Origin, enzymatic response and fate of dissolved organic matter during flood and non-flood conditions in a river-floodplain system of the Danube (Austria)

Spectroscopic techniques and extracellular enzyme activity measurements were combined with assessments of bacterial secondary production (BSP) to elucidate flood-pulse-linked differences in carbon (C) sources and related microbial processes in a river-floodplain system near Vienna (Austria). Surface connection with the main channel significantly influenced the quantity and quality of dissolved organic matter (DOM) in floodplain backwaters. The highest values of dissolved organic carbon (DOC) and chromophoric DOM (CDOM) were observed during the peak of the flood, when DOC increased from 1.36 to 4.37 mg l^?1 and CDOM from 2.94 to 14.32 m^?1. The flood introduced DOC which consisted of more allochthonously-derived, aromatic compounds. Bacterial enzymatic activity, as a proxy to track the response to changes in DOM, indicated elevated utilization of imported allochthonous material. Based on the enzyme measurements, new parameters were calculated: metabolic effort and enzymatic indices (EEA 1 and EEA 2). During connection, bacterial glucosidase and protease activity were dominant, whereas during disconnected phases a switch to lignin degradation (phenol oxidase) occurred. The enzymatic activity analysis revealed that flooding mobilized reactive DOM, which then supported bacterial metabolism. No significant differences in overall BSP between the two phases were detected, indicating that heterogeneous sources of C sufficiently support BSP. The study demonstrates that floods are important for delivering DOM, which, despite its allochthonous origin, is reactive and can be effectively utilized by aquatic bacteria in this river-floodplain systems. The presence of active floodplains, characterized by hydrological connectivity with the main channel, creates the opportunity to process allochthonous DOC. This has potential consequences for carbon flux, enhancing C sequestration and mineralization processes in this river-floodplain system.     

Effect of organic matter source and salinity on dissolved organic matter isolation via ultrafiltration and solid phase extraction

In this study, samples were taken from three contrasting freshwater sources and amended with salt in order to determine the influence of salinity and dissolved organic matter (DOM) composition on DOM recovery via ultrafiltration and solid phase extraction (SPE) with C₁₈ disks. Salt addition caused variable recovery of DOM when using C₁₈ SPE, and ultraviolet–visible spectroscopic characterization of the extracted material showed spectral responses that varied among sample sources. In contrast, increasing sample salinity from 0 to 30 ppt consistently caused a 15–25% reduction in the amount of high molecular weight DOM isolated by ultrafiltration for both dissolved organic carbon (DOC) and chromophoric DOM (CDOM), regardless of DOM composition. We hypothesize that a change in conformation (such as coiling or disaggregation) of DOM molecules occurs in the presence of salt, allowing them to pass through the ultrafiltration membrane and thereby decreasing the DOM retained by ultrafiltration. These results are important because they demonstrate that changes in salinity can influence DOM recovery in estuaries. Interpretation of DOM characteristics along estuarine gradients needs to account for potential artifacts introduced by sample isolation techniques.     

Distributions and characteristics of dissolved organic matter in temperate coastal waters (Southern North Sea)

The spatial and temporal distributions of chromophoric dissolved organic matter (CDOM) and dissolved organic carbon (DOC) was studied in the East-Frisian Wadden Sea (Southern North Sea) during several cruises between 2002 and 2005. The spatial distribution of CDOM in the German Bight shows a strong gradient towards the coast. Tidal and seasonal variations of dissolved organic matter (DOM) identify freshwater discharge via flood-gates at the coastline and pore water efflux from tidal flat sediments as the most important CDOM sources within the backbarrier area of the Island of Spiekeroog. However, the amount and pattern of CDOM and DOC is strongly affected by various parameters, e.g. changes in the amount of terrestrial run-off, precipitation, evaporation, biological activity and photooxidation. A decoupling of CDOM and DOC, especially during periods of pronounced biological activity (algae blooms and microbial activity), is observed in spring and especially in summer. Mixing of the endmembers freshwater, pore water, and open sea water results in the formation of a coastal transition zone. Whilst an almost conservative behaviour during mixing is observed in winter, summer data point towards non-conservative mixing.     

Diurnal variability in riverine dissolved organic matter composition determined by in situ optical measurement in the San Joaquin River (California,USA)

Dissolved organic matter (DOM) concentration and composition in riverine and stream systems are known to vary with hydrological and productivity cycles over the annual and interannual time scales. Rivers are commonly perceived as homogeneous with respect to DOM concentration and composition, particularly under steady flow conditions over short time periods. However, few studies have evaluated the impact of short term variability (<1 day) on DOM dynamics. This study examined whether diurnal processes measurably altered DOM concentration and composition in the hypereutrophic San Joaquin River (California) during a relatively quiescent period. We evaluated the efficacy of using optical in situ measurements to reveal changes in DOM which may not be evident from bulk dissolved organic carbon (DOC) measurement alone. The in situ optical measurements described in this study clearly showed for the first time diurnal variations in DOM measurements, which have previously been related to both composition and concentration, even though diurnal changes were not well reflected in bulk DOC concentrations. An apparent asynchronous trend of DOM absorbance and chlorophyll‐a in comparison to chromophoric dissolved organic matter (CDOM) fluorescence and spectral slope S_290–350 suggests that no one specific CDOM spectrophotometric measurement explains absolutely DOM diurnal variation in this system; the measurement of multiple optical parameters is therefore recommended. The observed diurnal changes in DOM composition, measured by in situ optical instrumentation likely reflect both photochemical and biologically‐mediated processes. The results of this study highlight that short‐term variability in DOM composition may complicate trends for studies aiming to distinguish different DOM sources in riverine systems and emphasizes the importance of sampling specific study sites to be compared at the same time of day. The utilization of in situ optical technology allows short‐term variability in DOM dynamics to be monitored and serves to increase our understanding of its processing and fundamental role in the aquatic environment. Copyright © 2007 John Wiley & Sons, Ltd.     

Processing of humic-rich riverine dissolved organic matter by estuarine bacteria: effects of predegradation and inorganic nutrients

The bioavailability of predegraded dissolved organic matter (DOM) from a humic-rich, boreal river to estuarine bacteria from the Baltic Sea was studied in 39-day bioassays. The river waters had been exposed to various degrees of bacterial degradation by storing them between 0 and 465 days in dark prior to the bioassay. The resulting predegraded DOM was inoculated with estuarine bacteria and the subsequent changes in DOM quantity and quality measured. During the incubations, dissolved organic carbon (DOC) and oxygen concentrations decreased, indicating heterotrophic activity. Coloured DOM was degraded less than DOC, indicating a selective utilization of DOM, and humic-like fluorescence components increased during the incubations. The amount of DOC degraded was not affected by the length of DOM predegradation. The percentage of bioavailable DOC (%BDOC) was higher in experiment units with added inorganic nitrogen and phosphorus than without addition (on average 13.5 % and 9.0, respectively), but had no effect on the degradation of fresh, non-predegraded, DOC (%BDOC 12.0 %). Bacterial growth efficiency (BGE) was highest (65 ± 2 %) in the units with fresh DOM, and lowest in units with predegraded DOM and no added inorganic nutrients (11 ± 4 %). The addition of inorganic nutrients increased the BGE of predegraded DOM units by an average of 28 ± 4 %. There was no significant effect on BGE by length of predegradation after the initial drop (<3 months). This study suggests that both the length of predegradation and the inorganic nutrient status in the receiving estuary has consequences to carbon cycling and will determine the amount of terrestrial-derived DOC being ultimately assimilated into marine food webs.     

The freshwater dissolved organic matter fluorescence–total organic carbon relationship

The fluorescent properties of dissolved organic matter (DOM) enable comparisons of humic‐like (H‐L) and fulvic‐like (F‐L) fluorescence intensities with dissolved organic carbon (DOC) in aquatic systems. The fluorescence‐DOC relationship differed in gradient, i.e. the fluorescence per gram of carbon, and in the strength of the correlation coefficient. We compare the fluorescence intensity of the F‐L and H‐L fractions and DOC of freshwater DOM in north Shropshire, England, featuring a river, wetland, spring, pond and sewage DOM sources. Correlations between fluorescence and DOC varied between sample sites. Wetland water samples for the F‐L peak gave the best correlation, r = 0·756; the lowest correlation was from final treated sewage effluent, r = 0·167. The relationship between fluorescence and DOC of commercially available International Humic Substances Society standards were also examined and they generally showed a lower fluorescence per gram of carbon for the F‐L peak than the natural samples, whereas peat wetland DOM gave a greater fluorescence per gram of carbon than river DOM. Here, we propose the strength of the fluorescence–DOC correlation to be a useful tool when discriminating sources of DOM in fresh water. Copyright © 2007 John Wiley & Sons, Ltd.     

Photochemical degradation of dissolved organic matter from streams in the western Lake Superior watershed

The input and fate of dissolved organic matter (DOM) can have important consequences for coastal zone productivity in large lakes and oceans. Chromophoric DOM (CDOM) is often delivered to coastal zones from rivers and streams and affects light penetration in a water column. CDOM can protect biota from damaging ultraviolet (UV) light by acting as sunscreen, resulting in increased ecosystem productivity. Alternatively, CDOM can decrease ecosystem productivity by absorbing light needed for photosynthesis and forming photoreaction products that are harmful to coastal zone biota. Increased urbanization of watersheds and seasonal differences in weather patterns change the delivery pathways, reactivity, input, and energy flow of DOM (and its CDOM component) into aquatic systems. This study investigated the effects of watershed and season on the concentrations and potential photodegradation of stream-derived DOM in Lake Superior tributaries, chosen to be geographically and geologically similar but differing in land use. Organic carbon analysis, UV–Visible spectrophotometry, and terrestrial (land use) analysis were used to investigate differences among samples and sample treatments. The major differences in DOM concentration and photochemical response appeared seasonal rather than site specific, with snow-melt samples showing stronger and more consistent changes in UV–Visible parameters while base-flow samples showed stronger and more consistent losses in DOC.     

Bioavailability and radiocarbon age of fluvial dissolved organic matter (DOM) from a northern peatland-dominated catchment: effect of land-use change

The radiocarbon age and biodegradability of dissolved organic matter (DOM) from a northern peat-dominated river system was studied and the effects of land-use were compared. Samples were obtained from streams and ditches comprising sub-catchments of the Kiiminki River, Northern Finland. Sample sites included areas of natural mire, areas subjected to moderate disturbance (ditching to enhance forestry), and areas subjected to serious land use change (agriculture and peat excavation). The study employed a 55 day bioassay that measured the biodegradation potential of surface-water DOM. We identified release of modern (mean 6–13 year old) DOM from natural sites, and material aged up to 1,553 years from disturbed sites. The proportion of biodegradable DOC ranged from 4.1 to 17.9 %, and bacterial DOC removal was modelled using twin-pool and reactivity-continuum (beta distribution) approaches. Bacterial growth efficiency ranged from 0.11 to 0.26 between areas of different land use, and these relatively low values reflect the humic-rich DOM released from boreal peatland. Despite the range of land-use types studied, including intensive peatland excavation areas, there was no detectable relationship between the biological lability of DOM and its radiocarbon age.     

Variation in the photochemical lability of dissolved organic matter in a large boreal watershed

Boreal watersheds contain a vast quantity of terrestrially derived dissolved organic matter (DOM) originating from wetland and forest soils, yet variation in the potential for photochemical transformation of boreal aquatic DOM sources remains poorly understood. Laboratory solar radiation exposure experiments were conducted on DOM samples collected in three seasons, across nine sites, representing contrasting catchment composition and watershed position to assess variation in the photochemical lability of boreal DOM source and stable carbon isotopic signature (δ¹³C) of photomineralized DOM. Dissolved organic carbon (DOC) loss rates during laboratory exposure were lowest in summer, suggesting that DOM may have been more photo-degraded during summer. DOM from upstream portions of forested stream sites and wetland-influenced sites was more photolabile relative to downstream portions and the river DOM, suggesting potential losses in photolabile DOM downstream and in the lower reaches of the watershed. Increased a₂₅₄:a₃₅₀ and spectral slope following sample exposure suggest photoproduction of low molecular weight (LMW) CDOM and/or a higher photoreactivity of high molecular weight versus LMW compounds. Photomineralization of nitrogen was regulated by organic nitrogen concentration and resulted in NH₄ ⁺-photoproduction rates between 0.01 and 0.3?μM N?h^?1 and ecologically significant increases in NH₄ ⁺ for these waters. The δ¹³C of the photomineralized DOM was positively correlated to initial DOC concentration and generally lower when initial DOC concentrations were lower, suggesting variation in photomineralized DOM δ¹³C may be a result of kinetic isotope fractionation. Results from this study demonstrate significant variation in the photochemical lability of boreal watershed sources of DOM. Such variation suggests landscape and environmental change has the potential to alter the biogeochemical role photochemical transformations play in downstream portions of boreal watersheds.     

太湖有色可溶性有机物(CDOM)对COD及BOD5的指示意义

化学需氧量(COD)、五日生化需氧量(BOD_5)及溶解性有机碳(DOC)是指示湖泊水质的重要指标,然而上述指标测定通常耗费大量时间、试剂及人力物力且排放大量有害废液.有色可溶性有机物(CDOM)是溶解性有机物(DOM)中可以强烈吸收光谱中的紫外光和可见光的部分,数据测定耗时短、方便快捷,且样品处理过程环境友好,能在很大程度上反映湖泊水质.本研究基于2016年2、5和8月在太湖均匀布设的32个采样点进行样品采集,运用光谱吸收与三维荧光-平行因子分析(EEMs-PARAFAC)探究太湖CDOM的光谱吸收和荧光组分,探讨CDOM光谱指标对湖泊BOD_5、COD及DOC浓度等湖泊环境质量指标的可替代性.结果表明:(1)运用EEMs-PARAFAC方法解析出3种荧光组分:类腐殖酸C1、类酪氨酸C2和类色氨酸C3.(2) COD和BOD_5和DOC在空间上呈现出相似的分布趋势,不同水期的最高值均出现在竺山湾和梅梁湾,由西北湖区至中部敞水区、东南湖湾递减.(3)在不同水期,COD、BOD_5、DOC浓度和C1组分均表现为丰水期极显著大于枯水期和平水期,a_(254)在丰、平、枯水期间无显著性差异,最大值出现在丰水期;C2与C3组分均在枯水期和平水期极显著大于丰水期.(4)在不同水文时期,COD、BOD_5和DOC浓度均与a_(254)、类腐殖酸C1呈显著正相关,丰水期太湖COD、BOD_5和DOC浓度与CDOM光谱指标的线性相关性要优于枯水期和平水期.(5) CDOM光谱指标在不同水文时期均能很好地替代COD、BOD_5和DOC等作为反映太湖水体中有机物污染程度及湖泊水质的指标.     

Coastal blue carbon: Concept,study method,and the application to ecological restoration

Coastal blue carbon refers to the carbon taken from atmospheric CO2; fixed by advanced plants(including salt marsh,mangrove, and seagrass), phytoplankton, macroalgae, and marine calcifiers via the interaction of plants and microbes; and stored in nearshore sediments and soils; as well as the carbon transported from the coast to the ocean and ocean floor. The carbon sequestration capacity per unit area of coastal blue carbon is far greater than that of the terrestrial carbon pool. The mechanisms and controls of the carbon sink from salt marshes, mangroves, seagrasses, the aquaculture of shellfish and macroalgae, and the microbial carbon pump need to be further studied. The methods to quantify coastal blue carbon include carbon flux measurements, carbon pool measurements, manipulative experiments, and modeling. Restoring, conserving, and enhancing blue carbon will increase carbon sinks and produce carbon credits, which could be traded on the carbon market. The need to tackle climate change and implement China's commitment to cut carbon emissions requires us to improve studies on coastal blue carbon science and policy. The knowledge learned from coastal blue carbon improves the conservation and restoration of salt marshes,mangroves, and seagrasses; enhances the function of the microbial carbon pump; and promotes sustainable aquaculture, such as ocean ranching.     

Distribution and photoreactivity of chromophoric dissolved organic matter in northern Gulf of Mexico shelf waters

The distribution and photoreactivity of chromophoric dissolved organic matter (CDOM) in the northern Gulf of Mexico along the Louisiana coastal shelf were examined during three cruises in summer 2007, fall 2007, and summer 2008. The influence of the Mississippi River plume was clearly evident as CDOM levels (defined as a₃₀₅) and dissolved organic carbon (DOC) concentrations were well-correlated with salinity during all cruises. Elevated CDOM and CDOM:DOC ratios of surface samples collected offshore of Atchafalaya Bay and the Breton-Chandeleur Sound complex indicated emanations of organic-rich waters from coastal wetlands are also an important source to nearshore shelf waters. Generally, CDOM and DOC levels were highest in surface waters and decreased with depth, but during summer 2007 and summer 2008, CDOM levels in near-bottom samples were occasionally higher than at mid-depths without concomitant increases in DOC. CDOM photobleaching was measured during 24 irradiations using a SunTest XLS+ solar simulator with photobleaching rate coefficients (k₃₀₅) ranging from 0.011 to 0.32 h⁻¹. For fall 2007 and summer 2008, higher k₃₀₅ values were generally observed in samples with higher initial CDOM levels. However, samples collected during summer 2007 did not exhibit a similar pattern nor were there differences in photobleaching rates between surface and bottom samples. Spectral slope coefficients (S_275-295 or S_350-400) and DOC levels were largely unchanged after 24 h irradiations. Modeled CDOM photobleaching for northern Gulf of Mexico mid-shelf waters predicts that during the summer when solar irradiance is high and the water column becomes stratified, nearly 90% of the CDOM in the upper 1 m may be lost to photobleaching, with losses up to 20% possible even at 10 m depth.     

Organically Bound Nutrients in Dissolved Organic Matter Fractions in Seepage and Pore Water of Weakly Developed Forest Soils

Previous field and laboratory studies showed that organically bound nutrients can contribute largely to the export of N, P, and S from soil into aquatic systems. One possible determinant for the losses of dissolved organic nutrients leaving the soil environment could be their distribution between dissolved organic matter (DOM) fractions of different mobility in soil. To elucidate the potential influence of DOM fractions under varying flow conditions on the vertical translocation of organically bound nutrients, we determined the concentrations and fluxes of dissolved organic C (DOC) and nutrients (DON, DOP, DOS) in soil water under a Scots pine (Pinus sylvestris L.) and a European beech (Fagus sylvatica L.) forest. We sampled seepage water from the organic forest floor layer and the mineral subsoil using zero‐tension lysimeters and soil pore water using tension lysimeters and suction cups. DOM in soil water was fractionated into hydrophilic and hydrophobic compounds by XAD‐8 at pH 2. We found that the organic forest floor layers were large sources for DOC, DON, DOP, and DOS. The dissolved organic nutrients were mainly concentrated in the hydrophilic DOM fraction which proved to be more mobile in mineral soil pore water than the hydrophobic one. Consequently, the concentrations and fluxes of dissolved organic nutrients decreased less with depth than those of DOC. Concentrations as well as fluxes in subsoil pore water of DOC and dissolved organic nutrients in the studied weakly developed soils were high as compared with literature data on deeply developed forest soils. Under conditions of rapid water flow through the strongly structured mineral soil at the beech site, almost no retention of DOM took place and thus the influence of the distribution of organically bound nutrients between the DOM fractions on the export of DON, DOP, and DOS was negligible.     

Changes in dissolved organic matter (DOM) amount and composition along nested headwater stream locations during baseflow and stormflow

Amount and composition of dissolved organic matter (DOM) were evaluated for multiple, nested stream locations in a forested watershed to investigate the role of hydrologic flow paths, wetlands and drainage scale. Sampling was performed over a 4‐year period (2008–2011) for five locations with drainage areas of 0.62, 3.5, 4.5, 12 and 79 ha. Hydrologic flow paths were characterized using an end‐member mixing model. DOM composition was determined using a suite of spectrofluorometric indices and a site‐specific parallel factor analysis model. Dissolved organic carbon (DOC), humic‐like DOM and fluorescence index were most sensitive to changes with drainage scale, whereas dissolved organic nitrogen, specific UV absorbance, Sr and protein‐like DOM were least sensitive. DOM concentrations and humic‐like DOM constituents were highest during both baseflow and stormflow for a 3.5‐ha catchment with a wetland near the catchment outlet. Whereas storm‐event concentrations of DOC and humic DOM constituents declined, the mass exports of DOC increased with increasing catchment scale. A pronounced dilution in storm‐event DOC concentration was observed at peak stream discharge for the 12‐ha drainage location, which was not as apparent at the 79‐ha scale, suggesting key differences in supply and transport of DOM. Our observations indicate that hydrologic flow paths, especially during storms, and the location and extent of wetlands in the catchment are key determinants of DOM concentration and composition. This study furthers our understanding of changes in DOM with drainage scale and the controls on DOM in headwater, forested catchments. Copyright © 2014 John Wiley & Sons, Ltd.     

Dissolved organic matter variations in coastal plain wetland watersheds: The integrated role of hydrological connectivity,land use,and seasonality

Dissolved organic matter (DOM) is integral to fluvial biogeochemical functions, and wetlands are broadly recognized as substantial sources of aromatic DOM to fluvial networks. Yet how land use change alters biogeochemical connectivity of upland wetlands to streams remains unclear. We studied depressional geographically isolated wetlands on the Delmarva Peninsula (USA) that are seasonally connected to downstream perennial waters via temporary channels. Composition and quantity of DOM from 4 forested, 4 agricultural, and 4 restored wetlands were assessed. Twenty perennial streams with watersheds containing wetlands were also sampled for DOM during times when surface connections were present versus absent. Perennial watersheds had varying amounts of forested wetland (0.4–82%) and agricultural (1–89%) cover. DOM was analysed with ultraviolet–visible spectroscopy, fluorescence spectroscopy, dissolved organic carbon (DOC) concentration, and bioassays. Forested wetlands exported more DOM that was more aromatic‐rich compared with agricultural and restored wetlands. DOM from the latter two could not be distinguished suggesting limited recovery of restored wetlands; DOM from both was more protein‐like than forested wetland DOM. Perennial streams with the highest wetland watershed cover had the highest DOC levels during all seasons; however, in fall and winter when temporary streams connect forested wetlands to perennial channels, perennial DOC concentrations peaked, and composition was linked to forested wetlands. In summer, when temporary stream connections were dry, perennial DOC concentrations were the lowest and protein‐like DOM levels the highest. Overall, DOC levels in perennial streams were linked to total wetland land cover, but the timing of peak fluxes of DOM was driven by wetland connectivity to perennial streams. Bioassays showed that DOM linked to wetlands was less available for microbial use than protein‐like DOM linked to agricultural land use. Together, this evidence indicates that geographically isolated wetlands have a significant impact on downstream water quality and ecosystem function mediated by temporary stream surface connections.     

A Preliminary Study of Chromophoric Dissolved Organic Matter (CDOM) in Lake Taihu,a Shallow Subtropical Lake in China

The optical properties and spatial distribution of chromophoric dissolved organic matter (CDOM) in Meiliang Bay of Lake Taihu were evaluated and compared to the results in literature. Concentrations of dissolved organic carbon (DOC) ranged from 8.75 to 20.19 mg L^?1 with an average of (13.10 ± 3.51) mg L^?1. CDOM absorption coefficients a(λ) at 280 nm, 355 nm, and 440 nm were in the range 11.28...33.46 m^?1 (average (20.95 ± 5.52) m^?1), 2.42...7.90 m^?1 (average (4.92 ± 1.29) m^?1), and 0.65...2.44 m^?1 (average (1.46 ± 0.44) m^?1), respectively. In general, CDOM absorption coefficient and DOC concentration were found to decrease away from the river inflow to Meiliang Bay towards the lake center. The values of the DOC‐specific absorption coefficients a*(λ), given as absorption coefficient related to mass concentration of organic carbon (C) ranged from 0.28 to 0.47 L mg^?1 m^?1 at 355 nm. The determination coefficients between CDOM absorption and DOC concentration decreased with the increase of wavelength from 280 to 550 nm. The linear regression relationship between CDOM absorption at 280 nm and DOC concentration was following: a(280 nm) = 1.507 L mg^?1 m^?1 · DOC + 1.215 m^?1. The spectral slope S values were dependent on the wavelength range used in the regression. The estimated S values decreased with increasing wavelength range used. A significant negative linear relationship was found between CDOM absorption coefficients, DOC‐specific absorption coefficients and estimated S values especially in longer wavelength range. The linear regression relationship between DOC‐specific absorption coefficients at 440 nm and estimated S values during the wavelength range from 280 to 500 nm was following: a*(440 nm) = (–0.021 μm · S + 0.424) L mg^?1 m^?1.     

Effects of precipitation on summer epilimnion thickness in tropical lakes

This study aimed to evaluate the importance of physical, chemical and biological factors (e.g., chlorophyll-a, and carbon) in the regulation of the summer epilimnion thickness (Zmix) of the tropical lakes focusing on trends of the current environmental variability. We sampled a set of tropical lakes from the middle Rio Doce lacustrine system (Minas Gerais, Brazil) for two consecutive summers. Besides Zmix, we analyzed the visible light attenuation (Kd_PAR), dissolved organic carbon (DOC), chlorophyll-a (Chl-a), total suspended solids (TSS), and fetch. We also analyze the quality of the dissolved organic matter (DOM) through colored dissolved organic matter (CDOM), spectral slope (S), DOC-specific absorbance (SUVA) and the ratio between the absorbance at 250 nm and 365 nm (E_250:365). Our results showed quite different results during the two years studied. In the summer of 2012, when there was higher rainfall, variations in Zmix were correlated to the optical factors associated with DOM quality, while in the drier summer of 2013 Chl-a, TSS and fetch were the variables that more explained Zmix. It suggests that DOM regulates the Zmix in the rainy periods in the studied tropical lakes and that control is determined by the balance between the DOM input (runoff) and output (such as photodegradation) of the aquatic systems. In reduced rain conditions (summer 2013), the factors responsible by chemistry photodegradation of DOM were predominant, and the organic matter was no longer controlling Zmix, which was driven by internal mixing, productivity, and the wind (Chl-a, TSS and fetch). In this study we showed how changes in precipitation might cause shifts in the factors that regulate the epilimnion thickness in tropical lakes.