COBISS: 1.01 Diel cycling and flux of HCO3- in a typical karst spring-fed stream of southwestern China Dnevne spremembe in tok bikarbonata v tipičnem kraškem potoku na jugozahodu Kitajske Cheng ZHANG1,2, Jinliang WANG1,2, Jun YAN3 & Jianguo PEI1,2 Abstract UDC 546.26:556.53(513) 556.53:551.44(513) Cheng Zhang, Jinliang Wang, Jun Yan & Jianguo Pei: Diel cy­cling and flux of HCO3- in a typical karst spring-fed stream of southwestern China We investigated the diel variations of the dissolved inorganic carbon, isotopic composition, and partial CO2 pressure from a karst spring (Guangcun Village, Guangxi, Southwest China) to the 1,350 m downstream profile of the stream. In addition, the carbon loss and CO2 exchange flux at the water-gas inter­face were also estimated. The results showed that the pH value and DO in the stream varied regularly on a daily basis with the temperature of stream water, suggesting that the photosynthe­sis of aquatic plants and algae is the controlling factor for the diel variations of the pH and DO. During the monitoring pe­riod, while the DIC (mainly in HCO3-) input (at spring) was relatively stable at about 4.46 mmol L-1, the concentrations of HCO3- and Ca2+ at downstream showed a diel cycle of daytime decrease and nighttime increase, with an amplitude of 22.4 %. We also found out that the CO2 degassing mainly occurred in the upper reach of the surface stream right after groundwater is exposed to the surface. The total CO2 exchange flux of the entire monitoring stream section was calculated to be 29.83 kg d-1, accounting for 17.8 % of the DIC loss, which means that ap­proximately 4/5 of the loss was converted into organic carbon or calcite precipitation. Compared with the total carbon input at spring, this carbon loss only accounts for 6.5 % of the to­tal carbon amount (1.4 % of which was converted into organic carbon and 1.1 % of which was degassed to the atmosphere), indicating that the DIC of karst groundwater in low order sur­face stream of Guancun is stable in general, with 1 % being lost to the atmosphere. This suggests that on a daily timescale, car­bon loss in the form of CO2 of low order karst streams with lower gradient is much less pronounced. Key words: inorganic carbon cycle, spring-fed stream, aquatic vegetation photosynthesis, CO2 degassing, inorganic carbon flux, karst. Izvleček UDK 546.26:556.53(513) 556.53:551.44(513) Cheng Zhang, Jinliang Wang, Jun Yan & Jianguo Pei: Dnevne spremembe in tok bikarbonata v tipičnem kraškem potoku na jugozahodu Kitajske Preučevali smo dnevno nihanje raztopljenega anorganskega ogljika, izotopsko sestavo in parcialni tlak CO2 na 1350 m dolgem odseku kraškega potoka v vasi Guangcun, Guangxi na jugozahodu Kitajska. Ocenjevali smo tudi izgubo ogljika in izmenjavo CO2 na stiku vodne gladine z ozračjem. Dne­vna nihanja pH in raztopljenega kisika so očitno povezana s fotosi­ntezo vodnih rastlin, saj so se v času spremljanja, ko je bil dotok DIC (raztopljen organski ogljik, predvsem HCO3-) skozi izvir relativno stabilen (4,46 mmol l-1), koncentracije HCO3- in Ca2+ spreminjale v dnevnem ciklu. Največje dnevne in najmanjše nočne vrednosti so se razlikovale za 22,4 %. Raz­plinjanje CO2 je največje v zgornjem toku, tik za izvirom. Sku­pna izmenjava CO2 za celoten odsek potoka je bila ocenjena na 29,83 kg CO2 d-1, kar je 17,8 % celotne izgube raztopljenega organskega ogljika, iz česar je mogoče sklepati, da se približno 4/5 pretvorji v organski ogljik oziroma v izločanje kalcita. Izgu­ba predstavlja 6,5 % skupnega dotoka ogljika na,od katerega je 1,4 % pretvorjenega v organski ogljik, 1,1 % pa razplinjenega v ozračje. Vrednost anorganskega ogljika v potoku je relativno stalna, izhajanje ogljika v obliki CO2 je v opazovanem potoku z relativno majhnim strmecem, precej neizrazito. Ključne besede: spremembe anorganskega ogljika, izvirni po­tok, fotosinteza vodnega rastlinja, CO2 razplinjevanje, tok ano­rganskega ogljika, kras. 1 Key Laboratory of Karst Dynamics, MLR/GZAR, Institute of Karst Geology, CAGS, Guilin 541004 China, e-mail: chzhang@karst.ac.cn, jlwang@karst.ac.cn, peijg@karst.ac.cn 2 International Research Center on Karst under the Auspices of UNESCO, Guilin 541004 China 3 Department of Geography and Geology, Western Kentucky University, Bowling Green, 42101, USA, e-mail: jun.yan@wku.edu Received/Prejeto: 25.01.2016 ACTA CARSOLOGICA 45/1, 107–122, POSTOJNA 2016 Cheng Zhang, Jinliang Wang, Jun Yan & Jianguo Pei Introduction As one of three processes of the Earth’s critical zone (Lin 2010), biogeochemical process connects biotic process with abiotic process. Coupled with hydrological process, it supports the sustainability of ecological process and in turn determines the overall morphology and func­tion of the critical zone (Yang et al. 2014). In the fields of global change and karst carbon cycle, a better under­standing of biogeochemical process, e.g. its influencing factors and mechanisms, are of vital importance to the study of the time-scale of karst processes and the stabil­ity of karst carbon sink (Martin et al. 2013; Zhang 2011). CO2 consumed in carbonate dissolution, directly origi­nated from rainfall as well as indirectly from decomposi­tion of organic matter, are transferred into karst aquifers. The result is also chemical denudation of carbonate rocks (White 2013). During CO2 transfer through karst aqui­fers, some portion of the carbon is released back to the atmosphere via the degassing of CO2 as well as CaCO3 deposition as a result of speleothems deposition (e.g. sta­lagmites) and tufa downstream of karst spring. However, some inorganic carbon (mainly HCO3-) (Madsen 1983) is usually converted to organic carbon by aquatic vegeta­tion photosynthesis and some, with the assistance of mi­crobial carbon pump, might further become semi-labile dissolved organic carbon (SLDOC) or recalcitrant dis­solved organic carbon (RDOC) in water bodies such as reservoirs or lakes (Jiao et al. 2013; He et al. 2010; Chen et al. 2012; Mermillod-Blondin et al. 2015). Hence, bio­geochemical process in karst reservoirs and rivers can reflect weathering processes of watershed (Kanduc et al. 2007), and moreover could be counted as one of natural carbon sinks, that is, carbon sequestration. However, the research on the magnitude and mechanisms of its influ­ence on carbon sink as well as the role of microbial proc­esses in fresh water carbon storage is still in its infancy stage and we just start to gain some level of understand­ing. Thanks to the development of high-resolution auto­matic online monitoring equipment and high-frequency automatic sampling techniques, more and more studies have been conducted on diel biogeochemical processes since the 1990s (Nimick et al. 2011). At least five pro­cesses have influence on the variations in amount and flux of dissolved inorganic carbon (DIC) in river waters: photosynthesis, respiration, water-gas exchange (degas­sing), groundwater recharge, and geochemical process (namely carbonate mineral precipitation or dissolution) (Tobias et al. 2011). Based on the findings from various studies in the past 20 years (Nimick et al. 2005; Nagorski et al. 2003; Waldron et al. 2007; Poulson et al. 2010; Liu et al. 2006; Liu et al. 2008), we now know that aquatic vegetation activity can significantly affect the variations of hydrochemical parameters of stream waters, such as pH, DIC, dissolved oxygen (DO), and specific conduc­tivity (SpC). The short time scale research on diurnal or seasonal variations is valuable to the investigation of the relatively rapid biogeochemical processes in waters (e.g. processes in stream flows). In addition, it also helps to reckon or to quantify watershed processes in upstream recharge areas. The existing studies (Dandurand et al. 1982; Spiro & Pentecost 1991; Guasch et al. 1998; Reich­ert 2001; Lorah & Herman 1988; Finlay 2003) on differ­ent stream orders show that biological processes (pho­tosynthesis and respiration) and geochemical processes (bicarbonate equilibration and calcite precipitation) are two main controlling factors for diurnal variations of pH value, SpC, and concentrations of Ca2+ and HCO3- in streams. The majority of these studies focused on small tributary streams, particularly karst spring-fed small streams (discharge < 0.1m3/s) as documented in Nimick et al. (2011), while some were conducted on streams with moderate flows (discharge < 5m3/s) (Parker et al. 2010) or large rivers (discharge > 10m3/s) (Hayashi et al. 2012). During carbonate rock dissolution, CO2 is con­sumed and HCO3- is generated. Accordingly, high amount of HCO3- is one of the distinctive characteris­tics of karst groundwater. It was found that carbon sink generated by karst processes could be an important part of the missing carbon sink on a global scale (Jiang et al. 2012). In karst areas, surface rivers are often rich in aquatic plants. Hence, the transformation of DIC to DOC caused by aquatic vegetation photosynthesis usu­ally leads to the loss of DIC (mainly HCO3-) along riv­ers’ course, which presents a natural sink of carbon (De Montety et al. 2011; Zhang et al. 2012). Nevertheless, the intensity, seasonal variation and major controlling fac­tors of this carbon sink should be targeted by further investigation. The research on diel variations in hydro­chemical composition and biogeochemical processes in surface karst rivers can not only reveal controlling fac­tors for diel cycling of hydrochemical inorganic com­positions, but also help us better understand the rate of transfer of inorganic and organic carbon in karst pro­cesses so as to provide insights about the biogeochemical nature (e.g. short time-scale features) of karst processes. The findings could significantly improve accuracy of karst carbon flux. Furthermore, they can help the imple­mentation of long term water quality monitoring plan in drainage basins. With a case study on a karst spring-fed stream at Guangcun Village, Guangxi, Southwest China, this paper deals with the diel variations of the dissolved inorganic carbon, isotopic composition, and partial CO2 pressure from the outlet to the 1,350 m downstream section of the underground stream. In addition, it also estimates the carbon loss and CO2 exchange flux at the water-gas interface. Diel cycling and flux of HCO3- in a typical karst spring-fed stream of southwestern China Study Area The three monitoring sites are located along the stream section from the outlet of an underground stream at Guancun Village (in Daliang Town, Rongshui County, Guangxi Province, China) to the junction of its lower reach at Leiya Village. This stream section is 1,350 m long (Fig. 1) and 2-5 m wide, with a depth of 0.2-1.0 m (gen­erally 0.5 m). The flow velocity in this section is about 0.2 m s-1 and the average hydraulic gradient is less than 2 m/km. This stream is about 60 km away north from Liuzhou City, a major industrial city in Southwest China. Its spring is located at 109° 20' 3.41" E and 24° 52' 5.34" N with an elevation of 160 m. The drainage area is about 30.5 km2 with higher elevations in the northeast and northwest. This section is characterized by karst peak-cluster depressions, with well-developed dolines and sinkholes. This underground stream is developed from the recharge area and composed of limestone and dolo­mite of Rongxian Formation (D3r). The annual average temperature of the study area is 20 oC and the annual av­erage rainfall is 1,750 mm, with the wet season from May to July and the dry season from September to February. The rainfall in six months of the dry season accounts for only 11 % of the total of the whole year. The water of this underground stream is rich in calcium and bicarbonate with low contents of Mg2+, Na+, K+, Cl-, SO42- and NO3-. The ranges and mean values of their concentrations can be seen in Tab. 1, showing relatively low variations in the monitoring period. Ca2+, HCO3- and Mg2+ mainly come from the Devonian limestones and dolomites of Rongx­ian Formation while K+ and Na+ are from soil. During the monitoring period, the discharge at spring, the source of the underground stream was rela­tively stable, with a range of 149.5-156.4 L/s. Water tem­perature (T), pH, SpC, and dissolved oxygen (DO) re­mained about the same as well (21.26°C, 7.47, 418 µS/cm, and 7.18 mg L-1 respectively) during the monitoring pe­riod. Fig. 1: Locations of the study area and monitoring sites. Cheng Zhang, Jinliang Wang, Jun Yan & Jianguo Pei Tab. 1: Diel variations and mean values of concentration of ma­jor ions at spring (mg L-1). Ion Maximum Mean Minimum Maximum-Minimum K+ 0.55 0.53 0.5 0.05 Na+ 0.72 0.70 0.68 0.04 Ca2+ 85.85 84.78 83.22 2.63 Mg2+ 10.37 10.21 9.96 0.41 Cl- 2.76 2.74 2.72 0.04 SO42- 14.58 14.54 14.46 0.12 HCO3- 278.45 272.30 265.8 12.65 NO3- 9.83 9.54 8.24 1.59 TOC 1.41 0.65 0.39 1.02 DOC 0.79 0.52 0.38 0.41 Materials and Methods Monitoring, Sampling, Analytical Methods Diel monitoring was carried out during August 21-26, 2011, a total of six days. Three monitoring sites were set up along the stream section: (1) the outlet site of the un­derground stream (CK): the input section site to obtain the physiochemical indicators as the initial values of the water; (2) the Guangcun Bridge site (GCQ): 880 m downstream from the outlet; (3) the junction site just be­fore confluence with Shimen river at Leiya Village (LY): 1,350 m downstream from the outlet, which is used to evaluate the variations, diel cycling and ranges of the biogeochemical parameters along the flow path, and ul­timately how biological processes influence variations of bicarbonates temporally and spatially. An automatic on­line monitoring instrument was installed at each moni­toring site. The in-situ titration analysis and sampling work were started at the outlet site first. It lasted only one day (Tab.1) at the outlet site because of the relatively constant ion concentrations. The work conducted at the Guangcun Bridge site and the junction site at Leiya Vil­lage lasted all six days. T, pH, SpC, water level, and DO were monitored by a YSI 6920 at 5-minute intervals with the accura­cies of 0.1 °C, 0.2, 1 µS cm-1, 0.01 m, and 0.01 mg L-1 respectively. The concentrations of Ca2+ and HCO3- were obtained through in-situ titration analysis using Aqua­merck testing box at 1-hour intervals with the accuracy of 0.05 mmol L-1. Hydrochemical samples and car­bon isotopic samples for chemical and isotopic analy­sis were collected at 1-hour intervals by an automatic sampler ISCO. Three flow discharge measurements were conducted on August 21, 23, and 26 respective­ly. In line with the standards of GB/T8538-2008 and DZ/T 0184.1-0184.22-1997 (geological sample isotopic analytical methods), the determination of hydrochemi­cal compositions and .13C isotopes was performed in the Karst Geological Resources, Environmental Monitoring and Testing Center, Ministry of Land and Resources of China using an IRIS Intrepid II XSP plasma spectrom­eter and a MAT253 stable isotope mass spectrometer. Data Processing The values of calcite saturation indices (SIc) and partial pressure of dissolved CO2 (pCO2) were calculated using the program WATSPEC (Wigley 1977). The water tem­perature and the pH were extracted from the available online monitoring data. The concentrations of Ca2+ and HCO3- were determined by in-situ titration data and the ion concentrations of Mg2+, K+, Na+, Cl- and SO42- were the analytical data tested from the water samples in the laboratory. The CO2 exchange flux at the water-gas interface was estimated by the molecular diffusion model: F = k × .C = k × (Cwater - Cair) (1) where F is the exchange flux at the water-gas interface (mmol m-2 d-1). When F > 0, water releases CO2 to the at­mosphere, while water absorbs CO2 from the atmosphere when F < 0. Cwater and Cair are the CO2 concentrations (ppm) in water and gas respectively and hence .C repre­sents the difference of gas concentrations across the inter­face (Cole et al. 1994; Galy-Lacaux et al. 1997; Jones et al. 2001), k, the gas exchange coefficient, is a function of the boundary diffusion layer thickness (z), i.e. k=D/z, where D is the diffusion coefficient of gas. Cole et al. (1994) found that the k value for CO2, was 0.5 m d-1 in most lakes and the corresponding boundary diffusion layer thick­ness was equivalent to 300 µm (in summer) and 200 µm (in winter). Researchers from the UK (Galy-Lacaux et al. 1997) and Germany (Schmidt & Conrad 1993) used 100 µm (in winter) and 200 µm (in summer) as the values of boundary diffusion layer thickness when studying water reservoirs and lakes. Some researchers adopted a uni.ed boundary layer thickness of 200 µm for estimating (Jones et al. 2001; Wang et al. 2012). The underground stream basin in Guancun Village, Rongshui County, Guangxi, is located in low-elevation karst peak-cluster depressions with low perennial wind speeds, accordingly, a uni.ed boundary diffusion layer thickness of 200 µm was used for this research. The diffusion coefficient of CO2 in water were 1.26 ×10-5 cm2 s-1 (in winter with the water temper­ature of 10°C) and 1.93 × 10-5 cm2 s-1 (in summer with the water temperature of 25°C), and the corresponding CO2 exchange coefficient was 0.5-0.8 m d-1 (Wang et al. 2012). Diel cycling and flux of HCO3- in a typical karst spring-fed stream of southwestern China Results Water Temperature, pH, and DO The water average temperature, pH and DO were con­stant at the outlet site of the underground stream, 21.3 °C, 7.5 and 7.17 mg/L respectively (Fig. 2). In contrast, these parameters at the Guangcun Bridge site (GCQ) and the Leiya Village site (LY) showed prominent diel variations (Fig. 2). At GCQ, the pH value increased to 7.9-8.3 and the average daily values of the water temperature, pH and DO (from August 21 to 25) were 23.2 °C, 8.0 and 8.41 mg L-1 respectively, with the respective diel ampli­tudes of 5.0 °C, 0.4 and 4.83 mg L-1. As the water tem­perature rose during daytime, the pH and DO values of the water increased simultaneously, with the maximum values of 8.3 and 11.85 mg L-1 (13:00-14:30). Low val­ues occurred at nighttime, with the minimum values of 7.87 and 6.69 mg L-1, respectively (Tab. 2 and Fig. 2), thus showing typical diel variations of pH and DO led by photosynthesis. Comparatively, similar variations oc­curred at LY. The pH increased to 8.0-8.3 and the aver­age daily values of water temperature, pH and DO (from August 22 to 25) were 23.6 °C, 8.1 and 8.11 mg L-1 re­spectively, with the respective diel amplitudes of 5.0°C, 0.3 and 3.58 mg L-1. The variation amplitudes of pH and DO were smaller slightly at LY when compared to those at GCQ. The maximum values during daytime were 8.3 and 10.59 mg L-1 and the minimum values at night were 8.0 and 6.61 mg L-1. The underground stream has a flow path of 880 m from CK to GCQ. As shown in Fig. 2, the increased range of the pH values along this flow path can be divided into two parts: from 7.5 (which is the constant pH value at CK) to 7.9, and a diel range from 7.9 to 8.3. This may suggest that the degassing and the photosynthesis of aquatic plants and photosynthetic organisms both affect the increase of the pH values. HCO3-, Ca2+, and .13CDIC The hydrochemical indices at CK were relatively con­stant. The average daily concentrations of Ca2+, Mg2+ and HCO3- were 84.8, 10.2 and 272.3 mg L-1 respec­tively, with the respective amplitudes of 2.6, 0.4 and 12.7 mg L-1. The diel average value of the inorganic car­bon isotope (.13CDIC) in water was -15.31 ‰, with an amplitude of 0.68 ‰. The average daily concentrations of Ca2+, Mg2+ and HCO3- at GCQ were 78.6, 10.2 and 262.7 mg L-1 respectively, with the respective ampli­tudes of 13.1, 0.6 and 49.0 mg L-1. The diel average value of .13C was -13.56 ‰, with an amplitude of 2.55 ‰. The average daily concentrations of Ca2+, Mg2+ and HCO3- at LY were 79.4, 10.3 and 259.3 mg L-1 respectively, with the respective amplitudes of 13.6, 0.6 and 40.1 mg L-1. The diel average value of .13C was -13.34 ‰, with an amplitude of 2.64 ‰. Contrary to the diel variations of pH and DO, the concentrations of Ca2+ and HCO3- at GCQ and LY showed a daytime decrease and nighttime increase cy­cling (Fig. 3), but the peaks were not completely syn­chronized with each other. The low values appeared at around 17:00 with the minimum values of 69.0 mg L-1 and 222.6 mg L-1 respectively, which may be consis­tent with the high values of water temperature, almost 3 hours later than the time when the maximum values of pH and DO appeared. After that, the concentrations of Ca2+ and HCO3- at GCQ and LY gradually increased and the maximum values appeared in the early morning of the following day. The diel variations of the .13CDIC were the total opposite to those of the HCO3-, showing a day­time increase and nighttime decrease. The high .13CDIC value corresponds to the low DIC. The maximum diel amplitude of the isotope was around -3.10 ‰ (from the minimum value of -14.40 ‰ at night to the maximum value of -11.30 ‰ during daytime). SIc and pCO2 At CK, the calcite saturation index (SIc) and partial CO2 pressure (pCO2) were relatively constant, with the aver­age values of 0.32 and 8,494 µatm respectively and the amplitudes of 0.04 and 397 µatm respectively. The values of SIc and pCO2 showed sharp increases and decreases downstream, respectively. At GCQ, the average daily val­ues of SIc and pCO2 were 0.84 (range 0.69-1.07) and 2500 µatm (range 1,110-3,510 µatm) and the average diel am­plitudes were 0.31 and 2144 µatm, a significant increase in variation amplitude compared to those at CK. At LY, the average daily values of SIc and pCO2 were 0.94 (range 0.85-1.12) and 2360 µatm (range 950-2,490 µatm) and the average diel amplitudes were 0.23 and 1,245 µatm. Possibly influenced by rainfall (early morning in the day of 23rd August), these diel amplitudes showed slight de­creases when compared to those at GCQ. The diel variation of pCO2 was the opposite of that of pH, showing a decrease during daytime, with the min­imum value of 946µatm. The dissolved CO2 partial pres­sure rose at night, with the maximum value of 2,487 µatm and -the diel amplitude amounts of 1,220-1,270 µatm (Fig. 4). Due to the consumption of CO2 by photosyn­thesis in the morning at 7:00, pCO2 decreased and the water gradually became supersaturated, suggesting that Ca began to precipitate. After 17:00, pCO2 gradually rose again and reached a relatively constant high value after midnight. The values of SIc were larger than 0, indicat­ing that the water stayed in the state of supersaturation. The values showed a daytime increase and nighttime de­crease cycling, with the maximum of 1.12 and the mini­mum of 0.85. SIc rose steadily in the morning from 8:00, with an associated decrease of Ca2+ and HCO3- (Fig. 3). The maximum value occurred around 14:00 and then decreased slowly during 14:00-17:00, corresponding to the constant low values of Ca2+ and HCO3-. After that, SIc sharply dropped while concentration of Ca2+ and HCO3- sharply increased. After 22:00, SIc reached a relatively constant low value but Ca2+ and HCO3- increase gradu­ally. Tab. 2: Diel variations of water parameters at the Guancun Bridge monitoring site and the Leiya Village monitoring site. Index Spring(LY) Guancun Bridge(GCQ) Leiya Village(LY) Note Mean Min Max Max-Min Mean Min Max Max-Min Mean Twater (°C) 21.3 21.5 27.5 6.0 23.7 21.8 28.0 6.2 24.3 monitoring data on August 22, 2011 SpC (µS cm-1) 418 412 452 40 438 395 454 59 433 pH 7.5 7.9 8.3 0.4 8.0 8.0 8.3 0.3 8.1 DO (mg L-1) 7.18 6.69 11.85 5.16 8.51 6.61 10.59 3.98 8.15 Ca2+ (mg L-1) 84.8 73.0 87.1 14.1 80.7 69.0 84.9 15.9 77.8 monitoring data on August 25, 2011 Mg2+ (mg L-1) 10.2 10.3 10.7 0.5 10.5 10.1 10.5 0.4 10.3 HCO3- (mg -1L) 272.3 222.6 279.5 56.9 259.1 227.8 278.5 50.6 254.7 .13CDIC (‰) -15.31 -14.55 -11.95 2.60 -13.50 -14.40 -11.30 3.10 -13.21 SIc 0.32 0.74 1.07 0.33 0.88 0.85 1.12 0.27 0.96 pCO2 (µatm) 8494 1137 3355 2218 2305 946 2166 1220 1652 Cheng Zhang, Jinliang Wang, Jun Yan & Jianguo Pei Fig. 2: Diel variations in water temperature, pH, and DO. CK-the outlet monitoring site of the underground stream (spring). GCQ-the Guancun Bridge monitoring site. LY-the Leiya Village monitoring site. The grey areas represent nighttime. Fig. 3: Diel variations in HCO3-, Ca2+ and .13CDIC. Diel cycling and flux of HCO3- in a typical karst spring-fed stream of southwestern China Fig. 4: Diel variations in SIc and pCO2. Cheng Zhang, Jinliang Wang, Jun Yan & Jianguo Pei Discussion Variations in Water Temperature, pH, and DO Karst groundwater is characterized by high alkalinity as a result of high concentration of HCO3- and Ca2+. In this weakly alkaline environment, the geochemical processes that influence the diel variations of river hydrochemistry mainly include: photosynthesis and respiration of aquat­ic plants and algae (Odum 1956), biomass, heat exchange at the water-gas interface, water pCO2 values and related CO2 degassing, etc. During daytime, high air temperature and strong daylight make rise of the water temperature. Aquatic plants consume CO2 and produce O2 with pho­tosynthesis as the dominant process. pH value, DO, and redox potential (Eh) tend to increase and the nitrifica­tion process is enhanced accordingly (ammonium nitro­gen transform into nitrate nitrogen). During nighttime, air temperature drops and aquatic plants assimilate O2 and release CO2 as a result of respiration. pH value, DO, and Eh tend to decrease and the denitrification process is enhanced accordingly (that is, nitrate nitrogen con­verted into molecular nitrogen) (Brick & Moore 1996; Burns 1998; Gammons et al. 2011; Grimm 1987; Hayashi et al. 2012; Heffernan & Cohen 2010; Hessen et al. 1997; Johnson & Tank 2009, Manny & Wetzel 1973; Mulhol­land 1992; Roberts & Mulholland 2007; Rusjan & Mikoš 2010; Scholefield et al. 2005). In Fig. 2, the diel variations in stream water tem­perature are mainly associated with the heat transfer over the water-gas interface. Water temperature rises with high air temperature in daytime. The pH value and DO vary regularly on a daily basis, suggesting that the photosynthesis of aquatic plants and algae is indeed the controlling factor for the diel variations of the pH and DO of the stream at the Guancun Village. The temperature of stream water changes on a daily basis as a result of higher air temperature and direct im­pact of solar radiation and other factors (Nimick et al. 2011). Generally speaking, the amplitudes of diel varia­tions of water temperature tend to be larger during sum­mer, during lower .ows, and in streams that are wider and shallower (Nimick et al. 2005). Thus compared to surface streams in the usual sense, the hyporheic dis­charge commonly buffers the diel range in surface stream water temperature due to the relatively constant temper­atures of shallow groundwater (Arrigoni et al. 2008). The pH value influences not only carbonate geobio­chemistry, but also broad variety of other abiotic pro­cesses. The diel variations in pH value thus lead to the diel variations in a series of hydrochemical parameters. At GCQ, the amplitude of diel pH variations was 0.4. The amplitudes of diel pH variations typically are less than 1 pH unit and increase as season warms, reported by Nimick et al. (2005). Whereas, during summer low-.ow conditions in streams rich in aquatic plants, ampli­tudes of diel pH variations can be as much as 2 pH units (Jarvie et al. 2000). Variations in DIC and Carbon Isotopes During the monitoring period, the DIC input (CK) was relatively stable of about 272.3 mg L-1 (4.46 mmol L-1). The concentrations of HCO3- and Ca2+ at GCQ and LY showed a diel cycle of daytime decrease and nighttime increase stagnation, with the amplitude of 22.4 % (Zhang et al. 2015). The minimum values were 222.6 mg L-1 and 67.6 mg L-1 respectively, observed during the period of 14:00 to 15:00. At night, the ion concentration levels cor­responding to groundwater recharge recurred, and the values remained high (Fig. 3). The DIC content in the karst river waters is mainly in the form of HCO3-. The amplitude of DIC of the stream in Guancun Village was 22.4 %. This value is reasonable as it was found in the existing studies that the amplitude could reach as high as 30 % (Nagorski et al. 2003; Wal­dron et al. 2007; Poulson et al. 2010). The daytime de­crease nadir and the nighttime wide stagnation of Ca2+ and HCO3- correspond to the daytime rise stagnation and the nighttime wide nadir of pH and DO, showing a remarkable inversed correlation. This indicates that when stream water temperature rises up, aquatic veg­etation photosynthesis consumes HCO3-, leading to de­crease in DIC content and CaCO3 precipitation. This is supported by the previous observations that when the dissolved CO2 in waters is not available for carbon se­questration by aquatic plants (for photosynthesis), near­ly half of aquatic plants would take carbon directly from the HCO3- in water for photosynthesis (Axelsson et al. 1999; Larsson et al. 1999). In this study, the maximum diel amplitude of the isotope was about 3.10 ‰. This value is not only less than 4.5 ‰, the maximum diel amplitude of the .13CDIC in surface water according to the study by Parker et al. (2010), but also less than 5 ‰, the amplitude of .13CDIC variation as a result of CO2 degassing in karst springs reported by Michaelis et al. (1985) and surface streams fed by karst water (Doctor et al. 2008). As shown in Fig. 5, a significantly positive correlation exists between the reciprocal of DIC and the carbon isotope .13CDIC, in­dicating that the carbon loss from streams during day­time is prominently controlled by biological processes, i.e. aquatic vegetation photosynthesis and calcification. Photosynthesis consumes dissolved CO2 or (HCO3-+H+) and shifts the following chemical equilibrium (Eq.(2)) to the right (Hayashi et al. 2012; McConnaughey 1998). As a result, the activity of H+ decreases and pH increases, accompanied by synchronous increases in DO concen­tration. H+ + HCO3- ‹› H2CO3(aq) ‹› H2O + CO2(g) › CH2O + O2 (2) Aquatic plants use bicarbonate for photosynthesis by conversion of HCO3- into CO2 with H+ depletion. H+ can be generated through extracellular or intracellular acid secretion, or ATPase powered calcification (McCon­naughey 1998) (reaction 3). Ca2+ + CO2+ H2O › CaCO3(s) + 2H+ (3) Adding reaction 2 and 3 yields the following result: Ca2++ 2HCO3- › CaCO3(s) + CH2O + O2 (4) Accordingly, in mildly alkaline karst water, where HCO3- is the dominant species of DIC, calcification might protonate HCO3- to CO2. The two H+ derived from calcification protonate 2 HCO3- to produce 2 CO2. One CO2 is used for calcification, leaving one for photosyn­thesis (McConnaughey 1998), thus resulting in-stream diel variations of concentrations of Ca2+ and HCO3-. The studies on the characteristics of the temporal and spatial variations of .13CDIC value can help evaluate degassing, biological processes, carbonate dissolution and precipitation processes with respect to ecosystems (De Montety et al. 2011; Finlay 2003; Gammons et al. 2011; Jiang et al. 2013; Parker et al. 2005, 2007, 2010; Poulson & Sullivan 2010; Smith et al. 2011; Spiro & Pen­tecost 1991; Waldron et al. 2007). Jiang et al. (2013) stated that aquatic plants raise .13CDIC by using DIC, the same as the effect of degassing. But degassing usually leads to only a small increase in .13CDIC. As a result, respiration is the dominant factor causing the decrease of .13CDIC. CO2 Fluxes at the Water-Gas Interface The values of pCO2 of the waters at the outlet of the un­derground stream were high, with the average daily value of 8,494 µatm (Tab. 3). After the underground stream ris­ing up and discharging into downstream of spring, the calculated pCO2 gradually decreased with flow distances (Arrow I in Fig. 6). At 880 m away from the outlet, the pCO2 dropped to 2,565 µatm, with the decreasing am­plitude of 67.4 µatm every 10 m. At 1,350 m, it dropped to only 1,844 µatm, with the decreasing amplitude of 15.3 µatm every 10 m, indicating that the decreasing rate of pCO2 reduces with distance. The pCO2 value exhibits a significant inversed correlation with DO at both GCQ and LY. The sharp rise of DO during daytime leads to the significant drop of pCO2, indicating that the diel varia­tions are in fact dominated by aquatic vegetation photo­synthesis (Arrow II in Fig. 6). As shown in Tab. 3, the calculated pCO2 value of water at the outlet of the underground stream was much larger than the atmospheric pCO2 concentra­tion (390 ppm), suggesting the high water-gas ex­change flux (FCO2). According to the calculation us­ing Equation 1 (k=0.8), the value of FCO2 through the water-gas interface at the outlet of the underground stream was as high as 289.43 mmol m-2 d-1, while those at GCQ and LY dropped to 77.71 mmol m-2 d-1 and 51.94 mmol m-2 d-1, respectively, with the decreasing amplitudes of 2.41 mmol m-2 d-1 and 0.55 mmol m-2 d-1 every 10 m respectively. This shows that the CO2 degassing of the surface stream fed by karst groundwa­ter mainly occurred in the upper reaches of the surface stream right after groundwater exposed to the surface. The values (at GCQ and LY) of FCO2 at the water-gas interface of a low order stream is similar to the average value in tropics (79.5 mmol m-2 d-1), but higher than the average value in a temperate climate (31.8 mmol m-2 d-1) (St. Louis et al. 2000), and much higher than the values in Loch Ness in Scotland (12.3 mmol m-2 d-1, Jones et al. (2001)), and the Hongfeng Lake in Guizhou Province of China (13.2 mmol m-2 d-1, Wang et al. (2012)) as well as the global average value (~16.2 mmol m-2 d-1, Cole et al. (1994)). According to Equation 1 and the above two de­creasing rates with flow distance, the FCO2 at the water-gas interface of the stream section per 10 m could also be calculated (if the average river width is assumed to be 3.5 m). From this, the total CO2 exchange flux of the en­tire monitoring stream section (from CK to LY) was cal­culated to be 29.83 kg d-1 (677.95 mol d-1). Considering the fact that the degassing mainly occurred around the outlet of underground streams, this exchange flux calcu­lated on the basis of the equivalent pCO2 decreasing am­plitudes could be possibly larger than the actual value. Carbon Fluxes and CO2 Degassing The groundwater (at CK) originating from karst areas was saturated with respect to calcite (Fig. 6). It remained saturated with respect to calcite and has high partial CO2 pressure so it can lead to calcium carbonate precipitation and CO2 escape (pH rise) (Arrow I in Fig. 7). The aver­age daily Ca2+ declined from 4.24 mmol L-1 (at CK) to 4.04 mmol L-1 (at GCQ) and 3.89 mmol L-1 (at LY), an average decrease of 0.023-0.032 mmol L-1 every 100 m. The pH-HCO3- values of the stream waters at both GCQ and LY had their projection points all above the dolomite saturation curve, indicating that the water was supersaturated regarding to calcite and dolomite (Fig. 7). The pH and HCO3- values at both GCQ and LY showed significant diel variations. Influenced by aquatic vegeta­tion photosynthesis, the pH increased while the HCO3- decreased during daytime. Influenced by both the CO2 produced by respiration and the recharge of groundwater rich in HCO3-, the pH decreased while HCO3- increased during nighttime (Arrow II in Fig. 7). The HCO3- content of the groundwater at CK was 4.46 mmol L-1. Along the flow path, it dropped to 4.25 mmol L-1 (at GCQ) and 4.18 mmol L-1 (at LY) as a result of CO2 degassing, an average decrease of 0.015-0.024 mmol/L every 100 m. During daytime, the HCO3- decrease reached 0.83-0.93 mmol/L at both GCQ and LY. In addition, as shown in Fig. 7 the HCO3- variations caused by the process II were much greater than those by the process I, suggesting that the HCO3- consumption of rivers is mainly influenced by aquatic vegetation photosynthesis (Equation 4) and to less extent by degassing. The positive correlation be­tween the partial CO2 pressure and the DO in the lower reaches is in agreement with this (Fig. 6). Based on the calculation of the bicarbonate fluxes at CK (input) and LY (output), the reduction of bicar­bonate in water influenced by biological processes can be described in mass balance terms (in unit of kg d-1). Dur­ing the monitoring period, the average discharge was 152.9 L s-1 (Wang et al. 2012). Calculated by the HCO3- content of water samples and the average discharge, the input amount of dissolved inorganic carbon by the un­derground stream was 3,597.5 kg HCO3- d-1 (Tab. 4). The output amount of dissolved inorganic carbon at LY was 3,365.2 kg HCO3- d-1. Along the flow path and downstream from CK to LY, the loss of dissolved inorganic carbon was 232.3 kg HCO3- d-1, namely 172.1 g HCO3- m-1 d-1, about 45.7 kg C d-1. This shows that, dominated by pho­tosynthesis and calcification (McConnaughey 1998), in­organic carbon was indeed converted partly to organic carbon downstream along the flow path, which should be considered as a natural carbon sink as a part of the carbon flux of water. The CO2 exchange flux at the water-gas interface of the monitoring stream section from CK to LY was 677.95 mol d-1, equivalent to 8.14 kg C d-1, accounting for 17.8 % of the DIC loss. This indicates that the varia­tion of inorganic carbon in the stream is mainly caused by ecosystem activity, namely aquatic vegetation photo­synthesis and calcification, accounting for 21.6 % and 60.6 % of the loss amount, respectively (Tab. 4). Thus ap­proximately 4/5 of the loss was converted into organic carbon and inorganic carbon in form of Ca precipitation. Eventually they would be stored in streambed sediments and form karst carbon sink. Compared with the total carbon input by the under­ground stream, this carbon loss only accounts for 6.5 % of the total carbon amount (1.4 % of which was convert­ed into organic carbon and 1.1 % of which was degassed to the atmosphere), indicating that the DIC of the karst groundwater in low order surface steams is stable in gen­eral, with roughly 1% being lost to the atmosphere. This finding is similar to the results from the karst spring fed surface rivers in Florida (De Montety et al. 2011). The impact of CO2 degassing on DIC variations can be negli­gible on a daily scale as the loss induced by degassing in this study is much lower than the carbon flux on a catch­ment scale. This suggests that carbon loss in the form of CO2 of low-order Guangcun karst stream is not signifi­cant on a daily timescale (Zavadlav et al. 2013). Diel cycling and flux of HCO3- in a typical karst spring-fed stream of southwestern China Cheng Zhang, Jinliang Wang, Jun Yan & Jianguo Pei Fig. 5: Relation between .13CDIC and 1/DIC. Tab. 3: The partial CO2 pressures in water and CO2 fluxes at the water-gas interface at the monitoring sites. Monitoring site Flow distance (m) pCO2 (µatm) .pCO2 (µatm/10m) FCO2 (mmol/m2.d) .FCO2 (mmol/m2.d.10m) CK 0 8494 289.43 67.4 2.41 GCQ 880 2565 77.71 15.3 0.55 LY 1350 1844 51.94 Diel cycling and flux of HCO3- in a typical karst spring-fed stream of southwestern China Fig. 6: Relation between dissolved oxygen and CO2 partial pres­sure. The arrow I shows the drop of partial CO2 pressure due to the CO2 degassing and the two-headed arrow II indicates the diel variations in pCO2 and DO due to biotic processes. Tab. 4: The C mass balance of the monitoring stream profile. Ca2+ HCO3- degassing bio uptake mg L-1 CK 84.8 272.3 GCQ 80.7 259.1 LY 77.8 254.7 kg d-1 CK 1120.1 3597.5 GCQ 1066.4 3423.3 LY 1027.7 3365.2 kmoles d-1 CK 28.0 59.0 GCQ 26.7 56.1 LY 25.7 55.2 mg L-1 CK-LY 7.0 17.6 kg d-1 CK-LY 92.3 232.3 kmoles d-1 CK-LY 2.3 3.8 kgC d-1 CK-LY 27.7 45.7 8.1 9.9 % respect to DIC loss 60.6 17.8 21.6 100.0 % respect to initial DIC 3.9 1.1 1.4 6.5 precipitation degassing bio-uptake Total Cheng Zhang, Jinliang Wang, Jun Yan & Jianguo Pei Fig. 7: Relation between pH and HCO3- concentrations in stream water. The two bold dashed lines indicate the saturating curves of calcite and dolomite respectively. The fine dashed line represents the equilibrium line of CO2 (–log pCO2), and the arrow I and the two-headed arrow II indicate the variations of pH and HCO3- along the flow path and daily respectively. Diel cycling and flux of HCO3- in a typical karst spring-fed stream of southwestern China Conclusion Results from high-resolution data logger monitoring and high frequency sampling indicated that the changes in aqueous chemistry of spring-fed stream in Guancun are closely associated with biogeochemical processes. pH, SpC, DO, HCO3- and .13CDIC all showed diel variations, reflecting strong influence of photosynthesis and calcite precipitation. The concentrations of HCO3- and Ca2+ at downstream showed a diel cycle of daytime decrease and nighttime increase, with an amplitude of 22.4 %. Diel DIC change indicate that the stream is losing inorganic carbon along its flow path, the daily loss of Ca2+ and DIC are estimated to be about 92.3 kg d-1 and 232.3 kg d-1 (namely 45.7 kg C d-1), respectively along the 1350 m of the Guancun River. The total CO2 exchange flux of the entire monitoring stream profile was calculated to be 8.14 kg C d-1, accounting for 17.8 % of the DIC loss. This indicates that the variation of inorganic carbon in the stream is mainly caused by ecosystem activity, namely aquatic vegetation photosynthesis and calcification, ac­counting for 21.6 % and 60.6 % of the loss amount, re­spectively. Compared with the total carbon input by the underground stream, this carbon loss only accounts for 6.5 % of the total carbon amount (1.4 % of which was converted into organic carbon and 1.1 % of which was degassed to the atmosphere), suggesting that carbon loss in the form of CO2 of low-order Guangcun karst stream is not significant on a daily timescale. Acknowledgement This work was supported by the Special Fund for Pub­lic Benefit Scientific Research of Ministry of Land and Resources of China (No. 201111022), IGCP/SIDA 598, fund from the Department of Science and Technology of Guangxi (15-140-09) and the China Geological Survey Projects (No. 12120114006301 and 12120113006700). Special thanks are given to Dr. Lu Qian for her valuable comments and suggestions on the original manuscript, two anonymous reviewers whose constructive comments and suggestions have greatly improved this manuscript. References Arrigoni, A.S., Poole, G.C., Mertes, L.A.K., O’Daniel, S.J., Woessner, W.W. & S.A. Thomas, 2008: Buff­ered, lagged, or cooled? Disentangling hyporheic influences on temperature cycles in stream chan­nels.- Water Recourses Research, 44, W09418. doi10.1029/2007WR006480. Axelsson, L., Larsson, C. & H. Ryberg, 1999: Affinity, ca­pacity and oxygen sensitivity of two different mech­anisms for bicarbonate utilization in Ulva lactuca L.(Chlorophyta).- Plant, Cell and Environment, 22, 969-978. Brick, C.M. & J.N. Moore, 1996: Diel variations in the upper Clark Fork River, Montana.- Environmental Science and Technology, 30, 1953-1960. Burns, D.A., 1998: Retention of NO3- in an upland stream environment: a mass balance approach.- Biogeo­chemistry, 40, 73-96. Chen, X., Zeng, Y., Jian J., Lu Y. & Z. Wu, 2012: Genetic diversity and quantification of aerobic anoxygenic phishanototrophic bacteria in Hugangyan Maar Lake based on pufM DNA and mRNA analysis.- Microbiology China, 39(11), 1560-1572. Cole, J.J., Caraco, N.F., Kling, G.W. & T.K. Kratz, 1994: Carbon dioxide supersaturation in the surface wa­ters of lakes.- Science, 265, 1568-1570. Dandurand, J.L., Gout, R., Hoefs, J., Menschel, G., Schott, J. & E. Usdowski, 1982: Kinetically controlled varia­tions of major components and carbon and oxygen isotopes in a calcite-precipitating spring.- Chemical Geollogy, 36, 299-315. Doctor, D.H., Kendall, C., Sebestyen, S.D., Shanley, J.B., Ohte, N. & E.W. Boyer, 2008: Carbon isotope frac­tionation of dissolved inorganic carbon (DIC) due to outgassing of carbon dioxide from a headwater stream.- Hydrology Processes, 22, 2410-2423. Finlay, J.C., 2003: Controls of streamwater dissolved in­organic carbon dynamics in a forested watershed.- Biogeochemistry, 62, 231-252. Galy-Lacaux, C., Delmas, R., Jambert, C., Dumestre, J., Labroue, L., Richard, S. & P. Gosse, 1997: Gaseous emissions and oxygen consumption in hydroelec­tric dams: a case study in French Guiyana.- Global Biogeochemical Cycles, 11, 471-483. Gammons, C.H., Babcock, J.N., Parker, S.R. & S.R. Poul­son, 2011: Diel cycling and stable isotopes of dis­solved oxygen, dissolved inorganic carbon, and nitrogenous species in a stream receiving treated municipal sewage.- Chemical Geology, 283, 44-55. Grimm, N.B., 1987: Nitrogen dynamics during succes­sion in a desert stream.- Ecology, 68, 1157-1170. Guasch, H., Armengol, J., Martí, E. & S. Sabater, 1998: Diurnal variation in dissolved oxygen and carbon dioxide in two low-order streams.- Water Research, 32, 1067-1074. Hayashi, M., Vogt, T., Mahler, L., Mächler, L. & M. Schirmer, 2012: Diurnal fluctuations of electrical conductivity in a pre-alpine river: Effects of pho­tosynthesis and groundwater exchange.- Journal of Hydrology, 450-451, 93-104. Heffernan, J.B. & M.J. Cohen, 2010: Direct and indirect coupling of primary production and diel nitrate dy­namics in a large spring-fed river.- Limnology and oceanography, 55, 677-688. Hessen, D.O., Henriksen, A. & A.M. Smelhus, 1997: Seasonal fluctuations and diurnal oscillations in nitrate of a heathland brook.- Water Resources, 31, 1813-1817. He, Y., Zeng, Y., Yuan, B., Liu, H. & F. Feng, 2010: Phy­logenetic diversity of aerobic anoxygenic pho­totrophic bacteria in eutrophic zone of Lake Ulan­suhai based on gene pufM.- Microbiology China, 37(8), 1138-1145. Johnson, L.T. & J.L. Tank, 2009: Diurnal variations in dissolved organic matter and ammonium uptake in six open-canopy streams.- Journal of the North American Benthological Society, 28, 694-708. Jarvie, E.H.P., Neal, C., Tappin, A.D., Burton, J.D., Hill, L., Neal, M., Harrow, M., Hopkins, R., Watts, C. & H. Wickham, 2000: Riverine inputs of major ions and trace elements to the tidal reaches of the Riv­er Tweed, UK.- Science of the Total Environment, 251/252, 55-81. Jiang Z., Yuan D., Cao J., Qin X., He S. & C. Zhang, 2012: A study of carbon sink capacity of karst processes in China.- Acta Geoscientica Sinica, 2012, 33(2), 129-134. Jiang, Y. Hu, Y. & M. Schirmer, 2013: Biogeochemical controls on daily cycling of hydrochemistry and .13C of dissolved inorganic carbon in a karst spring-fed pool.- Journal of Hydrology, 478, 157-168. Jiao N., Tang K., Zhang Y., Zhang R., Xu D. & Q. Zheng, 2013: Microbial processes and mechanisms in car­bon sequestration in the ocean.- Microbiology Chi­na, 40(1), 71-86. Jones, R.I., Grey, J., Quarmby, C. & D. Sleep, 2001: Sources and fluxes of inorganic carbon in a deep, oligotrophic lake (Loch Ness, Scotland).- Global Biogeochemical Cycles, 15, 863-870. Kanduc, T., Szramek, K., Ogrinc, N. & L.M. Walter, 2007: Origin and cycling of riverine inorganic carbon in the Sava River watershed (Slovenia) inferred from major solutes and stable carbon isotopes.- Biogeo­chemistry, 86, 137-154. Larsson, C. & L. Axelsson, 1999: Bicarbonate uptake and utilization in marine macroalgae.- European Jour­nal of Phycology, 34, 79-86. Lin, H., 2010: Earth’s critical zone and hydropedology: concepts, characteristics, and advances.- Hydrology and Earth System Sciences, 14, 25-45. Liu, Z., Li, Q., Sun, H., Liao, C., Li, H., Wang, J. & K. Wu, 2006: Diurnal variations of hydrochemistry in a travertine-depositing stream at Baishuitai, Yunnan, SW China.- Aquatic Geochemistry, 12, 103-121. Liu, Z., Liu, X. & C. Liao, 2008: Daytime deposition and nighttime dissolution of calcium carbonate con­trolled by submerged plants in a karst spring-fed pool: insights from high time-resolution monitor­ing of physico-chemistry of water.- Environmental Geology, 55, 1159-1168. Lorah, M.M. & J.S. Herman, 1988: The chemical evolu­tion of a travertine-depositing stream: geochemical processes and mass transfer reactions.- Water Re­sources Research, 24, 1541-1552. Madsen, T.V., 1983: Growth and photosynthetic acclima­tion by ranunculus aqutuatilis L. in response to in­organic carbon availability.- New Phytology, 125(4), 707- 715. Manny, B.A. & R.G. Wetzel, 1973: Diurnal changes in dissolved organic and inorganic carbon and nitro­gen in a hardwater stream.- Freshwater Biology, 3, 31-43. Martin J.B., Brown A. & J. Ezell, 2013: Do carbonate karst terrains affect the global carbon cycle?.- Acta Carsologica, 42/2-3, 187-196. McConnaughey, T., 1998: Acid secretion, calcifica­tion, and photosynthetic carbon concentrating mechanisms.- Canadian Journal of Botany, 76, 1119-1126. Mermillod-Blondin, F., Simon, L., Maazouzi, C., Foul­quier, A., Delolme, C. & P. Marmonier, 2015: Dy­namics of dissolved organic carbon (DOC) through stormwater basins designed for groundwater re­charge in urban area: Assessment of retention effi­ciency.- Water Research, 81, 27-37. Michaelis, J., Usdowski, E. & G. Menschel, 1985: Parti­tioning of 13C and 12C on the degassing of CO2 and the precipitation of calcite—Rayleightype fraction­ation and a kinetic model.- American Journal of Science, 285, 318-327. De Montety, V., Martin, J.B., Cohen, M.J., Foster, C. & M.J. Kurz, 2011: Influence of diel biogeochemical cycles on carbonate equilibrium in a karst river.- Chemical Geology, 283, 31-43. Mulholland, P.J., 1992: Regulation of nutrient concentra­tions in a temperate forest stream—roles of upland, riparian, and instream processes.- Limnology and oceanography, 37, 1512-1526. Nagorski, S.A., Moore, J.J., Mclinnon, T.E. & D.B. Smith, 2003: Scale-dependent temporal variations in stream water geochemistry.- Environmental Sci­ence and Technology, 37, 859-864. Nimick, D.A., Cleasby, T.E. & R.B. McCleskey, 2005: Sea­sonality of diel cycles of dissolved trace-metal con­centrations in a Rocky Mountain stream.- Environ­mental Geology, 47, 603-614. Nimick, D.A., Gammons, C.H. & S.R. Parker, 2011: Diel biogeochemical processes and their effect on the aqueous chemistry of streams: A review.- Chemical Geology, 283, 3-17. Odum, H.T., 1956: Primary production in flowing wa­ters.- Limnology and Oceanography, 1, 102–117. Parker, S.R., Poulson, S.R., Gammons, C.H. & M.D. DeGrandpre, 2005: Biogeochemical controls on diel cycling of stable isotopes of dissolved O2 and dissolved inorganic carbon in the Big Hole River, Montana.- Environmental Science and Technology, 39, 7134-7140. Parker, S.R., Gammons, C.H., Poulson, S.R. & M.D. De­Grandpre, 2007: Diel variations in stream chemis­try and isotopic composition of dissolved inorganic carbon, upper Clark Fork River, Montana, USA.- Applied Geochemistry, 22, 1329-1343. Parker, S.R., Poulson, S.R., Smith, M.G., Weyer, C.L. & K.M. Bates, 2010: Temporal variability in the con­centration and stable carbon isotope composition of dissolved inorganic and organic carbon in two Montana, USA rivers.- Aquatic Geochemistry, 16, 61-84. Poulson, S.R. & A.B. Sullivan, 2010: Assessment of diel chemical and isotopic techniques to investigate bio­geochemical cycles in the upper Klamath River, Or­egon, USA.- Chemical Geology, 269, 3-11. Reichert, P., 2001: River water quality model no. 1 (RWQM1): case study II. Oxygen and nitrogen con­version processes in the River Glatt (Switzerland).- Water Science and Technology, 43, 51-60. Roberts, B.J. & P.J. Mulholland, 2007: In-stream biotic control on nutrient biogeochemistry in a forested stream, West Fork of Walker Branch.- Journal of Geophysical Research, 112, G04002. http://dx.doi.org/10.1029/2007JG000422. Rusjan, S. & M. Mikoš, 2010: Seasonal variability of di­urnal in-stream nitrate concentration oscillations under hydrologically stable conditions.- Biogeo­chemistry, 97, 123-140. Schmidt, U. & R. Conrad, 1993: Hydrogen, carbon mon­oxide, and methane dynamics in Lake Constance.- Limnology and oceanography, 38, 1214-1226. Scholefield, D., Le Goff, T., Braven, J., Ebdon, L., Long, T. & Butler, M., 2005: Concerted diurnal patterns in riverine nutrient concentrations and physical conditions.- Science of the Total Environment, 344, 201-210. Smith, M.G., Parker, S.R., Gammons, C.H., Poulson, S.R. & F.R. Hauer, 2011: Tracing dissolved O2 and dis­solved inorganic carbon stable isotope dynamics in the Nyack aquifer: Middle Fork Flathead River, Montana, USA.- Geochimica et Cosmochimica Acta, 75, 5971-5986. Spiro, B. & A. Pentecost, 1991: One day in the life of a stream-a diurnal inorganic carbon mass balance for travertine-depositing stream (Waterfall Beck, York­shire).- Geomicrobiology Journal, 9, 1-11. St. Louis, V.L., Kelly, C.A., Duchemin, E., Rudd, J.W.M. & D.M. Rosenberg, 2000: Reservoir surfaces as sources of greenhouse gases to the atmosphere: a global estimate.- Bioscience, 50, 766-775. Tobias, C. & J.K. Böhlke, 2011: Biological and geochemi­cal controls on diel dissolved inorganic carbon cycling in a low-order agricultural stream: Impli­cations for reach scales and beyond.- Chemical Ge­ology, 283, 18-30. Waldron, S., Scott, E.M. & C. Soulsby, 2007: Stable iso­tope analysis reveals lower-order river dissolved inorganic carbon pools are highly dynamic.- Envi­ronmental Science and Technology, 41, 6156-6162. Wang, S., Yeager, K.M., Wan G., Liu C., Wang Y. & Y. Lü, 2012: Carbon export and HCO3- fate in carbon­ate catchments: A case study in the karst plateau of southwestern China.- Applied Geochemistry, 27, 64-72. White, W.B., 2013: Carbon fluxes in karst aquifers: sources, sinks, and the effect of storm flow.- Acta Carsologica, 42/2-3, 177-186. Wigley, T.M.L., 1977: WATSPEC-a computer program for determining the equilibrium of aqueous solu­tions.- British Geomorphology Research Group Technical Bulletin, 20, 1-46. Yang, J. & C. Zhang, 2014: Earth’s critical zone: a holistic framework for geo-environmental researches.- Hy­drogeology & Engineering Geology, 41(3), 98-104. Zavadlav S., Kanduc T., McIntosh J. & S. Lojen, 2013: Isotopic and chemical constraints on the biogeo­chemistry of dissolved inorganic carbon and chem­ical weathering in the karst watershed of Krka River (Slovenia).- Aquatic Geochemistry, 19, 209-230. Zhang C., 2011: Time-scale of karst processes and the carbon sink stability.- Carsologica Sinica, 30(4), 368-371. Zhang C., Wang J., Pu J. & J. Yan, 2012: Bicarbonate daily variations in a karst river: the carbon sink effect of subaquatic vegetation photosynthesis.- Acta Geo­logica Sinica (English Edition), 86(4), 973-979. Zhang C., Wang J. & J. Pu, 2015: Diel aqueous chemical cycling in a typical karst spring-fed stream: controls of biogeochemical processes.- Acta Geoscientica Si­nica, 36(2), 197-203. Cheng Zhang, Jinliang Wang, Jun Yan & Jianguo Pei Diel cycling and flux of HCO3- in a typical karst spring-fed stream of southwestern China Cheng Zhang, Jinliang Wang, Jun Yan & Jianguo Pei