Microbial degradation of terrigenous dissolved organic matter and potential consequences for carbon cycling in brown-water streams

Streams receive substantial terrestrial deliveries of dissolved organic matter (DOM). The chromophoric (CDOM) fraction of terrestrial deliveries confers the brown colour to streamwater, often understood as browning, and plays a central role in aquatic photochemistry and is generally considered resistant to microbial metabolism. To assess the relevance of terrigenous DOM for carbon fluxes mediated by stream microorganisms, we determined the bioavailable fraction of DOM and microbial carbon use efficiency (CUE), and related these measures to partial pressure of CO2 in headwater streams spanning across a browning gradient. Fluorescence and absorbance analyses revealed high molecular weight and aromaticity, and elevated contributions from humic-like components to characterize terrestrial CDOM. We found that microorganisms metabolized this material at the cost of low CUE and shifted its composition (from fluorescence and absorbance) towards less aromatic and low-molecular weight compounds. Respiration (from CUE) was related to CO2 supersaturation in streams and this relationship was modulated by DOM composition. Our findings imply that terrigenous DOM is respired by microorganisms rather than incorporated into their biomass, and that this channelizes terrigenous carbon to the pool of CO2 potentially outgassing from streams into the atmosphere. This finding may gain relevance as major terrigenous carbon stores become mobilized and browning progresses.

nm/ Ex < 250 (305) nm) is putatively associated with biological production in situ and therefore diagenetically younger 7 .
To characterize DOM properties and composition we computed the following indices from DOM absorbance and fluorescence: The humification index (HIX) was calculated as the peak area of the emission wavelengths from 435 to 480 nm divided by the peak area of the emission wavelengths from 300 to 445 nm, at an excitation wavelength of 254 nm 8 . Higher values indicate higher humic substance content or extent of humification 8 .The ß/α index was computed as the ratio of emission intensity at 380 nm (ß) to the maximum emission intensity between 420 and 435 nm (α) at an excitation wavelength of 310 nm 9 . High ß/α values indicate autochthonous inputs, i.e., recently microbially produced DOM 10 . The fluorescence index (FI) was calculated as the ratio of emission intensity at 450 nm to that at 500 nm for an excitation wavelength of 370 nm 11 ; it is commonly used as an indicator of DOM source (i.e. terrigenous vs. microbially derived DOM) 11 . The slope ratio (S R ) was calculated as the ratio of S 275-295 to S 350-400 ; it is reported to correlate inversely with DOM molecular weight 12 . The specific UV absorption (SUVA 254 ) was calculated as the absorption coefficient at 254 nm (m -1 ) relative to the DOC concentration (mg l -1 ) 13 . The SUVA 254 was reported to correlate positively with increasing DOM aromaticity 13 . Additionally we calculated the ratio of absorbance at 254 and 365nm (a254/a365), which was shown to be negatively correlated to molecular weight of DOM 14,15 .
Alkalinity was determined by Titration with 0.1 mM HCl.

Streamwater ΔCO 2 and respiration
To determine the streamwater CO 2 concentration we carefully submersed triplicate 50-ml glass vials (pre-conditioned with precipitated NaN 3 , 0.02 % final concentration) into the streamwater, and allowed them to overflow before closing them under water with a 1 cm thick gas-tight rubber stopper.
While closing the vials a needle was inserted into the rubber stopper to avoid disturbance and allow excess water to emerge from the vial. Triplicate atmospheric samples (10 ml each) were obtained using a gas-tight syringe and injected into pre-evacuated 7-ml headspace vials, fitted with silicon sealed rubber stoppers and aluminum caps. Gas samples were stored at 4°C in the dark for a maximum of 2 days. Prior to measurement a headspace of known volume was generated with N 2 in the liquid samples. Equilibration of the gaseous and the liquid phase was then achieved by means of a water bath at controlled temperature aided by shaking for 45 min. Samples were analysed using gas chromatography (Agilent Technologies 7890A) and a headspace autosampler. Temperature was kept constant during measurement.
During sample equilibration at lab conditions, CO 2 evades from the liquid phase into the N 2headspace, causing the equilibria of the carbonate fractions (CO 2 , HCO 3 and CO 3 2-) to shift. Thus, the CO 2 concentration measured in the headspace cannot be directly related to the CO 2 concentration in the liquid phase in the stream and at the respective field conditions. We accounted for this by calculating the total DIC concentration of the sample assuming constant alkalinity and employing the equilibrium constants for the respective carbonate fractions 16,17 and respective temperature and pressure conditions. Briefly, we computed the amount of moles of CO 2 in the liquid phase from measured CO 2 partial pressure in the headspace and Henry´s law assuming full equilibration. This computed sample CO 2 concentration (mol l -1 ), alkalinity and the respective equilibrium constants for the carbonate fractions (adjusted for temperature and ionic strength) were used to infer pH. Then the concentration of HCO 3 and CO 3 2could be calculated using pH and the adequate equilibration constants (adjusted for temperature and ionic strength). The sum of all carbonate fractions (in moles) in the liquid phase and the total amount of CO 2 (in moles) in the headspace gave the total amount of DIC (in moles).
We then calculated the CO 2 concentration (mol l -1 ) in the streamwater from computed DIC values and measured pH values using the equilibrium constants (adjusted for temperature and ionic strength). A potential streamwater CO 2 concentration in mol l -1 at equilibrated conditions was calculated from the measured atmospheric CO 2 concentrations using Henry´s law. The difference of the CO 2 concentration in the stream in mol l -1 to the concentration expected from the atmospheric equilibrium gives the degree of CO 2 oversaturation (ΔCO 2 ).