Iron-dependent AOM

blobs

New research using sediments from a Dutch canal has yielded important insights into the anaerobic oxidation of methane (AOM).  A group of scientists from the Institute for Water and Wetland Research at Radboud University identified archaea of the order Methanosarcinales as being responsible for AOM coupled to iron and manganese reduction.

AOM proceeds as methane is oxidised with various terminal electron acceptors including sulphate, nitrite and nitrate, and the importance of oxidised metals in the process has been investigated, but the microorganisms involved have remained unknown.  To address this, a culture of methanotrophs was established that was enriched with canal sediment.  When this culture was supplied with nitrate as the only available electron acceptor, it became dominated by AOM-associated archaea (AAA).  This culture linked methane oxidation to nitrate reduction, with N2 being the main end product, although approximately 10% of nitrate was converted to ammonium.

The enrichment culture with AAA was used for further experiments, where methane oxidation was observed following the addition of ferric citrate, nanoparticulate ferrihydrite (Fe3+) or birnessite (Mn4+).  Analysis of the AAA genome showed that it could couple methane oxidation to nitrate reduction or Fe3+ reduction, or that the reverse methanogenesis pathway could also operate.  The authors therefore suggest that AAA could act as a versatile methanotroph, switching electron acceptors depending on their availability.

The paper was jointly led by Katharina Ettwig and Baoli Zhu. One of the co-authors, Boran Kartaldescribed the possible application of their findings: “A bioreactor containing anaerobic methane and ammonium oxidizing microorganisms can be used to simultaneously convert ammonium, methane and oxidized nitrogen in wastewater into harmless nitrogen gas and carbon dioxide, which has much lower global warming potential.”  The authors conclude that their work “may also shed light on the long-standing discussion about Fe2+ -producing processes on early Earth, when AAA-related organisms may have thrived under the methane-rich atmosphere in the ferruginous Archean oceans.”

Ettwig, K.F., Zhu, B., Speth, D., Keltjens, J.T., Jetten, M.S.M., Kartal, B. 2016. Archaea catalyse iron-dependent anaerobic oxidation of methane. PNAS, doi: 10.1073/pnas.1609534113.

Image is from the paper and shows fluorescence in situ hybridization of biomass from the enrichment culture of AAA and M. oxyfera-like bacteria.

Advertisement

Rising atmospheric methane and its isotopic shifta

gbc20468-fig-0004

A paper published in Global Biogeochemical Cycles has clarified the importance of different sources contributing to the atmospheric methane burden during 2007-2014. The research used measurements taken by the USA National Oceanic and Atmospheric Administration (NOAA) Cooperative Global Air Sampling Network, and investigated changes in mole fractions and isotopic composition.

The results showed that the globally averaged mole fraction of CH4 increased by 5.7 ± 1.2 ppb yr-1 up to 2013, then by 12.5 ± 0.4 ppb in 2014. Accompanying this change was a shift in δ13CCH4 to more negative values. At the remote monitoring station at Ascension Island, this shift was −0.24 ± 0.02‰, but the same trend was observed globally.  These global changes will potentially be driven by varying processes in different regions, which the authors attempt to untangle.

In 2007 there was an increase in CH4 concentrations at high northern latitudes, which was caused by an increase in late summer emissions from wetlands, stimulated by higher temperatures. However, during 2008-2013 this Arctic increase fell below global means, indicating that no new CH4 emissions developed during this time.  In 2014, strong increases commenced again, but this time in line with global means. For the entire period, isotopic data shows a trend to more depleted δ13CCH4, indicative of wetland sources.

For Ascension Island, methane increases have been sustained for the duration of monitoring. Assuming a linear trend, isotopic change is in a negative direction and of approximately −0.03‰ yr−1. Measurements from Cape Point (South Africa) and the South Pole were similar to the Ascension ones. The authors point out that the southern Amazon is a distant source of air for Ascension Island.  They therefore posit that increased rainfall and warmth may have been responsible for increased CH4 emissions from tropical wetlands, as recent years have seen flood levels above long-term averages. The paper also emphasise the importance of emissions from ruminants in the tropics, but point to a lack of data regarding the δ13CCH4 values of such emissions.

Lead author, Professor Euan Nisbet at Royal Holloway University of London said: “Our results go against conventional thinking that the recent increase in atmospheric methane must be caused by increased emissions from natural gas, oil, and coal production. Our analysis of methane’s isotopic composition clearly points to increased emissions from microbial sources, such as wetlands or agriculture.”

To conclude, the authors stress the magnitude of the global increase in methane concentrations; 60 ppb in nine years. They write: “if methane growth continues, and is indeed driven by biogenic emissions, the present increase is already becoming exceptional, beyond the largest events in the last millennium.”

Reference. Nisbet, E.G., Dlugokeencky, E.J., Manning, M.R., et al. Rising atmospheric methane: 2007-2014 growth and isotopic shift. 2016. Global Biogeochemical Cycles,DOI: 10.1002/2016GB005406.

Image is from the above paper, under a Creative Commons Attribution 4.0 International License.

Deep microbial communities created through fracking

nmicrobiol2016146-f4

A new study published in Nature Microbiology has shed light on the microbial communities inhabiting hydraulically fractured shale, at a depth of approximately 2.5 km.  The process of hydraulic fracturing relies on the high-pressure injection of water and various chemicals deep underground.

The investigation took place in two Appalachian basin shales in the USA, and involved metagenomic and metabolite analyses on input fluids to the well, as well as analyses on produced fluids over a period of 328 days.  This approach enabled changes in microbial communities and metabolites to be observed over time.

The analyses of the Marcellus shale showed that changes in microbial community corresponded to increases in salinity, and persisting halotolerant and thermotolerant members were found in various bacterial and archaeal taxa.  Additionally, the authors discovered one apparently unique bacteria and proposed the genus name Candidatus Frackibacter.  Over time, there was an increase in the concentration of glycine betaine (GB), which is used by microbes to survive osmotic stress, as well as evidence of uptake and de novo synthesis of GB by microbes.  GB could be degraded by obligate fermenters from the genera Halanaerobium or Candidatus Frackibacter which would produce trimethylamine (TMA).  In turn, it was hypothesised that TMA would be used as a methanogenic substrate by organisms in the genera Methanolous and Methanohalophilus.

The second shale (Utica) investigated was geologically and geographically distinct from the Marcellus shale.  The authors experimentally amended produced fluids from the Utica shale with GB, which resulted in enrichment of Methanohalophilus and Halanaerobium.  TMA production was detected, and the amended samples produced 6.5 times more methane per day when compared to controls that were not amended with GB.  The genomes of Methanohalophilus and Halanaerobium from the Marcellus and Utica shales were closely related, demonstrating that even though the two ecosystems were different, similarities arose in the microbial communities.

Kelly Wrighton, last author on the paper, said: “We think that the microbes in each well may form a self-sustaining ecosystem where they provide their own food sources.  Drilling the well and pumping in fracturing fluid creates the ecosystem, but the microbes adapt to their new environment in a way to sustain the system over long periods.”

Much of the previous research on hydraulic fracturing has focussed on economics and environmental impacts.  However, this new research shows that hydraulic fracturing also creates the necessary physical and chemical conditions for microbial life to persist, and that much of this is implicated in methane cycling.

Daly, R.A., Borton, M.A., Wilkins, M.J. et al. 2016. Microbial metabolisms in a 2.5km deep ecosystem created by hydraulic fracturing in shales. Nature Microbiology, 1, article number 16146.

Image is from the above paper, under a Creative Commons Attribution 4.0 International License.

EGU 2016 summary – forests and methane

snowforest

The European Geosciences Union is one of the largest scientific conferences in the world and between the 17th and 22nd April this year saw 13,650 scientists from 109 countries descend upon Vienna for a week of cutting edge science. This year for the first time there was a session on ‘Forests and the Methane Cycle’ chaired by MethaneNet’s director, Vincent Gauci, reflecting the rapid growth in interest in the field.

The first talk was by the solicited speaker Patrick Megonigal of the Smithsonian Environmental Research Centre, discussing research done by his lab and collaborators in China into upland forests and their contribution to the global methane sink. At their sites in Maryland, USA, they found that warm and wet soils resulted in high stem methane fluxes which offset 5% of the annual ecosystem sink at one site and 3% at a more westerly site. Stem fluxes also showed diurnal patterns, with fluxes increasing during the daytime and decreasing at night which shows the importance of considering plant metabolism and physiology as a control. The Chinese upland site had higher methane fluxes than their US sites but did receive higher precipitation and warmer temperatures meaning the stem fluxes had a far more significant effect at an ecosystem scale. The key message from Megonigal is that the global methane sink has been overestimated by 3-60%.

Vincent Gauci, of the Open University, then presented the results of a major sampling campaign in the Brazilian Amazon on behalf of Sunitha Pangala who was unable to attend. The campaign in flooded forests along the edges of the Amazon sampled over 2,400 trees at three heights across 13 sites and found substantial tree stem fluxes at all of them. The stem fluxes dominated at each site, accounting for 70-80% of ecosystem fluxes, with young stems being the largest single source of methane.

The focus then switched from living trees to what happens after they die. Kris Covey, a PhD student at the Yale School of Forestry & Environmental Studies has been studying cores of deadwood from hundreds of sites across the USA to look at two important greenhouse gases: methane and nitrous oxide. Deadwood methane production declines with age however still produces above ambient levels, highlighting that detritus and deadwood needs to be factored into the ecosystem models. Nitrous oxide was uniformly consumed within deadwood which again could be important at ecosystem and regional scales.

Kateřina Macháčová of the Global Change Research Institute CAS in Brno, Czech Republic, presented a summary of her group’s work into methane fluxes in boreal forests. They compared three common species: Scots pine, Norway spruce and silver birch. Birch stem fluxes were the greatest of the species studied and likely due to differing physiology to the coniferous species. She also reported seasonal variation in fluxes with two distinct pulses of methane release: the first in February (likely due to snow melt) and the second during the growing season of May to October.

Keeping with the boreal theme, Mari Philatie from the University of Helsinki, presented the results of an experiment attempting to constrain the size of fluxes from component parts of the forest (forest floor, lower trunk, mid trunk and shoots/leaves). Using LiDAR to examine microtopographic features they were able to correlate topography and soil water content to the methane fluxes measured. They also noted that fluxes from tree shoots did not show seasonal variation, nor were correlated with other physical features measured. As with Machacova’s study they also found that birch at their site produced more methane than spruce.

Taken together, the research presented during this exciting session shows the growing interest in the role that trees and their stems and shoots play in the global cycling of greenhouse gases. Hopefully the session will act as a springboard for future research directions and collaborations.

————————–

Written by Bertie Welch. Bertie is a PhD student at the Open University, researching emissions of methane and nitrous oxide from temperate and tropical forests. He tweets as @WelchEcology and has blogged for MethaneNet previously.

EGU 2016 – forests and the methane cycle

parc

The European Geosciences Union (EGU) General Assembly 2016 in Vienna will shortly be underway (17th-22nd April). Session BG2.10, “forests and the methane cycle” will be convened by MethaneNet’s director, Vincent Gauci of the Open University, UK. Co-conveners are Mari Pihlatie of the University of Helsinki, Katerina Machacova of the Global Change Research Institute in Czech Republic, and Sunitha Pangala and Karen Olsson-Francis, both of the Open University. The session abstract is:

Cycling of methane in terrestrial ecosystems has traditionally focused on exchanges between wetlands with relatively low herbaceous or shrubby canopies and the atmosphere. New findings are now suggesting that forests occupying permanently and seasonally inundated soils may be important conduits of soil-produced methane to the atmosphere, and trees occupying drier soils may also be playing a role in determining net exchanges of this powerful greenhouse gas. Forests are also vulnerable to fire which itself is a poorly quantified source of methane. This session seeks to bring together researchers working on the exchange of methane in forest ecosystems at any relevant scale and from the full climatic and hydrological forest range. We therefore welcome contributions on microbial processes in soils, plant tissues and microtopographic forms, measurement campaigns on the forest floor, on tree stems and at the leaf and canopy level; micrometeorological measurements using flux towers, and satellite, inverse and numerical modelling studies that seek to integrate our understanding of methane exchange in these ecosystems.

Methane emissions from upland boreal forest soils

talltrees

Large uncertainties still exist in the magnitude of the sources and sinks that contribute to the global methane budget.  Boreal landscapes typically contain a mixture of forests, wetlands and peatlands, and the total catchment flux of methane will be determined by the source/sink behaviour of each ecosystem.  Upland forests are typically considered to be net sinks of methane, although periods of methane release have been observed after heavy rainfall.  Contrary to this, peatlands and wetlands are well known to emit large quantities of methane.  However, there is a still a lack of knowledge concerning how inter-annual variations in climate can affect the total catchment methane budget.

To address this knowledge gap, a group of researchers led by the Finnish Meteorological Institute conducted a twenty-eight month study in the Pallaslompolo catchment in northern Finland.  Methane fluxes were measured from a minerotrophic fen and a nearby upland forest.  The forest soil was observed to be a small methane sink (-0.18 to -2.3 mg CH4 m-2 d-1) from September 2010 to August 2011.  Following three months of heavy rainfall, the forest then became a large methane source until January 2012 (max = 92 mg m-2 d-1).  After this spike, the forest returned to acting as a methane sink for the remainder of the study (which ran until January 2013).

When upscaling flux measurements to the entire catchment, the forest consumed the equivalent of 10% of the methane emitted by the fen in the dry year.  During the wet year when the large forest methane spike was observed, the forest was a source equivalent to 57% of fen emissions. During August and September 2011, monthly methane fluxes from the forest were twice as large as the emissions from the fen.  The authors hypothesise that high forest methane emissions occur during wet conditions, but also require other factors which include: 1) high soil temperature; 2) an input of soil carbon from roots/litter; 3) high methanogenic activity.  This explains why high forest fluxes where not observed during spring snowmelt when soil moisture values peaked.

A signal of the high forest emission was also detected in the atmospheric methane concentration measured nearby. In the long-term data, the mean September concentration anomaly was well explained by the water level of the nearby lake, suggesting that the water level could act as a better proxy for soil methanogenic potential than a point measurement of soil moisture.

This paper shows that methane fluxes in boreal forests can show extreme variation between years according to differences in precipitation.  It demonstrates the importance of taking frequent, year-round flux measurements. Lead author, Annalea Lohila said: “We were surprised to see how differently the forest and wetland ecosystems responded to excess wetness, and how important the upland forests are for methane balances after upscaling to the landscape level.”

Lohila, A., Aalto, T., Aurela, M., et al. 2016. Large contribution of boreal upland forest soils to a catchment-scale CH4 balance in a wet year. Geophysical Research Letters, DOI: 10.1002/2016GL067718.

Forests and the Methane Cycle Workshop

????????????????????????????????????

During the first week of June a workshop took place covering the topic of “Forests and Methane Cycling.”  The meeting was jointly organised by MethaneNet and Mari Pihlatie of the University of Helsinki, and was held on the idyllic banks of Lake Kuivajärvi at Hyytiälä Forestry Field Station in Finland.  The aim of the event was to synthesise current knowledge on forest methane cycling from various ecoregions, and at different spatial and temporal scales.

A general introduction set the scene for the workshop, but also raised important questions concerning methods for upscaling fluxes, and the appropriateness of different metrics in reporting data.  After this it was on to individual presentations, beginning with a boreal focus.  Speakers discussed both micrometeorological and chamber approaches to measuring methane in northern coniferous forests, which allow complex methane dynamics to be elucidated.  For example, at Hyytiälä there is evidence that suggests the forest floor is a methane sink, whilst the trees act as a small source; similar results from trees have been reported elsewhere.  This introduced the question: “are trees the missing source of methane in boreal forests?”

For the afternoon session, the science switched from the boreal zone to the tropics.  It was noted that carbon cycling in tropical wetlands is still poorly understood.  Furthermore, the possibility was presented that unquantified methane emissions from tree stems might be the reason for the regional discrepancy between bottom-up and top-down estimates of atmospheric methane sources.  The mediating role of tree physiology on greenhouse gas emissions was also broached.

Day 2 of the workshop was a more interactive affair, with conversations about knowledge gaps and the future direction of forest methane research.  There was time for two field trips, however.  The first was to SMEAR II; an atmospheric research facility where measurements of greenhouse gases from soil and trees are ongoing using chamber methods.  Greenhouse gases are also measured using micrometeorological methods on a 124 m tower.  Our second trip was to the scenic Siikaneva peatland, where greenhouse gases are measured using chamber and eddy covariance methods.

The workshop organizer Dr Mari Pihlatie of the University of Helsinki said “this workshop was an eye opener to see how big uncertainties we still have in understanding the role of trees in methane dynamics in forest ecosystems. Also, the meeting gave an excellent opportunity to initialize collaboration between research groups and disciplines to work towards a comprehensive understanding of the methane cycling in forest ecosystems.” Dr Vincent Gauci of MethaneNet added: “the workshop highlighted that we are at a relatively early stage of developing fundamental knowledge of how forested wetlands and upland ecosystems participate in the methane cycle.  The future of this field of research will have important implications for characterisation and modelling of ecosystem sources of this gas under a range of global change scenarios.”

The Uncertain Climate Footprint of Wetlands Under Human Pressure

Wicken flux tower

New research published in PNAS has attempted to untangle the differences in methane, carbon dioxide, and nitrous oxide fluxes from natural and managed wetlands. As MethaneNet readers will know, wetlands play an important role in global methane dynamics but many questions remain unanswered. The new study, which primarily used FLUXNET sites, analysed data collected using both static chamber and eddy covariance methods in temperate, boreal, and arctic locations. As such, the dataset is likely to be the first large-scale synthesis of wetland flux data generated by eddy covariance.

As is frequently observed for wetlands, most of the studied sites were net sources of methane, and net sinks of carbon dioxide. For temperate sites, it was found that drainage and conversion to agriculture reduced methane fluxes but increased carbon dioxide fluxes. Rewetting or restoring such sites then creates a methane source that is not offset by the associated carbon dioxide sink. To fully quantify the impacts of management, sites with annual greenhouse gas budgets (including nitrous oxide from agricultural sites) were used to calculate radiative forcing (RF) on a 100 year time scale. These calculations also accounted for carbon offtakes, such as forestry or crop harvesting. The calculations suggested that conversion of arctic and boreal wetlands to agriculture resulted in a positive RF. Temperate wetlands converted to agriculture generally had a positive RF, though there was variation due to the type and intensity of management.

In their conclusions, the international group of authors stress the importance of long-term monitoring to fully understand ecosystem responses to both natural and anthropogenic changes. They further add that such responses may not be adequately captured by manual static chamber sampling, and that focus should be directed towards eddy covariance and quasi-continuous chamber measurements. They conclude: “Our results prove that management intensity strongly influences the net climate footprint of wetlands and in particular the conversion of natural ecosystems to agricultural land ultimately leads to strong positive RF.” These results therefore have important policy implications regarding land management and climate.

Reference: Petrescu, A.M.R., Lohila, A., Tuovinen, J-P., et al. 2015. The uncertain climate footprint of wetlands under human pressure. PNAS, doi: 10.1073/pnas.1416267112.

Photo: Mike Peacock

New Forum to Support Inland Water GHG Flux Related Studies

elsi

A message from David Bastviken:

Dear all,

For your information The Inland Water Greenhouse Gas FLUX Forum (IWFLUX) was just made public and open at www.tema.liu.se/tema-m/iwflux

IWFLUX is a web based resource for and by the active community where we can jointly gather and share information we find useful to learn more about inland water greenhouse gas fluxes or related topics. At present, this is just a structure to be filled with content by us all – this grass root initiative was developed at a workshop on inland water greenhouse gas fluxes in Finland in September last year.

IWFLUX has:

(1) a number of webpages for established/referenced information expected to be reasonably stable over time,

(2) an open discussion forum for any IWFLUX related discussions/questions including aspects not yet established, and

(3) a possibility to join an e-mail list for IWFLUX related communication (if you want to join please follow the instructions on the Contact information page to sign up).

The format of all parts is deliberately kept simple to minimize maintenance work given that this is a non-funded initiative at present. I also quickly realized there is no perfect structure for all webpages so we now have a start that may develop based on upcoming needs.

IWFLUX is also a global effort on all GHGs where we can gather links to related networks that are regional or concern specific GHGs.

The original idea is not primarily to fill the site with papers or data…but to gather information on where to find key papers or data, and to provide extended information to help us coordinate methods and efforts for maximized future learning and impact. Very fundamental information is also welcome as IWFLUX could be an important resource for students and starters in the field. …but regardless of the original ideas, IWFLUX will develop over time and become whatever we want it to be depending on the contributions.

Contributions can be various types of files and internet links that will be posted in file archives. To make contributions please send them to the IWFLUX coordinator.

So please take a look, spread the news, and hope you will find it useful…and more so when we have filled the present IWFLUX structure with real content!

Many thanks to all who have so far contributed to develop IWFLUX!

All the best,

David

Blog: Trace Gas Dynamics in Tropical Forests

trop

By Bertie Welch. 18th December 2013.

After carbon dioxide, methane is the second most important greenhouse gas, yet how much is produced and transferred to the atmosphere by tropical rainforests is not fully understood, because much of the research in the tropics focuses on rice paddies and wetlands, as these are significant sources of anthropogenic methane. Rice et al. (2010) suggest that globally, hardwood trees could account for emissions of ~60 Tg Ch4 yr-1.

These potentially significant methane emissions are a result of trees allowing the methane produced in deeper, anaerobic soils to bypass the methanotrophs and nitrifying bacteria in the aerobic upper soils (Milich, 1999). We know from temperate forest and mesocosm studies that trees can act as conduits for soil methane into the atmosphere.  Trees can adapt to cope with flooding-induced soil anoxia in a variety of ways: hypertrophied lenticels, adventitious roots and enlarged aerenchyma (Havens, 1994; Kozlowksi, 1997).  Studies of temperate trees such as alders have shown that they use all the above adaptations when in waterlogged soils. This results in increased methane fluxes from the trees as there is increased surface area for emission and easier transmission from soil to atmosphere (Rusch and Rennenberg, 1998; Gauci et al., 2010).

Pangala et al. (2013) showed that in a tropical peat forest in Borneo, tree stems were responsible for 60-80% of total ecosystem methane fluxes, demonstrating that trees can be a significant source of emissions. This new study in lowland evergreen tropical rainforest in Panama aims to expand on this work and investigate the extent of tree stem emissions in an area of relatively free draining soils on a fortnightly basis for 5 months, covering the dry to wet season transition starting March 2014.

There is a second aspect to the study – we don’t know how trace greenhouse gas biosphere-atmosphere exchange will be affected by an atmosphere that is being enriched with CO2. It is thought that rising atmospheric CO2 concentrations will increase primary productivity in rainforests resulting in greater amounts of litterfall. The fieldwork will be done at the Smithsonian Tropical Research Institute, Panama, on plots that are part of an ongoing litter manipulation experiment meant to simulate elevated atmospheric CO2. There are 15 plots in total: 5 control, 5 with litter removed and 5 with litter added. Sayer et al. (2011) discovered soil respiration to be significantly higher in the litter addition plots compared to the control and litter removal plots. We hypothesise that due to increased litter input (and therefore more source carbon for methanogenesis) methane emissions will be greatest in the addition plots.

Gauci, V., Gowing, D.J.G., Hornibrook, E.R.C., Davis, J.M. & Dise, N.B. (2010) Woody stem methane emission in mature wetland alder trees. Atmospheric Environment, 44, 2157-2160.

Havens, K.J. (1994) The formation of hypertrophied lenticels, adventitious water roots, and an oxidized rhizosphere by Acer rubrum seedlings over time along a hydrologic gradient. Virginia Institute of Marine Science, Gloucester Point, Va.

Kozlowski, T.T. (1997) Responses of woody plants to flooding and salinity. Tree Physiology, Monograph No. 1, 29.

Milich, L. (1999) The role of methane in global warming: where might mitigation strategies be focused. Global Environmental Change, 9, 179-201.

Pangala, S.R., Moore, S., Hornibrook, E.R.C. & Gauci, V. (2013) Trees are major conduits for methane egress from tropical forested wetlands. New Phytologist, 197, 524-531.

Rice, A.L., Butenhoff, C.L., Shearer, M.J., Teama, D., Rosenstiel, T.N. & Khalil, M.A.K. (2010) Emissions of anaerobically produced methane by trees. Geophysical Research Letters, 37, L03807.

Rusch, H. & Rennenberg, H. (1998) Black Alder (Alnus glutinosa (L.) Gaertn.) trees mediate methane and nitrous oxide emission from the soil to the atmosphere. Plant and Soil, 201, 1-7.

Sayer, E.J., Heard, M.S., Grant, H.K., Marthews, T.R. & Tanner, E.V.J. (2011) Soil carbon release enhanced by increased tropical forest litterfall. Nature Climate Change, 1, 304-307.

Image: Forest of Taman Negara National Park by Vladimir Yu. Arkhipov, Arkhivov under Creative Commons CC BY-SA 3.0.