Operational Effect Variables and Functional Ecosystem Classifications – a Review on Empirical Models for Aquatic Systems along a Salinity Gradient

This paper presents a comparative study based on a very comprehensive set of empirical data from many international data bases including fresh water systems, coastal brackish water areas and marine coastal areas. We present a general trophic level classification system (oligotrophic, mesotrophic, eutrophic and hypertrophic categories) for sites/areas characterised by a wide range of salinities. This classification system targets on the following operational effect variables (bioindicators), which are meant to reflect key structural and functional aspects of aquatic ecosystems and characteristic (median) values for entire defined areas (the ecosystem scale) for the growing season: Secchi depth (as a standard measure of water clarity), chlorophyll‐a concentrations (a measure of primary phytoplankton biomass), the oxygen saturation in the deep‐water zone (an indicator reflecting sedimentation, oxygen consumption, oxygen concentrations and the habitat conditions for zoobenthos, an important functional group) and the macrophyte cover (an important variable for the bioproduction potential, including fish production, and the “biological value” of aquatic systems). For a wide range of systems, these bioindicators can be predicted using practically useful models, i.e., models based on variables that can be accessed from standard monitoring programs and maps. These bioindicators are regulated by a set of abiotic factors, such as salinity, suspended particulate matter (SPM), nutrient concentrations (N and P), morphometry and water exchange. Empirical data ultimately form the basis for most ecological/environmental studies and this work uses maybe the most comprehensive data set ever related to trophic level conditions. It also gives compilations of empirically‐based (statistical) models quantifying how the variables are interrelated and how they reflect fundamental aspects of aquatic ecosystems. (© 2007 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Testing relationships between macroalgal cover and Secchi depth in the Baltic Sea

The present study tests whether relationships between macroalgal cover and water quality, recently developed for Danish coastal waters, are more universal and also applies at the other extreme of the Baltic Sea in Finnish coastal waters. We found that algal cover increases as a function of Secchi depth according to the same logarithmic function in Danish and Finnish coastal waters. Algal cover at a given depth (here modelled for 4 m) increases with increasing Secchi depth and approaches a maximum at the high Secchi depths found in the clearest areas of the Danish coastal waters. For a given Secchi depth the combined Danish/Finnish algal model thus predicts a similar cover of the algal community at a given water depth at both extremes of the Baltic Sea which represent quite different algal habitats. These results suggest that light limitation, and thus shading effects of eutrophication may cause similar reductions of macroalgal cover across ecosystems.     

Modeling the foodweb in coastal areas: a case study of Ringkøbing Fjord,Denmark

This work utilizes the CoastWeb model, a foodweb model for coastal areas that also includes a mass-balance model (CoastMab) for phosphorus and many abiotic/biotic interactions, to study the development in Ringkøbing Fjord, Denmark, from 1985 to 2004. This shallow coastal lagoon has an area of 300 km² and a mean depth of 1.9 m. The water exchange between the lagoon and the North Sea is regulated by a sluice. In 1996 there was a major regime shift in this lagoon with drastic reductions in chlorophyll-a concentrations, significant increases in water clarity (Secchi depth) and major changes in the number and biomass of clams as well as in macrophyte cover. Regime shifts is a “hot” topic in aquatic ecology and in this work the CoastWeb model is used as a tool to understand and quantify the causes behind this regime shift. The CoastWeb model is general and can also be used for other coastal areas. The basic model calculates monthly production values and changes in biomasses of ten functional groups of organisms (phytoplankton, bacterioplankton, herbivorous, and predatory zooplankton, benthic algae, macrophytes, jellyfish, zoobenthos and prey and predatory fish) and in Ringkøbing Fjord, also for clams (Mya arenaria). In spite of its complexity, the model is relatively simple to use, since all driving variables may be readily accessed from maps or monitoring programs. The model includes much abiotic/biotic feedback and it can also be used to address other causes for regime shifts other than the changes in salinity and nutrient inflow, which have caused the changes in Ringkøbing Fjord. The model has previously been tested for more than 20 smaller coastal areas and was shown to predict variations in foodweb characteristics very well. The focus of this paper is on temporal variations within one well-studied coastal area. The paper compares modeled values to empirical data for Ringkøbing Fjord and discusses fundamental ecosystem features such as regime shifts and compensatory effects in a way that is not practically feasible without the use of quantitative models.     

Morphometry and sedimentation as regulating factors for nutrient recycling and trophic state in coastal waters

The aim of this work is to quantify the importance of morphometry and sedimentation/resuspension on nutrient recycling and trophic characteristics in coastal waters. Extensive field work has been carried out in 23 coastal areas in the Swedish and Finnish part of the Baltic Proper. Sediment traps were deployed for two one-week periods in all areas. On average, 56% of the total sedimentation in sediment traps 3 m below the water surface (SedS) and 62% of the total sedimentation on sediment traps 1 m above the bottom (SedB) was resuspended material. Coastal morphometric parameters, surface water retention time and bottom dynamic conditions were determined for all areas. There is a marked relationship between SedS and inorganic-N concentration in the surface water. The relationship was improved significantly by using sedimentation of the resuspended fraction at 3 m water depth (SedR) instead of SedS.This led to the hypothesis that increased concentration of inorganic nitrogen in the surface water results from increased mineralisation of resuspended organic particles. A model describing SedS is presented where inorganic nitrogen concentration, the water surface area and the surface water retention time can explain 82% of the variation in SedS. In another model inorganic nitrogen and water surface area can explain as much as 93% of the variation in SedR.These results emphasise the importance of resuspension for nutrient recycling and trophic state in coastal waters. The importance of coastal morphometry and surface water retention time on total sedimentation and nutrient recycling makes it possible to classify coastal areas in terms of potential nutrient recycling capacity/trophic state from these simple sensitivity parameters.     

A Model to Predict How Individual Factors Influence Secchi Depth Variations among and within Lakes

This work presents a model to predict lake Secchi depth. It has been developed as a part of a comprehensive ecosystem model, LakeWeb. The Secchi depth model is based on a modelling technique using dimensionless moderators. It gives, e.g., weekly variations and accounts for how various factors influence Secchi depth individually in spite of the fact that many of them are interrelated in a complex manner. The model has been tested over a wide range of lakes. It gives predictions that agree well with empirical reference regressions, and also expected and requested divergences when the regressions do not provide sufficient resolution. The model is driven by the following data which are easily accessed from standard monitoring programs or maps: Total phosphorus, colour, pH, mean depth and lake area reflecting the three main processes influencing variations of Secchi depth within and among lakes, allochthonous inputs, autochthonous production and resuspension.     

A dynamic model for suspended particulate matter (SPM) in rivers

Aim To present a general, process‐based river model for suspended particulate matter (SPM). Location General approach based on processes; data from Europe and Israel. Methods The model has been tested and calibrated using an empirical river model for SPM and validated (blind‐tested) using data from seven European sites. This modelling gives mean monthly SPM concentrations in water for defined river sites. The model is based on processes in the entire upstream river stretch (and not for given river segments) and calculates the transport of SPM from land to water, primary production of SPM (within the upstream river stretch), resuspension, mineralization and retention of SPM in the upstream river stretch (but not bed load of friction materials, such as sand). The catchment area is differentiated into inflow (～ dry land) areas and outflow area (～ wetland areas dominated by relatively fast horizontal SPM‐fluxes). The model is simple to apply in practice as all driving variables may be accessed readily from maps. The driving variables are: latitude, altitude, continentality, catchment area and mean annual precipitation. Results Modelled values have been compared to independent empirical data from sites covering a relatively wide domain (catchment areas from 93 to 5250 km², precipitation from 400 to 660 mm year^?1, altitudes from ?210 to 150 m a.s.l., latitudes from 47 to 59° N and continentalities from 200 to 1000 km from the ocean). When blind‐tested, the model predicts annual SPM‐fluxes well. Conclusion When modelled values are compared to empirical data, the slope is almost perfect (1.03) and the r²‐value is 0.9996. This is good, given the fact that there are several simplifications in the model structure. It must, however, be stressed that there are only seven validation cases and that this model has not been tested for small catchments.     

Seasonality and trends in the Secchi disk transparency of Lake Ladoga

Data on Secchi disk transparency in Lake Ladoga, the largest lake in Europe, collected between 1905 and 2003, were used to detect climatic (interannual) trends for lake regions with various depths. The seasonal variations in Secchi depth (D _s) during the ice-free period both for limnetic regions with large differences in bathymetric characteristics and for the whole lake were estimated by more than 7000 transparency measurements. The two-dimensional data sets have a spatial resolution of approximately 20 km and are geo-referenced by latitude and longitude in Lake Ladoga. Monthly mean spatial transparency distributions and their variances were calculated from May to October. The spatial distributions of the transparency for each month are discussed within the context of lake bathymetric patterns. The maximum values of Secchi depth (more than 4 m) occur during May and October in deep regions. Both the minimum mean value of water transparency and minimum horizontal gradients of D _s for the lake occur in August. The regions with significant interannual (climatic) decreasing trends of D _s have been identified. These areas increase in summertime and cover approximately half the lake area. In spring and autumn the areas decrease and occur in the southern near-shore regions. The mean downward climatic trend of water transparency in Lake Ladoga is 0.02 m/year.     

Secchi disk — chlorophyll relationships in a lake with highly variable phytoplankton biomass

In meseutrophic Lake Constance mean euphotic phytoplankton chlorophyll concentrations vary about 100-fold over the year. Concomitant fluctuations in euphotic depth (Z_eu) and Secchi depth (Z_s) are related to each other in a non-linear fashion that as a rough approximation can be expressed by Z_eu 5 Z_s.Secchi depth is to a great extent a function of beam attenuation of light which depends on the inherent optical properties of the water and is highly sensitive to light scattering from particles. Euphotic depth, by contrast, is a function of the vertical light attenuation coefficient which also depends on absorption and scattering, but is less sensitive to the latter than beam attenuation. Algal cells both absorb and scatter light and therefore influence Secchi depth and euphotic depth, however, in different fashions.Whenever the lake is clear due to scarce phytoplankton, scattering is small and beam attenuation only exceeds vertical light attenuation by a relatively small factor. As a consequence, the ratio of euphotic depth to Secchi depth is small (1.5–2.5). When the lake is turbid due to high algal density, enhanced scattering from algal cells and detrital particles causes beam attenuation to rise more than vertical light attenuation, thus leading to high ratios of euphotic depth to Secchi depth (3–5). The relatively close relationships between Secchi depth and chlorophyll in Lake Constance are due to (1) high influence of chlorophyll concentration on water transparency, (2) co-variation of phytoplankton and other suspended particles, and (3) limited variation of cellular chlorophyll contents.     

The seasonality of monsoonal primary productivity in Sri Lanka

The relationship between phytoplankton primary production and seasonality of physico-chemical parameters were examined for five man-made lakes in the dry-zone of Sri Lanka. Sri Lanka experiences two monsoons dividing the year into four meteorological seasons: — the North East (October–December) and South West (April–June) monsoons and the two inter-monsoons. A significant log linear relationship was found between Secchi disc depth and the depth of the euphotic zone which was lowest during the NE monsoon. Maximum mean photosynthetic rate ranged from 0.935 ± 0.067 SE to 0.479 mg O₂l^–1 h^–1 ± 0.115. Gross primary productivity which ranged from 0.378 g O₂ m^–2 h^–1 in the NE monsoon to 0.980 g O₂ m^–2 h^–1 in the SW monsoon showed significant season variation. This is shown to be determined either directly or indirectly by the light regime.     

Seasonality in Nutrients and Phytoplankton Production in Two Shallow Lakes: Waigani Lake,Papua New Guinea,and Barton Broad,Norfolk, England

Waigani Lake, near Port Moresby, Papua New Guinea and Barton Broad, Norfolk, England are both shallow lakes nutrient-enriched from sewage effluent disposal. In Waigani Lake phytoplankton biomass varied seasonally with lower levels (100-200 mg chlorophyll α m⁻³) during the wet season increasing to over 400 mg chlorophyll α m⁻³ at the end of the dry season. Secchi disc depths varied between 0. 11 and 0. 34 m. Phytoplankton productivity in Waigani Lake was very high throughout the year (range: A_max 4,370-21,000 mg O₂ m⁻³ h⁻¹) but production was lower during the wet season (range: A_max 4,370-12,700 mg O₂ m⁻³ h⁻¹). High surface productivity was recorded from August to December except on sampling days when the weather was overcast. Productivity throughout the year declined rapidly with depth. Algal biomass in Barton Broad varied from 3-10 mg chlorophyll α m⁻³ in winter but increased in spring and was very high in summer (200-500 mg chlorophyll α m⁻³). Secchi disc depth varied from 0.21 m in August 1976 to 1.76 m in December. Phytoplankton production in Barton Broad was low in winter (range: A_max 247-1,250 mg O₂ m⁻³ h⁻¹) but increased markedly in spring and summer with the highest rate (A_max 6,850 mg O₂ m⁻³ h⁻¹) being recorded in August. Surface inhibition was observed during summer except when the weather was overcast. Seasonality in nutrients and phytoplankton in Waigani Lake appear to be related to rainfall. Nutrient concentrations in Barton Broad are more closely related to phytoplankton activity which, in turn, correlates with seasonality in solar radiation.     

Pre-classification improves relationships between water clarity, light attenuation, and suspended particulates in turbid inland waters

Relationships between water clarity, light attenuation, and concentration of suspended particulates are important in water optics and remote sensing, but are not well described yet, especially for optically complex turbid inland waters. In this study, based on 3-year data from Chinese lakes (Lake Taihu, Lake Chaohu, and Three Gorges reservoir), we propose a new approach to describe the inter-relationships of these bio-optical variables. This approach includes a pre-classification step of the waters into three types based on a semi-analytical parameter C _s before establishing the relationships. Our results showed that the pre-classification of waters increased model accuracies both for Z _SD (Secchi depth) versus K _d (diffuse attenuation coefficient) and K _d versus TSM (total suspended matter concentration). The quasi-theoretical model described better the relationship between Z _SD and K _d than the empirical model. For the K _d versus TSM relationship, linear models proved suitable for the Type 2 and Type 3 waters, whereas the power function model gave a better fit for the Type 1 water. Testing of the proposed relations with an independent dataset showed mean absolute percentage errors (MAPE) mostly below 30%. The findings of this study clarify the relationships between Z _SD, K _d, and TSM, and improve our bio-optical understanding of complex turbid inland waters.     

A dynamic mass balance model for phosphorus fluxes and concentrations in coastal areas

This paper presents a general, process-based mass balance model (CoastMab) for total phosphorus (TP) in defined coastal areas (at the ecosystem scale). The model is based on ordinary differential equations and calculates inflow, outflow and internal fluxes on a monthly basis. It consists of four compartments: surface water, deep water, erosion/transportation areas for fine sediments and accumulation areas for fine sediments. The separation between surface water and deep water is not done based on water temperature, but on sedimentological criteria instead (from the theoretical wave base). There are algorithms for all major internal TP fluxes (sedimentation, resuspension, diffusion, mixing and burial). Validations were performed using data from 21 different Baltic coastal areas. The results show that the model predicts monthly TP in water and chlorophyll a very well (generally within the uncertainty bands of the empirical data). The model has also been put through sensitivity tests, which show that the most important factor regulating the predictions of the model is generally the TP concentration in the sea beyond the coast. The model is simple to apply, since all driving variables may be accessed from maps or monitoring programs. The driving variables include coastal area, section area (between the defined coastal area and the adjacent sea), mean and maximum depths, latitude (used to predict water temperatures, stratification and mixing), salinity and TP concentration in the sea. Many of the model structures are general and could be used for areas other than those included in this study, e.g., for open coasts, estuaries or tidal coasts, as well as for other substances than phosphorus.     

Particle sources and trace element reactivity in the Humber plume

Samples of suspended particulate matter (SPM) collected from the Humber Estuary had higher concentrations of particulate metals than SPM from Holderness coastal waters (U.K.). Characterised SPM from both sources was used in laboratory experiments involving the uptake of radiotracer¹⁰⁹Cd,¹³⁷Cs,⁵⁴Mn and⁶⁵Zn. Kinetic experiments, over five days, showed that the rate and extent of uptake was highly dependent on particle type, with¹⁰⁹Cd,⁵⁴Mn and⁶⁵Zn being more reactive with Humber Estuary particles than those from Holderness and¹³⁷Cs having the opposite trend. Adsorption experiments were also carried out on suspensions in which SPM from the Humber Estuary and Holderness coastal water were mixed in various proportions. These experiments revealed that K_d for⁶⁵Zn increased linearly with the proportion of Humber SPM, K_d for¹³⁷Cs decreased linearly with increase in Humber SPM and K_d for⁵⁴Mn and¹⁰⁹Cd displayed non-linear behaviour. The results of the study were used to develop an algorithm for predicting the partition coefficients in the Humber Plume based on the extent of particle mixing from the two source regions. The use of^206/207Pb ratios in determining the extent of particle mixing is discussed, along with the application of the algorithm to the modelling of particulate trace metal behaviour in the Humber-Wash coastal zone.     

Decrease in water clarity of the southern and central North Sea during the 20th century

Light in the marine environment is a key environmental variable coupling physics to marine biogeochemistry and ecology. Weak light penetration reduces light available for photosynthesis, changing energy fluxes through the marine food web. Based on published and unpublished data, this study shows that the central and southern North Sea has become significantly less clear over the second half of the 20th century. In particular, in the different regions and seasons investigated, the average Secchi depth pre‐1950 decreased between 25% and 75% compared to the average Secchi depth post‐1950. Consequently, in summer pre‐1950, most (74%) of the sea floor in the permanently mixed area off East Anglia was within the photic zone. For the last 25+ years, changes in water clarity were more likely driven by an increase in the concentration of suspended sediments, rather than phytoplankton. We suggest that a combination of causes have contributed to this increase in suspended sediments such as changes in sea‐bed communities and in weather patterns, decreased sink of sediments in estuaries, and increased coastal erosion. A predicted future increase in storminess (Beniston et al., 2007; Kovats et al., 2014) could enhance the concentration of suspended sediments in the water column and consequently lead to a further decrease in clarity, with potential impacts on phytoplankton production, CO₂ fluxes, and fishery production.     

The effect of non-algal turbidity on the relationship of Secchi depth to chlorophyll a

Data from four reservoirs representative of different trophic states and with different apparent optical properties were analyzed to determine the relationship of Secchi depth to algal biomass as measured by chlorophyll a. In the eutrophic reservoir Secchi depth was determined partially by the chlorophyll a content (r² = 0.31) but only when chlorophyll a data from bloom conditions are included. In the two mesotrophic reservoirs, Secchi depth was entirely determined by non-algal turbidity. In the oligotrophic reservoir, Secchi depth was determined neither by chlorophyll a nor non-algal turbidity and was probably determined by dissolved color. When data from the four reservoirs were pooled (N = 205), 53% of the variation in Secchi depth was explained by: SD = 2.55–0.52 ln (Turbidity) + 0.005 (Chlorophyll a). It is apparent that attempts to estimate algal biomass for trophic state classification or other management practices from Secchi depth data are inappropriate even where moderate amounts of non-algal turbidity are present.     

Effects of light on seagrass photosynthesis,growth and depth distribution

The relationships between light regime, photosynthesis, growth and depth distribution of a temperate seagrass, Zostera marina L. (eelgrass), were investigated in a subtidal eelgrass meadow near Woods Hole, MA. The seasonal light patterns in which the quantum irradiance exceeded the light compensation point (H_comp) and light saturation point (H_sat) for eelgrass photosynthesis were determined. Along with photosynthesis and respiration rates, these patterns were used to predict carbon balances monthly throughout the year. Gross photosynthesis peaked in late-summer, but net photosynthesis peaked in spring (May), due to high respiration rates at summer temperatures. Predictions of net photosynthesis correlated with in situ growth rates at the study site and with reports from other locations.The maximum depth limit for eelgrass was related to the depth distribution of H_comp, and a minimum annual average H_comp (12.3 h) for survival was determined. Maximum depth limits for eelgrass were predicted for various light extinction coefficients and a relationship between Secchi disc depth and the maximum depth limit for survival was established. The Secchi disc depth averaged over the year approximates the light compensation depth for eelgrass. This relationship may be applicable to other sites and other seagrass species.     

Below-halocline oxygen consumption in the Kattegat

Sediment and seston oxygen consumption rates below the sharp halocline in the south-eastern part of the shallow Kattegat were measured and compared to calculated rates of carbon addition through the halocline. The mean rate of decrease in deep-water oxygen concentrations between March and September 1988 was 1.0 ml O₂ M^–3 h^–1. Measurements of benthic oxygen uptake using laboratory-incubated sediment cores from depths 30 m gave a mean value of 7.8 ml O₂ m^–2 h^–1. Below-halocline water (from 20 m, 30 m and 1 m above bottom) incubated in bottles showed oxygen consumption rates varying from 0.5 ml O₂ m ^–3 h^–1 in March to 2.8 ml O₂ M^–3 h^-1 in late August. The sum of benthic and deep-water oxygen consumption was equivalent to a mean oxygen decrease rate of 1.7 ml O₂ m^–3 h^–1 below the halocline. Of the total oxygen consumption below the halocline 65% was due to oxygen up-take in the water and 35% was due to benthic oxygen consumption. The sum of oxygen consumption measured in sediment cores and in bottles corresponds to a carbon utilisation of 80.1 g C m^–2 (respiratory quotient (RQ), assumed 1.0 and 1.4 for water and sediment, respectively), while the decrease in deep-water oxygen concentration was equivalent to 43.0 g C m^–2 (RQ assumed = 1.0). Using published values for the external N loading (including deep-water supply), ¹⁵NO₃-uptake, ¹⁴CO₂-uptake in combination with % ¹⁵NO₃-uptake of total ¹⁵N-uptake (nitrate, ammonia and urea) and a Redfield C/N ratio of 6.6, rates of carbon addition (new or export production) through the halocline were calculated to 31.9, 46.7 and 36.3 g C m^–2, respectively, with a mean value of 38.3 g C m^–2 for the 8 month period March–September. This is somewhat less than the value (50.5 g C m^–2) calculated from a published empirical relationship between total and export production. The fact that the calculated carbon addition through the halocline was appreciably less than the carbon equivalent of the measured below-halocline respiration may be an effect of sediment focusing (horizontal transport of sedimenting material to deeper areas), since the bottom area below the halocline is much smaller than the total area of the Kattegat. A lower observed decrease in the oxygen concentration below the halocline compared to the sum of measured sediment and deep-water oxygen consumption on the other hand indicates oxygen supply to below-halocline waters through advection and/or vertical entrainment.     

Prediction of Secchi disc depths in Florida lakes: Impact of algal biomass and organic color

A model for the prediction of Secchi disc depths in Florida lakes was developed and tested using data from 205 lakes. A statistical analysis showed that the best estimate of lake Secchi disc depths could be obtained by In (SD) = 2.01 ? 0.370 In (Chla) ? 0.278 In (C) where SD is Secchi disc depth (m), Chla is the chlorophyll a concentration (mg/m³) and C is the organic color concentration (mg/l as Pt). The model yields unbiased estimates of lake Secchi disc depths over a wide range of algal and organic color concentrations and has a 95% confidence interval of 47 to 224% of the calculated Secchi disc depth. Other published Secchi-Chlorophyll models are less precise but can be used almost equally well. This indicates organic color concentrations do not affect lake Secchi disc depths as much as algal levels. Further reductions in the remaining error term, however, might be accomplished by including a variable for suspended inorganic sediment.     

Global DNA methylation in a population with aflatoxin B1 exposure

We previously reported that global DNA hypomethylation, measured as Sat2 methylation in white blood cells (WBC), and aflatoxin B₁ (AFB₁) exposure were associated with increased hepatocellular carcinoma risk. In this study, we assessed the association between AFB₁ exposure and global DNA methylation. We measured LINE-1 and Sat2 methylation in WBC DNA samples from 1140 cancer free participants of the Cancer Screening Program (CSP) cohort. Blood and urine samples were used to determine the level of AFB₁-albumin (AFB₁-Alb) adducts and urinary AFB₁ metabolites. In continuous models, we found reverse associations of urinary AFB₁ with LINE-1 and Sat2 methylation. The odds ratio (OR) per 1 unit decrease were 1.12 (95%CI = 1.03–1.22) for LINE-1 and 1.48 (95%CI = 1.10–2.00) for Sat2 methylation. When compared with subjects in the highest quartile of LINE-1, we found that individuals in the 2nd and 3rd quartiles were less likely to have detectable AFB₁-Alb adducts, with ORs (95%CI) of 0.61 (0.40–0.93), 0.61 (0.40-.94), and 1.09 (0.69–1.72), respectively. The OR for detectable AFB₁-Alb was 1.81 (95%CI = 1.15–2.85) for subjects in the lowest quartile of Sat2 methylation. The OR for detection of urinary AFB₁ for those with LINE-1 methylation in the lowest quartile compared with those in the highest quartile was 1.87 (95%CI = 1.15–3.04). The corresponding OR was 1.75 (95%CI = 1.08–2.82) for subjects in the lowest quartile of Sat2 methylation. The association between AFB₁ exposure and global DNA methylation may have implications for the epigenetic effect of AFB₁ on hepatocellular carcinoma development and also suggests that changes in DNA methylation may represent an epigenetic biomarker of dietary AFB₁ exposure.