Acid-base characteristics of the Grass Pond watershed in the Adirondack Mountains of New York State , USA : interactions among soil , vegetation and surface waters

Grass Pond watershed is located within the southwestern Adirondack Mountain region of New York State, USA. This region receives some of the highest rates of acidic deposition in North America and is particularly sensitive to acidic inputs due to many of its soils having shallow depths and being generally base poor. Differences in soil chemistry and tree species between seven subwatersheds were examined in relation to acid-base characteristics of the seven major streams that drain into Grass Pond. Mineral soil pH, stream water BCS (base-cation surplus) and pH exhibited a positive correlation with sugar maple basal area ( p = 0.055; 0.48 and 0.39, respectively). Black cherry basal area was inversely correlated with stream water BCS, ANC (acid neutralizing capacity) c and NO − 3 (p = 0.23; 0.24 and 0.20, respectively). Sugar maple basal areas were positively associated with watershed characteristics associated with the neutralization of atmospheric acidic inputs while in contrast, black cherry basal areas showed opposite relationships to these same watershed characteristics. Canonical correspondence analysis indicated that black cherry had a distinctive relationship with forest floor chemistry apart from the other tree species, specifically a strong positive association with forest floor NH4, while sugar maple had a distinctive relationship with stream chemistry variables, specifically a strong positive association with stream water ANC c, BCS and pH. Our results provide evidence that sugar maple is acid-intolerant or calciphilic tree species and also demonstrate that black cherry is likely an acid-tolerant tree species.


Introduction
The Adirondack Mountain region of New York State is characterized by high elevations, granitic bedrock and is particularly sensitive to acidic atmospheric inputs due to many of the soils in this region having shallow depths and generally being base poor (Ito et al., 2002).Acidic deposition alters soils through the depletion of calcium (Ca 2+ ) and other plant nutrient cations (Mg 2+ , K + ) and the mobilization of inorganic monomeric aluminum (Al) (Lawrence et al., 1997).Both of these factors may stress forest vegetation and deleteriously affect water quality (Cronan and Grigal, 1995;Horsley et al., 2000;MacAvoy and Bulger, 1995).Countering natural forces that may reduce the deleterious effects of acidic deposition include nutrient recycling, mineral weathering, and the release of exchangeable base cations (McFee et al., 1977).
It is generally recognized that within the northern hardwood forests, sugar maple (Acer saccharum) and other species such as American basswood (Tilia americana), eastern hophornbeam (Ostrya virginiana) and white ash (Fraxinus americana) require base-rich soils and are acid intolerant, while American beech (Fagus grandifolia), and red spruce (Picea rubens) are acid-tolerant species (Christopher et al., 2006;Duchesne et al., 2005;Fujinuma et al., 2005;Mitchell et al., 2003).Sugar maple often grows poorly in soils that are low in Ca and Mg, high in Al and have a low pH (Burns and Honkala, 1990;Van Breemen et al., 1997).Acidintolerant species such as sugar maple have been shown to respond to liming with significantly increased diameter growth, improved crown vigor, increased stem basal area increment, and increased seedling density, while acid-intolerant species such as black cherry and American beech have not shown a response to liming (Long et al., 1997;Moore et al., 2012).Previous studies indicate black cherry (Prunus serotina) might be an acid-tolerant species although research on this tree species in relation to site acid-base characteristics is very limited (Aguilar and Arnold, 1985;Godefroid et al., 2005;Long et al., 2009;Lorenz et al., 2004).
In addition to species-specific nutrient requirements, various tree species will differ in their cycling of nutrients and base cations, which may result in differences in nutrients and base cation concentrations of soils in forest stands.For instance, sugar maple litter generally has low lignin and C : N concentrations and relatively high N concentrations causing rapid decomposition and higher rates of net nitrification compared to American beech, red maple, black cherry and red spruce that have relatively high lignin concentrations (Aber et al., 2003;Adams and Angradi, 1996;Lovett and Rueth, 1999;Melillo et al., 1982;Mitchell et al., 1992;Page and Mitchell, 2008).In general, it has been suggested that there may be connections between tree species, soil and litter nutrient cycling on sites, whereby tree species with highly labile litter yield high rates of nutrient cycling and soil nutrient concentrations (Hobbie, 1992;Page and Mitchell, 2008;Van Breemen, 1995).Additionally, Dijsktra (2003) found that in forest stands in Connecticut net Ca mineralization and Ca concentrations in the mineral soil was significantly higher beneath sugar maple than beneath beech and white ash.
Within the Adirondack Park, sulfate and nitrate concentrations in precipitation generally decrease from the southwest to the northeast and increase with elevation (Ito et al., 2002).Stream and lake nitrate concentrations are generally consistent with this pattern of nitrogen deposition with decreasing values from west to east across New York and New England (Aber et al., 2003).Grass Pond watershed, located within the southwestern Adirondack Mountains region of New York, receives some of the highest rates of acidic deposition in the region (Driscoll et al., 1991(Driscoll et al., , 2003;;Ito et al., 2002).Despite receiving similar rates of acidic deposition, subwatersheds within Grass Pond watershed differentially neutralize acidic surface waters (Ito et al., 2007).Understanding the relationships between watershed characteristics and acid neutralizing capacity provides insight into the sensitivity or resilience of watersheds and the relative importance of different tree species with respect to the deleterious impacts of acidic deposition and the potential for different forest management scenarios and strategies.
The objective of our study was to evaluate the relationships between stream and soil acid-base chemistry as a function of the presence and abundance of acid-tolerant and acidintolerant tree species within Grass Pond subwatersheds.Black cherry and sugar maple tree species were the primary focus of our research study due to the presence of both tree species within Grass Pond subwatersheds, the economic  importance of black cherry for the timber industry and sugar maple for the maple syrup industry as well as the noted decline of sugar maple trees in a number of regions including the Adirondacks (Burns and Honkala, 1990;Van Breemen et al., 1997).Recent studies have suggested the importance of black cherry especially in sites with low base status (Aguilar and Arnold, 1985;Godefroid et al., 2005;Lorenz et al., 2004).The researchers predicted that acid-tolerant tree species, such as black cherry, would exhibit an inverse relationship with soil and stream base cation concentrations and stream acid neutralizing capacity while acid-intolerant tree species, such as sugar maple, would exhibit a positive relationship with soil and stream base cation concentrations and stream acid neutralizing capacity.

Site description
Grass Pond watershed is located within the Ha-De-Ron-Dah Wilderness Area at 43 • 41 25 N, 75 • 3 54 W in the southwestern Adirondack Mountains of New York State.The watershed is divided into seven adjacent subwatersheds that exhibit substantial differences in stream water acid-base chemistry (Ito et al., 2007) (Fig. 1).The entire Grass Pond watershed has an area of 237 ha with elevation ranges from 552 to 684 m.The parent material has variable thickness and is derived from glacial till.Soils are predominantly Spodosols.Vegetation is typical of a northern hardwood mixed forest containing red maple, yellow birch (Betula alleghaniensis), American beech, and sugar maple, with some black cherry, red spruce and eastern hemlock.A more detailed description of the seven Grass Pond subwatersheds is available in Ito et al. (2007).

Site history
In the 1800s to the early 1900s, large portions of the southwestern Adirondacks were cleared for timber harvest and for agriculture (Schneider, 1998).It is likely that during this time landowners logged Grass Pond watershed.In 1903, a 25 000 acre fire burned the Ha-De-Ron-Dah area (Gray, 1995).From 1909 to 1981, New York State acquired parcels of land including the Grass Pond watershed and in 1986 the State designated this area as the 26 600 acre Ha-De-Ron-Dah Wilderness Area (Gray, 1995).

Stream water
In October 2008, a period for which discharge was reflecting base flow conditions, 37 stream water samples were collected along an elevational gradient along the seven streams located within the Grass Pond watershed.Stream samples were collected starting at each stream inlet to Grass Pond and taken at ∼ 150 m longitudinal increments going upstream.Streams 2 and 5 had two main tributaries that joined and drained into the inlet to Grass Pond.For streams 2 and 5, samples were collected at the inlet downstream from where the two tributaries joined as well as upstream both tributaries.Stream water samples were collected and stored in 250 mL polyethylene bottles at 4 • C until chemical analysis.Stream samples were filtered through 0.45 µm HV (high volume) membrane filters to remove particulate matter prior to analysis.

Vegetation
Tree vegetation surveys were conducted during August, September and October 2008.Each subwatershed was sampled using 9 m radius plots selected from a grid of transects, accounting for ∼ 1 % of the area of each subwatershed with a total of 80 plots.All tree species ≥ 5 cm in diameter at 1.4 m above the ground (dbh, diameter at breast height) within each selected sample plot were identified and dbh measured to determine total and relative (%) basal area for each species within plots and subwatersheds.

Soil
Soil samples were collected at Grass Pond watershed in October 2008 within the selected vegetation plots.In each plot, the forest floor (Oe/Oa horizon) and upper mineral soil (0-10 cm) were sampled from three locations, homogenized in a bucket and a sub-sample was sealed in a polyethylene bag.The forest floor was sampled by removing litter (Oi horizon) and by cutting a small square, approximately 20 cm × 20 cm, into the Oe/Oa horizon with a knife.The upper mineral soil was sampled using a "bulb planter corer" under the location of the forest floor sample.All collected samples were kept refrigerated at 4 • C until further sample preparation and chemical analysis.A total of 59 mineral soil and 63 forest floor samples were collected.At locations where insufficient soil development occurred, samples were not collected and at locations with forest floor soil layers deeper than approximately a meter, mineral soil samples were not collected.

Stream water
Stream water samples were analyzed for pH potentiometrically, using a MI-410 combination pH electrode (Microelectrodes, Inc.) in conjunction with an Accumet ® , AR50, dual channel pH/ion/conductivity meter.Stream water samples were analyzed for dissolved organic carbon (DOC) by persulfate oxidation using a Tekmar Phoenix 8000 carbon analyzer.Ammonium (NH 4 ) and total N were determined using the Autoanalyzer3.Stream water samples were analyzed for cation (Al, Ca, K, Mg, Na and Si) concentrations using a Perkin-Elmer Optima DIV 3300 ® , inductively coupled plasma atomic emission spectrophotometer (ICP-AES) and for anion (Cl − , NO − 3 and SO 2− 4 ) concentrations using a Dionex ICS-2000 ® ion chromatograph (IC).
Stream water acid neutralizing capacity (ANC c ) was calculated using the equation in which concentrations are expressed in µeq L −1 .The base-cation surplus (BCS) was calculated using the equation in which concentrations are expressed in µeq L −1 (Lawrence et al., 2007).In the equation for BCS, [RCOO − s ] equals the concentration of strong organic anions.Estimates of the contribution of organic anions to a solution charge were calculated using the equation [RCOO − s ] = 0.071[DOC] − 2.1 (Lawrence et al., 2007).Where [RCOO − s ] is the (µeq L −1 ) organic anion concentration, and [DOC] is the (µeq L −1 ) dissolved organic carbon concentration.

Soil
Fresh forest floor and mineral soil samples were homogenized and sieved to 6.4 mm to remove coarse fragments.Percent organic matter was determined as loss-on-ignition at 470 • C for 16 h (Wilde et al., 1972).Soil pH was determined potentiometrically in 2 : 1 slurry of 0.01 molar CaCl 2 : fresh soil sample (Sparks et al., 1996).
Total extractable N and NH + 4 were determined by combining soil samples (approximately 10 g for forest floor and 20 g for mineral soil) with 70 mL of 2 molar KCl.The solution was then shaken for 1 h, filtered through Whatman 42 ® ashless filter paper (pore size 2.5 µm), rinsed three times and then raised to 100 mL volume with 2 molar KCl (modified from Blume et al., 1990;Page and Mitchell, 2008).A 20 mL sub-sample of the filtrate was decanted and analyzed using continuous flow colorimetry on a Bran-Luebbe AutoAnalyzer3 ® for NH + 4 and total N. Forest floor soil samples were ground in a Wiley ® mill using a #20 (0.85 mm) screen and were analyzed for percent C and N on a Thermo Electron Flash EA 1112 ® elemental analyzer to determine the C : N mass ratio.Forest floor soil samples were analyzed for total elemental Al, Ca, K, Mg, Na, and P by ashing a 1 g homogenized sample at 470 • C for 16 h.The ash was then dissolved in 10 mL of 6 molar HCl, evaporated to dryness on a hot plate, re-dissolved in 10 mL of 6 molar HCl, filtered through a Whatman 42 ® ashless filter paper, rinsed three times with deionized, distilled water (DDW) and raised to 100 mL with DDW (Page and Mitchell, 2008;modified from Sparks et al., 1996).Total elemental concentrations (Al, Ca, K, Mg, Na, and P) were then determined using ICP-AES.
Mineral soils were analyzed for exchangeable elements (Al, Ca, K, Mg, Na, and Si) by extracting approximately 10 g dry soil in 50 mL of 1 M NH 4 Cl.The solution was then filtered through a Whatman 42 ® ashless filter paper, rinsed three times with 10 mL of 1 molar NH 4 Cl, and then raised to a volume of 100 mL with 1 molar NH 4 Cl (Blume et al., 1990).Samples were then frozen until analyzed for exchangeable cation concentrations using ICP-AES.

Statistical analyses
Pearson correlations were used to determine the significance of relationships between soil chemical measurements and tree species basal areas, as well as between stream chemistry and tree species basal areas.Analyses were based on sample sizes of 60 and 55 for the forest floor and mineral soil, respectively, after removal of 7 outliers (< 6 % of all samples) to help facilitate multivariate analysis assumptions.We also did the statistical determinations with the outliers and found there was no substantive effect on the overall results.The plot data and stream chemistry were averaged by the 7 subwatersheds for analysis (sample size of 7).Variables used in the analysis include the chemical constituents of the forest floor, mineral soil, and stream (pH, Ca 2+ , etc.) and the total basal areas of specific tree species.For interpreting "p values" we followed the approach recommended by Sterne and Smith (2001).
Canonical correspondence analysis (CCA) is a wellestablished multivariate technique to analyze speciesenvironment relationships (Ter Braak, 1986;Khattree and Naik, 2000).It is a two-step gradient analysis, in which ordination axes are extracted from species abundance data, followed by the identifying gradient from environment variables.Thus, the dominant pattern of variation in species community can be directly related to variation in environmental variables.The results of CCA are commonly displayed by a biplot, which is a graphical presentation of the data matrix by two sets of plots overlaid on the same coordinate system, one plot representing species and the other plot representing environment variables (Hodge and Naik, 1999).In addition, the percentage variations accounted for by the two ordination axes were computed as the proportion of the first two eigen values to the sum of all eigen values of species-environment correlations (Ter Braak, 1986).
In our study, CCA was used to relate tree species basal areas to forest floor, mineral soil, and stream chemistry, respectively.In the resultant biplots, species are shown as points, while chemical variables are represented by arrows from the origin of the biplot.Important chemical variables tend to be represented by longer arrows than less important chemical variables, and the relative locations of species points to the direction of arrows (chemical variables) indicate the association and correlation of each species with respect to each chemical variable (Ter Braak, 1986;Khattree and Naik, 2000).

GIS mapping
The density analysis function in the Spatial Analyst tool of ArcMap version 8.3 (ESRI, Inc.) was used to create maps that geographically distribute tree species basal area data collected from each plot and expand it throughout the entire watershed.This allowed for a visual representation of tree species and stream chemistry data.

Watershed characteristics
The major tree species within Grass Pond watershed included American beech, black cherry, eastern hemlock, red maple, red spruce, striped maple (Acer pensylvanicum), sugar maple and yellow birch (Fig. 2).Red maple and American beech were the most abundant species within the watershed with 28 and 34 % relative basal area, respectively.Black cherry, with 4 % relative basal area, occurred in nine plots, primarily in the upper elevations of subwatersheds 4 and 5. Sugar maple, with 9 % relative basal area occurred primarily in subwatershed 3.
Within Grass Pond watershed, streams 4, 5, 6 and 7 were acidic with ANC < 0 µeq L −1 (Table 1 and Fig. 3a).All stream sample sites had moderate to low nitrate levels (< 40 µeq L −1 ) (   60 µeq L −1 (Aber et al., 2003) and hence the nitrate values for the streams of the Grass Pond watershed were within the low to moderate range.Only two stream sample sites exhibited high DOC levels (> 600 µmol C L −1 ) (Driscoll and Newton, 1990), which were located in the headwater sample sites of stream 5 (Table 1).Forest floor and mineral soils within Grass Pond watershed were generally acidic with mean pH values less than 4.0 (Fig. 3b).Forest floor mean C : N mass ratios were below 20 indicating elevated nitrogen levels (Driscoll et al., 2003).It has generally been established that monomeric Al concentrations in soil and surface waters > 10 µmol L −1 are toxic (Driscoll et al., 2001).The concentrations of exchangeable Ca found in the Grass Pond watershed soils were within the range (< 0.1 mg Ca g −1 soil) that has been attributed to surface water acidification throughout the Adirondacks (Table 2).These relatively low Ca soil concentrations were also coupled with high Al soil concentrations in the Grass Pond watershed (Table 2) (Sullivan et al., 2006).

Canonical correspondence analysis
CCA is based on the correlations between tree species basal areas to forest floor, mineral soil and stream chemistry, in which these correlations are transformed to two ordination axes, similar to the first two principal components, to display the relationships between the species and chemistry in a biplot.However, the re-computed relationships are on the relative basis.Thus, the relative locations and directions are important and interesting.For example, in Fig. 5c, sugar maple is on the same direction as stream water ANC c , BCS and pH and American beech is on the same direction as stream water NO − 3 , which indicated positive association or relationships (as also shown in Table 4).But the length of the arrows for the stream chemistry was relative to their correlations with the species.For example, the arrow length of DOC was longest because it had the largest (negative) correlation with sugar maple (thus opposite direction) in Table 4. On the other hand, the biplot was overlaid from the two ordination systems (one for species and one for chemistry).The scaling of the two axes system may be different.Therefore, the relative directions of species vs. chemistry are more important than the relative locations of species vs. chemistry (e.g., how far away from each other).
The most important environmental variables associated with tree species basal area in the forest floor CCA biplot, as indicated by the longer arrows, were forest floor Ca and NH 4 (Fig. 5a).The most important environmental variables associated with tree species basal area in the mineral soil CCA biplot were mineral soil Ca, NH 4 and pH (Fig. 5b).Forest floor Al and mineral soil Al were the least important environmental variables as indicated by the shortest arrows in the biplots (Fig. 5a, b).Stream water DOC was the most important environmental variable associated with tree species basal area in the stream water CCA biplot (Fig. 5c).
The relative location of black cherry in the CCA biplot indicated a strong positive association with forest floor NH 4 (Table 3 and Fig. 5a).Both the forest floor and mineral soil CCA biplots indicated a strong positive association between yellow birch and Ca levels, a strong positive association between sugar maple and pH and a strong negative association between black cherry and pH (Fig. 5a, b).The relative location of sugar maple in the CCA biplot indicated a strong positive association between sugar maple and stream water ANC c , BCS and pH, and a strong positive association between American beech and stream water NO − 3 (Fig. 5c).

Discussion and conclusion
Northern hardwood forests are comprised of a mix of acid-intolerant and acid-tolerant species, primarily American beech, basswood, black cherry, eastern hemlock, red maple, red spruce, sugar maple, white ash, white pine and yellow birch (Braun, 1950).In northern hardwood stands in Pennsylvania and New York, sugar maple growth was positively related to foliar concentrations of Ca and Mg while co-occurring black cherry showed opposite growth responses with greater growth in stands with below-threshold Ca and Mg compared with above-threshold stands (Long et al., 2009).Acidic deposition has accelerated the loss of exchangeable Ca and the mobilization of monomeric Al in forest stands resulting in the decline of sugar maple stands in the northeastern United States (Lawrence et al., 1999).Sugar maple decline has most predominantly occurred on ridge tops and on upper slopes, where soil base availability is much lower than at mid and low slopes on the landscapes (Bailey et al., 2004).In forest stands with substantial tree mortality between 1989 and 1994, within the Alleghany National Forest, Pennsylvania, the dominant tree species has shifted from sugar maple to black cherry and red maple (McWilliams et al., 1996).Such a shift in three species composition would be consistent with a shift to more acid-tolerant tree species.
In Grass Pond watershed, forest composition is dominated by red maple and American beech with black cherry and sugar maple only comprising 4 and 9 % relative basal area, respectively.At Grass Pond watershed, sugar maple basal area was associated with watershed attributes associated with the neutralization of atmospheric acidic inputs, including high stream water ANC c , BCS and pH, low stream water DOC and high soil pH.These locations were in contrast to those in which black cherry was found that showed more acidic attributes (Fig. 6).Other research has suggested that sugar maple tree characteristics, such as high Ca concentrations and low C : N ratio leaf litter, will result in high mineralization and nitrification rates within the soil, contributing to increased nutrient cycling and nitrate leaching (Christ et al., 2002;Lovett and Rueth, 1999;Page and Mitchell, 2008).The associations between sugar maple basal area and high stream water high ANC and low stream water DOC indicates deep sub-surface hydrological flow paths through thick mineral soil layers and parent material (Fig. 6) (Chen et al., 1984;Mitchell et al., 2013;Driscoll and Peters, 1987).In general, sites with thick soil layers with relatively high soil pH and Ca concentrations are more likely to be colonized by sugar maple trees than other less suitable more acidic sites because of the relatively high nutrient demands of this tree species (Bailey et al., 2004;Burns and Honkala, 1990;Christopher et al., 2006;Van Breemen et al., 1997).
As a shade intolerant, early successional species, black cherry is associated with shallow soils on steep slopes where blowdowns and other disturbances create gaps that allow adequate sunlight for black cherry regeneration and growth (Aguilar and Arnold, 1985).At Grass Pond watershed, black cherry basal area was associated with those watershed attributes associated with the inability to neutralize atmospheric acidic inputs, including low stream water ANC c , BCS and pH, high stream water DOC and total Al, low soil Ca concentrations, and high C : N ratios (Fig. 6).The negative association between black cherry basal area and stream water ANC c and BCS and the positive association with stream water DOC is likely due to black cherry stands occurring on relatively steep slopes with thin soils and due to the predominance of shallow flow paths through forest floor  The limited significant statistical relationships between tree species basal area, stream and soil chemistry was most likely due to the limited presence of sugar maple and black cherry within Grass Pond watershed, generally low soil Ca concentrations, low variations in soil Ca concentrations among plots as related to species composition and the lack of mineral and forest floor soil development in the high elevation and steep sloped plots resulting in a decreased soil sample size and a small watershed stream chemistry sample size.These soil conditions combined with influences of the vegetation components are the major drivers affecting the spatial variation in the surface water chemistry of the Grass Pond watershed.Although correlations have been identified between tree species basal area, soil chemistry and stream chemistry, it is unclear whether the presence of sugar maple (conversely, black cherry) in base-rich soils was due to sugar maple contributing base-rich nutrients to the forest floor or if it was due to base-rich soils promoting sugar maple growth.
Documented disturbance history indicates that trees within Grass Pond watershed were likely logged and burned over 100 years ago, potentially influencing soil, vegetation and stream characteristics of the watershed.Other studies have found that in the short term following a disturbance, northern hardwood watersheds in New Hampshire exhibited an increase in stream water NO − 3 , Al and base cations, and a decrease in stream water SO 4 and pH (Lawrence et al., 1987).Additionally, other studies have found that forests in the Adirondack Mountains of New York, with varying disturbance histories have similar soil C : N ratios, net N mineralization rates and net nitrification rates (Latty et al., 2004).
Previous research has shown that sugar maple is acid intolerant or a calciphilic tree species, and our results provide additional information that supports that black cherry is an acid-tolerant tree species and hence may be associated with those areas with low surface water ANC and pH values.Similarly, sugar maple presence is a good indicator of basic soil and stream conditions compared to stands containing black cherry.These soil and site characteristics are not only important in determining which tree species colonize an area but also ultimately affect how these tree species will concomitantly influence soil and forest patterns including variations in litter nutrient concentrations and mineralization rates and hence surface water characteristics (Christ et al., 2002;Lovett and Rueth, 1999;Page and Mitchell, 2008).
Our results show that black cherry has a distinctive positive relationship with forest floor NH 4 , while sugar maple has a distinctive inverse relationship with stream water DOC concentrations.Understanding the interactions among tree species and watershed acid-base status provides insight into the sensitivity or resilience of the watershed to inputs of acidic deposition.The continual depletion of base cations from northern hardwood forest soil, primarily due to the effects of acidic deposition, will reduce the area favorable for growing sugar maple and increase area available for black cherry and other acid-tolerant tree species.Shifts in the tree species from those that require relatively high soil base concentrations such as sugar maple to those acid-tolerant species such as black cherry could have major impacts on these forest ecosystems and their associated surface waters.Future shifts in tree species composition and soil nutrient dynamics will presumably be accompanied by concomitant shifts in stream chemistry throughout the watershed.These evaluations of the spatial and temporal patterns of soil and surface water chemistry in northern hardwood forests provide new insights into the continued and long-term effects of acidic deposition on forested watersheds.
Figure 1.The Adirondack Park located within New York State.The star indicates the location of Grass Pond watershed in the southwestern region of the Adirondacks.

Fig. 1 .
Fig. 1.The Adirondack Park located within New York State.The star indicates the location of Grass Pond watershed in the southwestern region of the Adirondacks.

Figure 2 .
Figure 2. Grass Pond watershed mean tree species total basal area (m 2 /ha) calculated from plot data collected August-October 2008.Bars indicate one standard error.

Fig. 2 .
Fig. 2. Grass Pond watershed mean tree species total basal area (m 2 ha −1 ) calculated from plot data collected August-October 2008.Bars indicate one standard error.

Figure 3 Figure 3
Figure 3. a) Base-cation surplus means (µeq L -1 ) for each Grass Pond stream with ba indicating one standard deviation.Values ≤ 0 µeq L -1 indicate a low capacity to neut inputs.b) Forest floor and mineral soil sample pH (units) for each Grass Pond subwa

Figure 4
Figure 4. a) Stream base-cation surplus (BCS) and sugar maple basal area upstream from the sample location.b) Stream base-cation surplus (BCS) and black cherry basal area upstream from the stream sample location.

Table 1 )
. Regional surveys of surface waters in forested watersheds in the northeast United States have suggested that nitrate concentrations range from 0 to

Table 1 .
Grass Pond stream chemistry means (standard deviation) analyzed from stream samples collected in October 2008.

Table 2 .
Forest floor (total) and mineral soil (exchangeable) chemistry means (standard deviation) for sample plots within Grass Pond subwatersheds.

Table 3 .
Factor analysis.Forest floor data and tree species basal area Pearson correlation coefficients (n = 60) and mineral soil data and tree species basal area Pearson correlation coefficients (n = 55) with p values.

Table 4 .
Factor analysis.Watershed average stream chemistry data and tree species basal area Pearson correlation coefficients (n = 7) with p values.