- Email: [email protected]
The molecular properties of humic substances isolated from a UK upland peat system
Environment International 27 (2001) 449 – 462 www.elsevier.com/locate/envint
The molecular properties of humic substances isolated from a UK upland peat system A temporal investigation M.J. Scotta,1, M.N. Jonesa,*, C. Woof b,2, B. Simonb,2, E. Tippingb,2 a
School of Biological Sciences, Stopford Building, University of Manchester, Oxford Road, Manchester M13 9PL, UK b Institute of Freshwater Ecology, Ferry House, Ambleside, Cumbria LA22 0LP, UK Received 20 September 2000; accepted 26 July 2001
Abstract The study concerns the possible changes in the molecular characteristics of humic materials isolated from the same source as a function of time. A great deal of data has been reported concerning the contrast in molecular characteristics of humic substances isolated from different environments. This has primarily been an attempt to identify source-specific molecular characteristics. However, data presented in this paper suggests that humic substances isolated from a single catchment have significant changes in molecular characteristics over time. Two naturally occurring peat pools (X and Y) situated upon a small organic catchment on Great Dun Fell, Cumbria, UK were sampled monthly between November 1994 and November 1996. Dissolved organic matter (DOM) from the pool water samples was fractionated using macroporous nonionic resins (XAD8 and 4), and the humic, fulvic and hydrophilic acids were collected. These fractions were analysed for elemental composition (C, H and N), weight average molecular weight, functional group content and adsorption (340 nm) of a 1 g l 1 solution measured in a 1-cm spectrophotometer cell. The molecular characteristics were compared to those of natural DOM described by Scott et al. (1998). Scott et al. reported that drought conditions and seasonal climatic changes could have appreciable effects upon molecular characteristics of natural DOM. Results showed that the atomic H/C ratio of the humic substances increased immediately after strong drought conditions experienced in the summer of 1995. This change was temporary with atomic H/C ratio decreasing gradually over the following months. A similar decrease was observed in the carboxyl group content of the isolated compounds. The data set suggested that atomic H/C ratio in the fulvic and hydrophilic fractions exhibited seasonal characteristics of higher ratios during the late summer/early autumn months. This was not observed in the humic fraction. Humic acids exhibited a seasonal pattern of higher weight average molecular weight during the summer months. These trends were explained in terms of summer production of DOM in the catchment soils, their sequestering in the soil due to limited soil water movement during the summer months and their relative ease of dissolution when rainfall and soil water movement increased during the late summer/early autumn period. The results were found to support seasonal and long-term patterns observed in natural DOM as reported by Scott et al. (1998). D 2001 Elsevier Science Ltd. All rights reserved. Keywords: Dissolved organic matter (DOM); Humic substances; Fulvic acids; Hydrophilic acids; Great Dun Fell, Cumbria; Seasonal characteristics of DOM; DOM molecular characteristics
1. Introduction Dissolved organic matter (DOM) plays an integral biogeochemical role in the soil and aquatic environments.
* Corresponding author. Tel.: +44-161-275-5093; fax: +44-161-2755082. E-mail address: [email protected] (M.N. Jones). 1 Tel.: + 44-161-275-5093; fax: + 44-161-275-5082. 2 Tel.: + 44-15394-42468; fax: + 44-15394-46914.
pH buffering, nutrient cycling, metal leaching, mineral weathering and pollutant behavior and toxicity are all influenced by DOM (Hemond, 1980; Cronan and Aiken, 1985; Pohlman and McColl, 1988; Vance and David, 1992; Liechty et al., 1995). Peat systems being primarily organic matter are considered large sources of DOM (Urban et al., 1989). Estimates of DOM flux from peatlands include: 7 – 15 g C m 2 a 1 (Scott et al., 1998); 13 –26 g C m 2 a 1 (Tipping et al., 1998); and 6– 8.5 g C m 2 a 1 (Tegen and Dorr, 1996).
0160-4120/01/$ – see front matter D 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 1 6 0 - 4 1 2 0 ( 0 1 ) 0 0 1 0 0 - 3
450
M.J. Scott et al. / Environment International 27 (2001) 449–462
These authors reported that DOM flux exhibited a seasonal pattern of maximum flux during the late summer/early autumn period. The reasoning behind this observation was that DOM production is primarily microbially driven (McKnight et al., 1985; Scott et al., 1998; Tipping et al., 1998). Soil microbes are more active during the warm and relatively dry periods associated with the summer months (Anderson, 1973; Grieve, 1984; Batomalaque et al., 1992). It is therefore reasonable to assume that, during the summer, increased microbial degradation of organic matter lead to a rise in the production of potential DOM. Due to greater evaporation, limited rainfall and, subsequently, soil water movement, the release of potential DOM is restricted until the onset of the rainfall associated with the late summer/ early autumn period (Scott et al., 1998; Tipping et al., 1998). Evidence presented by Scott et al. (1998) suggested that the molecular characteristics of DOM could exhibit both seasonal and long-term trends. During a 4-year monitoring period, it was shown that DOM in the surface waters of a peat system became more hydrophilic during the summer months, and that DOM aromaticity changed notably after a prolonged drought period. It was suggested that the changes in hydrophobicity might have been a consequence of the ease of dissolution of DOM produced in the summer months, whereas changes in the aromaticity of DOM after a drought period may have been due to oxygenation of primarily anoxic peat layers. These data suggest that the molecular characteristics of DOM found in surface waters may be susceptible to changes in climatic conditions, both seasonally and more long term. Leenheer (1981) outlined a procedure for the isolation of individual components of DOM using a separation technique involving macroporous nonionic resins. Using Amberlite XAD8 and XAD4 resins, it is possible to separate DOM into five main components (Fig. 1). Humic, fulvic and hydrophilic acids constitute approximately 75% of DOM (Malcolm, 1985), and it is reasonable therefore to assume
that these compounds contribute significantly to the molecular characteristics of DOM, as a whole. It has been shown that the molecular characteristics of the constituent molecules of DOM can control its interactions in the environment. Changes in aromaticity, molecular weight, elemental composition and functional group content can alter solubility characteristics (Stevenson, 1985), buffering capacity (Wilson, 1979), metal ion/pesticide interactions (Thurman, 1985), light attenuation (Tipping et al., 1998) and bioavailability (Hessen, 1985). Such changes in molecular behaviour may have major implications for the soil and aquatic environments. Aiken (1997) showed that there is a strong correlation between molar absorptivities at 340 nm and the aromaticity, molecular weight and pyrene-binding coefficient of humic substances. The A340 g 1 C l 1 cm 1 has been found to have a significant controlling influence upon the reactivity of humic substances with metals, hydrophobic organic compounds and chemical oxidants such as chlorine. It is well documented that humic substances from different origins exhibit significant, almost characteristic variations in molecular make up. Elemental composition, functional group content, molecular weight and aromaticity have all been shown to differ in samples collected from contrasting environments (Rice and MacCarthy, 1991; Thurman, 1985; Thurman et al., 1982; Meier et al., 1999, respectively). The aim of this study was to identify any seasonal or longer-term patterns evident in the molecular characteristics of DOM constituents isolated from a small peat and ranker catchment situated upon Great Dun Fell, Cumbria, UK. A previously reported work (Scott et al., 1998) identified patterns in the molecular characteristics of DOM at this site. The primary aim of this study was to identify if corresponding molecular changes are evident in the humic materials isolated from the same DOM. As far as the authors are aware, this is the first study of its kind.
Fig. 1. Constituents of DOM isolated. Average taken from a number of North American streams (Malcolm, 1985).
M.J. Scott et al. / Environment International 27 (2001) 449–462
2. Methods Samples were collected from the Great Dun Fell site at monthly intervals (weather permitting) between November 1994 and November 1996. Water samples were taken from two naturally occurring peat pools named X and Y. The pools (approximate volume 1 m3, depth 30 cm) are situated on a small catchment (400 m2), which consists of a 250-m2 slope draining into a 150-m2 plateau region. Rainfall is the sole source of water input into the catchment. The pools receive water only from direct soil water flow and are therefore considered natural soil water samplers. An automatic weather station was installed close to the pool site and recorded average daily air temperatures and daily rainfall amounts. Samples from X and Y were taken using a 12-V peristaltic pump, with the inlet pipe held 2 – 3 cm below the surface to avoid contamination from the surface or sediment. On each occasion, 25-l samples were collected from both pools X and Y, from which fractions were isolated. Aquatic humic substances were isolated according to the method outlined by Leenheer (1981). The nonionic macroporous resins used were Amberlite XAD8 and XAD4. To minimise bleed, resins were Soxhlet-washed twice for 24 h with methanol. Only the humic, fulvic and hydrophilic acid fractions were obtained. Isolated samples were then analysed for weight average molecular weight, functional group content, elemental composition and absorbance (340 nm) of a 1 g l 1 solution measured in a 1-cm spectrophotometer cell (A340 g 1 C l 1 cm 1). Twenty-five-litre water samples from X and Y were taken for isolation. The samples were acidified with HCl to pH 2, filtered through a Whatman GF/F filter and then through a 0.45-mm Millipore HA filter to remove the particulate organic matter (POM) fraction. The sample was then passed through a 400-ml Amberlite XAD8 column, at a flow rate of approximately 15 ml min 1. The column was back-eluted using 250 ml 1 M NaOH to remove the adsorbed hydrophobic fraction. This was then collected and, using HCl, immediately acidified to pH 1, to avoid oxidation of the organic compounds and to separate the hydrophobic fraction into the fulvic and humic fractions. At pH 1, humic acids precipitate from the solution, whereas the fulvic acids remain in the solution. The extract was transferred into a 380-ml Beckman centrifuge tube and spun at 10,000 rpm for 60 min in a Beckman J2-21 centrifuge equipped with a JA10 rotor head. The supernatant (fulvic acids) was separated from the sediment (humic acids). The humic acids were transferred into 1-cm-diameter cellulose dialysis membrane (Medicell International) and dialysed for 72 h against stirred deionised water to remove the NaCl. The fulvic acids were readsorbed onto the XAD8 column at a flow rate of approximately 10 ml min 1. Deionised water was back-eluted through the column and the effluent was tested for chloride ions using silver nitrate solution. Once back-eluted, the fulvic acids
451
were immediately passed through a 100-ml AG-MP-50 column, which removed the Na + , and lowered the pH to avoid oxidation. The effluent water from the XAD8 resin was then passed through a 400-ml XAD4 column at a flow rate of 15 ml min 1 to remove the hydrophilic acids. Once complete, the column was then back-eluted with deionised water to remove chloride ions. Once desalinated, the deionised water was replaced with 250 ml 1 M NaOH and the subsequent effluent was collected and run through a 100-ml AG-MP-50 column. The samples collected (humic acids, fulvic acids and hydrophilic acids) were individually rotary evaporated at 30 C under reduced pressure until only 25 ml remained, then transferred to 100-ml vials, frozen at 30 C and freeze-dried. All samples were protected from light at every available opportunity in order to minimise photochemical UV oxidation during this procedure (Allard et al., 1994). Once isolated, the humic substances were characterised for elemental composition (C, H and N content), weight average
Fig. 2. Average daily air temperature (Panel 1) and total monthly rainfall (Panel 2) for Great Dun Fell.
452
M.J. Scott et al. / Environment International 27 (2001) 449–462
Fig. 3. Weight average molecular weight data (Pool X – open symbols; Y— closed symbols) for humic, fulvic and hydrophilic acids (panels top, middle and bottom, respectively).
M.J. Scott et al. / Environment International 27 (2001) 449–462
molecular weight, aromaticity (A340 g 1 C l 1 cm 1) and functional group content. Carbon, hydrogen and nitrogen analysis was carried out using a Carlo-Erba model 1106 C, H, N and S analyser. Oxygen was calculated by difference. The weight average molecular weight was calculated using a sedimentation equilibrium method with data obtained from a Beckman Optima XL-A analytical ultracentrifuge. Small subsamples of isolated humic substances were dissolved into phosphate-buffered saline (PBS) until an optical density (OD) of 0.5 was observed on a LKB Ultraspec 4050 UV spectrophotometer (W/L 280 nm). Humic acids were spun at 12,000 rpm, fulvic acids at 20,000 rpm and hydrophilic acids at 23,000 rpm, because of their differences in molecular weight (lower molecular weight compounds need to be spun faster to attain equilibrium). The samples were centrifuged until sedimentation had come to equilibrium (usually within 16 h). Functional group content was measured using a direct acid titration method. The isolated humic substance was accurately weighed to four decimal places and then placed into 20 ml of nitrogen purged double distilled water. All experiments were carried out under nitrogen to exclude CO2, because of the buffering effect of dissolved HCO3 or CO32 . The sample was then placed into the titration vessel (a triple necked quickfit flask) along with a known volume of NaNO3 (to raise the ionic strength to I = 0.05). Nitrogen was passed continuously through the solution until the completion of the experiment. The sample was stirred continuously using a Chemlab SS3 stirrer and magnetic stirrer bar. To allow full degassing of CO2 from the solution the pH was lowered to 2 and left for 1 h. After the hour, the pH of the solution was then raised to 11 –12 using nitrogen-purged NaOH (prepared no earlier than 20 min before the experiment and tightly stoppered). All titrations were performed using neutralisation of charge with protons because of the difficulty of maintaining a standard CO2-free NaOH solution (Bryan, 1994). When the required pH was obtained, the sample was left to equilibrate for an hour. Then aliquots of either 1 M or 0.1 M HCl were titrated into the humic solution. After each addition, the solution was left to equilibrate before the pH reading was taken. The equilibration criterion was that the pH did not drift by > 0.1 pH units/min. The experimental blank was taken by a titration of 20 ml of NaNO3 solution by the procedure outlined above. The moles of acid added to the humic substance solution in order to change the pH from 11 to 7 and from 7 to 3 were calculated (DTG (titratable groups)7 – 11 and DTG3 – 7, respectively). Blank titration values were subtracted and the final result was given as milliequivalents per gram of isolate. Sample analysis involved the accurate weighing of approximately 0.005 g of humic substance. The sample was dissolved in 3 ml of PBS, filtered through a 0.45-mm polycarbonate filter and analysed using a Dohrmann DC190 high-temperature total organic carbon (TOC) analyser. The absorbance at 340 nm was quantified in a 1-cm quartz cell
453
and blanked against PBS. Measurements were taken in a LKB Ultraspec 4050 UV spectrophotometer.
3. Results Fig. 2 shows the average daily air temperature (Panel 1) and monthly rainfall amounts (Panel 2) measured close to the site of pools X and Y. The average daily air temperature exhibited a seasonal temperature fluctuation from 5 C in approximately January/February rising to 20 C in July/ August. Rainfall patterns were erratic with a tendency towards lower values during the summer months. It was noted that a prolonged period of reduced rainfall and higher air temperatures caused both pools to dry completely for approximately 2 months during the summer of 1995. A similar but somewhat shorter (1 month) period was observed in the summer of 1996. Fig. 3 shows the weight average molecular weight data for the humic (Panel 1), fulvic (Panel 2) and the hydrophilic acids (Panel 3). The weight average molecular weight of the isolated humic acids showed a seasonal trend, with the molecular weights increasing before and after the dry spells experienced in both the summers of 1995 and 1996. Molecular weight appeared to increase from approximately 8000 to 10,000 in the winter and spring months up to 14,000 to 16,000 before or after a period when pool X dried. However, for pool Y lower molecular weights ( 6000 –8000) were found after the drought. The fulvic and hydrophilic acid fractions indicated no obvious seasonal or long-term trend. Weight average molecular weight varied from between 1600 and 2800 and 1400 and 3500, respectively. Both X and Y samples showed a good correlation between one another, having comparable values and patterns of change. Table 1 shows the mean weight average molecular weight for the humic, fulvic and hydrophilic acids. From the data, it is evident that aquatic humic acids have higher weight average molecular weight than fulvic acids, which are, in turn, marginally heavier than hydrophilic acids. Fig. 4 shows the atomic hydrogen to carbon ratio for the humic, fulvic and hydrophilic acids (Panels 1, 2 and 3, respectively). It was evident that upon rewetting after the prolonged dry spell of 1995 the atomic H/C ratio was greater than the previous value. This trend was evident in both pools and all three isolated compounds. The fulvic and hydrophilic acids exhibited a seasonal trend of slightly higher atomic H/C ratio during the late summer/early autumn period. This trend was more prominent in the pool
Table 1 Mean weight average molecular weight Fraction
Pool X
Pool Y
Literature
Reference
Aquatic humic acid Fulvic acid Hydrophilic acid
9136 2105 1954
10033 2211 1625
6581 2816 411
Birkett, 1996 Jones et al., 1995 Aiken et al., 1992
454
M.J. Scott et al. / Environment International 27 (2001) 449–462
Fig. 4. Atomic H/C ratio for humic, fulvic and hydrophilic acids. Panels: top, middle and bottom, respectively.
Y samples. The mean atomic H/C ratio for the fulvic acids appeared substantially lower than that for the humic and hydrophilic fractions (Table 2). Fig. 5 shows the atomic nitrogen to carbon ratio for the humic, fulvic and hydrophilic acids (Panels 1, 2 and 3, respectively). The atomic N/C ratio exhibited little or no
seasonal or long-term variations in any of the isolated compounds or pools. The fulvic acids had substantially lower atomic N/C ratios than the humic and hydrophilic acids (Table 2). Fig. 6 shows the content of functional groups titratable between pH 3 and 7 (DTG3 – 7) for the humic, fulvic and
M.J. Scott et al. / Environment International 27 (2001) 449–462 Table 2 Mean elemental composition data Sample
Pool
C
H
N
H/C
N/C
Humic acid
X Y X Y X Y
45.6 45.3 49.7 47.66 44.6 45.3
4.40 4.30 4.05 4.06 4.29 4.42
1.95 1.95 0.91 1.09 2.21 1.84
1.1483 1.1296 0.9809 1.0369 1.1630 1.1693
0.0364 0.0366 0.0159 0.0194 0.0443 0.0350
Fulvic acid Hydrophilic acid
hydrophilic acids (Panels 1, 2 and 3, respectively). There appeared to be little evidence to support a seasonal pattern within the data set. However, it was evident that although there are some initial postdrought rises, all three isolated compounds exhibited a decline in the DTG3 – 7 content after the prolonged drought period of the summer of 1995. This decline continued until the dry period of July 1996. Humic acid DTG3 – 7 fell from approximately 2 meq gm 1 to approximately 0.7 meq gm 1. This trend was also exhibited in both the fulvic (3 –1.2 meq gm 1) and the hydrophilic acids (3– 1 meq gm 1). After the summer drought of 1996, the DTG3 – 7 content of the hydrophilic acid samples isolated from both pools rose sharply to approximately 3 meq gm 1. After which, the DTG3 – 7 declined rapidly to 1.3 meq gm 1 by November 1996. This trend was not observed in either the humic or fulvic acid isolates. The humic acid isolates exhibited substantially lower mean DTG3 – 7 than the fulvic and hydrophilic acids (Table 3). Fig. 7 shows the content of functional groups that titrate between pH 7– 11 (DTG7 – 11) for the humic, fulvic and hydrophilic acid isolates (Panel 1, 2 and 3, respectively). The data showed no evidence of a seasonal or long-term pattern. No appreciable difference in mean DTG7 – 11 was evident between the humic, fulvic and hydrophilic fractions (Table 4). Fig. 8 shows the A340 g 1 C l 1 cm 1 for the humic, fulvic and hydrophilic acids (Panels 1, 2 and 3, respectively). The A340 g 1 C l 1 cm 1 data indicated no obvious seasonal or long-term trend. The mean A340 g 1 C l 1 cm 1 value for the humic was appreciably higher than those for the fulvic and hydrophilic acids (Table 4).
4. Discussion 4.1. Seasonal and long-term patterns Scott et al. (1998) showed that the DOM present in the pool water appeared to undergo a considerable change in characteristic after the extended drought period of 1995. The optical absorbance (340 nm) g 1 C l 1 cm 1 is a method of assessing to what degree a compound is aromatic (Aiken, 1997). Assuming that the DOM present solely caused the A340 in the pool waters, then the organic compounds appeared to become 50% less coloured after the drought
455
of 1995. The DOM molecules appeared to have undergone a significant change in molecular structure. Strong water deficits associated with the droughts of 1992 and 1995 may have allowed the influx of oxygen into layers of peat that were normally anoxic, possibly causing a major change in the decomposition process. Evidence of an oxygen influx is apparent from the rise in sulphate concentration after both prolonged drought events. Evidence of a similar change was found in the elemental composition data of the DOM isolates. The humic, fulvic and hydrophilic fractions all exhibited an increase in the atomic H/C ratio immediately after the 1995 drought, when compared to the previous reading. The rise in the atomic H/C ratio may have been indicative of a reduction in the aromaticity of the compounds (Steelink, 1985). After the prolonged drought of 1995, both atomic H/C and DTG3 – 7 exhibited gradual declines. It appeared that the humic substances present in the soil water gradually became more aromatic with less carboxyl group content per unit weight. This may have been a consequence of oxygenated peat becoming gradually more anoxic after rewetting. The decomposition process may have changed, as oxygen became less available. Higher aromaticity and lower carboxyl content may be characteristics expected for humic substances originating from more anoxic environments. Alternatively, this observation may have been a consequence of ease of dissolution upon rehydration of less aromatic humic substances with higher carboxyl group content. Tipping and Woof (1990) showed that the release of humic materials into soil water from acid organic soils was integrally linked to molecular charge and functional group content. Hayes (1985) stated that aromaticity can play a role in humic dissolution properties. In fact, Y fulvic acid and both hydrophilic isolates exhibited a similar elevation in atomic H/C in the following summer. Humic substances produced during the summer months are effectively trapped in the catchment soil because of restricted rainfall and therefore soil water flow. During the late summer/early autumn period, rainfall becomes more frequent effectively flushing potential DOM from the catchment (Scott et al., 1998). When soil water flow resumes, more hydrophilic components of DOM will be preferentially dissolved. Less aromatic humic substances are more soluble and therefore their prevalence in soil water during the autumn flush is greater relative to the more aromatic and possibly aggregated materials. The authors acknowledge that the evidence to suggest that the aromaticity of the humic isolate changes after prolonged dry periods is somewhat questioned by the lack of a corresponding drop in observed A340 g 1 C l 1 cm 1. The weight average molecular weights of humic acids isolated from the pool waters appeared to exhibit seasonal variations. The molecular weight of material isolated from both X and Y increased during the summer, particularly just before and after periods when the pools were dry. This trend was not evident in either the fulvic or hydrophilic acid
456
M.J. Scott et al. / Environment International 27 (2001) 449–462
Fig. 5. Atomic N/C ratio for humic, fulvic and hydrophilic acids. Panels: top, middle and bottom, respectively.
fractions. The humic acid molecules themselves appeared to remain relatively consistent in terms of molecular characteristic. The rise in molecular weight was not accompanied by
a corresponding change in the elemental composition and A340 g 1 C l 1 cm 1, suggesting that there was no change in the level of aromaticity. From the evidence available it
M.J. Scott et al. / Environment International 27 (2001) 449–462
Fig. 6. DTG3 – 7 (Carboxyl group meq g
1
457
) for the humic, fulvic and hydrophilic acids. Panels: top, middle and bottom, respectively.
was difficult to hypothesise potential reasons for such a change. Humic acids are known to complex with one another in the presence of certain metal ions and at higher concentrations (Reid et al., 1991). However, the isolation
procedure removes most metal ions, and all isolates were analysed for molecular weight at the same concentration and in a buffer solution to combat intermolecular complex formation. It appeared that there was a change in the
458
M.J. Scott et al. / Environment International 27 (2001) 449–462
Table 3 Mean functional group content DTG3 – 7 (meq g 1)
DTG7 – 11 (meq g 1)
DTG3 – 11 (meq g 1)
Sample
X
Y
X
Y
X
Y
Humic acid Fulvic acid Hydrophilic acid
1.1 1.9 2.1
1.3 1.9 2.2
1.2 1.2 1.1
1.5 1.3 1.4
2.3 3.1 3.2
2.8 3.2 3.6
WHAM humic WHAM fulvic
2.1 2.9
1.4 1.3
3.5 4.2
production process of humic acids during the summer months. Warmer, drier conditions appeared to alter the chemical and biochemical processes, producing humic acids that were significantly higher in weight average molecular weight. Unfortunately, due to the relatively short monitoring period and the absence of information available on the formation of humic substances it was difficult to make more than an observation of this trend. It has long been acknowledged that humic substances isolated from contrasting environments have significantly different, almost characteristic molecular make-ups. However, the authors are not aware of a comparable study to this one outlining that humic substances isolated at different dates from the same source can have considerable variation in their molecular characteristics. Table 5 illustrates the variation observed in each molecular characteristics exhibited by the humic substances over the monitoring period. Weight average molecular weight exhibits a two- to threefold variation and a fivefold variation is observed in A340 g 1 C l 1 cm 1 during the monitoring period. There are also significant variations in H/C and N/C ratios and DTG7 – 11 and DTG3 – 7. Such changes indicate that DOM production may be somewhat dynamic. A complex interaction of different processes may be constantly changing the production of DOM. Processes such as microbial degradation, chemical degradation, solute dissolution, etc. may all be subtly influenced by external factors such as temperature, oxygen availability, moisture content, etc. Changes in molecular characteristics may also be an artefact of the isolation procedure. As no one isolation is exactly identical to its predecessor subtle changes from one procedure to the next may influence the type of DOM adsorbed to the XAD resins or may influence the amount of resin leached into the humic substance during the isolation procedure. 4.2. Comparison of the molecular characteristics of the DOM isolates The fulvic acid fractions had appreciably lower atomic H/C and N/C ratios than the humic and hydrophilic fractions. Closer examination of the elemental composition showed the fulvic fractions contained less hydrogen and nitrogen, and more carbon than the humic and hydrophilic
acids (Table 2). Comparison of the data with those published by Rice and MacCarthy (1991) shows the fractions to be of similar elemental composition to those isolated elsewhere. However, the Great Dun Fell data set was unusual in that the fulvic acid fraction had a lower mean atomic H/C ratio than the humic acid, the reverse is usually the case. Though unexpected, the data for the humic and fulvic acids fell well within the range reported by Rice and MacCarthy. Evidence suggested that aquatic humic acids are considerably more similar to fulvic acid than their soil borne counterparts in terms of molecular weight (Wershaw and Aiken, 1985), elemental composition (Steelink, 1985; Rice and MacCarthy, 1991) and functional group content (Perdue, 1985). This suggested that this observation may just have been a characteristic of this system. Presently, the literature is very limited on examples of elemental analysis performed upon the hydrophilic fractions. McKnight et al. (1985) analysed the hydrophilic fraction of DOM taken from Thoreau’s Bog, MA, and reported elemental compositions of 44.7% carbon, 3.6% hydrogen and 1.27% nitrogen. McKnight’s results were closely comparable to those reported in this study. The weight average molecular weight data for the humic and fulvic acids compared well with the values contained within the published literature (Table 1). The large contrast with the hydrophilic fractions was probably due to differences in the method used for analysis. Aiken et al. (1992) measured number average molecular weights which are lower than weight average molecular weights for polydisperse samples such as hydrophilic acids (Wershaw and Aiken, 1985). The results of mean A340 g 1 C l 1 cm 1 indicated that the humic acid fraction had more colour per unit weight than the fulvic or the hydrophilic fractions. These data were consistent with the idea that humic acids are more aromatic than fulvic acids (Stevenson, 1985). It appeared from the data that fulvic acids were more aromatic than the hydrophilic acids, which would agree with the preliminary findings of characterisations carried out by Malcolm (1985). Most data in the literature for direct titrations carried out upon humic substances did not report measuremen ts between pH 3 and 11. This made it difficult to find measurements that were comparable to the results obtained in this study. However, it was possible to use the Windermere Humic Aqueous Model (WHAM) (Tipping, 1994) to calculate the change in charge between pH 3 – 7 and 7– 11 for both humic and fulvic acids. The model parameters were averages derived from titrations performed upon eight fulvic acids (water and soil origin) and six humic acids (soil origin). The model produced values of 2.9 meq g 1 for a fulvic acid and 2.1 meq g 1 for a humic acid which appear slightly higher than the mean DTG3 – 7 values for the pool isolates (Table 3). However, the results were consistent with the idea that fulvic acid contains more carboxyl groups than humic acid (Stevenson, 1985; Mathur and Farnham, 1985). The content of functional groups that could be titrated between pH 7 and 11 (DTG7 – 11) appeared to be of a similar
M.J. Scott et al. / Environment International 27 (2001) 449–462
Fig. 7. DTG7 – 11 (Phenol group meq g
1
459
) for the humic, fulvic and hydrophilic acids. Panels: top, middle and bottom, respectively.
value in each of the individual fractions (Table 3). The mean values varied from 1.1 to 1.5 meq g 1. Humic acid had slightly higher DTG7 – 11 than fulvic acid. Using WHAM, the calculated changes in charge between pH 7 and 11 were 1.28 meq g 1 for fulvic acid and 1.37 meq g 1 for humic acid. The average DTG7 – 11 for the humic and fulvic
fractions agree closely with those generated by the WHAM model. The model also supports the observation that the humic fractions have slightly higher DTG7 – 11 than the fulvic fractions. Unfortunately, WHAM does not consider hydrophilic acid and therefore no comparison could be made.
460 Table 4 Mean A340 g C
M.J. Scott et al. / Environment International 27 (2001) 449–462
1
l
1
cm
4.3. Summary
1
Fraction
Pool X
Pool Y
Aquatic humic acid Fulvic acid Hydrophilic acid
11.24 10.53 6.83
11.86 10.66 5.26
Fig. 8. A340 g C
1
l
1
cm
1
The observations made in this paper support the ideas put forward by Scott et al. (1998). Scott et al. showed that long-term wetting and drying patterns, and seasonal temperature variations and rainfall patterns, affected soil acid-
for the humic, fulvic and hydrophilic acids. Panels: top, middle and bottom, respectively.
M.J. Scott et al. / Environment International 27 (2001) 449–462 Table 5 Intra isolate molecular characteristic range
References
Characterisation
Humic
Fulvic
Hydrophilic
H/C ratio N/C ratio Mw DTG3 – 7 DTG7 – 11 A340 g C
0.93 – 1.74 0.005 – 0.08 6,000 – 18,000 0.2 – 2.6 0.4 – 3.0 1.4 – 7.0
0.82 – 1.25 0.009 – 0.05 1,500 – 2,750 1.0 – 3.0 0.9 – 2.6 1.0 – 4.5
0.85 – 1.6 0.017 – 0.138 1,300 – 3,300 1.0 – 3.4 0.2 – 2.9 0.8 – 3.8
1
l
1
cm
1
461
ity, sulphate dynamics, DOM production, its flux and molecular characteristics. Humic substances extracted from the same samples as Scott et al. (1998) showed comparable (if not as clear) changes in molecular characteristics. Elemental composition and carboxyl group content exhibited associations with long-term drying patterns. This corroborated well with changes in DOM A340 g C 1 l 1 cm 1 and hydrophilic DOM to total DOM ratio. However, changes in DOM A340 g C 1 l 1 cm 1 were not mirrored by isolate A340 g C 1 l 1 cm 1. Such contradictions may have been a consequence of the isolation procedure. Excessive exposure to alkaline solutions could have caused oxidation of humic substances (Hayes, 1985). The isolation procedure itself is selective to molecules with certain properties. The acidification of samples precolumn may have caused some DOM to precipitate, thus losing less soluble components. Characteristics such as lower functional group contents (Tipping and Woof, 1990), higher molecular weights (Leenheer, 1985) and higher aromaticity (Stevenson, 1985) can reduce humic substance solubility. Adsorption onto the XAD resin is by its nature selective (Leenheer, 1985). XAD8 would selectively adsorb hydrophobic compounds (Aiken, 1985), thus selecting molecules that had higher molecular weights, higher A340 g C 1 l 1 cm 1, lower atomic H/C and lower functional group content. These explanations may have played roles in the difference in molecular characteristics of natural DOM and its corresponding isolated humic substances. Numerous authors have used XAD resins to isolate humic substances from DOM originating from different environments, and found the technique capable of identifying ascertainable differences in molecular characteristics. However, it may have been that changes caused by climatic fluctuations in DOM originating from the same environment may have been too subtle to not be masked by the selective nature of the isolation procedure.
Acknowledgments We would like to Mr K. Taylor and Mr M. Garnett (ITE, Merlewood) for the meteorological data. English Nature for allowing access to the Moor House Nature Reserve. Finally, we would like to thank the Natural Environment Research Council for funding the CASE studentship to MJS.
Aiken GR. Isolation and concentration techniques for aquatic humic. In: Aiken GR, McKnight DM, Wershaw RL, MacCarthy P, editors. Humic substances in soils, sediment and water. New York: Wiley-Interscience, 1985. pp. 363 – 86. Aiken GR. Factors controlling the nature and reactivity of organic matter in aquatic systems. Abstract: Humus. Nordic Humus Newsletter 1997;4(1) 8. 6th Nordic Symposium on Humic Substances Humic Substances as an Environmental Factor, Hameenlinna, Finland. Aiken GR, McKnight DM, Thorn KA, Thurman EM. Isolation of hydrophilic organic acids from water using nonionic macroporous resins. Org Geochem 1992;18(4):567 – 73. Allard B, Boren H, Petterson C, Zang G. Degradation of humic substances by UV radiation. Environ Int 1994;20(1):97 – 101. Anderson JM. Carbon dioxide evolution from two temperate, deciduous woodland soils. J Appl Ecol 1973;10:361 – 78. Batomalaque AE, Kikuma M, Seki H. Population dynamics of attached bacteria in a mesotrophic swampy bog of Japan. Water, Air, Soil Pollut 1992;63:371 – 8. Birkett JW. The physical characterisation of humic substances from peat and natural waters. PhD Thesis, University of Manchester, 1996. Bryan ND. Metal uptake by humic substances: a combined experimental and modelling study. PhD Thesis, University of Manchester, Manchester, 1994. Cronan CS, Aiken GR. Chemistry and transport of soluble humic substances in forested watersheds of the Adirondack Park, New York. Geochim Cosmochim Acta 1985;49(8):1697 – 705. Grieve IC. Concentrations and annual loading of dissolved organic matter in a small moorland stream. Freshwater Biol 1984;14:533 – 7. Hayes M. Extraction of humic substances from soil. In: Aiken GR, McKnight DM, Wershaw RL, MacCarthy P, editors. Humic substances in soils, sediment and water. New York: Wiley-Interscience, 1985. pp. 329 – 62. Hemond HF. Biogeochemistry of Thoreau’s Bog, Concord, Massachusetts. Ecol Monogr 1980;50:507 – 26. Hessen DO. The relationship between bacterial carbon and dissolved humic compounds in oligotrophic lakes. FEMS Microbiol Ecol 1985;31(4): 215 – 24. Jones MN, Birkett JW, Wilkinson AE, Hesketh N, Livens FR, Bryan ND, Lead JR, Hamilton-Taylor J, Tipping E. Experimental determinations of partial specific volumes of humic substances in aqueous solutions. Anal Chim Acta 1995;314:149 – 59. Leenheer JA. A comprehensive approach to preparative isolation and fractionation of dissolved organic carbon from natural waters and waste waters. Environ Sci Technol 1981;15(5):578 – 87. Leenheer JA. Fractionation techniques for aquatic humic substances. In: Aiken GR, McKnight DM, Wershaw RL, MacCarthy P, editors. Humic substances in soils, sediment and water. New York: Wiley-Interscience, 1985. pp. 409 – 32. Liechty HO, Kuuseoks E, Mroz GD. Dissolved organic carbon in northern hardwood stands with differing acidic inputs and temperature regimes. J Environ Qual 1995;24(5):927 – 33. Malcolm RL. Geochemistry of stream fulvic and humic substances. In: Aiken GR, McKnight DM, Wershaw RL, MacCarthy P, editors. Humic substances in soil, sediment and water. Geochemistry, isolation and characterisation. New York: Wiley-Interscience, 1985. pp. 181 – 210. Mathur SP, Farnham RS. Geochemistry of humic substances in natural and cultivated peatlands. In: Aiken GR, McKnight DM, Wershaw RL, MacCarthy P, editors. Humic substances in soil, sediment and water. Geochemistry, isolation and characterisation. New York: Wiley, 1985. pp. 53 – 86. McKnight D, Thurman EM, Wershaw RL, Hemond H. Biochemistry of aquatic humic substances in Thoreau’s Bog, Concord, Massachusetss. Ecology 1985;66(4):1339 – 52. Meier M, Namjesnik Dejanovic K, Maurice PA, Chin YP, Aiken GR.
462
M.J. Scott et al. / Environment International 27 (2001) 449–462
Fractionation of aquatic natural organic matter upon sorption to geothite and kaolinite. Chem Geol 1999;157(3 – 4):275 – 84. Perdue EM. Acidic functional groups of humic substances. In: Aiken GR, McKnight DM, Wershaw RL, MacCarthy P, editors. Humic substances in soil sediment and water. Geochemistry, isolation and characterisation. New York: Wiley-Interscience, 1985. pp. 493 – 526. Pohlman AA, McColl JG. Soluble organics from forest litter and their role in metal dissolution. Soil Sci Soc Am J 1988;52(1):265 – 71. Reid PM, Wilkinson AE, Tipping E, Jones MN. Aggregation of humic substances in aqueous media as determined by light-scattering methods. J Soil Sci 1991;42:259 – 70. Rice JA, MacCarthy P. Statistical evaluation of the elemental composition of humic substances. Org Geochem 1991;17(5):635 – 48. Scott MJ, Woof C, Jones MN, Tipping E. Concentrations and fluxes of dissolved organic carbon in drainage water from an upland peat system. Environ Int 1998;24(5 – 6):537 – 546. Steelink C. Implications of elemental analysis of humic substances. In: Aiken GR, McKnight DM, Wershaw RL, MacCarthy P, editors. 1st ed. Humic substances in soil, sediment and water. Geochemistry, isolation and characterisation, vol 1. New York: Wiley, 1985. pp. 457 – 76. Stevenson FJ. Geochemistry of soil humic substances. In: Aiken GR, McKnight DM, Wershaw RL, MacCarthy P, editors. Humic substances in soil, sediment and water. New York: Wiley-Interscience, 1985. pp. 13 – 52. Tegen I, Dorr H. Mobilization of caesium in organic rich soils: correlation with production of dissolved organic carbon. Water, Air, Soil Pollut 1996;88:133 – 44.
Thurman EM, editor. Organic geochemistry of natural waters. Boston: Niyhoff/Junk, 1985. Thurman EM, Wershaw RL, Malcolm RL, Pinckney DJ. Molecular sizes of aquatic humic substances. Org Geochem 1982;4:27 – 35. Tipping E. WHAM—a chemical equilibrium model and computer code for waters, sediments and soils incorporating a discrete-site/electrostatic model of ion binding by humic substances. Comp Geosci 1994;20: 973 – 1023. Tipping E, Woof C. Humic substances in acid organic soils—modeling their release to the soil solution in terms of humic charge. J Soil Sci 1990;41(4):573 – 86. Tipping E, Woof C, Ineson P, Taylor K, Bol R, Harkness DD, Benham D, Poskitt J, Rigg E. Climatic influences on the leaching of dissolved organic matter from upland UK moorland soils, investigated by a field manipulation experiment. Environ Int 1998;25(1):83 – 95. Urban NR, Bayley SE, Eisenreich SJ. Export of dissolved organic carbon and acidity from peatlands. Water Resour Res 1989;25(7):1619 – 28. Vance GF, David MB. Dissolved organic carbon and sulphate sorption by spodsol mineral horizons. Soil Sci 1992;154(2):136 – 44. Wershaw RL, Aiken GR. Molecular size and weight measurements of humic substances. In: Aiken GR, McKnight DM, Wershaw RL, MacCarthy P, editors. Humic substances in soil sediment and water. Geochemistry, isolation and characterisation. New York: Wiley, 1985. pp. 477 – 92. Wilson DE. The influence of humic compounds on titrimetric determinations of total organic carbon in freshwater. Arch Hydrobiol 1979;87(3): 379 – 84.
The molecular properties of humic substances isolated from a UK upland peat system A temporal investigation M.J. Scotta,1, M.N. Jonesa,*, C. Woof b,2, B. Simonb,2, E. Tippingb,2 a
School of Biological Sciences, Stopford Building, University of Manchester, Oxford Road, Manchester M13 9PL, UK b Institute of Freshwater Ecology, Ferry House, Ambleside, Cumbria LA22 0LP, UK Received 20 September 2000; accepted 26 July 2001
Abstract The study concerns the possible changes in the molecular characteristics of humic materials isolated from the same source as a function of time. A great deal of data has been reported concerning the contrast in molecular characteristics of humic substances isolated from different environments. This has primarily been an attempt to identify source-specific molecular characteristics. However, data presented in this paper suggests that humic substances isolated from a single catchment have significant changes in molecular characteristics over time. Two naturally occurring peat pools (X and Y) situated upon a small organic catchment on Great Dun Fell, Cumbria, UK were sampled monthly between November 1994 and November 1996. Dissolved organic matter (DOM) from the pool water samples was fractionated using macroporous nonionic resins (XAD8 and 4), and the humic, fulvic and hydrophilic acids were collected. These fractions were analysed for elemental composition (C, H and N), weight average molecular weight, functional group content and adsorption (340 nm) of a 1 g l 1 solution measured in a 1-cm spectrophotometer cell. The molecular characteristics were compared to those of natural DOM described by Scott et al. (1998). Scott et al. reported that drought conditions and seasonal climatic changes could have appreciable effects upon molecular characteristics of natural DOM. Results showed that the atomic H/C ratio of the humic substances increased immediately after strong drought conditions experienced in the summer of 1995. This change was temporary with atomic H/C ratio decreasing gradually over the following months. A similar decrease was observed in the carboxyl group content of the isolated compounds. The data set suggested that atomic H/C ratio in the fulvic and hydrophilic fractions exhibited seasonal characteristics of higher ratios during the late summer/early autumn months. This was not observed in the humic fraction. Humic acids exhibited a seasonal pattern of higher weight average molecular weight during the summer months. These trends were explained in terms of summer production of DOM in the catchment soils, their sequestering in the soil due to limited soil water movement during the summer months and their relative ease of dissolution when rainfall and soil water movement increased during the late summer/early autumn period. The results were found to support seasonal and long-term patterns observed in natural DOM as reported by Scott et al. (1998). D 2001 Elsevier Science Ltd. All rights reserved. Keywords: Dissolved organic matter (DOM); Humic substances; Fulvic acids; Hydrophilic acids; Great Dun Fell, Cumbria; Seasonal characteristics of DOM; DOM molecular characteristics
1. Introduction Dissolved organic matter (DOM) plays an integral biogeochemical role in the soil and aquatic environments.
* Corresponding author. Tel.: +44-161-275-5093; fax: +44-161-2755082. E-mail address: [email protected] (M.N. Jones). 1 Tel.: + 44-161-275-5093; fax: + 44-161-275-5082. 2 Tel.: + 44-15394-42468; fax: + 44-15394-46914.
pH buffering, nutrient cycling, metal leaching, mineral weathering and pollutant behavior and toxicity are all influenced by DOM (Hemond, 1980; Cronan and Aiken, 1985; Pohlman and McColl, 1988; Vance and David, 1992; Liechty et al., 1995). Peat systems being primarily organic matter are considered large sources of DOM (Urban et al., 1989). Estimates of DOM flux from peatlands include: 7 – 15 g C m 2 a 1 (Scott et al., 1998); 13 –26 g C m 2 a 1 (Tipping et al., 1998); and 6– 8.5 g C m 2 a 1 (Tegen and Dorr, 1996).
0160-4120/01/$ – see front matter D 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 1 6 0 - 4 1 2 0 ( 0 1 ) 0 0 1 0 0 - 3
450
M.J. Scott et al. / Environment International 27 (2001) 449–462
These authors reported that DOM flux exhibited a seasonal pattern of maximum flux during the late summer/early autumn period. The reasoning behind this observation was that DOM production is primarily microbially driven (McKnight et al., 1985; Scott et al., 1998; Tipping et al., 1998). Soil microbes are more active during the warm and relatively dry periods associated with the summer months (Anderson, 1973; Grieve, 1984; Batomalaque et al., 1992). It is therefore reasonable to assume that, during the summer, increased microbial degradation of organic matter lead to a rise in the production of potential DOM. Due to greater evaporation, limited rainfall and, subsequently, soil water movement, the release of potential DOM is restricted until the onset of the rainfall associated with the late summer/ early autumn period (Scott et al., 1998; Tipping et al., 1998). Evidence presented by Scott et al. (1998) suggested that the molecular characteristics of DOM could exhibit both seasonal and long-term trends. During a 4-year monitoring period, it was shown that DOM in the surface waters of a peat system became more hydrophilic during the summer months, and that DOM aromaticity changed notably after a prolonged drought period. It was suggested that the changes in hydrophobicity might have been a consequence of the ease of dissolution of DOM produced in the summer months, whereas changes in the aromaticity of DOM after a drought period may have been due to oxygenation of primarily anoxic peat layers. These data suggest that the molecular characteristics of DOM found in surface waters may be susceptible to changes in climatic conditions, both seasonally and more long term. Leenheer (1981) outlined a procedure for the isolation of individual components of DOM using a separation technique involving macroporous nonionic resins. Using Amberlite XAD8 and XAD4 resins, it is possible to separate DOM into five main components (Fig. 1). Humic, fulvic and hydrophilic acids constitute approximately 75% of DOM (Malcolm, 1985), and it is reasonable therefore to assume
that these compounds contribute significantly to the molecular characteristics of DOM, as a whole. It has been shown that the molecular characteristics of the constituent molecules of DOM can control its interactions in the environment. Changes in aromaticity, molecular weight, elemental composition and functional group content can alter solubility characteristics (Stevenson, 1985), buffering capacity (Wilson, 1979), metal ion/pesticide interactions (Thurman, 1985), light attenuation (Tipping et al., 1998) and bioavailability (Hessen, 1985). Such changes in molecular behaviour may have major implications for the soil and aquatic environments. Aiken (1997) showed that there is a strong correlation between molar absorptivities at 340 nm and the aromaticity, molecular weight and pyrene-binding coefficient of humic substances. The A340 g 1 C l 1 cm 1 has been found to have a significant controlling influence upon the reactivity of humic substances with metals, hydrophobic organic compounds and chemical oxidants such as chlorine. It is well documented that humic substances from different origins exhibit significant, almost characteristic variations in molecular make up. Elemental composition, functional group content, molecular weight and aromaticity have all been shown to differ in samples collected from contrasting environments (Rice and MacCarthy, 1991; Thurman, 1985; Thurman et al., 1982; Meier et al., 1999, respectively). The aim of this study was to identify any seasonal or longer-term patterns evident in the molecular characteristics of DOM constituents isolated from a small peat and ranker catchment situated upon Great Dun Fell, Cumbria, UK. A previously reported work (Scott et al., 1998) identified patterns in the molecular characteristics of DOM at this site. The primary aim of this study was to identify if corresponding molecular changes are evident in the humic materials isolated from the same DOM. As far as the authors are aware, this is the first study of its kind.
Fig. 1. Constituents of DOM isolated. Average taken from a number of North American streams (Malcolm, 1985).
M.J. Scott et al. / Environment International 27 (2001) 449–462
2. Methods Samples were collected from the Great Dun Fell site at monthly intervals (weather permitting) between November 1994 and November 1996. Water samples were taken from two naturally occurring peat pools named X and Y. The pools (approximate volume 1 m3, depth 30 cm) are situated on a small catchment (400 m2), which consists of a 250-m2 slope draining into a 150-m2 plateau region. Rainfall is the sole source of water input into the catchment. The pools receive water only from direct soil water flow and are therefore considered natural soil water samplers. An automatic weather station was installed close to the pool site and recorded average daily air temperatures and daily rainfall amounts. Samples from X and Y were taken using a 12-V peristaltic pump, with the inlet pipe held 2 – 3 cm below the surface to avoid contamination from the surface or sediment. On each occasion, 25-l samples were collected from both pools X and Y, from which fractions were isolated. Aquatic humic substances were isolated according to the method outlined by Leenheer (1981). The nonionic macroporous resins used were Amberlite XAD8 and XAD4. To minimise bleed, resins were Soxhlet-washed twice for 24 h with methanol. Only the humic, fulvic and hydrophilic acid fractions were obtained. Isolated samples were then analysed for weight average molecular weight, functional group content, elemental composition and absorbance (340 nm) of a 1 g l 1 solution measured in a 1-cm spectrophotometer cell (A340 g 1 C l 1 cm 1). Twenty-five-litre water samples from X and Y were taken for isolation. The samples were acidified with HCl to pH 2, filtered through a Whatman GF/F filter and then through a 0.45-mm Millipore HA filter to remove the particulate organic matter (POM) fraction. The sample was then passed through a 400-ml Amberlite XAD8 column, at a flow rate of approximately 15 ml min 1. The column was back-eluted using 250 ml 1 M NaOH to remove the adsorbed hydrophobic fraction. This was then collected and, using HCl, immediately acidified to pH 1, to avoid oxidation of the organic compounds and to separate the hydrophobic fraction into the fulvic and humic fractions. At pH 1, humic acids precipitate from the solution, whereas the fulvic acids remain in the solution. The extract was transferred into a 380-ml Beckman centrifuge tube and spun at 10,000 rpm for 60 min in a Beckman J2-21 centrifuge equipped with a JA10 rotor head. The supernatant (fulvic acids) was separated from the sediment (humic acids). The humic acids were transferred into 1-cm-diameter cellulose dialysis membrane (Medicell International) and dialysed for 72 h against stirred deionised water to remove the NaCl. The fulvic acids were readsorbed onto the XAD8 column at a flow rate of approximately 10 ml min 1. Deionised water was back-eluted through the column and the effluent was tested for chloride ions using silver nitrate solution. Once back-eluted, the fulvic acids
451
were immediately passed through a 100-ml AG-MP-50 column, which removed the Na + , and lowered the pH to avoid oxidation. The effluent water from the XAD8 resin was then passed through a 400-ml XAD4 column at a flow rate of 15 ml min 1 to remove the hydrophilic acids. Once complete, the column was then back-eluted with deionised water to remove chloride ions. Once desalinated, the deionised water was replaced with 250 ml 1 M NaOH and the subsequent effluent was collected and run through a 100-ml AG-MP-50 column. The samples collected (humic acids, fulvic acids and hydrophilic acids) were individually rotary evaporated at 30 C under reduced pressure until only 25 ml remained, then transferred to 100-ml vials, frozen at 30 C and freeze-dried. All samples were protected from light at every available opportunity in order to minimise photochemical UV oxidation during this procedure (Allard et al., 1994). Once isolated, the humic substances were characterised for elemental composition (C, H and N content), weight average
Fig. 2. Average daily air temperature (Panel 1) and total monthly rainfall (Panel 2) for Great Dun Fell.
452
M.J. Scott et al. / Environment International 27 (2001) 449–462
Fig. 3. Weight average molecular weight data (Pool X – open symbols; Y— closed symbols) for humic, fulvic and hydrophilic acids (panels top, middle and bottom, respectively).
M.J. Scott et al. / Environment International 27 (2001) 449–462
molecular weight, aromaticity (A340 g 1 C l 1 cm 1) and functional group content. Carbon, hydrogen and nitrogen analysis was carried out using a Carlo-Erba model 1106 C, H, N and S analyser. Oxygen was calculated by difference. The weight average molecular weight was calculated using a sedimentation equilibrium method with data obtained from a Beckman Optima XL-A analytical ultracentrifuge. Small subsamples of isolated humic substances were dissolved into phosphate-buffered saline (PBS) until an optical density (OD) of 0.5 was observed on a LKB Ultraspec 4050 UV spectrophotometer (W/L 280 nm). Humic acids were spun at 12,000 rpm, fulvic acids at 20,000 rpm and hydrophilic acids at 23,000 rpm, because of their differences in molecular weight (lower molecular weight compounds need to be spun faster to attain equilibrium). The samples were centrifuged until sedimentation had come to equilibrium (usually within 16 h). Functional group content was measured using a direct acid titration method. The isolated humic substance was accurately weighed to four decimal places and then placed into 20 ml of nitrogen purged double distilled water. All experiments were carried out under nitrogen to exclude CO2, because of the buffering effect of dissolved HCO3 or CO32 . The sample was then placed into the titration vessel (a triple necked quickfit flask) along with a known volume of NaNO3 (to raise the ionic strength to I = 0.05). Nitrogen was passed continuously through the solution until the completion of the experiment. The sample was stirred continuously using a Chemlab SS3 stirrer and magnetic stirrer bar. To allow full degassing of CO2 from the solution the pH was lowered to 2 and left for 1 h. After the hour, the pH of the solution was then raised to 11 –12 using nitrogen-purged NaOH (prepared no earlier than 20 min before the experiment and tightly stoppered). All titrations were performed using neutralisation of charge with protons because of the difficulty of maintaining a standard CO2-free NaOH solution (Bryan, 1994). When the required pH was obtained, the sample was left to equilibrate for an hour. Then aliquots of either 1 M or 0.1 M HCl were titrated into the humic solution. After each addition, the solution was left to equilibrate before the pH reading was taken. The equilibration criterion was that the pH did not drift by > 0.1 pH units/min. The experimental blank was taken by a titration of 20 ml of NaNO3 solution by the procedure outlined above. The moles of acid added to the humic substance solution in order to change the pH from 11 to 7 and from 7 to 3 were calculated (DTG (titratable groups)7 – 11 and DTG3 – 7, respectively). Blank titration values were subtracted and the final result was given as milliequivalents per gram of isolate. Sample analysis involved the accurate weighing of approximately 0.005 g of humic substance. The sample was dissolved in 3 ml of PBS, filtered through a 0.45-mm polycarbonate filter and analysed using a Dohrmann DC190 high-temperature total organic carbon (TOC) analyser. The absorbance at 340 nm was quantified in a 1-cm quartz cell
453
and blanked against PBS. Measurements were taken in a LKB Ultraspec 4050 UV spectrophotometer.
3. Results Fig. 2 shows the average daily air temperature (Panel 1) and monthly rainfall amounts (Panel 2) measured close to the site of pools X and Y. The average daily air temperature exhibited a seasonal temperature fluctuation from 5 C in approximately January/February rising to 20 C in July/ August. Rainfall patterns were erratic with a tendency towards lower values during the summer months. It was noted that a prolonged period of reduced rainfall and higher air temperatures caused both pools to dry completely for approximately 2 months during the summer of 1995. A similar but somewhat shorter (1 month) period was observed in the summer of 1996. Fig. 3 shows the weight average molecular weight data for the humic (Panel 1), fulvic (Panel 2) and the hydrophilic acids (Panel 3). The weight average molecular weight of the isolated humic acids showed a seasonal trend, with the molecular weights increasing before and after the dry spells experienced in both the summers of 1995 and 1996. Molecular weight appeared to increase from approximately 8000 to 10,000 in the winter and spring months up to 14,000 to 16,000 before or after a period when pool X dried. However, for pool Y lower molecular weights ( 6000 –8000) were found after the drought. The fulvic and hydrophilic acid fractions indicated no obvious seasonal or long-term trend. Weight average molecular weight varied from between 1600 and 2800 and 1400 and 3500, respectively. Both X and Y samples showed a good correlation between one another, having comparable values and patterns of change. Table 1 shows the mean weight average molecular weight for the humic, fulvic and hydrophilic acids. From the data, it is evident that aquatic humic acids have higher weight average molecular weight than fulvic acids, which are, in turn, marginally heavier than hydrophilic acids. Fig. 4 shows the atomic hydrogen to carbon ratio for the humic, fulvic and hydrophilic acids (Panels 1, 2 and 3, respectively). It was evident that upon rewetting after the prolonged dry spell of 1995 the atomic H/C ratio was greater than the previous value. This trend was evident in both pools and all three isolated compounds. The fulvic and hydrophilic acids exhibited a seasonal trend of slightly higher atomic H/C ratio during the late summer/early autumn period. This trend was more prominent in the pool
Table 1 Mean weight average molecular weight Fraction
Pool X
Pool Y
Literature
Reference
Aquatic humic acid Fulvic acid Hydrophilic acid
9136 2105 1954
10033 2211 1625
6581 2816 411
Birkett, 1996 Jones et al., 1995 Aiken et al., 1992
454
M.J. Scott et al. / Environment International 27 (2001) 449–462
Fig. 4. Atomic H/C ratio for humic, fulvic and hydrophilic acids. Panels: top, middle and bottom, respectively.
Y samples. The mean atomic H/C ratio for the fulvic acids appeared substantially lower than that for the humic and hydrophilic fractions (Table 2). Fig. 5 shows the atomic nitrogen to carbon ratio for the humic, fulvic and hydrophilic acids (Panels 1, 2 and 3, respectively). The atomic N/C ratio exhibited little or no
seasonal or long-term variations in any of the isolated compounds or pools. The fulvic acids had substantially lower atomic N/C ratios than the humic and hydrophilic acids (Table 2). Fig. 6 shows the content of functional groups titratable between pH 3 and 7 (DTG3 – 7) for the humic, fulvic and
M.J. Scott et al. / Environment International 27 (2001) 449–462 Table 2 Mean elemental composition data Sample
Pool
C
H
N
H/C
N/C
Humic acid
X Y X Y X Y
45.6 45.3 49.7 47.66 44.6 45.3
4.40 4.30 4.05 4.06 4.29 4.42
1.95 1.95 0.91 1.09 2.21 1.84
1.1483 1.1296 0.9809 1.0369 1.1630 1.1693
0.0364 0.0366 0.0159 0.0194 0.0443 0.0350
Fulvic acid Hydrophilic acid
hydrophilic acids (Panels 1, 2 and 3, respectively). There appeared to be little evidence to support a seasonal pattern within the data set. However, it was evident that although there are some initial postdrought rises, all three isolated compounds exhibited a decline in the DTG3 – 7 content after the prolonged drought period of the summer of 1995. This decline continued until the dry period of July 1996. Humic acid DTG3 – 7 fell from approximately 2 meq gm 1 to approximately 0.7 meq gm 1. This trend was also exhibited in both the fulvic (3 –1.2 meq gm 1) and the hydrophilic acids (3– 1 meq gm 1). After the summer drought of 1996, the DTG3 – 7 content of the hydrophilic acid samples isolated from both pools rose sharply to approximately 3 meq gm 1. After which, the DTG3 – 7 declined rapidly to 1.3 meq gm 1 by November 1996. This trend was not observed in either the humic or fulvic acid isolates. The humic acid isolates exhibited substantially lower mean DTG3 – 7 than the fulvic and hydrophilic acids (Table 3). Fig. 7 shows the content of functional groups that titrate between pH 7– 11 (DTG7 – 11) for the humic, fulvic and hydrophilic acid isolates (Panel 1, 2 and 3, respectively). The data showed no evidence of a seasonal or long-term pattern. No appreciable difference in mean DTG7 – 11 was evident between the humic, fulvic and hydrophilic fractions (Table 4). Fig. 8 shows the A340 g 1 C l 1 cm 1 for the humic, fulvic and hydrophilic acids (Panels 1, 2 and 3, respectively). The A340 g 1 C l 1 cm 1 data indicated no obvious seasonal or long-term trend. The mean A340 g 1 C l 1 cm 1 value for the humic was appreciably higher than those for the fulvic and hydrophilic acids (Table 4).
4. Discussion 4.1. Seasonal and long-term patterns Scott et al. (1998) showed that the DOM present in the pool water appeared to undergo a considerable change in characteristic after the extended drought period of 1995. The optical absorbance (340 nm) g 1 C l 1 cm 1 is a method of assessing to what degree a compound is aromatic (Aiken, 1997). Assuming that the DOM present solely caused the A340 in the pool waters, then the organic compounds appeared to become 50% less coloured after the drought
455
of 1995. The DOM molecules appeared to have undergone a significant change in molecular structure. Strong water deficits associated with the droughts of 1992 and 1995 may have allowed the influx of oxygen into layers of peat that were normally anoxic, possibly causing a major change in the decomposition process. Evidence of an oxygen influx is apparent from the rise in sulphate concentration after both prolonged drought events. Evidence of a similar change was found in the elemental composition data of the DOM isolates. The humic, fulvic and hydrophilic fractions all exhibited an increase in the atomic H/C ratio immediately after the 1995 drought, when compared to the previous reading. The rise in the atomic H/C ratio may have been indicative of a reduction in the aromaticity of the compounds (Steelink, 1985). After the prolonged drought of 1995, both atomic H/C and DTG3 – 7 exhibited gradual declines. It appeared that the humic substances present in the soil water gradually became more aromatic with less carboxyl group content per unit weight. This may have been a consequence of oxygenated peat becoming gradually more anoxic after rewetting. The decomposition process may have changed, as oxygen became less available. Higher aromaticity and lower carboxyl content may be characteristics expected for humic substances originating from more anoxic environments. Alternatively, this observation may have been a consequence of ease of dissolution upon rehydration of less aromatic humic substances with higher carboxyl group content. Tipping and Woof (1990) showed that the release of humic materials into soil water from acid organic soils was integrally linked to molecular charge and functional group content. Hayes (1985) stated that aromaticity can play a role in humic dissolution properties. In fact, Y fulvic acid and both hydrophilic isolates exhibited a similar elevation in atomic H/C in the following summer. Humic substances produced during the summer months are effectively trapped in the catchment soil because of restricted rainfall and therefore soil water flow. During the late summer/early autumn period, rainfall becomes more frequent effectively flushing potential DOM from the catchment (Scott et al., 1998). When soil water flow resumes, more hydrophilic components of DOM will be preferentially dissolved. Less aromatic humic substances are more soluble and therefore their prevalence in soil water during the autumn flush is greater relative to the more aromatic and possibly aggregated materials. The authors acknowledge that the evidence to suggest that the aromaticity of the humic isolate changes after prolonged dry periods is somewhat questioned by the lack of a corresponding drop in observed A340 g 1 C l 1 cm 1. The weight average molecular weights of humic acids isolated from the pool waters appeared to exhibit seasonal variations. The molecular weight of material isolated from both X and Y increased during the summer, particularly just before and after periods when the pools were dry. This trend was not evident in either the fulvic or hydrophilic acid
456
M.J. Scott et al. / Environment International 27 (2001) 449–462
Fig. 5. Atomic N/C ratio for humic, fulvic and hydrophilic acids. Panels: top, middle and bottom, respectively.
fractions. The humic acid molecules themselves appeared to remain relatively consistent in terms of molecular characteristic. The rise in molecular weight was not accompanied by
a corresponding change in the elemental composition and A340 g 1 C l 1 cm 1, suggesting that there was no change in the level of aromaticity. From the evidence available it
M.J. Scott et al. / Environment International 27 (2001) 449–462
Fig. 6. DTG3 – 7 (Carboxyl group meq g
1
457
) for the humic, fulvic and hydrophilic acids. Panels: top, middle and bottom, respectively.
was difficult to hypothesise potential reasons for such a change. Humic acids are known to complex with one another in the presence of certain metal ions and at higher concentrations (Reid et al., 1991). However, the isolation
procedure removes most metal ions, and all isolates were analysed for molecular weight at the same concentration and in a buffer solution to combat intermolecular complex formation. It appeared that there was a change in the
458
M.J. Scott et al. / Environment International 27 (2001) 449–462
Table 3 Mean functional group content DTG3 – 7 (meq g 1)
DTG7 – 11 (meq g 1)
DTG3 – 11 (meq g 1)
Sample
X
Y
X
Y
X
Y
Humic acid Fulvic acid Hydrophilic acid
1.1 1.9 2.1
1.3 1.9 2.2
1.2 1.2 1.1
1.5 1.3 1.4
2.3 3.1 3.2
2.8 3.2 3.6
WHAM humic WHAM fulvic
2.1 2.9
1.4 1.3
3.5 4.2
production process of humic acids during the summer months. Warmer, drier conditions appeared to alter the chemical and biochemical processes, producing humic acids that were significantly higher in weight average molecular weight. Unfortunately, due to the relatively short monitoring period and the absence of information available on the formation of humic substances it was difficult to make more than an observation of this trend. It has long been acknowledged that humic substances isolated from contrasting environments have significantly different, almost characteristic molecular make-ups. However, the authors are not aware of a comparable study to this one outlining that humic substances isolated at different dates from the same source can have considerable variation in their molecular characteristics. Table 5 illustrates the variation observed in each molecular characteristics exhibited by the humic substances over the monitoring period. Weight average molecular weight exhibits a two- to threefold variation and a fivefold variation is observed in A340 g 1 C l 1 cm 1 during the monitoring period. There are also significant variations in H/C and N/C ratios and DTG7 – 11 and DTG3 – 7. Such changes indicate that DOM production may be somewhat dynamic. A complex interaction of different processes may be constantly changing the production of DOM. Processes such as microbial degradation, chemical degradation, solute dissolution, etc. may all be subtly influenced by external factors such as temperature, oxygen availability, moisture content, etc. Changes in molecular characteristics may also be an artefact of the isolation procedure. As no one isolation is exactly identical to its predecessor subtle changes from one procedure to the next may influence the type of DOM adsorbed to the XAD resins or may influence the amount of resin leached into the humic substance during the isolation procedure. 4.2. Comparison of the molecular characteristics of the DOM isolates The fulvic acid fractions had appreciably lower atomic H/C and N/C ratios than the humic and hydrophilic fractions. Closer examination of the elemental composition showed the fulvic fractions contained less hydrogen and nitrogen, and more carbon than the humic and hydrophilic
acids (Table 2). Comparison of the data with those published by Rice and MacCarthy (1991) shows the fractions to be of similar elemental composition to those isolated elsewhere. However, the Great Dun Fell data set was unusual in that the fulvic acid fraction had a lower mean atomic H/C ratio than the humic acid, the reverse is usually the case. Though unexpected, the data for the humic and fulvic acids fell well within the range reported by Rice and MacCarthy. Evidence suggested that aquatic humic acids are considerably more similar to fulvic acid than their soil borne counterparts in terms of molecular weight (Wershaw and Aiken, 1985), elemental composition (Steelink, 1985; Rice and MacCarthy, 1991) and functional group content (Perdue, 1985). This suggested that this observation may just have been a characteristic of this system. Presently, the literature is very limited on examples of elemental analysis performed upon the hydrophilic fractions. McKnight et al. (1985) analysed the hydrophilic fraction of DOM taken from Thoreau’s Bog, MA, and reported elemental compositions of 44.7% carbon, 3.6% hydrogen and 1.27% nitrogen. McKnight’s results were closely comparable to those reported in this study. The weight average molecular weight data for the humic and fulvic acids compared well with the values contained within the published literature (Table 1). The large contrast with the hydrophilic fractions was probably due to differences in the method used for analysis. Aiken et al. (1992) measured number average molecular weights which are lower than weight average molecular weights for polydisperse samples such as hydrophilic acids (Wershaw and Aiken, 1985). The results of mean A340 g 1 C l 1 cm 1 indicated that the humic acid fraction had more colour per unit weight than the fulvic or the hydrophilic fractions. These data were consistent with the idea that humic acids are more aromatic than fulvic acids (Stevenson, 1985). It appeared from the data that fulvic acids were more aromatic than the hydrophilic acids, which would agree with the preliminary findings of characterisations carried out by Malcolm (1985). Most data in the literature for direct titrations carried out upon humic substances did not report measuremen ts between pH 3 and 11. This made it difficult to find measurements that were comparable to the results obtained in this study. However, it was possible to use the Windermere Humic Aqueous Model (WHAM) (Tipping, 1994) to calculate the change in charge between pH 3 – 7 and 7– 11 for both humic and fulvic acids. The model parameters were averages derived from titrations performed upon eight fulvic acids (water and soil origin) and six humic acids (soil origin). The model produced values of 2.9 meq g 1 for a fulvic acid and 2.1 meq g 1 for a humic acid which appear slightly higher than the mean DTG3 – 7 values for the pool isolates (Table 3). However, the results were consistent with the idea that fulvic acid contains more carboxyl groups than humic acid (Stevenson, 1985; Mathur and Farnham, 1985). The content of functional groups that could be titrated between pH 7 and 11 (DTG7 – 11) appeared to be of a similar
M.J. Scott et al. / Environment International 27 (2001) 449–462
Fig. 7. DTG7 – 11 (Phenol group meq g
1
459
) for the humic, fulvic and hydrophilic acids. Panels: top, middle and bottom, respectively.
value in each of the individual fractions (Table 3). The mean values varied from 1.1 to 1.5 meq g 1. Humic acid had slightly higher DTG7 – 11 than fulvic acid. Using WHAM, the calculated changes in charge between pH 7 and 11 were 1.28 meq g 1 for fulvic acid and 1.37 meq g 1 for humic acid. The average DTG7 – 11 for the humic and fulvic
fractions agree closely with those generated by the WHAM model. The model also supports the observation that the humic fractions have slightly higher DTG7 – 11 than the fulvic fractions. Unfortunately, WHAM does not consider hydrophilic acid and therefore no comparison could be made.
460 Table 4 Mean A340 g C
M.J. Scott et al. / Environment International 27 (2001) 449–462
1
l
1
cm
4.3. Summary
1
Fraction
Pool X
Pool Y
Aquatic humic acid Fulvic acid Hydrophilic acid
11.24 10.53 6.83
11.86 10.66 5.26
Fig. 8. A340 g C
1
l
1
cm
1
The observations made in this paper support the ideas put forward by Scott et al. (1998). Scott et al. showed that long-term wetting and drying patterns, and seasonal temperature variations and rainfall patterns, affected soil acid-
for the humic, fulvic and hydrophilic acids. Panels: top, middle and bottom, respectively.
M.J. Scott et al. / Environment International 27 (2001) 449–462 Table 5 Intra isolate molecular characteristic range
References
Characterisation
Humic
Fulvic
Hydrophilic
H/C ratio N/C ratio Mw DTG3 – 7 DTG7 – 11 A340 g C
0.93 – 1.74 0.005 – 0.08 6,000 – 18,000 0.2 – 2.6 0.4 – 3.0 1.4 – 7.0
0.82 – 1.25 0.009 – 0.05 1,500 – 2,750 1.0 – 3.0 0.9 – 2.6 1.0 – 4.5
0.85 – 1.6 0.017 – 0.138 1,300 – 3,300 1.0 – 3.4 0.2 – 2.9 0.8 – 3.8
1
l
1
cm
1
461
ity, sulphate dynamics, DOM production, its flux and molecular characteristics. Humic substances extracted from the same samples as Scott et al. (1998) showed comparable (if not as clear) changes in molecular characteristics. Elemental composition and carboxyl group content exhibited associations with long-term drying patterns. This corroborated well with changes in DOM A340 g C 1 l 1 cm 1 and hydrophilic DOM to total DOM ratio. However, changes in DOM A340 g C 1 l 1 cm 1 were not mirrored by isolate A340 g C 1 l 1 cm 1. Such contradictions may have been a consequence of the isolation procedure. Excessive exposure to alkaline solutions could have caused oxidation of humic substances (Hayes, 1985). The isolation procedure itself is selective to molecules with certain properties. The acidification of samples precolumn may have caused some DOM to precipitate, thus losing less soluble components. Characteristics such as lower functional group contents (Tipping and Woof, 1990), higher molecular weights (Leenheer, 1985) and higher aromaticity (Stevenson, 1985) can reduce humic substance solubility. Adsorption onto the XAD resin is by its nature selective (Leenheer, 1985). XAD8 would selectively adsorb hydrophobic compounds (Aiken, 1985), thus selecting molecules that had higher molecular weights, higher A340 g C 1 l 1 cm 1, lower atomic H/C and lower functional group content. These explanations may have played roles in the difference in molecular characteristics of natural DOM and its corresponding isolated humic substances. Numerous authors have used XAD resins to isolate humic substances from DOM originating from different environments, and found the technique capable of identifying ascertainable differences in molecular characteristics. However, it may have been that changes caused by climatic fluctuations in DOM originating from the same environment may have been too subtle to not be masked by the selective nature of the isolation procedure.
Acknowledgments We would like to Mr K. Taylor and Mr M. Garnett (ITE, Merlewood) for the meteorological data. English Nature for allowing access to the Moor House Nature Reserve. Finally, we would like to thank the Natural Environment Research Council for funding the CASE studentship to MJS.
Aiken GR. Isolation and concentration techniques for aquatic humic. In: Aiken GR, McKnight DM, Wershaw RL, MacCarthy P, editors. Humic substances in soils, sediment and water. New York: Wiley-Interscience, 1985. pp. 363 – 86. Aiken GR. Factors controlling the nature and reactivity of organic matter in aquatic systems. Abstract: Humus. Nordic Humus Newsletter 1997;4(1) 8. 6th Nordic Symposium on Humic Substances Humic Substances as an Environmental Factor, Hameenlinna, Finland. Aiken GR, McKnight DM, Thorn KA, Thurman EM. Isolation of hydrophilic organic acids from water using nonionic macroporous resins. Org Geochem 1992;18(4):567 – 73. Allard B, Boren H, Petterson C, Zang G. Degradation of humic substances by UV radiation. Environ Int 1994;20(1):97 – 101. Anderson JM. Carbon dioxide evolution from two temperate, deciduous woodland soils. J Appl Ecol 1973;10:361 – 78. Batomalaque AE, Kikuma M, Seki H. Population dynamics of attached bacteria in a mesotrophic swampy bog of Japan. Water, Air, Soil Pollut 1992;63:371 – 8. Birkett JW. The physical characterisation of humic substances from peat and natural waters. PhD Thesis, University of Manchester, 1996. Bryan ND. Metal uptake by humic substances: a combined experimental and modelling study. PhD Thesis, University of Manchester, Manchester, 1994. Cronan CS, Aiken GR. Chemistry and transport of soluble humic substances in forested watersheds of the Adirondack Park, New York. Geochim Cosmochim Acta 1985;49(8):1697 – 705. Grieve IC. Concentrations and annual loading of dissolved organic matter in a small moorland stream. Freshwater Biol 1984;14:533 – 7. Hayes M. Extraction of humic substances from soil. In: Aiken GR, McKnight DM, Wershaw RL, MacCarthy P, editors. Humic substances in soils, sediment and water. New York: Wiley-Interscience, 1985. pp. 329 – 62. Hemond HF. Biogeochemistry of Thoreau’s Bog, Concord, Massachusetts. Ecol Monogr 1980;50:507 – 26. Hessen DO. The relationship between bacterial carbon and dissolved humic compounds in oligotrophic lakes. FEMS Microbiol Ecol 1985;31(4): 215 – 24. Jones MN, Birkett JW, Wilkinson AE, Hesketh N, Livens FR, Bryan ND, Lead JR, Hamilton-Taylor J, Tipping E. Experimental determinations of partial specific volumes of humic substances in aqueous solutions. Anal Chim Acta 1995;314:149 – 59. Leenheer JA. A comprehensive approach to preparative isolation and fractionation of dissolved organic carbon from natural waters and waste waters. Environ Sci Technol 1981;15(5):578 – 87. Leenheer JA. Fractionation techniques for aquatic humic substances. In: Aiken GR, McKnight DM, Wershaw RL, MacCarthy P, editors. Humic substances in soils, sediment and water. New York: Wiley-Interscience, 1985. pp. 409 – 32. Liechty HO, Kuuseoks E, Mroz GD. Dissolved organic carbon in northern hardwood stands with differing acidic inputs and temperature regimes. J Environ Qual 1995;24(5):927 – 33. Malcolm RL. Geochemistry of stream fulvic and humic substances. In: Aiken GR, McKnight DM, Wershaw RL, MacCarthy P, editors. Humic substances in soil, sediment and water. Geochemistry, isolation and characterisation. New York: Wiley-Interscience, 1985. pp. 181 – 210. Mathur SP, Farnham RS. Geochemistry of humic substances in natural and cultivated peatlands. In: Aiken GR, McKnight DM, Wershaw RL, MacCarthy P, editors. Humic substances in soil, sediment and water. Geochemistry, isolation and characterisation. New York: Wiley, 1985. pp. 53 – 86. McKnight D, Thurman EM, Wershaw RL, Hemond H. Biochemistry of aquatic humic substances in Thoreau’s Bog, Concord, Massachusetss. Ecology 1985;66(4):1339 – 52. Meier M, Namjesnik Dejanovic K, Maurice PA, Chin YP, Aiken GR.
462
M.J. Scott et al. / Environment International 27 (2001) 449–462
Fractionation of aquatic natural organic matter upon sorption to geothite and kaolinite. Chem Geol 1999;157(3 – 4):275 – 84. Perdue EM. Acidic functional groups of humic substances. In: Aiken GR, McKnight DM, Wershaw RL, MacCarthy P, editors. Humic substances in soil sediment and water. Geochemistry, isolation and characterisation. New York: Wiley-Interscience, 1985. pp. 493 – 526. Pohlman AA, McColl JG. Soluble organics from forest litter and their role in metal dissolution. Soil Sci Soc Am J 1988;52(1):265 – 71. Reid PM, Wilkinson AE, Tipping E, Jones MN. Aggregation of humic substances in aqueous media as determined by light-scattering methods. J Soil Sci 1991;42:259 – 70. Rice JA, MacCarthy P. Statistical evaluation of the elemental composition of humic substances. Org Geochem 1991;17(5):635 – 48. Scott MJ, Woof C, Jones MN, Tipping E. Concentrations and fluxes of dissolved organic carbon in drainage water from an upland peat system. Environ Int 1998;24(5 – 6):537 – 546. Steelink C. Implications of elemental analysis of humic substances. In: Aiken GR, McKnight DM, Wershaw RL, MacCarthy P, editors. 1st ed. Humic substances in soil, sediment and water. Geochemistry, isolation and characterisation, vol 1. New York: Wiley, 1985. pp. 457 – 76. Stevenson FJ. Geochemistry of soil humic substances. In: Aiken GR, McKnight DM, Wershaw RL, MacCarthy P, editors. Humic substances in soil, sediment and water. New York: Wiley-Interscience, 1985. pp. 13 – 52. Tegen I, Dorr H. Mobilization of caesium in organic rich soils: correlation with production of dissolved organic carbon. Water, Air, Soil Pollut 1996;88:133 – 44.
Thurman EM, editor. Organic geochemistry of natural waters. Boston: Niyhoff/Junk, 1985. Thurman EM, Wershaw RL, Malcolm RL, Pinckney DJ. Molecular sizes of aquatic humic substances. Org Geochem 1982;4:27 – 35. Tipping E. WHAM—a chemical equilibrium model and computer code for waters, sediments and soils incorporating a discrete-site/electrostatic model of ion binding by humic substances. Comp Geosci 1994;20: 973 – 1023. Tipping E, Woof C. Humic substances in acid organic soils—modeling their release to the soil solution in terms of humic charge. J Soil Sci 1990;41(4):573 – 86. Tipping E, Woof C, Ineson P, Taylor K, Bol R, Harkness DD, Benham D, Poskitt J, Rigg E. Climatic influences on the leaching of dissolved organic matter from upland UK moorland soils, investigated by a field manipulation experiment. Environ Int 1998;25(1):83 – 95. Urban NR, Bayley SE, Eisenreich SJ. Export of dissolved organic carbon and acidity from peatlands. Water Resour Res 1989;25(7):1619 – 28. Vance GF, David MB. Dissolved organic carbon and sulphate sorption by spodsol mineral horizons. Soil Sci 1992;154(2):136 – 44. Wershaw RL, Aiken GR. Molecular size and weight measurements of humic substances. In: Aiken GR, McKnight DM, Wershaw RL, MacCarthy P, editors. Humic substances in soil sediment and water. Geochemistry, isolation and characterisation. New York: Wiley, 1985. pp. 477 – 92. Wilson DE. The influence of humic compounds on titrimetric determinations of total organic carbon in freshwater. Arch Hydrobiol 1979;87(3): 379 – 84.