| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a SUNY College of Environmental Science and Forestry, Syracuse, NY 13210
b Woods Hole Research Center, Woods Hole, MA 02543
c Yale Univ., New Haven, CT 06511
* Corresponding author (rdyanai{at}mailbox.syr.edu).
Received for publication July 9, 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: BEF, Bartlett Experimental Forest • HBEF, Hubbard Brook Experimental Forest • WMNF, White Mountain National Forest
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The affinity of soil organic matter for Pb (Schnitzer and Skinner, 1967; Strawn and Sparks, 2000) has protected aquatic systems and public water supplies from much of the Pb deposited in precipitation. In forested ecosystems, Pb is strongly retained by the forest floor. High concentrations of Pb in the forest floor were reported at the height of Pb deposition, even in remote areas (Parker et al., 1978; Heinrichs and Mayer, 1980; Smith and Siccama, 1981). Samples collected in 1980 from nine northeastern states ranged from 0.7 to 1.8 g Pb m–2, depending on the forest type (Johnson et al., 1982). Sites at higher elevations tended to have more Pb in the forest floor, presumably because precipitation increases with elevation (Miller and Friedland, 1994), and older forest floors had more Pb, reflecting the duration of the exposure to deposition (Johnson et al., 1982).
Forest floors were observed to accumulate Pb over time from the 1960s to the 1970s. White pine (Pinus strobus L.) stands in central Massachusetts had nearly doubled the Pb content of the forest floor between 1962 and 1978 (Siccama and Smith, 1980). In northern hardwood and boreal forest at Camels Hump, VT, Pb concentrations about doubled between 1966 and 1977 (Friedland et al., 1984). In both studies, the majority of the Pb deposited from the atmosphere over the measured interval was retained in the forest floor (Siccama and Smith, 1980; Friedland et al., 1984).
The regulation of Pb in gasoline, beginning in 1975, reduced the atmospheric deposition of this element. For example, at the Hubbard Brook Experimental Forest (HBEF) in New Hampshire, precipitation in 1989 contained only 3% as much Pb as it did in 1977 (Johnson et al., 1995). The forest floor was expected to continue to accumulate Pb at a low rate, because Pb residence times were estimated to be hundreds of years (Friedland and Johnson, 1985; Turner et al., 1985). Surprisingly, the forest floor at this site had 29% less Pb in 1987 as in 1977 (Johnson et al., 1995). Similarly, in a study of 30 sites in the northeastern USA, 12% of Pb in the forest floor was lost between 1980 and 1990 (Friedland et al., 1992). The fact that Pb is disappearing from the forest floor could be cause for concern, although soil water and stream export of Pb has been observed to be low in a variety of systems (Van Hook et al., 1977; Heinrichs and Mayer, 1980; Turner et al., 1985; Wang et al., 1995).
We hypothesized that Pb would be found deeper in the forest floor over time, because the majority of the litter input, now nearly free of Pb, enters from above. Previous studies of changes in Pb have not distinguished horizons within the forest floor. The effect of stand age and elevation on the retention of Pb in the forest floor is also unknown, although the amount of organic matter and Pb stored in the forest floor is known to be greater in older stands and in those at higher elevation (Johnson et al., 1982).
We studied changes in Pb in the forest floor in two independent sets of samples initially collected for other purposes. At the HBEF, we collected forest floor samples at seven different dates from 1976 to 1997 in a single watershed. This data set provides the best remeasurement sequence of forest floors in the region. Two other watersheds at HBEF provide ancillary data. In the second study, we measured changes in the forest floor at two points in time in 13 additional stands of different ages, all in the White Mountains of New Hampshire. We distinguished horizons within the forest floor at the HBEF and at six of the additional stands, which allowed us to test whether Pb was moving down through the forest floor. We also tested the importance of elevation and stand age in the loss or retention of Pb by the forest floor. This information should be useful to predicting the fate and transport of Pb in forested ecosystems.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Site Descriptions
Hubbard Brook Experimental Forest
The HBEF is located in the central White Mountains of New Hampshire, USA (43°56' N, 71°45' W). Samples for this study were collected from Watershed 6, a 13.2-ha natural area which serves as the reference watershed for the HBEF and has not been experimentally manipulated. Species composition varies along an elevational gradient, dominated by American beech (Fagus grandifolia Ehrh.), yellow birch (Betula alleghaniensis Britton), and sugar maple (Acer saccharum Marsh.) at the lower elevations (550–630 m); sugar maple and beech at mid-elevations (630–700 m); and balsam fir [Abies balsamea (L.) Mill.], red spruce (Picea rubens Sarg.), and white birch [Betula papyrifera var. cordifolia (Marsh.) Regal] at higher elevations (>700 m). Soils are acidic, well-drained Spodosols (mostly Typic Haplorthods and Fragiorthods) derived from glacial till and of varying depth. Further descriptions of the HBEF are available in various publications (e.g., Whittaker et al., 1974; Bormann and Likens, 1979; Johnson et al., 1991; Likens and Bormann, 1995; Likens et al., 1998).
Three watersheds at the HBEF—W6, W5, and W1—have been sampled for forest floor properties at different times and somewhat different intensities (Table 1). The reference watershed, W6, was sampled on seven occasions between 1976 and 1997. At three of the dates (1976, 1977, and 1982) the forest floor was collected as a whole, while at the others (1978, 1987, 1992, and 1997) Oie and Oa-A horizons were distinguished. Individual plots consisted of 15 by 15 cm square blocks of excavated forest floor material and each sample date included between 58 and 87 plots (Table 1), randomly located within elevational strata (16–27 at low elevation, 21–26 at mid-elevation, and 21–34 at high elevation). An adjacent watershed, W5, was sampled using the same methods in 1982 before it was whole-tree harvested. We did not include in our analysis another data set from W5 (Huntington et al., 1988) from the same date, because it used different sampling methods. Finally, W1 was sampled in 1996, 1998, and 2000, in association with an experimental addition of calcium silicate, which we assumed would not affect the behavior of Pb in the forest floor.
|
The BEF study involves six even-aged stands spanning a range of ages representing a number of developmental stages. The stands were sampled in 1979 (Federer, 1984) and again 15 yr later in 1994 (Yanai et al., 1999, 2000). All sampled stands originated following clearcutting between about 1875 and 1985. This remeasured chronosequence provides observations from stands of different ages (two stands were the same age), from 11 to 119 yr following treatment. Forest floor samples at the BEF were separated into Oi, Oe, Oa, and A horizons. Plots (n = 10) were established at regular intervals along six systematically located transect lines in 1979 and along five of the previously established lines in 1994. Each plot consisted of a 10 by 10 cm square block of excavated forest floor material. Plots within a line were composited by horizon, providing six and five samples per stand in 1979 and 1994, respectively. We also report on combined Oie and Oa-A layers for the BEF because these were collected at the HBEF.
White Mountain National Forest
Seven additional stands were sampled in the White Mountain National Forest (WMNF), ranging in latitude from 43°56' to 44°13' N, in longitude from 71°14' to 71°44' W, and in elevation from 455 to 550 m. Soils were coarse-loamy, mixed, frigid Typic Haplorthods. Sampling methods were similar to those used at the BEF: five transects of 10 plots each were sampled in 1980 (Federer, 1984) and 1995 (Yanai et al., 1999, 2000). The forest floor blocks were not subdivided into horizons in 1980, so only the forest floor as a whole (O-A) can be assessed for change over time at these sites.
Laboratory Analyses
Both sets of samples from BEF were analyzed at the same time, as were both sets of samples from the WMNF, to control for variation in laboratory procedures. Samples from the HBEF were analyzed over a longer period of time. Forest floor samples from all sites were analyzed using similar procedures. Samples were dried to a constant weight, screened to remove coarse fragments in the case of BEF and WMNF samples (Yanai et al., 1999), then ground with a Wiley mill. Loss on ignition was used to estimate organic mass. Ashed subsamples from each horizon were dissolved in 6 mol L–1 nitric acid and analyzed by inductively coupled plasma spectroscopy (ICP). We report Pb concentrations on a dry weight basis. Because chemical analyses were performed on individual horizons, mass-weighted concentrations were calculated for all combined layers.
Statistical Analyses
The HBEF data set was analyzed separately from the BEF and WMNF data sets, because the amounts of Pb and the changes over time differed between the two studies. Mean values at the stand level were used to test hypotheses about changes in forest floor properties over time. Dependent variables did not require transformation to meet tests for normality of the residuals.
Relationships between forest floor properties and stand age, elevation, or time of measurement were tested using linear regression. Paired t tests were used to test for change over time in stands at the BEF and WMNF, which were sampled only twice. Differences were determined by comparing stand level means at the two sampling dates. Minimum detectable differences were determined based on inverse t tests with 
= 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Like Pb contents, concentrations of Pb at HBEF declined by 29 mg kg–1 (ppm), or 32%, in the forest floor as a whole between 1976 and 2000 (p < 0.001, linear regression) (Fig. 1b). The change in Pb concentrations at BEF and WMNF was not statistically distinguishable from zero (p = 0.10, paired t test) and was less than the rate of loss at HBEF (a mean change of 7 mg kg–1 would have been statistically detectable).
The amount and concentration of Pb in the forest floor was higher at the HBEF than at the other sites by about 40% for content and 22% for concentration (Fig. 1). The mass of the forest floor was similar at the HBEF, BEF, and WMNF sites (Fig. 1c). Older stands tended to have more massive forest floors than younger stands (r2 = 0.28; p < 0.001 for linear regression on all observations; r2 = 0.30; p = 0.02, using average values for each stand). There was no significant effect of stand age on Pb contents or concentrations or changes in these variables (Fig. 1a and b).
Changes within Horizons
For eight of the sites, the forest floor was divided into Oie and Oa-A horizons. The underlying Oa-A is more massive (Fig. 2c) and has greater Pb contents (Fig. 2a) than the Oie. Declines in Pb were more evident in the upper portions of the forest floor (Oie) than in the Oa-A (Fig. 2a and b).
|
Like Pb contents, concentrations of Pb in the Oie horizon declined dramatically (Fig. 2b), by almost 65% at the HBEF (p < 0.0001) and by 45% at the BEF (p = 0.002). In contrast, in the Oa-A horizon, Pb concentrations were stable within detection limits at HBEF and increased by 29% at BEF (p = 0.03). Concentrations of Pb were greater in the Oie than Oa horizon at the beginning of the measurement period, but were lower in the Oie than the Oa at both locations at the most recent observations.
The forest floor was divided into four horizons (Oi, Oe, Oa, A) at the BEF. Changes within these horizons provide finer resolution of Pb dynamics, and confirm that Pb was found deeper in the profile over the period of observation (Fig. 3) . The uppermost Oi horizon declined in Pb concentration (Fig. 3b) by 58%, on average, during the 15-yr period (p < 0.0001); the Oe horizon experienced an average 44% loss (p = 0.004). In contrast, Pb concentrations increased by 18% in Oa and 37% in A horizons (p = 0.05 and 0.01, respectively). Changes in Pb content (Fig. 3a) were not as consistent, with the loss being statistically significant only in the Oe horizon (p = 0.01). Changes in organic matter content within horizons (Fig. 3c) were also mostly insignificant (the A was significantly less massive in 1994 than 1979; p = 0.04).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Much of the downward movement of Pb that we observed can be attributed to litter additions at the top of the forest floor. Some may be due to migration of Pb in the soil solution or in particulate matter (Wang and Benoit, 1996). At Camel's Hump in Vermont, Pb has migrated from the forest floor into the upper 10 cm of the mineral soil (Kaste et al., 2003). At the HBEF, the Pb lost from the upper forest floor was not retained within the forest floor: the lower Oa-A remained about constant, and the Pb contents of the forest floor as a whole declined significantly from 1976 to 1997. In contrast, at the BEF sites, declines in concentrations of Pb in the Oi and Oe were compensated by significant increases in the Oa and A. Although the observed increases in Pb contents were not statistically significant in the Oa-A, they were sufficient to prevent detectable losses of Pb from the forest floor as a whole. Similarly small changes in Pb contents took place in forest floors of the additional WMNF sites. One difference between the HBEF and the other studies is that forest floor sampling commenced earlier, in 1976, instead of in 1979 or 1980. The most rapid losses of Pb at the HBEF occurred in the earliest portion of the record (Fig. 1). We can imagine that losses from the forest floor might have been greater at the BEF and WMNF sites if they had been sampled somewhat earlier.
Another difference between the HBEF and the BEF-WMNF sites, besides the timing of sampling, is their elevation. The greater Pb content of the forest floor at Hubbard Brook is consistent with previous observations of forest floor Pb as a function of elevation in the Taconic, Green, and White Mountains (Johnson et al., 1982) (Fig. 5) . The pattern of greater Pb in the forest floor at higher elevations is explained by greater wet and dry deposition velocities (Miller and Friedland, 1994). Within the HBEF sites, Pb concentrations increased significantly with elevation. The most surprising result is perhaps the absence of a consistent pattern in Pb content with elevation at HBEF, in spite of consistent increases in concentration. Similarly, in testing for patterns with time, concentrations of Pb changed more systematically than Pb contents. This raises questions about the consistency of forest floor sampling over time.
|
When monitoring change over time in the forest floor, the choice of concentration or contents as the indicator variable depends on the depth distribution of the element in question. When concentrations vary dramatically with depth, it is possible to introduce spurious but statistically significant changes over time in elemental concentrations merely by sampling to different depths (Yanai et al., 1999). In the case of Pb, since concentrations are similar in the upper and lower portions of the forest floor (Fig. 2b), variation in the depth of samples collected has a greater effect on Pb contents, and Pb concentration values are less sensitive to variation in sampling methods. Indeed, Pb concentrations were more consistent than Pb contents in our data sets (Fig. 1– 4, comparing a and b). For elements that decline with depth in the forest floor, such as C, N, P and exchangeable bases, contents will be less sensitive to sampling error than concentrations. Since it is far easier to collect "grab" samples for chemical analysis than "quantitative" samples for calculation of contents, concentration may be the better parameter for future monitoring of the fate of Pb in the forest floor.
In the future, we can anticipate that the upper portions of the forest floor will continue to decline in Pb content and concentration, as litter low in Pb is added to the surface. In the sites we studied, peak concentrations of Pb are now in the Oa horizon, in contrast to two decades ago, when Pb concentrations were highest at the surface. The fate of Pb in the Oa horizon is linked to humus decomposition and complexation; the rate of loss of Pb from the forest floor could provide insight into the dynamics of C cycling as well as that of heavy metals.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. M. Kaste, B. C. Bostick, A. J. Friedland, A. W. Schroth, and T. G. Siccama Fate and Speciation of Gasoline-Derived Lead in Organic Horizons of the Northeastern USA Soil Sci. Soc. Am. J., August 22, 2006; 70(5): 1688 - 1698. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Soil Science Society of America Journal | Journal of Plant Registrations | The Plant Genome | |||
