Environ. Sci. Technol. 2006, 40, 6580-6586 Accumulation of Persistent Organic Pollutants in Canopies of Different Forest Types: Role of Species Composition and Altitudinal-Temperature Gradient L U C A N I Z Z E T T O , †,§ K E V I N C . J O N E S , ‡ PAOLA GRAMATICA,† ESTER PAPA,† BRUNO CERABOLINI,† AND A N T O N I O D I G U A R D O * ,§ Department of Structural and Functional Biology, University of Insubria, Via J. H. Dunant 3, 21100 Varese VA, Italy, Centre for Chemicals Management and Environmental Science Department, Lancaster Environment Centre, Lancaster University, Lancaster, LA1 4YQ, UK, and Department of Chemical and Environmental Sciences, University of Insubria, Via Valleggio, 11, 22100 Como CO, Italy Leaves from the dominant tree species in three different alpine forests were sampled along an altitudinal gradient and analyzed for HCB, R- and γ-HCH, and PCBs. The mean canopy concentration was calculated, considering the relative abundance of each species in the respective forest. Compound fractionation occurred in the vegetation along the altitudinal/temperature gradient. Results were compared with air concentrations and in-field plant/air partition coefficients (KPA) were calculated for each species; this showed differences between broadleaves and needles. The mean canopy/air partition coefficient (KCA) was also calculated by averaging results from single species. The variability of persistent organic pollutants distribution in canopies is discussed considering two main factors, the altitudinal/temperature gradient and the species composition. The latter is responsible for most of the concentration variability of the more volatile compounds. A model to calculate dry gaseous deposition to different forest canopies is presented. Introduction Important advances in our knowledge of the global distribution of organic pollutants have been made by studying vegetation (1). Persistent organic pollutants (POPs) have been measured in the vegetal biomass across latitudes and with altitude, reflecting the accumulation of organic pollutants from the atmosphere (2, 3) and supplying terrestrial food webs (4, 5). Exposure of forest ecosystems to semivolatile organic pollutants has also been investigated (6-10). Forest canopies can influence air concentrations temporally (11, 12) and spatially (13). Given the proportion of the earth’s surface covered by forests, their role in influencing pollutant * Corresponding author phone: +39 031 2386480; fax: +39 031 2386449; e-mail: antonio.diguardo@uninsubria.it. † Department of Structural and Functional Biology, University of Insubria. ‡ Lancaster University. § Department of Chemical and Environmental Sciences, University of Insubria. 6580 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 21, 2006 fate on a regional or global scale merits consideration (1417). POPs were recently studied in vegetation in the Canadian Rocky Mountains (18). Concentrations of some organochlorine pesticides increased along the altitudinal gradient. Alpine systems provide the opportunity to compare different forest types which change in specific composition with the increasing elevation over relatively small distances (compared to latitudinal gradients) and where air concentrations may be more uniform. Mountain environments were therefore recently pinpointed as ideal sites to comparatively investigate some of the transport phenomena (such as role of temperature or species composition) associated with the global distribution of pollutants (19). In this study, concentrations of PCBs, R- and γ-HCH, and HCB were determined in leaves and needles from three different forest sites in the Italian Alps. Previously, these sampling sites were used to compare air concentrations within and outside the forest (13, 20) and the deposition fluxes to the canopies. The main aims of this investigation were to evaluate the distribution of POPs in vegetation in three different ecosystems and assess the contribution of dominant species to the overall load of POPs in forest canopies. Experimental Section Sampling Sites. Leaves were sampled from three sites in the Lys Valley, Aosta, Italy, located at 1100, 1420, and 1790 m above sea level (asl), respectively (Figure 1). Each site is characterized by a different forest type, according to the altitude, as described in Table 1. Sampling. Leaf samples were collected from the dominant tree species. Table 1 summarizes the species sampled and their abundance at each site. Samples were collected in October 2002, at the end of the growth season. For spruce (Picea abies) new needles (year 0) were collected, to normalize exposure time to deciduous species. Leaves were collected from 21 transects using latex gloves at least 1.5 m from the soil. At each location g3 trees of the same species were sampled in an area of 400 m2 at least 100 m from the forest edge. Samples therefore reflected the most abundant foliar biomass in the inner canopy. Samples were kept in solventrinsed glass jars and frozen at -20 °C, until extraction. Analysis. Samples were freeze-dried, homogenized, and spiked with PCB 40 and PCB 128 as recovery standards. Extraction (typically of 5 g) was performed in precleaned cellulose thimbles (Schleicher & Schuell, Dassel, Germany) using an all glass Soxtech-type automatic extractor (Velp Scientifica, Usmate, Italy) with n-hexane/acetone 6:1 for 6 h. Solvents were pesticide residue grade from Sigma-Aldrich, Seelze, Germany. Gel permeation cleanup was then performed in accordance with method USEPA 3640A (1998) (21), using a 600 mm length, 26 mm i.d. glass column filled with SX-3 beads (200-400 mesh, Bio-Rad, Hercules, CA) and coupled to an Agilent 1100 series HPLC system (flow rate 3.5 mL/min, run time 72 min). Fractions were collected using retention times of chemical markers according to the USEPA procedure (21). Calibration standards were injected every 6 samples. Samples were subsequently fractionated in 15 cm × 5 mm i.d. glass columns packed with activated Florisil for residue analysis (60-100 mesh) purchased from Merck, Darmstadt, Germany. Samples were eventually concentrated to 50 µL and PCB 30 and PCB 141 were added as internal standards. Analyses were performed using a Hewlett-Packard 5890 series 2 gas chromatograph (GC) equipped with dual ECD. Injection was split via a Y connector into two parallel 10.1021/es0605523 CCC: $33.50  2006 American Chemical Society Published on Web 09/28/2006 Calculation of Mean Canopy Concentration (MCC). The MCC (expressed in pg g-1 dw) was calculated as follows, n MCC ) ∑f Ci (2) i i)1 where fi is the frequency of the species i and Ci is the mean concentration of a certain chemical in that species. Due to interspecific variability in accumulation behavior (6, 23), the concentration distribution in samples of different species collected in the same forest was not expected to be normal. It was assumed that an estimation of the error associated with the MCC calculation is the following, x∑ n EMCC ) fj(ci,j - MCC)2 i,j)1 (3) xN where EMCC is the error associated with MCC calculation, ci,j is the concentration of the i observations of the j species and N is the total number of observations. FIGURE 1. Geographical location of sampling sites. 1100 m: Broadleaf forest; 1400 m: Mixed broadleaf/conifer forest; 1800 m: coniferous forest. columns: a 60 m DB-5 (J&W Scientific, Folsom, CA, i.d. 0.25 mm, film thickness 0.25 µm) and a 60 m BP-50 (SGE International, Melbourne, Australia, i.d. 0.25 mm, film thickness 0.25 µm). This facilitated peak recognition when compounds coeluted on one column. Carrier gas (He) flow rate was 1 mL/min. The GC oven temperature program was as follows: initially 90 °C hold for 1 min, 25 °C min-1 to 170 °C, 1 °C min-1 to 260 °C, then 15 °C min-1 to 300 °C, and hold for 10 min. PCB congeners included (according to homologue groups) the following: tri-PCB 28/31; tetra-PCB 52, 44, 49, 64, 74, 70; penta-PCB 104, 101, 99, 97, 87, 110, 118, 105; hexaPCB 136, 151, 149, 132/153, 138, 156; hepta-PCB 188, 187, 183, 174, 177, 180; octa-PCB 201, 203, 194. Single-compound analytical standards were purchased from AccuStandard, New Haven, CT. Purities were >98%. QA/QC. Blanks were included as 1 in every 4 samples. MDL was set as 3 times the blank values and varied from 7 to 25 pg/g dry weight (dw). No blank correction was performed. Recoveries were 81 ( 11% for PCB 40 and 95 ( 7% for PCB 128. All samples were corrected individually for recovery. Analysis of standard reference material was performed using a certified sewage sludge. Vegetation Survey and Temperature. Canopy composition was estimated in terms of species frequency (fj) in the forest in a 2500 m2 plot. The parameter fj is calculated for each species as follows, fj ) Aj (1) n ∑A j j)1 where Aj is the surface of the canopy projection to the ground of a given species (j) in a plot divided by the total canopy projection for all the species of the same plot. Specific leaf area (SLA) was measured according to ref 22. Temperature was recorded hourly at each clearing and forest site throughout the campaign using 6 Testo 174 data loggers (Testo, Lenzkirch, Germany) equipped with a NTC-temperature sensor with resolution of 0.1 °C. Results and Discussion Levels of POPs in Vegetation and General Comments on Trends. Table 2 summarizes the POPs concentrations determined in each dominant species. HCB and R-HCH levels were within the range reported for spruce and pine needles collected in the Canadian Rocky Mountains (150-400 pg g1 dw) (18) at altitudes similar to those of the present study. HCB concentrations in spruce needles were also consistent with data from Austrian high-altitude sites (300-1100 pg g-1 dw) (24). Concentrations generally reflect remote and diffuse contamination, rather than local sources, in agreement with the air concentrations measured at the same sites (13). The remote origin of the contamination is also confirmed by the R-HCH/γ-HCH ratio that averaged 1.3 with the highest values at the upper sites. These values are a factor of 6-7 times higher than recent data from air measurements in the north of Italy (25) where some of the most elevated concentrations of γ-HCH in Europe have been detected. Davidson et al. (26) measured concentrations of organochlorine pesticides along an altitudinal-temperature gradient and showed that R- and γ-HCH have opposite trends with decreasing temperature, with the overall effect that the R/γ ratio increases with altitude. Increasing R/γ ratios with altitude were also generally observed elsewhere (24). These findings suggest that the agricultural and urbanized areas in northern Italy do not directly affect the concentrations recorded in the elevated mountainous sites in the southern part of the Alps, probably because of the lower contribution of local sources with height as shown by the back trajectory analysis of air masses (13). There are not many previous studies on PCB accumulation by different tree species (30). Most of the available data relate to conifers. PCB levels in spruce needles in the present study exceed the values reported for Austrian mountains (24) by a factor of 3-5. This could reflect the fact that the western Alps tend to receive air masses coming from the highly urbanized areas of the Po Valley and the northern and central part of France (13). Unfortunately, no comparable measurements of atmospheric concentrations in these areas are available. Nevertheless, the Austrian Alps appear to be “protected” from the higher inputs from southern and western sites (24). Recent data for pine needles collected in the Central Pyrenean Mountains (27) also fall into the same range as observed here. PCB levels measured here also exceed the concentrations reported by Ockenden et al. (6) for spruce collected in remote VOL. 40, NO. 21, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 6581 TABLE 1. Sampling Sites and Dominant Speciesa log KOA vs DF regression parameters** a* b* species SLA (m2 kg-1) relative abundance LAI* (m2 m-2) 1100 m chestnut: Castanea sativa hazel tree: Corylus avellana maple: Acer pseudoplatanus 0.013 0.020 0.012 0.33 0.33 0.33 3.9 -0.069 1.192 11.6 1400 m beech: Fagus sylvatica spruce: Picea abies (pn) 0.033 0.008 0.75 0.25 4.8 -0.079 1.312 8.5 1800 m larch: Larix decidua (dn) spruce: Picea abies (pn) 0.010 0.008 0.80 0.20 1.7 -0.032 0.969 7.6 site elevation (m asl) mean site temp. (°C) a SLA (specific leaf area) represents the foliar area corresponding to the unit of dry weight; LAI (leaf area index) represents the foliar surface area corresponding to the corresponding ground unit. *From ref 13. **a and b can be used to calculate the respective forest specific depletion factor as DF ) a log KOA + b. (pn): persistent needle species; (dn) deciduous needle species. TABLE 2. Concentrations (Mean of Three Replicates ( Standard Errora) and Mean Canopy Concentrations (MCC) in pg g-1 of Dry Weight HCB R-HCH γ-HCH PCB 28/31 PCB 52 PCB 101 PCB 118 PCB 138 PCB 153 PCB 180 a chestnut maple hazel beech spruce (1400 m) spruce (1800 m) larch MCC (1100 m) MCC (1400 m) MCC (1800 m) 43 ( 6 69 ( 25 23 ( 8 117 ( 7 118 ( 9 370 ( 100 214 ( 35 97 ( 21 493 ( 72 162 ( 25 205 ( 92 129 ( 35 109 ( 42 244 ( 25 242 ( 39 492 ( 130 261 ( 8 400 ( 120 465 ( 27 151 ( 15 63 ( 2 76 ( 11 125 ( 22 187 ( 71 335 ( 114 264 ( 43 77 ( 11 311 ( 51 420 ( 21 163 ( 8 190 ( 77 156 ( 59 199 ( 77 160 ( 8 228 ( 11 388 ( 62 166 ( 7 149 ( 25 406 ( 28 156 ( 8 442 ( 109 198 ( 50 387 ( 150 336 ( 52 169 ( 34 280 ( 55 179 ( 6 184 ( 37 281 ( 21 81 ( 13 481 ( 263 266 ( 141 213 ( 118 241 ( 120 202 ( 117 287 ( 58 240 ( 34 162 ( 23 362 ( 110 93 ( 34 609 ( 53 152 ( 5 185 ( 70 284 ( 48 317 ( 37 409 ( 147 202 ( 93 213 ( 76 590 ( 257 174 ( 49 104 ( 37 91 ( 16 93 ( 21 183 ( 29 232 ( 47 375 ( 59 184 ( 30 269 ( 52 460 ( 25 159 ( 9 253 ( 40 166 ( 35 246 ( 47 204 ( 25 213 ( 11 361 ( 53 169 ( 35 158 ( 49 374 ( 53 137 ( 22 584 ( 54 175 ( 121 191 ( 118 275 ( 29 294 ( 29 385 ( 65 209 ( 54 203 ( 66 544 ( 105 158 ( 40 In the case of MCC the calculation of the error is performed as EMCC; see eq 3 for definition. areas along the latitudinal transect in Norway by a factor of 3-5. This presumably reflects differences in the distance from sources and the same factor difference in air concentrations (13, 28). Forest Ecosystems and Altitudinal Distribution. Forest ecosystem zonation with altitude is mainly driven by average temperatures. Jaward et al. (13) showed that air concentrations within the forest canopy are strongly influenced by the vegetation itself. Kylin et al. (29) measured different concentrations in needles within a planted pine forest and attributed them to the effect of biomass density. Leaf concentrations for a given species collected in different parts of the transect or different forests will be influenced by many factors (e.g., density and presence of other species). The MCC was therefore calculated for the “forest canopy” compartment as a whole. Data are presented in Table 2 and Figure 2, for different compounds on the altitudinal gradient. The HCB MCC significantly (P < 0.05) increased with altitude, as did the air concentration (13). This is in agreement with concentrations in needles collected in the Rockies (18) where Davidson et al. related this to the orographic temperature gradient. In their study forest species composition was quite constant with altitude. In the present study the same trend occurs, but across varying canopy compositions. The concentrations of other organochlorine pesticides and total PCB did not vary significantly with altitude (Figure 2). However, a trend of fractionation of PCB congeners in leaves occurred with altitude (see Figure 3), even if not significant. This was also reported in a recent study in pine canopies (27) even if the authors suggested that the temperature dependence may have been a result of the combined seasonal and altitudinal contributions. Previous studies have shown a similar fractionation of PCBs with latitude in soil and air (31-34). It is interesting to note that (i) the altitudinal temperature gradient and distance from sources are smaller than those 6582 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 21, 2006 FIGURE 2. Mean canopy concentration (MCC) of selected compounds along the altitudinal gradient. Bars report EMCC (error in MCC, see eq 3 for definition). ΣPCBs ) sum of PCB 28/31, 52, 101, 118, 138, 153, 180. where latitudinal fractionation has been observed and (ii) no fractionation of PCBs was observed in the air sampled on this transect (13). Fractionation was observed here in a medium (the overall forest canopy) that changes composition along the altitudinal gradient. The progressive replacement of broadleaf species by conifer species with increasing altitude occurs in the Alps (as well as along the latitudinal gradient from the tropics to the sub-arctic region) could result in different accumulation behaviors in different forest canopies. To evaluate the pattern of accumulation in canopies, plant-air partition coefficients (KPA) were calculated for each species. FIGURE 3. Fractionation of PCBs congeners along the altitudinal gradient. Points represent single sample values. Bars report the weighted average for each forest and the EMCC (error in MCC, see eq 3 for definition). Lines indicate the trend of means. Plant-Air Partition Coefficients (KPA) Derived in the Field for Different Species. To allow the contribution of different species to the MCC to be compared, their concentrations (CP) were normalized by dividing them by the air concentrations (CA). Air concentrations were derived from passive sampler data reported by Jaward et al. (13) by assuming an uptake rate of 3.5 m3 day-1 as directly measured and suggested previously (13). Such normalization gave an in-field derived KPA. The air concentrations measured within the forest are influenced by the forest biomass itself, resulting in lower concentrations than observed in clearings (13). Therefore, the concept of an air concentration depletion factor (DF) was introduced. This is a compound and canopy specific parameter, defined as the ratio between the mean atmospheric concentration measured within the forest canopy and the one measured outside the forest (CA,in/CA,out) (13) (Table 1). Leaves were collected from the inner canopy for this study, thereby being in contact with “depleted air”. The KPA values (CP/CA) were therefore corrected by the depletion factors DF. Thus, assuming CA ) CA,out where CA,out and CA,in are the concentrations outside and inside the canopy, respectively: KPA ) CP/(CADF) ) CP/CA,in (4) KPA values derived for each species and chemical therefore reflect the specific bioconcentration factor, assuming equilibrium conditions prevail. Previous studies reported different times to reach plant-air equilibrium, depending on species, experimental conditions, etc., ranging between a few days and several months (35, 36). As far as we are aware, no BCF data were reported for most of these species in the literature, so it is unclear when such conditions prevail. Nevertheless, in this study, mature leaves were sampled in autumn after 4-6 months of exposure (depending on altitude), at the end of their life cycle (the canopies are mainly comprised of deciduous leaves), so if equilibrium was not reached, it may be assumed that this condition was approached as much as temporally possible. Data for log KPA are plotted against log KOA and corrected for the average site temperature and linear regressions were obtained: log KPA ) n log KOA + c VOL. 40, NO. 21, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY (5) 9 6583 Physical-chemical properties and regressions for PCBs and HCB were taken from Li et al. (37) and Harner and Mackay (38), respectively. Table S1 (Supporting Information) shows the regression parameters for each sampled species. Good linear correlations were found in all cases. Intercepts ranged between -1.03 and -4.54, with the highest values for the needle species. The slopes were lower for the needle species (ranging between 0.38 and 0.43) than for the broadleaves (0.53-0.74) and were highest for the chestnut (Castanea sativa). This suggests that broadleaves better reflect the accumulation characteristics of octanol since they have KPA slopes which are closer to 1 when plotted against KOA. These trends are similar to previous reports. Brorström-Lunden and Löfgren (9) present data for spruce and air concentrations which allow an intercept (-1.76) and slope (0.37) to be calculated while Ockenden et al. (6) reported lowest slopes for pine needles (0.51) compared to spruce and larch needles. However, caution must be taken in comparing the accumulation behavior of different types of needles (e.g., pine vs spruce) except where exposure conditions are similar. Di Guardo et al. (39) compared DDT concentrations in pine and spruce needles which were exposed to contaminated air near a point source. They reported differences of a factor of 2-3 in the sequestered amounts in pine needles >2 years old, while concentrations in spruce needle did not significantly vary with age. It was hypothesized that morphological differences in the position of the resin channel within the needle can result in a higher long-term accumulation capacity for pine needles. Concerning broadleaves, few data are available in the literature and most relate to monocotyledon grass species that present very different structural and physiological characteristics. Slopes varying between 0.16 and 0.52, with intercept ranging between -0.8 and -3.8, were found in a field uptake experiment (35). Figure S1 (Supporting Information) presents a comparison of regression lines for all the species in the present study. It appears that needles are more efficient at collecting the more volatile compounds than broadleaves, generally when log KOA < 9.5-10. If the temperature dependence of KPA and KOA are known, it is possible to compare the accumulation behavior of different plant species independently of temperature. Komp and McLachlan (40) showed that KPA increased considerably at lower temperatures in ryegrasses. They derived the enthalpy of phase transfer (∆HPA) experimentally. The application of such approaches to tree leaves is not known. However, to a first approximation, KPA values measured at different temperatures were normalized to the temperature recorded at the 1400 m site using reported ∆HPA (40) and plotted with the respective temperature-corrected KOA. Results are reported in Figure S1b (Supporting Information). Comparing the trend reported in Figures S1a and S1b (Supporting Information), it appears that the temperature gradient does not account for a large part of the observed variability. Figure S1b (Supporting Information) indicates that the “cross point” of the regression model is shifted to log KOA values of about 10-10.5. Spruce and larch showed a net accumulation of more volatile POPs during the exposure times at a rate approximately double that of the broadleaf species, while for less volatile compounds KPA values were closer. The concentrations in each species (values taken from Table 2) can be normalized by multiplying them by their relative abundance in the respective forests (data reported in Table 1). This shows that, at the 1400 m site, spruce represents only 25% of the total foliar biomass but accounts for 40% of the total load of HCB or PCB 28/31 in the canopy, yet only 14% for PCB 180. Similarly, at the 1100 m site, the accumulation of more volatile compounds is due primarily to the maple, which captures about 60% of the total HCB in the canopy. 6584 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 21, 2006 These results show that the specific composition of the forest and the relative abundance of species foliar biomass are important in controlling the distribution and load of POPs in the canopy, together with deposition velocities and the forest ecological parameters such as density or volume. All these parameters will influence the distribution and uptake of POPs in forests (7-10, 15-17). A Model for Dry Gaseous Deposition to the Forest Canopy. The results presented above showed that in complex forest systems, such as those occurring in temperate areas and characterized by large biodiversity and high density, forest structural parameters will be important for the development of accumulation models. McLachlan and Horstmann (10) have described the dry gaseous deposition of organic chemicals as a diffusive process that can be treated as a one-compartment model as follows: [ ( kaV ∆t CV ) KVACA 1 - exp KVA )] (6) where CV and CA are the concentrations in the vegetation and in air (mol m-3), respectively. The term KVA is the vegetation-air partition coefficient and could be derived from eq 1 as a function of KOA. k (m s-1) is the mass-transfer coefficient describing transport from the air to the vegetation and aV is the specific surface area of the vegetation AV/VV (area of vegetation (m2)/volume of vegetation (m3)). In this model, the canopy is a well-mixed monospecies compartment. In a multispecies forest, the accumulation behavior of different species can affect both the pollutant uptake rate and partitioning with the atmosphere, depending on the plant species and their contribution to the forest canopy biomass, as shown in the previous paragraphs. The KVA should therefore refer to the canopy as a whole, and for the sake of clarity will be called KCA, where KCA ) MCCV CADF (7) where MCCV (ng m-3) is the mean canopy concentration on a volume basis, derived using the contribution of each species leaf volume to the total canopy volume. Measures of DF and its dependence on log KOA are reported elsewhere (13) for three different forest types. Thus, eq 6 can be rewritten as follows: [ ( kaV MCCV ) KCACADF 1 - exp ∆t KCA )] (8) Most of the problems of canopy parametrization arise in estimating the volumetric concentrations, which requires knowledge of the vegetation density F (kg m-3 based on dry weight). Variability in this parameter can introduce a large uncertainty in the predictive model. Specifically, estimating values for the average leaf thickness (17) to represent the overall canopy density (FC) is problematic. Intraspecies leaf thickness variability of up to 40% depends on ecological conditions, notably the total daily irradiance received by leaves during their development (41). It is therefore useful to obtain a F-independent model. The parameter aV (specific surface area of vegetation) is F-dependent and it can be expressed in terms of the specific leaf area (SLA) (m2 kg-1), a parameter that expresses the amount of leaf surface (of one face) of a certain species per unit of dry weight, as follows, aV ) FC2SLAC (9) where the subscript indicates that the SLA refers to its mean value in the canopy. The SLA is commonly used in plant ecology as an index of the photosynthetic efficiency. It can be easily measured (22) and obtained from the literature. Thus, to obtain a model independent of leaf density (F), eq 9 was introduced into eq 8 and both members were divided by FC: [ ( MCCV KCA kFC2SLAC ) CADF 1 - exp ∆t FC FC KCA )] (10) The ratio MCCV/FC and KCA/FC represent the MCC (mol kg-1) and the canopy/air partition coefficient K/CA on a massvolume basis (m3 kg-1), respectively. Thus, ( [ MCC ) K/CACADF 1 - exp - k2SLAC ∆t K/CA )] (11) Equation 11 describes the dry gaseous deposition to the canopy of a multispecies dense forest at a given time, independent of F. It is easy now to describe the dry gaseous deposition to the forest canopy per unit of ground surface N (mol m-2 s-1), by introducing the leaf area index (LAI) (m2 m-2), a measure of the amount of vegetation surface per unit of ground surface, as follows: LAI N ) MCC ) SLAC [ ( k2SLAC LAI / K C DF 1 - exp ∆t SLAC CA A K/ CA )] (12) LAI/SLA (kg m-2) is a measure of the foliar biomass. As for SLA, LAI values are commonly available in spatially and monthly resolved forms (42). They can also be obtained using remote sensing techniques for regional application. As an example, the data reported in Table 2 and Table S1 (Supporting Information) can be used to calculate N for the three forests of this study, assuming they have reached the partitioning equilibrium with the atmosphere. In that case eq 12 can be simplified as follows: LAI N ) MCC SLAC SLAC was obtained by averaging specific SLA values considering the relative abundance of each species. Results are reported in Figure 4. Sites at 1100 and 1400 m appear to be the most efficient in sequestering POPs. When a comparison is performed by considering the same LAI value ) 1 to each forest type (Figure 4b), the coniferous forest at 1800 m appears much more efficient than the others. This normalization makes it possible to assess the storage capacity of the canopy due to density-independent factors. The enhanced efficiency of the coniferous canopy in storing POPs is mainly due to the high storage capacity of needles for accumulating gasphase POPs, even though their SLA is on average 50-70% lower than that of the broadleaves. The SLA therefore does not explain the differences in accumulation behavior, which are presumably due to other structural/chemical features of the leaves. Considering the parameters expressed in eq 12, it appears that the accumulation of gaseous POPs by canopies is a process strongly influenced by the dynamics of canopy growth and structure. Most of the parameters employed vary with time and ecological conditions. In temperate deciduous forests, for example, LAI can range from <1 to 4-5 (m2 m-2) during the growing season, influencing the DF and the mass balance of pollutant reaching the canopy. SLA is subject to both inter- and intraspecific variability that can reach 50% (43) depending on light availability. A similar variability of SLA with time was also observed in a study of the carbon FIGURE 4. (a) Pollutant load in forest canopy (ng m-2 of ground surface). (b) Density-independent pollutant load in forest canopy (LAI ) 1 m2 m-2 for each forest type). cycle of a tropical forest (44). KPA and consequently KCA may also be subject to variability, related to the life span of different leaves, due to the structural and chemical changes occurring in leaves during the growing period. An accurate evaluation of the impact of forest biomass on the environmental fate of POPs therefore needs to take account for the spatial and temporal variability of forest structural parameters. Acknowledgments MIUR-COFIN 2001 is acknowledged for financial support. We thank Mary Chiara (Mont Mars Regional Reserve, Valle d’Aosta Region) for help with site selection, quality control at the sampling sites, and general logistics. Supporting Information Available An additional table and figure illustrating regressions parameters for the log KPA equation and interspecific comparison of log KPA vs log KOA regression lines is reported. This material is available free of charge via the Internet at http:// pubs.acs.org. Literature Cited (1) Calamari, D.; Bacci, E.; Focardi, S.; Gaggi, C.; Morosini, M.; Vighi, M. Role of plant biomass in the global environmental partitioning of chlorinated hydrocarbons. Environ. Sci. Technol. 1994, 28, 429-434. (2) Eriksson, G.; Jensen, S.; Kylin, H.; Strachan. W. The pine needles as monitor of atmospheric pollution. Nature 1989, 341, 42-44. (3) Gaggi, C.; Bacci, E.; Calamari, D.; Fanelli, R. Chlorinated hydrocarbons in plant foliage: an indication of the troposheric contamination level. Chemosphere 1985, 14, 1673-1686. (4) McLachlan, M. S. Bioaccumulation of hydrophobic chemicals in agricultural food chains. Environ. Sci. Technol. 1996, 30, 252259. (5) McLachlan, M. S. A simple model to predict accumulation of PCDD/Fs in an agricultural food chain. Chemosphere 1997, 34, 1263-1276. (6) Ockenden, W. A.; Steinnes, E.; Parker, C.; Jones, K. C. Observations on persistent organic pollutants in plants: implications for their use as passive air samplers and POP cycling. Environ. Sci. Technol. 1998, 32, 2721-2726. VOL. 40, NO. 21, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 6585 (7) Horstmann, M.; Bopp, U.; McLachlan, M. S. Comparison of the bulk deposition of PCDD/F in a spruce forest and adjacent clearing. Chemosphere 1997, 34, 1245-1254. (8) Horstmann, M.; McLachlan, M. S.; Atmospheric deposition of semivolatile organic compounds to two forest canopies. Atmos. Environ. 1998, 32, 1799-1809. (9) Brorström-Lundén, E.; Löfgren, C. Atmospheric fluxes of persistent semivolatile organic pollutants to a forest ecological system at the Swedish west coast and accumulation in spruce needles. Environ. Pollut. 1998, 102, 139-149. (10) McLachlan, M. S.; Horstmann, M. Forests as filters of airborne organic pollutants: a model. Environ. Sci. Technol. 1998, 32, 413-420. (11) Hornbuckle, K. C.; Eisenreich, S. J. Dynamics of gaseous semivolatile organic compounds in a terrestrial ecosystem effects of diurnal and seasonal climate variations. Atmos. Environ. 1996, 30, 3935-3945. (12) Gouin, T.; Thomas, G. O.; Cousins, I.; Barber, J.; Mackay, D.; Jones, K. C. Air-surface exchange of polybrominated diphenyl ethers and polychlorinated biphenyls. Environ. Sci. Technol. 2002, 36, 1426-1434. (13) Jaward, F. M.; Di Guardo, A.; Nizzetto, L.; Cassani, C.; Raffaele, F.; Ferretti, R.; Jones, K. C. PCBs and selected organochlorine compounds in Italian mountain air: the influence of altitude and vegetation type. Environ. Sci. Technol. 2005, 39, 34553463. (14) Simonich, S. L.; Hites, R. A. Importance of vegetation in removing polycyclic aromatic hydrocarbons from the atmosphere. Nature 1994, 370, 49-51. (15) Wania, F.; Mclachlan, M. S. Estimating the Influence of Forests on the Overall Fate of Semivolatile Organic Compounds Using a Multimedia Fate Model. Environ. Sci. Technol. 2001, 35, 582590. (16) Wegmann, F.; Scheringer, M.; Möller, M.; Hungebühler, K. Influence of vegetation on the environmental partitioning of DDT in two global multimedia models. Environ. Sci. Technol. 2004, 38, 1505-1512. (17) Yushan, S.; Wania, F. Does the Forest Filter Effect Prevent Semivolatile Organic Compounds from Reaching the Arctic? Environ. Sci. Technol. 2005, 39, 7185-7193. (18) Davidson, D.; Wilkinson, A.; Blais, J. M. Orographic cold-trapping of persistent organic pollutants by vegetation in mountains of western Canada. Environ. Sci. Technol. 2003, 37, 209-215. (19) Kallenborn, R. Persistent organic pollutant (POPs) as environmental risk factors in remote high-altitude ecosystems. Ecotox. Environ. Safety 2006, 63, 108-112. (20) Nizzetto, L.; Cassani, C.; Di Guardo, A. Deposition of PCBs in mountains: The forest filter effect of different forest ecosystem types. Ecotox. Environ. Safety 2006, 63, 75-83. (21) SW-846, Test Methods for Evaluating Solid Waste, Physical/ Chemical Methods; U.S. Environmental Protection Agency, U.S. Government Printing Office: Washington, DC, 1998; Vol. 1. (22) Garnier, E.; Shipley, B.; Roumet, C.; Laurent, G. A standardized protocol for determination of specific leaf area and leaf dry matter content. Funct. Ecol. 2001, 15, 688-695. (23) Böhme, F.; Welsch-Pausch, K.; McLachlan, M. S. Uptake of semivolatile organic compounds in agricultural plants: field measurements of interspecies variabilities. Environ. Sci. Technol. 1999, 33, 1805-1813. (24) Weiss, P.; Lorbeer, G.; Scharf, S. Regional aspects and statistical characterization of the load with semivolatile organic compounds at remote Austrian forest sites. Chemosphere 2000, 40, 1159-1171. (25) Jaward, F. M.; Farrar, N. J.; Harner, T.; Sweetman, A. J.; Jones, K. C. Passive air sampling of PCBs, PBDEs and organochlorine pesticides across the Europe. Environ. Sci. Technol. 2004, 38, 34-41. (26) Davidson, D. A.; Wilkinson, A. C.; Kimpe, L. E.; Blais, J. M. Persistent organic pollutants in air and vegetation in Canadian Rocky Mountains. Environ. Toxicol. Chem. 2004, 23, 540-549. (27) Grimalt, J. O.; van Drooge, B. L. Polychlorinated biphenyls in mountain pine (Pinus uncinata) needles from Central Pyrenean 6586 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 21, 2006 (28) (29) (30) (31) (32) (33) (34) (35) (36) (37) (38) (39) (40) (41) (42) (43) (44) high mountains (Catalonia, Spain). Ecotox. Environ. Safety 2006, 63, 61-67. Jaward, F. M.; Meijer, S. N.; Steinnes, E.; Thomas, G. O.; Jones, K. C. Further study in the latitudinal and temporal trends of persistent organic pollutant in Norwegian and U.K. background air. Environ. Sci. Technol. 2004, 38, 2523-2530. Kylin, H.; Hellstrom, A.; Nordstrand, E.; Zaid, A. Organochlorine pollutants in scots pine needles - biological and site related variation within a forest stand. Chemosphere 2003, 51, 669675. Barber, L.; Thomas, G. O.; Parkman, S. A.; Jones, K. C. Variability in the PCB concentrations of vegetation. Organohalogen Compd. 1999, 41, 379-382. Agrell, C.; Okla, L.; Larsson, P.; Backe, C.; Wania, F. Evidence of latitudinal fractionation of polychlorinated biphenyl congeners along the Baltic sea region. Environ. Sci. Technol. 1999, 33, 1149-1156. Meijer, S. N.; Ockenden, W. A.; Steinnes, E.; Corrigan, B. P.; Jones, K. C. Spatial and temporal trends of POPs in norwegian and UK background air: implications for global cycling. Environ. Sci. Technol. 2003, 37, 454-461. Meijer, S. N.; Ockenden, W. A.; Sweetman, A.; Breivik, K.; Grimalt, J. O.; Jones, K. C. Global distribution and budget of PCBs and HCB in Background surface soils: implication for sources and environmental processes. Environ. Sci. Technol. 2003, 37, 667672. Ockenden, W. A.; Breivik, K.; Meijer, S. N.; Steinnes, E.; Sweetman, A. J.; Jones, K. C. The global re-cycling of persistent organic pollutants is strongly retarded by soils. Environ. Pollut. 2003, 121, 75-80. Thomas, G. O.; Smith, K. E. C.; Sweetman, A. J.; Jones, K. C. Further studies of the air-pasture transfer of polychlorinated biphenyls. Environ. Pollut. 1998, 102, 119-128. Barber, J. L.; Thomas, G. O.; Kerstiens, G.; Jones, K. C. Air-side plant -side resistances influence the uptake of airborne PCBs by evergreen plants. Environ. Sci. Technol. 2003, 36, 32243229. Li, N.; Wania, F.; Ley, D.; Daly, G. A. Comprehensive and critical compilation, evaluation, and selection of physical-chemical property data for selected polychlorinated biphenyls. J. Phys. Chem. Ref. Data 2003, 32, 1545-1590. Harner, T.; Mackay, D. Measurement of octanol-air partition coefficients for Chlorobenzenes, PCBs and DDT. Environ. Sci. Technol. 1995, 29, 1599-1606. Di Guardo, A.; Zaccara, S.; Cerabolini, B.; Acciari, M.; Terzaghi, G.; Calamari, D. Conifer needles as passive biomonitors of the spatial and temporal distribution of DDT from a point source. Chemosphere 2003, 52, 789-797. Komp, P.; McLachlan, M. S. Influence of temperature on the plant/air partitioning of semivolatile organic compound. Environ. Sci. Technol. 1997, 31, 886-890. Nobel, P. S. Internal leaf area and cellular CO2 resistance: photosynthetic implications of variations with growth conditions and plant species. Physiol. Plant. 1977, 40, 137-144. Meeson, B. W.; Corprew, F. E.; McManus, J. M. P.; Myers, D. M.; Closs, J. W.; Sun, J.-K.; Sunday, D. J.; Sellers, P. J. Global Data Sets for Land-Atmosphere Models Available on CD and On-line; LSLSCP Initiative 1: 1987-1988; American Geophysical Union: March 1995; Vols. 1-5, http://www.agu.org/eos_elec/ 95212e.html. Gratani, L. Canopy structure, vertical radiation profile and photosynthetic function in a Quercus ilex evergreen forest. Photosynthetica 1997, 33, 139-149. Nebel, G.; Dragsted, J.; Vega, A. S. Litter fall biomass and net primary production in flood plain forests in the Peruvian Amazon. For. Ecol. Manage. 2001, 150, 93-102. Received for review March 8, 2006. Revised manuscript received August 8, 2006. Accepted August 23, 2006. ES0605523