- Email: [email protected]
Impact of the contemporary environment on denitrifying bacterial communities
Ecological Engineering 82 (2015) 469–473
Contents lists available at ScienceDirect
Ecological Engineering journal homepage: www.elsevier.com/locate/ecoleng
Short communication
Impact of the contemporary environment on denitrifying bacterial communities Sarah K. Hathawaya,1, Matthew D. Portera,2 , Luis F. Rodríguezb , Angela D. Kentc, Julie L. Zillesa,* a b c
Department of Civil and Environmental Engineering, University of Illinois at Urbana–Champaign, Urbana, IL 61081, USA Department of Agricultural and Biological Engineering, University of Illinois at Urbana–Champaign, Urbana, IL 61801, USA Department of Natural Resources and Environmental Sciences, University of Illinois at Urbana–Champaign, Urbana, IL 61801, USA
A R T I C L E I N F O
A B S T R A C T
Article history: Received 13 October 2014 Received in revised form 7 April 2015 Accepted 23 May 2015 Available online xxx
To engineer a microbial ecosystem, an understanding of the relative influence of legacy effects versus the contemporary environment on the microbial community is required. In this work, the influence of these factors on both the denitrifying bacterial community structure and the overall bacterial community structure was assessed through comparison of agricultural soils, denitrifying bioreactors, and natural and constructed wetlands. Terminal restriction fragment analysis (tRFLP) of the nosZ gene and automated ribosomal intergenic spacer analysis (ARISA) were used. The results suggest distinct communities are characteristic of soil, bioreactors, and wetlands and show a strong influence of the contemporary environment on the bacterial and denitrifying bacterial communities in the bioreactors. ã2015 Elsevier B.V. All rights reserved.
Keywords: Denitrifying bioreactors Legacy effects Microbial community composition nosZ Subsurface drainage Wetlands
1. Introduction Many engineered ecosystems rely on the activity of specific microbes to carry out their designed function; hence successful performance depends on the creation of environmental conditions that encourage the growth and activity of the desired microorganisms. However, microbial communities sometimes show legacy effects, defined here as persistent effects of a species or activity that is no longer present (Cuddington, 2011). For example, distinct soil microbial communities were observed between oldgrowth temperate forests and sites that had been logged or farmed and were returned to forest 50–75 years ago (Fraterrigo et al., 2006). Legacy effects have also been observed in microbial communities for soils cultivated seven years prior (Buckley and Schmidt, 2001) and in restored wetlands (Peralta et al., 2010). However, other studies of soil microbial communities suggest a
* Corresponding author at: 3230C Newmark Civil Engineering Laboratory, MC250, 205 North Mathews Avenue, Urbana, IL 61801, USA. Tel.: +1 217 244 2925. E-mail addresses: [email protected] (S.K. Hathaway), [email protected] (M.D. Porter), [email protected] (L.F. Rodríguez), [email protected] (A.D. Kent), [email protected] (J.L. Zilles). 1 Present address H2 M architects + engineers, 538 Broad Hollow Road, 4th Floor East, Melville, NY 11747, USA. 2 Present address: Environmental Resources Management, 75 Valley Stream Parkway, Suite 200, Malvern, PA 19355, USA. http://dx.doi.org/10.1016/j.ecoleng.2015.05.005 0925-8574/ ã 2015 Elsevier B.V. All rights reserved.
stronger role for the contemporary environment (e.g. (Jesus et al., 2009; Waldrop et al., 2000)). If legacy effects are strong in a particular environment, changes in environmental conditions alone are unlikely to restore or achieve desired activities in ecological engineering projects. On the other hand, when the effects of the contemporary environment are dominant, designing for changes in environmental conditions should be sufficient. Understanding the relative influence of contemporary environmental parameters versus legacy effects is therefore an important consideration for ecological engineering. To investigate the influence of environmental parameters versus legacy effects, this work focused on the engineered ecosystem of denitrifying bioreactors, which treat subsurface agricultural drainage, as reviewed by Schipper et al. (2010). Subsurface or tile drainage systems consist of perforated plastic tubes installed 0.6–1.2 m below the soil surface and are used to lower the water table; they are widespread in the Midwestern United States. While tile drainage increases the land available for agriculture, it also has the negative effect of increasing the transport of nitrate to surface waters (Blann et al., 2009). To remove nitrate from the drainage water, denitrifying bioreactors can be constructed inline with or at the outlet of the tile drainage system. These engineered systems are intended to mimic the nutrient removal service of wetland ecosystems. They provide appropriate conditions for denitrification through flow control
470
S.K. Hathaway et al. / Ecological Engineering 82 (2015) 469–473
wetlands allowed comparison with another engineered environment. Constructed wetlands achieve varying degrees of success in delivery of denitrification and other ecosystem services, and differences in ecosystem services are correlated with differences in microbial community composition between the natural and constructed ecosystems (Flanagan, 2009; Peralta et al., 2010). In one previous wetland study, 40% of the variation in denitrification potential could be accounted for by differences in bacterial community composition (Peralta et al., 2010). Agricultural fields were selected because they represent the prior land use of the bioreactor locations and contribute to inoculation of and immigration into the bioreactors. These habitats represent a succession of land uses in the Illinois landscape (wetland to agricultural field to constructed wetland or bioreactor) and provide
structures that provide sufficient retention time within the bioreactor, woodchips that provide organic carbon, and rapid depletion of oxygen through microbial activity. As currently implemented, the bioreactors are not deliberately inoculated, instead containing native microorganisms from the surrounding soil and woodchips, and receive agricultural soil microorganisms with the influent water (immigration). The objective of this work was to test whether or not the contemporary environment controls the bacterial and denitrifying bacterial communities in denitrifying bioreactors through comparison with communities from natural and constructed wetlands and agricultural fields. The bioreactors were compared to wetlands because wetlands are environments with high rates of denitrification and because the inclusion of both natural and constructed
A
0.6 agriculture wetland
0.4
constructed wetland bioreactor
0.2
0 -0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
-0.2
-0.4
-0.6 1.5
B
agriculture wetland 1
constructed wetland bioreactor
0.5
0 -1.5
-1
-0.5
0
0.5
1
1.5
-0.5
-1
Fig. 1. Partial correspondence analysis on (a) ARISA data and (b) nosZ data representing microbial communities from different habitats. Points show the average location, or centroid, of all samples in the habitat group; error bars represent one standard deviation of sample scores on each axis. The vertical error bars for the field habitat in panel a are obscured by the symbol, as the standard deviation was only 0.012 in this dimension. Partial correspondence analysis removed variance explained by latitude and longitude, to allow visualization of patterns in community composition based on habitat type. The correspondence analysis axes represent theoretical environmental gradients, and the distance between points represents the dissimilarity of their microbial communities.
S.K. Hathaway et al. / Ecological Engineering 82 (2015) 469–473
an excellent setting in which to examine the impact of land use legacy vs. contemporary environmental and management conditions on microbial community structure. If land use legacy and/ or immigration effects dominate, then the bioreactor microbial communities should be more similar to the agricultural soils from which they originate and are hydrologically connected to. Alternatively, if the contemporary environment has a stronger effect, then the bioreactor microbial communities should be more similar to other denitrifying environments, such as wetlands. 2. Materials and methods 2.1. Sampling Samples were collected for microbial community analysis from four different habitats: field-scale denitrifying bioreactors, constructed wetlands, natural wetlands, and agricultural fields. Bioreactor woodchip samples were collected from three different reactors in central Illinois in July and August 2010 using existing sampling ports; a full description of sampling methods can be found in Porter (2011). Two water samples from bioreactor sampling ports were also included. In addition, for one of the bioreactors, soil samples were collected from the field that drains into the bioreactor and from a wetland near the bioreactor discharge location. For each of the locations in the field and wetland, three 1.9 cm diameter cores were collected within a 1 m2 sampling quadrat; the upper 6 cm from the three cores were combined. In addition to these closely located sampling sites, wetland sediment samples were collected from six pairs of natural and restored wetlands in Illinois in June 2007 (Flanagan, 2009). Each sediment sample was the composite of eight soil cores (1.9 cm diameter) collected from within a 1 m2 quadrat at a depth to 12 cm. Two samples from each wetland were used in this analysis. Agricultural soil samples were also collected from a field under a corn-soybean rotation near one of the wetland pairs near Dekalb, IL in June 2007, from the Morrow Plots (a long-term research field on the University of Illinois, Urbana–Champaign campus) in summer 2009, Miscanthus and switchgrass at Dixon Springs Agricultural Center in Simpson, IL, and a field in a corn-soy rotation at University of Illinois South Farms in Champaign, IL. The last three fields were sampled in 2009. At the Morrow Plots, 6 cores (1.9 cm) were collected to a depth of 12 cm from within a 1 m2 sampling quadrat and composited. The remaining three agricultural sites were sampled to plow depth (24 cm). All samples were stored at 20 C until DNA extraction. 2.2. Molecular analysis Bioreactor DNA was extracted using the FastDNA Spin Kit (MP Biomedicals, Solon, OH), and DNA from soil and sediment samples was extracted using the FastDNA Soil Spin Kit (MP Biomedicals, Solon, OH), both according to the manufacturer’s directions. The extracted DNA was purified using cetyl trimethyl ammonium bromide (CTAB) cleanup (Peralta et al., 2010) to remove humic acid contamination. The abundance of atypical or clade II nosZ was measured using primers nosZ-II-F and nosZ-II-R (Jones et al., 2013) for qPCR on a 7900HT Real-time Quantitative PCR machine (Applied Biosystems, Grand Island, NY). Bacterial and denitrifying bacterial community analysis was conducted using automated ribosomal intergenic spacer analysis (ARISA) and terminal restriction fragment analysis (tRFLP) of the clade I nitrous oxide reductase gene (nosZ), respectively, as described previously (Andrus et al., 2014). The abundance of atypical or clade II nosZ was measured using a published qPCR method (Jones et al., 2013). The chromatographs were analyzed using GeneMarker 2.2.0 (SoftGenetics, LLC, State College, PA). For nosZ tRFLP, restriction
471
enzymes AluI and HhaI were used. Minimum fluorescence thresholds of 200 and 80 were used, respectively, for the ARISA and nosZ tRFLP analyses in GeneMarker. 2.3. Statistical analyses Partial correspondence analysis was carried out using Canoco for Windows 4.5.1 (Plant Research International, Wageningen, Netherlands) (ter Braak and Smilauer, 2002). The influence of location was removed from the analysis with latitude and longitude as covariables. PERMANOVA, a non-parametric multivariate analysis of variance that generates p-values using permutations (Anderson, 2001; McArdle and Anderson, 2001), was used to test the effect of location and habitat type on bacterial and denitrifier composition and was carried out using the adonis function from the vegan package (Oksanen et al., 2012) in the R statistical environment (R Development Core Team, 2011). 3. Results and discussion The comparison of bioreactors, agricultural soils, and restored and natural wetlands sediments showed that each habitat had distinct communities of bacteria, as assessed using ARISA. In this plot from partial correspondence analysis (Fig. 1A), samples from the same habitat type have been grouped together and represented by their average location, or centroid. Based on the distance between centroids, the agricultural samples and the bioreactor samples contained microbial communities that were distinct from each other and from the wetland microbial communities (Table 1). PERMANOVA confirmed that these differences included an effect of habitat type in addition to the well-known effect of geographic location (Bacteria: Location R2 = 0.34, p < 0.001 and Habitat R2=0.14, p < 0.001). The same pattern was seen when analyzing only those samples collected on a single day and from a single geographic location, namely the field, bioreactor, and wetland that were hydrologically connected (Fig. 2A), further confirming that this result was not an artifact resulting from temporal variation or geographic distribution of the sampling sites. Perhaps most importantly, the bioreactor bacterial communities were more similar to those in wetlands than those in agricultural fields. To assess the denitrifying community, tRFLP of the clade I nosZ was used. This approach targets Gram-negative bacteria that perform complete denitrification. The denitrifying community composition was not analyzed using the atypical or clade II nosZ more often found in Gram positive bacteria (Jones et al., 2013; Sanford et al., 2012) because atypical nosZ was very low in abundance (<8 copies ng 1 DNA, as quantified by qPCR) in the bioreactor samples. This is in contrast to results from soil samples, where clade II nosZ was generally at similar or greater abundance than clade I nosZ, as measured by qPCR (Jones et al., 2013) and by analysis of soil metagenomes (Orellana et al., 2014). This difference could provide insight into the ecological niches preferred by these two types of bacterial denitrifiers. Fungal denitrification could also contribute to nitrate removal, but this was not considered here Table 1 Centroid distances between each habitat’s communities. Habitats
Centroid Distances ARISA
nosZ
Wetland, restored wetland Bioreactor, wetland Bioreactor, restored wetland Field, wetland Field, restored wetland Bioreactor, field
0.207 0.526 0.564 0.685 0.716 1.211
0.641 1.262 0.950 1.302 0.806 1.586
472
S.K. Hathaway et al. / Ecological Engineering 82 (2015) 469–473
Total Bacteria (ARISA) A
Denitrifying Bacteria (nosZ T-RFLP) B
Axis 1
Position Bioreactor Field Wetland
Axis 2
Axis 2
Fig. 2. Non-metric multidimensional Scaling (MDS) plots of comparison of (A) total bacteria (ARISA), and (B) denitrifying bacteria (nosZ T-RFLP) site comparisons among a tile-drained agricultural field, the denitrifying bioreactor receiving this field’s drainage, and a wetland slightly downstream of the bioreactor outlet. Points represent the microbial community from individual samples. The axes represent theoretical environmental variables, and the distance between points represents the dissimilarity of their microbial communities.
because inhibition studies suggest that denitrification in the bioreactors is primarily bacterial (Appleford et al., 2008). In addition, specific molecular methods for analyzing the fungal denitrifier community are not, to our knowledge, available. Focusing therefore on the clade I nosZ bacterial denitrifier community, the same general pattern was observed as for the bacterial community. Each habitat had distinct communities (Fig. 1B), and the bioreactor communities were more similar to wetlands than to agricultural fields (Table 1). PERMANOVA again confirmed an effect of habitat type (habitat R2 = 0.15, p < 0.001, Location R2 = 0.33, p < 0.001). The same pattern was again observed when only those field, bioreactor, and wetland samples from the same sampling day and geographic location were considered (Fig. 2B). The distinct clusters of samples in our correspondence analyses are striking, particularly since a wide range of locations were sampled across the state of Illinois for both wetlands and agricultural fields. Together with the PERMANOVA results, these clusters suggest there are microbial assemblages that are characteristic of each habitat type, despite spatial differences in microbial assemblages and changes over time. Characteristic microbial assemblages could result from differences in basic properties such as organic matter and hydrology. Since the microbial communities in engineered systems (constructed wetlands and bioreactors) are distinct from the source community (agricultural fields in each case), this suggests that contemporary environmental conditions are a strong determinant of microbial community composition in these environments. Despite prior land use and continual immigration of soil microbes from the agricultural field via the subsurface drainage water, the bioreactors have established bacterial and denitrifier assemblages that are distinct from the source community, and more similar to microbial assemblages observed in wetland habitats (the habitat that shares the microbial ecosystem services provided by the bioreactors). The results of this habitat comparison, along with published performance data demonstrating nitrate removal as reviewed by Schipper et al. (2010), indicate that denitrifying microbial communities can successfully be developed through the implementation of favorable environmental conditions. The strong effect of the contemporary environment observed here suggests that optimization of denitrifying bioreactors is likely to result from refinements in design and management parameters that influence the environmental conditions, rather than through inoculation or other direct manipulations of the microbial community.
Acknowledgements We thank A. Yannarell for contributing samples and landowners and the Illinois Department of Transportation for allowing sampling. Funding was provided by United States National Science Foundation project CBET-0853820. References Anderson, M.J., 2001. A new method for non-parametric multivariate analysis of variance. Aust. Ecol. 26, 32–46. Andrus, J.M., Porter, M.D., Rodríguez, L.F., Kuehlhorn, T., Cooke, R.A.C., Zhang, Y., Kent, A.D., Zilles, J.L., 2014. Spatial variation in the bacterial and denitrifying bacterial community in a biofilter treating subsurface agricultural drainage. Microb. Ecol. 67, 265–272. Appleford, J.M., Rodriguez, L.F., Cooke, R.A.C., Zhang, Y., Kent, A.D., Zilles, J., 2008. Characterization of microorganisms contributing to denitrification in tile drain biofilters in Illinois. 2008 ASABE Annual Meeting, St. Joseph, MI. Blann, K.L., Anderson, J.L., Sands, G.R., Vondracek, B, 2009. Effects of agricultural drainage on aquatic ecosystems: a review. Crit. Rev. Environ. Sci. Technol. 39, 909–1001. Buckley, D.H., Schmidt, T.M., 2001. The structure of microbial communities in soil and the lasting impact of cultivation. Microb. Ecol. 42, 11–21. Cuddington, K., 2011. Legacy effects: the persistent impact of ecological interactions. Biol. Theory 6, 203–210. Flanagan, D.F., 2009. Microbial community structure and function in wetland ecosystems. Natural Resources and Environmental Sciences. University of Illinois at Urbana–Champaign, Urbana, IL. Fraterrigo, J., Balser, T., Turner, M, 2006. Microbial community variation and its relationship with nitrogen mineralization in historically altered forests. Ecology 87, 570–579. Jesus, E., Marsh, T., Tiedje, J., Moreira, F, 2009. Changes in land use alter the structure of bacterial communities in Western Amazon soils. ISME J. 3, 1004–1011. Jones, C.M., Graf, D.R., Bru, D., Philippot, L., Hallin, S., 2013. The unaccounted yet abundant nitrous oxide-reducing microbial community: a potential nitrous oxide sink. ISME J. 7, 417–426. McArdle, B.H., Anderson, M.J., 2001. Fitting multivariate models to community data: a comment on distance-based redundancy analysis. Ecology 82, 290–297. Oksanen J., Blanchet F.G., Kindt R., Legendre P., Minchin P.R., O’Hara R.B., Simpson G. L., Solymos P., Stevens M.H.H., Wagner, H. vegan. Community Ecology Package. R package version 2. 0–4, 2012. Orellana, L.H., Rodriguez-R, L.M., Higgins, S., Chee-Sanford, J.C., Sanford, R.A., Ritalahti, K.M., Löffler, F.E., Konstantinidis, KT, 2014. Detecting nitrous oxide reductase (nosZ) genes in soil metagenomes: method development and implications for the nitrogen cycle. mBio 5. Peralta, A.L., Matthews, J.W., Kent, AD, 2010. Microbial community structure and denitrification in a wetland mitigation bank. Appl. Environ. Microbiol. 76, 4207–4215. Porter, M.D., 2011. Microbial community dynamics in denitrifying biofilters receiving agricultural drainage. Civil and Environmental Engineering. University of Illinois Urbana–Champaign, Urbana, IL. R Development Core Team, R.R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing, 2011. Sanford, R.A., Wagner, D.D., Wu, Q., Chee-Sanford, J.C., Thomas, S.H., Cruz-Garcia, C., Rodriguez, G., Massol-Deya, A., Krishnani, K.K., Ritalahti, K.M., Nissen, S., Konstantinidis, K.T., Loffler, FE, 2012. Unexpected nondenitrifier nitrous oxide
S.K. Hathaway et al. / Ecological Engineering 82 (2015) 469–473 reductase gene diversity and abundance in soils. Proc. Nat. Acad. Sci. U. S. A 109, 19709–19714. Schipper, L.A., Robertson, W.D., Gold, A.J., Jaynes, D.B., Cameron, SC, 2010. Denitrifying bioreactors–an approach for reducing nitrate loads to receiving waters. Ecol. Eng. 36, 1532–1543.
473
CANOCO reference manual and CanoDraw for Windows user’s guide: software for canonical community ordination (version 4.5). In: ter Braak, C.J.F., Smilauer, P. (Eds.), Wageningen University and Research Center Publications, Wageningen, Netherlands. Waldrop, M., Balser, T., Firestone, M., 2000. Linking microbial community composition to function in a tropical soil. Soil Biol. Biochem. 32, 1837–1846.
Contents lists available at ScienceDirect
Ecological Engineering journal homepage: www.elsevier.com/locate/ecoleng
Short communication
Impact of the contemporary environment on denitrifying bacterial communities Sarah K. Hathawaya,1, Matthew D. Portera,2 , Luis F. Rodríguezb , Angela D. Kentc, Julie L. Zillesa,* a b c
Department of Civil and Environmental Engineering, University of Illinois at Urbana–Champaign, Urbana, IL 61081, USA Department of Agricultural and Biological Engineering, University of Illinois at Urbana–Champaign, Urbana, IL 61801, USA Department of Natural Resources and Environmental Sciences, University of Illinois at Urbana–Champaign, Urbana, IL 61801, USA
A R T I C L E I N F O
A B S T R A C T
Article history: Received 13 October 2014 Received in revised form 7 April 2015 Accepted 23 May 2015 Available online xxx
To engineer a microbial ecosystem, an understanding of the relative influence of legacy effects versus the contemporary environment on the microbial community is required. In this work, the influence of these factors on both the denitrifying bacterial community structure and the overall bacterial community structure was assessed through comparison of agricultural soils, denitrifying bioreactors, and natural and constructed wetlands. Terminal restriction fragment analysis (tRFLP) of the nosZ gene and automated ribosomal intergenic spacer analysis (ARISA) were used. The results suggest distinct communities are characteristic of soil, bioreactors, and wetlands and show a strong influence of the contemporary environment on the bacterial and denitrifying bacterial communities in the bioreactors. ã2015 Elsevier B.V. All rights reserved.
Keywords: Denitrifying bioreactors Legacy effects Microbial community composition nosZ Subsurface drainage Wetlands
1. Introduction Many engineered ecosystems rely on the activity of specific microbes to carry out their designed function; hence successful performance depends on the creation of environmental conditions that encourage the growth and activity of the desired microorganisms. However, microbial communities sometimes show legacy effects, defined here as persistent effects of a species or activity that is no longer present (Cuddington, 2011). For example, distinct soil microbial communities were observed between oldgrowth temperate forests and sites that had been logged or farmed and were returned to forest 50–75 years ago (Fraterrigo et al., 2006). Legacy effects have also been observed in microbial communities for soils cultivated seven years prior (Buckley and Schmidt, 2001) and in restored wetlands (Peralta et al., 2010). However, other studies of soil microbial communities suggest a
* Corresponding author at: 3230C Newmark Civil Engineering Laboratory, MC250, 205 North Mathews Avenue, Urbana, IL 61801, USA. Tel.: +1 217 244 2925. E-mail addresses: [email protected] (S.K. Hathaway), [email protected] (M.D. Porter), [email protected] (L.F. Rodríguez), [email protected] (A.D. Kent), [email protected] (J.L. Zilles). 1 Present address H2 M architects + engineers, 538 Broad Hollow Road, 4th Floor East, Melville, NY 11747, USA. 2 Present address: Environmental Resources Management, 75 Valley Stream Parkway, Suite 200, Malvern, PA 19355, USA. http://dx.doi.org/10.1016/j.ecoleng.2015.05.005 0925-8574/ ã 2015 Elsevier B.V. All rights reserved.
stronger role for the contemporary environment (e.g. (Jesus et al., 2009; Waldrop et al., 2000)). If legacy effects are strong in a particular environment, changes in environmental conditions alone are unlikely to restore or achieve desired activities in ecological engineering projects. On the other hand, when the effects of the contemporary environment are dominant, designing for changes in environmental conditions should be sufficient. Understanding the relative influence of contemporary environmental parameters versus legacy effects is therefore an important consideration for ecological engineering. To investigate the influence of environmental parameters versus legacy effects, this work focused on the engineered ecosystem of denitrifying bioreactors, which treat subsurface agricultural drainage, as reviewed by Schipper et al. (2010). Subsurface or tile drainage systems consist of perforated plastic tubes installed 0.6–1.2 m below the soil surface and are used to lower the water table; they are widespread in the Midwestern United States. While tile drainage increases the land available for agriculture, it also has the negative effect of increasing the transport of nitrate to surface waters (Blann et al., 2009). To remove nitrate from the drainage water, denitrifying bioreactors can be constructed inline with or at the outlet of the tile drainage system. These engineered systems are intended to mimic the nutrient removal service of wetland ecosystems. They provide appropriate conditions for denitrification through flow control
470
S.K. Hathaway et al. / Ecological Engineering 82 (2015) 469–473
wetlands allowed comparison with another engineered environment. Constructed wetlands achieve varying degrees of success in delivery of denitrification and other ecosystem services, and differences in ecosystem services are correlated with differences in microbial community composition between the natural and constructed ecosystems (Flanagan, 2009; Peralta et al., 2010). In one previous wetland study, 40% of the variation in denitrification potential could be accounted for by differences in bacterial community composition (Peralta et al., 2010). Agricultural fields were selected because they represent the prior land use of the bioreactor locations and contribute to inoculation of and immigration into the bioreactors. These habitats represent a succession of land uses in the Illinois landscape (wetland to agricultural field to constructed wetland or bioreactor) and provide
structures that provide sufficient retention time within the bioreactor, woodchips that provide organic carbon, and rapid depletion of oxygen through microbial activity. As currently implemented, the bioreactors are not deliberately inoculated, instead containing native microorganisms from the surrounding soil and woodchips, and receive agricultural soil microorganisms with the influent water (immigration). The objective of this work was to test whether or not the contemporary environment controls the bacterial and denitrifying bacterial communities in denitrifying bioreactors through comparison with communities from natural and constructed wetlands and agricultural fields. The bioreactors were compared to wetlands because wetlands are environments with high rates of denitrification and because the inclusion of both natural and constructed
A
0.6 agriculture wetland
0.4
constructed wetland bioreactor
0.2
0 -0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
-0.2
-0.4
-0.6 1.5
B
agriculture wetland 1
constructed wetland bioreactor
0.5
0 -1.5
-1
-0.5
0
0.5
1
1.5
-0.5
-1
Fig. 1. Partial correspondence analysis on (a) ARISA data and (b) nosZ data representing microbial communities from different habitats. Points show the average location, or centroid, of all samples in the habitat group; error bars represent one standard deviation of sample scores on each axis. The vertical error bars for the field habitat in panel a are obscured by the symbol, as the standard deviation was only 0.012 in this dimension. Partial correspondence analysis removed variance explained by latitude and longitude, to allow visualization of patterns in community composition based on habitat type. The correspondence analysis axes represent theoretical environmental gradients, and the distance between points represents the dissimilarity of their microbial communities.
S.K. Hathaway et al. / Ecological Engineering 82 (2015) 469–473
an excellent setting in which to examine the impact of land use legacy vs. contemporary environmental and management conditions on microbial community structure. If land use legacy and/ or immigration effects dominate, then the bioreactor microbial communities should be more similar to the agricultural soils from which they originate and are hydrologically connected to. Alternatively, if the contemporary environment has a stronger effect, then the bioreactor microbial communities should be more similar to other denitrifying environments, such as wetlands. 2. Materials and methods 2.1. Sampling Samples were collected for microbial community analysis from four different habitats: field-scale denitrifying bioreactors, constructed wetlands, natural wetlands, and agricultural fields. Bioreactor woodchip samples were collected from three different reactors in central Illinois in July and August 2010 using existing sampling ports; a full description of sampling methods can be found in Porter (2011). Two water samples from bioreactor sampling ports were also included. In addition, for one of the bioreactors, soil samples were collected from the field that drains into the bioreactor and from a wetland near the bioreactor discharge location. For each of the locations in the field and wetland, three 1.9 cm diameter cores were collected within a 1 m2 sampling quadrat; the upper 6 cm from the three cores were combined. In addition to these closely located sampling sites, wetland sediment samples were collected from six pairs of natural and restored wetlands in Illinois in June 2007 (Flanagan, 2009). Each sediment sample was the composite of eight soil cores (1.9 cm diameter) collected from within a 1 m2 quadrat at a depth to 12 cm. Two samples from each wetland were used in this analysis. Agricultural soil samples were also collected from a field under a corn-soybean rotation near one of the wetland pairs near Dekalb, IL in June 2007, from the Morrow Plots (a long-term research field on the University of Illinois, Urbana–Champaign campus) in summer 2009, Miscanthus and switchgrass at Dixon Springs Agricultural Center in Simpson, IL, and a field in a corn-soy rotation at University of Illinois South Farms in Champaign, IL. The last three fields were sampled in 2009. At the Morrow Plots, 6 cores (1.9 cm) were collected to a depth of 12 cm from within a 1 m2 sampling quadrat and composited. The remaining three agricultural sites were sampled to plow depth (24 cm). All samples were stored at 20 C until DNA extraction. 2.2. Molecular analysis Bioreactor DNA was extracted using the FastDNA Spin Kit (MP Biomedicals, Solon, OH), and DNA from soil and sediment samples was extracted using the FastDNA Soil Spin Kit (MP Biomedicals, Solon, OH), both according to the manufacturer’s directions. The extracted DNA was purified using cetyl trimethyl ammonium bromide (CTAB) cleanup (Peralta et al., 2010) to remove humic acid contamination. The abundance of atypical or clade II nosZ was measured using primers nosZ-II-F and nosZ-II-R (Jones et al., 2013) for qPCR on a 7900HT Real-time Quantitative PCR machine (Applied Biosystems, Grand Island, NY). Bacterial and denitrifying bacterial community analysis was conducted using automated ribosomal intergenic spacer analysis (ARISA) and terminal restriction fragment analysis (tRFLP) of the clade I nitrous oxide reductase gene (nosZ), respectively, as described previously (Andrus et al., 2014). The abundance of atypical or clade II nosZ was measured using a published qPCR method (Jones et al., 2013). The chromatographs were analyzed using GeneMarker 2.2.0 (SoftGenetics, LLC, State College, PA). For nosZ tRFLP, restriction
471
enzymes AluI and HhaI were used. Minimum fluorescence thresholds of 200 and 80 were used, respectively, for the ARISA and nosZ tRFLP analyses in GeneMarker. 2.3. Statistical analyses Partial correspondence analysis was carried out using Canoco for Windows 4.5.1 (Plant Research International, Wageningen, Netherlands) (ter Braak and Smilauer, 2002). The influence of location was removed from the analysis with latitude and longitude as covariables. PERMANOVA, a non-parametric multivariate analysis of variance that generates p-values using permutations (Anderson, 2001; McArdle and Anderson, 2001), was used to test the effect of location and habitat type on bacterial and denitrifier composition and was carried out using the adonis function from the vegan package (Oksanen et al., 2012) in the R statistical environment (R Development Core Team, 2011). 3. Results and discussion The comparison of bioreactors, agricultural soils, and restored and natural wetlands sediments showed that each habitat had distinct communities of bacteria, as assessed using ARISA. In this plot from partial correspondence analysis (Fig. 1A), samples from the same habitat type have been grouped together and represented by their average location, or centroid. Based on the distance between centroids, the agricultural samples and the bioreactor samples contained microbial communities that were distinct from each other and from the wetland microbial communities (Table 1). PERMANOVA confirmed that these differences included an effect of habitat type in addition to the well-known effect of geographic location (Bacteria: Location R2 = 0.34, p < 0.001 and Habitat R2=0.14, p < 0.001). The same pattern was seen when analyzing only those samples collected on a single day and from a single geographic location, namely the field, bioreactor, and wetland that were hydrologically connected (Fig. 2A), further confirming that this result was not an artifact resulting from temporal variation or geographic distribution of the sampling sites. Perhaps most importantly, the bioreactor bacterial communities were more similar to those in wetlands than those in agricultural fields. To assess the denitrifying community, tRFLP of the clade I nosZ was used. This approach targets Gram-negative bacteria that perform complete denitrification. The denitrifying community composition was not analyzed using the atypical or clade II nosZ more often found in Gram positive bacteria (Jones et al., 2013; Sanford et al., 2012) because atypical nosZ was very low in abundance (<8 copies ng 1 DNA, as quantified by qPCR) in the bioreactor samples. This is in contrast to results from soil samples, where clade II nosZ was generally at similar or greater abundance than clade I nosZ, as measured by qPCR (Jones et al., 2013) and by analysis of soil metagenomes (Orellana et al., 2014). This difference could provide insight into the ecological niches preferred by these two types of bacterial denitrifiers. Fungal denitrification could also contribute to nitrate removal, but this was not considered here Table 1 Centroid distances between each habitat’s communities. Habitats
Centroid Distances ARISA
nosZ
Wetland, restored wetland Bioreactor, wetland Bioreactor, restored wetland Field, wetland Field, restored wetland Bioreactor, field
0.207 0.526 0.564 0.685 0.716 1.211
0.641 1.262 0.950 1.302 0.806 1.586
472
S.K. Hathaway et al. / Ecological Engineering 82 (2015) 469–473
Total Bacteria (ARISA) A
Denitrifying Bacteria (nosZ T-RFLP) B
Axis 1
Position Bioreactor Field Wetland
Axis 2
Axis 2
Fig. 2. Non-metric multidimensional Scaling (MDS) plots of comparison of (A) total bacteria (ARISA), and (B) denitrifying bacteria (nosZ T-RFLP) site comparisons among a tile-drained agricultural field, the denitrifying bioreactor receiving this field’s drainage, and a wetland slightly downstream of the bioreactor outlet. Points represent the microbial community from individual samples. The axes represent theoretical environmental variables, and the distance between points represents the dissimilarity of their microbial communities.
because inhibition studies suggest that denitrification in the bioreactors is primarily bacterial (Appleford et al., 2008). In addition, specific molecular methods for analyzing the fungal denitrifier community are not, to our knowledge, available. Focusing therefore on the clade I nosZ bacterial denitrifier community, the same general pattern was observed as for the bacterial community. Each habitat had distinct communities (Fig. 1B), and the bioreactor communities were more similar to wetlands than to agricultural fields (Table 1). PERMANOVA again confirmed an effect of habitat type (habitat R2 = 0.15, p < 0.001, Location R2 = 0.33, p < 0.001). The same pattern was again observed when only those field, bioreactor, and wetland samples from the same sampling day and geographic location were considered (Fig. 2B). The distinct clusters of samples in our correspondence analyses are striking, particularly since a wide range of locations were sampled across the state of Illinois for both wetlands and agricultural fields. Together with the PERMANOVA results, these clusters suggest there are microbial assemblages that are characteristic of each habitat type, despite spatial differences in microbial assemblages and changes over time. Characteristic microbial assemblages could result from differences in basic properties such as organic matter and hydrology. Since the microbial communities in engineered systems (constructed wetlands and bioreactors) are distinct from the source community (agricultural fields in each case), this suggests that contemporary environmental conditions are a strong determinant of microbial community composition in these environments. Despite prior land use and continual immigration of soil microbes from the agricultural field via the subsurface drainage water, the bioreactors have established bacterial and denitrifier assemblages that are distinct from the source community, and more similar to microbial assemblages observed in wetland habitats (the habitat that shares the microbial ecosystem services provided by the bioreactors). The results of this habitat comparison, along with published performance data demonstrating nitrate removal as reviewed by Schipper et al. (2010), indicate that denitrifying microbial communities can successfully be developed through the implementation of favorable environmental conditions. The strong effect of the contemporary environment observed here suggests that optimization of denitrifying bioreactors is likely to result from refinements in design and management parameters that influence the environmental conditions, rather than through inoculation or other direct manipulations of the microbial community.
Acknowledgements We thank A. Yannarell for contributing samples and landowners and the Illinois Department of Transportation for allowing sampling. Funding was provided by United States National Science Foundation project CBET-0853820. References Anderson, M.J., 2001. A new method for non-parametric multivariate analysis of variance. Aust. Ecol. 26, 32–46. Andrus, J.M., Porter, M.D., Rodríguez, L.F., Kuehlhorn, T., Cooke, R.A.C., Zhang, Y., Kent, A.D., Zilles, J.L., 2014. Spatial variation in the bacterial and denitrifying bacterial community in a biofilter treating subsurface agricultural drainage. Microb. Ecol. 67, 265–272. Appleford, J.M., Rodriguez, L.F., Cooke, R.A.C., Zhang, Y., Kent, A.D., Zilles, J., 2008. Characterization of microorganisms contributing to denitrification in tile drain biofilters in Illinois. 2008 ASABE Annual Meeting, St. Joseph, MI. Blann, K.L., Anderson, J.L., Sands, G.R., Vondracek, B, 2009. Effects of agricultural drainage on aquatic ecosystems: a review. Crit. Rev. Environ. Sci. Technol. 39, 909–1001. Buckley, D.H., Schmidt, T.M., 2001. The structure of microbial communities in soil and the lasting impact of cultivation. Microb. Ecol. 42, 11–21. Cuddington, K., 2011. Legacy effects: the persistent impact of ecological interactions. Biol. Theory 6, 203–210. Flanagan, D.F., 2009. Microbial community structure and function in wetland ecosystems. Natural Resources and Environmental Sciences. University of Illinois at Urbana–Champaign, Urbana, IL. Fraterrigo, J., Balser, T., Turner, M, 2006. Microbial community variation and its relationship with nitrogen mineralization in historically altered forests. Ecology 87, 570–579. Jesus, E., Marsh, T., Tiedje, J., Moreira, F, 2009. Changes in land use alter the structure of bacterial communities in Western Amazon soils. ISME J. 3, 1004–1011. Jones, C.M., Graf, D.R., Bru, D., Philippot, L., Hallin, S., 2013. The unaccounted yet abundant nitrous oxide-reducing microbial community: a potential nitrous oxide sink. ISME J. 7, 417–426. McArdle, B.H., Anderson, M.J., 2001. Fitting multivariate models to community data: a comment on distance-based redundancy analysis. Ecology 82, 290–297. Oksanen J., Blanchet F.G., Kindt R., Legendre P., Minchin P.R., O’Hara R.B., Simpson G. L., Solymos P., Stevens M.H.H., Wagner, H. vegan. Community Ecology Package. R package version 2. 0–4, 2012. Orellana, L.H., Rodriguez-R, L.M., Higgins, S., Chee-Sanford, J.C., Sanford, R.A., Ritalahti, K.M., Löffler, F.E., Konstantinidis, KT, 2014. Detecting nitrous oxide reductase (nosZ) genes in soil metagenomes: method development and implications for the nitrogen cycle. mBio 5. Peralta, A.L., Matthews, J.W., Kent, AD, 2010. Microbial community structure and denitrification in a wetland mitigation bank. Appl. Environ. Microbiol. 76, 4207–4215. Porter, M.D., 2011. Microbial community dynamics in denitrifying biofilters receiving agricultural drainage. Civil and Environmental Engineering. University of Illinois Urbana–Champaign, Urbana, IL. R Development Core Team, R.R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing, 2011. Sanford, R.A., Wagner, D.D., Wu, Q., Chee-Sanford, J.C., Thomas, S.H., Cruz-Garcia, C., Rodriguez, G., Massol-Deya, A., Krishnani, K.K., Ritalahti, K.M., Nissen, S., Konstantinidis, K.T., Loffler, FE, 2012. Unexpected nondenitrifier nitrous oxide
S.K. Hathaway et al. / Ecological Engineering 82 (2015) 469–473 reductase gene diversity and abundance in soils. Proc. Nat. Acad. Sci. U. S. A 109, 19709–19714. Schipper, L.A., Robertson, W.D., Gold, A.J., Jaynes, D.B., Cameron, SC, 2010. Denitrifying bioreactors–an approach for reducing nitrate loads to receiving waters. Ecol. Eng. 36, 1532–1543.
473
CANOCO reference manual and CanoDraw for Windows user’s guide: software for canonical community ordination (version 4.5). In: ter Braak, C.J.F., Smilauer, P. (Eds.), Wageningen University and Research Center Publications, Wageningen, Netherlands. Waldrop, M., Balser, T., Firestone, M., 2000. Linking microbial community composition to function in a tropical soil. Soil Biol. Biochem. 32, 1837–1846.