Introduction to DOM
A short introduction to dissolved organic matter (DOM) taken from the introduction chapter of my PhD. (Stedmon 2004)
What is DOM and why is it relevant?
The reservoir of dissolved organic matter (DOM) in the world’s oceans is the second largest dynamic carbon pool on Earth, and it plays an important role in the global carbon cycle. More than 97% of the organic carbon in the ocean exists as DOM (Benner, 2002). The amount of carbon in DOM is similar to that present in the atmosphere as carbon dioxide (CO2) (Siegenthaler and Sarmiento, 1993). For comparison, the net oxidation within a year of 1% of the carbon bound in DOM would lead to a greater production of CO2 than that arising from the burning of fossil fuels (Hedges, 2002). In addition to playing an important role in global carbon budgets, DOM is also important on smaller scales. Its biological, physical and chemical properties greatly influence the ecology of aquatic ecosystems. DOM forms the base of the microbial food web, being a source of energy (carbon) and nutrients (nitrogen and phosphorous) to bacteria. In many systems DOM is also the major light absorbing component (Figure 1), having a combined positive and negative feedback on the aquatic organisms. DOM plays a photoprotective role due to the fact that it strongly absorbs ultraviolet (UV) light, which can be harmful to aquatic organisms. Conversely, when present in high concentrations, it may also limit the penetration of photosynthetically available radiation (PAR) and therefore influence the distribution, composition and productivity of algae and macrophytes. DOM also influences trace metal speciation and as a result their transport and bioavailability in aquatic ecosystems. The presence of a large number of charged functional groups in the structures of the organic compounds present in DOM, gives DOM the ability to complex with metals. In freshwaters and coastal waters a considerable proportion of metals are bound within the DOM pool. These metals can be either toxins and pollutants (e.g. Pb) or micronutrients (e.g. Fe).
What is dissolved?
In nature there is no clear boundary between dissolved and particulate phases, but more a continuum of increasing particle sizes from water molecules up to whales (Sharp, 1973). A common definition of the dissolved phase is that that can pass through GF/F glass fibre filters, as these are readily available to the aquatic science community and easy to clean, to avoid contamination of samples. GF/F filters have an average pore size of approximately 0.7 μm, and DOC measurements therefore also include some colloids (0.001-1 μm) and small particles. However this fractionation is thought to be satisfactory as it excludes most living organisms, apart from small bacteria and viruses (Hedges, 2002). In addition particles below 1μm in diameter do not tend to sink (Duursma, 1961). Therefore, in commonly the dissolved phase is defined as that passing through a specific easily available glass fiber filter (so called GF/F filter).
DOM sources
In aquatic ecology sources of DOM are generally classed into two groups; allochthonous DOM, representing the material which is produced outside of a system and transported to it, and autochthonous material which is produced within the system. The relative importance of the allochthonous and autochthonous fractions varies depending on season, location and trophic status of the water body. Allochthonous material is generally dominated by terrestrially derived DOM, which arises from the degradation of terrestrial plants in soils. Soil organic matter is dissolved in rain and ground water and subsequently transported through riverine, estuarine and coastal environments. Terrestrial DOM characteristically contains lignin degradation products and other macromolecules unique to terrestrial (vascular) plants (Moran et al., 1991). The quality and quantity of terrestrial DOM is dependent on the climate, land use and drainage characteristics of the catchment (Aitkenhead-Peterson et al., 2003; Stedmon et al 2006). In addition microbial and photochemical degradation processes are continually acting on allochthonous DOM whilst it is transported through rivers, lakes and estuaries, transforming its characteristics. There is also a supply of DOM to the water column from the sediments (Burdige et al., 1992). DOM released from the sediment is derived from the microbial degradation of particulate organic matter originating from both terrestrial sources and aquatic plant matter. In this case I have considered it to be an allochthonous source as it is not formed in the water column itself but derived from particle bound material in the sediments. In reality the organic material in the sediments contains a combination of both terrestrially derived and autochthonous material.
Autochthonous DOM is derived from algae and macrophytes and the re-cycling of this material through the microbial food web (Figure 1). Approximately 50% of primary production by phytoplankton is released as DOM (Nagata, 2000). A number of processes are responsible for this release. Both living and dead algae release DOM from their cells either passively (leakage, Bjørnsen, 1988) or actively (overflow of photosynthate, Fogg, 1983). The infection and subsequent lysis (rupture) of bacterial and algae cells by viruses, releases the cell solutes and fragments of cells into the surrounding water (Gobler et al., 1997). During grazing across all levels of the food web, DOM is released through “sloppy feeding”, excretion or egestion (Nagata, 2000). Finally, there is also a release of DOM via dissolution of sinking non-living particulate organic material also referred to as detritus (Jumars et al., 1989).
Figure 1. Schematic of carbon flow through the pelagic food web.
DOM composition
The carbon content of DOM (dissolved organic carbon, DOC) is commonly used as a quantitative measurement of DOM. There are considerable variations in DOC concentrations across different environments. The lowest concentrations are generally found in the deep ocean, where concentrations can vary between 35−45 μM (Benner, 2002). Highest concentrations are found in rivers and streams (2501000 μM), while estuarine and coastal seas have intermediate concentrations. Although DOM consists predominantly of carbon, there are also significant amounts of nitrogen and phosphorous bound within it (e.g. Paper V). These concentrations are referred to as dissolved organic nitrogen (DON) and dissolved organic phosphorous (DOP). Our understanding of the dynamics of DON and DOP is limited when compared to that of DOC, primarily due to analytical difficulties. However, recently, the importance of DON and DOP in aquatic systems is becoming more apparent. It is clear that both the content and availability of organically bound N and P varies depending on the source. In estuarine and marine systems over 50 % of the total dissolved N and P exists as DOM (Berman and Bronk, 2003; Karl and Björkman, 2002, Paper V). A recent review of the chemical composition of DOM in oceanic environments revealed that molecular analyses can only account for 4−11% and 1−3% of dissolved organic carbon (DOC) in surface and deep waters respectively (Benner, 2002). Among the identified groups of compounds are neutral sugars, amino sugars and amino acids. In contrast, studies in some lakes where autochthonous DOM dominated, have shown that these groups of compounds can explain up to a third of the DOC (Tranvik and Jørgensen, 1995; Weiss and Simon, 1999). However, the bulk of DOM in all aquatic environments remains uncharacterised on a molecular level. The high concentrations of other dissolved constituents in water (e.g. salts) and the low concentrations of specific dissolved organic moieties make further molecular characterisation difficult.
The composition of DOM has also been described by other, more general, characteristics. The development of ultrafiltration techniques has allowed the classification of DOM into different size fractions (Sharp, 1973; Benner, 2002). Currently two operationally defined size classes exist; high molecular weight (HMW) material which is retained on a 1 nm pore size membrane, and low molecular weight (LMW) material, which passes through. In the marine environment 60-80% of DOM is in the LMW class (Benner, 2002). In contrast riverine DOM is dominated by HMW DOM (45-80%) (Benner, 2003). This HMW fraction is characterised as consisting of more aromatic components than autochthonous HMW material (Benner et al., 1992). This is due to the fact that terrestrial plant material contains more aromatic structures (for example degradation products of lignin’s and tannin’s) than its algae derived counterpart. A major fraction of HMW DOM consists of humic substances (Harvey et al., 1983; Malcom and McCarthy, 1992). They are a series of yellow/brown organic compounds formed via biotic and abiotic processes (humification) in both soils and the water column. There are two groups of humics in DOM, operationally defined by their solubility at different pHs. Fulvic acids are soluble at all pH values and represent 45-90% of the DOC in allochthonous DOM (Thurman, 1985). The other fraction, which precipitates at low pH, is referred to as humic acid and is less dominant in both freshwaters and marine environments. In autochthonous DOM, fulvic acid represents 10-47% of DOC, whilst humic acid represents less than 5% (Harvey et al., 1983).
Removal of DOM
Microbial degradation
Microbial consumption of DOM is by far the most intensely studied removal process, due to the fact that it is considered to be the most important sink of DOM (Williams, 2000). Bacteria readily consume the fraction of LMW DOM which can pass through their cell membranes. This is termed labile DOM. The carbon consumed is either respired as CO2 or incorporated into cell material. Via the production of hydrolytic enzymes, bacteria are also able to degrade some of the larger molecular weight fraction. The enzymes enable the bacteria to cleave larger structures into smaller molecules suitable for uptake. However, as mentioned earlier, the majority of DOM is present in complex macromolecules which can be resistant to enzyme cleavage. Indeed the fact that both bacteria and DOM are found to co-exist in all aquatic ecosystems suggests that not all of the DOM is readily available for consumption by bacteria. However other factors, such as nutrient limitation and grazing, also play a role. The uptake of DOM by bacteria forms the base of the flow of carbon through aquatic food webs. Bacteria are in turn consumed by protozoa and zooplankton, and as a result carbon is channelled from the DOM pool through bacteria and onwards up the food web (Figure 1). This concept is termed the microbial loop and represents a significant link in the transfer of carbon to higher trophic levels, which until the 1970’s was not included in food chain models. It is also important to note that at each trophic level there is a production of DOM. The DOM produced from extracellular release by phytoplankton consists of carbohydrates, proteins and amino acids. These products are considered to be relatively labile and readily available to microbes. Grazing and excretion by protozoa and zooplankton, is thought to supply between 10-30% of ingested organic carbon, as DOM. In addition viral lysis of bacteria and algae cells also releases both labile cell solutes and relatively more recalcitrant DOM derived from cell structures.
In addition to the fact that microbial degradation of DOM can be limited by its lability, bacterial activity can also be restricted by nutrient limitation. As well as needing carbon for production of cell material, bacteria also depend on nitrogen and phosphorous. Bacteria therefore compete with phytoplankton for available nutrients. Nutrient limitation of bacteria has been shown to explain the seasonal accumulation of labile DOM in surface waters (Williams, 1995).
Photodegradation
What about photochemistry? The traditional view has been that microbial degradation is its dominant removal process. However this is changing as a result of research into the importance of photodegradation processes. It is now clear that photochemical processes are both a sink (removal process) for, and transformer of, DOM, acting in concert with bacteria. The global input of DOM via rivers is enough to sustain the turnover of the entire pool of DOM in the oceans (Hedges et al., 1997). Riverine DOM is thought to be relatively resistant to microbial degradation, due to the fact that it consists of high molecular weight material. However in the oceans the DOM measured does not appear to be of terrestrial origin (Druffel et al., 1989; Hedges et al., 1997). So the question arises; where does the terrestrial DOM go? Photochemical reactions could explain the additional removal of terrestrial DOM. Miller and Zepp (1995) estimated that photo-mineralisation could be responsible for the removing 10-20% of terrestrially derived DOC in continental shelf waters. In addition, the structural changes occurring lead to alterations in biological, chemical and physical properties and can in turn substantially decrease terrestrial DOM lifetime. Photodegradation can have an extensive influence on aquatic ecosystems resulting in: i) A loss in the light absorbing properties and hence increased exposure of the water column to UV and visible light. ii) Direct mineralisation of organic carbon (e.g. CO and CO2 production). iii) Oxygen consumption. iv) Production of labile organic compounds and compounds rich in nitrogen and phosphorous. v) Production of biologically resistant DOM. vi) Destruction of organic ligand-binding capacity, causing the release of trace metals/ micronutrients/toxins (e.g. Cu, Fe, Mn).
Absorption of light (energy) by a molecule results in its transition into an excited state. Immediately after this transition, relaxation to the ground state then occurs through loss of the absorbed energy via internal conversion (heat and molecular motion) and fluorescence or phosphorescence. During a primary photochemical reaction, a chemical alteration occurs whilst the molecule is in the excited state. The products of the reaction can be a new stable molecule(s) and/or new reactive species, which then go on to initiate secondary reactions. Whereas primary reactions only concern the organic compounds that can absorb light (is coloured, CDOM), secondary reactions can affect all compounds present (i.e. the whole DOM pool), depending on the lifetimes, concentrations and reactivity of the reactive species. Photoreactions can result in molecular cleavage and/or rearrangement of the molecules (Zika, 1980). As a result of the complex nature of DOM in aquatic environments, there are many possible secondary reactions, however, the reactions of organic and inorganic free radicals produced as a consequence of primary reactions, are major degradation pathways for DOM. Radicals are highly reactive and influence both the chemistry and biology of sunlit waters. For example, light absorption by the coloured fraction of DOM (CDOM) is the principal source of a variety of reactive oxygen species (ROS) in surface waters, which play a pivotal role in the degradation of DOM. Inorganic chromophores (e.g. nitrate and nitrite) are also a significant source of ROS, if they are present in high enough concentrations (Blough and Zepp 1995).
Figure 2. The variation in the energy of a photon of light at different wavelengths. Also plotted are the bond dissociation energies (i.e. the energy needed to break/cleave a chemical bond in a molecule) for a selection of organic bonds.
The underwater light environment can be divided into three wavebands; UVB (280-315 nm), UVA (315-400 nm) and visible light (400-800 nm). The energy associated with a photon of light decreases with increasing wavelength (Figure 2). For example, a photon at 300 nm has 33% more energy than a photon at 400 nm. This, in combination with increasing awareness of decreasing stratospheric ozone concentrations (the major UVB light absorber in the atmosphere), resulted in research focused on the effects of UVB light on photodegradation processes. However UVB light is rapidly attenuated in natural waters due to CDOM absorption. Figure 3 shows the depth at which there is 1% of surface irradiance left for different sampling sites in this project. One can see that UVA light and blue light (400 nm) penetrates much further into the water column than UVB and therefore has the potential to expose a greater volume. Although the energy per photon in the UVB is high, the combined effect of a comparatively lower total energy for the waveband (accounting for <1% of incident radiation) and rapid attenuation suggests that the UVA and visible wavelengths could dominate photodegradation processes. In addition it is clear that photons in the UVA and blue region have enough energy to initiate various photochemical reactions in sunlit waters (Figure 3). So when investigating light induced degradation of DOM in natural waters it is clear that one has to consider the trade-off between photon energy, in situ light intensity and CDOM’s light absorption.
Figure 3. The photic depth (depth of 1% of surface light) for UVB, UVA and blue light in four localities sampled during the project.
Interaction and competition between microbial and photodegradation.
Photodegradation reactions compete with bacteria to degrade a fraction of DOM and also aid bacterial degradation via the cleavage of non-bioavailable components of DOM. Bacteria use DOM for growth and as an energy source (see earlier). As they can only exploit energy from bond cleavages taking place inside their cell, DOM must first be taken up (i.e. penetrate their cell membranes). The fragmentation of high molecular weight DOM via extracellular enzymes (enzymes attached to the outside of their cell membranes) and photodegradation processes therefore control the availability of DOM to the microbial population. Research over the last decade has generally shown that the photodegradation of terrestrially derived DOM has a positive effect on its bioavailability, reducing its average molecular weight and producing a suite of labile organics compounds, which are readily available to the microbial food web (Wetzel et al., 1995; Moran and Zepp, 1997). On the contrary, results have shown that photodegradation of autochthonous DOM can render the pool more resistant to microbial degradation, suggesting that the two processes compete for a degradable fraction (Bidanda and Benner, 1998; Tranvik and Kokalj, 1998; Tranvik and Bertilsson, 2000, Stedmon & Markager 2005b).
Abiotic dissolved-particulate phase transformations
There are primarily two abiotic pathways for the transformation of DOM to the particulate phase; adsorption and flocculation. Adsorption is the process by which dissolved organic compounds are attracted to the surface of inorganic particles. These can be mineral particles or metal oxides and hydroxides. If the number of mineral sites is limited different fractions of the DOM pool compete for adsorption and the more aromatic fractions are preferentially adsorbed (Gu et al., 1996a; 1996b). Flocculation is the aggregation of colloidal DOM to form particulate material. This can occur as a result of changes in ion concentrations during saline mixing in estuaries (Sholkovitz, 1976; Mulholland, 1981). The majority of removal occurs at salinity’s between 0 and 5 (Sholkovitz et al., 1978; Powell et al., 1996). Overall this process leads to a reduction in the average molecular weight of DOM during mixing (Brown, 1975; Sholkovitz et al., 1978; Powell et al., 1996). Flocculation has also been shown to occur spontaneously in a water body (Wells and Goldberg, 1996; Chin et al., 1998; Kerner et al., 2003), indicating a possible direct non-bacterial “supply route” of carbon from DOM to higher trophic levels, which is currently not included in existing food web models (Kerner et al., 2003). The importance of these abiotic pathways for the quantitative removal of DOM is still debatable and has been reported to vary between 0-25% of DOC, possibly depending on the source of DOM (Sholkovitz, 1976; Powel et al., 1996; Forsgren et al., 1996; Søndergaard et al 2005). However it is clear that salinity driven flocculation processes do lead to a restructuring of the DOM fraction, with the preferential removal of a humic fraction (Sholkovitz, 1976; Powel et al., 1996; Søndergaard et al 2005).
This text is taken from my Ph.D. thesis introduction chapter (2004).