Assessing Effects of Climate Change on Biogeochemical Cycling of Trace Metals in

Assessing the impacts of climate changes on water quality requires an understanding of the biogeochemical cycling of trace metals. Evidence from research on alluvial aquifers and coastal watersheds shows direct impacts of climate change on the fate and transformation of trace metals in natural environments. The case studies presented here use field data and numerical modeling techniques to test assumptions about the effects of climate change on natural arsenic contamination of groundwater in alluvial aqui fers and mercury bioaccumulation in coastalsalt marshes. The results show that the rises of sea level and river base during the warm Holocene period has led to an overall increase in groundwater arsenic concentration due to the development of reducing geo chemical


INTRODUCTION
Trace metals (e.g., mercury, arsenic, lead, cadmium, etc.) are of particular concern in alluvial and coastal watersheds because in excess they are toxic to wildlife and humans [1].As these elements cycle through natural environments, they undergo biogeochemical reactions that are often coupled to climate change.Many of these toxic heavy metals (e.g., mercury and arsenic) are found at low background concentrations throughout earth's crust [2,3].These metals may be concentrated locally by natural weathering of bedrocks or anthropogenic activities [4], and their fate and transformation in natural environments often depends on climatic, hydrological and biogeochemical conditions.Climate change can affect every component of the global hydrological cycle, including precipitation, evaporation, groundwater, and runoff [5].An increase in precipitation, for example, would lead to increases in surface runoff, weathering rates, sediment yields, nutrient loading, and release of heavy metals from both natural and anthropogenic sources.
Various biogeochemical conditions developed in hosting geological environments can naturally cause metals to mobilize, contaminating water resources far from metals' ultimate sources.For example, arsenic released from chemical weathering of source rocks (e.g., black shale's, metamorphic slates, sulfide-rich coals and igneous rocks) would be absorbed by hydrous iron oxides (HFOs) transported by surface streams [6].Climatic or hydrologic changes may lead to their deposition in stream or coastal environments along with transported organic matter.Subsequently, anaerobic conditions developed in groundwater hosted by these deposits and iron-reducing bacteria could cause the reductive dissolution of As-bearing HFO, releasing arsenic to groundwater [7,8].Moreover, rising sea level and saltwater intrusion may also lead to substantial desorption of arsenic from hydrous oxides in affected aquifers due to pH effects and ionic competition for HFO sorbing sites [9,10].Naturally occurring high levels of groundwater arsenic in alluvial aquifers have caused major health problems in many countries [11][12][13][14].
Mercury is also of particular concern when methylated because methyl mercury (MeHg) bioaccumulates in organisms and becomes more toxic than inorganic forms.Mercury methylation can be facilitated by sulfate reducing bacteria (SRB) under sulfide-forming (strongly reducing) conditions [15][16][17][18] which are themselves promoted by an influx of electron acceptors (i.e., sulfate) and organic carbon [19].Saltwater intrusion accompanying global sea-level rise provides additional sulfate for bacterial sulfate reduction in coastal sediments.However, sulfide produced by sulfate reduction may also react with dissolved metals to form insoluble metal sulfides; such reaction would decrease the availability of dissolved Hg for methylation by sequestering it in insoluble HgS [18] or Fe-sulfides [14,[20][21].Fe-sulfides formed in Fe-rich systems also sequester many other dissolved metals (e.g., Cu, Ni, Co, Zn, U, Pb, As, etc.).Moreover, droughts and saltwater intrusion could lead to salinity increase in estuaries and coastal wetlands [22].Compeau and Bartha [23] showed that bacterial sulfate reduction and Hg methylation may be inhibited in high-salinity environments where sulfate reducing bacteria compete with methanogenic bacteria for limited organic matter produced by fermentation organisms.Climate change and resulting hydrological modification (e.g., droughts, sea level rise) thus could significantly alter the mobility and transformation of Hg and many other metals in aquatic ecosystems.
In this paper, we present case studies on the concentrations, mobility, and transformation of trace metals within alluvial aquifers and coastal watersheds in response to changes in climatic, hydrologic, and biogeochemical conditions.Our main objective is to demonstrate how field data and numerical modeling techniques can be extended and applied to understand the effects of climate change on contaminant fluxes and their transformation in alluvial and coastal environments.

CLIMATIC EFFECTS ON LEVEL OF ARSENIC IN AQUIFERS
The first case study examines the link between climate conditions and arsenic level in groundwater of alluvial aquifers over time.During the dry and cold glacial period of Late Pleistocene (120,000-12,000 yr BP), low sea-level stand and greater hydraulic gradients facilitate faster hydrologic flushing of the alluvial aquifers [24,25].Due to fast hydrologic flushing, groundwater in most Pliocene-Pleistocene alluvial aquifers has very low arsenic contents (Fig. 1).Following the end of the last glacial period, climate warmed up and sealevel gradually and steadily rose during the Holocene (12,000-6,000 yr BP).As a result, finegrain sediments along with organic matter deposited and filled the river channels.During this warmer Holocene period, high sea level, high river base level, and very gentle hydraulic gradients favor the accumulation and degradation of organic matter, leading to the development of reducing geochemical conditions and sluggish groundwater movement.The reducing conditions triggered the reduction of Fe-and Mn-oxyhydroxides and the release of As, Fe, and Mn into groundwater.The sluggish hydrologic conditions allow little hydrologic flushing and result in an enrichment of arsenic in groundwater in shallow Holocene-age alluvial aquifers (Fig. 1).This sea-level highstand and sluggish hydrologic conditions continued into the civilization and modern day.Multi-century sea-level records reveal a significant acceleration of sea-level rise from the 20th century to the present day [27].Morhange et al. [28] indicated that the ocean has reached its highest level at present over the last 10,000 years.Thus, slow arsenic flushing in alluvial aquifers will likely continue if this acceleration of sea-level rise remains its current course.The widespread arsenic contamination of groundwater in glacial tills worldwide [6] also indicates that arsenic mobilization and enrichment could be enhanced by global warming and glacial melting.Previous workers [29,30] have proposed that the retreat of continental glaciers at the end of the Pleistocene led to rise of sea and river base levels, and deposition of the Holocene alluvium which hosts arsenic derived from glacial erosion of arsenic-bearing bedrocks.

Hydrologic Flushing in Transport-Dominated Systems (Low Sea Level Stand)
To gain insight into how sea level and hydraulic gradient may control level of groundwater arsenic in alluvial aquifers, we have carried out 1-D simulations of freshwater flushing of an As-contaminated aquifer.The following solute transport equation takes into account the effect of advection, hydrodynamic dispersion, adsorption, and chemical reactions on plume migration: ( Where C, D, and R i represent concentration, dispersion coefficient, and chemical reaction rate (by production and consumption) of solute i in a flow system with velocity v.The reaction term R i for all possible mineral precipitation-dissolution, speciation, redox, and surface complexation reactions [31,32].Traditional Freundlich and Langmuir sorption theories use distribution coefficients K d to calculate the ratios of sorbed to dissolved ions.The more comprehensive surface complexation model accounts for electrical charges on the sorbent surface and prescribes mass balance on the sorbing sites.In our model, an alluvial aquifer containing 300 µg/kg of arsenic is infiltrated by dilute rainwater at 25ºC.Assuming a Darcian flow rate of 0.1 m/day, we calculate the distribution of arsenic along the flowpath of 1000 m.To calculate the minimum time required to flush out existing arsenic loads dissolved in groundwater, the reaction term R i is set to zero (i.e., no arsenic is released from chemical The widespread arsenic contamination of groundwater in glacial tills worldwide [6] also indicates that arsenic mobilization and enrichment could be enhanced by global warming and glacial melting.Previous workers [29,30] have proposed that the retreat of continental glaciers at the end of the Pleistocene led to rise of sea and river base levels, and deposition of the Holocene alluvium which hosts arsenic derived from glacial erosion of arsenic-bearing bedrocks.

Hydrologic Flushing in Transport-Dominated Systems (Low Sea Level Stand)
To gain insight into how sea level and hydraulic gradient may control level of groundwater arsenic in alluvial aquifers, we have carried out 1-D simulations of freshwater flushing of an As-contaminated aquifer.The following solute transport equation takes into account the effect of advection, hydrodynamic dispersion, adsorption, and chemical reactions on plume migration: ( Where C, D, and R i represent concentration, dispersion coefficient, and chemical reaction rate (by production and consumption) of solute i in a flow system with velocity v.The reaction term R i for all possible mineral precipitation-dissolution, speciation, redox, and surface complexation reactions [31,32].Traditional Freundlich and Langmuir sorption theories use distribution coefficients K d to calculate the ratios of sorbed to dissolved ions.The more comprehensive surface complexation model accounts for electrical charges on the sorbent surface and prescribes mass balance on the sorbing sites.In our model, an alluvial aquifer containing 300 µg/kg of arsenic is infiltrated by dilute rainwater at 25ºC.Assuming a Darcian flow rate of 0.1 m/day, we calculate the distribution of arsenic along the flowpath of 1000 m.To calculate the minimum time required to flush out existing arsenic loads dissolved in groundwater, the reaction term R i is set to zero (i.e., no arsenic is released from chemical The widespread arsenic contamination of groundwater in glacial tills worldwide [6] also indicates that arsenic mobilization and enrichment could be enhanced by global warming and glacial melting.Previous workers [29,30] have proposed that the retreat of continental glaciers at the end of the Pleistocene led to rise of sea and river base levels, and deposition of the Holocene alluvium which hosts arsenic derived from glacial erosion of arsenic-bearing bedrocks.

Hydrologic Flushing in Transport-Dominated Systems (Low Sea Level Stand)
To gain insight into how sea level and hydraulic gradient may control level of groundwater arsenic in alluvial aquifers, we have carried out 1-D simulations of freshwater flushing of an As-contaminated aquifer.The following solute transport equation takes into account the effect of advection, hydrodynamic dispersion, adsorption, and chemical reactions on plume migration: ( Where C, D, and R i represent concentration, dispersion coefficient, and chemical reaction rate (by production and consumption) of solute i in a flow system with velocity v.The reaction term R i for all possible mineral precipitation-dissolution, speciation, redox, and surface complexation reactions [31,32].Traditional Freundlich and Langmuir sorption theories use distribution coefficients K d to calculate the ratios of sorbed to dissolved ions.The more comprehensive surface complexation model accounts for electrical charges on the sorbent surface and prescribes mass balance on the sorbing sites.In our model, an alluvial aquifer containing 300 µg/kg of arsenic is infiltrated by dilute rainwater at 25ºC.Assuming a Darcian flow rate of 0.1 m/day, we calculate the distribution of arsenic along the flowpath of 1000 m.To calculate the minimum time required to flush out existing arsenic loads dissolved in groundwater, the reaction term R i is set to zero (i.e., no arsenic is released from chemical reactions).The modeling results show that most initial mobile arsenic loads in a considerable part of the aquifer can be flushed out in a few thousand years (Fig. 2).The modeling results are consistent with the ages of As-rich groundwater estimated for Bangladesh Holocene floodplain aquifers [33].However, longer groundwater residence times (with 14 C groundwater ages of 10-40 ka) have been reported in the alluvial plain of Taiwan [34].It should be noted that the time required to flush the aquifer would be considerably longer for slow flow rates and higher K d values (not shown), or as long as the geochemical conditions allow continuous arsenic release from solid phases in a reaction-dominated system.The modeling results indicate that climate and related sea level changes could set the stage of regional hydrologic gradients that ultimately control the level and cycling of arsenic in alluvial aquifers.High sea level stand and very gentle hydraulic gradients often lead to accumulation of organic matter and the development of reducing (anaerobic) geochemical conditions in river floodplains.We used Geochemist's Workbench (GWB) [35] to trace the sequence of biogeochemical reactions that occur during the bacterial Fe(III) and Mn(IV) oxide reduction, which releases metals and subsequently induces the precipitation of Fe carbonate and sulfide minerals.The program can trace how a fluid's chemistry evolves and which minerals precipitate or dissolve over the course of biogeochemical processes.The purpose of the modeling is to provide insights on the sequence of mineral reactions during the reductive dissolution of Fe and Mn oxides and how mineral reactions affect metal mobility in alluvial aquifers under similar biogeochemical conditions.We begin by equilibrating a typical coastal plain groundwater upstream of an iron reduction zone under aerobic conditions at 25ºC.The calculation uses the water chemical data collected from the Eutaw aquifer of Alabama [36] and assumes that the initial concentrations of Fe and Mn reflect equilibrium with hematite reactions).The modeling results show that most initial mobile arsenic loads in a considerable part of the aquifer can be flushed out in a few thousand years (Fig. 2).The modeling results are consistent with the ages of As-rich groundwater estimated for Bangladesh Holocene floodplain aquifers [33].However, longer groundwater residence times (with 14 C groundwater ages of 10-40 ka) have been reported in the alluvial plain of Taiwan [34].It should be noted that the time required to flush the aquifer would be considerably longer for slow flow rates and higher K d values (not shown), or as long as the geochemical conditions allow continuous arsenic release from solid phases in a reaction-dominated system.The modeling results indicate that climate and related sea level changes could set the stage of regional hydrologic gradients that ultimately control the level and cycling of arsenic in alluvial aquifers.

Bacterial Reduction of Fe(III) and Mn(IV) Oxides in Reaction-Dominated System (Sea Level High Stand)
High sea level stand and very gentle hydraulic gradients often lead to accumulation of organic matter and the development of reducing (anaerobic) geochemical conditions in river floodplains.We used Geochemist's Workbench (GWB) [35] to trace the sequence of biogeochemical reactions that occur during the bacterial Fe(III) and Mn(IV) oxide reduction, which releases metals and subsequently induces the precipitation of Fe carbonate and sulfide minerals.The program can trace how a fluid's chemistry evolves and which minerals precipitate or dissolve over the course of biogeochemical processes.The purpose of the modeling is to provide insights on the sequence of mineral reactions during the reductive dissolution of Fe and Mn oxides and how mineral reactions affect metal mobility in alluvial aquifers under similar biogeochemical conditions.We begin by equilibrating a typical coastal plain groundwater upstream of an iron reduction zone under aerobic conditions at 25ºC.The calculation uses the water chemical data collected from the Eutaw aquifer of Alabama [36] and assumes that the initial concentrations of Fe and Mn reflect equilibrium with hematite reactions).The modeling results show that most initial mobile arsenic loads in a considerable part of the aquifer can be flushed out in a few thousand years (Fig. 2).The modeling results are consistent with the ages of As-rich groundwater estimated for Bangladesh Holocene floodplain aquifers [33].However, longer groundwater residence times (with 14 C groundwater ages of 10-40 ka) have been reported in the alluvial plain of Taiwan [34].It should be noted that the time required to flush the aquifer would be considerably longer for slow flow rates and higher K d values (not shown), or as long as the geochemical conditions allow continuous arsenic release from solid phases in a reaction-dominated system.The modeling results indicate that climate and related sea level changes could set the stage of regional hydrologic gradients that ultimately control the level and cycling of arsenic in alluvial aquifers.

Bacterial Reduction of Fe(III) and Mn(IV) Oxides in Reaction-Dominated System (Sea Level High Stand)
High sea level stand and very gentle hydraulic gradients often lead to accumulation of organic matter and the development of reducing (anaerobic) geochemical conditions in river floodplains.We used Geochemist's Workbench (GWB) [35] to trace the sequence of biogeochemical reactions that occur during the bacterial Fe(III) and Mn(IV) oxide reduction, which releases metals and subsequently induces the precipitation of Fe carbonate and sulfide minerals.The program can trace how a fluid's chemistry evolves and which minerals precipitate or dissolve over the course of biogeochemical processes.The purpose of the modeling is to provide insights on the sequence of mineral reactions during the reductive dissolution of Fe and Mn oxides and how mineral reactions affect metal mobility in alluvial aquifers under similar biogeochemical conditions.We begin by equilibrating a typical coastal plain groundwater upstream of an iron reduction zone under aerobic conditions at 25ºC.The calculation uses the water chemical data collected from the Eutaw aquifer of Alabama [36] and assumes that the initial concentrations of Fe and Mn reflect equilibrium with hematite (Fe 2 O 3 , a proxy of Fe(III) oxyhydroxides) and pyrolusite (MnO 2 , a proxy for Mn(IV) oxyhydroxides) in the sediments.The model then simulates the geochemical effects of titration of organic matter into the system.We consider transformation of pyrolusite and hematite by the following redox reactions: In the simulation, fluid reactants containing 500 µmol of acetate (CH 3 COO -) displace existing fluid from the system and the values of Eh slide from +700 mV to -200 mV over the reaction path.
The predicted mineral reactions of manganese and iron oxides (Fig. 3) follow the well-known Ostwald's step rule, which suggests that the solid first formed from a solution would be the least stable polymorph, followed by the more stable phases [37,38].Pyrolusite in the initial system first becomes thermodynamically unstable during bacterial reduction and transforms over time to a sequence of progressively more stable manganese minerals and species (Fig. 3a) at lower oxidation states: Once the reduction of Mn minerals has nearly been completed, the iron redox reactions start (Fig. 3a) and hematite (Fe 2 O 3 ) begins to dissolve to form more stable siderite (FeCO 3 ) or pyrite at low oxidation states.The reduction of Fe(III) oxides occurs under more reducing conditions than Mn minerals.At the later stage of the reaction, reduced metal species also combine with HCO 3 -mineralized from organic sources to form minerals such as rhodochrosite (MnCO 3 ) and siderite (FeCO 3 ).Under highly reducing conditions, aqueous Fe 2+ reacts with H 2 S to form pyrite, which can remove trace elements such as Co, Ni, and As from groundwater by co-precipitation [39,40].It should be noted that Fe-and Mn-rich groundwaters in many alluvial aquifers are actually saturated with siderite and rhodochrosite [36,41].Fig. 3b shows the calculated Mn and Fe concentrations in fluid over the same reaction path.It clearly demonstrates how the precipitation and dissolution of various Mnand Fe-minerals control the mobility of metals in groundwater.Rapid rise and fall of metal concentrations observed over a short distance along the flow path from the recharge zone may be explained by the transformation of various iron and manganese minerals.This modeling result implies that mobilization and re-adsorption (or re-precipitation) of arsenic may occurrapidly over small distances, leading to a dramatic change of As in groundwater from high-As into low-As facies over a very short groundwater flow distance (within meters) [42].This poses a major challenge to predicting spatio-temporal variations in arsenic concentrations at a local scale.Moreover, the modeling results imply that the transformation of iron and manganese minerals could control mobility of trace metals such as arsenic in alluvial aquifers and requires continued investigation.

Effects of Saltwater Intrusion on Desorption and Mobilization of Arsenic
We use GWB and surface complexation model [31] to simulate the adsorption and desorption of arsenic species associated with oxide minerals in response to sea level rise and saltwater intrusion.The intrusion of seawater in shallow alluvial aquifers may lead to desorption of arsenic from surface of hydrous oxides due to pH effects and ionic competition for HFO sorbing sites.The initial system contains 1 kg of fresh groundwater equilibrated with a small amount (1 g) of Fe(OH) 3 at pH 7. The groundwater contains 1 µg/kg of As(III), 1 µg/kg of As(V), and 0.05 molal of NaCl.The calculation prevents the redox coupling between As(III) and As(V).More stable ferric minerals hematite, goethite, and jarosite are suppressed as they are unlikely to reach equilibrium with shallow groundwater at near surface temperatures.Ferric hydroxides used in the simulation have high specific surface areas (600 m 2 /g) for sorption reactions.The surface of HFOs is composed of weakly sorbing sites (density = 0.4 mol/mol mineral) and strongly sorbing sites (density = 0.01) [35].

Effects of Saltwater Intrusion on Desorption and Mobilization of Arsenic
We use GWB and surface complexation model [31] to simulate the adsorption and desorption of arsenic species associated with oxide minerals in response to sea level rise and saltwater intrusion.The intrusion of seawater in shallow alluvial aquifers may lead to desorption of arsenic from surface of hydrous oxides due to pH effects and ionic competition for HFO sorbing sites.The initial system contains 1 kg of fresh groundwater equilibrated with a small amount (1 g) of Fe(OH) 3 at pH 7. The groundwater contains 1 µg/kg of As(III), 1 µg/kg of As(V), and 0.05 molal of NaCl.The calculation prevents the redox coupling between As(III) and As(V).More stable ferric minerals hematite, goethite, and jarosite are suppressed as they are unlikely to reach equilibrium with shallow groundwater at near surface temperatures.Ferric hydroxides used in the simulation have high specific surface areas (600 m 2 /g) for sorption reactions.The surface of HFOs is composed of weakly sorbing sites (density = 0.4 mol/mol mineral) and strongly sorbing sites (density = 0.01) [35].

Effects of Saltwater Intrusion on Desorption and Mobilization of Arsenic
We use GWB and surface complexation model [31] to simulate the adsorption and desorption of arsenic species associated with oxide minerals in response to sea level rise and saltwater intrusion.The intrusion of seawater in shallow alluvial aquifers may lead to desorption of arsenic from surface of hydrous oxides due to pH effects and ionic competition for HFO sorbing sites.The initial system contains 1 kg of fresh groundwater equilibrated with a small amount (1 g) of Fe(OH) 3 at pH 7. The groundwater contains 1 µg/kg of As(III), 1 µg/kg of As(V), and 0.05 molal of NaCl.The calculation prevents the redox coupling between As(III) and As(V).More stable ferric minerals hematite, goethite, and jarosite are suppressed as they are unlikely to reach equilibrium with shallow groundwater at near surface temperatures.Ferric hydroxides used in the simulation have high specific surface areas (600 m 2 /g) for sorption reactions.The surface of HFOs is composed of weakly sorbing sites (density = 0.4 mol/mol mineral) and strongly sorbing sites (density = 0.01) [35].
The modeling results show that the initial As sorbed onto Fe(OH) 3 is about 19.50 mg/kg (19.22 mg/kg as As(V) and 0.28 mg/kg as As(III)) in equilibrium with a freshwater containing 2 ug/kg of total dissolved As at pH of 7.This result suggests that As is strongly adsorbed by iron oxyhdroxides under aerobic, near-neutral pH conditions.Iron oxide with positive surface charges has a high capability to adsorb negatively-charged anions of arsenic (V) (e.g., HAsO 4 2-, AsO 4 3-).By contrast, As(III), in the form of a non-ionic neutral species As(OH) 3 , is a poorly sorbing complex.It should be noted that concentrations of arsenic in alluvial and coastal sediments depend on mineralogy (the amounts of Fe oxides and sulfides), texture, and sorbing competition from other ions.For example, the sorbed As contents in Holocene alluvial sediments are in the range of 10-22 mg/kg and 1.2-5.9mg/kg in parts of the USA [36] and Bangladesh [4], respectively.Seawater pH is in range of 7.5 to 8.5.The simulations assess the desorption of various arsenic species in response to pH increases caused by saltwater intrusion.The modeling results show that significant desorption of As(V) and minor desportion of As(III) occur as pH increases from 7 to 9 in a sliding activity path (Figs.4a-b).The final arsenic concentration in the fluid increases to more than 250 µg/kg while the As sorbed onto Fe(OH) 3 drops to 19.28 mg/kg (19.00 mg/kg of As(V) and 0.28 mg/kg of As(III)).A substantial increase in dissolved as concentrations (> 100 µg/kg) occur when pH > 8.5 (Fig. 4b).

Fig. 4. (A) Dissolved total dissolved arsenic concentrations (B) and the desorption of main As species (As(OH) 3 as As(III) and HAsO 4 2as As(V)) versus pH calculated by HFO desorption simulations with pH sliding from 7 to 9
The second geochemical model assesses how various ions in seawater may replace arsenic sorbed onto the surface of HFOs.The modeling results show that the desorption may result in an increase of a few to a few tens of µg As kg -1 in solution when the initial system reacts with more than one pore volumes of seawater (Fig. 5).Concurrent increases of arsenic and salinity levels are found in the alluvial aquifers of Taiwan [43] and Bangladesh [26], indicating that many ions (e.g., silicate, carbonate, calcium, sulfate, etc.) in salty groundwater or seawater would compete with arsenic for sorption sites of hydrous oxides.Global warming may cause an additional +0.57 to +1.10 meter of sea level rise by 2100 [44,45].If the rates of global warming and sea level rise accelerate, more arsenic may be mobilized in aquifers along the coast by ionic competition processes and pH effect.

CLIMATIC EFFECTS ON METAL CYCLING IN COASTAL SALT MARSHES
Recent studies suggest that direct atmospheric deposition and riverine input are the primary sources of trace metals (e.g., Hg and As) to estuaries and coastal watersheds [46][47][48][49].
Since the brackish water is not potable, metals associated with coastal salt marshes do not pose direct health risks as those in alluvial aquifers.However, in aqueous environments, inorganic mercury is taken up by anaerobic micro-organisms [50] and transformed into the toxic form methyl mercury (CH 3 Hg).The health risks associated with biotic accumulation of Hg and other toxic metals underscore the necessity to understand fate and transformation of metals in coastal watersheds.Previous studies have shown that the mixing of large influx of freshwater with seawater in estuarines creates a favorable environment (i.e., the presence of warm, acidic, and low-salinity waters) for mercury methylation [18,23,[50][51].This case study examines the influences of climatic, hydrologic, and microbial processes on biogeochemical cycles of trace metals in coastal watersheds.The Weeks Bay (Fig. 6) in Alabama was chosen because of its close proximity to the mixing zone of freshwater and seawater.The Weeks Bay is impacted by metal pollution like many other estuaries and coastal environments.Largemouth Bass in the Weeks Bay contain mercury concentrations > 1 mg/kg [52,53], the limit set by the Food and Drug Administration.The nearby Wolf Bay (Fig. 6), which has not shown elevated biotic metal concentration, was chosen as a control site for comparison.The main objectives were to (1) analyze total mercury deposition from atmospheric sources and its seasonal variations, and (2) evaluate sediment geochemical profiles and microbial community to study the source, distribution, and fate of trace metals with depth.and 8) and AL24 (not shown) near Mobile Bay.Both sites show seasonal trends of highest mercury deposition occurring during wet months of spring, whereas lower mercury deposition generally occurs in the dry summer months (Fig. 7).Clearly higher mercury deposition correlates with periods of higher precipitation (Fig. 8) (correlation coefficients = 0.81 and 0.78 for AL02 and AL24, respectively), supporting the hypothesis that atmospheric deposition represents an important source of mercury and perhaps other airborne metals in the coastal watersheds.Since both Weeks Bay and Wolf Bay receive about the same amount of precipitation, higher levels of Hg and other trace metals found in sediments of the Weeks Bay (see sections below) are likely derived from its higher riverine inputs from Fish River.The Weeks Bay receives far more freshwater inputs from Fish River, the largest surface stream in Baldwin County study area.
The mercury data can be used to assess the impact of climate change on key atmospheric processes that control the transport of mercury from emission to deposition.Atmospheric deposition and riverine input appear to provide the primary pathways by which mercury and other trace metals enter aquatic environments in which bacterial methylation and subsequent accumulation in the food chain can occur.Incremental impacts of climate change on riverine transport and atmospheric deposition of trace metals can be quantified by geochemical analysis of sediments recovered from coastal salt marshes.and 8) and AL24 (not shown) near Mobile Bay.Both sites show seasonal trends of highest mercury deposition occurring during wet months of spring, whereas lower mercury deposition generally occurs in the dry summer months (Fig. 7).Clearly higher mercury deposition correlates with periods of higher precipitation (Fig. 8) (correlation coefficients = 0.81 and 0.78 for AL02 and AL24, respectively), supporting the hypothesis that atmospheric deposition represents an important source of mercury and perhaps other airborne metals in the coastal watersheds.Since both Weeks Bay and Wolf Bay receive about the same amount of precipitation, higher levels of Hg and other trace metals found in sediments of the Weeks Bay (see sections below) are likely derived from its higher riverine inputs from Fish River.The Weeks Bay receives far more freshwater inputs from Fish River, the largest surface stream in Baldwin County study area.
The mercury data can be used to assess the impact of climate change on key atmospheric processes that control the transport of mercury from emission to deposition.Atmospheric deposition and riverine input appear to provide the primary pathways by which mercury and other trace metals enter aquatic environments in which bacterial methylation and subsequent accumulation in the food chain can occur.Incremental impacts of climate change on riverine transport and atmospheric deposition of trace metals can be quantified by geochemical analysis of sediments recovered from coastal salt marshes.and 8) and AL24 (not shown) near Mobile Bay.Both sites show seasonal trends of highest mercury deposition occurring during wet months of spring, whereas lower mercury deposition generally occurs in the dry summer months (Fig. 7).Clearly higher mercury deposition correlates with periods of higher precipitation (Fig. 8) (correlation coefficients = 0.81 and 0.78 for AL02 and AL24, respectively), supporting the hypothesis that atmospheric deposition represents an important source of mercury and perhaps other airborne metals in the coastal watersheds.Since both Weeks Bay and Wolf Bay receive about the same amount of precipitation, higher levels of Hg and other trace metals found in sediments of the Weeks Bay (see sections below) are likely derived from its higher riverine inputs from Fish River.The Weeks Bay receives far more freshwater inputs from Fish River, the largest surface stream in Baldwin County study area.
The mercury data can be used to assess the impact of climate change on key atmospheric processes that control the transport of mercury from emission to deposition.Atmospheric deposition and riverine input appear to provide the primary pathways by which mercury and other trace metals enter aquatic environments in which bacterial methylation and subsequent accumulation in the food chain can occur.Incremental impacts of climate change on riverine transport and atmospheric deposition of trace metals can be quantified by geochemical analysis of sediments recovered from coastal salt marshes.

Sediment Characteristics and Chemistry
Grain size distribution data [54,55] indicate that the Weeks Bay sediments are dominated by fine-grain mud (49 to 53% of clay and 47 to 51% of silt) whereas the Wolf Bay core consists mostly of fine-grain sand (69-85%) with much lower contents of clay and silt.The presence of fine-grain materials is important for retention of metals in the Weeks Bay, as heavy metals tend to concentrate in fine-grain sediments with high-surface sorbing areas.
A total of 20 sediment samples from different depths in the sediment cores have been processed and analyzed for their total organic carbon (TOC) and trace metal contents.The organic carbon contents (TOC) of bulk sediments are higher in the Weeks Bay (2.040.24%) than those in the Wolf Bay (1.760.37%).Higher levels of organic carbon in sediments in

Sediment Characteristics and Chemistry
Grain size distribution data [54,55] indicate that the Weeks Bay sediments are dominated by fine-grain mud (49 to 53% of clay and 47 to 51% of silt) whereas the Wolf Bay core consists mostly of fine-grain sand (69-85%) with much lower contents of clay and silt.The presence of fine-grain materials is important for retention of metals in the Weeks Bay, as heavy metals tend to concentrate in fine-grain sediments with high-surface sorbing areas.
A total of 20 sediment samples from different depths in the sediment cores have been processed and analyzed for their total organic carbon (TOC) and trace metal contents.The organic carbon contents (TOC) of bulk sediments are higher in the Weeks Bay (2.040.24%) than those in the Wolf Bay (1.760.37%).Higher levels of organic carbon in sediments in

Sediment Characteristics and Chemistry
Grain size distribution data [54,55] indicate that the Weeks Bay sediments are dominated by fine-grain mud (49 to 53% of clay and 47 to 51% of silt) whereas the Wolf Bay core consists mostly of fine-grain sand (69-85%) with much lower contents of clay and silt.The presence of fine-grain materials is important for retention of metals in the Weeks Bay, as heavy metals tend to concentrate in fine-grain sediments with high-surface sorbing areas.
A total of 20 sediment samples from different depths in the sediment cores have been processed and analyzed for their total organic carbon (TOC) and trace metal contents.The organic carbon contents (TOC) of bulk sediments are higher in the Weeks Bay (2.040.24%) than those in the Wolf Bay (1.760.37%).Higher levels of organic carbon in sediments in the Weeks Bay are likely derived from high rates of erosion and riverine transport of terrestrial carbon associated with Fish River.Organic carbon flux is often higher in watersheds with great weathering and sediment yielding rates [56,57].
Geochemical results from total digestion geochemical analysis of the sediment samples show variations in concentrations of trace elements with depth (Table 1).Enrichment of trace metals and sulfur in salt marsh sediments is revealed by high values aluminum-normalized enrichment factors ANEF(see Appendix).In the Weeks Bay, concentrations of some trace metals (e.g., Cu, Pb, Zn, Fe, Hg, As, etc.) are generally greatest near the water-sediment interface but decline with depth (Fig. 9).The enrichment of trace metals in shallow (postindustrial) sediments is directly related to riverine inputs derived from industrial sources [48].The Wolf Bay does not show this enrichment pattern.Sedimentation rates in the Gulf coastal marshes are generally less than 0.3 cm/yr [58,59], thus the top 20 cm of sediments (with ages ranging from modern to about 60-70 years) mostly likely exhibit higher levels of trace metals derived from industrial sources.The abundance of fine-grain mud and organic matter in the Weeks Bay further facilitates the retention of trace metals there.The sediment trace metals levels in the Weeks Bay are much higher than those found in the Wolf Bay (Table 2).Sediment mercury concentrations exhibit large depth dependence and range from 0.04 to 0.08 and 0 to 0.02 mg/kg, respectively, at the Weeks Bay and Wolf Bay sites.Levels of As, Pb, Co, Ni, Zn, Cu, Sr, Ba, P, and S in the Weeks Bay sediments are significantly higher than those in the Wolf Bay.Interestingly, concentrations of trace metals are generally low in pore-water extracted from sediments [54].We hypothesize that high organic matter content and bacterially-mediated sulfate reduction facilitate metal retention via formation of sulfide solids in the marsh sediments.

Bacterial Sulfate Reduction and Its Impacts on Metal Cycling and Atmosphere CO 2 and O 2
Scanning Electron Microscope (SEM) and laser ablasion inductively coupled plasma mass spectrometry (LA-ICP-MS) analyses reveal the mineralogy, chemical compositions, grain size, and authigenic nature of solids precipitated from pore-water under reducing conditions.Analyzed core samples were chosen from a shallow depth interval (< 20 cm below surface) in the Weeks Bay where metals and sulfur-rich sediments occur.Biogenic sulfide minerals such as pyrite with distinct framboidal form (Fig. 10) are present in the Weeks Bay core samples.The occurrence of aggregates of equi-dimensional spheroids of about 0.5-2 µm (Fig. 10) of iron sulfide has been interpreted as the pyritized corpses of nannobacterial cells embedded in decaying organic matter [60].The biomineralization of framboidal pyrite was likely performed by sulfate reducing bacteria fueled by organic matter.The formation of framboidal pyrite indicates that sulfate reducing conditions are well established in salt marsh sediments [61,62].Concentrations of selected trace metals (Table 3) were measured in individual framboid and detrital pyrite grains recovered from the Weeks Bay cores using LA-ICP-MS (Fig. 11).Average S and Fe concentrations in framboid pyrite are 37.9 and 57.4 %, respectively.Mn, Pb, Zn, and V are the most abundant trace elements (average concentrations > 1,000 mg/kg) in the analyzed framboid pyrites.The samples also contain various amounts of Co, Ni, Cu, As, Se, Hg, Mo, and Cd.Hg and Te are present but their concentrations are not calculated due to the lack of standard references.The large variations in chemical composition also suggest a biogenic origin for framboid pyrite.Chemical compositions of detrital pyrite (Table 3) are different from those of framboid pyrite grains.The average trace metal contents are much lower in detrital pyrites than those in the framboid form.Mn, Zn, Ni, and As are the most abundant trace elements in the analyzed detrital pyrites; but their maximum concentrations are less than 65 mg/kg.The early stage framboid iron sulfide in amorphous form is capable of adsorbing and incorporating trace metals into their structure.However over time iron sulfide grains may lose their sorbing capability as they age and reorganize into larger and better-defined atomic structure.Opportunistic microorganisms, such as sulfate reducers, may take advantage of favorable geochemical conditions (i.e., with abundant electron donors/acceptors and available carbon sources) to grow quickly and randomly assemble nearby elements into pyrite with unique framboid form.
Bacterial sulfate reduction may be expressed as one of the following reactions: Where CH 2 O represents organic matter.Hydrogen sulfide (H 2 S) produced by bacterial sulfate reaction can quickly react with dissolved metals (e.g., Fe 2+ ) in pore-waters to precipitate iron sulfides (Fig. 10   Reaction (7) not only limits the mobility of trace metals and their dissolved loads in porewater, it may also regulate CO 2 and O 2 oxygen levels in the atmosphere.The roles of sulfur cycling in regulating atmospheric carbon and oxygen have often been ignored.This study shows that pyrite-like sulfide forms in salt marshes (Fig. 10) or sea floor [20] when microbes take up sulfur in the form of sulfate (with four bound oxygen atoms) in seawater and consequently release sulfide along with extra oxygen as free O 2 or CO 2 .More research is needed to assess how bacterial sulfate reduction and sulfide formation may regulate global oxygen and carbon cycles under current changing climate conditions and rising sea levels.
Elucidating the bacterial sulfate reduction can help understand the climatic and hydrological factors controlling the contamination and bioaccumulation of Hg and other metals in aquatic ecosystems.Sulfide produced from sulfate reduction may also react with dissolved metals to form insoluble metal sulfides (Fig. 10); such reaction would decrease the availability of dissolved Hg for methylation by sequestering it in insoluble metal sulfides [18,[20][21].Moreover, Compeau and Bartha [23] showed that Hg methylation may be inhibited in highsalinity environments where sulfate reducing bacteria compete with methanogenic bacteria for limited organic matter produced by fermentation organisms.Sea level rise has resulted in salinity increases of many coastal estuaries and bays and intrusion of seawater into fresh coastal ground water resources [63].Climate change models have predicted that the global sea level could increase between 18 cm and 59 cm during this century [64]; other worst-case scenario predictions have forecasted even higher sea-level rise of up to 180 cm in part of the Gulf of Mexico Coast regions [65].We argue that climate change and related hydrological modification will significantly impact the biogeochemical cycles of heavy metals and carbon in the coastal aquatic ecosystems.

FUTURE CHALLENGES
One of the most noteworthy predicted biogeochemical changes in response to climate change and sea-level rise is a widespread increase in groundwater arsenic levels in alluvial and coastal watersheds.Our field data and computer models link the enrichment of groundwater arsenic with high sea-level stand, high river base level, sluggish hydrologic flushing, and reducing geochemical conditions.The biogeochemical cycling of trace metals in alluvial and coastal environments is very complex because it has to be considered within the context of geology (e.g., grain size and mineralogy) and many climatic, hydrologic, and microbial proccesses, such as changes in temperature, salinity, pH, and redox conditions, bacterial-mediated biogeochemical reactions (i.e., iron and sulfate reduction, methylation), biomineralization (e.g., formation of iron sulfide), and adsorption/desorption reactions.Hydrologic and geochemical models are particularly useful for providing insights into these complex interactions but are limited to represent processes to the extent that they are quantitatively formulated.Some of the modeling results may be better constrained by field data when available.Measuring climate events, hydrological modifications, and biogeochemical processes in the field over time could be challenging because they may occur at different temporal and spatial scales in different geological environments.Perhaps the greatest challenge is in our understanding of the feedback loops between the climate system and biogeochemical cycling.For example, climate-induced biogeochemical processes may provide a source or a sink for oxygen or greenhouse gases in the atmosphere.An integrated research framework consisting of numerical modeling, long-term monitoring, laboratory experiments will be necessary for building a comprehensive understanding of the complex response of biogeochemical cycling of trace metals to climate change.

Fig. 1 .
Fig. 1.Diagram shows the depth-wise variations in groundwater arsenic concentration in Pliocene, Pleistocene, and Holocene alluvial aquifers in Bangladesh [26].Arsenic levels are elevated during Holocene high sea level stand

Fig. 1 .
Fig. 1.Diagram shows the depth-wise variations in groundwater arsenic concentration in Pliocene, Pleistocene, and Holocene alluvial aquifers in Bangladesh [26].Arsenic levels are elevated during Holocene high sea level stand

Fig. 1 .
Fig. 1.Diagram shows the depth-wise variations in groundwater arsenic concentration in Pliocene, Pleistocene, and Holocene alluvial aquifers in Bangladesh [26].Arsenic levels are elevated during Holocene high sea level stand

Fig. 2 . 3 /
Fig. 2. Calculated breakthrough curves showing the relative concentration of arsenic along a 1000 m-long aquifer in response to infiltration of freshwater after 50, 1000, 2000, and 4000 years.Parameters used in the model: k d = 5 cm 3 /g, =10 m, = 0.20, (of sediment) = 2.2 g/cm 3 , v = 0.1 m/day.Without new arsenic mobilized from sediments, the initial dissolved arsenic loads will be flushed out in about 4000 years 2.2 Bacterial Reduction of Fe(III) and Mn(IV) Oxides in Reaction-Dominated System (Sea Level High Stand)

Fig. 2 . 3 /
Fig. 2. Calculated breakthrough curves showing the relative concentration of arsenic along a 1000 m-long aquifer in response to infiltration of freshwater after 50, 1000, 2000, and 4000 years.Parameters used in the model: k d = 5 cm 3 /g, =10 m, = 0.20, (of sediment) = 2.2 g/cm 3 , v = 0.1 m/day.Without new arsenic mobilized from sediments, the initial dissolved arsenic loads will be flushed out in about 4000 years

Fig. 2 . 3 /
Fig. 2. Calculated breakthrough curves showing the relative concentration of arsenic along a 1000 m-long aquifer in response to infiltration of freshwater after 50, 1000, 2000, and 4000 years.Parameters used in the model: k d = 5 cm 3 /g, =10 m, = 0.20, (of sediment) = 2.2 g/cm 3 , v = 0.1 m/day.Without new arsenic mobilized from sediments, the initial dissolved arsenic loads will be flushed out in about 4000 years

Fig. 3 .
Fig. 3. (A) Predicted sequence of mineralogical reactions resulting from bacteria reduction of Fe and Mn oxides in equilibrium with Moundville groundwater near the Eutaw aquifer outcrop.The plot shows changes in mineral volume as acetate is titrated into the system and Eh decreases with time.Positive changes indicate precipitation, and negative changes show dissolution.(B) Calculated total Mn and Fe concentrations in fluid predicted by the same reaction path model

Fig. 3 .
Fig. 3. (A) Predicted sequence of mineralogical reactions resulting from bacteria reduction of Fe and Mn oxides in equilibrium with Moundville groundwater near the Eutaw aquifer outcrop.The plot shows changes in mineral volume as acetate is titrated into the system and Eh decreases with time.Positive changes indicate precipitation, and negative changes show dissolution.(B) Calculated total Mn and Fe concentrations in fluid predicted by the same reaction path model

Fig. 3 .
Fig. 3. (A) Predicted sequence of mineralogical reactions resulting from bacteria reduction of Fe and Mn oxides in equilibrium with Moundville groundwater near the Eutaw aquifer outcrop.The plot shows changes in mineral volume as acetate is titrated into the system and Eh decreases with time.Positive changes indicate precipitation, and negative changes show dissolution.(B) Calculated total Mn and Fe concentrations in fluid predicted by the same reaction path model

Fig. 5 .
Fig. 5. Calculated dissolved concentrations of arsenic in water as seawater is incrementally added into a freshwater-bearing aquifer, calculated accoring to the Dzombak and Morel (1990) surface complexation model

Fig. 6 . 3 . 1
Fig. 6.Simplified location map of the Weeks Bay and Wolf Bay salt marsh study sites in Alabama 3.1 Source of Mercury Analysis of 2001-2009 data of weekly precipitation and associated mercury levels reveal possible sources of Hg in the Weeks Bay and surrounding areas.The data, obtained from the Mercury Deposition Network (MDN) (http://nadp.sws.uiuc.edu/mdn/),include weekly precipitation (mm), total mercury concentrations (ng/L) in precipitation, and total mercury wet deposition (ng/m 2 ) collected from two Alabama MDN sites AL02 (Figs. 7and 8) and AL24 (not shown) near Mobile Bay.Both sites show seasonal trends of highest mercury deposition occurring during wet months of spring, whereas lower mercury deposition generally occurs in the dry summer months (Fig.7).Clearly higher mercury deposition correlates with periods of higher precipitation (Fig.8) (correlation coefficients = 0.81 and 0.78 for AL02 and AL24, respectively), supporting the hypothesis that atmospheric deposition represents an important source of mercury and perhaps other airborne metals in the coastal watersheds.Since both Weeks Bay and Wolf Bay receive about the same amount of precipitation, higher levels of Hg and other trace metals found in sediments of the Weeks Bay (see sections below) are likely derived from its higher riverine inputs from Fish River.The Weeks Bay receives far more freshwater inputs from Fish River, the largest surface stream in Baldwin County study area.

British
Journal of Environment & Climate Change, 3(1): 44-66, 2013 53 atmospheric sources and its seasonal variations, and (2) evaluate sediment geochemical profiles and microbial community to study the source, distribution, and fate of trace metals with depth.

Fig. 6 . 3 . 1
Fig. 6.Simplified location map of the Weeks Bay and Wolf Bay salt marsh study sites in Alabama 3.1 Source of Mercury Analysis of 2001-2009 data of weekly precipitation and associated mercury levels reveal possible sources of Hg in the Weeks Bay and surrounding areas.The data, obtained from the Mercury Deposition Network (MDN) (http://nadp.sws.uiuc.edu/mdn/),include weekly precipitation (mm), total mercury concentrations (ng/L) in precipitation, and total mercury wet deposition (ng/m 2 ) collected from two Alabama MDN sites AL02 (Figs. 7and 8) and AL24 (not shown) near Mobile Bay.Both sites show seasonal trends of highest mercury deposition occurring during wet months of spring, whereas lower mercury deposition generally occurs in the dry summer months (Fig.7).Clearly higher mercury deposition correlates with periods of higher precipitation (Fig.8) (correlation coefficients = 0.81 and 0.78 for AL02 and AL24, respectively), supporting the hypothesis that atmospheric deposition represents an important source of mercury and perhaps other airborne metals in the coastal watersheds.Since both Weeks Bay and Wolf Bay receive about the same amount of precipitation, higher levels of Hg and other trace metals found in sediments of the Weeks Bay (see sections below) are likely derived from its higher riverine inputs from Fish River.The Weeks Bay receives far more freshwater inputs from Fish River, the largest surface stream in Baldwin County study area.

British
Journal of Environment & Climate Change, 3(1): 44-66, 2013 53 atmospheric sources and its seasonal variations, and (2) evaluate sediment geochemical profiles and microbial community to study the source, distribution, and fate of trace metals with depth.

Fig. 6 . 3 . 1
Fig. 6.Simplified location map of the Weeks Bay and Wolf Bay salt marsh study sites in Alabama 3.1 Source of Mercury Analysis of 2001-2009 data of weekly precipitation and associated mercury levels reveal possible sources of Hg in the Weeks Bay and surrounding areas.The data, obtained from the Mercury Deposition Network (MDN) (http://nadp.sws.uiuc.edu/mdn/),include weekly precipitation (mm), total mercury concentrations (ng/L) in precipitation, and total mercury wet deposition (ng/m 2 ) collected from two Alabama MDN sites AL02 (Figs. 7and 8) and AL24 (not shown) near Mobile Bay.Both sites show seasonal trends of highest mercury deposition occurring during wet months of spring, whereas lower mercury deposition generally occurs in the dry summer months (Fig.7).Clearly higher mercury deposition correlates with periods of higher precipitation (Fig.8) (correlation coefficients = 0.81 and 0.78 for AL02 and AL24, respectively), supporting the hypothesis that atmospheric deposition represents an important source of mercury and perhaps other airborne metals in the coastal watersheds.Since both Weeks Bay and Wolf Bay receive about the same amount of precipitation, higher levels of Hg and other trace metals found in sediments of the Weeks Bay (see sections below) are likely derived from its higher riverine inputs from Fish River.The Weeks Bay receives far more freshwater inputs from Fish River, the largest surface stream in Baldwin County study area.

Fig. 7 . 2 )Fig. 8 . 2 )
Fig. 7. Fluctuations of total mercury wet deposition (in ng/m 2 ) responding to rainfall (precipitation, in mm) in southwestern Alabama from December 2008 to December 2009.The diagram shows increased mercury deposition during periods of higher precipitation, suggesting atmospheric deposition of mercury pollution

British 54 Fig. 7 . 2 )Fig. 8 . 2 )
Fig. 7. Fluctuations of total mercury wet deposition (in ng/m 2 ) responding to rainfall (precipitation, in mm) Alabama from December 2008 to December 2009.The diagram shows increased mercury deposition during periods of higher precipitation, suggesting atmospheric deposition of mercury pollution

British 54 Fig. 7 . 2 )Fig. 8 . 2 )
Fig. 7. Fluctuations of total mercury wet deposition (in ng/m 2 ) responding to rainfall (precipitation, in mm) in southwestern Alabama from December 2008 to December 2009.The diagram shows increased mercury deposition during periods of higher precipitation, suggesting atmospheric deposition of mercury pollution

Fig. 9 .
Fig. 9. Depth-wise variations in the concentrations of selected trace metals in sediments of Weeks Bay

Fig. 11 .
Fig. 11.LA-ICP-MS spectra for weeks bay pyrite samples Web1 (a), Web6a (b), Web6b (c), and Web 7(d).The average trace element concentrations are shown in Table 3.Samples show high levels of Fe and S, along with varying quantities of Pb, Zn, As, Hg, etc. Framboid sample Web 7 shows zoning of several metals (e.g., Fe, Zn, Pb, and As).While trace metal contents show large variations in framboid pyrites, detrital pyrite grains web6a and web6b exhibit "fixed" Fe/S molar ratios ranging from 0.52-0.58