalacidity" or generalizations about naturalacids do not enrich or advance the pres-

ent debate about the acid rain problem(13).

NOYE M. JOHNSONDepartment of Earth Sciences,Dartmouth College,Hanover, New Hampshire 03755

GENE E. LIKENSInstitute of Ecosystem Studies,New York Botanical Garden,Millbrook, New York 12545

MICHAEL C. FELLERFaculty of Forestry,University of British Columbia,Vancouver, Canada V6T I W5

CHARLES T. DRISCOLLDepartment of Civil Engineering,Syracuse University,Syracuse, New York 13210

References and Notes

1. G. E. Likens, N. M. Johnson, J. N. Galloway,F. H. Bormann, Science 194, 643 (1976).

2. The solubility product, K,P, of Al(OH)3 is vari-able depending on the crystalline state of thealumina [see (6)]. However, the amount of alu-minum ion in solution at equilibrium varies asthe third power of hydrogen ion concentration,that is [Al"]J = K%p [H']'.

3. Textbooks of solution chemistry invariably be-gin with chapters on the acid-base aspects ofaqueous solutions and the role of pH. See, forexample, W. Stumm and J. J. Morgan, Aquat-ic Chemistry (Wiley-Interscience, New York,1970).

4. G. E. Likens, R. F. Wright, J. N. Galloway, T.J. Butler, Sci. Am. 241, 43 (October 1979).

5. Geology and precipitation chemistry for theJamieson Creek area and the Hubbard BrookExperimental Forest have been described by L.J. Zeman ["Chemistry of tropospheric falloutand streamflow in a small mountainous water-shed near Vancouver, British Columbia,' the-sis, University of British Columbia, Vancouver(1973)] and by G. E. Likens, F. H. Bormann, R.S. Pierce, J. S. Eaton, and N. M. Johnson[Biochemistry of a Forested Ecosvstem (Spring-er-Verlag, New York, 1977)], respectively.

6. N. M. Johnson et al., Geochim. Cosmochim.Acta 45, 1421 (1981).

7. C. T. Driscoll and G. E. Likens, Tellus 34, 283

(1982); G. E. Likens et al., in (5).8. J. P. Baker and C. L. Schofield, Water Air Soil

Pollut. 18, 289 (1982); C. T. Driscoll, J. P.Baker, J. J. Bisogni, C. L. Schofield, Nature(London) 284, 161 (1982); J. P. Born, in AcidRain/Fisheries, R. E. Johnson, Ed. (AmericanFisheries Society, Bethesda, Md., 1982); C. L.Schofield and J. R. Trojnar, in Polluted Rain, T.Y. Toribara, M. W. Miller, T. E. Morrow, Eds.(Plenum, New York, 1980).

9. G. E. Likens et al., Ecol. Monogr. 40, 23 (1970);F. H. Bormann, G. E. Likens, D. W. Fisher, R.S. Pierce, Science 159, 882 (1968); G. E. Likens,F. H. Bormann, N. M. Johnson, ibid. 163, 1205(1 969).

10. G. E. Likens, Verh. Int. Verin. Limnol., inpress; F. H. Bormann and G. E. Likens, Pat-terns and Process in a Forested Ecosystem(Springer-Verlag, New York, 1979).

11. D. Drabl0s and A. Tollan, Eds., Proceedings ofthe International Conference on the EcologicalImpacts of Ac id Rain (Sandefjord, Norway,

1980); T. A. Haines, Trans. Am. Fish. Soc. 110,669 (1981); W. Dickson, Inst. Freshwater Res.Drottingholm Rep. 54 (1975), p. 8; C. L. Scho-field, in Acid Rain/Fisheries, R. E. Johnson, Ed.(American Fisheries Society, Bethesda, Md.,1982); P. J. Dillon et al., J. Fish. Res. BoardCan. 35, 809 (1978).

12. Acid Deposition and Atmospheric Processes inEastern North America (National AcademyPress, Washington, D.C., 1983); E. Marshall,Science 221, 241 (1983); G. E. Bangay and C.Riordan, Chairmen, Working Group I. ImpactAssessment. Final Report (U. S./Canada Coordi-nating Committee responsible for implementingthe Memorandum of Intent between Canada andthe United States of America, 1980, on Trans-boundary Air Pollution, U.S. Department ofState and Department of External Affairs, Cana-da, 1983); Atmosphere-Biosphere Interactions:Toward a Better Understanding of the Conse-quences of Fossil Fuel Combustion (NationalAcademy Press, Washington, D.C., 1981).

13. Contribution of the Hubbard Brook EcosystemStudy. Financial support provided by the Na-tional Science Foundation and the Andrew W.Mellon Foundation. Hubbard Brook is operatedby the Forest Service, U.S. Department ofAgriculture, Broomall, Pa. The help of theGreater Vancouver Water District is gratefullyacknowledged.

Krug and Frink present some interest-ing arguments but do not give a balancedview of the importance of various causesof acidification of freshwaters. More-over, it seems inappropriate to say thatthey present a new perspective. Impor-tant arguments on this topic were raisedby Rosenqvist (1) several years ago, andthey have since been under continuousdiscussion. The mechanisms involved infreshwater acidification are certainlycomplicated (2), but very likely acidicprecipitation plays an important role inmost cases (2, 3).Krug and Frink discuss a number of

studies carried out in Norway, but oftentheir description is incomplete. Theysay, for example, that runoff from mini-catchments that received acidic deposi-tion was twice as acidic as that in therain. Although this is correct for theperiod they mention (an event in Novem-ber 1977), it is an incomplete representa-tion. The final report from the Norwe-gian project on "Acid Precipitation-Effects on Forests and Fish" (the SNSFproject) (3) states, "It is seen that therunoff is, in general, more acid than theprecipitation [during the event]. Howev-er, there is apparently a wash-out ofsulphate in the same period. The sul-phate may originate from an accumula-tion of air-borne sulphate or oxidation ofsulphur compounds during summer."The SNSF project final report alsoshows that accumulation of airborne sul-phate does, in fact, occur during summerand early fall.Krug and Frink suggest that the "mo-

bile anion" concept (2, 4), that is, thatsulfuric acid falling on soils may movethrough the soils "causing increased cat-ion leaching proportional to the in-creased flux of So42-" has a "majorflaw" because organic anions tend todecrease as other anions increase. They

1425