Development of associations between ... - Wiley Online Library

RESEARCH ARTICLE

Development of associations between microalgae and denitrifying bacteria in streams of contrasting anthropogenic in£uence Christopher G. Peterson1, Allison D. Daley2, Shannon M. Pechauer2, Kathryn N. Kalscheur3, Malachy J. Sullivan2, Samantha L. Kufta2, Miguel Rojas2, Kimberly A. Gray3 & John J. Kelly2 1

Department of Environmental Science, Loyola University Chicago, Chicago, IL, USA; 2Department of Biology, Loyola University Chicago, Chicago, IL, USA; and 3Department of Civil and Environmental Engineering, Northwestern University, Evanston, IL, USA

Correspondence: Christopher G. Peterson, Department of Environmental Science, Loyola University Chicago, 6460 N. Kenmore Ave., Chicago, IL 60660, USA. Tel.: 11 773 508 2950; fax: 11 773 508 2983; e-mail: [email protected] Received 6 January 2011; revised 26 April 2011; accepted 4 May 2011. Final version published online 16 June 2011. DOI:10.1111/j.1574-6941.2011.01131.x

MICROBIOLOGY ECOLOGY

Editor: Riks Laanbroek Keywords algal exudates; stream biofilms; algal/bacterial interactions; organic fingerprinting; diatom species; denitrifying bacteria.

Abstract We compared the development of microalgal and bacterial-denitrifier communities within biofilms over 28 days in a restored-prairie stream (RP) and a stream receiving treated wastewater effluent (DER). Inorganic nutrient concentrations were an order of magnitude greater in DER, and stream waters differed in the quality of dissolved organics (characterized via pyrolysis-GC/MS). Biofilm biomass and the densities of algae and bacteria increased over time in both systems; however, algal and denitrifier community composition and the patterns of development differed between systems. Specifically, algal and denitrifier taxonomic composition stabilized more quickly in DER than RP, whereas the rates of algal and denitrifier succession were more closely coupled in RP than DER. We hypothesize that, under unenriched conditions, successional changes in algal assemblages influence bacterial denitrifiers due to their dependence on algal exudates, while under enriched conditions, this relationship is decoupled. Between-system differences in organic signatures supported this, as RP biofilms contained more labile, aliphatic compounds than DER. In addition, potential denitrification rates (DNP) were negatively correlated with the percentage of aromatic compounds within the biofilm organic signatures, suggesting a significant relationship between algal exudate composition and denitrification. These results are significant because anthropogenic factors that affect biofilm community composition may alter their capacity to perform critical ecosystem services.

Introduction In aquatic ecosystems, biofilms are important sites of microbially mediated biogeochemical processes, as accrual of biomass and increases in mat density reduce the degree to which external factors influence internal processes. Within developed biofilms, activity of resident macro- and microorganisms produces transient areas of microenvironmental heterogeneity that vary in the concentrations of exoenzymes, dissolved nutrients or oxygen, the form and amount of dissolved carbon present, and pH (Lock, 1993; Kühl et al., 2007). Thus, microorganisms capable of anaerobic physiological processes, such as denitrification, are exposed episodically to conditions conducive for running these biochemical reactions. When such conditions do occur, the amount and nature of the proximate supply of dissolved FEMS Microbiol Ecol 77 (2011) 477–492

carbon can influence which bacterial phylotypes are most active (Haynes et al., 2007; Murray et al., 2007). Evidence of biological interactions among bacteria and algae is well documented, in both planktonic (Cole, 1982; Baines & Pace, 1991) and benthic (Haack & McFeters, 1982; Bruckner et al., 2008) communities, with bacterial cells often using algae-derived exudates as a carbon source to fuel their metabolism (Malinsky-Rushansky & Legrand, 1996; Bellinger et al., 2009). Such associations can be quite specific, as individual algal species can support taxonomically distinct consortia of satellite bacteria (Schäfer et al., 2002; Grossart et al., 2005; Lachnit et al., 2011). Organic compounds released from algal cells of different species can differ chemically (Widrig et al., 1996; Hamels et al., 2004; Bahulikar & Kroth, 2008), inducing species-specific chemotaxic responses in herbivores (Hamels et al., 2004) and bacteria (Grossart et al., 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

478

2007; Stocker et al., 2008). As such, the taxonomic composition of algal communities with which bacterial consortia are associated may, via the production of unique combinations of exudates, modify the rate and efficiency of biochemical processes for which these bacteria are responsible. However, the links between algal and bacterial processes can decouple or weaken under certain environmental conditions (e.g. Scott et al., 2008; Rusanov et al., 2009). With aquatic ecosystems increasingly modified by human activities, a better understanding of the mechanisms through which bacterially mediated biogeochemical processes are influenced by microalgal community attributes will enhance our ability to predict and ultimately mitigate human impacts on ecosystem services. Here, we present the results of an investigation comparing the rates and patterns of succession in benthic algae and consortia of denitrifying bacteria during biofilm development on artificial substrata in two streams differing in the nature of human activity affecting them. Our objectives were to (1) determine whether the rates at which benthicalgal assemblages and consortia of denitrifying bacteria develop differ between stream systems with substantially different human impacts; (2) determine whether patterns of algal and denitrifier community development differed between stream systems with substantially different human impacts; and (3) explore the potential links among algal and denitrifier community structure and function and the dissolved organic carbon (DOC) signatures within biofilms in these systems.

C.G. Peterson et al.

matrix of connected 5.1-cm2 tiles), were introduced into each stream on June 23, 2008. One sampling unit from each block, constituting three replicate samples per date, were collected after 3, 7, 14, 21, and 28 days of colonization and material was removed using a stiff-bristle toothbrush into a container. These samples were brought up to a known volume upon return to the laboratory, homogenized, and 25-mL aliquots were drawn from the slurry for the determination of ash-free dry mass (AFDM), algal density and taxonomic structure, total bacterial densities via direct counts, density of denitrifying bacteria via quantitative real-time PCR (Q-PCR), diversity of denitrifying bacteria via terminal restriction fragment length polymorphism (T-RFLP) analysis, and, for 28-day samples, characterization of organics via pyrolysis-GC/MS (Py-GC/MS). Water samples were collected for nutrient analyses 0, 7, and 21 days from the start of the experiment, frozen until filtered (prerinsed 0.45-mm Tuffryns membrane filters), and analyzed for dissolved nitrate (NO3-N) via the UV secondderivative spectroscopy method (Crumpton et al., 1992), dissolved ammonia (NH3-N) via the phenate method (Solorzano, 1969), and dissolved phosphate (PO4-P) via the ascorbic acid method (Murphy & Riley, 1962). At the start of the experiment, stream water was also collected and filtered (prerinsed 0.45-mm Tuffryns membrane filters) for the determination of DOC by high-temperature catalytic oxidation [APHA (1998) Standard Method 5310 B; Dohrmann Apollo 9000] and organic material characterization via Py-GC/MS, as described below.

Materials and methods Study sites The two study streams are located within the DuPage County (IL) Forest Preserve system, and both, while not contiguous, are named ‘Spring Brook’. The first, located within the city of Naperville, runs through a restored prairie within the Springbrook Prairie Preserve and has no riparian canopy. This site will hereafter be referred to as ‘RP’ (restored prairie). The second is located within the St James Farm Forest Preserve, 3.2 km downstream of the outflow of the Wheaton Sanitary District Wastewater Treatment Plant, and is partially shaded by a canopy of mixed hardwoods. This site will be referred to hereafter as ‘DER’ (downstream of effluent release). Photosynthetic Irradiation Loggers (Data Flow Systems, Christchurch, NZ) were deployed at each site to quantify photosynthetically active radiation (PAR) just above the stream surface.

Experimental design, sample collection, and processing Three 30.5 30.5 7.6 cm cinder blocks, each supporting five 10.2 10.2 cm ceramic-tile sampling units (each a 4 4 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

Biofilm characterization Subsamples for the determination of AFDM were dried at 105 1C for 24 h, weighed to obtain dry mass, and combusted at 500 1C for 4 h. Subsamples for algal analysis were preserved in Lugol’s solution. Some of this material was digested in 30% hydrogen peroxide and potassium permanganate and mounted in Naphraxs on microscope slides for the identification of diatom species before conducting quantitative algal counts. A second aliquot, of known volume, was drawn through 25-mm, 0.45-mm Metricel membrane filters, which were then cleared in immersion oil on microscope slides, and sealed along coverslip edges with clear fingernail polish. Species-specific estimates of algal density were based on counts at 1000 of at least 500 chloroplast-containing cells (i.e. presumed ‘live’ at time of sampling), using an Olympus BX light microscope. Algal biovolume was calculated using formulae of geometric shapes that approximated species-specific cell dimensions (Hillebrand et al., 1999). The number of cells measured for each taxon was based on the protocol of the Phycology section, Patrick Center for Environmental Research, The Academy of Natural Sciences (http://diatom.acnatsci.org/ FEMS Microbiol Ecol 77 (2011) 477–492

479

Algal/denitrifying-bacterial interactions in urban streams

nawqa/Biovol2001.aspx). Subsamples for bacterial counts were preserved in a filtered (0.2 mm) solution of 3.7% formaldehyde and 0.1 M tetrasodium pyrophosphate and stored at 4 1C until analyzed. Bacterial cells were separated by sonication for 30 s on ice. Sample aliquots were stained with SYBR Gold (Invitrogen, Carlsbad, CA) and filtered onto 0.2-mm Anodisc filters (Whatman, Piscataway, NJ). Filters were mounted on glass slides and bacterial cells were enumerated in a minimum of 10 fields using an Olympus BH2 epifluorescence microscope (Olympus, Center Valley, PA).

Characterization of denitrifying bacteria Subsamples for the quantification and characterization of bacterial denitrifiers were pelleted by centrifugation at 10 000 g for 30 s and stored at 80 1C until processing. DNA was extracted from pellets using MoBio Microbial DNA Isolation Kits (MoBio Inc., Solana, CA). Successful DNA isolation was confirmed by agarose gel electrophoresis. The amount of DNA isolated from each sample was determined using the Quant-iT DNA Assay Kit (Invitrogen), and ANOVA confirmed that there was no significant difference (P = 0.871) in DNA yields normalized by the total bacterial cell counts for biofilms from our two study sites. Denitrifying bacteria were quantified based on the copy numbers of nirS genes determined by a Q-PCR following Mincer et al. (2007). Primers nirS1F and nirS6R (Braker et al., 1998) were used to amplify an 890-bp fragment of the nirS gene. The standard used for quantification was genomic DNA isolated from Pseudomonas stutzeri ATCC 11607, which was assumed to have a genome size of 4 Mbp and 1 nirS copy per genome (Gruntzig et al., 2001). The P. stutzeri dilution series included 10-fold dilutions ranging from 1.2 105 to 12 copies of nirS. Q-PCR reactions were run using an MJ Research DNA Engine Opticon 1 thermal cycler equipped with OPTICON MONITOR software version 3.1 (Bio-Rad, Hercules, CA). The conditions for all Q-PCR reactions were as follows: 12.5 mL QuantiTect SYBR Green PCR Master Mix (Qiagen, Valencia, CA), 0.5 mM final concentration of each primer, 1 mL template, and water were added to a final 25 mL volume. All primers were synthesized by Eurofins MWG Operon (Huntsville, AL). All reactions were performed in low-profile 0.2-mL white strip tubes with optical ultra clear strip caps (Bio-Rad). Three analytical replicates were run for each sample. The specificity of Q-PCR reactions was confirmed by melting curve analysis and agarose gel electrophoresis. Thermal cycling was conducted as follows: initial denaturation at 95 1C for 10 min, 40 cycles of denaturation at 95 1C for 1 min, primer annealing at 57 1C for 1 min, extension at 72 1C for 1 min, hold at 78 1C for 1 s, and plate read. Finally, a melting curve was run from 50 to 95 1C with a read every 1 1C and a hold of 1 s between reads. Copy FEMS Microbiol Ecol 77 (2011) 477–492

numbers were normalized based on the surface area of the tile supporting the biofilm. Amplification of nirK genes was attempted using primers nirK1F and nirK5R and the PCR cycling parameters specified by Braker et al. (1998). Although the nirK gene could be amplified from genomic DNA of the reference strain Achromobacter xylosoxidan (ATCC 15173), amplification of this gene from DNA isolated from our biofilms was unsuccessful, despite repeated attempts to optimize PCR by modifying the template concentrations, reagent concentrations, and cycling parameters. Denitrifier community composition was assessed by T-RFLP analysis of nosZ genes. Primers nosZF and nosZ1622R and the PCR conditions and cycling parameters described previously (Throback et al., 2004) were used to amplify a 453-bp fragment of the nosZ gene. PCR products were analyzed by T-RFLP as described previously (Baniulyte et al., 2009), with the exception that digestions were performed using AluI and HinfI (New England Biolabs, Beverly, MA) according to the manufacturer’s instructions.

Characterization of organics The quality of organics within periphyton collected from ceramic tiles and of stream water was characterized using Py-GC/MS. The protocol for this procedure was developed specifically for the analysis of aquatic organic matter and is described in detail by Sirivedhin (2002). This technique allows the characterization of the chemical features of complex mixtures of nonvolatile, macromolecular organic material that cannot be analyzed using traditional methods. It also allows the sources of organics to be determined, and allows the tracking of transformations of molecular components of the organic matrix within biofilms. In the pyrolysis step, the organic material is thermally cleaved into volatile fragments, which are then swept into the GC for separation and the MS for identification. When pyrolysis is carried out under a set of controlled conditions, the parent structures produce predictable and reproducible sets of fragments, creating a chemical fingerprint reflecting the mixture’s organic characteristics. The fragments in the pyrochromatogram can provide structural information about the chemical building blocks of the organic material. This structural information allows samples to be differentiated based on material origin, system inputs, and/or biogeochemical processing (Irwin, 1982; Moldoveanu, 1998). The preparation of water samples involved first distilling approximately 4 L of the filtered water sample in a series of evaporations using rotary vacuum evaporation (Buchi Rotavapor Model R114 with B480 Waterbath) operating at 25 1C and under a vacuum pressure of approximately 27.5 in. Hg. To reduce salt interferences, which can exert matrix effects during pyrolysis, the concentrated liquid 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

480

samples were dialyzed using a 2000 molecular weight cutoff membrane (Thermo Scientific Slide-A-Lyzer G2 Dialysis Cassette). Both solid and concentrated water samples were lyophilized, homogenized, loaded into pyrolysis tubes (unsealed quartz tube, Scientific Instrument Services Inc.) to approximately 0.8–1.0 mg of carbon, and plugged with deactivated glass wool. Before analysis, 3 mg of an internal standard solution (2 g L1 polystyrene, MW = 3350, in 2-butanone) was added to the middle of each sample tube to monitor reproducibility among replicates. Polystyrene served as the internal standard because, under our experimental conditions, the GC elution time of styrene is at 30 min, which falls within our analytical window, and yet, in an area of the pyrochromatogram where few of the pyrolysis fragments produced by environmental samples elute. In addition, styrene is rarely found as a major pyrolysis product of aquatic organic matter, although minor amounts can be present. External standards of polystyrene and poly-L-tyrosine were run before each analysis series and within a series to verify proper instrument functioning. Once the sample or the standard was prepared, it was placed inside the platinum filament coil of the pyrolysis probe (Chemical Data Systems Pyroprobe 2000), which was then inserted into the interface (Chemical Data Systems 1500 valved GC interface). Conditions of the pyrolysis unit used to ensure the reproducibility of the analysis include: (1) pyrolysis interface temperature of 250 1C, (2) final pyrolysis temperature of 625 5 1C, (3) total pyrolysis time of 1 min, and (4) ramping rate of 20 1C ms1. After flash pyrolysis, the volatile pyrolyzates were directly swept onto a GC column to be separated (60-m, 0.25-mm internal diameter, crossbond, carbowax column; Restek: Stabilwaxs). The gas chromatograph (Fisons 8030) was operated in a splitless injection mode with a column head pressure of 25 psi. Oven temperature was held at 45 1C for 15 min, then ramped up to 240 1C at 2 1C min1, and finally, held at 240 1C for 10 min. The separated fragments were identified by MS (Fisons MD 800) that operated at 70 eV and scanned from 20 to 400 amu at 1 scan per second. Positive electron ionization (EI1) mode was used as an ion source. The source temperature was set at 200 1C and the GC/MS interface temperature was 250 1C. A GC/MS data acquisition software, XCALIBUR version 1.2, was used to collect the mass to charge (m/z) scan and produce a pyrochromatogram. The National Institute of Standard Technology Library was then used to identify the fragments in the pyrochromatogram.

Measurement of denitrification potential (DNP) The DNP of 28-day-old biofilms was determined using a modified version of the standard acetylene inhibition method for ecological research (Groffman et al., 1999; Sirivedhin 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c


& Gray, 2006; Arnon et al., 2007a, b). Intact biofilms were placed in 125-mL gas-tight jars (I-CHEM septa jar) and filled with 70 mL of nutrient solution to provide enriched conditions for denitrification. The nutrient solution contained 40 mg L1 NO3-N as KNO3, 100 mg L1 carbon as glucose, and 225 mg L1 chloramphenicol to inhibit microbial growth. The incubation started immediately after the jars were flushed with N2 for 3 min and acetylene was added (10% v/v) to inhibit the transformation of N2O to N2. During the incubation, the jars were agitated gently at 150 r.p.m. under dark conditions at room temperature (23 1 1C). Headspace samples were measured using a gas chromatograph (Hewlett Packard 5890) equipped with a 63Ni electron capture detector at an operating temperature of 320 1C. A stainless-steel Porapak Q (80/100 mesh) column was used to separate the gases at 60 1C with highpurity N2 as a carrier gas (18–20 mL min1). Denitrification rates were calculated from linear regression of N2O accumulation in the headspace, after the concentrations were corrected for solubility using the Bunsen coefficient (Venterink et al., 2003).

Data analyses Between-stream differences in AFDM, ln-transformed bacterial cell numbers, ln-transformed nirS copy numbers, algal cell densities, algal biovolumes, and changes in these variables over the 28 days of biofilm development were assessed using two-factor (age stream) ANOVA. Algal species richness and Shannon diversity (H 0 ; Zar, 1974) were determined for all samples and assessed for differences between systems and with age using two-factor ANOVA, followed by Holm– Sidak multiple comparisons (SIGMASTAT 3.5, Statistical Software, Point Richmond, CA). Variation in both T-RFLP profiles and microalgal community structure between streams and among sampling dates was examined using nonmetric multidimensional scaling (NMDS) analysis via the PRIMER software package (Primer-E Ltd, Plymouth, UK). Ordinations were based on Bray–Curtis similarity matrices calculated from the presence/absence of all unique nosZ terminal restriction fragments (TRFs) encountered for denitrifiers and, for algal communities, from the relative biovolumes of algal taxa comprising at least 5% of the total algal biovolume from at least one system on one sampling date. NMDS [using the vegan package in R (R Development Core Team, 2008)] was also used to assess between-system differences in the organic signatures of stream water and 28day-old biofilms, which were comprised of the 50 major (by area) pyrolysis fragments, creating a list of 99 chemicals in each 28-day-old biofilm pyrochromatogram and 75 chemicals in each stream-water pyrochromatogram. In addition, we categorized chemical pyrolysis fragments identified from pyrochromatograms into four broad chemical groups FEMS Microbiol Ecol 77 (2011) 477–492

481


[aliphatic, aromatic, nitrogen (N)-containing aliphatic and N-containing aromatic], and determined the percent contribution of these groups to signatures from different streams and different sources. We compared these data to examine the possible links between the organic content of biofilms and the content (nirS copy# cm2, genotype number) and function (DNP/nirS density) and organic content via linear regression. Rates of change in taxonomic structure (i.e. ‘succession rates’) for microalgal and denitrifier communities were calculated, as per Stevenson et al. (1991), via linear regression of Bray–Curtis percent dissimilarities of all pair-wise comparisons among samples collected vs. age difference between the samples compared. For example, communities collected from replicate tiles on the same sampling date have an age-difference value of zero, a value of seven for comparisons between 7- and 14-, 14- and 21-, and 21- and 28-day-old communities, a value of four of comparisons between 3- and 7-day-old communities, and so forth. Regression coefficients (r2) and comparison of slopes using Student’s t-test (Zar, 1974) were used to compare the succession rates between streams. Regressions of algal percent taxonomic dissimilarity against dissimilarity values for denitrifier consortia for all pair-wise sample comparisons for which both algal and denitrifier data were available were used to assess the degree to which successional changes in algal communities were coupled to changes in denitrifiers in the two study systems.

Results Study site characterization The sites, while similar in current velocity and exposure to Photosynthetic Active Radiation (PAR), differed in discharge, depth, the amount of DOC, and, most notably, concentrations of dissolved inorganic nutrients, which were an order of magnitude higher in DER than RP (Table 1). The organic quality of stream water samples collected at the beginning of the study (June 23, 2008) was characterized using Py-GC/MS. The general features of the pyrolysis fingerprints of the two water samples were similar; both predominantly contained aromatic compounds bearing a strong phenolic signature (Fig. 1). The samples were

distinct, however, in their secondary characteristics. Stream water from RP had a greater presence of pyrolysis products of polycarboxylic acid, 2-cyclopenten-1-one (likely derived from soil polycarboxylic acids), and bore a stronger aliphatic signature (33% of peak area) than the stream water at DER (21% of peak area) (Wilson et al., 1983). Similar to previous research on effluent-derived organic matter (EfOM), the fragments from the DER site had a marked nitrogencontaining signature (12%) associated with proteins and amino sugars (Peschel & Wildt, 1988; Sirivedhin & Gray, 2005). Several markers characteristic of wastewater inputs (Sirivedhin & Gray, 2005) were also present in the DER sample, including benzaldehyde, benzonitrile, and furancarboxaldehyde, indicating the persistent influence of EfOM on the organic quality in this stream water. The NDMS ordination (Supporting Information, Fig. S1) of the pyrolysis fragments comprising the organic signatures of stream water from our two study systems showed excellent within-site reproducibility in the analyses and a high degree of separation between sites (P o 0.00001), illustrating clear differences in the organic quality of these samples.

Development and characterization of biofilms Biomass increased in both systems over 28 days of biofilm development (P o 0.001), but was significantly higher in RP than DER (P = 0.017) (Fig. 2a). Bacterial cell densities also increased over time (P = 0.001), but did not differ between streams (P = 0.226) (Fig. 2b). Algal cell densities increased significantly over time (P = 0.042), attaining significantly higher levels late in colonization in DER (system effects: P o 0.001) (Fig. 3). The temporal pattern of algal cell density change differed between systems (age system; P o 0.001), however, as algal densities increased throughout the 28-day period of biofilm development in DER, but stabilized in RP after day 21 (Fig. 3). Patterns of change in algal biovolume also differed between systems (age system; P = 0.005), but biovolume decreased from high values in both systems within the first week of colonization to generally stabilize at lower values thereafter (Fig. 3). These biovolume reductions, concurrent with increases in algal cell density, were generated by shifts in the taxonomic composition, with an early presence of comparatively large-celled diatom taxa [e.g. Rhoicosphenia curvata

Table 1. Range of values for physicochemical attributes measured at study sites over the 28-day experiment (June 23–July 21, 2008)

Site

DOC (mg L1)

PO4-P (mg L1)

NH4-N (mg L1)

DER RP

8.0 6.8

689–1197 2.5–3.8

119.6–130.3 6.3–54.2

NO3-N (mg L1) 10.5–15.9 0.003–0.2

N:P

Depth at block (cm)

Discharge (m3 s1)

Water temperature ( 1C)

Current velocity (cm s1)

19.7–51.6 31.8–137.7

14.8–39.3 0.2–12.5

0.053–0.591 0.031–0.169

20–23 26–28

5.3–59.1 3.1–16.9

Water-column DOC measurements were taken only on June 23.

FEMS Microbiol Ecol 77 (2011) 477–492

2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

482


% Full scale

100 80 60 40 20 20

40

60

80

20 40 100 Retention time (min)

(Kütz.) Grun. in RP and Gomphonema parvulum Kütz. in both systems] diminishing as biofilm development progressed (Fig. 3). Algal species richness (Fig. 4a) and diversity exhibited similar patterns of change, with the highest values in both systems occurring within the first week of colonization, with subsequent curvilinear patterns of decline thereafter. Both algal richness and diversity were significantly higher in RP than in DER [system effects: P = 0.01 (richness), P = 0.029 (diversity)], but these differences were not manifested until day 7 for diversity (Holm–Sidak multiple comparison, P = 0.005) and day 21 for species richness [P = 0.023 (21 days), P = 0.007 (28 days)]. Organic signatures of 28-day-old biofilms from our two study streams differed slightly in the relative proportions of organic classes into which the pyrolysis products were categorized (Fig. 5). Organics within periphyton from RP were characterized by primarily aliphatic (39%) compounds, with simple aromatic pyrolysis fragments comprising about 32%. By comparison, the contributions of simple aliphatic and aromatic compounds were 30% and 34%, respectively, in 28-day-old DER periphyton (Fig. 5). As expected, organic signatures of biofilm samples from both systems exhibited large peaks for pyrolysis products of typical biopolymers such as proteins, including, pyrroles, phenols, and indoles. Both periphyton samples showed very strong phenol and cresol signatures, which is, in part, associated with amino acids such as tyrosine. While the basic organic makeup of biofilms from the two sites was generally similar, they differed in the relative contributions among the assortment of pyrolysis fragments. Differences in the organic signatures between the sites can be attributed, in part, to the greater presence of pyrolysis products such as 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

60

80

100

Fig. 1. Characterization of the organic content of stream water from RP (left panels) and DER (right panels). Upper panels illustrate the relative contributions of broad categories of organic molecules to each carbon pool (ALI, aliphatic; ARO, aromatic; NAL, nitrogen-substituted aliphatic; NAR, nitrogen-substituted aromatic). Lower panels illustrate the Py-GC/MS profiles of a more specific organic content in stream water from RP (left) and DER (right). Chemical species associated with numbered peaks are as follows: 1, 2-cyclopenten-1-one,2-methyl-; 2, acetic acid; 4, 2-cyclopenten-1-one,2,3-dimethyl-; 5, propanoic acid; 6, benzonitrile; 7, 1H-indene,1, 1-dimethyl-; 8, naphthalene; 9, acetamide; 10, naphthalene,1-methyl-; 11, naphthalene, 2-methyl-; 13, phenol; 14, P-cresol.

polycarboxylic acid, 2-cyclopenten-1-one (possibly derived from soil polycarboxylic acids and secondary polysaccharides) in biofilm samples from the RP site. DER biofilm samples contained a greater proportion of pyrolysis fragments containing nitrogen and propanoic acid compounds, possibly illustrating the influence of N-containing and aliphatic compounds characteristic of stream water downstream of wastewater treatment plants (WWTPs). As observed in the chemical profiles of stream water, anthropogenic markers such as benzonitrile were also present in the organic signatures of biofilm samples from both sites, but were present at a greater proportion in samples from biofilm developed at DER.

Microalgal succession Ordination of algal community structure based on relative biovolume of common taxa showed clear separation between assemblages in DER vs. RP (Fig. 6a). Most notably, the diatom Amphora pediculus (Kütz.) Grun. ex A. Schmidt, a species virtually absent in RP, comprised between 19.3–65.9% of algal biovolume and 11.9–73.5% of all cells in DER samples, and Achnanthidium minutissimum (Kütz.) Czarnecki, which comprised 0.03–31.0% of biovolume and 12.7–60.9% of cell densities in RP, but was encountered in DER in low numbers only in 3-day-old communities (Fig. 3). Algal community composition changed throughout the 28 days of biofilm development in RP due, primarily, to an increase in the relative biovolume of filamentous green algae and decreases in the relative biovolume of a suite of Gomphonema species and R. curvata. In DER, algal succession was less directional and community structure appeared FEMS Microbiol Ecol 77 (2011) 477–492

483


1.4

(a)

AFDM (mg cm–2)

1.2 1.0 0.8 0.6 0.4 0.2 0.0

Ln bacterial density (cells cm–2)

28.0

(b)

27.5 27.0 26.5 DER RP

26.0 25.5

Ln (nir S copy number cm–2)

18

(c)

16 14

consistently different between the two streams (Fig. 6b). One hundred and forty-seven distinct TRFs, representing distinct nosZ genotypes, were detected collectively from our two study systems. Of these, 56 were unique to DER and 36 were encountered only in RP. There was no significant difference in the total numbers of nosZ genotypes between DER and RP (ANOVA, system effects: P = 0.716), and only marginally significant (age effects: P = 0.058) changes in the total numbers were observed over time. However, systemspecific second-order linear regressions of numbers of nosZ genotypes vs. community age showed a highly significant (P o 0.001) increase with time (R2 = 0.782) for RP data and an insignificant relationship in DER (R2 = 0.121) (Fig. 4b). Examination of the appearance of new nosZ genotypes that had not been detected in samples from previous sampling dates showed that, in DER, the number of newly observed nosZ genotypes peaked at 67 in 14-day-old biofilms and declined to nearly half that number by day 28. In contrast, the appearance of unique nosZ genotypes in RP increased progressively with biofilm age, so that 37 of the 44 nosZ genotypes encountered in 28-day-old biofilms in that system were established after 3 weeks of biofilm development. The patterns observed in NMDS ordinations of T-RFLP data were similar to those observed for microalgae, but with a higher variance (Fig. 6b). Consortia of denitrifiers differed taxonomically between systems and, unlike algal communities (Fig. 6a), changed with community age in the same direction along NMDS axis 2, indicating the influence of the changing identity of denitrifiers common to both systems.

12

Algal and denitrifying-bacterial succession rates

10 8 6 0

7

14 21 Community age (days)

28

Fig. 2. Changes in biomass, as AFDM (a), bacterial cell denisity (b), and nirS copy numbers (c), over 28 days of biofilm development in two study streams.

to stabilize by day 14 to a community where A. pediculus and filamentous green algae comprised a large portion of algal biovolume (Figs 3 and 6a).

Succession of denitrifying bacteria The numbers of denitrifying bacteria within biofilms (as indicated by nirS copy numbers) increased over time (P = 0.001), but did not differ between streams (P = 0.914) (Fig. 2c). However, the composition of denitrifier communities (indicated by T-RFLP analysis of nosZ genes) was FEMS Microbiol Ecol 77 (2011) 477–492

The rates of algal succession, measured as Bray–Curtis dissimilarity regressed against the difference in community age in all pair-wise sample combinations, were more rapid in RP (Fig. 7a) than DER (Fig. 7b), reflecting continued change in RP algal community structure throughout the 28 days of biofilm development. Comparison of communities collected from the same sampling block over the five collection dates indicated that these communities were neither more nor less similar to each other than communities developed on other sampling blocks. Consortia of bacterial denitrifiers exhibited a similar, but less robust pattern of dissimilarity with age difference than algae (Fig. 7c and d), perhaps due to fewer data points available for the analysis of TRF dissimilarities. As with algal communities, denitrifier assemblages became more dissimilar in taxonomic structure with age difference between samples in RP (Fig. 7c), but, in DER, this relationship was considerably weaker and more variable (Fig. 7d). Analyses regressing algal dissimilarity values against denitrifier dissimilarities from the same sample pairs (Fig. 8) showed that the succession rates of algal assemblages were significantly coupled to those 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

484


1.6e+7

(a)

DER

1.2e+7 1.0e+7

1000

8.0e+6 6.0e+6

500

4.0e+6

Algal cell density (cells cm–2)

Algal biovolume (µm3 cm–2)

1.4e+7 1500

2.0e+6 0

0.0 1.6e+7

(b)

RP

1.2e+7 1.0e+7 1000

8.0e+6 6.0e+6

500

4.0e+6

Algal cell density (cells cm–2)

Algal biovolume (µm3 cm–2)

1.4e+7 1500

2.0e+6 0 0

5

10

15 20 Community age Achnanthidium lanceolatum Achnanthidium minutissimum Amphora pediculus Green algal filaments Green algal basal cells Cocconeis placentula v. euglypta Gomphonema spp. Gomphonema parvulum Navicula cryptocephala Rhoicosphenia curvata v. abbreviatum Others

25

of denitrifier consortia (P o 0.001 that r2 4 0) in RP, but not in DER (P = 0.081). Comparison of the slopes of these regression lines confirmed that algal communities and consortia of denitrifying bacteria changed in taxonomic structure at the same rate in RP, but were less strongly coupled in DER (t = 4.99, d.f. = 56; P o 0.001).

Biofilm DNP DNP measured in 28-day-old biofilms did not differ significantly between systems [DER = 27.1 10.4 (s.e.) mg N2O h1 cm2; RP = 22.2 10.4 mg N2O h1 cm2], but there was a considerable variation in DNP among replicate 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

0.0 30

Fig. 3. Changes in algal biovolume (bars) and algal cell densities (open circles) during 28 days of biofilm development in streams within a DER (a) and downstream of WWTP effluent release (RP) (b). Areas with a distinct color/pattern within each biovolume bar represent the contributions of common algal taxa within biofilms to the total algal biovolume.

samples from each system. A significant positive relationship was detected between biofilm biomass and DNP normalized by nirS copy number among the three replicate 28-day-biofilm samples in both systems (DER, r2 = 0.986; RP, r2 = 0.892) (Fig. 9a). nirS-specific DNP rates within biofilms decreased with increases (r2 = 0.464) in the proportion of pyrolysis fragments derived from aromatic compounds within biofilm organic signatures (Fig. 9b). When the one replicate biofilm sample from RP that contained the highest percentage of aromatic organic compounds and the lowest measured DNP was excluded from the analysis, a strong (r2 = 0.899) positive linear relationship was observed between natural-log-transformed DNP and nirS-genotype FEMS Microbiol Ecol 77 (2011) 477–492

485


40

differences among algal species in the chemical form of their organic exudates (Hamels et al., 2004; Bahulikar & Kroth, 2008), we propose that variation in algal community structure controls, in part, the quality of dissolved organics available for bacterial uptake.

(a)

Algal species richness

35 30 25 20

Biofilm development

15 10 5 0 60

(b)

nosZ genotypes (#)

50 40 30 20 10 0 0

7

14 21 Community age (days)

28

Fig. 4. Changes in the numbers of algal species (a) and distinct nosZ genotypes (as indicated by distinct TRFs) (b) within samples collected over 28 days of biofilm development in RP (open symbols) and DER (closed symbols). Lines result from second-order curvilinear regressions of number with biofilm development time [RP (solid lines) and DER (broken lines)].

richness (Fig. S2). The regression coefficient resulting from the inclusion of all six samples in the analysis indicated no relationship (r2 = 0.044).

Discussion The results of this study provide insight into the factors affecting the diversity and patterns of succession in a bacterial guild responsible for the biogeochemical cycling of nitrate. Our study streams differed substantially in the quantity of inorganic and organic nutrients, as well as the quality of dissolved organic compounds in the water column potentially available to residents within developing epilithic biofilms. These differences were likely due to the input of effluent from a WWTP several kilometers upstream of the DER site. Biofilms in these streams differed in the taxonomic content and rates of successional change of microalgal assemblages, and ultimately differed in the numerically dominant algal. Based on between-system differences in the organic signatures of 28-day-old biofilms, and documented FEMS Microbiol Ecol 77 (2011) 477–492

Biomass accrual on artificial substrata in our study was significantly, but not substantially, greater at the RP site than that receiving effluent from a WWTP (DER), despite much higher concentrations of dissolved inorganic nutrients at the latter and differences between streams in the amount and quality of DOC in stream water. Algal cell densities were ultimately higher in DER, however, and algal biovolume did not differ between sites, suggesting greater retention of nonalgal particulate organic matter in RP. Lower algal biovolume in assemblages collected from both systems in the second half of the 4-week colonization period reflects a shift in dominance from larger diatom species in the early stages of community development to small-celled taxa. This inverse relationship between algal biovolume and cell size within benthic biofilms is consistent with the findings of Passy (2008) from analysis of US Geological Survey data from streams throughout the continental United States. The increase in the collective cell surface area resulting from a shift in dominance from large- to small-celled taxa may, potentially, influence both the volume of algal exudates produced and the degree of bacterial access to them, although this remains to be tested. Algal taxonomic structure and species richness differed substantially between streams in both the content and the pattern of change. Taxa that comprised a large portion of algal biovolume or cell densities in one system were absent, or nearly so, in the other. Species richness declined precipitously over the course of biofilm development in DER, but not in RP. Finally, changes in the taxonomic structure of biofilms in RP were relatively uniform, with little variation among replicate samples in RP, but, in DER, the composition of algal assemblages was much more variable within sample dates and, unlike RP, stabilized from 14 through 28 days. In systems with an extremely high nutrient load, like DER, a relatively rapid decline in algal species richness and diversity with consequent stabilization of community structure is expected, as taxa best able to compete under nutrient-replete conditions, here the small-celled diatom species A. pediculus, proliferate more rapidly than other taxa (Stevenson et al., 1991; Peterson & Grimm, 1992).

Establishment of algal/bacterial interactions In both systems, heterogeneity of the physical/chemical conditions within biofilms likely increased as communities developed, with increased light attenuation (Johnson et al., 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

486


13

100

% Full scale

14

3

80

14 13

3

15

60 40

15

4

2

1 2

20

6 20

40

8 9

60

1012

1 80

20 100 Retention time (min)

40

9 5 4 6

8 60

12 10 80

100

Fig. 5. Characterization of the organic content of 28-day-old biofilms from RP (left panels) and DER (right panels). Upper panels illustrate the relative contributions of broad categories of organic molecules to each carbon pool (ALI, aliphatic; ARO, aromatic; NAL, nitrogen-substituted aliphatic; NAR, nitrogen-substituted aromatic; HAL, halogen-substituted aliphatic). Lower panels illustrate the Py-GC/MS profiles of a more specific organic content in stream water from RP (left) and DER (right). Chemical species associated with numbered peaks are as follows: 1, 2-cyclopenten-1-one,2-methyl-; 2, acetic acid; 3, pyrrole; 4, 2-cyclopenten-1-one,2,3-dimethyl-; 5, propanoic acid; 6, benzonitrile; 8, naphthalene; 9, acetamide; 10, naphthalene,1methyl-; 12, benzyl nitrile; 13, phenol; 14, P-cresol; 15, indole.

1997), reduced rates of diffusion of inorganic nutrients (Arnon et al., 2007a, b; Renslow et al., 2010), and increases in spatial/temporal heterogeneity in the distribution of transient microzones that vary in oxygen concentration, pH, and/or exoenzyme activity (Lock et al., 1984). Such changes likely increase the diversity of bacterial niches (Jackson, 2003). Pohlon et al. (2010) reported that the bacterial production of exoenzymes that allow cells to exploit dissolved organic molecules can occur within 4 h of attachment. Establishment of algal colonists further increases both spatial and chemical resource availability. Grossart et al. (2007) reported that bacteria colonized agar spheres embedded with diatoms much more rapidly than spheres without algae, and that bacteria clustered around diatom cells, suggesting that algal/bacterial associations within developing biofilms form quickly. Bacterial protease activity in that study was 10–20 times higher in attached bacteria compared with free-living cells, and individual 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

bacterial strains could increase enzyme production within 2 h of exposure to organic substrates. Benthic-algal exudates provide a direct autogenic carbon source for bacterial metabolism (Haack & McFeters, 1982); the uptake and incorporation of radiolabeled algal polysaccharides into bacterial fatty acids has been detected within 4 h of release (Bellinger et al., 2009). Successional changes in the taxonomic content of algal assemblages increase spatial heterogeneity within biofilms, likely increasing further the diversity of bacterial niches. Because algae typically reproduce asexually, individual cells ultimately give rise to colonies, resulting in aggregates of different algal species throughout the biofilm (Hoagland et al., 1982; Lock et al., 1984), generating a patchwork of potential ‘hot spots’ of concentrated organic exudates of different quality. As such, biofilms that support higher algal species diversity likely provide a more diverse array of organic substrates. Because bacterial species vary in their FEMS Microbiol Ecol 77 (2011) 477–492

487


(a)

Stress = 0.07

3

21 21 14 28 28

3

7

28

21

7 14

7

7

14

7

3

3 3 7 14 14 21 21 21

28

28 28

(b) Stress = 0.16

DER RP

7

7

14 14

14 14

14

14

28

21 21 28

28 28 28

28 21

Fig. 6. MDS ordination of changes in microalgal community structure (a) and bacterial-denitrifier community structure (b) over 28 days of biofilm development in RP (open symbols) and DER (closed symbols). Numbers adjacent to each data point identify the age of the biofilm sampled.

ability to exploit various carbon substrates, this should translate into a higher diversity of bacterial residents. Haynes et al. (2007) demonstrated that bacterial assemblages in treatments amended with algae-derived extracellular polymeric substances increased in density and diverged in taxonomic structure from assemblages not so amended, indicating differential performance among taxa in enzyme specificity and production. Thus, algal species diversity may, to some degree, dictate the diversity of bacteria that rely on their exudates as a carbon source. Our results lend support to this hypothesis, and fit predictions put forth by Jackson (2003) relating developmental changes in microenvironmental conditions within biofilms to increases in niche diversity and consequent increases in bacterial species richness and the appearance of new bacterial species as biofilms develop. Biofilms that developed in our study systems exhibited little or no difference in biomass, algal biovolume, total bacterial densities, or numbers of denitrifiers, and yet, the FEMS Microbiol Ecol 77 (2011) 477–492

rate and pattern of change in both algal and denitrifier community structure differed substantially between systems. Analysis of algal succession rates, and the results of MDS ordinations, indicated robust patterns of directional change in species composition as biofilms developed in RP. Concurrently, the number of newly encountered denitrifier genotypes increased steadily; no such increase was observed in DER, where algal community structure and, thus, algalgenerated niche space, stabilized within 2 weeks. These differences support the contention that the degree to which internal or external factors influence community dynamics and biogeochemical processes within biofilms is context dependent. Differences between systems in the relationship between biofilm biomass and DNP illustrate this point. In both systems, our DNP assay of 28-day-old biofilms showed that individual bacterial bearing the nirS gene were much more actively denitrifying in biofilms of higher biomass than lower. While this relationship was consistent between systems, the observed displacement in manifestation toward higher biomass in RP suggests that the biomass threshold necessary to facilitate denitrification in relatively nutrientpoor systems exceeds that required in systems that are nutrient replete.

Influence of nutrients and dissolved organics on algal/denitrifier coupling Coupling of changes in the taxonomic content of consortia of bacterial denitrifiers to those of benthic microalgae differed in strength between study streams, with apparent strong coupling in biofilms developed in the RP stream, but little evidence of strong algal/bacterial linkage below the WWTP. Nutrient and organic concentrations in stream water were much lower in RP and the external supply of DOC bore a stronger terrestrial chemical signature, which is generally less labile, than the DER site. Despite the more recalcitrant allochthonous carbon supply, the organic signature of periphyton in RP revealed a much stronger aliphatic, more labile, nature than that developed at the DER site. This suggests that denitrifiers in RP biofilms likely relied most heavily on organic carbon produced autogenically. These results are consistent with those from other studies illustrating that the reliance of bacteria on algalderived organic carbon increases as the allogenic nutrient supplies decrease (e.g. Murray et al., 1986; Cook et al., 2007; Scott et al., 2008; Ziegler et al., 2009), and support the conclusions of Mulholland et al. (1991) that, in the absence of excessive allogenic nutrient supply, the reliance on internal nutrient cycling increases. The total bacterial cell densities did not differ between systems, nor did the total numbers of denitrifiers as indicated by nirS gene copy numbers. In this study, nirK genes could not be amplified from DNA isolated from biofilms, 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

488


(a)

(b)

80 60 40 20 R 2 = 0.856

0 % Dissimilarity in denitrifier / consortia

100

R 2 = 0.175 (d)

(c)

80 60 40 20 R 2 = 0.650

0 0

5

10

15

20 25 0 5 Age difference (days)

despite the fact that the same primer set has been used successfully by other researchers to amplify nirK genes from environmental samples (Prieme´ et al., 2002; Wolsing & Prieme´ , 2004). The failure to amplify nirK genes suggests that nirK-containing organisms were either not present in the biofilms or were present at a level below the limit of detection of the PCR assay, indicating that they were not the dominant denitrifiers in these ecosystems. This result fits with the results of several groups that have suggested that nirK-containing denitrifiers are less abundant than nirScontaining denitrifiers in aquatic ecosystems (Braker et al., 2000; Nogales et al., 2002; Angeloni et al., 2006). The identity and number of distinct denitrifier genotypes (indicated by T-RFLP analysis of nosZ genes) differed between systems. DER supported a greater number of system-specific denitrifier genotypes than RP, as might be expected, given that effluent from WWTP contains a high diversity of dissolved organic compounds, many of which are generally not common in systems under natural controls (Heim et al., 2003; Sirivedhin & Gray, 2005; Parnaudeau & Dignac, 2007). Similar to the patterns noted in algal assemblages, however, the number of denitrifier genotypes associated with biofilm samples appeared to stabilize in older DER biofilms, while distinct denitrifier genotypes detected in RP samples increased in number throughout the 28 days of biofilm development. Although the characterization of DOC in the water column of our study streams was based on samples collected on a single date, it is reasonable to suggest that the overall quality of DOC within WWTP effluent was more uniform than in RP, which likely varied over time, with the relative contribution of aquatic and terrestrial sources changing with runoff from precipitation, variation 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

R 2 = 0.295 10

15

20

25

Fig. 7. Regressions of taxonomic dissimilarity in community structure between communities of different ages (i.e. succession rate) for algal communities in RP (a) and DER (b), and for denitrifier communities in RP (c) and DER (d). Values of zero represent the taxonomic dissimilarities between replicate samples collected on the same date. Closed symbols on the algal panels represent the dissimilarity values calculated from communities collected from the same sampling unit on different dates; open symbols represent the taxonomic differences between communities collected from different sampling blocks.

100 Denitrifier dissimilarity

% Algal dissimilarity

100

80 60 40 20

DER RP

R = 0.145 R = 0.353

0 0

20

40 60 Algal dissimilarity

80

100

Fig. 8. Regressions of taxonomic dissimilarity in algal assemblages between pairs of sampled biofilm against the dissimilarity of denitrifier assemblages in the same biofilms in RP (open symbols) and DER (closed symbols).

in discharge, and season (Biber et al., 1996; Sirivedhin & Gray, 2005). In addition, the EfOM from the WWTP may have been more readily assimilable by microorganisms in DER biofilms than the allogenic carbon sources available at the RP site, which bore a strong terrestrial, and hence more refractory, organic signature. The linear increase in new denitrifier taxa suggests increasing diversity of unique carbon sources, likely provided by changes in algal taxonomic composition in RP. Although limited in extent, our results suggest a link between the chemical quality of carbon sources within biofilms and denitrification activity. The negative relationship we observed between the percentage of the carbon pool comprised of aromatic molecules and reduced per cell (i.e. nirS copy number) denitrification FEMS Microbiol Ecol 77 (2011) 477–492

489

(µg N2O h–1 /nir S copy # cm–2)


0.014

(a)

(b)

0.012 0.010 0.008 0.006 0.004 0.002 0.000 0.6

0.8

1.0

1.2

1.4

AFDM (mg cm–2)

1.6

0.25

0.30

0.35

0.40

0.45

0.50

% Aromatic/100

Fig. 9. Denitrification (as DNP) rates per nirS copy in 28-day-old biofilms differing in (a) biomass and (b) the percentage of within biofilm organics comprised of aromatic compounds in samples from RP (open symbols) and DER (closed symbols).

activity is consistent with the findings of Sirivedhin & Gray (2006), who reported that denitrification rates of bacteria from wetland sediments decreased in tandem with a reduction in the aliphatic : aromatic ratio of the carbon source available for assimilation. The strength of this relationship may vary with algal species identity, with some taxa providing exudates that are more readily assimilated by a wider variety of denitrifiers than others. For example, Ishida et al. (2008) documented that the magnitude of DNP in biofilms developed on mesh substrata in a group of wetlands correlated positively with the relative biovolume of some diatom taxa, but not others. Thus, bacterial performance may depend on which combinations of coexisting algal taxa are present. Our data do not allow the assessment of the relative importance of individual algal taxa to bacterial-denitrifier efficiency, but this is an avenue of research that should be investigated further. Furthermore, while our observations confirmed a distinct anthropogenic influence on the quality of carbon available to microorganisms at DER, we cannot determine whether this exerts a stress on the system that may select for characteristics within biofilms that decouple algal and bacterial communities.

Implications for ecosystem functioning Taxonomic diversity within bacterial guilds provides functional redundancy, which may serve as a buffer against species loss and consequent loss of their functional contribution, and may increase ability to adapt to changing environmental conditions. Species that perform the same ecosystem function, however, are likely not perfectly ‘redundant’. That is, they may provide the same service, but may vary in efficiency depending on the proximate conditions, thus providing an essential service across a much broader range of environmental conditions. With recent advances in molecular technology, the influence of FEMS Microbiol Ecol 77 (2011) 477–492

taxonomic content of consortia of bacterial denitrifiers on biogeochemical performance is becoming more evident (Wolsing & Priemé, 2004). Tiedje (1988), for example, demonstrated that species of denitrifying bacteria differ in the oxygen-concentration thresholds at which they switch to nitrate as a terminal electron acceptor. Cavigelli & Robertson (2000) observed a similar variation in oxygen-concentration thresholds in consortia of bacterial denitrifiers in soils with different disturbance histories, but also differences in the sensitivity of denitrification to variation in soil pH. In biofilms, such differences are particularly important, because microscale heterogeneity in resource availability can, potentially, provide suitable niche space for denitrifying bacterial phylotypes that differ in efficiency and exploit otherwise unexploited resources. Thus, denitrification efficiency within biofilms that support a higher diversity of denitrifiers may exceed that of biofilms supporting less diverse assemblages, over a range of environmental conditions. Our results suggesting that denitrification activity may increase as a function of denitrifier ‘species’ richness, while based on a limited data set, are intriguing and illustrate the potential importance of a small-scale variation in regulating ecosystem processes. The amount and chemical nature of DOC in stream water, as well as inorganic nutrient content, is heavily altered by anthropogenic influence (Williams et al., 2005; Petrone et al., 2009) and is a major determinant of taxonomic structure in epilithic bacterial communities (Kobayashi et al., 2009; Lear & Lewis, 2009). Kirchman et al. (2004) illustrated that the most influential determinant of taxonomic structure and exoenzyme activity of bacterioplankton at sites along the Hudson River and its tributaries was the content and concentration of DOC in the source of water taken from those sites. The disruption of algal/bacterial interactions via human-induced alteration of chemical attributes of surface waters will inevitably be manifested at much broader scales, through the modification of 2011 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

490

biogeochemical cycles. A better understanding of the underlying mechanisms through which microbial interactions within biofilms are established, are maintained, and translate up to broader scales of ecosystem function will aid in the development of plans to mitigate anthropogenic degradation of aquatic systems.

Acknowledgements We thank John Oldenburg for allowing access to study sites within the DuPage County Forest Preserve system. Comments by two anonymous reviewers improved the manuscript and were much appreciated. Sam Saliba, Robert Bednarzck, Sarah Akinde, and Drew Harvey assisted in the laboratory and field. This research was funded by NSF Ecosystem Grant DEB 0640717 to C.G.P. and J.J.K., NSF Ecosystem Grant DEB 0640459 to K.A.G., and an REU Supplement to NSF Ecosystem Grant DEB 0640717. M.J.S. was also supported by an Undergraduate Research Award from the Department of Biology at Loyola University Chicago and by fellowships provided by the Loyola Undergraduate Research Opportunities Program.

References Angeloni NL, Jankowski KJ, Tuchman NC & Kelly JJ (2006) Effects of an invasive cattail species (Typha glauca) on sediment nitrogen and microbial community composition in a freshwater wetland. FEMS Microbiol Lett 263: 86–92. A.P.H.A. (1998) Standard Methods for the Examination of Water and Wastewater, 20th edn. American Public Health Association, Washington, DC. Arnon S, Gray KA & Packman AI (2007a) Biophysicochemical process coupling controls nitrate use in biofilms. Limnol Oceanogr 52: 1665–1671. Arnon S, Packman AI, Peterson CG & Gray KA (2007b) Effects of overlying velocity on periphyton structure and denitrification. J Geophys Res – Biogeosciences 112, G01002. Bahulikar RA & Kroth PG (2008) The complex extracellular polysaccharides of mainly chain-forming freshwater diatom species from epilithic biofilms. J Phycol 44: 1465–1475. Baines SB & Pace ML (1991) The production of dissolved organic matter by phytoplankton and its importance to bacteria – patterns across marine and freshwater systems. Limnol Oceanogr 36: 1078–1090. Baniulyte D, Favila E & Kelly JJ (2009) Shifts in microbial community composition following surface application of dredged river sediments. Microb Ecol 57: 160–169. Bellinger BJ, Underwood GJC, Ziegler SE & Gretz MR (2009) Significance of diatom-derived polymers in carbon flow dynamics within estuarine biofilms determined through isotopic enrichment. Aquat Microb Ecol 55: 169–187.


c


Biber MV, Gulacar FO & Buffle J (1996) Seasonal variations in principal groups of organic matter in a eutrophic lake using pyrolysis/GC/MS. Environ Sci Technol 30: 3501–3507. Braker G, Fesefeldt A & Witzel KP (1998) Development of PCR primer systems for amplification of nitrite reductase genes (nirK and nirS) to detect denitrifying bacteria in environmental samples. Appl Environ Microb 64: 3769–3775. Braker G, Zhou J, Wu L, Devol AH & Tiedje JM (2000) Nitrite reductase genes (nirK and nirS) as functional markers to investigate diversity of denitrifying bacteria in Pacific Northwest marine sediment communities. Appl Environ Microb 66: 2096–2104. Bruckner CG, Bahulikar R, Rahalkar M, Schink B & Kroth PG (2008) Bacteria associated with benthic diatoms from Lake Constance: phylogeny and influences on diatom growth and secretion of extracellular polysaccharides. Appl Environ Microb 74: 7740–7749. Cavigelli MA & Robertson GP (2000) The functional significance of denitrifier community composition in a terrestrial ecosystem. Ecology 81: 1402–1414. Cole JJ (1982) Interactions between bacteria and algae in aquatic ecosystems. Annu Rev Ecol Syst 13: 291–314. Cook PI, Veuger B, Boer S & Middelburg JJ (2007) Effect of nutrient availability on carbon and nitrogen incorporation and flows through benthic algae and bacteria in near-shore sandy sediment. Aquat Microb Ecol 49: 165–180. Crumpton WG, Isenhart TM & Mitchell PD (1992) Nitrate and organic N analyses with 2nd-derivative spectroscopy. Limnol Oceanogr 37: 907–913. Groffman PM, Holland EA, Myrold DD, Robertson GP & Zou X (1999) Denitrification. Standard Soil Methods for Long-Term Ecological Research, (Roberston GP, Coleman DC, Bledsoe CS & Sollins P, eds), pp. 272–288. Oxford University Press, Oxford. Grossart HP, Levold F, Allgaier M, Simon M & Brinkhoff T (2005) Marine diatom species harbour distinct bacterial communities. Environ Microbiol 7: 860–873. Grossart HP, Tang KW, Kiorboe T & Ploug H (2007) Comparison of cell-specific activity between free-living and attached bacteria using isolates and natural assemblages. FEMS Microbiol Ecol 266: 194–200. Gruntzig V, Nold SC, Zhou JZ & Tiedje JM (2001) Pseudomonas stutzeri nitrite reductase gene abundance in environmental samples measured by real-time PCR. Appl Environ Microb 67: 760–768. Haack TK & McFeters GA (1982) Nutritional relationships among microorganisms in an epilithic biofilm community. Microb Ecol 8: 115–126. Hamels I, Mussche H, Sabbe K, Muylaert K & Vyverman W (2004) Evidence for constant and highly specific active food selection by benthic ciliates in mixed diatom assemblages. Limnol Oceanogr 49: 58–68. Haynes K, Hofmann TA, Smith CJ, Ball AS, Underwood GJC & Osborn AM (2007) Diatom-derived carbohydrates as factors


491


affecting bacterial community composition in estuarine sediments. Appl Environ Microb 73: 6112–6124. Heim S, Schwaubauer J, Kronimus A, Littke R, Woda C & Mangini A (2003) Geochronology of anthropogenic pollutants in riparian wetland sediments of the Lippe River (Germany). 21st International Meeting on Organic Geochemistry, pp. 1409–1425. Pergamon-Elsevier Science Ltd, Kidlington, UK. Hillebrand H, Dürselen C-D, Kirschtel D, Pollingher U & Zohary T (1999) Biovolume calculation for pelagic and benthic algae. J Phycol 35: 403–424. Hoagland KD, Roemer SC & Rosowski JR (1982) Colonization and community structure of two periphyton assemblages, with emphasis on the diatoms (Bacillariophyceae). Am J Bot 69: 188–213. Irwin WJ (1982) Analytical Pyrolysis: A Comprehensive Guide. Marcel Dekker, New York. Ishida CK, Arnon S, Peterson CG, Kelly JJ & Gray KA (2008) Influence of algal community structure on denitrification rates in periphyton cultivated on artificial substrata. Microb Ecol 56: 140–152. Jackson CR (2003) Changes in community properties during microbial succession. Oikos 101: 444–448. Johnson RE, Tuchman NC & Peterson CG (1997) Changes in vertical microdistribution of diatoms during periphyton mat development. J N Am Benthol Soc 16: 503–519. Kirchman DL, Dittel AI, Findlay SEG & Fischer D (2004) Changes in bacterial activity and community structure in response to dissolved organic matter in the Hudson River, New York. Aquat Microb Ecol 35: 243–257. Kobayashi Y, Kim C, Yoshimizu C, Kohzu A, Tayasu I & Nagata T (2009) Longitudinal changes in bacterial community composition in river epilithic biofilms: influence of nutrients and organic matter. Aquat Microb Ecol 54: 135–153. Kühl M, Rickelt LF & Thar R (2007) Combined imaging of bacteria and oxygen in biofilms. Appl Environ Microb 73: 6289–6295. Lachnit T, Meske D, Wahl M, Harder T & Schmitz R (2011) Epibacterial community patterns on marine macroalgae are host-specific by temporally variable. Environ Microbiol 13: 655–665. Lear G & Lewis GD (2009) Impact of catchment land use on bacterial communities within stream biofilms. Ecol Indic 9: 848–855. Lock MA (1993) Attached microbial communities in rivers. Microbiology: An Ecological Approach (Ford TE, ed), pp. 113–138. Blackwell Scientific Publishers, Boston. Lock MA, Wallace RR, Costerton JW, Ventullo RM & Charlton SE (1984) River epilithon: toward a structural–functional model. Oikos 42: 10–22. Malinsky-Rushansky NZ & Legrand C (1996) Excretion of dissolved carbon by phytoplankton of different sizes and subsequent bacterial uptake. Mar Ecol-Prog Ser 132: 249–255. Mincer TJ, Church MJ, Taylor LT, Preston C, Karl DM & DeLong EF (2007) Quantitative distribution of presumptive archaeal


and bacterial nitrifiers in Monterey Bay and the North Pacific Subtropical Gyre. Environ Microbiol 9: 1162–1175. Moldoveanu S (1998) Analytical Pyrolysis of Natural Organic Polymers, 1st edn. Elsevier Science, Amsterdam, The Netherlands. Mulholland PJ, Steinman AD, Palumbo AV, Elwood JW & Kirschtel DB (1991) Role of nutrient cycling and herbivory in regulating periphyton communities in laboratory streams. Ecology 72: 966–982. Murphy J & Riley JP (1962) A modified single solution method for determination of phosphate in natural waters. Anal Chim Acta 26: 31–36. Murray AE, Arnosti C, De La Rocha CL, Grossart HP & Passow U (2007) Microbial dynamics in autotrophic and heterotrophic seawater mesocosms. II. Bacterioplankton community structure and hydrolytic enzyme activities. Aquat Microb Ecol 49: 123–141. Murray RE, Cooksey KE & Priscu JC (1986) Stimulation of bacterial DNA synthesis by algal exudates in attached algal–bacterial consortia. Appl Environ Microb 52: 1177–1182. Nogales B, Timmis KN, Nedwell DB & Osborn AM (2002) Detection and diversity of expressed denitrification genes in estuary sediments after reverse transcription-PCR amplification from mRNA. Appl Environ Microb 68: 5017–5025. Parnaudeau V & Dignac MF (2007) The organic matter composition of various wastewater sludges and their neutral detergent fractions as revealed by pyrolysis-GC/MS. J Anal Appl Pyrol 78: 140–152. Passy SI (2008) Species size and distribution jointly determine diatom densities in streams. Ecology 89: 475–484. Peschel G & Wildt T (1988) Humic substances of natural and anthropogeneous origin. Water Res 22: 105–108. Peterson CG & Grimm NB (1992) Temporal variation in enrichment effects during periphyton succession in a nitrogenlimited desert stream ecosystem. J N Am Benthol Soc 11: 20–36. Petrone KC, Richards JS & Grierson PF (2009) Bioavailability and composition of dissolved organic carbon and nitrogen in a near coastal catchment of south-western Australia. Biogeochemistry 92: 27–40. Pohlon E, Marxsen J & Küsel K (2010) Pioneering bacterial and algal communities and potential extracellular enzyme activities of stream biofilms. FEMS Microbiol Ecol 71: 364–373. Priemé A, Braker G & Tiedje JM (2002) Diversity of nitrite reductase (nirK and nirS) gene fragments in forested upland and wetland soils. Appl Environ Microbiol 68: 1893–1900. R Development Core Team (2008) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria, ISBN 3-900051-07-0. Available at http://www.R-project.org. Renslow RS, Majors PD, McLean JS, Fredrickson JK, Ahmed B & Beyenal H (2010) In situ effective diffusion coefficient profiles in live biofilms using pulse-field gradient nuclear magnetic resonance. Biotechnol Bioeng 106: 928–937.


c

492

Rusanvo AG, Grossart HP & Pusch MT (2009) Periphytic bacterial and algal response to a hydraulic gradient under different light levels: test of algal–bacterial coupling in the laboratory. Fund Appl Limnol 175: 339–353. Schäfer H, Abbas B, Witte H & Muyzer G (2002) Genetic diversity of ‘satellite’ bacteria present in cultures of marine diatoms. FEMS Microbiol Ecol 42: 25–35. Scott JT, Back JA, Taylor JM & King RS (2008) Does nutrient enrichment decouple algal–bacterial production in periphyton? J N Am Benthol Soc 27: 332–344. Sirivedhin T (2002) Qualitative and Quantitative Characterization of the Organic Matrix in Surface Water Systems: Methodology and Application. Northwestern University, Evanston, IL. Sirivedhin T & Gray KA (2005) Part I. Identifying anthropogenic markers in surface waters influenced by treated effluents: a tool in potable water reuse. Water Res 39: 1154–1164. Sirivedhin T & Gray KA (2006) Factors affecting denitrification rates in experimental wetlands: field and laboratory studies. Ecol Eng 26: 167–181. Solorzano L (1969) Determination of ammonia in natural waters by phenolhypochlorite method. Limnol Oceanogr 14: 799–801. Stevenson RJ, Peterson CG, Kirschtel DB, King CC & Tuchman NC (1991) Density-dependent growth, ecological strategies, and effects of nutrients and shading on benthic diatom succession in streams. J Phycol 27: 59–69. Stocker R, Seymour JR, Samadani A, Hunt DE & Polz MF (2008) Rapid chemotactic response enable marine bacteria to exploit ephemeral microscale nutrient patches. P Natl Acad Sci USA 105: 4209–4214. Throback IN, Enwall K, Jarvis A & Hallin S (2004) Reassessing PCR primers targeting nirS, nirK and nosZ genes for community surveys of denitrifying bacteria with DGGE. FEMS Microbiol Ecol 49: 401–417. Tiedje JM (1988) Ecology of denitrification and dissimilatory nitrate reduction to ammonium. Biology of Anaerobic Microorganisms, Vol. 170 (Zehnder AJB, ed), pp. 179–244. John Wiley & Sons, New York, NY. Venterink HO, Hummelink E & Van Den Hoorn MW (2003) Denitrification potential of a river floodplain during flooding with nitrate-rich water: grasslands versus reedbeds. Biogeochemistry 65: 233–244. Widrig DL, Gray KA & Mcauliffe KS (1996) Removal of algalderived organic material by preozonation and coagulation:


c


monitoring changes in organic quality by pyrolysis–GC–MS. Water Res 30: 2621–2632. Williams M, Hopkinson C, Rastetter E, Vallino J & Claessens L (2005) Relationships of land use and stream solute concentrations in the Ipswich River basin, northeastern Massachusetts. Water Air Soil Poll 161: 55–74. Wilson MA, Philp RP, Gillam AH, Gilbert TD & Tate KR (1983) Comparison of the structures of humic substances from aquatic and terrestrial sources by pyrolysis–gas chromatography–mass spectrometry. Geochim Cosmochim Ac 47: 497–502. Wolsing M & Priemé A (2004) Observation of high seasonal variation in community structure of denitrifying bacteria in arable soil receiving artificial fertilizer and cattle manure by determining T-RFLP of nir gene fragments. FEMS Microbiol Ecol 48: 261–271. Zar JH (1974) Biostatistical Analysis. Prentice Hall, Englewood Cliffs, NJ. Ziegler SE, Lyon DR & Townsend SL (2009) Carbon release and cycling within epilithic biofilms of two contrasting headwater streams. Aquat Microb Ecol 55: 285–300.

Supporting Information Additional Supporting Information may be found in the online version of this article: Fig. S1. NMDS ordination of pyrolysis fragments derived from 28-day-old periphyton samples (three replicates from each site) from two study streams [RP (open symbols) and DER (closed symbols)]. Fig. S2. Ln-transformed denitrification (as DNP) rates (mg N2O h1) in 28-day-old biofilms varying in the number of distinct nirS genotypes present in RP (open symbols) and DER (closed symbols). Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.
