Characterization of water extractable organic matter in a deep soil profile

Transcription

1 Chemosphere 62 (2006) Characterization of water extractable organic matter in a deep soil profile Maddalena Corvasce a, *, Adam Zsolnay b, Valeria DÕOrazio a, Raffaele Lopez a, Teodoro M. Miano a a Dipartimento di Biologia e Chimica Agroforestale ed Ambientale, Univ. di Bari, Via Amendola 165/A, Bari, Italy b Inst. Soil Ecol., GSF-Res. Center Environ. Health, Neuherberg bei München, Germany Received 6 September 2004; received in revised form 28 June 2005; accepted 5 July 2005 Available online 19 September 2005 Abstract The aim of this study was to identify qualitative and quantitative differences of water extractable organic matter (WEOM) isolated from each horizon along a deep soil profile and to evaluate any relationship between the WEOC and the total organic carbon (TOC) content. The soil profile Monte Pietroso is located in the Murge area, Apulia region in Southern Italy. Samples from the eight horizons (Ap1, Ap2, Ab1, Ab2, Bt1, 2B, 2Bt2, and 2B/C) were collected in October The WEOM characterization was carried out by means of UV absorbance, fluorescence spectroscopy in the emission and excitation/emission matrix (EEM) modes, and additional spectroscopic derived indexes. Soil organic carbon was shown to accumulate in the top horizons (Ap) and, in general, to decrease with depth, whereas, the WEOM/TOC ratio increases with increasing depth. The aromaticity and the humification index of the WEOM decrease dramatically downward the soil profile, whereas the fluorescence efficiency index tends to increase markedly. The WEOM fractions feature three main fluorophores with different wavelength and relative intensity. In general WEOM transport phenomena are suggested to occur downward the soil profile, depending on the nature of the organic material and on the chemical and mineral characteristics of the various horizons. Ó 2005 Elsevier Ltd. All rights reserved. Keywords: Soil organic matter; Water extractable organic matter; Total organic carbon; Humification index; Fluorescence spectroscopy 1. Introduction Soil organic matter (SOM) is the final result of the combined effect of decomposition and humification processes driven by microorganisms on plant and animal residues (Stevenson, 1986). Nature, properties and * Corresponding author. Tel.: ; fax: address: m.corvasce@agr.uniba.it (M. Corvasce). dynamics of SOM depend on numerous variables, including environmental and pedological conditions, types and activity of biological populations, soil use and anthropogenic impacts (Stevenson, 1996). Microbial biomass, including bacteria, fungi and algae, temperature, nutrients and content and availability of water affect markedly mineralization of SOM thatõs transformed in both mobile and immobile fractions. In general, most OM is decomposed within the top 10 cm of soil yielding a soluble portion, which migrates to a certain extent towards deeper soil horizons, and a /$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved. doi: /j.chemosphere

2 1584 M. Corvasce et al. / Chemosphere 62 (2006) portion that is immobilized in the top soil (Tate, 1987). Mineralization and humification processes are key factors influencing SOM translocation to deeper horizons (Nelson et al., 1994). Only a small portion of SOM is mobile and distributed along the soil profile within the liquid phase (Boyer and Groffman, 1996). Dissolved organic matter (DOM) represents the most active and mobile form of SOM. Although several studies have been conducted and several information have been obtained in recent years, additional investigation on DOM evolution, stability and dynamics in soils is still needed, in order to better understand its role in SOM translocation and mineralization processes, humification phenomena and interactions with organic and inorganic compounds (Kalbitz et al., 2000). DOM can be considered the major soluble component of natural aquatic systems, but its definition in soil-sediments matrices is still poorly defined. Further, the ecological fate and availability of DOM may be affected by its location in the different pore structures (Zsolnay, 2003). A common approach defines DOM through filtration, by establishing a size limit (0.45 lm) to differentiate the latter from particulate organic matter. In general, the term DOM is used unspecifically to indicate organic material truly dissolved in situ, whereas WEOM (water extractable organic matter) is the corrected acronym, which provides information as to what type of DOM fraction is being referred to. In particular, WEOM is the fraction of DOM extracted with gentle conditions, and conceptually consisting of the mobile and available portion of the total DOM pool (Zsolnay, 1996). Monodimensional fluorescence spectroscopy has been applied extensively to the molecular and functional characterization of DOM of various origins. Recently, total luminescence spectroscopy (TLS) is revealed as a powerful tool to obtain detailed information on fluorescent compounds occurring in DOM (Coble, 1996; Chen et al., 2002, 2003; Goslan et al., 2004). However, very few studies have focused on the dynamics of DOM as well as on its composition, molecular structure and functional properties in the deep soil horizons. The objectives of this study were: (i) to investigate WEOM occurrence and composition with increasing soil depth by means of fluorescence spectroscopy; (ii) to evaluate the existence of possible relationship between the WEOM and the total organic carbon (TOC) content in soil and, (iii) to study changes in humification and aromaticity of WEOM along a deep soil profile. 2. Materials and methods 2.1. Site description The selected site Monte Pietroso is located in the Murge area, Apulia region, in southern Italy, at an altitude of about 470 m above sea level. This area is SE NW oriented, and characterized by a moderate relief, a scarcely developed river network, limestone rock outcrops, and deep soils in the lower areas and shallow ones in the higher ones. The Murge area is a high plain, homogeneous geomorphological unit where calcareous Mesozoic rocks are widely distributed and deeply weathered by epigeous and hypogeal karst phenomena (Lopez and Miano, 2000). The climate is typically Mediterranean with rainy winters and dry and hot summers. The mean annual rainfall is approximately 600 mm and the average temperature is 14.7 C (observations from 1970 to 1994). Natural vegetation is also typically Mediterranean and is characterized by oaks, wild pear, olive and almond trees, and bushes. A soil pit approximately 2.0 m deep was excavated by means of a digger in one of the several small valleys, named lama, crossing the Murge high plain. The soil profile was characterized by several horizons designated as Ap1, Ap2, Ab1, Ab2, Bt1, 2B, 2Bt2, and 2B/C, according to the Soil Taxonomy (Soil Survey Staff, 1998). Samples from each horizon were collected in October 2002 and characterized for their main chemical and physical parameters according to the methods proposed by the International Union of Soil Sciences Organic matter extraction An aliquot of each 2 mm-sieved soil sample was added with a 10 mm CaCl 2 solution at a ratio of 2:1 = solution:soil and gently stirred for about 10 min at room temperature. The suspension was then centrifuged at 4000 rpm for 10 min and filtered through a 0.4 lm pore polycarbonate filter (Zsolnay, 2003). Solutions were then acidified to ph 2 with 2 M HCl prior to DOC analyses (Zsolnay, 2003). This ph value was considered to be the standard condition in the spectroscopic analyses in order to avoid possible artefacts subsequent to further ph adjustments. The WEOM was analysed for DOC concentration and several spectroscopic parameters. All measures were carried out in triplicate. The WEOM content was quantified by determining the amount of water extractable organic carbon (WEOC) in each sample after filtration. The WEOC concentration was measured by using a Shimadzu total organic carbon analyser, mod. TOC-5050, operating at high temperature and catalytic oxidation. Carbonates were automatically removed by the instrument Spectroscopic analyses Absorbance was measured on a Varian spectrometer, model Cary 50 at 254 nm in 5 ml quartz cuvette.

3 M. Corvasce et al. / Chemosphere 62 (2006) Fluorescence emission and excitation/emission matrix (EEM, contour map) spectra of WEOM extracts were recorded on a Varian Cary Eclipse spectrofluorimeter equipped with a PC, and elaborated by means of Noesys 2.4 software. The slit widths were 10 nm and 20 nm for excitation and emission monochromators, respectively, and the scan speed was kept at 4800 nm/ min for the emission spectra and at 6000 nm/min for the EEM spectra. All data were normalized to the intensity of the water Raman scatter peak, determined daily with freshly distilled water, and corrected for inner filter effects by means of absorbance at 254 nm. Emission spectra were obtained in the range of nm at a fixed excitation wavelength of 254 nm, whereas the EEMs were obtained by subsequent scanning the emission spectra from 300 to 500 nm by varying the excitation wavelength from 250 to 400 nm at 10 nm increments Spectroscopic indexes Additional parameters derived from spectroscopic analyses were: (a) the aromaticity index (AI), obtained by dividing the UV specific absorption at 254 nm (cm 1 ) multiplied per 100 (m 1 ) by the WEOC content (mg l 1 ); (b) the humification index (HIX), defined as the ratio between the area in the upper quarter (R nm) of the usable fluorescence emission spectrum and the area in the lower usable quarter (R nm) (Zsolnay et al., 1999). This calculation is based on the fact that humification is considered to be associated with an increase in the C/H ratio, and with a resulting shift to higher fluorescence emission wavelengths; and, (c) the fluorescence efficiency index (F eff ), calculated as the ratio between the maximum fluorescence intensity (F max ) of each emission spectrum and the relative UV absorption at 254 nm (F max /Abs). The F eff is considered to be proportional to the fluorophores quantum efficiency (Zsolnay, 2003). 3. Results and discussion Selected chemical and physical parameters of the soil profile studied are reported in Table 1. A slight increase of ph values is observed in the lower portion of the profile, even though ph is almost constant below 60 cm of depth. The electrical conductivity (EC) shows only small changes with the exception of a marked increase in the Ab horizons. Total carbonates content appears to be neglectable in the clay-silty and clayey horizons (from Ap1 to Ab2, and 2B/C), whereas it reaches up to 9 g kg 1 in the sandy horizons (Bt1, 2B and 2Bt2). As expected, total organic carbon (TOC), total nitrogen (N tot ) contents and C/N ratio decrease with increasing depth, ideally dividing the whole profile in three main zones: (1) the top one, including horizons Ap1, Ap2 (0 30 cm); (2) the intermediate one, including horizons Ab1, Ab2 and Bt1 ( cm) and (3) the bottom one, that includes horizons 2B, 2Bt2 and 2B/C ( cm). The soil texture values indicate high variability throughout the profile, with the predominance of the finer fractions (>30%) in horizons Ap1, Ap2, Ab1, Ab2 and Bt1, and an increase of the sandy fraction (45%) in horizon 2B and silty fraction (45%) in horizon 2Bt2. Table 2 shows the water extractable organic carbon (WEOC) content measured in each horizon, expressed either in weight (lg g 1 of air dried soil) or in volume (mg l 1 of water extract) and the WEOC/TOC ratios, expressed in percentage. Similar to TOC, the WEOC values decrease significantly with increasing soil depth, and show marked changes between the three different zones of the profile. In particular, in the top horizons the amount of WEOC is approximately three times greater than the value in the deepest ones (Table 2). The decrease of the WEOC concentration in deep soil horizons may be attributed to adsorptive interaction of WEOM with reactive surfaces of the mineral matrix (minerals and oxides) (Kaiser and Guggenberger, 2000; Kalbitz et al., 2000; Shen, 1999). Table 1 Main chemical and physical parameters of the Monte pietroso soil profile Horizon Depth (cm) ph (H 2 O) EC (ds m 1 ) Carbonates (g kg 1 ) Sand (%) Site (%) Clay (%) TOC (g kg 1 ) N tot (g kg 1 ) Ap Trace Ap Trace Ab Trace Ab Trace Bt B Bt B/C Trace EC = electrical conductivity determined in water soil extract 1:2 at 25 C; TOC = total organic carbon; N tot = total nitrogen; CN = carbon/total nitrogen weight ratio. C/N

4 1586 M. Corvasce et al. / Chemosphere 62 (2006) Table 2 Water extractable organic carbon (WEOC) content and WEOC/TOC ratio measured along the soil profile Horizon Depth (cm) WEOC (lgg 1 ) WEOC (mg l 1 ) Ap Ap Ab Ab Bt B Bt B/C WEOC/TOC (%) The WEOC/TOC ratios increase approximately four times along the profile. Results indicate that in spite of the decreasing amount of total C with depth, the relative abundance of WEOC tends markedly to increase, thus suggesting specific dynamics between TOC and WEOC related to the physical and/or chemical characteristics of each horizon. In general, the amount of WEOC decreases with depth, as already observed, showing no apparent relationship with the textural characteristics of each horizon along the entire profile (Fig. 1a and b). At the same time, a closer view to selected portions of the soil profile indicates very distinct trends. In the lower portion (2B, 2Bt2, 2B/C), in fact, the amount of WEOC increases almost linearly with the decrease of clay (R 2 = 0.90), and with the increase of sand percentages (Fig. 1b). The result is possibly consistent with strong physico chemical interactions occurring between the nature of the dissolved organic matter and the specific mineralogical and chemical properties of the mineral surfaces in the lower portion of the profile. The pattern is rather different from Ap1 to Bt1 horizons (Fig. 1a). The amount of WEOC slightly increases with the decrease of clay and sand contents whereas an opposite behaviour is observed for the silty fraction. Results show a bimodal distribution of the WEOC along the profile. In depth (2B, 2Bt2, 2B/C), WEOC appears strongly connected to soil clay particles, whereas in the upper portion of the profile the WEOC behaviour is more independent on the textural characteristics. This result may also indicate a different pedogenetical path of the deeper soil portions which is possibly related to severe changes in the environmental parameters occurred during the formation of the soil in the selected area of study. The aromaticity index (AI) of WEOM is much higher in the top most horizon, Ap1, than in all other horizons (Fig. 2). This result may be attributed to the decay byproducts of the organic residues applied to the soil surface. Plant tissues contain various classes of aromatic compounds, which can be released into the soil top layer together with the products of microbial origin (Qualls and Haines, 1992). Further, aromatic compounds tend to be preferentially adsorbed onto the soil particles rather than dissolving into the liquid phase (McKnight et al., 1992). In horizons below the surface, a preferential retention of highly condensed and substituted aromatic molecules, which are recalcitrant to mineralization processes, would occur, thus decreasing their amount in WEOM (Kalbitz, 2001), whose aromaticity tends to remain almost constant, i.e. between 0.5 and 0.7 l mg 1 m 1. Fluorescence emission spectra of WEOM in the eight horizons (Fig. 3) show a similar shape; featuring a unique broad band with a maximum of different varying relative fluorescence intensity (RFI), which progressively shifts from 406/413 nm to 361 nm with increasing soil depth. This result is consistent with a progressive decreasing of molecular size and structural complexity of WEOM with increasing soil depth, particularly in the deeper horizons. Similar effects were previously observed for fulvic acids of diverse origins (Miano et al., 1988, 1991). The humification index (HIX) of WEOM decreases sharply with increasing soil depth, especially when passing from the top two horizons to deeper horizons Fig. 1. Relations between WEOC and soil texture. (a) Relations between WEOC and soil texture of the Ap1, Ap2, Ab1, Ab2, Bt1 horizons (0 115 cm). (b) Relations between WEOC and soil texture of the 2B, 2Bt2, 2B/C horizons ( cm).

5 M. Corvasce et al. / Chemosphere 62 (2006) Fig. 2. Aromaticity Index of WEOM in soil horizons. Error bars show SD of the mean. Fig. 4. Humification index of WEOM in soil horizons. Error bars show SD of the mean. on this soil. Therefore, the deeper horizons tend to retain preferentially more humified matter because of physical and chemical interaction mechanisms. In other words, it is feasible that more complex, aromatic and/or hydrophobic components of WEOM are relatively less mobile and, thus are preferentially retained onto soil mineral surfaces; whereas only smaller and more hydrophilic molecules are more mobile and can migrate towards deeper soil horizons. The fluorescence efficiency index (F eff ) is operationally related to the maximum value of the fluorescence intensity of each spectrum and the UV absorption measured at the excitation wavelength (254 nm). The WEOM in top and intermediate horizons exhibits similar F eff values (Fig. 5), which tend to increase markedly in the deeper horizons. This result can be attributed to the selective downward migration of WEOM components characterized by highly efficient and simple fluorophores. Excitation emission matrix (EEM, contour maps) spectra of WEOM isolated from each horizon of the soil profile are shown in Fig. 6, where the x-axis and the Fig. 3. Fluorescence emission spectra of WEOM in soil horizons. (Fig. 4). This result follows closely the pattern of total organic carbon along the profile. The higher the amount of C, the higher is the amount of WEOM occurring in the top horizon. In addition, horizons Ap1 and Ap2 are highly disturbed by agricultural practices performed Fig. 5. Fluorescence efficiency index of WEOM in soil horizons. Error bars show SD of the mean.

6 1588 M. Corvasce et al. / Chemosphere 62 (2006) y-axis represent the scan range of emission and excitation wavelengths, respectively. Colours represent the relative fluorescence intensity (RFI) of WEOM, and the relative scale is reported on the left side of the Fig. 6. In general, three different fluorophores, named a, b and c, not necessarily occurring in all horizons, can be observed in the EEM spectra. Each fluorophore is characterized by an excitation/emission wavelength pair (WWP) and by a specific RFI. The fluorophore a is centered at a WWP 250 ex / em, the fluorophore b at a WWP ex / em, and the fluorophore c at a WWP 270 ex /300 em (Table 3). Based on previous results, these fluorophores may be attributed to specific molecular units of WEOM. In particular, fluorophore a represent structurally simple compounds, humic- and fulvic-like moieties and protein derived materials (Baker and Genty, 1999). Fluorophore b can be likely related to structures such as simple phenolic-like hydroxy-substituted benzoic and cinnamic acid derivatives, coumarins, and alkaloid-like hydroxyquinolines (Senesi et al., 1991; Wolfbeis, 1985). Fluorophore c is typical of aromatic aminoacids, simple phenolic compounds with the conjugated electronic system limited to one aromatic ring, and nucleic acids (Blaser et al., 1999; Wolfbeis, 1985). Changes of various extent in the WWPs and RFI values of each WEOM fluorophore can be observed along the soil profile (Table 3). Fluorophores a and b are typical of WEOM from the first two horizons, which have a similar WWP value but a RFI value that decreases markedly from Ap1 to Ap2. The EEM spectra of WEOM from the intermediate horizons (Ab1, Ab2 and Bt1) feature the presence of fluorophores a and c, both showing low RFI values. Further, with increasing depth, fluorophore a is slightly shifted towards shorter k em. The EEM spectra of WEOM isolated from horizons 2B, 2Bt2 and 2B/C exhibit the three fluorophores. In particular, with increasing depth fluorophores a and b show a small shift of WWP towards shorter wavelengths, whereas fluorophore c feature a constant WWP value. Further, the RFI of fluorophore c increases towards the bottom of the profile, whereas fluorophores a and b exhibit their maximum RFI in WEOM from horizon 2Bt2. 4. Conclusions Fig. 6. Contour Maps of WEOM in soil horizons a, b, c indicate the main flourophores (see Table 3). The WEOM fraction isolated from the various horizons of a Mediterranean soil profile shows significant qualitative and quantitative differences. In general, the fluorescence intensity, and the aromaticity and humification indexes of WEOM decrease with soil depth. These results suggest the preferential flow of structurally simple molecules towards deeper soil horizons, and the feasible occurrence of retention phenomena of more complex and aromatic WEOM molecules by mineral components of deeper soil horizons. This is confirmed by the fact that the fluorescence efficiency increase with depth (Fig. 5). With increasing soil depth WEOM amount increases, relatively to TOC content. In particular the amount of WEOM show a bimodal distribution along the profile. In depth (2B, 2Bt2, 2B/C), WEOM appears strongly connected to soil clay particles, whereas

7 M. Corvasce et al. / Chemosphere 62 (2006) Table 3 Excitation/emission wavelength values and corresponding relative fluorescence intensity (RFI) values of the main fluorophores (a,b,c) in WEOM along the soil profile Depth Horizon a (k ex /k em ) RFI b (k ex /k em ) RFI c (k ex /k em ) RFI 0 4 Ap1 250/ / Ap2 250/ / Ab1 250/ / Ab2 250/ / Bt1 250/ B 250/ / / Bt2 250/ / / B/C 250/ / / in the upper portion of the profile the WEOM behaviour is more independent of the textural characteristics. Slight but important differences in WEOM properties, can be highlighted by fluorescence spectroscopy analysis both in the fluorophores types and in their relative fluorescence intensity and efficiency. In particular, EEM data suggest the existence of three main zones along the whole profile, namely: Ap1 and Ap2; Ab1, Ab2 and Bt1; and 2B, 2Bt2 and 2B/C. In particular the upper zone is richer in WEOM characterized by a more complex chemical structure and greater aromaticity; the intermediate zone features a homogeneous mixture of small WEOM molecules with weak fluorescence performances; and the third zone shows the highest variety of WEOM molecules featuring low size and complexity but high fluorescing efficiency and intensity. References Baker, A., Genty, D., Fluorescence wavelength and intensity variations of cave waters. Journal of Hydrology 217, Blaser, P., Heim, A., Luster, J., Total luminescence spectroscopy of NOM-typing samples and their aluminium complexes. Environment International 25, Boyer, J.N., Groffman, P.M., Bioavailability of water extractable organic carbon fractions in forest and agricultural soil profiles. Soil Biology and Biochemistry 28 (6), Chen, J., Gu, B., LeBoeuf, E.J., Pan, H., Dai, S., Spectroscopic characterization of the structural and functional properties of natural organic matter fractions. Chemosphere 48, Chen, J., LeBoeuf, E.J., Dai, S., Gu, B., Fluorescence spectroscopic studies of natural organic matter fractions. Chemosphere 50, Coble, P.G., Characterization of marine and terrestrial DOM in seawater using excitation emission matrix spectroscopy. Marine chemistry 51, Goslan, E.H., Voros, S., Banks, J., Wilson, D., Hillis, P., Campbell, A.T., Parsons, S.A., A model for predicting dissolved organic carbon distribution in a reservoir water using fluorescence spectroscopy. Water Research 38, Kaiser, K., Guggenberger, G., The role of DOM sorption to mineral surfaces in the preservation of organic matter in soils. Organic Geochemistry 31, Kalbitz, K., Properties of organic matter in soil solution in a german fen area as dependent on land use and depth. Geoderma 104, Kalbitz, K., Solinger, S., Park, J.-H., Michalzik, B., Matzner, E., Controls on the dynamics of dissolved organic matter in soils: a review. Soil Science 165, Lopez, R., Miano, T.M., Rock fragmentation processes in soils of Mediterranean natural areas. In: Rubio, J.L., et al. (Eds.), Abstracts Book, 3rd International Congress European Society for Soil Conservation, Grupo Carduche, Valencia (Spain), 121. McKnight, D.M., Bencala, K.E., Zellweger, G.W., Aiken, G.R., Feder, G.L., Thorn, K.A., Sorption of dissolved organic carbon by hydrous aluminium and iron oxides occurring at the confluence of Deer Creek with the Snake river, Summit County, Colorado. Environmental Science and Technology 26, Miano, T.M., Senesi, N., Malcolm, R.L., Chemical and spectroscopic properties of river and marine fulvic acids and their metal complexes. In: Proceedings of the 3rd International Nordic Symposium on Humic Substances, Turku 3, Miano, T.M., Sposito, G., Martin, J.P., Fluorescence spectroscopy of humic substances. Soil Science Society of America Journal 52, Nelson, P.N., Dictor, M.C., Soulas, G., Availability of organic carbon in soluble and particle-size fractions from a soil profile. Soil Biology and Biochemistry 26, Qualls, R.G., Haines, B.L., Biodegradability of dissolved organic matter in forest through-fall, soil solution, and stream water. Soil Science Society of America Journal 56, Senesi, N., Miano, T.M., Provenzano, M.R., Fluorescence spectroscopy as a means of distinguishing fulvic and humic acids from dissolved and sedimentary acquatic sources and terrestrial sources. In: Humic Substances in the Acquatic and Terrestrial Environment. In: Allard, B., Boren, H., Grimwall, A. (Eds.), Lecture notes in Earth Sciences, 33, pp Shen, Y.H., Sorption of natural dissolved organic matter on soil. Chemosphere 38, Soil Survey Staff, Keys of Soil Taxonomy, eighth ed. USDA.

8 1590 M. Corvasce et al. / Chemosphere 62 (2006) Stevenson, F.J., Cycles of Soil. John Wiley and Sons, New York. Stevenson, F.J., Humus Chemistry: Genesis, Composition, Reactions. John Wiley and Sons, New York. Tate III, R.L., Soil Organic Matter: Biological and Ecological effects. Wiley-Interscience, New York. Wolfbeis, O.S., In: Molecular Luminescence Spectroscopy, Methods and Applications. Part 1. Wiley, New York, pp Zsolnay, A., Dissolved humus in soil waters. In: Piccolo, A. (Ed.), Humic Substances in Terrestrial Ecosystem. Elsevier, Amsterdam, pp Zsolnay, A., Dissolved organic matter: artefacts, definitions and function. Geoderma 113, Zsolnay, A., Baigar, E., Jimenez, M., Steinweg, B., Saccomandi, F., Differentiating with fluorescence spectroscopy the sources of dissolved organic matter in soils subjected to drying. Chemosphere 38,