Diffusive methane emissions from temperate semi-intensive carp ponds

Manuring and supplementary feeding are common practices used to sustain high fish production in temperate semi-intensive carp ponds. How ever, the low use efficiency of added nutrients and organic matter may cause carp ponds to be ‘hot spots’ of methane (CH4) production and emission. Surface CH4 concentrations were measured and diffusive CH4 flux was estimated using a wind-based transboundary layer model in 3 nursery and 3 main carp ponds with different feeding rates and organic loading during 1 growing season. Mean (±SD) concentrations of CH4 were 1.3 ± 0.9 μM and 0.8 ± 0.8 μM in nursery and main ponds, respectively. All ponds were sources of CH4, with diffusive CH4 fluxes of 9.1 ± 6.8 mg C m−2 d−1 in nursery ponds and 6.4 ± 6.9 mg C m−2 d−1 in main ponds. Lower CH4 concentration and diffusive flux in the main ponds were probably due to bioturbation caused by the larger carp and consequent oxidation of the sediment. Seasonal dynamics of CH4 were mainly related to temperature. Methane concentration and diffusive flux levels re corded in this study were within the range of those reported in natural water bodies worldwide. Our results provide information on the role of carp aquaculture in greenhouse gas emission in temperate regions.

CH 4 is known to be produced in anoxic sediments (Bast viken et al. 2004), but recent evidence also infers its production in the aerobic water column (Grossart et al. 2011, Bogard et al. 2014).CH 4 is transferred from water to the atmosphere through diffusion or released by ebullition or through aerenchym tissue of littoral emergent aquatic plants (Bastviken et al. 2004).Nutrients, organic matter, temperature, and sediment are the main drivers of CH 4 production in aquatic ecosystems (Huttunen et al. 2003).Oxygen is an important factor in CH 4 production and consumption (Huttunen et al. 2006, Juutinen et al. 2009); lack of oxygen enhances CH 4 production in sediment, while its presence promotes its microbial oxidation (Bastviken et al. 2002, Attermeyer et al. 2016).The characteristics of catchment features, including vegetation and land use (Maberly et al. 2013, Borges et al. 2015a,b), temperature, rainfall, and wind speed influence CH 4 production, transport, and emission from aquatic ecosystems (Natchimuthu et al. 2014, Emilson et al. 2018).
In some countries, fishponds are an important component of lentic ecosystems (Pechar 2000).Fishponds occupy a surface area of 1200 km 2 in France, 410 km 2 in the Czech Republic, 420 km 2 in Germany, 25 669 km 2 in China, and 87 500 km 2 worldwide (Pokorný & Hauser 2002, Four et al. 2017, Xiong et al. 2017).In addition to rearing fish, fishponds provide ecosystem functions such as flood regulation along with retention of water, sediments, organic matter, nutrients, and micropollutants and may be important in maintaining bio diversity (Oertli et al. 2005, Boyd et al. 2010, Gaillard et al. 2016).
Semi-intensive carp polyculture is the main aquaculture production system in the Czech Republic and Central Europe as a whole (Gál et al. 2016).In this system, common carp Cyprinus carpio L. represents approx.90% of the total fish production, with the remainder com prising predatory fishes such as northern pike Esox lucius L., perch Perca fluviatilis L., eel Anguilla an guilla L., wels catfish Silurus glanis L., grass carp Ctenopharyngodon idella Valenciennes 1844, silver carp Hypophthalmichthys molitrix Valenciennes 1844, bighead carp H. nobilis Richardson 1845, white fish of the genus Coregonus, and tench Tinca tinca L. (Potužák et al. 2007).A key component of this production system is its reliance on a combination of natural and artificial feed (Adámek 2014).In practice, young-of-the-year fish are kept in nursery ponds and, from the second year, are held in main ponds until harvesting (Pokorný & Pechar 2000).This system is intended to reduce competition for food and maximise the use of natural pond re -sources in the production of fish biomass (Pokorný & Pechar 2000, Rahman et al. 2006).
Practices employed for high fish production often lead to eutrophication and deterioration of pond ecosystems (Pechar 2000), raising environmental concerns including those associated with the release of greenhouse gases (Williams & Crutzen 2010).The production of easily degradable organic matter coupled with the development of anoxic conditions on the bottom of eutrophic water bodies enhances the production of CH 4 and its subsequent evasion to the atmosphere (Gelesh et al. 2016).Indeed, Juutinen et al. (2009) found that CH 4 concentrations were higher in lakes with an anoxic hypolimnion and higher concentrations of total phosphorus than in lakes with an oxic hypo limnion and low concentrations of total phosphorus.
Aquaculture ponds are highly supplemented with organic matter and nutrients through feed, manuring, and, often, high concentrations of nutrients in their supply water and runoff from the catchment area (Pokorný & Pechar 2000, Adámek 2014).Low efficiency in use of the added material is common and causes accumulation of organic matter (Potužák et al. 2007).Some authors have considered ponds as hotspots of CH 4 production (Yang et al. 2015(Yang et al. , 2018a,b),b).Liu et al. (2016) reported that the conversion of rice paddies into crab ponds combined with fishponds reduced CH 4 emissions by 50%.Based on sedimentation rates, Boyd et al. (2010) showed that aquaculture ponds sequester 0.21% of global carbon emissions annually, with high rates of carbon containment in tilapia and carp ponds.They further recommended that fishpond managers might receive incentives to mitigate emissions of greenhouse gases from fishponds into the atmosphere.However, they also stated that sufficient data are not available on CH 4 and CO 2 in situ emissions from fishponds to confirm a positive net carbon sequestration.Studies of CH 4 emissions in various pond types from different geographic and climatic re gions are needed to resolve these contrasting views.
Previous studies of the environmental impact of fishponds in temperate regions focussed on eutrophication and its impact on pond biodiversity and downstream water bodies (Pechar 2000, Banas et al. 2008, Všetičková et al. 2012, Všetičková & Adámek 2013, Hlaváč et al. 2014, Four et al. 2017).CH 4 emissions from temperate fishponds have to date not been addressed.
The aims of this study were to determine and compare levels of dissolved and diffusive CH 4 in nursery and main fishponds in the South Bohemia region (Czech Republic) and to investigate the factors influencing this.Main ponds receive higher doses of manure and grains, thus providing substrate for methanogenesis, in addition to producing excessive phytoplankton biomass.We assumed that environmental factors and fishery management practices are synergistic in creating conditions favourable for CH 4 emission and ex pected to find higher concentrations and emissions in main ponds than in nursery ponds.

Study site
The study was conducted from April to October 2017 in 3 nursery ponds: Beranov, Roubíček, and Zb ěhov, and 3 main ponds: Kvítkovický, Posm ěch, and Dehtář, located in the upper catchment of the Vltava River near České Bud ějovice in South Bo hemia, Czech Republic.These ponds were created during the 15 th and 16 th centuries and have been used for fish production since then.Sediments were removed from Beranov 12 yr ago and from Kvítkovick ý 15 yr ago whereas there have been no sediments removed from Roubíček, Zb ěhov, Posm ěch and Dehtář in the last 20 yr.Annually, nursery ponds receive up to 0.5 t of feed per hectare in the form of cereals and are not manured.Main ponds receive 1−2 t each of feed and manure per hectare.Descriptions of the ponds are given in Table 1.

Physico-chemical water characteristics
The physico-chemical characteristics of water were measured once a month at the deepest part of the ponds, near the outlet.Dissolved oxygen, temperature, and pH were recorded using a YSI Exo2 multi-parameter probe.Water transparency was measured with a Secchi disk, and depth was measured using a graduated stick.Depth-integrated water samples from the whole water column were taken with a Van Dorn water sampler and transported to the hydrochemistry laboratory of the Institute of Hydrobiology (Biology Centre of the Czech Academy of Science, České Bud ějovice) for further analyses.The samples for analyses of dissolved organic carbon (DOC), dissolved nitrogen (DN), nitrate nitrogen (NO 3 -N), soluble reactive phosphorus (SRP), and total suspended solids (TSS) were filtered through glass-fibre filters with nominal porosity of 0.4 µm (type GF5, Macherey-Nagel).Samples were analysed within 24 h or kept frozen at −20°C.
Levels of TSS were determined gravimetrically on GF5 filters dried to constant weight at 105°C.Total organic carbon (TOC), total nitrogen (TN), DOC, and DN were determined on a Shimadzu TOC-L CPH ana lyser, working on the principle of high-temperature (750°C) catalytic oxidation of water samples and detection of the combustion products CO 2 and NO x using non-dispersive infrared (NDIR) and chemiluminescence de tectors, respectively.Samples were acidified with HCl and sparged with oxygen to remove inorganic carbon before analysis.Total inorganic carbon (TIC) was determined on a Shimadzu TOC-L CPH analyser by sparging acidified samples with purified oxygen to convert the inorganic carbon compounds CO 2 , bicarbonate, and car bonate to gaseous CO 2 , which was detected by the NDIR detector.Particulate organic carbon (POC) was determined as the difference between the TOC in unfiltered samples and DOC in the samples filtered through GF5 filters.Total phos phorus (TP) was determined by the molybdate method after perchloric acid digestion according to Kopáček & Hejzlar (1993).SRP was analysed according to Murphy & Riley (1962).Ammonium nitrogen (NH 4 -N) was determined by the spectrophotometric meth od with bis-pyrazolon ac cording to Ko páček & Procház ková (1993).NO 3 -N was quantified using di rect spectrophotometry in the UV region at 220 and 270 nm with correction for organic substances (Carvalho et al. 1998, Kalinichenko & Demutskaya 2004).Chlorophyll a (chl a) was analysed spectrophotometrically after acetone extraction following Lorenzen (1967).

Sediments
Sediments were collected in July and October 2017 with a core tube sampler at the deepest point near the pond outlet and in a littoral shallow part of the pond.Three samples were collected at each site in tubes with a diameter of 5 cm, and the top 5 cm of sediment were sliced and pooled into a single sample.Samples were freeze-dried and analysed for sediment TN, TP, and TOC in the same laboratory as water analyses.TP sed was determined by the molybdate method after perchloric acid digestion according to Kopáček et al. (2001).TOC sed and TN sed were determined by elemental analysis on a varioMICRO Cube analyser (Elementar Analysensysteme).Samples were acidified with HCl before analysis, and inorganic carbon was removed as CO 2 (Kopáček et al. 2001).

Surface-water CH 4 concentration
Surface CH 4 concentrations were measured using the headspace technique as described by Bastviken et al. (2004).In the field, water samples were taken from 10 cm below the surface with a 50 ml syringe capped with a needle mounted on a 3-way valve.The first water sample was used to remove air, and a new water sample of 40 ml was drawn into the syringe and adjusted to 20 ml.A headspace was then created by adding 20 ml of ambient air and shaking for 1 min to equilibrate the CH 4 concentration in the water and air enclosed in the syringe.The headspace gas was then transferred into 12 ml pre-evacuated exetainer vials equipped with chlorobutyl septa (vial type 3, order code 839W/GL, LabCo).Ambient CH 4 concentrations were also determined from air samples collected on the same sampling day to correct for background concentrations of air in the headspace (Bastviken et al. 2010).Headspace CH 4 concentration was determined in the laboratory of the Department of Ecosystem Biology (Faculty of Science USB, České Bud ějovice) using an HP 6890 gas chromatograph (Agilent) equipped with a 0.53 mm × 30 m GS-Alumina column and a flame ionization detector.Calibration was done with certified CH 4 :N 2 mixtures (Linde) in concentrations of 1.7, 10, 100, 1000, and 10 000 ppm of CH 4 .The detection limit for CH 4 analysis was 0.1 ppm, and the precision of measurements was ± 3%.The quantity of CH 4 that remained dissolved in the syringe water sample was calculated from headspace CH 4 concentrations using Henry's law adjusted for in situ temperature according to Wiesenburg & Guinasso (1979).CH 4 concentration in the original water sample was then obtained by dividing total CH 4 quantity in the headspace and in the syringe water corrected for ambient air concentration by the volume of water sample (Bastviken et al. 2010).The results were considered representative for the month in which the samples were taken.

Surface-water CH 4 emissions
Gas exchange between air and water (F) was calculated indirectly using the 2-layer model with the equation F = k(C sur − C eq ), where C sur is the gas concentration in surface water in µmol l −1 , C eq is the gas concentration in surface water in equilibrium with the atmosphere in µmol l −1 , and k is the gas exchange constant (cm h −1 ).The value of k was calculated from the local wind speed according to Crusius & Wanninkhof (2003): k = k 600 (Sc/600) n , where k 600 is the gas transfer velocity for a Schmidt number of 600; Sc is the Schmidt number of CH 4 ; and n takes the value of −0.67 or −0.5 if the wind speed at 1 m height is lower or higher than 3 m s −1 , respectively (Crusius & Wanninkhof 2003).The value of k 600 (cm h −1 ) was calculated according to Crusius & Wanninkhof (2003) as ), where µ 10 is the local wind speed in m s −1 at a height of 10 m.The wind speed measured at 2 m was converted to a height of 10 m according to Crusius & Wanninkhof (2003): µ 10 = 1.22 µ 2 , where µ 2 is the wind speed at 2 m.Sc for CH 4 was calculated according to Wanninkhof (1992) with the following formula: Sc CH4 = 1897.8− 114.28t + 3.2902t 2 − 0.039061t 3 , where t (°C) is the water temperature at the time of CH 4 extraction.C eq was determined from equation: C eq = β pCH 4 , where β is the solubility of CH 4 computed according to Wiesenburg & Guinasso (1979), and pCH 4 is the partial pressure of CH 4 in the atmosphere.The measured surface water CH 4 concentrations were compared to their respective concentrations in equilibrium with the atmosphere to obtain the level of CH 4 saturation.

Statistical analysis
Generalized linear mixed models (GLMMs) were used to assess significant differences in water quality parameters between pond types (Zeger & Liang 1992, Breslow & Clayton 1993).Non-parametric analysis of longitudinal data (nparLD) was used to test the effect of pond type on organic carbon, nitrogen, and phosphorus content in pond sediment (Noguchi et al. 2012).A Wilcoxon signed rank test was used to evaluate differences in nutrient and organic matter content in sediment between the 2 sampling times.GLMM was also used to test the effect of pond type, sampling time, and their inter action on dissolved CH 4 in pond surface water, CH 4 saturation levels, and diffusive CH 4 flux.This analysis was followed by Tukey's post hoc tests to determine differences in CH 4 concentration, saturation, and flux within a pond type over time and differences between pond types at each sampling time.Partial least squares regression (PLSR) analysis was used to identify drivers of variation in CH 4 concentration and flux between pond types.Explanatory variables were log(x +1) transformed prior to regression analyses.Pond type, temperature, DO, DOC, POC, chl a, TP, TN in water, and TP sed and TN sed were selected as variables for regression analyses.The variable 'pond type' was considered as a nominal variable of 2 levels, i.e. nursery and main.The most important drivers of CH 4 concentration and flux were identified based on the weight of each predictor variable and total explanatory capacity (R 2 of Y and R 2 of X i ) of extracted components.GLMMs, nparLD, and the Wilcoxon test were performed in R version 3.4.4(R Core Team 2018), and PLSR was conducted using Statistica 13 (STATIS-TICA advanced, module STATISTICA Multivariate Exploratory Technique; Statsoft).

CH 4 concentrations and diffusive emissions
The mean (± SD) surface concentrations of dissolved CH 4 were 0.8 ± 0.8 and 1.3 ± 0.9 µM in main and nursery ponds, respectively.In nursery ponds, a 2-peak pattern was observed, with the minimum in April and September, intermediate values in June and July, and maximum values in May and August (Fig. 3a).The main ponds showed a peak in May and low consistent values in the remaining months of the season (Fig. 3a).Dissolved CH 4 concentrations ranged from 0.06 to 4.8 µM in all ponds.There was an effect of time of sampling (F 5, 416 = 77.8,p < 0.001, Fig. 3a) and an interaction of pond type and time of ponds.Variable abbreviations as in Table 2. Values are means ± SE sampling (F 5, 416 = 14.1, p < 0.001, Fig. 3a) influencing the surface CH 4 concentration.CH 4 concentration differed significantly between nursery and main ponds in June, July, and August (Fig. 3a).Nursery ponds exhibited higher dissolved CH 4 concentration than did main ponds throughout the monitored period.All in vestigated ponds were highly supersat-urated with CH 4 .The mean saturation degree was 41 824 ± 28 932% and 20 770 ± 22 791% in nursery and main ponds, respectively (Fig. 3b).CH 4 supersaturation showed the same temporal trend in both pond types.All ponds were sources of CH 4 in the atmosphere during the growing season.Diffusive emissions of CH 4 carbon (CH 4 -C) ranged from 0.19 to 32 mg m −2 d −1 in all ponds with a mean of 7.8 ± 7.0 mg m −2 d −1 .CH 4 flux rates differed significantly over time (F 5, 416 = 96.5, p < 0.001, Fig. 3c), with a significant interaction between sampling time and pond type (F 5, 416 = 27.9, p < 0.001, Fig. 3c).However, the interaction was weak, as flux rates differed significantly between pond types only in August.Flux rates of diffusive CH 4 -C were slightly higher in nursery ponds (9.1 ± 6.8 mg m −2 d −1 ) than in main ponds (6.4 ± 6.9 mg m −2 d −1 ) throughout the growing season.Nursery ponds exhibited peaks in May and August, while main ponds peaked in May (Fig. 3c).Unlike the trends in dissolved CH 4 and CH 4 saturation, the highest peak, recorded in May, was in the main ponds (Fig. 3c).

Factors affecting CH 4 concentration and diffusive emissions
The PLSR was used to reveal whether physicochemical properties of water and sediment composition can explain CH 4 concentration and diffusive emissions.The results indicated that 3 components explained 55% of the variation in CH 4 concentration in the investigated ponds (Table 3).The first component explained 40% of the total variance, and its information content was positively associated with water temperature and negatively associated with pond type.The second component, also positively associated with water temperature, and the third component, negatively associated with DOC, ac - counted for 11 and 4% of the total variance, respectively (Table 3).The PLSR analysis also indicated that only 1 component explained 37% of the variation in diffusive CH 4 flux (Table 3).This component was positively associated with water temperature, but the association was not significant.

DISCUSSION
The increase in CH 4 concentration and flux that occurred from April to May suggests that the increase in CH 4 concentration was primarily related to the increase in water temperature (Table 3).Water temperature influences CH 4 production in aquatic ecosystems as it stimulates activity of methanogenic bacteria (Hofmann et al. 2010, Musenze et al. 2014, Natchimuthu et al. 2014, Borges et al. 2018).In temperate regions, CH 4 concentration in water increases at the beginning of spring, triggered by the increase in sediment temperature and water temperature (Descloux et al. 2017).The peak of CH 4 flux recorded in the main ponds in May can be explained by sediment bioturbation by carp along with wind speed.At the beginning of the growing season, the feeding behaviour of carp enhances the release of CH 4 accumulated in sediment during the previous growing season and winter (Bhattacharyya et al. 2013, Xiong et al. 2017).The average wind speed in May was higher over the main ponds (2.3 ± 0.5 m s −1 ) than above nursery ponds (1.0 ± 0.5 m s −1 ).
CH 4 concentration and flux in the main ponds decreased in summer and became lower than in nursery ponds.Other studies have shown CH 4 emissions to be highly correlated with temperature throughout the growing season (Natchimuthu et al. 2014, Wik et al. 2014).The observed low CH 4 concentration and flux in the main ponds may be explained by CH 4 oxidation and the behaviour of carp over 1 kg body weight burrowing in search of food at the bottom of the ponds.CH 4 oxidation is an important pathway that reduces surface water CH 4 concentration and its emission from water bodies (Bastviken et al. 2008, Juutinen et al. 2009).Oxidation probably plays an important role in CH 4 dynamics in carp ponds as well, despite its short time exposure due to shallowness of the ponds (Table 1).In shallow lakes, CH 4 bubbles escape oxidation due to short travel time from sediment through a wellmixed water column to the surface (Bastviken et al. 2004, 2008, Juutinen et al. 2009, Natchimuthu et al. 2014).However, low concentrations of NO 3 (Table 2) may be a limiting factor in CH 4 oxidation occurring in deeper areas near pond outlets (Bastviken et al. 2008, Deutzmann et al. 2014, Roland et al. 2017).CH 4 oxidation rates are positively correlated to consumption of NO 3 under anoxic conditions (Roland et al. 2017).Bioturbation of the top sediment layer by carp may reduce CH 4 production by improving aerobic conditions of top sediment or by reducing the concentration of easily oxidised organic matter through the exposure of older sediment (Ritvo et al. 2004).
CH 4 concentration was negatively related to DOC (Table 3), implying that DOC was not the primary source of, or a factor strongly associated with, CH 4 production.The increase in CH 4 concentration and flux in the nursery ponds in August probably followed maturation and decomposition of fresh plant biomass rather than originating from old settled detritus (Kelly et al. 1997).CH 4 production in lakes of temperate and boreal regions might differ substantially depending on the chemical composition of sediments (Emilson et al. 2018).Sediments containing organic matter from macrophytes and aquatic plants produce more CH 4 than sediments containing organic matter of terrestrial origin.Nursery ponds had littoral zones largely covered by emergent macrophytes in addition to floating and submerged aquatic plants that could supply fresh organic matter for methanogenesis.Additionally, water bodies with higher abundance of macrophytes usually have significantly higher CH 4 concentration and flux than those without, or with low abundance, of macrophytes (Selvam et al. 2014) were rare in main ponds due to eutrophication, as indicated by low water transparency, as an effect of nutrient overload.Moreover, aquatic plants cannot establish in densely stocked fishponds due to carp feeding behaviour (Scheffer et al. 2001).Common carp, especially larger individuals, are known to interfere with aquatic plant growth both directly by mechanical uprooting and consumption and indirectly by increasing water turbidity causing reduction in photosynthesis (Miller & Crowl 2006).Diffusive CH 4 flux was not significantly related to any measured environmental factor, indicating that wind speed was the main factor regulating diffusive flux (Musenze et al. 2014).
Our findings of CH 4 concentration and diffusive flux were in general agreement with those obtained in other aquatic bodies worldwide (Table 4), although they deviated from some observations.CH 4 concentrations and flux were reported to be lower in Lake Erssjön in Sweden and higher in Indian ponds compared to our findings (Table 4).The primary difference between our ponds and the Indian ponds was higher organic matter supply and higher water temperature recorded in Indian ponds than in our ponds (Selvam et al. 2014).Lake Erssjön had lower nutrient concentrations and lower mean temperature compared to our ponds (Natchimuthu et al. 2016).Recent studies have re ported very diverse values for emissions of greenhouse gases from aquaculture systems (Yang et al. 2015, Ma et al. 2018).In agreement with our study, these authors confirmed that temperature and aquaculture management strongly influ-ence CH 4 emissions from ponds.However, they did not relate CH 4 emissions from ponds to the behaviour of cultured animals.We did not compare CH 4 flux rates from these studies to our results, since they did not distinguish diffusive flux from ebullitive flux.Our results represent only a portion of the CH 4 flux from the ponds because our study does not include ebullitive flux.The contribution of ebullitive CH 4 to total CH 4 emission ranges from 10 to more than 90% of total CH 4 emissions in temperate and boreal aquatic systems (Casper et al. 2000, Bastviken et al. 2004), hence it is not possible to make a reliable estimate of total emissions based on diffusive fluxes only.The level of ebullitive CH 4 from carp ponds remains uncertain until temporal and spatial data of ebullitive fluxes from them are available, as ebullitive CH 4 is system specific.In this study, the main ponds did not diffuse more CH 4 than the nursery ponds, possibly due to sediment disturbance by carp.This indicates that organic matter in the sediment of the main ponds might be processed more through oxic pathways than anoxic-methanogenic pathways.

CONCLUSIONS
Both the nursery ponds and the main carp ponds were significant sources of diffusive CH 4 into the atmosphere.Contrary to our expectations, the main ponds had lower CH 4 concentration and lower diffuse CH 4 flux m −2 than the nursery ponds, despite the higher loadings of organic matter they receive

Fig. 1 .
Fig. 1.Temporal dynamics of physico-chemical characteristics of water in main (dotted line) and nursery (continuous line) ponds.Variable abbreviations as inTable 2. Values are means ± SE

Table 3 .
Ma et al. (2018)t al. (2018)reported higher CH 4 flux from crab ponds with macrophytes than from those without.Macrophytes Results of the partial least square regression analysis, extracted components and weights of associated explanatory variables.COMP: component; other variable abbreviations as in Table2.R 2 Y: explained variability of dependent variables (CH 4 concentration or CH 4 diffusive flux); R 2 X: explained variability in independent variables.Significant correlations (p < 0.05) are highlighted in bold

Table 4 .
Methane (CH 4 ) concentration and flux (means ± SD or range of values) in lentic water bodies worldwide.Conc: concentration; Dif: diffusive; Ebul: ebullitive; nm: not measured.International country codes based on the ISO3166 standard published by the International Organisation for Standardisation (www.iso.org/iso-3166-country-codes.html) are in parentheses through fishery management.Common carp, being a benthic feeder in the main ponds, may reduce CH 4 production and release by disturbing sediment and maintaining the upper layer in oxic conditions.The CH 4 emissions from the carp ponds in our study are within the range found in other freshwater lentic water bodies.However, more studies are required to quantify ebullitive and other pathways of CH 4 release into the atmosphere in order to define the local and global role of carp ponds in CH 4 emissions.