Tree growth, foliar chemistry, and nitrogen cycling across a nitrogen deposition gradient in southern Appalachian deciduous forests

The declining health of high-elevation red spruce (Picea ruhens Sarg.) and Fraser lir (Abies fraseri (Pursh) Poir.) in the southern Appalachian region has long been linked to nitrogen (N) deposition. Recently. N deposition has also been proposed as a source of negative health impacts in lower elevation deciduous forests. In 1998 we established 46 plotc on six sites in North Carolina and Virginia dominated by American bcech (Fagus glaizdifoliu Ehrh.), sugar maple (AC*PF ~ ~ r c c ' l z u ~ ~ t n ~ Marsh.), and yellow birch (Befulu crlle~haniensis Britt). We evalualed several response variables across an N deposition gradient, including annual basal area growth; foliage percent N, Al, P, K, Mg, and Ca; and forest floor percent N. Mg. and C, pH, and potential net nitrification and N mineralization rates. We found a significant linear relationship between N deposition and basal area growth in sugar maple, but not in American beech or yellow birch. In addition. Mie found a significant relationship between N deposition and foliar chemistry (foliar %N in all species, foliar MgIN and % K i n sugar maple, and %P in sugar n~aple and yellow birch). Foliar %N of the three studied species was high relative to values reported in other st~tdies in the United States and Canada. Several forest floor response variables (%N, CIN, pH. 1\/1g/N, and potential net nitrilication and N mineralization rates and nitrification/mineralizalion fractions) were also correlated with N deposition. The correlations between the above response variables and N deposition are consistent with the ~nfluence of chronic N depositio~i on forested ecosystems measured in other regions and suggest that chronic N deposition rnay be influencing forest structure and chemistry within the southern region. K6sumC : Le dkpCrissement de 1'Cpinette rouge (Picea ruberzs Sarg.) et du sapin de Fraser (Abies .fraseri (Pursh) Poir.) croissant en altitude dans le sud des Aypalaches a depuis longte~nps kt6 associk aux dkpbts d'azote (N). Rkccmment. des impacts n6gatiSs sur la sante des foritts decidues croissant 5 pluh faible altitude ont aussi CtC attribuks aux dCp6ts de N. En 1998, nous avons i-tabli 46 placettes, en Caroline du Nord et en Virginie, dans six stations dominks par le lietre 5 grandes Seuilles (Fagus giundifolia Ehrh.). l'erable B sucre (Acer ~ac~charuurn Marsh.) et le bouleau jaune (Betula alleg~tarzien~i~ Britt.) Nous avow, kvalu6 le colnporternent de plusicurs variables le long d'un gradient de dkpbts de N, incluant : la croissance annuellc en surface ternere, le pourcentage de N, A1, P, IS, Mg et Ca dans le feuillage, le pourcentagc de N, Mg et C dans la couverture niorte. le pH et ler t aw potentiels de nitrification nette et de rninkralisation. Nous awns observe une relation linPaire significatice entre les dCpCits de N et la croissance en surface tenikre chez I'Crable a sucre mats non cher: le hetre i grandes feu~lles ni chez le bouleau jaune. De plus, nous avons observC une relation significative entrc les ddpitts de N et lcs caractkristiques chim~qucs du feuillage (pourcentage de N foliaire chez toutes les especer, Mg/N et pourcentape de K foliaire chez I'Crable ii sucre et pourcentage de P foliaire chez lYCrable & sucre et Ic bouleau jaunc. Le pourcentage de N dans les feuilles des trois esp&eer dtudikes Ctait 61evC cornparativement B ce qrti a CtC rapport& dans d'autres Ctudes aux E.-U. et au Canada. PluGeurs variables de la couverture morte (%N, CIN. pH, MglN et taux de minkralisatiorl et dc nitrification nette potcntiels et rapport nitrificationlmindralisation) 6taient Cgalenient correlke, alee les dkp8ts de N. Lea corrklations entre les variables ci-dessus et les dCpbts de N sont consistantes avec lcs effets dcs dkp6ts chroniques de N sur Ies ecosystkmes forestiers mesurCs dans d'autres regions et indiqueiit que le\ dkp8tr chi-ontques de N pourralcnt ~nfluencer la structure et les caractkristiq~tes chimiqrtes de la for& dans cette regton du rud. [Traduit par la RCdaction] Introduction highest N deposition values are found in the northern part of the region (National Atrnospheric Deposition Program 1998). In the northeastern United Slates, the highest nitrogen (N) The 1970 Clean Air Act and its 1990 amendments set redeposition values generally occur along the southwe5tcrn duced thresholds for annual N and sulfur (S) deposition in edge of the region, and in the southeastern United States, the response to increases in N deposition that have occurred I Received 13 September 2004. Acceptect 8 June 2005. Published on the NRC Research Press Web site at http:1lcjfr.n1*c.ca on 8 Septernbcr 2005. I J.L. Boggs,' S.G. RiIcNulty, M.J. Gavazzi, and J. Moore Myers. Southern Global Change Program, Southern Research Station, USDA Forest Service, 920 Main Campus Drive. Suite 300, Venturc Ccntcr 11, Raleigh, NC 27606, LJSA. 1 'Correspo~lding author (e-mail: [email protected]) Cart. J. For. Rcs. 35: 1901-191 3 (2005) doi: 10.1 I 39iX05128 O 2005 NRC Canada Can, J. For. Res. Vol. 35, 2005 'since the beginning of the industrial era (Hicks et al. 1990). Implementation of the Clean Air Acts helped reduce S emissions, but trcnd analyses of N deposition have not been significantly reduced from levels in the 1980s (NADP 5998). Historically, the greatest concerns about the in~pacts of N deposition on forests focused on tfie northeastern United States. particularly on higti-elevation spruce-fir (Picecr-Ahies) forests that rcccivc sonie of the highest levels of N deposition as a result of clctud deposition (Fleilman et al. 2000). Clouri deposition coupled with wet and dry deposition can result in a total N depositioii 6 to 20 linies higher than N cfepositio~i levels in lower elevation forests (Baumgardner ct al. 2003). Previous studies have shown that forest ecosystems in the northeasteni United States have been altered by N deposition. McN~lIty et al. (1991 f examined forest floor and foliar chemistry across northeast spruce-fir forests along a gradient of N deposition. Their results showed that there was a correlation bet ween N deposition and forest floor inorganic N and foliar lignin/N ratio. They suggested that these relationships across the region indicated that these forests may be progressing toward nlore advanced N-saturated conifitions (defined as N in excess of plant and microbial biological demand (Aber et al. 1989)). In another study, McNulty et al. (1996) used N amendments to simulate the potential impacts of increased N dcposition on spruce-fir forcst structure and function. Their results showed a decrease in tree growth rates associated with a foliar nutricnt inibalancc (i.e., N, calcium (Ca), magnesium (Mg)), increases in forest floor %N, and increases in forest floor potential net nitrification/~ni~icrdlization ratios. Foliar nutrient imbalances are generally thought to be caused by either soil depletion or the addition of nutrients. However, Schaberg et al. (2001) found that Ca, in particular. can be leached from the needles of red sprucc (Picm I-uherz.r Sarg.) during exposurc to acidic cieposition, thus predisposing the tree species to Ca deficiency. Schaberg et al. (1997) and Perkins et al. (2000) fbund that increases in fctrcst floor %N increase freezing injury to spruce foliage, thus making the tree species more susceptible to mortality. McLaughiin ct al. (1998) reported that N deposition in southern Appalachian spruce-fir forests can cause leaching of Ca and Mg from soils where base cation storcs are very low and tlie ability of the ecosystems to retain N is minimal. In another study on southern Appalachia11 spruce-fir forests, Flum and Nodvin (1995) found the highest occurrence of nitrate (NO3-N) leaching (an indicator of N saturation) i n watersheds dominated by a sprucc-fir overstory. Fenn ct al. (1998) found that N saturation conditions are more likely to develop in mature forests that receive high N deposition iuputs and have low soil CIN ratios. Since the 1980s, there have been indications that N deposition may be ~iegatively affecting forest ecosystems, particuIarly in the southern Appalachian region. Rruck (1984) and Bruck et al. (1989) indicated a relationship between N and S deposition and reduced growth and early senescence of needles. This relatioilship suggests a direct impact of N deposition on species dynamics that may prcdispose species to growth decline. Additional studies in the region (Southern Appalachian Man and Biosphere 1996; F'criiandez and Adams 2000) suggest that N saturation is not confined to highelevation conifer forests, but may also cicvelop in deciduous forests. The cited htudies suggest how N deposition may change forest structure and function and increase the rate of forest mortality relative to low N deposition areas. The southern Ap alachian physiographic region consists R of about 15. I x 10 ha; 70% of the region is forested, with deciduous trees accounting for 47% of the forested area (Roone and Aplei 1994). Southern Appalachian deciduous forests corltain the largest plant species diversity of any forcst ecosystem in North America (Grcat Smoky Mountain National Park, http://www.nps.gov/grsm/pphtml/facts.htm1). Thc area is also a centcr for recscation in the eastern United States, with over 9 million people visiting the Great Smoky Mountains National Park every year (Great Smoky Mount;lin National Park). The loss of deciduous forest tree species could have significant ecological and economic implications for the region. The objectives of this study are to correlate the response of tree species basal area growth rates, foliar chemistry, and N cycling with a gradient of N deposition in the southern Appalachians and to determine if any of these relationships could be characteristic of declining forest health. Material and methods Plot location During the sulnnler of 1998, we established 46 plots on six deciduous forest sites in the southern Appalachian Mountains (Table 1). Figure 1 shows the locations of tlie six sites with the estimated N deposition across the area. The 46 plots, each with a radius of 10 m, were located at elevations between 1 126 to 166 1 m and were divided evenly along each of the cardinal aspects on each site. At each site a semirandomized block design was applied to select plots that contained greater than 70% combined basal area of yellow birch, Arnerican beech, and sugar maple. Since site selection was based on the level of N deposition inputs (Fig, 1) and on occurrence and composition of these three deciduous tree species. the distance between plots and the slope of the plots varied among sites. Climate and estimated wet N deposition inputs Both temperature and precipitation data were extracted from the Vegetatiomosystem ModeIitig and Analysis Project (VEMAP) (Kittel et al. 2000) monthly climate data set. We identified VEMAP grid cell numbers for each sample plot by a GIs overlay. Precipitation and temperature data for years 1978 through 1993 were pulled from the historical climate table, and the data for years 1994 through 1998, from the Hadley Centre's climate change scenario table (http:Nwww.metoffice. conllrcsearcNhadleycentre). Annual averages were then calculated in centimetres for precipitation and in degrees Celsius for tcinpcraturc for each plot. The wet N deposition rates were determined for each plot by kriging mean annual N deposition data from 1978 through 1990 from 62 sites in thc southei-n 13 states from the National Atmospheric Deposition Program (NADP 1998). Several krig algorithms were evaluated. The spatial distribution of NADP sites was log skewed, and with consideration of the work of Finkelstein (1983), a spherical model with a 100-krn separation distance was selected to mininlize enor in the krig estimates. We then acquired the site iianles and coordinates (latitude, longitude) for all points from NADP; a point site coverage