Bacterial production and their role in the removal of dissolved organic matter from tributaries of drinking water reservoirs

Highlights

• DOM was considerably degraded in summer but not during flood events.

• The ratio of DOM removal to total DOM release was negatively related to discharge.

• Phosphorus addition did not stimulate bacterial DOM degradation.

• A proportion of 0.6–12% of the total DOM was degraded in the tributaries.

Abstract

Enhanced concentrations of dissolved organic matter (DOM) in freshwaters are an increasing problem in drinking water reservoirs. In this study we investigated bacterial DOM degradation rates in the tributaries of the reservoirs and tested the hypotheses that (1) DOM degradation is high enough to decrease DOM loads to reservoirs considerably, (2) DOM degradation is affected by stream hydrology, and (3) phosphorus addition may stimulate bacterial DOM degradation. Bacterial biomass production, which was used as a measure of DOM degradation, was highest in summer, and was usually lower at upstream than at downstream sites. An important proportion of bacterial production was realized in epilithic biofilms. Production of planktonic and biofilm bacteria was related to water temperature. Planktonic production weakly correlated to DOM quality and to total phosphorus concentration. Addition of soluble reactive phosphorus did not stimulate bacterial DOM degradation. Overall, DOM was considerably degraded in summer at low discharge levels, whereas degradation was negligible during flood events (when DOM load in reservoirs was high). The ratio of DOM degradation to total DOM release was negatively related to discharge. On annual average, only 0.6–12% of total DOM released by the catchments was degraded within the tributaries.

Introduction

Globally, streams and lakes receive an estimated amount of 2.0–2.7 billion tons of terrestrial carbon per year (Cole et al., 2007, Battin et al., 2008, Aufdenkampe et al., 2011). The majority of the terrestrial organic carbon entering freshwater systems is respired to CO₂ locally or buried in sediments, whereas only a fraction is discharged into the ocean (Aufdenkampe et al., 2011). The organic carbon in rivers and streams affects aquatic ecosystems and serves as a major energy source for microbes (Findlay et al., 1993, Jonsson et al., 2007). Particularly freshly leached, terrestrial dissolved organic matter (DOM) may influence stream ecosystem processes (Burrows et al., 2013). In addition, the dynamics of DOM mobilization is typically linked to hydrology. About 50% of the DOM was exported by the 10% highest discharges (Hilton et al., 1997). Snowmelt events and heavy rainfall events in summer causing high discharges mobilized the highest DOM fluxes in a small catchment, and spring areas with high DOM pools and pronounced hydrological dynamics were regarded as important sources (Andrews et al., 2011). The concept of hot spots and hot moments (e.g., McClain et al., 2003) may be applied to the DOM export from catchments: Pools of available DOM (e.g., upland areas) are hot spots only during certain periods (hot moments; e.g., heavy rainfall) when hydrological connectivity between the terrestrial sources of DOM and stream discharge is facilitated, enabling a relevant mass flux to surface waters (Casper et al., 2003, Stieglitz et al., 2003).

Once the DOM has entered the surface water ecosystem, instream heterotrophic processing of terrestrial derived carbon occurs along the river continuum gradient from the headwater downstream (River Continuum Concept, Vannote et al., 1980). DOM is mainly consumed by heterotrophic bacteria, whereby it is partly removed by respiration and partly transferred to bacterial biomass, the latter process making the carbon available to organisms of higher trophic levels. Terrigenous DOM was shown to be respired by microorganisms rather than incorporated into their biomass (Fashing et al., 2014). Besides bacteria suspended in water, bacteria also grow in biofilms which are substratum-associated consortia of microorganisms including microalgae, bacteria, fungi, protozoans and small metazoans, as well as their extracellular polymeric substances (EPS). Biofilms, e.g. on stones, are regarded as major sites of carbon cycling in streams (Romani et al., 2004, Battin et al., 2008).

The use of DOM as a bacterial energy source is controlled by the concentration of limiting nutrients (N and/or P), temperature, and the chemical composition of DOM (Stelzer et al., 2003, Lane et al., 2013). Nutrient enrichment increased biofilm development and reduced the bacterial use of autochthonous carbon within the biofilm (Ziegler and Lyon, 2010). Carbon uptake of biofilms was sensitive to temperature (Baldwin et al., 2014). Land use and vegetation type have a major impact on the composition of DOM, affecting the degree of humification within catchments. DOM in agricultural streams was shown to be more labile and thus more accessible for microbes than DOM in wetland streams (Williams et al., 2010). A positive relationship was found between the aromatic content of DOM and the proportion of forested area in the river network of the Bode catchment in Germany (Kamjunke et al., 2013). Regarding the relationship between DOM quality and bacterial processing, planktonic bacterial production was positively related to labile DOM concentration in streams in Southern Ontario (Williams et al., 2010), and microbial bioavailability of DOM was negatively related to the proportions of humic-like DOM in streams of Maryland (Hosen et al., 2014). Recently, Kamjunke et al. (2015) compared planktonic and biofilm bacterial production with patterns of DOC along a topographic and land-use gradient at 17 sampling sites on one occasion (base flow in late summer). It was found that planktonic production was weakly correlated to the total DOC concentration but strongly to DOM quality. Biofilm production, on the other hand, was independent of both DOC concentration and DOM quality. However, little is known about the effect of stream hydrology on bacterial activity. Bacterial production, measured only during one investigated period, was found to be low at high discharge in Danube floodplains (Peduzzi et al., 2008).

Increasing concentrations of dissolved organic matter in freshwaters due to rising input of organic carbon is a common phenomenon in many regions (Hejzlar et al., 2003, Freeman et al., 2004, Eimers et al., 2008). Monteith et al. (2007) observed such an increase in more than 70% of the investigated surface waters in Northern Europe, Great Britain, and North America. An increase in DOM in English streams of 65% between 1988 and 2000 was found to be particularly caused by DOM from the catchment area, as indicated by the enrichment of humic substances (Freeman et al., 2001); the enhancement was even threefold in the Hudson River (Findlay, 2005). In Germany, a significant increase in DOM concentration was observed in 55% of 86 streams (Sucker et al., 2011). The browning of surface waters leads to increasing problems regarding the production of drinking water. Increasing DOC chlorination cause the formation of mutagenic (Nieuwenhuijsen et al., 2000) and carcinogenic (Sadiq and Rodriguez, 2004) disinfection by-products, taste and odor problems as well as growth of potential pathogens, adversely affect the function of water treatment plants and increase costs for purification of drinking water (Ledesma et al., 2012). The amount of chemicals for flocculation and precipitation rises, the running time of filters decreases, and the amount of sludge in water works increases. The corresponding challenge is to find a cost-effective DOM removal technique which makes use of the instream bacterial production potential at different seasons and along the river-continuum gradient with nutrient addition as a possible management practice.

In light of these growing concerns regarding increased DOM inputs into aquatic ecosystems, it is crucial that we understand how much instream bacterial production can contribute to the removal of DOM along a downstream gradient at different seasons and flow conditions. Furthermore, there is a need to explore the feasibility of using nutrient additions (P) as a method of stimulating DOM removal. While one traditional aim of reservoir management is to limit phosphorus imports (Cooke et al., 2005) the mineralization of DOC by bacteria can be limited by phosphorus in particular at high DOC to P ratios (Vadstein and Olsen, 1989). The main goal of the present study is to investigate whether the DOM released from the catchments is transported mainly unchanged by the tributaries to the reservoirs, or if the DOM is degraded considerably within the inflow streams, leaving only a part of the released DOM for the load of the reservoir. If the latter were true, one might consider measures of stimulating DOM degradation to reduce the export to the reservoir. Therefore, we tested the following hypotheses that (1) bacterial DOM degradation rates in the tributaries of the reservoir are high enough to decrease DOM concentrations considerably, (2) DOM degradation is affected by stream hydrology, and (3) phosphate addition may release bacteria from phosphorus limitation and thereby stimulate bacterial DOM degradation.