Reprinted from

Forest Ecology andManagement

Forcs~ Ecology and Management 84(1994) 109- 121

Nitrogen saturation in a high elevation New England spruce-firstand ’

Steven G. McNulty a9 l , John D. Aber b, Steven D. Newman b’ COH eera Hydrologtc Loboraror). 3160 Com,eera L.ah Rd.. Oflo. NC 28734. L’SA

’ Complex Syuern, Re,eurch Cenrer. Morse Ha//. Unluersq oJNw Hampshrre. Durham. NH 03824. USA

Accepted 14 Februay 19%

ELSEVIER

Forest Ecologyand Management

Aims and scope. Foresr Ecology and Managemenf publtshes sclentrfic amclef concerned wtth forest management and conser-ia:lcn. and In parttcular the appltcatton of btologtcal, ecological and Soctal knowledge to the management of man-made and nat-ural forests. The scope of the journal Includes all forest ecosystems 01 the world. A relereetng process ensures the quality and,r:erna;ional Interest of the manuscripts accepted for publtcatton. The journal aims to encourage communrcatlon between scten-!is:s in clsparale ltelds who share a common interest in ecology and natural-resource management, and to bhdge the gapoetween research workers and forest managers in the field to the beneftt of both. The journal should be of Interest to researchworkers, managers and pokey makers In lorestry. natural resources, ecologrcal conservatton and related fields.

FOUNDING EDITORLaurence L. Roche. Murroe. Ireland

EDITORS-IN-CHIEFFor the Americas.. Australra. New Zealand and the Paalic: For the rest of the world:

R F FisherDeparmenf of Forest Sctence

Texas A&M UnlversttyCollege Statton, TX 77843-2135. USA

G.M.J. MohrenForest Productron Ecology GroupDepartment of Vegetation EcologyDLO-Institute for Forestry and Nature ResearchP.O. Box 236700 AA Wageningen, The Netherlands

BOOK REVIEW EDITORD. Bnkley, Forest and Wood Sctence. Colorado State Untverstty, Fort Colltns, CO 80523, USA

EDITORIAL ADVISORY BOARDG Abrahamsen. Agrcultural Unrversrty of Norway, As,

NorwayMR. Ahula. BFH, institute of Forest Genetics,

Grosshansoorf. GermanyR Alfaro. Canadian Forestry Servtce, Vtctorra. B.C.,

CanadaF Andersson. Swedtsh Unrversity of Agncultural Sctences,

Uppsala. SwedenP Attlwlll. Untversity of Melbourne, Parkvtlle, VIC., AustraltaJ C Calve, lnstltute of Technology, Cartago. Costa RtcaJ.0 Deans lnstltute of Terrestrial Ecology, Pentcutk.

Midlothlan. UKD P. Dykstra. CIFOR. Jakarta, IndonesiaE P. Farrell, Untverslty College Dublin. Dublin, IrelandP H FreerSmIth, Forestry Commtsslon, Farnham, UKR.A Goyer. Loursana State Untversty. Baton Rouge, LA,

USAJ B. Hall. Unlverslty College of Nonh Wales. Bangor. UKB M. Kumar, Kerala Agrtcultural Untversrty. Kerala, lndtaJ.N. Long, Utah State Untverstty, Logan, UT, USAA E. Lugo. InternatIonal Institute of Troptcal Forestry, Rio

Pledras. PR. USA

J.A. Maghembe, SADCCIICRAF Agroforestry Project,Zomba. Malawt

F. Makeschin, lnstttut fur Bodenkunde und Standonslehre,Tharandt, Germany

D.C. Malcolm, Untversity of Edinburgh, Edinburgh, UKE. Malkonen, Frnnrsh Forest Research Instrtute. Vanraa.

FinlandM.A.R. Nahuz. lnstituto de Pesqutsas Tecnologcas.

Sao Paulo, SP, BraztlR.F. Powers, Pacific Southwest Research Statton. Redding,

CA, USAL. Rasmussen, Danish Forest and Landscape Research

Institute. Hoersholm, DenmarkG. Ruark, USDA Forest Service. Washington, DC, USAR. Sands, Untversity of Canterbury, Chnstchurch. NZJ.A. Stanturf, Stonevrlle, .MS, USA0. Sztklai, Untverstty ol British Columbia, Vancouver, B.C.,

CanadaK. von Werssenberg, Untversity of Helsinki. Helslnkt. Ftnland

Publication information: Fores! Ecokgy and Management (ISSN 0378-l 127). For 1996 volumes 80-89 are scheduled for pub-IIcatron. Subscriptton prices are avariable upon request from the Publisher. Subscripttons are accepted on a prepaid basrs onlyand are entered on a calendar year basts. Issues are sent by surface mall except to the following countries where air delivery vtaSAL mall IS ensured: Argentina. Australia, Brazil, Canada, Hong Kong, India, Israel, Japan, Malaysia, Mexico, New Zealand.Pakls:an. PR China, Stngapore. South Afnca. South Korea, Taiwan, Thailand, USA. For all other countrtes airmail ratesare avaIlable on request. Claims for mtssng Issues should be made within SIX months of our publtcation (matling) date.Please address all your requests regarding orders and subscription queries to: Elsevier Science B.V., Journal Department,P.O. Box 211, 1000 AE Amsterdam, The Netherlands, tel. (+31-20) 4863642, fax (+31-20) 4863598.

In rhe USA and Canada: For further tnformatton on thts and other Elsevier journals please contact: Elsevier Science Inc.,Journal Information Center, 665 Avenue of the Americas, New York, NY 10010, USA. Tel. (212) 6333750; fax (212) 6333764;telex 420-643 AEP UI.

Back volumes: Please contact the Publtsher.

_ I

EUEVIER Forest Ecology and Management 84 ( 1996) It% 121

Pores~~~ology

Management

Nitrogen saturation in a high elevation New England spruce-firstand ’

Steven G. McNulty a l , John D. Aber b, Steven D. Newman b’ Cowctva Hydrologic Laboratory. 3160 Cowtcra Lub Rd., Orro, NC 28734. US4

’ Complex Systems Research Cmrer. Morse Hall. Uniuersiry ofNew Hampshire, Durham. NH 03824. US4

Accepted 14 February 19%

Abstract

High rates of nitrogen (N) deposition were first postulated as a cause of N saturation (i.e. the availability of NH,-N andNO,-N in excess of total combined plant and microbial nutritional demand) and spruce mortality during the 1980s. To testthis hypothesis, N addition plots were established in 1988, in a high elevation spruce-fir forest in southeastern Vermont, anarea of relatively low N deposition (5.4 kg N bulk deposition ha-’ year-‘). To test how the form of applied N mayinfluence forest growth and N-cycling, four replicated treatment plots received either NH&I-N, NaNO,-N. or a combinationof both N forms at rates ranging from 15.7 to 31.4 kg N ha- ’ year- ‘. The N additions were applied in three equal doseseach year between June and August from 1988 to 1994. In addition to N treatments, two control plots were also established.Between 1988 and 1990, annual in situ net N mineralization and net nitrification in the forest floor, litterfall and forest floormass and elemental concentration, foliar elemental concentration, and basal area growth by species were measured on eachplot. In July 1994, basal area growth by species, net N mineralization potential and net nitrification potential in the forestfloor. and foliar and forest floor elemental concentration were again measured on all plots. Inter-treatment and intra-tnat-merit basal area growth changed substantially between 1988 and 1994. Spruce, fir. and birch trees on the N addition plotsreceiving < 20 kg N ha- i year- ’ had the highest rate of growth between 1988 and 1990 and then had the highest rate ofdecline between 1991 and 1994. Spruce, fir, and birch trees on the N addition plots receiving > 25 kg N ha-’ year-’showed moderate rates of decline from 1988 to 1994. Numerous birch and maple sprouts were noted on the sites with thehighest rates of decline, but no spruce or fir seedlings were observed. In July 1994, net N mineralization potential washighest on the control plots and net nitrification potential of the forest floor was highest on the plots receiving 15.7 kg Nha-’ year-‘. A strong positive correlation existed between forest floor %N and net niuitication potential. Foliar %N waspositively con-elated with added N and negatively correlated with the change in net basal area growth. Foliar Ca:AIconcenvations may also be negatively related to changes in net basal area growth. Our results suggest that N saturation hascaused foliar nutrient imbalances on the N addition plots, and that the stands may be changing in species composition andstructure. No long-term effects of N-form additions on N saturation and forest health were observed. Continued N additions

’ Corrcspondmg author.’ The use of trade or firm names in this publication is for reader information and does not imply endorsement by the US Department of

A8nculture of any product or service

0378-l 1?7/%/515.00 Copynghl 0 1996 Elxvicr Scwxc B.V. All righti reserved.f’II SO?78- i 127(96)03742-S

I IO S.G. .Llc.Y~Iry er 01. /Forest Ecohgy and .Ualonagrmcnr 84 (1996) 109- 121

may change the stands from a slow growing and slow N-cycling coniferous forest, CO a fast N-cycling and fast growingdeciduous forest.

Keywords: N saturation. MI. Ascumey; Red spruce: Picea rrcbenr: Nument imbalance: Mortality; Fertilization

1. Introduction

During the last 2 decades, high rates of sprucedecline have occurred across the northern and south-em Appalachians (Pear? et al., 1992) and Europe(Nihlgard, 1985; Schulze et al., 1989). Possiblecauses of this decline are nitrogen (N) availability inexcess of plant demand (Nihlgard, 1985; Aber et al.,1989; Schulze et al., 1989), drought (Johnson andSiccama, 1983). Al toxicity in roots (Shortle andSmith, 1988L foliar nutrient imbalance (Zcettl andHuettl, 1986; Cronan and Grigal, 1995). reducedcold tolerance (Sheppard. 1994), and freezing injury(DeHayes, 1992). In areas where spruce decline isthe highest, many pollutants have high depositionrates (e.g. SO,. NO,. heavy metals). Increases inpollutant loading arc often accompanied by reduc-tions in forest floor pH (Shortle and Smith, 1988;Cronan and Schofield, 19901, increased forest floorheavy metal concentrations (Andresen et al.. 1980,Friedland et al., 1984a. Friedland et al., 1984b;Her-rick and Friedland, 19901, and leaching of soil Aland Ca (Cronan and Schofield, 1979; Berg, 1986;Johnson et al., 1994).

The interactions among pollutants and changes inforest processes make it difficult to identify thecause(s) for the decline. Although the 1990 UnitedStates (US) Clean Air Act has targeted a 50% reduc-tion in S deposition, only about a 10% reduction inN deposition is required (and expected). Therefore,N deposition is more likely to have potential impactson forest structure and function in the future. Toexamine the influence of N inputs alone on spruce-firforest N-cycling and forest health (i.e. tree mortalityand growth). we established a series of N additionplots on Mt. Ascutney, Vermont, in 1988 (McNultyand Aber, 1993). Our objective was to observechanges in N-cycling associated with varying ratesand forms of N inputs, to determine if and when Nsaturation (i.e. the availability of NH,-N and NO,-Nin excess of total combined plant and microbialnutritional demand) occurred within each treatment,

and to observe changes in forest health associatedwith N saturation. We used two types of N-fertilizerto assess the ecosystem’s ability to absorb anionicand cationic forms of N. As an anion, we hypothe-sized that NO,-N would not be absorbed by theforest floor, but would leach through the ecosystem.and deplete cations, such as Ca, from soil colloids.Therefore, NO,-N additions could have a minimalimpact on N availability, but a potentially largeimpact on cation availability, soil acidification, andAI mobility (A&r et al., 1989; Raloff. 1995). Con-versely, the cation NH,-N would more likely beretained by the ecosystem, thus conserving cationsand potentially increasing N availability. During thisperiod, annual in situ net N mineralization and netnitrification in the forest floor, litter-fall and forestfloor amount and elemental concentration, foliar ele-mental concentration, and basal area CBA) growth byspecies were measured on each plot. Results from1988 to 1991 suggest that N addition plots receiving>25 kgNha_’ year-’ may have been in the earlystages of N saturation, while plots receiving < 20 kgN ha-’ year” still had increased BA growth com-pared with control and high N input plots (McNultyand Aber, 1993). No substantial long-term (after 3years of treatment) changes were detected betweenthe form of N applied and forest health or N satura-tion. As N additions continued from 1991 to 1994,forest processes and stand composition changed. Thispaper discusses the effects of continued N additionson N-cycling and tree growth and mortality andpresents a possible scenario for future forest struc-tural and functional changes given continued addi-tions of N.

2. Materials and methods

2.1. study plots

During June 1988, ten 15 X 15 m plots wereestablished at an elevation of 762 m a.s.l., on Mount

S.G. McNulry CI a/./Forts! Ecology and Managrmcnr 84 (1596) 109-121 I I I

Ascumey. Vermont (42”26’N, 72”27’W), where redspruce (Picea rubens Sarg.) grows in large patches(> 1 ha) at elevations above 725 m. Red sprucecomprises > 80% of the total basal area; the remain-der is divided between balsam fir ( Abies bulsamecd,red maple (Acer rubrum), mountain maple (Acerspicancm), and birch (Betula spp. ) (McN1~1ty andAber. 1993). Study plots were randomly locatedwithin these patches. Before N additions, no statisti-cally significant (paired ?-test P > 0.05) differencesin stand structure, stand species composition, net Nmineralization and net nitrification of the forest flooror foliar chemistry existed among plots (McNulty,1991; McNulty and Aber, 1993).

These plots receive approximately 1000 mm ofprecipitation year-‘, 709 in frozen form as sleet orsnow (McNulty and Aber, 1993). The area is charac-terized by warm summers (average July air tempera-ture 22°C) and cold winters (average January airtemperature - 6°C) (USDC, 1%8X Located on broadareas between rock outcrops, the plots are onHoughtonville soils classified as frigid Typic Haplorrhods. The soils are well-drained and have 3 to15% slopes (USDA Soil Conservation Service, 1989).

2.2. Nitrogen additions

One pair of plots (controls) received no N-feti-lizer. Four pairs of plots were randomly selected tobe amended with NH&l-N and (or) NaNO,-N dur-ing June, July, and August from 1988 to 1994 (Table1). The rates of N application were comparable toannual N deposition rates recorded in spruce-firecosystems located in industrialized regions in theUS (e.g. 16 kg N ha-’ year- ‘, Friedland et al.,

1991) and western Europe (e.g. 30 to 40 kg N ha- ’year-‘, Grennfelt and Hultberg, 1986). Across NewEngland, N deposition increases from east to wesr(McNulty et al., 1990). The Mt. Ascutney plots arelocated slightly east of the area where signs of Nsaturation have been observed (McNulty et al., 19911,and the plots did not exhibit any of the initialsymptoms of N saturation presented by Aber et al.(1989). Bulk precipitation collectors measured 5.4 kgN ha-’ year-’ in open areas adjacent to the re-search plots (McNulty and Aber, 1993). By supple-menting the long-term inputs of anthropogenic Ndeposition with additional N-fertilizer at a rate equalto high rates of N deposition found elsewhere, wehypothesized that N saturation could be induced onthese plots.

2.3. Basal areu change

In June 1988. prior to the star of N additions,basal area CBA), species, and tree condition (i.e.living vs. dead) were recorded for all trees > 2 cmat 1.5 m above the forest floor. In November 1990,and July 1994, the trees within each plot were againmeasured and compared to initial measurements.Changes in BA were divided into two growth peri-ods: 1988 to 1990 (Period I) and 1991 to 1994(Period 2). Net BA growth was calculated by sub-tracting total BA death from total BA growth.

2.4. Forest floor net N mineruliwtion and net nims-cation

From July 1988 to June 1991, 40 forest floorsamples with an average area of 150 cm3 were

Table IDwgn of N addltron cxperimen~ on Mount Ascumey, Vermont Nitrogen was added in rluu qual doses during June. July, and August1988 to 1994

Plot number Monthly amendmeot (kg N ha- ’ per applicauon) Total annual amendment (kg N ha- ’ year- ‘)NH &I-N NaNO,-N

1.6 0.0 0.0 0.04.10 5.2 0.0 15.72.8 0.0 6.6 19.85.9 5.2 3.3 25.63.7 10.5 0.0 31.4

112 S.G. McNulry CI al./Forrsr Ecology and Managrmcnr 84 ~15961109-121

collected monthly (except over-winter samples whichwere incubated in situ from September to April),from each plot. Twenty samples from each plot wererandomly combined into five samples per plot andsieved through a No. 18 mesh screen to removetwigs, coarse roots, and stones. Combining samplesensured sufficient sample mass for analysis. Approx-imately 10 g of each of the five composite samplesfrom each plot were placed into 150 ml of 1 M KC1for 48 h. After 48 h. approximately 20 ml of solutionextract was removed and frozen at 0°C until ana-lyzed for NH,-N and NO,-N concentrations.

The remaining 20 samples from each plot wereplaced into l-ml (2.54 X lo-’ mm) thick poly-ethylene bags. replaced into the forest floor andincubated in situ for 28 days. The 20 incubatedsamples were then composite, sieved, and extractedas prevtously described. Initial and incubated ex-tracts were run on a TRUCS 800 Auto-analyzer forNH,-N (Technicon Industrial Systems, 1978) andNO,-N (Technicon Industrial Systems, 1977) con-centrations. Net annual N mineralization was calcu-lated as the sum of monthly incubated sample (NH,-N + NO,-N) minus initial monthly sample (NH,-N+ NO,-N). Net annual nin-ification was calculated asthe sum of monthly incubated sample NO,-N minusinitial monthly sample minus NO,-N.

In July 1994, 40 forest floor samples were col-lected and sieved as previously described. Half of thesamples were composite, placed in KC1 for 48 h,extracted, and run on the auto-analyzer to measureinitial NO,-N and NH,-N concentration, while theother half were lab. incubated at 22°C for 28 days.After 28 days the samples were composited as previ-ously described, placed in KCI for 48 h, extracted,and run on a auto-analyzer to measure incubatedNO,-N and NH,-N concentration. Net N mineraliza-tion and net nitrification potentials were calculatedas previously described.

2.5. Forest jloor chemistry

In June, 1988 (before the first N additions), and inJuly 1994, 20 forest floor samples, each with anaverage area of 150 cm3 were collected from eachplot and randomly composited into five samples perplot. The samples were sieved through a 5 X 5-mmmesh screen to remove all twigs, coarse roots, and

stones. Carbon (0, hydrogen (H), and N were mea-sured in these samples using a Perkin-Elmer model240B CHN analyzer. Two (of the five) forest floorsamples were randomly selected for pH measure-ments, using a 1:2 (w/v) forest floor to CaCl, (0.01M) solution.

2.6. Foliar chemistry

Three randomly selected dominant red spruce treeswere tagged on each plot at the beginning of theexperiment. In July 1988, 1990, and 1994, threerandomly oriented, sunlit, or understory branch sam-ples, each containing needles from all age classeswere removed from each tagged tree using a pruningpole. The clipped branches from each tree wereplaced in a large paper bag and dried for 48 h at70°C. After drying, all branch needles had fallen intothe bag. The contents of the bag were shaken tohomogenize the age classes, and a randomly selectedsub-sample was sieved through a 2 X 2-mm meshscreen to remove twigs and detritus, and groundthrough a No. 10 mesh using a Wiley Mill. Thesamples were placed in 60-m] acid-washed glass jars,dried for 24 h at 70°C. and stored in darkness at22°C.

Percent foliar N and lignin were measured usingnear-infrared reflectance specuoscopy (Wessman etal., 1988). After determination of N and lignin con-centration, cation concentration of the foliar sampleswere measured using 0.5 g of leaf tissue ashed at450°C for 4 h and added to IO ml of plant buffer.One liter of buffer consisted of 300 ml HCI, 100 mlHNO,, 20 ml of 1000 ppm MO Standard, and both tovolume (Jones, 1988). The extracts were analyzedusing inductively coupled argon plasma emissionspectroscopy (Jarrell-Ash %5 Atomcomp).

3. Results and discussion

3.1. Basal area change

A substantial shift in net BA growth occurred onthe fertilized plots between Period 1 (1988 to 1990)and Period 2 (1991 to 1994) (Fig. 1). During Period1, total plot net BA growth was highest for plotsreceiving < 20 kg N ha-’ year- ‘, suggesting a

S.G. McNuhy et ol./Forrsr Ecology and Monagcmcnr & (1996) 109-121 113

Fig. I. Average basal ma growth by species oo ICP spruce-firplols (four piurcd N addition u’c-auDcn& and one paired control)OII Mt. Ascumey. Vermont, during Period I (1988 IO 1990) andPeriod 2 (1991 to 1994).

fertilization response at these rates of N addition.Plots receiving > 25 kg N ha-’ year-’ showedlittle or no growth increase, suggesting that N was nolonger limiting growth. The 25.6 kg NaNO,-N ha-’year- ’ plots were the only treatments to have anegative total growth rate (i.e. net mortality), butrates of mortality were not significantly different(paired ;-test P > 0.05, n = 2) from the high 31.4kg NH,Cl-N ha-’ year-’ plots or the control plots.The plots receiving a combination of N-fertilizer at arate of 19.8 kg N ha-’ year- ’ had the highestaverage annual spruce BA growth.

During Period 2, net BA growth was the higheston the control plots ( +0.55 m2 ha- ’ year- ‘1 andlowest in the 19.8 kg N ha-’ year-’ addition plots(- 1.19 m2 ha- ’ year- ‘1. Most of the negative netBA growth on the 19.8 kg N ha- ’ year-’ plots canbe attributed to a large reduction in living balsam fir.The 15.7 kg N ha-’ year-’ plots were the only Nadditions plots to show a positive net BA growth rate( + 0.3 1 m2 ha- ’ year- ‘1, but net BA growth de-creased on all the plots receiving N-fertilizer, com-pared with Period 1 (Fig. 1). Red spruce net BAgrowth was positive for all plots except those receiv-ing 31.4 kg NH,Cl-N ha- ’ year-‘, and red sprucenet BA growth was inversely correlated to the amountof N added (Fig. I>.

3.2. Mortality

Stand mortality is defined as the fraction of treesin each stand, separated by species, that die each

period. Because mortality expresses change as apercentage, variation in initial stand stocking is re-moved from the assessment of tree mortality at theplot level. During both periods (1 and 2), the trees inthe two control plots had a mortality rate of approxi-mately 3% year-’ (standard error (a) = 4% (Period1). and 3% (Period 2)) (Fig. 2). In Period 1, allN-fertilized plots had very low rates of average standmortality ( < 1% year-’ ) and inter-plot variation inmortality (u = 1%). In the plots receiving > 25 kgN ha-’ year- ’ during Period 1. the number ofingrowth stems was greater than or equal to thenumber of dying trees, thus net stand mortality wass 0%.

During Period 2 (1991 to 19941, stand mortalityin the N addition plots increased to 5 to 9% year”(cr = 2 to 3%) because no species had sufficientingrowth of stems to replace dying individuals, mor-tality rates were positive for all plots and species(Fig. 2). N addition plots receiving < 20 kg N ha-’year-’ had the highest rates of mortality (> 8%year-’ 1, but because there were only two plots pertreatment, the differences in mortality rates betweentreatments were not statistically significant using apaired r-test (P > 0.05, n = 2).

Although not quantified, we observed increasedstump sprouting and seedling establishment of maple( Acer spp. ) and birch (Betula spp. ) species on theplots with the highest mortality. The reason for this

Fig. 2. Average annual monality by species on ten spruce-firplots (four paired N addition treatmen& and one paired control)on Mt. Ascumey. Vermonf during Period I (1988 IO 1990) andPeriod 2 (1991 10 1994).

114 S.G. McNulry cr al./ Fores! Ecology and Managtmer 84 (1996) 109-121

increase in deciduous species regeneration is notclear (i.e. increased light availability onto the forestfloor. increased forest floor temperature, or increasedN availability), but a shift in species compositionmay be occurring on the N addition plots.

3.3. Net N minerakafion and net nimj’kation poren-tial

The possible reason trees in the N-fertilized plotswere growing more slowly or had a higher mortalityrate than those in the control plots was then exam-ined. Net N mineralization and net nitrification pat-terns have changed since N additions began in 1988.The plot receiving > 25 kg N ha-’ year-’ had thehighest rates of annual in situ net N mineralizationduring Period 1 (1988-19!30), but annual in situ netN mineralization declined over time with the > 25kg N ha-’ year-’ plots having the lowest rates ofnet N mineralization at the end of Period 1 (McNultyand Aber. 1993). The same pattern of declining netN mineralization was observed in July 1994. whennet N mineralization potential was measured in thelaboratory (Table 2). In July 1994, the highest ratesof net Y mineralization potential were on the controland 15.7 kg NH&l-N ha-’ year- ’ plots (Table 2).Using the July 1994 forest floor samples, no patternwas observed regarding the form of N applied andnet N mineralization potential or net nitrificationpotential.

Over the years, as N additions continued on plots,the total amount of net N mineralization and net

nitrification potential decreased, but the percentageof mineralized N that was nibified increased. Beforethe N additions, no net nitrification was detected onany of the plots (McNulty and Aber, 1993) suggest-ing that forest floor N concentrations may have beentoo low to support nitrifiers. These results are sup-ported by McNulty et al. (19901, who found no netnirrification potential in undisturbed spruce-fir siteslocated in areas of low N deposition. Following thefirst year of N application, the fraction of net miner-alized N which was nitrified rose to approximately5% on plots receiving > 25 kg N ha-’ year-’ andno net nitrification occurred on the plots receiving< 20 kg N ha-’ year- ’ (McNulty and Aber, 1993).In the July 1994 sampling, the fraction of mineral-ized N that was niuified was approximately 12% oftotal net N mineralization potential on the plotsreceiving < 20 kg N ha- ’ year-‘, and a strongcorrelation was observed between forest floor %Nand net nitrification potential (Fig. 3). When forestfloor N was below 1.35% no net nitrification poten-tial was observed, but when forest floor N was> 1.35% net nitrification was linearly correlatedwith forest floor %N (r2 = 0.92, P < 0.05, n = 4)(Fig. 3). Both th e intercept and slope of this relation-ship are very similar to earlier work (McNulty et al.,19911, that related forest floor %N and net nitrifica-tion potential on I61 plots located in 11 areas acrossNew England (Fig. 3). The Mt. Ascumey data sup-pon the hypothesis that nitrifiers living in these veryacidic, organic soil horizons need forest floor Nconcentrations > 1.4% to nitrify. If forest decline

Table 2Change m forcsr floor 6N. C:N ratio, net N mineralization potential (N miner.), net nitification potential (nitrif.), and the percent ofmmcralizcd N hat was muificd (W nitnf.) on Mt. Ascumey, Vermont, spruce-fir plots between 1988 (prc-N xldltions) and 1994. Thestandard error for the mean of each value is in parentheses. Five forest floor samples were collected from exh of two plots per ucatment fora total of ten samples per treatment

Treatment(kg N ha- ’year- ’ 1

015.719.825.631.4

Forest floor N (96) Fores1 floor C:N (kg: kg) July. 195% colltxvd forest floor(mg N kg- ’ soil per 28 days)

I988 1994 1988 1994 N miner. Nitrif. % Nitrif.

1.22 (0.16) I .36 (0.05) 29.4 (0. I) 32.1 (0.8) 50.0 (12.8) 0.0 (0.0) 0I .47 (0.0 I ) 1.47 (0.05) 29.2 (0.4) 32.01 (1.0) 56.6 (21.5) 6.8 (3.3) 12I .44 (0.06) I .39 (0.07) 29.4 (0.8) 3 1.7 (0.9) 38.6t lS. l ) l.S(l.1) 51.35 (0.14) 1.28(0.11) 29.5 (0.4) 28.3 (0.4) 9.6 (5.7) 0.0 (0.4) 0I .30 (0.04) I .40 (0.06) 30.0 (0.9) 29.7 (0.4) 38.7 (27.9) 0.7 (0.6) 2

S.G. McNulry cr al./Forcsr Ecology and Management 84 f19961109-121

PLOTS

* N Deporltlon Grrdlont 9 Mt. A8cutnoy ’, ,“X

,,,,,,I,zi/

I’,

J

IIS

1 1.2 1.4 1.6 1.6 2 2.2

KN FOREST FLOOR

Fig. 3. Relationship befwcea average forest floor RN and 2&day eel nitrificrtion potential on ten sprua-fir plols (four paired N additiontxatments. and one paired conool) on ML Ascumcy. Vermont. sampkd in July 1994, and avenge forest floor %N and 28day netnltnficatioa potential for 1 I SIICS across a New England N deposition gradient sampled from June to August in 1987 and 1988.

(i.e. reduced N demand) occurs in conjunction withincreased net nitrification, significant losses of ni-trate are possible (Durka et al., 1994).

No relationships were found between the form ofN addition and the rate of net N mineralization or netnitriiication between 1988 and 1991, or between netN mineralization or net nitrification potential in July1994.

3.4. Forest floor chemistry

From 1988 to 1994, up to 220 kg NH&l-N ha-’(3 1.4 kg NH &l-N ha-’ year- ’ plots), or 180 kgNaNO,-N ha-’ (25.6 kg NaNO,-N ha” year-‘)representing up to a 20% addition to the existing1150 kg N ha-’ average forest floor N content(McNulty and Aber, 1993) have been added to theplots. Despite these additions, forest floor %N hasnot significantly changed on any of the plots (Table2). suggesting that N is immobili&d in decomposingwood, taken up by the vegetation, is leaching fromthe ecosystem or that variation in forest floor chem-istry is too large relative to the change in N to detectany noticeable differences. Loss of N as N,O-N was

also not significant because < 0.1 kg N,O-N ha-’per growing season is released from either the highN addition or control plots (Castro et al., 1993). Thechange in woody debris N concentration and masshave not been measured over time, but the signifi-cant increase in foliar N concentrations (Fig. 4)partially account for the lack of change in forest

Fig. 4. Avenge nd spruce foliar %N colle.cted in July 1988, 1990.and 1994 on IIXI spruce-fu plots (four paired N addition ucat-men& and one paired control) on Mt. Ascumey. Vermont.

116 S.G. McNulry CI al./Foresr Ecology and Mqnagcmcnr 84 (1996) 109-121

floor %N. During the first 2 years of treatment (1988and 19901, control and N addition plot litterfall massand quality were measured (McNulty and Aber,1993). Red spruce litterfall from the control plotsaveraged 1500 kg ha-’ year- ’ , and using a foliarretention time of 8 years, total spruce foliar biomasson control plots was equal to 12.0 t foliage ha-‘.Using 0.90% N for spruce foliage, the control plotscontained approximately 110 kg organic N ha-’ infoliage. Assuming that the high N addition plots alsocontained 12 t foliage ha- ’ , these plots would retain180 kg organic N ha-’ (given a 1.5% N of greenleaf tissue). The difference between control and highN addition plots foliar N storage is only 80 kg N.Because forest growth is now less on the treatmentplots than on controls, and the treatment plots werelosing more litter each year than control plots (MC-Nulty and Aber. 1993), the differences in N storagebetween control and N addition plots is likely < 80kg organic N.

retained in the forest floor. Across all plots, forestfloor pH was very low (2.7 to 3.0) at the beginningof N-fertilization in 1988 and had not changed sig-nificantly (paired r-test P > 0.05. n = 4) when mea-sured during the 1994 sampling (i.e. pH ranged from2.6 to 3.0).

3.5. Spruce foliar chemisrry

The greatest loss of N (especially N applied asNO,-N) may occur through leaching of N from theforest floor. Potential loss of N through leaching wassupported by McNulty and Aber (19931, who foundhigher levels of N in resin bags located at the base ofthe forest floor in N-fertilized plots (> 800 ppmNO,-N log-’ resin and > 600 ppm NH,-N log-’resin) than in control plots ( < 20 ppm NO,-N log-’resin and < 15 ppm NH,-N log-’ resin). Between1989 and 1991, ion exchange resin bags placed at thebase of the forest floor detected substantial differ-ences in HN,-N and NO,-N concenuations relatingto the form of N applied to each plot (McNulty andAber, 1993). No clear trend was observed betweenthe amount of N added and changes in the forestfloor C:N ratio, which also suggests that N is not

From the July 1994 sampling, the control plotshad the lowest average foliar %N, while the 25.6 and31.4 kg N ha-’ year- ’ plots had the highest averagefoliar %N (Table 3). Foliar N concentration in-creased significantly between the July 1990 samplingand the July 1994 sampling (paired r-test, T - 8.18,P < 0.0001, n = 12) on the N-fertilized plots, but thefoliar N concentrations on the control plots were notstatistically different between measurement periods(paired r-test P > 0.05, n = 12) (Fig. 4). Controlplot foliar N was significantly lower (paired r-test,T - 8.6, P < 0.0001, n = 12) than the foliar concen-trations of plots receiving > 20 kg N ha-’ year-‘,and the foliar N concentration of spruce from the25.6 kg N ha-’ year-’ was significantly higher(paired t-test, T - 16.9, P < 0.0001, n - 12) than allother treatment or conuol plots. These foliar N con-centrations are higher than any values reponed forred spruce across an N deposition gradient in NewEngland (Femandet and Struchtemeyer, 1984; Fried-land et al., 1988; Huntington et al., 1990; McNulty etal., 1991) and may represent the upper limit of thatred spruce foliar N concenuations under N additionfield conditions.

The July 1994 foliar N concenuation was in-versely correlated to BA growth up to the 25.6 kg N

Table 3Variaoon in red spruce foliar chemistry on Mt. Ascumey. Vermont spruce-fir plots collec~cd in July 1994. The standard error for the meanof each value is m parentheses. Six foliage samples were collecti from each of two plots per treatment for a total of 12 samples pertreatment

Trcatmenr (kg N ha- ’ year- ‘) N (%) Ligmn (%I Ca (g kg- ‘) Al (mg kg- ‘1 N:Ca (molar ratio) Ca:AI (molar ratio)

0 0.85 (0.03) 24. I (0.6) 3.5 (0.3) 34 (4) 7.1 (0.11) 69 (8)15.7 I .30 (0.05) 24.6 (0.9) 3.5 (0.5) 32 (4) I I .o (0.09) 73 (7)19.8 1.23 (0.06) 23.8 (0.7) 3.1 (0.4) 32 (3) 11.3 (0.12) 65 (7)25.6 1.51 (0.07) 23.4cl.l) 3.4 (0.4) 36 (4) 13.1 (0.18) 63 (6)31.6 I .47 (0.07) 24.2 (0.7) 3.1 (0.3) 38 (3) 14.0(0.l8) 55 (5)

I

S.G. McNvlry cr al./Forcsr Ecology andhioMgcmcru & (1996) 109-121 117

NET DECLINE , NflGROWTH

I -T-

Fig. 5. Relationshtp bctwttn foliar N% collected in July 1994 andred spruce annual BA growth from 1991 to 1994 on ten sprua-fitplots (four paired N addition v~~uIK~&, and ooe paired coalrol)on ML Ascumcy. Vermoot.

ha-’ year- ’ addition level (Fig. 5). Although the31.4 kg N ha-’ year-’ addition plots, had thehighest rate of decline, the foliar N concentrationwere not higher than in the 25.6 kg N ha-’ year-’addition plots. These results indicate that even rela-tively low additions of chronically applied N canpotentially cause reduction in forest growth and ulti-mately stand decline. No differences in BA growthor mortality rates relating to the form or N appliedwere observed.

In July 1994, red spruce foliar Ca concentrationsin control and fertilized plots did not differ signifi-cantly (paired t-test P > 0.05, n = 12) (Table 3).Our measured foliar Ca concentrations are in themid-to-upper range of values reported by others in-cluding MacLean and Robertson (1981) (1.8 to 3.1 gCa kg - ’ ); Femandez and Stnxhtemeyer ( 1984) (1.9to 11.6 g Ca kg- ’ , mean 4.5 g Ca kg-‘); Friedlandet al. (1988) (1.8 to 4.1 g Ca kg-‘); Joslin et al.(1992) (2.7 g Ca kg-’ (low elevation), 2.4 g Cakg- ’ (high elevation)); and Thornton et al. (1994)(0.9 g Ca kg” to 1.1 g Ca kg-‘). Foliar Caconcentrations are well above the minimum level of0.5 g Ca kg-’ to 0.8 g Ca kg-’ which serve as anindicator of Ca deficiency (Swan, 1971). However,these minimum deficiency levels were developedunder greenhouse conditions (Swan, 19711, and lev-els of nutrient deficiency may be different for maturetrees (Joslin et al., 1992). Thornton et al. (1994)

suggested that a minimum deficiency level of 0.17%may be a more accurate level for mature red spruce.However, the Mt. Ascutney red spruce foliar sampleswere also well above this Ca concentration (Table 3).

Foliar Al concentrations from the July 1994 col-lection were restricted to a narrow range across theplots (Table 3). and are within the spruce foliar Alconcentration averages reported by Joslin et al. (1992)for several low elevation (31 mg kg” Al) and highelevation (39 mg kg-’ Al) spruce stands. Any analy-sis of variance correlation between N added andfoliar Al concentration was not significant (pairedT-test P > 0.05, n - 12) in this study.

The proportion of nutrients (e.g. Al, N, Ca) maybe a bentr measure of forest health than the absoluteconcentration of each individual nutrient becausedisproportionate levels of nutrients could lead lo animbalance even if neither nutrient level in itself isdeficient (van Dijk and Roelofs, 1988; Cronan andGrigal, 1995). Foliar CzAI molar ratios from theJuly 1994 collection had a wider range of mean plotvalues (i.e. 55 to 69) because foliar Ca concentrationtended to decrease and foliar Al concentration tendedto increase with increased N additions 10 plots (Table3). Because there were only hvo plots per treatment,the intra-treatment standard deviation is large enoughthat the difference in foliar CzAl concentrations arenot statistically significant (paired T-test P > 0.05,n - 2). However, the average treatment foliar CxAl

NlTf DECLINE , NETQROWTH

r : 0.81P l 0.03

ROT

Fig. 6. Relationship between foliar C&AI molar ratios collcctcd inJuly 1994 and red spruce annual basal area growth from 1991 to1994 on ten spruce-fir plots (four paired N addition treatments.and one pairtd control) on ML Ascumcy. Vermont.

118 S.G. McNuliy cl al./Fortst Ecology and Managcmrru 84 (1996) 109-121

was inversely related to N-fertilization (r2 = 0.93,P = 0.04. n = 5) and net spruce BA growth (Fig. 6).A literature review by Cronan and Grigal (1995)found that foliar Ca:Al molar ratios between 11.4and 14 can adversely affect forest growth. Cronanand Grigal (1995) note that the reported levels ofimpact (i.e. threshold levels) were often observed atthe lowest level of foliar Al concentration and theactual threshold level may be higher. Also, of the 36studies of plant CzAl ratios affecting growth, onlythree studies used forest soils (but not necessarily insitu soils), indicating the need for increased researchunder field conditions. The lowest foliar spruce CxAlmolar ratio on Mt. Ascumey was 55 and occurred onthe highest N addition plot (Table 3). This ratio isapproximauzly four times higher than values repottedby Cronan and Grigal (1995) as causing adverselyaffecting forest growth.

On Mt. Ascumey plots, excessive N may beadversely affecting forest health before large in-creases in foliar Al, or decreases in foliar Ca orCazAl ratio occur. McLaughlin et al. (1991) con-cluded that low foliar CzAl ratios may be related todecreased leaf carbon balances among conifers grow-ing on sites altered by pollutant additions, includingN. Although mechanisms by which N inputs mightalter plant Ca nutrition have been suggested (MC-Laughlin and Kohut. 19921, how N availability, Ca,and Al interrelate and affect plant carbohydrate pro-duction and use remains unknown. Measurements ofphotosynthetic capacity have been used to evaluatethe carbon economies of red spruce to relate differ-ences in leaf N and Ca concentrations to tree vigor.Field assessments of red spruce in the southernextent of its range in the US have shown that trees athigh elevations and declining trees at any elevationhave higher rates of dark respiration, lower folk Caconcentrations, and higher folk Al concentrationsthan trees at lower elevations and healthy trees atany elevation (McLaughlin et al., 1990, McLaughlinet al., 1991). Nutrient deficiencies, imbalances, orboth may lead to carbon depletions which predisposetrees to the detrimental impacts of other stresses suchas &ought or freezing injury (DeHayes, 1992;McLaughlin and Kohut, 1992). Winter physiologicalstudies on Mt. Ascumey plots during January of1994 and 1995 indicate an inverse correlation be-tween N addition and cold tolerance of red spruce

foliage (T.D. Perkins and S.G. McNulty, unpub-lished data, 1995).

3.6. Potential future changes in plot structure andfunction

In areas of low N deposition, N availability isprimarily dependent on net N mineralization. How-,ever, as N deposition increases, a larger portion ofthe N needed for plant growth and development canbe supplied directly from atmospheric sources.’ AsN-fertilization continues, the linkage between net Nmineralization, N uptake, foliar N concentration, andlitter quality degenerates. From 1988 to 1990. fertil-ized plots had increased net N mineralization, foliar%N, forest growth, and litter quality. However, asfertilization continued, net N mineralization potentialand forest growth decreased while foliar %N re-mained high. These results suggest that N is nolonger limiting, and that excess N may be leachingfrom the ecosystem (McNulty and Aber, 1993) caus-ing foliar nutrient imbalances and forest decline. Ifanion leaching continues (McNulty and Aber, 1993).Ca and Mg may be leached and nutrient imbalancesincrease (Table 3). Much of the available Ca isobtained through the weathering of parent material(Woodwell and Whittaker, 1%7; Likens et al., 1970).If nutrient leaching could be reduced, cation weather-ing could increase Ca availability and reduce nutrientimbalances despite continued N-fertilization. On theplots receiving > 25 kg N ha-’ year-’ as fertilizer,a total of approximately 50 kg N ha-’ year- ’ (30 kgN ha-’ year-’ from fertilization and 5 kg N ha- ’year-’ from bulk N deposition, and 15 kg N ha-’year-’ from net N mineralization) is available forplant, mycorrhizae and microbial uptake. This levelof N availability is at the low end of measuredvalues for deciduous broad-leaved forests (Pastor etal., 1984). which typically have higher foliar Ncontents, better litter quality (lower lignin:NJ, andhigher N demand than needle-leaved evergreens.Greater N demand by deciduous broad-leaved species(e.g. Bench spp. and Acer spp.) could decreaseNO,-N leaching by nducing available NO,-Nthrough increased leaf-based nitrate reductase activ-ity (Smimoff et al., 1984; Smimoff and Stewart,1985). Downs et al. (1993) found higher rates offoliar nitrate reductase activity in Acer rubrum (red

S.G. McNulty et al./Forcst Ecology and Managrmcnt 84 (19%) 109-121 119

maple) seedlings exposed to levels of NO,-N equalto NO,-N deposition rates experienced across NewEngland than in seedlings not exposed to elevatedlevels of NO,-N.

In this study, growth declined in both coniferousand deciduous trees when fertilization continued af-ter 1990 (Period 2). However, in the plots with thegreatest percentage of tree mortality, numerous de-ciduous stump sprouts, and some deciduous seedlingsemerged, while spruce or fir did not regenerate.Heavy birch and maple regeneration was also ob-served in areas of high N deposition and spruce-firdecline across New England in 1987 and 1988 (S.G.McNulty and J.D. Aber, unpublished data, 1988).We propose that additional N will continue to reducespruce-fir dominance in the fettilized plots, and thatspruce-fir will be replaced by the cold tolerant, Ndemanding species already present on the plots inlow numbers (i.e. Befula spp. and Acer spp.) Withimproved litter quality (lower C:N ratio), net Nmineralization and nitrifkation may also increase.Eventually, the N-cycle and forest growth could bere-established at a new, higher equilibrium (Fig. 7).

4. Conclusion

The results of 7 years (1988 to 1994) of N-fenili-zation to spruce-fir stands on Mt. Ascumey, Ver-mont suggest that the early stages of N saturation areoccurring on all fertilized plots. Despite the low level

nlm 01T)o*wlcIpIc(

__._-_---.

TIME -

Fig. 7. Conceptual curve of changes in N availability (net Nmtncraltzation plus N deposition) and forest growth with contin-ued N addtttons to high elevatton spruce-fir forests.

of added N; spruce foliage %N, BA growth, net Nmineralization and net nitrification were significantlydifferent between treatment and control plots after 7years of N additions. These differences appear to berelated to the amount of N applied and not the formof N. We predict that these changes are only thebeginning of a series of ecosystem alterations thatcould eventually lead to a shift in species composi-tion (assuming no serious cation deficiencies developin replacement species), forest growth, and N-cy-cling. If N additions continue, both on Mt. Ascumeythrough N-fertilization, and across New England fromanthropogenic sources, high elevation spruce-firforests could eventually be converted into stands ofbirch, maple, or both.

References

Abcr. J.D., Nadelhoffer. K.J., Stcudler. P. and Melillo. J.M.. 1989.Nitrogen saturation in nonhem forest ecosystems. BioScieacc,39: 378-386.

Andmscn. A.M.. Johnson, A.H. and Siccama, T.G.. 1980. Levelsof lead, copper and zinc in the forest floor in the northeasternUnited States. J. Environ. Qual., 9f2): 293-296.

Berg, 8.. 1986. The influence of experimental acidification onnutrient release and decomposition rates of needle and rootlitter in the forest floor. For. Ecol. Manage., IS: 195-213.

Castro, MS., Stcudler, P.A., Melillo. J.M.. Abcr. I.D. and Mill-ham, S., 1993. Exchange of N,O and CH, between theatmosphere and soils in spruce-fir forests in the notshcastcmUnited States. Biogeochemistry, 18: l19- 135.

Cronan. C.S. and Grigal. D.F.. 1995. Use of caJcium/aJumiaumratios as indicators of stmss in forest ecosystems, 1. Envtron.QuaJ., 24: 209-226.

Cronan. C.S. and Schofield, C.L.. 1979. Aluminum leachingresponse to acid precipitation: effects on high-elevation water-sheds in the nonhcast. Science, 204: 305-306.

Cronan. C.S. and Schofield. C.L.. 19%. Relationships betweenqueous aluminum and acidic deposition in forested water-sheds of North America and Northern Europe. Environ. Sci.Tcchnol.. 24: 1100-l 105.

DcHayes, D.H.. 1992. Winter injury and developmental coldtoletancc of red spruce. In: C. Eagar and M.B. Adams (WI-ton). ‘The Ecology and Decline of Red Spruce in the EasternUnited States. Springer-Verlag. New York, NY, USA.

Downs, M.R.. Nadelhoffer. K.J.. Melillo. J.M. and Abcr. J.D.,1993. Folk and fine root nitrate rcductasc activity in seedlingsof four tree species in mlation to nitrogen availability. Trees.7: 233-236.

Durka. W., Schulze, E.-D., Gebauer, G. and Vocrkelius, S.. 1994.Effects of forest decline on uptake and leaching of depostted

120 S.G. Mc.~uliy et al./Forcst Ecology and Managemenr 84 119961109-121

niuale devrmmcd from “N and “0 measurements. Nature.

372. 765-767.

Femandcz. IJ. and Su-uchumeyer. R.A.. 1984. Correlations be-tween element conccnlralioos in spruce foliage and forest

soils. Soil Sci. Plan1 Anal., I SC IO). l243- 1255.

Friedland. A.J.. Johnson. A.H.. Siccama. T.G. and Madcr. D.L..

1984.a. Trace meti profiles in the forest floor of New Eng-

land. Soil Sci. Sot. Am. J.. 48: 422-425.

Friedland, A.J.. Gregory. R.A.. Karcnlampi. L. and Johnson.A.H., 1984.b. Winter damage 10 foliage as a factor in red

spruce decline. Can. J. For. Rcs.. 14: %3-%5.

Friedland. A.J.. Hawley. G.J. and Gregory, R.A.. 1988. Red

spruce (Plcea rubens Sarg.) fohar chemistry in nonhcm Ver-

mono and New York. USA. Plant Soil, 105: 189-193.Friedland. A.J.. Miller. E.K.. Batdes. JJ. JJ. and Theme. J.F., 1991.

Nitrogen deposition. dlsmbuuon and cycling in a subalpinespruce-fir forest in the Adirondacks, New York, USA. Bio-

gecchermsuy. 14, 31-55.Grcnnfell. P. and Hulrbcrg. H.. 1986. Effects of niaogen deposi-

tion on the acidrficadon of rerrcstnal and aquatic ecosystems.

Waur Air Soil Poll.. 30 945-%3.Hcrrick. G.T. and Fricdland. A.J.. 1990. Patums of u-ace metal

conccnuatlon and acIdIt) m monlanc fores1 soils of Ihe north-eastern Unlled Slatis. Water Air So11 Poll.. 53. 151-157.

Huntington. T.G.. Pean, D.R.. Hornig. 1.. Ryan. D.F. and Russ*

Savage, S.. 1990. Relationships bctwecn soil chcmisny, foliarchemlsuy, and condition of red spruce at Mount Moosilauke,

New Harnpshlre. Can. J. For. Res., 20: 1219-1227.Johnson. A.H. and Siccama. 1.G.. 1983. Acid deposition and

fores! declme. Environ. SCI. Techn.. l7(7): 294A-305A.Johnson, A.H.. Friedland. AJ.. Miller. E.K. and Siccama. T.G..

1994. Acid mm and soils soils inin tictic Adirondacks. III. Ravs of soilacidification in a momane spruce-fir forest aI Whiteface

Mounlilin. New York. Can. J. For. Res., 24. 663-669.Jones. J.B., Jr. (Ediror). 1988. Soil Testing and Plant Analysis:

Procedures and USC. Univemty of Georgia Technical Bullc~inNo. 109. Athens. Georgia. USA.

Joslin. J.D.. Kelly. J.M. and Van Miegrocl. H.. 1992. Soil chcm-isuy and numtion of North Arncncan Spruce-fir stands: Evi-dence for recent change. J. Environ. Qual.. 21(i): 12-30.

Likens. G.E.. Bormann, F.H.. Johnson, NM., Fisher. D.W. and

Plcrcc. R.S., 1970. Effects of forest cutting and herbicidetreatment on numcnr budgets in Ihe Hubbard Brook waur-shed-ecosyslem. Ecol. Mono., 40. 23-47.

Ma&can. K.S. and Rotxruon. R.G.. 1981. Vanadon in Ihe major

elcmcm conunr of red spruce foliage with season, crownposition, tree and nssuenssue age. Commun. Soil Sci. Plan1 Anal.,12(l). 39-49.

McLaughlin, S.B. and Kohut, R.J., 1992. The effects of atme

sphcnc deposluon and ozone on carbon allocation and associ-acd physiological processes in red spruce. In: C. Eagar and

M.B. Adams (l311ors).(l311ors). TheThe Ecology and Decline of RedSpruce in the Eastern Unctd Stares. Springer.Vcrlag, NewYork, NY, USA.

McLaughlin. S.B.. Anderson. C.P.. Edwards. N.T., Roy. W.K. and

Layron. P.A., 1990. Seasonal patterns of photosynthesis andrespiration of red spruce from IWO IWO elevauons in declining

soulhem Appalachian stands. Can. J. For. Res.. 20: 485-495.

McLaughlin, S.B., Anderson. C.P.. Hanson, PJ.. Tjoclkcr. M.G.

and Roy. W.K., 1991. Increased dark rcspintion and calcium

deficiency of red spruce in relation IO acidic deposition athigh-elevation southern Appalachian Mountain sites. Can. J.

For. Res.. 21: 12361244.

McNulty. S.G.. 1991. Nutricru cycling in New England Spruce-Fir

Forests. Ph.D. Disxmtion University of New Hampshix.

Durham, NH, USA.

McNulty. S.G. and Abcr, J.D.. 1993. Effects of chronic nioogcn

additions on nitrogen cycling in a high-elevaIion spruce-fir

stand. Can. J. For. Res.. 23: 1252-1263.

McNulty, S.. Abcr. J.D., McLcllan, T.M. and Katt. S.M.. 1990.Nitrogen cycling in high elevation forests of the nor&astern

US in relation IO nitrogen deposition. Ambio. 19: 30-40.

McNulty, S.G.. Akr. J.D. and Boone, R.D., 1991. Spatial changesin forest floor and foliar chemisuy of spruce-fir forcso across

New England. Biogcochemisuy, 14: 13-29.Nihlgard. B.. 1985. The ammonium hypothesis - an additional

explanation to Ihe forest dieback in Europe. Ambio. 14: I-E.

Pastor. J., Aber, J.D.. McClaughcrry. C.A. and Melillo, J.M..1984. Aboveground production and N and P cycling along a

nitrogen mineralization gradient on Blackhawk island, Wis-

consin. Ecology. 65( I ): 256268.Peti. D.R.. Nicholas. N.S., Zedaker, S.M.. Miller-Weeks, M.M.

and Siccama, T.G.. 1992. Conditions and ruent ocnds inhigh-elevation red spruce population. In: C. Eagar and M.B.

Adams (Editors). The Ecology and Decline of Red Spruce in

the Eastern Unlud States. Springer-Verlag. New York, NY.USA.

Raloff. J., 1995. When nitrate reigns: air pollution can damage

fotesrs more than ties xvcal. Sci. News, 147: 90-91.Schulzc. E.-D., Lange, O.L. and Oren, R., 1989. Fores.1 Decline

and Air Pollution. Springer-Verlag. Berlin, Germany.Sheppard. LJ.. 1994. Causal mechanisms by which sulphau.

muau and acidity influence frost hardiness in red spruce:

review and hypothesis. New Phytol.. 127: 69-82.Shonle, W.C. and Smith. K.T.. 1988. Aluminum-induced calcium

deficiency syndrome in d&ning red spruce. Science. 240

1017-1018.Smimoff, N. and Suwar~Suwar~ G.R., 1985. N&au assimilation and

translocation by higher plan&: Comparative physiology and

ecological consequences. Physiol. Plant., 64: 133- 140.Smimoff. N.. Todd, P. and SUWM, G.R.. 1984. The occurrence

of nivatc reduction in the leaves of woody plan& Ann. BO L.54: 363-374.

Swan, H.S.D.. 1971. Relationships Between N&em Supply,Growth and Nutrient Concentrations in the Foliage of white

and Red Spruce. Pulp and Paper Research lnsti~ulnsti~u of Canada,Poinu Claitc, Que.. Canada.

Technicon Industrial Sysums. 1977. Nitrate and Niuite in Waur

and Scawaur. Indusnial Method l58-74W/A. Technicon In-

dusuial Sysums. Tarrytown. NY. USA.Technicon Industrial Systems. 1978. Ammonia in Waler and

Scawalrr. lndustial Method I547lW/B. Technicon Indus-

trial Systems, Tanyiown, NY, USA.ThornIon, F.C.. Joslin, J.D., Pier. P.A., Neufeld. H.. Seilcr. J.R.

and Hutherson. J.D., 1994. Cloudwater and ozone effects upon

high elevation red spruce: a summary of study resulls from

S.G. McNulq er al./Forcsr Ecology and Management 84 (1996) 109-121 121

Uhlu~op Mountain. Vlrgmia J. Environ. Qual.. 23: 1158-1167.

USDA Soil Conscrvatlon Service. 1989. interim soil survey repotifor H’mdsor Counry. Vcrrnoa~. US Department of AgnculturcSod Consewation Scrvlce. Washington, DC. USA.

USIX. 1968. Chmau Adas of the United Staus. United SlavesDcpanmcnt of Commerce. Washington, DC, USA.

van Dijk. H.F.G. and Roelofs. J.G.M.. 1988. Effccls of excessiveammomum dcposltion on the nutritional status and conditionof plnc needles. Physiol. Plan~., 73. 494-501.

Wcssman. C.A., Abcr. J.D.. Peterson, D.L. and hlelillo. J.M..1988. Fohar analysis usmg near inkared rcflecrancc spcc-

uoscopy. Can. 1. For. Rcs.. 18: 6- I I.Woodwell. GM. and Whittaker, R.H.. 1967. Pnrnary production

and the canon budget of the Brookhaven Forcs~ In: H.E.Young (Editor). Symposium on Primary Productiwty and Mm-eraI Cycling in Natural Ecosystems. University of MainePress. Orono. Maine, USA. pp. 151-161.

Zoenl. H.W. and Hue~tl. R.F., 1986. Nutrient supply and fores1decline in Southwesr-Germany. Water Air Soil Poll.. 31:449-462.

Forest Ecologyand Management

Submission of manuscripts.Manuscrlp!s should be submrtted in tnplicate. Authors from the Americas.. Australia, New Zealand and the Pacrfrc are requested!o send !helr manuscripts 10 FOreSr Ecology and Managemenl. Dr. Richard F. Fisher, Depanment of Forest Scrence. Texas A&MUnrversrty. College Station, TX77843-2135, USA; all other authors are requested to send thetr manuscripts to the EdtonalOfkce. forest Ecology and Management. P.O. Box 181, 1000 AD Amsterdam, The Netherlands,

Authors in Japan please note: UPOn request, Elsevter Scrence Japan WIII provide authors wtth a ltst of people who cancheck and Improve the English of their paper (belore submission). Please contact our Tokyo otfrce: Eisevler Sctence Japan,9-15 Htgashr-Azabu 1-chome. Mrnato-ku. Tokyo 106, Japan; tel. (03)-5561-5032: tax (03) 5561-5045.

Electronic manuscripts: Electrontc manuscripts have the advantage that there is no need for the rekeyrng of text, therebyavoidlng the possibrltty oi introducrng errors and resulttng tn reliable and last dekvery of proofs,For the lnrtlal SubmIssron of manuscripts for constderation. hardcoptes are suffrcrent. For the processtng 01 accepted papers.elec!romc versons are preferred. After ha/ accepfance, your disk plus two, final and exactly matchtng printed versons shouldSe suamltted together. Double densrty (DD) or hrgh densty (HD) drskettes (3.5 or 5.25 Inch) are acceptable. It IS Important that!he file saved IS In the nattve format of the wordprocessor program used. Label the disk with the name of the computer andwordprocessrng package used, your name, and the name of the ftle on the dtsk. Further rntormatton may be obtarned from thePubltsher.

All questlons artstng after acceptance of the manuscript, espectally those relattng to proofs, should be drrected to ElsevierEdr!orlal Services, Maytleld House, 256 Banbury Road, Oxford OX2 7DH, UK, tel. (01865) 314900, lax (01865) 314990.

Advertising information: Advenlslng orders and enqurries may be sent to: International: Elsevrer Scrence, AdvertrsrngZsbartment, The Boulevard, Langlord Lane. Krdlington, Oxford, OX5 lGB, UK; tel. (01865) 843565; fax (01865) 843952.USA and Canada: Wes:on Medra Assoctates, Dan Lipner, P.O. Box 1110. Greens Farms, CT 06436-l 110, USA; tel. (203)2-1 2500; fax (203) 261 0101. Japan: Elsevler Science Japan, Advertising Department, 9-15 Hlgashl-Azabu l-chome, Minato-ku. Tokyo 106, Japan; tel. (03)5561-5033; fax (03)-5561-5047.

US mailing info: Forest Ecology and Management (0378-l 127) IS published monthly by Elsevrer Science B.V. (Molenwert 1.Postous 211, 1000 AE Amsterdam). Annual subscription price in the USA is USS 2043.00 (valid In North, Central and SouthAqerrca). rncludtng atr speed dellvery. Second class postage rate IS pard at Jamaica, NY 11431.USA POSTMASTERS: Send address changes to Forest Ecology and Management Publicattons Expedrtrng. Inc.. 200 MeachamAvenue, Elmont. NY 11003.AIRFREIGHT AND MAILING In the USA by Publicatrons Expedtting Inc., 200 Meacham Avenue, Elmont, NY 11003.

Forest Ecology and Management has no page chargesL

For a full and complete Guide for Authors, please refer toForest EcologyandManagemen?, Vol. 80, Nos. 80, Nos. l-3,pp.287-290

J

Copyright 0 1996, Elsevier Science B.V. All rights reserved. 0376.11271961$15.OQ

$3 The paper used In this publicatton meets the requrrements of ANSVNISO 239.48-1992 (Permanence of Paper).

Prtnted In The Netherlands

.I