NITROGEN AVAILABILITY INFLUENCES REGENERATION OF …...nitrogen inputs in temperate forests. In the present study, we investigated the inﬂuence of nitrogen availability on patterns

1056

Ecological Applications, 12(4), 2002, pp. 1056–1070q 2002 by the Ecological Society of America

NITROGEN AVAILABILITY INFLUENCES REGENERATION OFTEMPERATE TREE SPECIES IN THE UNDERSTORY SEEDLING BANK

S. CATOVSKY1 AND F. A. BAZZAZ

Department of Organismic and Evolutionary Biology, Harvard University, Cambridge, Massachusetts 02138 USA

Abstract. While we have a good mechanistic understanding of effects of light avail-ability on tree seedling regeneration in forests, we know little of the processes controllingseedling responses to changes in nutrient availability in the understory (low light). Toexamine how nitrogen availability may influence patterns of seedling regeneration in mixedtemperate forests in eastern North America, we investigated nitrogen impacts on the dy-namics of both coniferous and broad-leaved tree species in the understory. In addition, wecompared these regeneration responses across coniferous (eastern hemlock) and broad-leaved (red oak) stand types. We applied nitrogen (0, 2.5, or 7.5 g·m22·yr21) to replicatedunderstory plots in three hemlock and three red oak dominated stands, and examinedseedling survival and growth for three coniferous and three broad-leaved species over twoyears. Nitrogen addition influenced different demographic stages in contrasting ways, lead-ing to a complex series of seedling responses to increased nitrogen availability in theunderstory. Hemlock was the only species to show a positive survival response to nitrogenaddition, and these effects only became apparent after the first growing season. In contrast,a suite of midsuccessional species (red maple, white pine, and red spruce) underwentnitrogen-induced declines in survival, particularly at early stages of regeneration. Bothsuccessional position and species’ habitat preferences emerged as better determinants ofspecies’ responses to nitrogen deposition in the understory than did leaf habit (coniferousvs. broad-leaved). Effects were often, but not always, more marked in hemlock stands(characterized by very low light availability), suggesting nitrogen-induced declines in seed-ling abundance were likely due to an imbalance between above- and belowground resources.In the present study, we have clearly demonstrated that nitrogen availability can be animportant determinant of understory seedling bank dynamics in mixed temperate forestsin eastern North America. Nitrogen-induced changes in seedling performance in the un-derstory will ultimately influence the spatial and temporal patterns of regeneration of thesemixed forests.

Key words: contrasting stand types; forest understory; mixed conifer–broad-leaved forests; ni-trogen availability; seedling bank; seedling regeneration; temperate forest community dynamics.

INTRODUCTION

Light is typically regarded as the main resource lim-iting seedling regeneration in forests (Pacala et al.1996), and has been the focus of a large body of eco-logical research (reviewed in Bazzaz and Wayne 1994).However, recent studies in forests suggest that plantresponses to light are often contingent on soil nutrientavailability (Coomes and Grubb 2000). Disturbanceevents coupled with individual tree effects on the en-vironment generate substantial spatial and temporalheterogeneity in both aboveground (light) and below-ground (nutrients, water) resources (Carlton and Baz-zaz 1998, Battaglia et al. 2000). While we have con-siderable mechanistic information on how variation inlight availability affects seedling regeneration dynam-ics (Pacala et al. 1994, Kobe et al. 1995), we know

Manuscript received 31 January 2001; revised 24 August2001; accepted 3 September 2001; final version received 18 Oc-tober 2001.

1 Present address: NERC Centre for Population Biology,Imperial College at Silwood Park, Ascot, Berks SL5 7PY,UK.

little of the specific processes controlling seedling re-sponses to changes in nutrient availability under forestcanopies (low light). Variation in nutrient availability,even under low light conditions, could be an importantinfluence on performance at the very early stages of aseedling’s life. These effects may in turn influence for-est composition, because an important component ofthe regeneration strategy of many species is the de-velopment of a seedling bank in closed-canopy forest(Marks and Gardescu 1998). Nutrient effects on un-derstory seedling regeneration will be complex, asseedling responses to soil nutrient availability will becontingent both on species’ life history traits (Reich etal. 1998b) and the surrounding resource environment(Chapin et al. 1987).

Determining seedling responses to variation in soilnutrient levels will not only improve our understandingof how natural variation in nutrient availability affectsforest dynamics, but may also help to better establishimpacts of the widespread human-induced fertilizationof natural ecosystems (Tilman et al. 2001). Agriculturalintensification and fossil fuel burning are increasing

August 2002 1057NITROGEN AVAILABILITY IN FOREST UNDERSTORY

TABLE 1. Details of study species used for seedling and seed addition components of experiment.

Species Common name Leaf habit

Succes-sional

position†Seed mass

(mg)‡Seed collection

years

Tsuga canadensisPicea rubensPinus strobusAcer saccharumAcer rubrumBetula alleghaniensis

eastern hemlockred sprucewhite pinesugar maplered maple§yellow birch

evergreen coniferousevergreen coniferousevergreen coniferousdeciduous broad-leaveddeciduous broad-leaveddeciduous broad-leaved

543543

2.53.5

207515

0.5

1996, 199819961996

1996, 19981997, 19981996, 1997

† Based on Baker’s (1949) table, where 1 is earliest and 5 is latest successional.‡ Based on empirical measurements of seed samples.§ Red maple seeds collected in spring.

nutrient loading in waterways, rainfall, and atmospher-ic particulate matter (Galloway et al. 1995), a greatproportion of which finds its way into natural ecosys-tems (Lovett 1994). Ammonium and nitrate inputs intotemperate forests are now of particular concern, be-cause nitrogen deposition is concentrated in highlypopulated areas of the world: Europe, North America,and now parts of East Asia (Holland et al. 1999). Nat-ural ecosystems in all these temperate regions are com-monly nitrogen limited (Vitousek and Howarth 1991),and nitrogen deposition might exert profound effectson the structure and function of these systems (Aberet al. 1998). A better mechanistic understanding ofseedling responses to nitrogen may significantly en-hance our ability to predict the impacts of increasednitrogen inputs in temperate forests.

In the present study, we investigated the influenceof nitrogen availability on patterns of seedling regen-eration in mixed temperate forests in eastern NorthAmerica, and contrast responses in stands of differinglight availability. Soil nitrogen might be a particularlyimportant factor controlling seedling regeneration inthese mixed forests, because such forests contain a mix-ture of evergreen coniferous and deciduous broad-leaved trees, which differ greatly in many aspects oftheir basic biology (Walters and Reich 1999). As ev-ergreen species tend to exhibit more conservative pat-terns of nutrient use than deciduous species (Chapin1980, Aerts 1995), and as coniferous species are typ-ically more slow-growing initially than broad-leavedspecies (Bond 1989, Reich et al. 1998a), we mightexpect these two major groups of trees to respond dif-ferently to variation in soil nutrient availability.

As well as comparing the responses of coniferousand broad-leaved tree seedlings to changing nitrogenavailability, the present study also contrasts their re-sponses under coniferous and broad-leaved canopies.Mixed conifer–broad-leaved forests in eastern NorthAmerica are commonly composed of a mosaic of eitherconiferous or broad-leaved dominated stands (Pastorand Mladenoff 1992). The prevalent conifer in theseforests (eastern hemlock, Tsuga canadensis (L.) Carr.)frequently forms single-species stands within a matrixof broad-leaved tree species (Pastor and Broschart

1990). Light availability is significantly lower in hem-lock stands than in red oak stands (dominant broad-leaved species in southern New England) (Catovsky2000), which could in turn constrain species’ responsesto increased nitrogen availability in the understory(e.g., Latham 1992, Crabtree and Bazzaz 1993, Canhamet al. 1996, Finzi and Canham 2000).

Uncoupling the relative roles of light vs. nutrientavailability in forest regeneration dynamics is difficult,given the experimental challenges posed by resourcemanipulations in forest systems. Previous studies haveaddressed the importance of resource interactions inforests mostly through examination of seedling re-sponses to: (1) natural variation in resource availability(Kobe 1996, Walters and Reich 1997), (2) trenchingtreatments (Coomes and Grubb 1998, Lewis and Tanner2000), and (3) different resource combinations in acommon garden/greenhouse (Canham et al. 1996, Wal-ters and Reich 2000). In the present study, we exper-imentally applied nitrogen in the field to manipulatenitrogen availability, a method that remarkably fewcontemporary studies have used (Grubb 1994, but seeFahey et al. 1998). We chose nitrogen levels that wouldact as a substantial perturbation to the natural nitrogencycle. Thus, although our treatments were not designedto directly simulate the human-induced fertilization oftemperate forests, the results of the experiment do pro-vide valuable information on the mechanisms of seed-ling response to increasing nitrogen availability.

The current study tests the hypothesis that coniferousand broad-leaved tree seedlings show contrasting re-sponses to nitrogen addition in the understory of mixedtemperate forests, and that these responses are contin-gent on stand type. We used three coniferous and threebroad-leaved tree species whose natural distributionsoverlap in southern New England (our study location).As we did not wish to confound effects of leaf habitwith those related to successional position, we chosespecies within each group that ranged from mid- to latesuccessional (Table 1, with data from Baker [1949]).We predicted that coniferous and late successional spe-cies would show smaller increases in growth and sur-vival in response to nitrogen than would broad-leavedand early successional species, because their evergreen

1058 S. CATOVSKY AND F. A. BAZZAZ Ecological ApplicationsVol. 12, No. 4

habit and slow early growth rates would correlate withless flexible patterns of nutrient uptake (Bazzaz 1996,Aerts and Chapin 2000) and more conservative patternsof nutrient use (Reich et al. 1995, 1998b). We alsopredicted that seedling responses would be greater inthe understory of broad-leaved tree stands than in hem-lock stands, as the higher resource availability in broad-leaved stands should allow seedlings to take advantageof additional nitrogen to a greater extent (Latham 1992,Canham et al. 1996, Walters and Reich 2000).

MATERIALS AND METHODS

Nitrogen addition treatments

We set up experimental plots in three hemlock- andthree red oak-dominated forest stands at Harvard Forestin central Massachusetts, USA (428329 N, 728119 W,elevation 340 m). The sites were established in the TomSwamp tract, and were chosen so that hemlock and redoak contributed over 50% of the basal area in each ofthree stands. Details of the stands are clearly describedin an earlier paper (Catovsky 2000). At each site, weset up nine 12-m2 plots (2 3 6 m) within a 20 3 20m square. Each experimental plot received one of threenitrogen addition treatments (0, 2.5, or 7.5 gN·m22·yr21), with the result that each treatment wasreplicated three times per site. Plots were arranged ina 3 3 3 grid, and nitrogen treatments were applied ina Latin square design (each row and column receivedone of each treatment level). In both 1998 and 1999,plots were given nitrogen eight times per year at ap-proximately three-weekly intervals beginning mid-April and ending mid-September. At each addition, ni-trogen was applied as prilled ammonium nitrate pellets(J. T. Baker, Phillipsburg, New Jersey, USA) using ahand-held seed spreader to ensure adequate and evenbroadcast. After addition, each plot was given 5 L ofwater to partially dissolve the pellets. We chose to ap-ply nitrogen as pellets, as preliminary observationssuggested that this method substantially reduced leafburn (pellets were too heavy to settle on leaves), andallowed for a more gradual release of nitrogen into thesoil between additions. Our treatments were designedto provide a substantial perturbation to the natural ni-trogen cycle, representing nitrogen treatments that add-ed considerably to background nitrogen mineralizationrates (4–8 g N·m22·yr21; Catovsky 2000) and currentlevels of nitrogen deposition (0.6 g N·m22·yr21; Mungeret al. 1998) at Harvard Forest.

Assessment of seedling regeneration patterns

Each plot was divided into three 4-m2 subplots (2 32 m), in order to examine nitrogen addition impacts ona different demographic component of seedling bankregeneration dynamics. The soil was not scarified inany of the seedling treatments, and any establishmentfrom seed occurred as a result of natural patterns ofgermination and emergence through the upper litterlayers.

Planted seedling plots: growth and survival of es-tablished seedlings.—Seeds of each study species werecollected from multiple trees at Harvard Forest in theautumn of 1996, 1997, and 1998 (Table 1). Any seedsthat were not to be used the following year were air-dried and stored at 48C. In late autumn 1997 and 1998,seeds were placed in cloth bags and buried in trays ofwet, coarse sand. These trays were placed outsidethrough the winter to stratify the seeds, and were col-lected the following spring. In 10-cm deep germinationflats, seeds were spread out evenly over a peat-basedpotting mix with added perlite, and then covered witha thin layer of vermiculite. These flats were then placedin shade-houses (neutral shade-cloth filtering 95% in-coming PAR) in an experimental garden at HarvardUniversity (Cambridge, Massachusetts, USA) in mid-April, and the seeds were left to germinate. The flatswere monitored once every three days and wateredwhen necessary. Seedlings began to germinate in earlyto mid-May. In mid-June, the seedling flats were trans-ported to Harvard Forest. At this point, most seedlingshad between two and four true leaves. In both 1998and 1999, five seedlings of each species were trans-planted into each ‘‘seedling’’ plot. Plots were wateredimmediately afterwards to prevent early transplant-re-lated mortality. Two weeks later, any dead seedlingswere replaced with new transplants. Typically, one totwo seedlings of each species died per plot.

Seed addition plots: seedling establishment fromseed in the absence of dispersal limitation.—Seedswere collected in the same way as those used for theplanted seedling plots. At the end of the autumn in1997 and 1998, seeds were broadcast over each subplotusing a hand-held seed spreader to ensure even spread.The number of seeds added per plot varied accordingto species (Table 2), and depended on seed quantitycollected, probable seed rain inputs, and seed viabilityestimated from germination trials. Although seed ad-dition plots were an attempt to remove effects of spatialvariation in seed rain, in some cases, the backgroundseed fall was much higher than the number of seedsadded experimentally, e.g., hemlock and birch seeds inhemlock stands in 1999 (Table 2, a vs. b). In thosecases, dispersal limitation may still underlie emergencepatterns in contrasting stand types.

Natural regeneration plots: patterns of regenerationfrom the soil seed bank.—In the third of each subplots,no seeds or seedlings were added. Instead, the naturaldynamics of seedlings emerging from the soil seedbank were monitored.

From June 1998 until September 1999, we tookmonthly censuses of seedling emergence and survivalin all study plots during the growing season. Seedlingswere individually tagged once they were planted oronce they emerged from the seed bank, and their sur-vival was subsequently monitored monthly during themain growing seasons of 1998 and 1999, and once afterthe 1998–1999 winter.


TABLE 2. Details of seed input and seedling emergence patterns for both (a) seed additionand (b) natural regeneration plots.

a) Seed addition plots

Species Year Stand typeSeeds added(seeds/m2)†

Emergence(seedlings/m2)‡

Tsuga canadensis 1998

1999

hemlockred oakhemlockred oak

100100250250

1.740.22

18.410.00

Picea rubens 1998

1999


100100200200

1.480.100.720.00

Pinus strobus 1998

1999


5050

150150

0.260.040.050.00

Acer saccharum 1998

1999


50507575

0.000.000.000.00

Acer rubrum 1998

1999


100100250250

0.670.120.200.01

Betula alleghaniensis 1998

1999


500500250250

26.124.25

36.861.33

b) Natural regeneration plots

Species Year Stand typeSeed rain

(seeds/m2)†‡§Emergence

(seedings/m2)‡

Tsuga canadensis 1998

1999


200.1

250060

0.000.00

15.400.00

Acer rubrum 1998

1999


75451210

0.470.060.040.01

Betula spp. 1998

1999


500100

2000120

15.290.36

28.430.37

† As seed germination commonly takes place the year after seed fall, seed inputs refer tothe previous year, i.e., 1997 and 1998. The temporal relationship between seed rain and emer-gence is less well coupled in Acer rubrum, as a proportion of seeds that fall in spring germinatethat year.

‡ Values in boldface represent significant differences between stand types for each year–species combination (post hoc tests, P , 0.05).

§ Based on seed trap sampling. See Catovsky (2000) for full details.

Seedling biomass and foliar nitrogen measurements

In August 1999, in every planted seedling plot, leafsamples were taken from one surviving seedling ofeach species belonging to each seedling cohort (1998,1999). For maple seedlings with relatively large leaves,three leaf discs (6-mm diameter) were punched. For allother seedlings, three to six leaves were removed, de-pending on seedling size. In these cases of unknownleaf area, samples were photocopied and leaf area sub-sequently calculated using NIH Image software v1.6(NIH, Bethesda, Maryland, USA). To determine leafchlorophyll concentrations, two to four of these leafdiscs/samples were placed in 3 mL dimethyl-form-

amide and left for 72 h at 48C in complete darkness(Garcıa and Nicolas 1998). Absorbance of the resultingsolution at 664 and 647 nm was measured using a Spec-tronics 20D spectrophotometer (Spectronic Unicam,Rochester, New York, USA), and chlorophyll concen-trations calculated from standard equations (Wellburn1994). The remainder of each set of leaf samples wasweighed after drying in an oven for 48 h at 708C, andleaf mass per unit area calculated (LMA, g/m2). In thisway, chlorophyll concentration could be expressed onboth a leaf area and mass basis, and also at the whole-plant canopy level.

At the end of the second growing season (late Sep-


tember 1999), all remaining seedlings were harvested.All seedlings were gently extracted from the soil andwashed to remove any soil. Seedlings from seed ad-dition and natural regeneration plots were weighedwhole after one week drying in an oven at 708C. Plantedseedling cohorts were additionally separated intoleaves, stems, and roots.

Environmental measurements

To determine how nitrogen addition affected seed-ling resource environment, above- and belowgroundresource availability were measured for each plot dur-ing the 1999 growing season. We used methods andsampling protocols that would provide a good indi-cation of ‘‘plant available’’ resources. Season-long in-tegrated nitrogen availability was determined using ionexchange resin bags placed in each nitrogen plot fromJune until October 1999 (Binkley and Vitousek 1989).The bags were constructed with 1.5 tablespoons (22mL) of mixed bed strong acid (cation) and strong base(anion) gel resins (Sybron Chemicals, Birmingham,New Jersey, USA) sealed in nylon mesh, and placedat a depth of 5 cm in the soil. After removal from thesoil, 4 g of dried resin (at 708C, overnight) was ex-tracted with 100 mL of 2 N potassium chloride solution(258C, 24 h), and then frozen immediately followingsuction-filtration. Ammonium and nitrate in all soil andresin extracts were measured using a LaChat contin-uous flow ion analyzer using methods 12-107-06-1-Aand 12-407-04-1-B (LaChat Instruments, Milwaukee,Wisconsin, USA). Blanks were created from resin bagsthat had been sealed in polyethylene bags for the lengthof the growing season. These resins were extracted inthe same way as the field bags to determine the lowerthreshold for detection.

Light availability was measured using canopy hemi-spherical photographs (Canham 1988). On a uniformlyovercast day, photos were taken at 5 cm and 1 m abovethe ground in the center of each plot using a fish-eyelens (1808) attached to a Canon camera body (with self-timer). The camera was leveled and oriented towardnorth before each photo. Two-hundred speed color slidefilm was used to capture the photos. The resulting im-ages were scanned into the computer with a slide scan-ner (500 dpi resolution) and analyzed using GLI/C soft-ware (courtesy of C. D. Canham, Institute of EcosystemStudies, Millbrook, New York, USA). Thresholds wereset manually in both blue and green spectra on eachimage separately, and the software used to calculatethe percentage of open sky, percentage of direct beamradiation, percentage of diffuse beam radiation, and thepercentage of global radiation (direct plus diffuse).

Water availability was determined by taking a soilcore (2-cm diameter, 20-cm depth) from the edge ofeach plot. Sieved soil (2-mm mesh) was dried at 1058Cfor 48 h and weighed to give gravimetric soil moisture.Further soil cores (10-cm diameter, 4–12 cm deep) weretaken to determine soil pH in the organic horizon. The

pH of sieved soil (5.6-mm mesh) was measured on amixture of 2 g of air-dried soil in 20 mL distilled water(pH 5.5) using an Orion 250A pH meter (Orion In-struments, Boston, Massachusetts, USA).

Statistical analysis

Multifactor analyses of variance were used to in-vestigate influence of nitrogen addition on understoryseedling bank development. Linear models includednitrogen addition as a continuous factor, and differentcombinations (plus interactions) of a variety of fixed,discrete factors (stand type, species, year, month, seed-ling age). Models also included a nested series of ran-dom factors. Site was nested within stand, and plot wasnested within site (stand), with the result that standmean square was tested over site, and site was testedover plot mean square. Significant multifactor inter-actions involving nitrogen were investigated by ex-amining the magnitude and significance of regressionslopes (dependent variable vs. nitrogen addition) usingstandard errors. Post hoc comparisons of means werecarried out using Scheffe multiple comparisons (Dayand Quinn 1989), and comparisons between slope co-efficients were carried out using T9, GT2, and Tukey-Kramer methods (Sokal and Rohlf 1995).

Similar linear models were used to investigatechanges in allocation between roots and shoots underdifferent nitrogen treatments, with leaf mass ratio(LMR, leaf/total biomass) and root mass ratio (RMR,root/total biomass) calculated for individual seedlings.Then, to distinguish ontogenetic changes from true al-location responses, we examined changes in Model IIlinear regression slopes (geometric mean regression)for the natural logarithm of root or shoot biomassagainst natural logarithm of total biomass, accordingto the method described by McConnaughay and Cole-man (1999). Allometric analyses were carried out forindividual species, nitrogen treatments, and standtypes.

When necessary, data were transformed to ensurethat the assumptions of analysis of variance were met(normality of residuals, homoscedascity). Commontransformations included natural logarithm (biomass),natural logarithm of square root (emergence), and thelogit function (proportional survival) (Sokal and Rohlf1995). For figures, the level to which data were pooledacross treatments was determined by the highest ordersignificant interaction in the models.

RESULTS

Environmental conditions

Nitrogen addition significantly increased availabilityof both ammonium and nitrate (Fig. 1; F1,46 5 21.5 and44.6, respectively, P , 0.0001), and these effects wereconsistent across stand types (no N 3 stand interac-tions, F1,46 , 1.9, P , 0.17). In contrast, nitrogen ad-dition did not significantly affect availability of other


FIG. 1. Effects of nitrogen addition (white 5 0, light gray5 2.5, and dark gray 5 7.5 g N·m22·yr21) on (a) ammoniumand (b) nitrate availability in hemlock and red oak stands(mean 6 1 SE), determined from mixed anion–cation resinbags.

FIG. 2. Effects of nitrogen addition (white 5 0, light gray5 2.5, and dark gray 5 7.5 g N·m22·yr21) on seedling emer-gence in seed addition plots in (a) hemlock and (b) red oakstands (mean 6 1 SE, pooled across years). Note the log scaleof the y-axis. Means and standard errors were calculated fromnatural-logarithm square-root transformed data and thenback-transformed. Values above histogram bars for each spe-cies represent slope coefficients calculated using transformeddata. Slopes significantly greater than zero (P , 0.05) areshown with an asterisk.

resources (F1,46 , 0.7, P , 0.40). Some resources, how-ever, differed significantly between stand types (lightand soil pH, both F1,46 5 8.4, P 5 0.006). Both lighttransmitted through the canopy and soil pH were sig-nificantly higher in red oak stands compared with hem-lock stands (6.0% vs. 1.4% light transmission, 4.62 vs.4.23 pH) (see further details in Catovsky 2000).

Nitrogen effects on seedling demography

In many treatments, there were significant species 3stand (3 year) interactions, which followed the basicseedling dynamics observed in a parallel natural de-mography experiment in the same stands (Catovsky2000). For clarity, these basic patterns are not ad-dressed here, but are described in more detail in thisadditional paper. Here we report predominantly on sig-nificant effects caused by the nitrogen treatment itself.

Emergence

In seed addition plots, all study species emergedfrom added seed except sugar maple, which sufferedhigh seed predation (Table 2). In natural regenerationplots, the major species to emerge from the soil seedbank were birch, hemlock, and red maple (Table 2). Atthese early stages of the life cycle, birch species couldnot be distinguished beyond genus. Nitrogen additionaltered some of these natural emergence patterns, butmainly in plots where seeds were added (significant N3 species 3 stand interaction only in seed additionplots, F4, 450 5 2.52, P 5 0.041; and not in naturalregeneration plots, F4, 450 5 0.17, P 5 0.95). In seedaddition plots, effects of nitrogen on seedling emer-gence were only significant in hemlock stands (Fig. 2),where nitrogen addition enhanced birch emergence by

almost three-fold (from 16 to 46 seedlings/m2), but de-creased white pine emergence.

Survival

Increasing nitrogen availability led to overall chang-es in first-year seedling survival in both seed additionand natural regeneration plots, but did not affect sur-vival in planted seedling plots (no significant N 3 spe-cies interactions, F5, 2350 , 1.31, P 5 0.26). In seedaddition plots, the differential responses of species toincreased nitrogen were contingent on both stand typeand year (significant N 3 species 3 stand and N 3species 3 year interactions, F4, 1815 5 3.68 and 3.92, P5 0.0054 and 0.0036, respectively). Nitrogen additiondecreased first-year survival for red maple, white pine,and red spruce seedlings, but only in hemlock stands(Fig. 3a) and not in red oak stands (Fig. 3b). For bothred maple and white pine, increasing nitrogen exac-erbated the lower overall seedling survival in hemlockstands vs. red oak stands. Nitrogen effects were stron-ger for red maple and red spruce in 1998 (Fig. 3c), andfor white pine in 1999 (Fig. 3d). We also found sig-


FIG. 3. Effects of nitrogen addition (white 5 0, light gray 5 2.5, and dark gray 5 7.5 g N·m22·yr21) on seedling survival(mean 6 1 SE, pooled across months). For seed addition plots, effects are shown for (a) hemlock and (b) red oak standsseparately (pooled across years), and for (c) 1998 and (d) 1999 separately (pooled across stands). (e) For natural regenerationplots, there were no higher order interactions, so effects are shown pooled across stand types and years. Means and standarderrors were calculated from logit transformed data and then back-transformed. Values above histogram bars for each speciesrepresent slope coefficients calculated using transformed data. Slopes significantly greater than zero (P , 0.05) are shownwith an asterisk.

nificant declines in red maple survival with increasingnitrogen in natural regeneration plots (Fig. 3e), but theeffects were not contingent on stand type or year (sig-nificant N 3 species interaction, F2, 891 5 3.68, P 50.025; but no significant higher order interactions, F1–6, 891

, 1.52, P . 0.17 for all).Although we were unable to detect nitrogen effects

on planted seedlings in the first year, influences of in-creasing nitrogen appeared after the first growing sea-

son. Nitrogen significantly increased hemlock survivalover the seedlings’ first winter, and led to nonsignifi-cant declines in red maple and sugar maple survival(Fig. 4a; significant N 3 species interaction, F5, 171 53.06, P 5 0.011). Hemlock survival in the secondgrowing season was also enhanced by increasing ni-trogen, but only in stands dominated by hemlock trees(Fig. 4b; significant N 3 species 3 stand interaction,F3, 100 5 2.72, P 5 0.049). Nitrogen addition also de-


FIG. 4. Effects of nitrogen addition (white 5 0, light gray5 2.5, and dark gray 5 7.5 g N·m22·yr21) on survival ofplanted seedlings after the first growing season (mean 6 1SE). Effects are shown pooled across stand types for over-winter survival (a), and separately for second-year survival(across full length of growing season) in hemlock (b) and redoak (c) stands. Due to low seedling numbers, second-yearsurvival for birch and sugar maple seedlings (in hemlockstands) could not be calculated. Means and standard errorswere calculated from logit transformed data and then back-transformed. Values above histogram bars for each speciesrepresent slope coefficients calculated using transformeddata. Slopes significantly greater than zero (P , 0.05) areshown with an asterisk.

TABLE 3. Effects of nitrogen addition (0, 2.5, or 7.5g·m22·yr21) on seedling allocation to leaves (leaf mass ratio,g/g) and roots (root mass ratio, g/g), pooled across standtypes and age classes.

Species

Nitrogen addition (g·m22·yr21)

0 2.5 7.5

a) Leaf mass ratioRed mapleSugar mapleWhite pineRed spruceHemlock

0.220.240.570.530.47

0.200.230.590.530.46

0.250.250.590.520.45

b) Root mass ratioRed mapleSugar mapleWhite pineRed spruceHemlock

0.440.470.220.220.26

0.420.470.220.220.28

0.390.460.210.230.28

creased red maple seedling survival in the secondgrowing season, but in contrast to seed addition plots,this effect was only significant in red oak stands (Fig.4c). In hemlock stands, there was a tendency for ni-trogen to decrease survival in second-year planted

seedling plots for the same species that showed nitro-gen-induced declines in survival in seed addition plotsin the first year, i.e., red maple, white pine, and redspruce (Fig. 4b). High overall seedling mortality inhemlock stands, however, reduced sizes of seedlingpopulations at the start of the second growing season.Low sample sizes increased variability in nitrogen ef-fects and reduced significance of these species-specificnitrogen-induced declines. In seed addition and naturalregeneration plots, mortality both within the growingseason and over winter was so high that sample sizeswere inadequate to examine regeneration dynamics ofany second-year seedlings.

Biomass, allocation patterns, and tissue chlorophyll

By the end of 1999, only planted seedling plots con-tained enough seedlings to harvest. Nitrogen additiondid not, however, significantly influence final seedlingbiomass in these plots (no significant N interactionterms, F1–4, 560 , 0.78, P . 0.38), and the result wasno different whether we considered individual seedlingbiomass or total plot-level biomass (F1–4, 210 , 1.33, P. 0.25).

Although final seedling biomass was not affected byincreasing nitrogen availability, seedling allocationpatterns did change significantly following nitrogen ad-dition treatments. Basic linear models of leaf and rootmass ratios produced no significant N terms (F1–4, 610 ,0.55, P . 0.46), showing that nitrogen addition had noeffect on either of these measures of allocation (Table3). However, when we accounted for size-related shiftsin allometry, we were able to detect some changes inseedling allocation patterns between leaves and roots(Figs. 5 and 6). For allocation to leaves, significanteffects of nitrogen were only observed when poolingacross stand types (equivalent to no significant N 3species 3 stand interaction), while there were cleardifferences in nitrogen-induced changes in root allo-cation between contrasting stand types. Red maple,


FIG. 5. Effects of nitrogen addition (circles, triangles, and squares denote addition of 0, 2.5, or 7.5 g N·m22·yr21, re-spectively) on planted seedling leaf allocation patterns (pooled across different seedling ages and stand types). Effects areshown as slopes from model II linear regression analysis (geometric mean regression) of log leaf biomass vs. log totalbiomass for each species-by-nitrogen combination (dotted lines, thin lines, and thick lines denote additions of 0, 2.5, or 7.5g N·m22·yr21, respectively). Slope values (shown in inset) that were significantly different from one another within a speciesare shown by different letters (P , 0.05, planned comparisons, sequential Bonferroni corrected).

white pine, and red spruce all showed significant in-creases in allocation to leaves (when accounting forsize) in response to nitrogen addition (Fig. 5), and par-ticularly significant decreases in allocation to roots, butonly in hemlock stands (Fig. 6).

Species also demonstrated some flexibility in leafcomposition in response to increasing nitrogen avail-ability in the understory. Nitrogen addition altered spe-

cies’ foliar chlorophyll concentrations (on both an areaand mass basis), although the effects were contingenton stand type (significant N 3 species 3 stand inter-action, F4, 197 5 3.01 for area and 3.48 for mass, P 50.019 and 0.0090, respectively). In red oak stands, in-creasing nitrogen availability led to significantly in-creased foliar chlorophyll for red maple, red spruce,and hemlock (Fig. 7b). In hemlock stands, only white


FIG. 6. Effects of nitrogen addition (circles, triangles, and squares denote addition of 0, 2.5, or 7.5 g N·m22·yr21, re-spectively) on planted seedling root allocation patterns (pooled across different seedling ages). Results are shown for seedlingsin hemlock stands only, as there were no significant shifts in red oak stands. Effects are shown as slopes from model II linearregression analysis (geometric mean regression) of log root biomass vs. log total biomass for each species-by-nitrogencombination (dotted lines, thin lines, and thick lines denote addition of 0, 2.5, or 7.5 g N·m22·yr21, respectively). Slope values(shown in inset) that were significantly different from one another within a species are shown by different letters (P , 0.05,planned comparisons, sequential Bonferroni corrected).

pine and hemlock showed significant increases in chlo-rophyll concentration, while red maple showed a sig-nificant decline (Fig. 7a). In contrast, whole-plant chlo-rophyll and the ratio of chlorophyll a to b were bothunaffected by nitrogen treatments (F1–4, 222 , 1.33, P. 0.26 for all N effects).

DISCUSSION

Nitrogen effects on understory seedlingbank dynamics

Demographic impacts of nitrogen availability in theunderstory of mixed conifer–broad-leaved forests were


FIG. 7. Effects of nitrogen addition (white 5 0, light gray5 2.5, and dark gray 5 7.5 g N·m22·yr21) on foliar chlorophyllconcentrations (mean 6 1 SE, pooled across seedling ages),expressed on a leaf area basis, in both (a) hemlock and (b)red oak stands. Values above histogram bars for each speciesrepresent slope coefficients produced from linear models.Slopes significantly greater than zero (P , 0.05) are shownwith an asterisk.

much larger than growth impacts (Fig. 8). Both seed-ling emergence and survival were significantly affectedby increasing nitrogen availability, while there were nodetectable effects on seedling biomass. Nitrogen ad-dition influenced different demographic stages in con-trasting ways, leading to a complex series of seedlingresponses to increased nitrogen availability in the un-derstory (Fig. 8). At early stages of regeneration (seedaddition plots), nitrogen addition led to declines inseedling survival of a suite of mid-successional species(red maple, white pine, and red spruce). In natural re-generation plots (no seeds added), red maple survivalalso declined in response to increased nitrogen avail-ability. No data were available for white pine or redspruce in these plots, as neither species had a mast yearduring the experiment. The negative effects of nitrogenaddition under low light conditions continued beyondthe first year of seedling growth, although in a lessconsistent manner (Fig. 8). Only in planted seedlingplots did any seedlings survive past their first growingseason in sufficient numbers to assess their subsequentperformance. In these cases, red maple showed bothsignificant (second year) and nonsignificant (over win-ter) declines in survival in response to increased ni-

trogen availability, while for white pine and red spruce,these declines were reduced to trends after the first year.These nitrogen-induced declines in survival are per-haps surprising, given the established view that nitro-gen generally enhances plant performance, but fertil-ization has previously been shown to hasten seedlingmortality under low light conditions (Hutchinson 1967,Grubb et al. 1996).

Only one species (hemlock) showed positive surviv-al responses to nitrogen addition in the forest under-story, and these effects only became apparent after thefirst growing season. When planted as seedlings intoour experimental plots, no species showed any responseto nitrogen addition in their first growing season. Thiseffect could have arisen as early seedling growth underthe more favorable and uniform shade-house conditionsprior to planting was a more important determinant offirst-year seedling performance than were field con-ditions (75% survival in first-year planted seedlingplots vs. 26% in seed addition plots). However, as treat-ment effects accumulated over time, nitrogen impactsbecame more significant, with hemlock seedling sur-vival increasing over winter and during the secondgrowing season in response to increased nitrogen avail-ability.

Influence of stand composition

The contingency of species’ responses to nutrientavailability might be particularly marked in the forestunderstory, as canopy trees each create a unique com-bination of resource conditions beneath them (Boettch-er and Kalisz 1990, Vesterdal and Raulund-Rasmussen1998). In the present study, seedling responses to ni-trogen addition were often strongly dependent on can-opy composition (hemlock vs. red oak stands). Envi-ronmental measurements in this current study and inan earlier study (Catovsky 2000) clearly demonstratedthat both light availability and pH were significantlylower in hemlock stands than in red oak stands. In thecurrent experiment, we found that often, but not al-ways, these differences in resource environment in-creased species’ sensitivity to nitrogen addition in hem-lock stands compared with red oak stands. Birch emer-gence was only enhanced by nitrogen addition in hem-lock stands, while nitrogen-induced declines insurvival of mid-successional species were more markedin general in hemlock stands. Thus, rather than in-creasing shade tolerance of other potential seedlingbank species (e.g., Kobe 1996), such as red maple andwhite pine, nitrogen addition actually accentuates cur-rent seedling dynamics in hemlock stands in favor ofa hemlock positive feedback.

Mechanisms underlying seedlingresponses to nitrogen

Contrary to our hypothesis, successional positionemerged as a more important determinant of species’responses to nitrogen than did leaf habit (coniferous


FIG. 8. Summary of main seedling demographic responses to increased nitrogen addition under each experimental treat-ment, showing species-specific effects that were significantly positive, 1; significantly negative, 2; or marginally negative,(2) (determined from regression coefficients). All other nonsignificant changes are indicated by NS. If effects were contingenton stand type or year, they are shown separately. When data were not available, the box is shaded gray.

vs. broad-leaved). For example, red maple and redspruce (similar tolerance ranking, but contrasting leafhabit) both showed similar declines in survival follow-ing nitrogen addition. Hemlock was the only specieswhose survival was increased by nitrogen addition inthe forest understory, and also represents the only studyspecies that is able to persist for decades in the deeplyshaded forest understory as a sapling (Kobe et al.1995). However, our results do not give overriding sup-port to the importance of successional position in de-termining nitrogen responsiveness, due to (1) shortageof data for the other late successional species in thestudy (shortage of sugar maple data was due to pooroverall seed and seedling survival), and (2) the some-

what positive response of yellow birch to nitrogen(emergence only).

In these forests, leaf habit is not well correlated withspecies’ habitat preferences, perhaps explaining thelack of predictive power provided by this species’ traitfor understanding responses to nitrogen availability. Infact, species’ site preferences emerged as an equallyparsimonious explanation for our results than did theirshade tolerance class. For example, two of the mainspecies that showed negative nitrogen responses in ourexperiment (red maple and white pine) are typicallyfound on poorer soils in New England forests than anumber of our other study species (e.g., hemlock andyellow birch) (Whitney 1991). In addition, red maple


has lower foliar nitrogen concentrations than many co-occurring forest species, again supporting the notionthat it has a preference for lower nutrient sites (Abrams1998). Successional position and species’ nutrient re-quirements may interact with one another in complexways to determine seedling responses to nitrogen in theunderstory (e.g., Mitchell and Chandler 1939, Spurr1956). Clearly, establishing the underlying physiolog-ical mechanisms for these responses will provide thebest method for untangling these effects.

In the present study, hemlock was the only speciesto respond positively to nitrogen addition under lowlight conditions, and was the only species to show con-sistent increases in leaf chlorophyll concentrations withnitrogen (although no species showed significant in-creases in whole-plant chlorophyll). Species differ inthe degree to which low light availability may constraintheir responsiveness to nutrients (Walters and Reich1996, 1997, Catovsky and Bazzaz 2000), and late suc-cessional species might have a better capacity to re-spond to increased nutrient availability under low lightscenarios than earlier successional species (Walters andReich 1997, 2000, Finzi and Canham 2000). For lessshade-tolerant species, manufacturing additional chlo-rophyll in low light environments may represent a met-abolic cost that is not offset by the resulting smallincrease in carbon gain.

In contrast, the more marked nitrogen-induced de-clines in survival of a number of species in the un-derstory of hemlock stands when compared with redoak stands strongly suggest that resource imbalancemay be driving the negative effects of nitrogen on seed-ling regeneration. In hemlock stands, where light avail-ability is particularly low, nitrogen addition is morelikely to create an imbalance between above- and be-lowground resource availability. Nutrient amendmenthas been shown to reduce seedling survival at low lightavailability (Grubb et al. 1996), although the actualmechanism has still not been clearly established. In thepresent study, we found that the species whose survivalsignificantly declined in nitrogen addition plots showedthe greatest allocational flexibility in response to ni-trogen. Red maple, red spruce, and white pine all al-located more biomass to leaves and less biomass toroots with increasing nitrogen availability.

From an optimality standpoint, this strategy appearsbeneficial (Bazzaz 1997). As nitrogen becomes lesslimiting, plants allocate less belowground, and increasetheir leaf area ratio to improve their capacity for carbongain aboveground (Peace and Grubb 1982). The greaterallocation flexibility of the earlier successional speciesin the present study agrees with ecological theory onsuccession (Bazzaz 1996), and has been previouslyfound for seedlings of these specific species growingunder varying combinations of light and soil resources(Canham et al. 1996). This flexibility in response isoften advantageous for responding to rapid increasesin resource availability, but may be detrimental in a

forest understory where resources are severely limiting.The reasons for shifts in allocation leading to increasedmortality are not clear in this case, but could be relatedto seedlings allocating carbon to improve growth po-tential rather than survival (Walters and Reich 2000).Kobe (1997) has shown that more shade-tolerant spe-cies had higher concentrations of nonstructural car-bohydrates in their roots than less tolerant species, andsuggested that a trade-off exists between growth andsurvival. An alternative hypothesis is that species thatallocate less biomass to roots in response to nitrogenaddition could be more sensitive to drought episodes(Canham et al. 1999).

Implications for forest dynamics

The structure and dynamics of temperate forests arecurrently threatened by a suite of novel human-inducedperturbations, such as increased nutrient loading, in-creased frequency of biological invasions (e.g., hem-lock woolly adelgid), changing human land-use pat-terns, and altered atmospheric chemical composition(Foster et al. 1997). Predicting the exact nature of anyfuture impacts is challenging given the uncertainties inthe magnitude and timing of these changes. A multipleresource perspective is useful in this regard, as manynovel perturbations involve changes in seedlings’ re-source environment (Field et al. 1992, Bazzaz and Ca-tovsky 2001). If we can understand how species behavealong a suite of resource axes, we should be able topredict patterns of forest dynamics under a range offuture change scenarios.

We have clearly demonstrated that nitrogen avail-ability can be an important determinant of understoryseedling bank dynamics in mixed temperate forests ineastern North America. Seedling responses to soil ni-trogen levels were a complex function of individualspecies’ traits and local resource environment (lightavailability, mostly). Differential survival responses tonitrogen availability increased the success of hemlockseedlings in the understory relative to other commonseedling bank species, such as red maple and whitepine, whose survival declined following nitrogen ad-dition. These nitrogen-induced changes were oftenstronger when hemlock dominated the canopy, and thussuggest that resource imbalance may be driving effectson seedling regeneration. For many tree species, main-taining a seedling bank is a critical component of theirregeneration strategy, and thus nitrogen-inducedchanges in seedling performance in the understory willultimately influence the spatial and temporal patternsof regeneration of these mixed temperate forests.

ACKNOWLEDGMENTS

We thank Chrissy Frederick, Joe LaCasse, and Henry Schu-macher for assistance in the field, and Richard Cobb, SteveCurrie, Matt Kizlinski, and David Orwig for help with theresin bag analysis. We are very grateful to Guntram Bauer,Christine Muth, Eric Macklin, Amity Wilczek, Kristina Stin-son, Jonathan Levine, and Richard Norby, who all contributed


substantially to the design and interpretation of the experi-ment, and to David Foster, John O’Keefe, and Glenn Motzkin,who all provided invaluable assistance with locating studysites. This work was supported by a NASA Earth SystemScience Graduate Fellowship, a Student Dissertation Grantfrom the Department of Organismic and Evolutionary Biol-ogy, and a Mellon Foundation Grant from Harvard Forest (allto S. Catovsky), and by the Harvard Forest Long-Term Eco-logical Research Program (NSF Grant DEB 9411795 to F. A.Bazzaz).

LITERATURE CITED

Aber, J., W. McDowell, K. Nadelhoffer, A. Magill, G. Bernt-son, M. Kamakea, S. McNulty, L. Rustad, and I. Fernandez.1998. Nitrogen saturation in temperate forest ecosystems.BioScience 48:921–934.

Abrams, M. D. 1998. The red maple paradox. BioScience48:355–364.

Aerts, R. 1995. The advantages of being evergreen. Trendsin Ecology and Evolution 10:402–407.

Aerts, R., and F. S. Chapin, III. 2000. The mineral nutritionof wild plants revisited: a re-evaluation of processes andpatterns. Advances in Ecological Research 30:1–67.

Baker, F. S. 1949. A revised tolerance table. Journal of For-estry 47:179–181.

Battaglia, L. L., S. A. Fore, and R. R. Sharitz. 2000. Seedlingemergence, survival and size in relation to light and wateravailability in two bottomland hardwood species. Journalof Ecology 88:1041–1050.

Bazzaz, F. A. 1996. Plants in changing environments: linkingphysiological, population, and community ecology. Cam-bridge University Press, Cambridge, UK.

Bazzaz, F. A. 1997. Allocation of resources in plants: state-of-the-science and critical questions. Pages 1–37 in F. A.Bazzaz and J. Grace, editors. Plant resource allocation. Ac-ademic Press, San Diego, California, USA.

Bazzaz, F. A., and S. Catovsky. 2001. Plants—from cells toecosystems: impacts of global environmental change. Pages94–111 in H. A. Mooney and J. G. Canadell, editors. En-cyclopedia of global change. Volume 2. John Wiley andSons, Chichester, UK.

Bazzaz, F. A., and P. M. Wayne. 1994. Coping with envi-ronmental heterogeneity: the physiological ecology of treeseedling regeneration across the gap-understory continu-um. Pages 349–390 in M. M. Caldwell and R. W. Pearcy,editors. Exploitation of environmental heterogeneity byplants: ecophysiological processes above- and below-ground. Academic Press, San Diego, California, USA.

Binkley, D., and P. Vitousek. 1989. Soil nutrient availability.Pages 75–96 in R. W. Pearcy, J. R. Ehleringer, H. A. Moo-ney, and P. W. Rundel, editors. Plant physiological ecology:field methods and instrumentation. Chapman and Hall,London, UK.

Boettcher, S. E., and P. J. Kalisz. 1990. Single-tree influenceon soil properties in the mountains of eastern Kentucky.Ecology 71:1365–1372.

Bond, W. J. 1989. The tortoise and the hare: ecology ofangiosperm dominance and gymnosperm persistence. Bi-ological Journal of the Linnean Society 36:227–249.

Canham, C. D. 1988. An index for understory light levels inand around canopy gaps. Ecology 69:1634–1638.

Canham, C. D., A. R. Berkowitz, V. R. Kelly, G. M. Lovett,S. V. Ollinger, and J. Schnurr. 1996. Biomass allocationand multiple resource limitation in tree seedlings. CanadianJournal of Forest Research 26:1521–1530.

Canham, C. D., R. K. Kobe, E. F. Latty, and R. L. Chazdon.1999. Interspecific and intraspecific variation in tree seed-ling survival: effects of allocation to roots versus carbo-hydrate reserves. Oecologia 121:1–11.

Carlton, G. C., and F. A. Bazzaz. 1998. Resource congruence

and forest regeneration following an experimental hurri-cane blowdown. Ecology 79:1305–1319.

Catovsky, S. 2000. Linking community dynamics and eco-system function in mixed conifer broad-leaved forests. Dis-sertation. Harvard University, Cambridge, Massachusetts,USA.

Catovsky, S., and F. A. Bazzaz. 2000. The role of resourceinteractions and seedling regeneration in maintaining a pos-itive feedback in hemlock stands. Journal of Ecology 88:100–112.

Chapin, F. S., III. 1980. The mineral nutrition of wild plants.Annual Review of Ecology and Systematics 11:233–260.

Chapin, F. S., III, A. J. Bloom, C. B. Field, and R. H. Waring.1987. Plant responses to multiple environmental factors.BioScience 37:49–57.

Coomes, D. A., and P. J. Grubb. 1998. Responses of juveniletrees of above- and belowground competition in a nutrient-starved Amazonian rain forest. Ecology 79:768–782.

Coomes, D. A., and P. J. Grubb. 2000. Impacts of root com-petition in forests and woodlands: a theoretical frameworkand review of experiments. Ecological Monographs 37:171–207.

Crabtree, R. C., and F. A. Bazzaz. 1993. Seedling responseof four birch species to simulated nitrogen deposition: am-monium vs. nitrate. Ecological Applications 3:315–321.

Day, R. W., and G. P. Quinn. 1989. Comparisons of treatmentsafter an analysis of variance in ecology. Ecological Mono-graphs 59:433–463.

Fahey, T. J., J. J. Battles, and G. F. Wilson. 1998. Responsesof early successional northern hardwood forests to changesin nutrient availability. Ecological Monographs 68:183–212.

Field, C. B., F. S. Chapin, III, P. A. Matson, and H. A. Moo-ney. 1992. Responses of terrestrial ecosystems to thechanging atmosphere: a resource-based approach. AnnualReview of Ecology and Systematics 23:201–235.

Finzi, A. C., and C. D. Canham. 2000. Sapling growth inresponse to light and nitrogen availability in a southernNew England forest. Forest Ecology and Management 131:153–165.

Foster, D. R., J. D. Aber, J. M. Melillo, R. D. Bowden, andF. A. Bazzaz. 1997. Forest response to disturbance andanthropogenic stress. BioScience 47:437–445.

Galloway, J. N., W. H. Schlesinger, H. Levy, II, A. Michaels,and J. L. Schnoor. 1995. Nitrogen fixation: anthropogenicenhancement—environmental response. Global Biogeo-chemical Cycles 9:235–252.

Garcıa, A. L., and N. Nicolas. 1998. Influence of the degreeof solvent impurity on the spectrophotometric determina-tion of chlorophylls in 80% aqueous acetone and dimethylformamide. Application to non-abrasive extraction ofleaves of Citrus aurantium. Photosynthetica 35:545–550.

Grubb, P. J. 1994. Root competition in soils of different fer-tility: a paradox resolved? Phytocoenologia 24:495–505.

Grubb, P. J., W. G. Lee, J. Kollman, and J. B. Wilson. 1996.Interaction of irradiance and soil nutrient supply on growthof seedlings of ten European tall-shrub species and Fagussylvatica. Journal of Ecology 84:827–840.

Holland, E. A., F. J. Dentener, B. H. Braswell, and J. M.Sulzman. 1999. Contemporary and pre-industrial globalreactive nitrogen budgets. Biogeochemistry 46:7–43.

Hutchinson, T. C. 1967. Comparative studies of the abilityof species to withstand prolonged periods of darkness. Jour-nal of Ecology 55:291–300.

Kobe, R. K. 1996. Intraspecific variation in sapling mortalityand growth predicts geographic variation in forest com-position. Ecological Monographs 66:181–201.

Kobe, R. K. 1997. Carbohydrate allocation to storage as abasis of interspecific variation in sapling survivorship andgrowth. Oikos 80:226–233.


Kobe, R. K., S. W. Pacala, J. A. Silander, Jr., and C. D.Canham. 1995. Juvenile tree survivorship as a componentof shade tolerance. Ecological Applications 5:517–532.

Latham, R. E. 1992. Co-occurring tree species change rankin seedling performance with resources varied experimen-tally. Ecology 73:2129–2144.

Lewis, S. L., and E. V. J. Tanner. 2000. Effects of above-and belowground competition on growth and survival ofrain forest tree seedlings. Ecology 81:2525–2538.

Lovett, G. M. 1994. Atmospheric deposition of nutrients andpollutants in North America: an ecological perspective.Ecological Applications 4:629–650.

Marks, P. L., and S. Gardescu. 1998. A case study of sugarmaple (Acer saccharum) as a forest seedling bank species.Journal of the Torrey Botanical Society 125:287–296.

McConnaughay, K. D. M., and J. S. Coleman. 1999. Biomassallocation in plants: ontogeny or optimality? A test alongthree resource gradients. Ecology 80:2581–2593.

Mitchell, H. L., and R. F. Chandler, Jr. 1939. The nitrogennutrition and growth of certain deciduous trees of north-eastern United States. Black Rock Forest Bulletin 11:1–94.

Munger, J. W., S. M. Fan, P. S. Bakwin, M. L. Goulden, A.H. Goldstein, A. S. Colman, and S. C. Wofsey. 1998. Re-gional budgets for nitrogen oxides from continental sourc-es: variations of rates for oxidation and deposition withseason and distance from source regions. Journal of Geo-physical Research—Atmospheres 103(D7):8355–8368.

Pacala, S. W., C. D. Canham, J. Saponara, J. A. Silander, Jr.,R. K. Kobe, and E. Ribbens. 1996. Forest models definedby field measurements: estimation, error analysis and dy-namics. Ecological Monographs 66:1–43.

Pacala, S. W., C. D. Canham, J. A. Silander, Jr., and R. K.Kobe. 1994. Sapling growth as a function of resources ina north temperate forest. Canadian Journal of Forest Re-search 24:2174–2183.

Pastor, J., and M. Broschart. 1990. The spatial pattern of anorthern conifer-hardwood landscape. Landscape Ecology4:55–68.

Pastor, J., and D. J. Mladenoff. 1992. The southern boreal-northern hardwood forest border. Pages 216–240 in H. H.Shugart, R. Leemans, and G. B. Bonan, editors. A systemsanalysis of the global boreal forest. Cambridge UniversityPress, Cambridge, UK.

Peace, W. J. H., and P. J. Grubb. 1982. Interaction of lightand mineral nutrient supply in the growth of Impatiensparviflora. New Phytologist 90:127–150.

Reich, P. B., B. D. Kloeppel, D. S. Ellsworth, and M. B.Walters. 1995. Different photosynthesis-nitrogen relations

in deciduous hardwood and evergreen coniferous tree spe-cies. Oecologia 104:24–30.

Reich, P. B., M. G. Tjoelker, M. B. Walters, D. Vanderklein,and C. Buschena. 1998a. Close association of RGR, leafand root morphology, seed mass and shade tolerance inseedlings of nine boreal tree species grown in high and lowlight. Functional Ecology 12:327–338.

Reich, P. B., M. B. Walters, M. G. Tjoelker, D. Vanderklein,and C. Buschena. 1998b. Photosynthesis and respirationrates depend on leaf and root morphology and nitrogenconcentration in nine boreal tree species differing in rel-ative growth rate. Functional Ecology 12:395–405.

Sokal, R. R., and F. J. Rohlf. 1995. Biometry. W. H. Freemanand Company, New York, New York, USA.

Spurr, S. H. 1956. Forest associations in the Harvard Forest.Ecological Monographs 26:245–262.

Tilman, D., J. Fargione, B. Wolff, C. D’Antonio, A. Dobson,R. Howarth, D. Schindler, W. H. Schlesinger, D. Simberloff,and D. Swackhamer. 2001. Forecasting agriculturally driv-en global environmental change. Science 292:281–284.

Vesterdal, L., and K. Raulund-Rasmussen. 1998. Forest floorchemistry under seven tree species along a soil fertilitygradient. Canadian Journal of Forest Research 28:1636–1647.

Vitousek, P. M., and R. W. Howarth. 1991. Nitrogen limi-tation on land and in the sea: how can it occur? Biogeo-chemistry 13:87–115.

Walters, M. B., and P. B. Reich. 1996. Are shade tolerance,survival, and growth linked? Low light and nitrogen effectson hardwood seedlings. Ecology 77:841–853.

Walters, M. B., and P. B. Reich. 1997. Growth of Acer sac-charum seedlings in deeply shaded understories of northernWisconsin: effects of nitrogen and water availability. Ca-nadian Journal of Forest Research 27:237–247.

Walters, M. B., and P. B. Reich. 1999. Low-light carbonbalance and shade tolerance in the seedlings of woodyplants: do winter deciduous and broad-leaved evergreenspecies differ? New Phytologist 143:143–154.

Walters, M. B., and P. B. Reich. 2000. Seed size, nitrogensupply, and growth rate affect tree seedling survival in deepshade. Ecology 81:1887–1901.

Wellburn, A. R. 1994. The spectral determination of chlo-rophylls a and b, as well as total carotenoids, using varioussolvents with spectrophotometers of different resolution.Journal of Plant Physiology 144:307–313.

Whitney, G. G. 1991. Relation of plant species to substrate,landscape position, and aspect in north central Massachu-setts. Canadian Journal of Forest Research 21:1245–1252.

Documents

NITROGEN AVAILABILITY INFLUENCES REGENERATION OF …...nitrogen inputs in temperate forests. In the present study, we investigated the inﬂuence of nitrogen availability on patterns