Carbon cycle

Although particulates that are predominantly clay mineral in content are important attenuators of UVR in turbulent streams and rivers, dissolved and particulate organic substances largely control the penetration of UV into most lakes and the sea. Hence, this discussion of the carbon cycle begins with a discussion of the interactions of UV with CDOM, with emphasis on the optical properties of aquatic ecosystems and penetration of solar UVR into the water (see also Chapters 3 and 6).

5.2.1 UV-CDOM interactions and aquatic optical properties 5.2.1.1 CDOM sources and characteristics

Because CDOM is the most important UV-absorbing dissolved organic constituent in aquatic systems, it plays an important role in the interactions between UVR and aquatic carbon cycling. As used in this chapter, CDOM includes hydrophobic colored organic matter referred to as "humic substances" [7,13]. These substances are chemically complex and poorly characterized mixtures of anionic organic substances known to contain phenolic and carboxyl groups [13-18]. Terrestrially-derived CDOM originates through the decomposition of dead plant material and it is introduced into water via leaching and runoff from land. CDOM also can be produced through the decay of algal detritus; this source of CDOM has been referred to as "microbially-derived" [15,16]. There is some evidence that CDOM in the ocean can also be produced by photoreactions of triglycerides and fatty acids [19] in regions such as the sea-surface microlayer [5].

There are distinguishable differences between the optical and physicochemical properties of humic substances that are related to the molecular properties of these substances [13-18]. CDOM accounts for only a small fraction of the DOC in the open ocean and it has a significantly-lower aromatic content and specific absorption coefficients than terrestrially-derived CDOM [13]. These differences can be ascribed to variability in sources [13,15,16,18] and/or transformations [19-35] of the CDOM.

Light absorption by CDOM typically decreases in an approximately exponential fashion with increasing wavelength [13]. The absorption coefficient a(X) is defined as:

where A(X) is the absorbance of a water sample or aqueous solution of isolated CDOM at wavelength X in a cell of pathlength I. The relationship between a(X) and wavelength (in nm) can be represented by an equation of the following form a(A) = a(A0)es(A°-4» (2)

where S, the spectral slope coefficient, is computed by a non-linear least-squares fitting routine, and a(X0) is the absorption coefficient at a reference wavelength X0. S is a parameter that characterizes how rapidly the absorption decreases with increasing wavelength. The specific absorption coefficient, a(A)*, is defined by the relation:

where [DOC] is the DOC concentration expressed as mg CL_1.

The coefficients, a(X)* and S, vary both spatially and temporally [13,18] although the variation is not large for CDOM in freshwaters and coastal waters. Generally, values of a(X)* in the UV-B region have been observed to be lower for very clear oligotrophic seawaters than for coastal seawaters and freshwaters strongly influenced by terrestrial input. S for the UV region generally increases with decreasing a{X)*, ranging from as low as 0.012-0.013 nm-1 for some highly absorbing coastal waters to over 0.02 nm-1 for weakly absorbing oligotrophic waters [13]. Sharp changes in S can occur in coastal regions along seaward transects [36-38] where oceanic (microbially-derived?) sources of CDOM are becoming dominant along the transect, perhaps as terrestrially-derived CDOM is removed by photochemical and biological processes.

5.2.1.2 CDOM effects on UV penetration

Detailed concepts and elegant numerical models have been developed to describe the transmission of solar radiation in aquatic ecosystems and several thorough reviews of this literature have appeared [39-41]. Solar irradiance at a given wavelength decreases approximately exponentially as it penetrates into a region of uniform composition in a natural water body. The slope of natural logarithmic plots of irradiance versus depth has been referred to as the "diffuse attenuation coefficient [Xd(/l)]." The diffuse attenuation coefficient for down-welling irradiance at a given wavelength can be modeled as a function of the composition of seawater, focusing on absorption by water, chlorophyll-like pigments, and dissolved organic carbon (DOC) [40,42,43], In certain fresh-

waters, non-living particulate matter, which includes detritus and suspended sediments made primarily of clay minerals, can make a significant contribution to UV and visible light attenuation [41,44,45]. Microorganisms that are often exposed to UV-B radiation can develop cellular UV-protective substances such as mycosporine-like amino acids (MAAs, see Chapter 10) that absorb in the UV region. Such organisms or detritus derived from them can contribute significantly to UV attenuation in ecosystems that have low concentrations of DOC [41,46].

Studies of a wide range of freshwater and marine environments have shown that DOC, and, to a lesser extent in the open ocean, organic colloids, play an important role in the attenuation of solar UVR [13,45-54]. Not all of the DOC is responsible for such attenuation; CDOM is the UV-absorbing component. CDOM strongly absorbs UV-B and UV-A radiation and its absorption is sufficiently great in the visible (400-700 nm) region that it affects the remote sensing of ocean color [13,50-52,54]. One important feature of the absorption spectra of CDOM is the exponential increase in absorption with decreasing wavelength in the UV region. This feature enables CDOM to effectively protect aquatic ecosystems from harmful UV-B radiation while permitting beneficial photorepairing (UV-A) and PAR to be much more efficiently transmitted into the water. Efforts to model the spectral effects of CDOM on primary production [55] and microbial processes [56] have been reported.

Changes in CDOM concentrations in freshwaters and the sea cause significant changes in UV penetration. These changes occur seasonally [13,46,48-54,57] and they also likely are linked to climate [50, Chapter 17], acid deposition [50] and land use changes.

5.2.1.3 Effects of UV on CDOM optical properties

Microorganisms do not readily decompose the polymeric substances that make up CDOM, but CDOM transformation is accelerated when it is exposed to solar UVR. In this section, the effects of such transformations on CDOM optical properties are discussed. Also, see Section 5.2.3 and Chapter 6 for discussion of other aspects of the effects of UV on CDOM decomposition. It has been known for years that absorption of solar UVR by terrestrially-derived humic substances and CDOM results in a reduction in its light absorption and fluorescence (i.e. photobleaching) [3-8,13,19-35,52-54,57,58], The CDOM in water samples, freshwater and coastal regions impacted by riverine inputs is readily bleached on exposure to natural solar radiation with concurrent increases in the spectral slope coefficient [18,35]. A study of freshwater lakes in the USA concluded that acid-neutralizing capacity has particularly important effects on photobleaching rates [28], but other studies in South America and Antarctica have found no effects of alkalinity on the rates [Zagarese, personal communication, 2002], Factors such as oxygen concentration [59] and temperature [60] also affect photobleaching efficiency.

Measuring "apparent quantum yield" spectra can quantitate wavelength effects on photoreactions. The term "apparent quantum yield" denotes the moles of photoproduct formed per absorbed Einstein (mole photons) of radiation at a

2 CD

2 CD

280 300 320 340 360 380 400 Wavelength, nm

Figure 2. Apparent quantum yield spectra for the photobleaching of CDOM derived from the St. Lawrence Estuary. [Adapted from Whitehead et al. [29], Figure 4, p. 285, Copyright 2000, The American Society for Limnology and Oceanography, Inc.]

certain wavelength. Apparent quantum yields provide a useful unitless gauge for comparison of photoreactions of DOM from different natural waters. Recently published quantum yield spectra for CDOM photobleaching are shown in Figure 2, where the wavelength dependence of CDOM absorbance loss per absorbed Einstein is plotted. These and other quantum yield spectra indicate that the efficiency for the photobleaching of CDOM absorption and fluorescence is greatest in the UV region [23,29,35,61].

The excitation-emission matrix spectra (EEMS) of CDOM fluorescence also are altered on exposure to solar radiation, generally with reductions in fluorescence that approximately parallel absorption losses [36-38]. Hypsochromic shifts (shifts to shorter wavelengths) occur in both excitation and emission maxima on irradiation [36-38]. Interactions between photochemical and microbial degradation [38] are involved.

The photobleaching of CDOM involves two general classes of photoreactions that occur in aquatic environments: direct and indirect (photosensitized) [3,62,63, Chapter 8]. Direct photoreactions involve light absorption by the photoreactive constituent(s) of CDOM to produce reactive excited states. On the other hand, photoreactions also can involve indirect photoprocesses that also are initiated through light absorption by the CDOM. Indirect photoreactions occur through the intermediacy of various short-lived reactive transients, such as excited states, or species produced there from, that diffuse through the system and then react with the CDOM, sulfur compounds (see Section 4) or other constituents of the water. There is abundant evidence from continuous or laser flash photolysis experiments that various reactive transients are produced on irradiation of CDOM [62, Chapter 8]. Molecular oxygen often is involved in the formation of the reactive transients. CDOM photoreactions potentially can proceed by either direct or indirect pathways or by both pathways. The finding that superoxide reacts rapidly with CDOM [64] suggests that indirect photoreactions involving this transient may contribute to CDOM photochemistry. Reactions of CDOM with hydroxyl radicals account for a small fraction of the photobleaching [65]. A recent study by Del Vecchio and Blough [35] has shown that photobleaching of CDOM with monochromatic radiation occurs most rapidly at the irradiation wavelength, indicating that direct photolysis is at least partly involved in the photobleaching. With complex substances such as CDOM, it is conceivable that photoreactions may involve trapping of reactive transients before they can diffuse away into bulk aqueous solution.

5.2.2 UV interactions with aquatic carbon capture and storage

Phytoplankton communities are primarily responsible for the production of biomass in large lakes and the ocean. Submerged and partially-submerged aquatic vegetation play a central role in creation of biomass in many freshwater systems [66], Freshwaters also receive inputs of allochthonous (externally produced) organic matter derived from terrestrial plants and soils, and rivers transport large amounts (~400 teragrams (Tg) C yr_1; 1 Tg= 1012 g ) of terrestrial organic matter into coastal waters [67-71]. The effects of UV-B on terrestrial plant productivity [72] and submerged plants [45, Chapter 11] has been discussed elsewhere.

5.2.2.1 Phytoplankton photosynthesis

Given the vital role of phytoplankton biology in aquatic carbon capture, understanding UV interactions with phytoplankton is of critical importance. Phytoplankton in the sea carry out about half of the photosynthesis on Earth. Phytoplankton photosynthesis reduces the partial pressure of carbon dioxide in the upper ocean and thereby promotes the absorption of C02 from the atmosphere. The organic carbon produced by photosynthesis forms the base of the marine food web. About 25% of the organic carbon produced by phytoplankton photosynthesis is exported from the upper ocean into intermediate and deep water [73]. On a global basis this "new production" is estimated to be about 11 to 16 Gt C per year [1 gigaton (Gt) of carbon equals 1015 g] and it is believed that most of this organic carbon is remineralized in the top km of the sea [73]. This mechanism for movement of carbon from the surface to deep ocean has been referred to as the "biological pump" and it has been shown that this process keeps atmospheric C02 concentrations about 150 to 200 ppmv lower than if there were no marine phytoplankton [73].

Here I briefly consider aspects of the direct effects of UV on phytoplankton that are relevant to carbon cycling. For more detailed considerations of UV interactions with phytoplankton see the chapters by Villafane et al. (Chapter 11) and Neale et al. (Chapter 4) in this book. Indirect effects include changes in trophic level interactions as well as in the biological availability of micronut-rients such as iron or in interactions with damaging reactive oxygen species. The latter are discussed elsewhere in the chapter as are the role of phytoplankton in the effects of UV on sulfur cycling.

A variety of studies have demonstrated that exposure to UV can directly inhibit photosynthesis in phytoplankton [11,12,45,74]. Studies in several differ ent locations have shown that reductions in current levels of solar UV-B result in enhanced primary production, and Antarctic experiments under the ozone hole demonstrated that primary production is inhibited by enhanced UV-B. For example, investigations of depth-integrated in situ phytoplankton productivity in austral spring of 1990 [75] indicated that productivity was reduced 6 to 12% inside the ozone hole in the Bellinghausen Sea compared with productivity outside the hole. On an annual basis, this range corresponds to an estimated annual productivity loss of 7 to 14 teragrams, which is 2 to 4% of production in the Antarctic marginal ice zone.

Action spectra describe the wavelength dependency of radiation in producing some biological or chemical response [8,12,76,77]. Action spectra can be used to estimate the biological impacts of UV changes that result with changes in the ozone layer as well as changes in location, time-of-day, season, and depth. The term "biological weighting function" (BWF) has been used to distinguish a type of action spectrum measured using polychromatic U V and visible radiation with a series of cutoff filters [12], as originally described by Rundel [78]. Unlike action spectra measured using monochromatic radiation [76], the Rundel approach helps take into account the fact that there are interactions between various part of the spectrum, such as photorepair of UV-B damage by UV-A radiation. The evaluation of action spectra for UV inhibition of phytoplankton photosynthesis also must take into account the dependence of photosynthesis on exposure, in particular whether reciprocity applies. The term "reciprocity" applies to systems in which biological or chemical responses to UV depend on cumulative exposure alone, independent of the duration of exposure or the irradiance [11,12,79]. Reciprocity does not apply to phytoplankton that rapidly repair UV damage. Instead, a steady state that reflects a balance between damage and repair is attained with continuous UV exposure [12,80,81]. This steady state can be described as a function of weighted irradiance. Elegant procedures for modeling these effects have been developed over the past decade [11,12,80,81]. Using these procedures, a recent study has shown that seasonally-averaged action spectra for phytoplankton inhibition by UV in a mid-latitude estuary (Rhode River in Maryland, USA) are remarkably similar to action spectra for photoinhibition of Antarctic phytoplankton (Figure 3) [80]. Interestingly, the action spectra observed in the mid-latitude studies did not exhibit large seasonal changes as was expected, but they did indicate major short-term changes that may be attributable to factors such as changes in species composition, nutrient availability, temperature, and light acclimation.

In addition to its direct inhibitory effects on production, UV may also be involved in indirect effects on oceanic new production by its effects on the biological availability and chemical reactivity of micronutrients, iron, manganese, and copper, in particular. UVR also strongly affects the production of ROS and such effects can be mediated through the reactions of ROS with trace metals such as copper and iron. The specific effects are discussed in more detail below.

5.2.2.2 Other UV interactions with carbon capture and storage

Ozone depletion may influence the ability of the ocean to take up atmospheric

Wavelength (nm)

Figure 3. Comparison of biological weighting functions for UV inhibition of photosynthesis for Rhode River (mid-latitude site, North America) and Antarctic phytoplankton. [Reprinted with permission from Banaszak and Neale [80], Figure 2, p. 597, Copyright 2001, The American Society for Limnology and Oceanography, Inc.]

Wavelength (nm)

Figure 3. Comparison of biological weighting functions for UV inhibition of photosynthesis for Rhode River (mid-latitude site, North America) and Antarctic phytoplankton. [Reprinted with permission from Banaszak and Neale [80], Figure 2, p. 597, Copyright 2001, The American Society for Limnology and Oceanography, Inc.]

C02, but the net impact of a reduction in primary production on the ocean sink for atmospheric C02 is uncertain. In addition to the factors influencing phytoplankton photosynthesis that were discussed earlier, aquatic circulation (Chapter 4), microbial cycling and photodegradation, macro- and micronutrient availability and other factors affect net carbon storage. Moreover, other indirect effects involving trophic level interactions may also affect ecosystem productivity [82]. For example, the vertical migration of zooplankton has been recently shown to be sensitive to UVR [83,84]. This finding has important implications for the flux of carbon through the microbial food web, which involves transfer of biomass from the primary producers to metazoa and bacteria. Sulfur cycling also may be affected by UV-induced changes in zooplankton grazing. Thus, the net impact on carbon capture is clearly not a linear function of increased UV exposure.

Moreover, it is likely that interactions with other global-scale environmental changes will affect the biological pump and other aspects of global biogeochemi-cal cycles. For example, changes in atmospheric circulation associated with climate change likely will affect aquatic mixing dynamics and thus the impact of UV on photosynthesis as well as decomposition (Figure 1). Moreover, recent remote sensing observations indicate that changes in thermocline depth in the tropical Pacific Ocean occur during El Niño/La Niña events that strongly influence new production and carbon export from the upper to deep ocean [85]. These changes can affect the impact of UV on phytoplankton photosynthesis and microbial decomposition as well as the air-sea exchange of gases. The unprecedented strength of the 1982/83 and 1997/98 El Niño events has indicated that they are becoming strongly influenced by anthropogenic activities and thus that strong El Niños may become more frequent in the future.

In the case of freshwaters, the past effects of climate change on UV exposure have impacted sedimentary records in a remarkable way. Analysis of fossil diatom assemblages in Canadian subarctic lake sediments has provided evidence of the interactive impacts of climate change and solar UVR on CDOM concentrations during the Holocene [86].

5.2.3 UV effects on decomposition

In Section 2.1.3 the effects of solar UVR on the optical properties of CDOM were discussed. Here the broader effects of UV on decomposition are considered. Detailed considerations of the effects of UVR on chemical and biological decomposition were initiated during the 1980s [63,87]. These early studies revealed that DOC plays a central, multifaceted role in aquatic photochemistry and photo-biology. DOC photoreacts to produce atmospherically-important trace gases and biologically-available carbon- and nitrogen-containing compounds, and to initiate free radical and photosensitized reactions that affect aquatic composition.

UV effects on decomposition of aquatic organic matter are caused by inhibition of microbial activity, by direct photodegradation of the CDOM and particulate organic carbon (POC) to C02 and other gases, and by UV-induced photodegradation of the persistent, polymeric components of the DOC to readily decomposable compounds. These effects are discussed in the following sections and then, to complete this section on carbon cycling, modeling and experimental techniques are described and then used to evaluate the role of UV-induced decomposition of organic matter in selected freshwater and marine environments.

5.2.3.1 Microbial photoinhibition

Bacterial activity is inhibited by UV-B radiation [10,56,88] and direct DNA damage (pyrimidine dimerization) has been demonstrated in field studies [89,90]. The greatest damage is observed in poorly-mixed, stratified waters. However, observations in lakes, coastal waters, and the Gulf of Mexico and modeling studies showed that the reduction in microbial activity is attenuated with increased winds and surface layer mixing and the activity is rapidly restored in the dark (within a few hours) via repair and regrowth [56,89,90]. A modeling study concluded that changes in UVR caused by ozone depletion can have a more serious net impact on bacterial activity than UV increases attributable to decreased CDOM concentrations [56]. The radiation amplification factor (RAF) for UV damage to bacterioplankton is close to that computed for generalized DNA damage (see Section 8) and thus DNA damage for these microorganisms must generally be more susceptible to ozone depletion than phytoplankton photosynthesis. However, like phytoplankton, bacterioplankton also can repair

DNA damage by photoreactivation and excision repair. The model indicated that damage caused by decreased CDOM is partly offset by increased photo-reactivation that is related to the modeled increase in repairing UV-A radiation caused by CDOM depletion [56]. It should be noted, however, that these results are a sensitive function of the assumed action spectra for damage and photoreactivation as well as the spectral slope coefficient for CDOM. The latter is considerably higher in seawater than the value of 0.014 that was assumed in the modeling study [13,14] and thus the study likely underestimated the dependence of changing CDOM concentrations on net UV damage. These studies indicate that adverse effects of UVR on microbial activity can change the timing and location of microbial decomposition of labile organic matter in the upper ocean. The amount and distribution of marine viruses also are affected by UV-B radiation in the sea [91,92]. Viruses can influence microbial diversity and activity, including decomposition. Light-induced repair of sunlight-damaged viruses, probably by photoreactivation, can be effected in the presence of bacteria [92].

As is the case for decomposers on terrestrial plant litter [6,72], UV exposure also affects bacterial and fungal growth on aquatic macrophyte detritus [93]. The effects are evidenced in part by changes in the attached microbial communities, which, for example, became dominated by bacteria in irradiated microcosms compared to shaded systems. Enzymatic activity of the microorganisms also was changed in UV-B irradiated systems, where significantly higher beta-glucosidase activity was observed [93].

5.2.3.2 Direct photodecomposition

In addition to its effects on microbial activity, solar UVR has direct effects on decomposition. A variety of recent studies have provided evidence that CDOM undergoes a complex array of other photoreactions that can involve a decrease in average molecular weight accompanied by cleavage to a variety of photo-products [3-5,8,24,25,68,70,74,94-118], changes in isotopic content [60,119] and consumption of oxygen [59,68,106,116]. These reactions include the direct photochemical mineralization of the CDOM to carbon monoxide and dissolved inorganic carbon (DIC). Of these various direct pathways, the photoproduction of DIC is most efficient.

It has long been known that intense short-wavelength UVR can mineralize DOC [3,7,8]. Mineralizations achieved under such extreme conditions, however, are irrelevant to natural conditions. Only very recently have several reports indicated that DOC in freshwaters and seawater can be directly mineralized on exposure to sunlight [3,8,59,70,106,110-115,118]. Miles and Brezonik [106] were first to report this reaction in a natural freshwater system. They presented evidence that this process included photoreactions of DOC-iron complexes. More recent studies have provided a more detailed understanding of photoreactions involving iron in the natural photooxidation of DOC [59,107-109]. Appreciable iron concentrations are sometimes found in high-DOC, acidic fresh-waters and iron can be introduced into the sea via riverine inputs, wet deposition, and deposition of Aeolian dust. It seems likely that future research will demonstrate an important role for iron in enhancing photochemical mineralization of

DOC in natural waters. However, other pathways for DIC photoproduction that do not involve iron must also be available [59,70]. Whatever the mechanism for DIC photoproduction, recent studies have shown that this process potentially could account for the global annual production of anywhere from 1 to 12 Gt of C02 C in the ocean [3,118].

In addition to effects on DOM, UV exposure also impacts the decomposition of POC [120]. Photoproduction of DIC has been observed from the sterilized detritus of several aquatic macrophytes in both air and immersed in water [120]. The highest production rates were observed in water. Although UVR was most effective at inducing detritus decomposition, visible light also played a role.

Carbon monoxide (CO) is also formed in aquatic environments from the photochemical degradation of DOM [3,4,8,22,94-105], Strong gradients of CO have been observed in the lowest 10 metres of the atmosphere over the Atlantic Ocean [97]. The samples nearest the ocean surface were some 50 ppb higher than at the 10-metre altitude-sampling inlet. This implies that the ocean is a source of CO to the atmosphere and that this source can increase the atmospheric concentration. CO is reactive in the troposphere and thus its emissions from the ocean may influence the hydroxyl radical (OH) and ozone concentrations in the marine atmospheric boundary layer that is remote from strong continental influences.

Although the sea is thought to be a net source of CO, this source has been subject to a wide range of estimates. The most recent estimate has come from Zafiriou et al. [105] who, based on modeled results derived using CO quantum yields that were measured using Pacific water samples during 1994, concluded that the global open ocean photochemical source of CO is approximately 50 ± 10 Tg CO carbon per year. An approximate estimate of the coastal ocean source was about 10 Tg CO carbon annually. Most of this CO production was estimated to be consumed by microorganisms rather than escaping to the atmosphere. The microbial sink estimates were based on a series of incubations that quantified CO loss in freshly collected seawater samples. Previous, much higher estimates of CO photoproduction in the sea were based primarily on CO photoproduction from terrestrially-derived CDOM [8,105]. As shown in Figure

300 320 340 360 380 400 420 440 460 Wavelength, nm

Figure 4. Comparison of apparent quantum yield spectra for production of CO from terrestrially-derived CDOM (•) [59,103] and from CDOM in the open ocean (O) [105].

300 320 340 360 380 400 420 440 460 Wavelength, nm

Figure 4. Comparison of apparent quantum yield spectra for production of CO from terrestrially-derived CDOM (•) [59,103] and from CDOM in the open ocean (O) [105].

4, apparent quantum yields for terrestrially-derived CDOM are much larger than those observed with open ocean water, especially in the UV-A and visible spectral regions. This difference in quantum yields, which largely accounts for the lower estimated fluxes in the open ocean, suggests that the photoreactivity of bluewater, algal-derived CDOM may be quite different from that of terrestrially-derived CDOM. The open ocean CO fluxes estimated by Zafiriou et al. [105] agree approximately with earlier estimates of Bates et al. [104], but are at the lower end of global flux estimates based on extensive studies during the 1980s of CO emissions in the Atlantic Ocean by Conrad and co-workers [100] and of CO emissions in the Pacific by Gammon and Kelley [102]. The differences in these estimates may reflect the fact that CO concentrations exhibit great spatial and temporal variability. But they may also reflect periodic large-scale changes in the nature of the CDOM in the upper ocean, related to El Niño/La Niña events [46,54,85]. For example, the higher CO fluxes reported by Gammon and Kelley [102] were based on Pacific observations during the 1987-1988 El Niño event, whereas the estimates of Zafiriou et al. [105] were based on observations during 1994 when El Niño conditions were not prevalent.

5.2.3.3 UV effects on lability of microbial substrates

UVR can potentially affect carbon cycling through modification of the biological availability of microbial substrates and microbial activity [4,6,8,9,38,67,112,116,121-138]. This effect, referred to here as photochemically-altered microbial degradation, is well documented in the case of terrestrially-derived DOC, where stimulation of microbial activity is usually observed. The relative importance of this pathway, compared to direct photodegradation, seems to strongly depend on the DOC source. Initial comparisons using early quantitative studies of identifiable biologically labile photoproducts indicated that other carbon photoproducts such as DIC are produced at rates many-fold higher than biologically-available photoproducts (BLPs) [8,9], The photoproduction of BLPs has been quantified using microbial growth indicators (e.g., uptake of tritiated leucine), or cumulative bacterial oxygen consumption during post irradiation-incubation as an index (e.g., respiratory activity per absorbed photon). Recent studies, however, indicate that the quantum yields for formation of BLPs in coastal waters of the Southeastern United States are of the same order of magnitude as that for DIC photoproduction [67,125]. Likewise, direct and photochemically-stimulated microbial decomposition of the DOC from the Adriatic Sea and coastal North Sea were estimated to be approximately equivalent [128].

Not all results are consistent with a stimulating effect, however [128,132-137]. Surface water DOM in the open ocean [131,132] and a subtropical seagrass meadow [133] were not sources of biologically labile photoproducts. In the case of the seagrass meadow, BLPs were produced by exposure to solar radiation but the effect was attributed to algal exudates [133] not DOC photodecomposition. Photoreactions can reduce the microbial availability of certain organic substrates such as peptone and algal exudates [134-136], possibly via light-induced cross-linking between the CDOM and algal exudates [135]. Decreased bacterial activity also was observed on the leachate from UV-exposed detritus from vascular plants and this effect was attributed to decreased availability of the DOM and possibly the release of inhibitory substances from the detritus [137].

5.2.3.4 Modeling UV-induced decomposition

Large-scale models provide a useful technique for estimating global-scale fluxes through carbon pools and how environmental changes such as ozone depletion affect the fluxes. Models of the effects of UVR on decomposition and trace gas production require equations based on field or laboratory measurements under varying natural conditions and/or experimental manipulations that relate rates of UV-induced processes to changing environmental parameters. Time series observations of various indicators of UV effects such as atmospheric and aquatic concentrations of trace gases, aquatic CDOM concentrations and UV absorption coefficients, and isotopic and lignin composition of DOC can provide large-scale integrators of changes in biogeochemical cycling in aquatic ecosystems and serve as checks on fluxes inferred from models.

Various approaches have been used to model the rates of photoreactions in aquatic environments [3-6,8-10,67,68,139-141]. To illustrate the utility of the modeling approach, the wavelength dependence for CO photoproduction fluxes simulated for a mid-latitude location is shown in Figure 5. Quantum yield spectra that are shown in Figure 4 were used in these calculations. The equations and assumptions involved in these calculations are discussed in more detail in

Wavelength, nm

Figure 5. Spectral dependence of potential fluxes of CO photoproduction from terrestrially-derived CDOM (solid line) and from CDOM in the Pacific Ocean (broken line), computed using equation (8), the data in Figure 6 and the TUV model of Madronich [235]. Computed for late July, midday at equator; integrated fluxes are: terrestrially-derived, 13 nanomoles m~2 s-1; Pacific, 2.4 nanomoles m~2 s-1 (integrated over 290450 nm).

Wavelength, nm

Figure 5. Spectral dependence of potential fluxes of CO photoproduction from terrestrially-derived CDOM (solid line) and from CDOM in the Pacific Ocean (broken line), computed using equation (8), the data in Figure 6 and the TUV model of Madronich [235]. Computed for late July, midday at equator; integrated fluxes are: terrestrially-derived, 13 nanomoles m~2 s-1; Pacific, 2.4 nanomoles m~2 s-1 (integrated over 290450 nm).

the Appendix. The results demonstrate the much larger area under the plot for terrestrially-derived CDOM, and thus much higher depth-integrated photoproduction, compared to that for open ocean CDOM. Longer wavelength UV is much more heavily involved in photoproduction of CO from the terrestrial organic matter. The difference may be due to an inherent difference in the source of the CDOM (derived from microbial processing of dead organic matter from terrestrial plants vs. phytoplankton) and/or to extensive photooxidation of the open ocean CDOM, decomposing the more reactive components and leaving less reactive substances in the residual CDOM.

5.2.3.5 UV-induced oxidation in coastal areas

Marine scientists have long puzzled over the fate of riverine organic matter on entry to the ocean [69,71]. Isotopic studies indicate that the DOC in the open ocean is primarily of marine origin [142,143], although some terrestrial character would have been expected. Models have been used to evaluate the potential loss of organic matter in coastal regions.

Quantum yield spectra determined for BLP production in coastal waters can be used in conjunction with simulated solar spectral irradiance to estimate potential seasonal and annual BLP fluxes as a function of latitude (Figure 6) [67]. Potential annual consumption of terrestrially-derived organic matter in coastal areas can then be calculated as the cross-product of the annual fluxes and coastal ocean areas for various 10° latitude bands. Using this approach it was estimated that most coastal BLP formation potentially occurs in the Northern Hemisphere, with substantial contributions from high-latitude coastal regions [67]. The estimated potential annual production of biologically labile photo-products from coastal regions worldwide is 253 x 1012 g C, a value that approximately corresponds to the annual global input of riverine DOM (220 x 1012 g C [69]. Using quantum yield spectra for photochemical oxygen demand as a

Month

Figure 6. Seasonal changes estimated for potential biologically labile photoproduct production from terrestrially-derived CDOM at various latitudes in the Northern Hemisphere [67].

Month

Figure 6. Seasonal changes estimated for potential biologically labile photoproduct production from terrestrially-derived CDOM at various latitudes in the Northern Hemisphere [67].

means to estimate efficiencies for direct photooxidation of CDOM, it was computed that, on an annual basis, the direct photooxidation of CDOM would potentially consume about 3.5 x 1014 moles 02 yr-1 in the global ocean [68]. Assuming that coastal waters make up about 7% of the ocean's area and that 0.5 moles of 02 are consumed per altered DOC carbon, this estimate corresponds to a potential consumption of 600 x 1012 g C per year in coastal regions.

These modeling results indicate that direct photooxidation combined with photochemically-stimulated microbial degradation can potentially consume about 850 x 1012 g C per year in coastal areas. The results indicate that even high-latitude coastal DOC is subject to major UV-induced oxidation in the latitudes of its input to the ocean. These likely are overestimates of the role of photochemistry in oxidizing terrestrially-derived organic matter. As noted earlier, these considerations apply only to the photoreactive component of the DOM (i.e., the CDOM). As noted above, riverine inputs of DOM appear to include a substantial non-reactive component. Estimates for the open ocean are clouded at this point by the previously-discussed findings that bluewater CDOM may be less photoreactive than terrestrially-derived CDOM and much less efficient in the net photoproduction of BLPs.

Other recent observations are consistent with the modeling results. The <513C isotopic composition of DOC in the open ocean, — 20%o to — 21%o, is about the same is that found in marine plankton, but is significantly (5% to 7%) more positive than that of freshwater (terrestrially-derived) DOC [142,143]. Thus, organic carbon created originally by phytoplankton appears to be the primary source of the DOC in the open ocean, even though riverine inputs of terrestrially-derived DOC should be evident. These comparisons suggest that a large sink for the terrestrially-derived DOC must exist in the ocean or perhaps in estuaries [69,71], If photooxidation plays an important role in this sink activity, as predicted by the models, then UV exposure of freshwater DOC should increase its c)13C. Recent studies have shown that this is indeed the case. Exposure of terrestrially-derived CDOM in water from coastal estuaries [60] or in mid-latitude lakes [119] resulted in significant changes in stable isotope composition: the <513C of DIC that was produced was isotopically "light" relative to the initial DOC, leaving a residual fraction of DOC that was isotopically "heavy" (Figure 7). It is likely that this effect involves photochemically-stimulated microbial degradation. Seasonal changes in the isotopic signature of the DOM in the upper layers of lakes also have been attributed to photooxidation [119]. These results are consistent with the model predictions and suggest that UV-induced photooxidation plays a role in the observed changes in DOC isotopic composition that occur in estuaries [144].

In addition to changes in stable isotope composition, photoreactions also cause decreases in lignin content of DOC [21,145,146]. Such decreases, especially loss (compared to freshwater DOC) of photochemically-sensitive molecular indicators such as syringyl phenols [142], also are observed in coastal DOC samples, indicating that UV-induced oxidation of part of the terrestrial DOC is rapid.

Time (days)

Figure 7. Changes in the stable isotopic composition of CDOM in a coastal river water sample on exposure to natural solar radiation . The change is consistent with observed changes in DOC isotopic composition that occur when terrestrially-derived DOM moves from land into the sea.

0 0

Post a comment