Abstract
Biochemical nitrogen removal relies heavily on wetland microbes, prior to entering downstream ecosystems. Microbes utilize oxy-reductases that contain metal-cofactors to reduce NO3 and NO2 to N2. These transformations are linked also to other C cycling processes and may emit N2O, CH4 and CO2 as well. In this incubation study we tested the effects of supplementing wetland sediments with trace-metals (Cu, Mo, Fe) at μM levels on nitrogen kinetics, gaseous emissions (N2O, CO2 and CH4) and genes abundances for the overall microbial community, related to NO2 reduction (denitrification and DNRA), N2O reduction and methanogenesis. Trace metal availability and specifically Cu, enhances the reduction of N2O to N2, increases the production of CH4 and has no effect on CO2 emissions from wetland sediments. When sediments were supplied with Cu, higher gene abundances were observed for Cu-reductases nirK and nosZ. Mo addition increased CH4 emissions from sediments and higher mcrA gene abundances were observed. Emerging from this investigation, trace-metals bio-availability may directly or indirectly regulate denitrification in the environment; an effect observed so far only in pure microbial cultures.
Introduction
The removal of many upstream anthropogenic pollutants (industrial, urban and agricultural) prior to entering downstream coastal and marine ecosystems relies on wetland microbes. Out of those pollutants, nitrogen (N) pollutants (exceeding 60 Tg N y-1) are of major concern because they are linked to water quality degradation, eutrophication and increased greenhouse emissions (carbon dioxide (CO2), methane (CH4), and nitrous oxide (N2O)) (Anderson et al., 2010;Tian et al., 2015). Wetland microbes reduce the N load through various biochemical pathways i.e. assimilation, reduction and oxidation, but often this ecosystem service comes with the release of significant amounts of CO2, N2O and CH4.
Microbial processes of nitrification, denitrification or nitrifier-denitrification, methanogenesis and methanotrophy are responsible for the production of N2O and CH4 in wetland sentiments (Soares et al.,2016). In particular N2O is a very potent greenhouse gas, with a global warming potential ∼ 300 and ∼ 60 greater than CO2 and CH4, respectively, with important implications in ozone depletion. N2O can be further reduced to inert N2 by microbes that possess the only known enzyme to reduce N2O; the nitrous oxide reductase (NOS)(Thomson et al., 2012). It is estimated that ∼ 17 Tg N p.a. is returned to the atmosphere via denitrification in wetlands and terrestrial ecosystems globally, releasing ∼ 48 Tg N p.a. to coastal and marine ecosystems (Schlesinger, 2009). Besides N cycling, wetlands are important ecosystems in C cycling as well. Despite their relative small coverage (∼ 10 % globally), wetlands store large amounts of C and emit considerable amounts of CH4 (144 Tg CH4 p.a.)(IPPC, 2014). CH4 emissions are the trade-off balance between the microbial processes of methanogenesis and methanotrophy.
The production and consumption of N2O and CH4 is catalyzed by oxy-reductases that utilize a metal cofactor (Mo, Cu, Fe and S) (Glass and Orphan, 2012). The reduction of NO3’ is catalyzed by a periplasmic or membrane bound molybdenum (Mo) nitrate reductase (NAR) (Schwarz et al., 2009), NO2- is reduced to NO by either a non-metal cytochrome cdi (Einsle et al., 1999) or Cu-cofactor (Adman et al., 1995) nitrite reductase (NIR; expressed by nirS and nirK respectively). NO is reduced to N2O by nitric oxide reductase (NOR) and contains heme-Fe co-factor; two types of NOR are distinguished, cNOR which is found in denitrifying bacteria and receives electrons from cytochrome c and qNOR which is typically found in pathogenic non-denitrifying bacteria as well as in denitrifiers and receives electrons directly from quinol (Shiro, 2012). Finally N2O is reduced by the Cu-containing NOS (Rosenzweig, 2000). NO2‘, besides being a denitrification intermediate, it is also a substrate for dissimilatory nitrite reduction to ammonium (DRNA) and this permutation has been linked to C/N (Kraft et al., 2014;Yoon et al., 2015;van den Berg et al., 2017). Likewise, denitrification may be linked or co-occur with other C cycling processes.
Interestingly, members of the NC1O phylum (e.g. M. oxyfero) are able to couple methane oxidation to denitrification (NO2--dependent methane oxidation (N-DAMO)) utilizing a Cu-dependent particulate methane monoxygenase (pMMO) and a Fe-rich cdi nitrite reductase (Deutzmann et al., 2014). While extensive research has tested the relative contribution of key environmental factors (pH, NO3-/NO2-, O2, C/N) in regulating anaerobic microbial processes and specifically denitrification (Thomson et al., 2012), we argue that the abundance and bio-availability of trace metals in the environment could considerably control greenhouse gas emissions in the C and N biochemical cycle.
Metals in the environment have been traditionally seen as unwanted pollutants. This is because trace metal accumulation is linked to toxicity and inhibition of ecosystem processes (Samanidou and Papadoyannis, 1992). Contrariwise, low levels of trace metals may limit crop production, and impede animal and human nutrition (Teklić et al., 2013). Concentrations exceeding mg.L-1 range for Cu, Mo, Fe, Zn and Pb severely inhibited denitrification in soils, sediments, water-bodies and waste waters (Labbé et al., 2003; Magalhães et al., 2007; Liu et al., 2016). The recent advances in our understanding of denitrification using model organisms demonstrated the complete reduction of NO3- to N2 requires trace metals at µM to pM levels. In fact, in vitro studies have shown that lack of Cu, Mo or Fe severely inhibits denitrification or methane cycling, due to the formation of non-functional enzymes typically lacking the respective metal co-factor. In soil bacteria P. denitrificans and Pseudomonas stutzeri Cu is required to express to a functional Cu-containing NOS dimer (Granger and Ward, 2003;Felgate et al., 2012;Black et al., 2016). In methanotrophs oxidizing CH4to methanol (CH3OH) by MMO, the switch over between a Cu-or Fe-dependent MMO is regulated by the environmental availability of Cu and Fe respectively (Murrell et al., 2000; Bollinger Jr, 2010). All three classes of methanogens utilize a common cobalt dependent Methyl-coenzyme M reductase (MCR; mcrA), whereas in the earlier steps of methanogenesis are catalyzed by Fe-, Ni-and Zn-dependent ferrodoxins and dehydrogenases (Glass and Orphan, 2012). Based on the above observations, few studies have optimized the trace metal concertation for denitrification in a culture medium (Hahnke et al., 2014) and suggested the optimal levels of trace metals in biological wastewater treatment plants (He et al., 2015;Wintsche et al., 2018).
Despite the great progress in metal ecotoxicity and metalloenzyme biochemistry, our understanding of trace-metal availability in ecosystem C and N biochemical cycles is generally limited to aquatic ecosystems primarily focusing on productivity, NH47NO3- and CO2 assimilation rates (Twining et al., 2007;Glass et al., 2012;Moore et al., 2013;Romero et al., 2013;Schoffman et al., 2016). Here, we investigated the effects of trace metal (μM) addition on the microbial biochemistry of wetland sediments focusing primarily on NO2-reduction and greenhouse gas kinetics. We applied molecular techniques (qPCR) to quantify changes in the microbial ecology of wetland sentiment in an effort to better understand the intrinsic link between microbial ecology and process rates. We hypothesized that (i) the bio-availabilitv of metals (Mo. Fe and Cu) at uM levels will increase denitrification rates and specifically Cu availability will reduce N7Q emissions. Also, considering C respiration, liia) CO2 emissions will be driven bv denitrification and liib) no effect on CHa emissions is expected.
Materials and Methods
Soil samples and experimental set-up
A combined (5 sites; 5-10 cm from surface) soil sample and soil water were collected from the tidal-fluvial point bar deposit of the Pamunkey River (N 37.557451, W −76.972521) using a PVC pipe (Ø 10 cm). Soil sediments had 25 ±4 %DW organic matter (OM), 79 ±6 % moisture, 0.18 ±0.05 g.crrf-3 bulk density, C/N ratio of 11 ±1.5 and pH 6.1 ±0.3. Soil water, at the collection time, contained 0.1 ±0.03 mM N03-, 0.1 ±0.02 mM N02-, 0.16 ±0.04 mM S04-, 0.4 ±0.1 mM Cl", and 0.3 ±0.1 mM NH4+, 0.11 ±0.01 mM K+, 0.26 ±0.1 mM Mg2+ and 0.29 ±0.1 mM Ca2+. Samples were kept cool (ice-cooler) during transport and stored at 4 °C overnight prior to processing. Any large stones and roots were removed, then 1000 g wet soil was homogenized with an additional 1500 ml of filtered soil-water (Cellulose, 10 (im, Millipore, USA). Soil slurry aliquots of 50 ml were transferred to 125 ml bottles (Wheaton, USA). Bottles were crimped with rubber septa and were kept for 3 days in dark (25 °C) to let any residual 02 to be consumed by basal respiration. Then, all treatments received 250 nl aliquots of 2 M KNO3 reaching a final concentration of 10 mM. A 400 nl volume of ammonium molybdate ((NH4)MO7O24; 1 mM), iron sulfate (FeS04; 18.5 mM) and copper sulfate (CuS04; 6.5 mM) stock solution were injected to the bottles, adding 4, 74 and 26 μM respectively, according to the treatment plan (Table 1). Bottles were vortexed briefly (30 s at 6000 rpm), flushed with N2 (130 mins) and incubated at 25 °C in dark. Throughout the incubation several gas and liquid samples were taken anaerobically (as indicated in the graphs). At 96 hours, bottles were un-crimped in an O2-free chamber, then 1 g of wet soil was taken anaerobically and was frozen instantly for further molecular analyses.
Analytical techniques
Gas samples of 5 ml were stored in 3 ml exetainer vials (Labco, UK), that were previously flushed (N2; 5 mins) and vacuumed. Gas samples were analyzed for N2O, CO2, CH4 and O2 concentration with gas chromoatography (GC -2014 equipped with FID, TCD and ECD; Shimadzu, USA) as in Morrissey and Franklin (2015). For each gas, total production was determined as the sum of the gas accumulated in the headspace and the gas dissolved in the slurry; the latter was estimated using the relationships described by Heincke and Kaupenjohann (1999). From each bottle, 1.5 ml slurry samples were collected, centrifuged (10000 x g for 5 mins) and the supernatant was diluted lOx, filtered (0.22 (im) and stored (−4 °C) prior to ion analysis. The concentration of NO3-, NO2-, SO42- and NH4+ was determined by ion-chromatography (ICS -5000+, Dionex, USA) according to the manufacturer’s instructions using columns Dionex lonPac AS17 (2 × 250 mm) and lonPac CS12 (A -5 nm, 3 × 150 mm) and a Dionex conductivity sensor. Soil water content and OM were determined gravimetrically, soil pH was determined with a glass electrode pH probe (Laquact; Horiba, USA) in soil-water extract (after centrifugation 4000 x g for 20 mins).
Molecular techniques
DNA was extracted from frozen soil samples (0.3 g) using the DNEasy Kit from Qiagen (USA) following the manufacturer’s instructions. The DNA yields were, on average, 42 ±2 ng.nl-1. The retrieved DNA was quantified using a Nanodrop spectrometer (Thermo, USA) and visualized on a 1.2 % agarose gel for integrity. DNA samples were diluted to 3 ng.μl-1 and were further tested (see qPCR protocols on Table 2) for the abundance of microbial groups typically found in wetlands: total eubacteria (eub-518), nitrite reducers (denitrification: nirK, nirS and DNRA: nrfA), nitrous oxide reducers (nosZ) and methanogens (mcrA). SensiFAST™ SYBR® No-ROX Kit (Bioline, UK) polymerase x2 mix was used according to the manufacturer’s instructions and real time data were collected using a Bio-rad CFX-384 (Bio-rad, USA) real time PCR engine.
Statistical analyses
Kinetic and molecular datasets were tested for normality (RStudio [base]-, Shapiro.test). Non-parametric statistical tests were applied to assess any differences in the mean of ranks between the treatments (RStudio [agricolae]-, Kruskal-Wallis with Bonferroni correction). Basic statistics (mean and standard error) were summarized with RStudio [dyplr] and plotted with RStudio [ggplot2]. For all statistical tests p<0.05 was considered significant.
Results
The effects of trace metal addition on NO3‘, NO2‘ and NH4+ kinetics
At the end of the incubation, all treatments had visible gas bubbles on the surface and within the slurry, indicating microbial activity and gas production. We observed yellow and black patches within the slurry possibly due to Fe and Mn reduction indicating anaerobic redusive conditions. Denitrfication commenced with out any lag between the control and trace metal additions. Approximately, 12 mM N03‘ were consumed and 0.3 mM NO2- remained, while NH4+ ranged from 200 to 750 μM at the end of the incubation (Figure 1). At 48 h we observed a reduction in the remaining NO2-and a simultaneous increase in the NH4+ concentration for all treatments. The remaining NO3- and NO2- at the end of the incubation did not differ significantly when compared to the control. The treatments receiving additonal Mo + Fe (p<0.01) and Mo + Fe + Cu (p<0.01) had significantly higher NH4+ than the control, and an increasing trend in NH4+ concentration for all treatments was observed towards the end of the incubation, perhaps due to OM mineralization.
The effects of trace metal addition on CO2, CH4and N2O cumulative gas kinetics
Cumulative CO2 production increased rapidly between 12 to 24 h, and continued to increase up to ∼ 24 mM C-C02 at the end of the incubation. There were no significant differences between the control and each metal treatment, except the Cu and the Mo + Fe treatment (Figure 2.a). A faster rate of CH4 accumulation in the microcosms was observed between 6 to 12 h of incubation, earlier than C02. CH4 continued to accumulate during the incubation and reached levels ranging from ∼ 4 to 6.5 mM C-CH4. At the end of the incubation, we observed that the Cu treatment had significantly accumulated more CH4 than the control treatment (Figure 2.b). Cumulative N2O emissions rapidly increased between 24 to 48 h, matching the increased NO3- consumption. Interestingly, the control, Mo, Fe and Mo + Fe treatment accumulated significantly more N2O than the remaining treatments that contained Cu (Cu, Fe + Cu, Mo + Cu and Mo + Fe + Cu). These differences remained until the end of the incubation (Figure 2.c). Contrary to CO2 and CH4 emissions, N2O emissions reached a plateau between 48 to 96 h, indicating that N2O emissions were limited by the exhaustion of the available NO3- and not by C.
The effects of trace metal addition on targeted microbial groups
Average total DNA was ∼ 52 ng.ulμ.g-1 sediment with the Mo + Fe + Cu treatment having significantly (p=0.03) greater DNA yield (62 ±2 SE) that the control (44 ±5 SE). The average bacterial abundance was 6.75E+8 eub copies g1 sediment with the Fe, Fe + Cu and Mo + Cu treatments having significantly (p=0.005) more eub gene copies than the control (Figure 3.a). The microbial group of methanogens (mcrA) was relatively more abundant in the treatments containing additional Mo (Mo, Mo + Cu, Mo + Fe and Mo + Fe + Cu; Figure 3.c) when compared to the control. Regarding the microbial groups involved in N cycling we observed no effect of trace metals on DNRA N02- reducers (nrfA; Figure 3.b). Interestingly, trace metal addition affected the abundance of nirK (p<0.001; Figure 3.d), nirS (p<0.001; Figure 3.e) and nosZ (p=0.001; Figure 3.f) denitrifying microbial groups. N02- reducers utilizing Cu-NIR (nirK) were more abundant in all treatments when compared to the control, except when Fe was added. Microbial groups having a cytochrome cdi-NIR (n/-rS) were relatively more abundant in the Mo and Fe + Cu treatment. Lastly, the nosZ microbial group responsible for the reduction of N20 to N2 was more abundant in the treatments Fe+Cu, Mo+Cu and Mo+Fe+Cu than the control. Although nosZ abundance was relatively higher in all trace metal additions, the effects of single Cu addition on nosZ did not differ significantly.
Discussion
We observed that trace metal additions (Mo, Fe, Cu and their combinations) in wetland sediments regulated denitrification kinetics and greenhouse gas emissions, and altered the abundance of targeted microbial groups that are typically found in wetlands. The key interpretation of the incubation experiments is that trace metal availability and specifically Cu, enhances the reduction of N2O to N2, the production of CH4 and has no effect on CO2 emissions from wetland sentiments. The increase in the denitrifiers’ associated physiological activity to reduce N2O when supplied with additional Cu corresponded to the relative higher abundance of Cu-containing genes -nitrous oxide (nosZ) and nititrite (nirK) reductase, but not to the abundance of nitrite reduces nirS and nrfA, suggesting a direct relationship between Cu availiabilty and denitrification process in soils.
The absence of substantial amounts of NO2 and N2O in the bottles for the Cu-containing treatments (Cu, Fe + Cu, Mo + Cu and Mo + Fe + Cu) and an increase in the respective gene abundance indicates that a balanced denitrification process NO3- to N2. Microbes with a complete denitrification strategy harvest the energy from the last reduction step (N2O to N2) in addition to the rest of the denitrification steps, which may be particularly advantageous when there are limited electron acceptors available in the environment. We do not exclude other partial denitrification processes or other anaerobic processes simultaneously co-occurring with denitrification as subjected by the residual N20 in the treatments Cu-containing treatments (<2.3 mM N-N20). The increasing NH4+ concentration (<1 mM) towards the end of the incubation could indicate DNRA and /or N mineralization processes commonly co-occurring in wetland sediments (White and Reddy, 2009). However, the lack of significant differences in DNRA gene abundance indicates towards anaerobic N mineralization. Although we didn’t measure NO in the treatments, we assume that it was quickly utilized due to its potent cytotoxicity at nM levels (Chaudhari et al., 2017). In the treatments accumulating notable amounts of N20 (> ImM) N2O cytotoxicity may occur and it can be releaved by the addition of vitamin B12, Co or methionine (Sullivan et al., 2013). The continuous accumulation of CO2 in all treatments, as proxy for microbial activity, firstly, excludes any indirect acute cytotoxicity from NO due to NO2- reduction or N2O accumulation and secondly idicates that the availible electron donor (organic C) was in excess of the added NO3-. The magnitute of CO2 accumulation was predictable; treatments with complete dentrification, thus having less residual N2O had accumulated more CO2 by the end of the incubation. We assumed that the lack of additional Ni would inhibit MCR, key enzyme in methanogenesis (see Possible mechanisms and other regulators). Contrary to our hypothesis iib, trace metals had a significant effect on CH4 emissions.
Other studies and trace metal levels
Several studies have looked into the effects of trace metal abundance or addition on C and N cycling in sediments and peat soils. Typically those studies report a negative effect of trace metals on biogeochemical processes, however the range of trace metal concentration far exceed the levels of the current study. For instance, Keller and Wade amended peat slurries with ∼ 100 nM trace element solution and saw significant decrease in CH4 emission, probably due to Cu induced toxicity. Similarly, Magalhaes et al. (2007) observed an 85% inhibition of denitrification accompanied with accumulation of substantial N20 and N02’ amounts at 79 Cu ng.g-1 sediment (equiv. 1.2 mM Cu) in estuaries. In a follow-up study, lower denitrification rates due to Cu (∼ 0.9 mM) were accompanied with a decline in nirK, nirS and nosZ microbial group abundance and 0-diversity (Magalhães et al., 2011). A more recent study that tested the effects of Cu pollution (addition of 100 μ ng.L-1 ∼ 1.6 nM) on urban wetland sediments, found that Cu addition significantly reduced CH4 emissions, but had no effect on N2O emissions when compared to the control (freshwater sediments; Doroski et al. (2019)). This divergent observation could be due to low NO3- concentration (18 nM) in relation to available C or the presence of substantial amounts of soil Cu (9.4 ng.g-1 sediment) at the start of the experiment in both control and Cu treatment.
Possible mechanisms
Denitrification is an anaerboic respiratory process that utilizes NO3- as a terminal electron acceptor (Equation (2)) accompanied by energy transfer (Thomson et al., 2012). Thus, for a complete denitrification reaction, two molecules of NO3- will be converted to one molecule of N2 with the transferring of 10 electrons to the electron recipient (NO3-, N02- NO and N2O). Therefore, microbial groups unable to utlize the complete denitrification pathway will harvest less energy, and/or will have to compensate the energy loss by adjusting their metabolic rates for example by enhancing the rate of NO3- reduction. Such bioenergetic adjustment was observed in chemostat cultures of P. denitrificans; the nosZ deficient strain was unable to complete the last step of denitrification and had lower biomass and higher NO3- reduction rates (Giannopoulos et al., 2017). At RNA level, Cu-decifient cells had more transcripts of nosZ and Cu-scavenging genes to compensate for the loss of N20 reduction due to depleted Cu availiability (Felgate et al., 2012). And finally, strains deficient in Cu -transporters and chaperones were unable to reduce N2O to N2, although nosZ was expressed, forming a non-functional NOR (Sullivan et al., 2013). Therefore we postulate that, at community level, microbes utilizing the complete dentrification pathway will have a physiological advantage than those unbale to do so. This was supported by the increased abundances of N02- reducers (nirK) and partly of the N2O reducers (nosZ) and of the overal microbial abundance (eub); corroborating the fact that denitrifcation genes were up-regulated in single denitrifiers (P.stutezi at 0.05 Cu mg.L-1; Black et al. (2016)).
Considering net CH4 emissions, Mo clearly enhanced CH4 accumulation (Mo) and abundance of mcrA genes in the Mo containing treatments. A Mo co-factor is typically required for the formylmethanofuran dehydrogenase (fwd) that catalyzes the reduction of CO2 to formyl-methanofuran in the first step of methanogenesis by reduction of CO2 with electrons from H2 (Glass and Orphan, 2012). Mo availioability and the increasing partial pressure of CO2 in the bottles during the incubation could have triggered a shift towards a Mo-based methanogenesis pathway. Another interesting aspect in CH4 emissions and Cu, is the competition between methanotrophs and denitrifiers for the “ Cu monopoly” (Chang et al., 2018), which should be further investigated in environmental samples. Eventhough the purpose of this study wasn’t to quantify the relative contribution between methanotrophs (MMO) and methanogens (MCR) in the net CH4 emissions, the higher abundance of methanogens (mcrA) and noticable accumulation of CH4 in all metal treatments indicates the important role of trace metal availiability in CH4 cycling in environmental samples.
Trace metal limitation in the environment and other regulators
The abundance of bio-availiable metals in the environment, typically in the ion form, is often observed at several magnitute orders less compared to the routinely reported total metal content. Such trivial amounts could be unavailiable to microorganisms due to physiochemical interactions and plant uptake (Giller et al., 1998). At neutral and alkaline pH levels, metals tend to be effectively immobilized as inorganic compounds (metal – oxides, – hydroxides and – carbonates). Additionally, organic complexes such humic ligands may bind metal cations making them unavailiable to microbes and plants. For example, the particularly low concentrations of dissolved Cu, Fe and Mo (∼ 5, 1 and 0.2 μ g.L-1, respectively) and the copious dissolved organic C (∼ 80 mg.L-1) in Siberian peatlands (Raudina et al., 2017), could trigger significant emissions of CH4 and N20 as our climate is warming due to metal limitations (Basiliko and Yavitt, 2001;Voigt et al., 2017). On the other hand, in temperate salt marshes with reducing conditions, S04’ may suppress CH4 emissions, but the abundant H2S may reduce the availiability of trace metals with important implications in N cycling and N2O emissions (Gauci et al., 2004;Butterbach-Bahl et al., 2013). In this study we added 26 μM and 74 μM of SO4‘ as CuSO4 and FeSO4 respectively that could potentially be reduced to H2S, however the initial and final SO4- concentration remained at comperable levels (Table 4).
Other well known environmental factors controlling denitrification is O2, NH4+, NO3-/NO2- and C (Thomson et al., 2012). Synthesizing the findings from other chemostat (Baumann et al., 1996;Felgate et al., 2012;Giannopoulos et al., 2017;Hartop et al., 2017;Conthe et al., 2018) and incubation (Koike and Hattori, 1975;Okereke, 1993;Morley et al., 2008) studies, it appears that the relative importance of the aforementioned factors as control of denitrification is O2> NH4, NO3/NO2> C/N or C quality, with trace metal bio-availability exerting an additional effect on each factor. In this study we did not indent to quantify the relative contribution of each environmental factor on denitrification. Instead we demonstrated the importance of trace metal availability on ecosystem processes contributing to the release of N2O, CH4 and CO2.
Future challenges
Trace metal limitition in denitrification biochemistry and in general the C and N cycle may also be important in agricultural ecosystems besides wetlands. The continous fertilization with macronutrients (NPK) and intentisive cropping has resulted in low concentrations of microelements (trace metals) in soils and edible crops (hidden hunger; (Alloway, 2008); Teklić et al. (2013)). The additon of essential micronutrients particularly Cu, Mo and Fe to agricultural lands has been previously suggested as a possible strategy to mitigate N20 emissions (Richardson et al., 2009), but few assessments have been reported. Felgate et al. (2012) and Black et al. (2016) recomended a Cu concentration of 13 μM and 50 μM (equiv. 150-200 μg.g-1 Cu in soils) based on studies with P. denitrificans and Ps. Stutezi respectively. Building upon prior knowledge, we demonstrate that effects of metal addition (2.5 Cu, 26.7 Fe and 7.8 Mo μg.g-1 sediment) on greenhouse gas emissions in environmental samples (sediments) are due to subsequent changes in the microbial community structure. We are currently missing quantitative data on optimal levels of soluble trace metals for biochemical processes in C and N cycling in a variety of different soil types. Agricultural and flooded soils are particularly interesting because they tend to have higher N2O yields than marine and aquatic ecosystems (Beaulieu et al., 2011). Future studies should investigate the effects of trace metal bio-fertilization in soils under commercial crops to assess any beneficial effects in greenhouse gas mitigation and crop quality, as in Montoya et al. (2018). The type of trace metal fertilizer is particularly interesting to investigate because chelators based fertilizers may scavenge the metal co-factors and dissable key enzymes of the N cycle (urease, ammonia monooxygenase and nitrite reductase) prior to N2O reduction (Moffett et al., 2012;Montoya et al.,2018).
Assessing the abundance of microbial groups using qPCR showed significant differences between the control and the treatments. In this short term batch experiment, differences in microbial group abundance could also illustrate changes in in the diversity of the microbial groups, for example between the abundance of copiotrophic and / or fast growing (r strategists) microbes upon NO3- and trace metal addtion, relative to slow growing (k strategists) microbes (Fierer et al., 2012). Further information on microbial diversity and meta-transcriptomic profiles under comperative treatments is needed to understand and evaluate the ecological importance of trace metal limitation. Such’omic studies should be accompanied by high-throughput metabolite or process rate analysis either in the lab or in the field.
Conlcusion
Emerging from this investigation, our results suggest that trace metal bio-availability may, directly or indirectly, regulate or co-limit denitrification kinetics and induce a shift in the microbial community in environmental samples, favoring a more active microbial community that is able to quickly acquire the bio-available metals and incorporate them in functional oxy-reductases. This observation has important implications in wetland C and N cycling. Considering also agro-ecosystems that are becoming limited in trace metals trace metal amendments should be evaluated and considered as an alternative way to mitigate N2O emissions and improve the nutritional quality of crops.
Acknowledgements
I wish to thank Dr loannis Ipsilantis for his comments on an earlier version of this manuscript.