SEARCH

SEARCH BY CITATION

Keywords:

  • Calcite dendrites;
  • growth cycles;
  • hot spring;
  • isotopes;
  • Tengchong

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Geological setting
  5. Methodology
  6. Terminology
  7. Results
  8. Interpretations
  9. Discussion
  10. Conclusions
  11. Acknowledgements
  12. References

Hot springs at Gongxiaoshe and Zhuyuan (maximum temperatures of 73 to 84°C, respectively) are characterized by deposits formed of calcite, aragonite, non-crystalline Si–Mg–Fe deposits, and minor amounts of barite and gypsum. The deposits at Gongxiaoshe are formed largely of alternating calcite and aragonite laminae, whereas those at Zhuyuan are formed largely of calcite. The calcite is in the form of: (i) pseudodendrites that grew as sub-crystals stacked upon each other; and (ii) unattached euhedral and incompletely formed dodecahedral and rhombohedral crystals. Amorphous CaCO3, formed of nanoparticles <1 μm long, is common in some of the Zhuyuan deposits, but minor in the Gongxiaoshe deposits. The morphologically diverse arrays of aragonite crystals that lie parallel to bedding were not nucleated on a growth surface. Many substrates in these deposits are covered with reticulate coatings that are formed largely of Si and Mg with minor Fe and micro-granular coatings that are formed largely of Si and Fe. Biofilms, with their extracellular polymeric substances, and microbes are common at both springs. The compositionally and crystallographically diverse precipitates at these two springs are attributed to a biologically influenced model with precipitation taking place in micro-domains that developed in the extracellular polymeric substances. According to this model, precipitation varied at the micron-scale influenced by the elemental concentrations that developed in the hydrogel of extracellular polymeric substances. Critically, the very low preservation potential of the extracellular polymeric substance and its formative microbes means that the precipitates will rapidly lose evidence of their biotic origin. The compositional diversity of the precipitates, the crystallographic diversity of the calcite and aragonite with numerous incompletely formed crystals, and local concentrations of Si, Mg and Fe may, however, serve as proxies of that biologically influenced precipitation.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Geological setting
  5. Methodology
  6. Terminology
  7. Results
  8. Interpretations
  9. Discussion
  10. Conclusions
  11. Acknowledgements
  12. References

Hot springs with water temperatures above 70°C, found in geothermal areas throughout the world, are commonly the sites of spectacular silica and/or calcite/aragonite deposits. Although precipitation from such hot waters was commonly thought to be entirely abiotic, it is now becoming increasingly apparent that these springs are inhabited by diverse extremophilic microbial biota (Reysenbach et al., 2000; Blank et al., 2002; Hetzer et al., 2007). There is also growing evidence that many of these microbes are involved, in some manner, in the development of the opal-A precipitates that form in many of these springs (Cady & Farmer, 1996; Jones et al., 1997; Inagaki et al., 2001; Guidry & Chafetz, 2003; Peng et al., 2007). Such assessments are possible because many of these microbes are rapidly silicified and preserved in these deposits (Jones et al., 2004).

Spring deposits formed largely of calcite and/or aragonite are known from geothermal areas in Italy (Chafetz & Folk, 1984; Folk, 1994; Guo & Riding, 1998, 1999), the Kenya Rift Valley (Jones & Renaut, 1995, 1998; Renaut & Jones, 1997), Pamukkle, Turkey (Pentecost et al., 1997; Özkul et al., 2002), New Zealand (Jones et al., 2000; Campbell et al., 2002), Iceland (Jones et al., 2005) and Yellowstone National Park (Fouke et al., 2000; Chafetz & Guidry, 2003; Fouke, 2011). Many are formed of calcite and/or aragonite with complex crystal morphologies that include composite crystals and large (up to 10 cm long) dendrite crystals (Jones & Renaut, 1995; Jones et al., 2005). In stark contrast to the opal-A precipitates, microbes are rarely preserved in these deposits and their presence can only be inferred from textures evident in the calcite and aragonite, as has been suggested for bacterial shrubs (Chafetz & Guidry, 1999). After examining modern deposits in Yellowstone National Park, Fouke et al. (2000) concluded that calcite precipitation was largely controlled by CO2 degassing, temperature decreases and possibly evaporation rather than the microbial mats found in many parts of the depositional system. Fouke (2011) later acknowledged, however, that various biotic processes, mediated largely by the microbes that live on the discharge apron, also influenced precipitation and the textures developed in the spring deposits.

This article focuses on two hot springs (Gongxiaoshe and Zhuyuan), located in the Ruidian geothermal area in the western part of Yunnan Province, China (Fig. 1), where calcite, aragonite and various other precipitates are being actively deposited (Fig. 2). The presence of biofilms (Fig. 2B and C) opens the possibility that development of the mineralogically and crystallographically diverse deposits in these two springs may have been biologically mediated. Evaluation of this possibility is based on a detailed description of the constituent precipitates and associated biofilms. This research shows that these precipitates and their fabrics were probably controlled by the physicochemical conditions that existed in the micro-domains of the extracellular exopolymeric substances (EPS) that form an integral part of the biofilms.

image

Figure 1. Location of hot springs. Map showing location of (A) Tenchong in western Yunnan Province and (B) Ruidianxiang north of Tengchong. (C) Map of Ruidianxiang and surrounding area showing locations of Gongxiaoshe and Zhuyuan hot springs.

Download figure to PowerPoint

image

Figure 2. Gongxiaoshe and Zhuyuan hot springs. (A) General view of Gongxiaoshe. White letters ‘B’ and ‘C’ show locations of views shown in panels (B) and (C), respectively. Water temperature on these ledges is 73 to 75 °C. (B) Margin of Gongxiaoshe showing the lower and upper ledges flanking the main pool. Note the division of the lower ledge into inner and outer parts (arrows indicate the boundary) based on the extent of green microbial mat. (C) View of the pool margin showing the distinct dark green microbial mat (arrow) on the face of the upper ledge. (D) Zhuyuan hot spring–flowing water (arrow) at 73°C in the concrete channel lined with red precipitates on the wall beneath the hammer. (E) View of precipitates from the channel wall shown in panel (D). Note the distinct red colour and columns at the top. White letter ‘F’ indicates the position of panel (F). (F) Upper part of precipitates formed of columns. Water level is at change in colour. (G) Top of precipitates showing the narrow ledge covered with pillars.

Download figure to PowerPoint

Geological setting

  1. Top of page
  2. Abstract
  3. Introduction
  4. Geological setting
  5. Methodology
  6. Terminology
  7. Results
  8. Interpretations
  9. Discussion
  10. Conclusions
  11. Acknowledgements
  12. References

The Ruidian geothermal area, located in a remote area in the western part of Yunnan Province close to the border with Myanmar (Fig. 1B), is situated at the eastern end of the Tibet–Yunnan geothermal zone (Kearey & Wei, 1993; fig. 10). It is located on the northern margin of the Tengchong volcanic area (Zhang et al., 2008) which is part of the Tengchong block that developed along the eastern collision boundary between the Indian and Eurasian plates (Du et al., 2005). Until the early Cenozoic, this area experienced intense tectonic activity with numerous volcanoes and extensive faulting (Zhu & Tong, 1987; Du et al., 2005).

The Ruidian geothermal field, located 38 km north of Tengchong, includes numerous hot springs. Unfortunately, there is considerable confusion regarding the names of the individual springs. Named springs have been cited in the literature without precise locations, a situation complicated by the fact that many of these springs have changed names with time. Conversations with local inhabitants have failed to resolve this problem. The names used in this article are those that are in common use today, with attention being focused on Gongxiaoshe and Zhuyuan (Fig. 1C). The former may correspond to Gongxiaoshe as used by Zhang et al. (2008). The relation (if any) between the spring that Zhang et al. (2008) referred to as Jieming and the spring that is now called Zhuyuan remains unknown. Similarly, it has been impossible to determine whether any of the six springs named by Zhang et al. (1987) correspond in any way to the Gongxiaoshe and Jieming springs named in Zhang et al. (2008) or the Gongxiaoshe and Zhuyuan springs described in this article.

Gongxiaoshe (GPS location of 25°26′ 22·25″ N; 98°27′ 39·42″ E) is located in the middle of the village of Ruidianxiang (Figs 1C and 2A to C) with Zhuyuan (GPS location of 25°26′ 19·39″ N; 98°27′ 36·90″ E) being located ca 130 m to the south-west of Gongxiaoshe (Figs 1C and 2D to G). Although the villagers have modified both springs so that they can make use of the hot water, calcite and aragonite with various accessory precipitates are still actively forming in both springs. According to Liao & Guo (1986, fig. 6), most of the springs in this area are located on faults.

Methodology

  1. Top of page
  2. Abstract
  3. Introduction
  4. Geological setting
  5. Methodology
  6. Terminology
  7. Results
  8. Interpretations
  9. Discussion
  10. Conclusions
  11. Acknowledgements
  12. References

At Gongxiaoshe, hand samples of the precipitates were collected from the lower and upper ledges. Samples from the pool floor and walls were not collected because of the high water temperature. At Zhuyuan, large samples of the precipitates that had developed on the channel wall were collected (Fig. 2E to G).

The water temperature, pH and Eh measured in the field on 16 June 2011 and again on 15 October 2011 were essentially the same. Water samples were passed through a syringe filter with a 0·22 μm filtration membrane before being stored in polypropylene bottles. Unfortunately, problems with analyses of the water samples collected in June meant that the springs had to be sampled again on 15 October 2011. Those water samples were analysed for major cations and anions at the Saskatchewan Research Council (Saskatoon, Canada) about three weeks after collection. The elements Ca, Mg, Na, K, Si and S were determined by Inductively Coupled Plasma Atomic Emission Spectroscopy (ICP-AES); alkalinity (including p alkalinity) was determined by titration with sulphuric acid on an auto-titration system. The bicarbonate, carbonate and hydroxides were calculated from the pH and alkalinity results. The chloride was measured colorimetrically and fluoride was determined by ion selective electrode. Carbonate samples from Gongxiaoshe and Zhuyuan were analysed for δ18O(calcite) and δ13C(calcite), whereas δ18O(water) and δ2H(water) were obtained from water samples collected on 15 October 2011. The isotope analyses, performed by Isotope Tracer Technologies Inc. (Ontario, Canada), are reported in the Vienna-Pee Dee Belemnite (VDPB) notation. Carbonate samples were flushed with helium for 4 min, then dried in an oven for 15 min at a temperature of 50°C before being acidified with 4 drops of 100% phosphoric acid. After allowing 2 h reaction time, aliquots were injected into a HP6890 GC-C + Delta Plus XL IRMA system for analysis (Hewlett Packard Company, Palo Alto, California, USA). The standards, NIST-19, NIST-18 and IT2-21, were run every five samples. The reproducibility is ±0·15‰ for δ18O(calcite) VPDB and ±0·1‰ δ13C(calcite) VPDB. For δ18O(water) and δ2H(water), the water samples were vapourized by a flash process at 140°C and then analysed using a Picarro Cavity Ring Down Spectrometer (CRDS; Picarro Inc., Santa Clara, California, USA). The reproducibility for these analyses is ±0·1‰ for δ18O(water) and ±0·6‰ for δ2H(water).

A total of five large thin sections (3 × 2 cm) were made from samples collected from Gongxiaoshe and Zhuyuan springs. Each sample was first impregnated with blue epoxy to highlight porosity. Fluorescence micrographs of the microbial mats were examined and imaged on a LEICA DM4500P fluorescence microscope (Leica Microsystems Inc., Concord, Ontario, Canada). Small fracture samples were mounted on SEM stubs using conductive glue and then sputter coated with a thin layer of gold before being examined on a JOEL 6400FE field-emission scanning electron microscope (SEM; JOEL USA Inc., Peabody, Maine, USA). Energy-dispersive X-ray (EDX) analyses were performed with an accelerating voltage of 20 kV, whereas most imaging was performed with an accelerating voltage of 5 kV. The location and orientation of all samples were recorded so that the different fabrics could be related to each other. A total of 1027 SEM photomicrographs were used in the assessment of the composition and fabrics in the samples collected from the two springs.

Terminology

  1. Top of page
  2. Abstract
  3. Introduction
  4. Geological setting
  5. Methodology
  6. Terminology
  7. Results
  8. Interpretations
  9. Discussion
  10. Conclusions
  11. Acknowledgements
  12. References

Dendrite crystals are architecturally complex crystals characterized by many levels of branching (Keith & Padden, 1964; Lofgren, 1974; Casanova, 1986; Jones & Kahle, 1986, 1993; Jones & Renaut, 1995; Jones et al., 2000, 2005). Branches in crystallographic calcite dendrites and non-crystallographic dendrites (Keith & Padden, 1964; Jones & Kahle, 1986; Jones & Renaut, 1995; Jones et al., 2005) typically develop through crystal splitting or preferential growth on various nucleation sites on the parent branches (Jones & Renaut, 1995). The term ‘pseudodendrite’ has been applied if discrete calcite crystals are arranged in a branching pattern (Jones & Renaut, 1998, fig. 10) or if it cannot be demonstrated that the branches are genetically related to each other.

Crystals constructed of smaller crystals have been termed polycrystalline crystals (Towe, 1967), aggregate crystals (Binkley et al., 1980; Chafetz et al., 1985) and composite crystals (Given & Wilkinson, 1985; Sandberg, 1985; Jones & Renaut, 1996b). The constituent crystals have been referred to as sub-crystals (Sandberg, 1985; Jones & Renaut, 1996a; Jones et al., 2005) or crystallites. Herein, the terms ‘composite crystals’ and ‘sub-crystals’ are used.

Amorphous calcium carbonate (ACC), which is isotropic and does not diffract X-rays (Addadi et al., 2003; Obst et al., 2009), forms the skeletons of many different organisms (Addadi et al., 2003, table 1). Decho (2010) suggested that ACC was a combination of precipitate and organic matrix, whereas Addadi et al. (2003) and Meldrum & Cölfen (2008) suggested that some ACC is hydrated (CaCO3.H2O). Identifying the presence of ACC is difficult, especially if it is intimately associated with crystalline calcite and/or crystalline aragonite, because its presence may be disguised on the X-ray diffractograms and criteria for its recognition on the SEM have yet to be fully established. By analogy with opal-A spheres, calcareous particles are herein considered to be ACC if they are <0·5 μm long and with no evidence of crystal faces (cf. Jones & Peng, 2012).

A biofilm is herein regarded as a community of microbes that are embedded in EPS (Rosenberg, 1989; Neu, 1996; Decho, 2000, 2010). According to Neu & Lawrence (2009): “…EPS are organic polymers of microbiological origin which, in biofilm systems, are responsible for the interaction with interfaces, as well as with dissolved, colloidal and particulate compounds…”. Typically, the EPS is a hydrogel that allows microbes to attach themselves to substrates while buffering them from the immediate extracellular environment (Decho, 1990, 2010). Herein, thin films that coat substrates and/or span cavities are deemed to be EPS if they appear similar to EPS, as illustrated by Decho & Lopez (1993, figs 2 and 3), Allen et al. (2000, figs 3,6, 9 and 11), Westall et al. (2000, figs 6 and 8), McKenzie et al. (2001, fig. 3E), Dupraz et al. (2004, fig. 9) and Pedley & Rogerson (2010, figs 3a, 4a and 5c). The presence of microbes in such coatings also adds credence to their identification.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Geological setting
  5. Methodology
  6. Terminology
  7. Results
  8. Interpretations
  9. Discussion
  10. Conclusions
  11. Acknowledgements
  12. References

General features of Gongxiaoshe and Zhuyuan springs

Gongxiaoshe, located in the middle of Ruidianxiang (Fig. 1C), is confined by a concrete wall that is up to 2 m high with an iron railing on top that is 0·25 m high (Fig. 2A to C). The date of construction of the wall and the railing is unknown. The octagonal-shaped pool, ca 7 m at its widest part and up to 1·5 m deep, with constantly circulating water, has its main vent beneath a wooden platform (Fig. 2A). Maximum water ebullience is above the vent where the water has a temperature of 73°C. In contrast, the calmer water around the edge of the pool has a temperature of 50 to 55°C.

Distinctive upper and lower ledges characterize the margins of the pool (Fig. 2B and C). The origin of the ledges is unknown because they are now completely covered with carbonate precipitates and microbial mats. The architectural consistency of the two ledges around the pool, however, indicates that they are probably an inherent part of the concrete wall that was constructed around the pool. The upper ledge, 16 to 33 cm wide, abuts the outer concrete wall, whereas the inner edge has lobate protrusions extending into the pool. Although the surface of this ledge is generally cream to light orange, there are local patches, especially on the inner margin, which have a greenish hue (Fig. 2B). The surface of this ledge is always damp because spring waters gently wash over its surface.

The lower ledge, 16 cm wide and submerged in water 5 to 10 cm deep, is divided into the inner (closest to wall) and outer (closest to pool) parts (Fig. 2B). The inner part is green, whereas the inner part is cream to light orange. The green colour reflects the presence of actively growing microbial mats. Water in this part of the pool has temperatures of 51 to 61°C and a pH of 7·5. The pool floor, which is difficult to see because of the steam and circulating water, appears to be lighter coloured than the surfaces of the ledges. A sample dredged from the floor of the pool was soft and lacked structures.

Zhuyuan Spring has been modified substantially because it provides hot water and steam for a bathhouse. The hot spring water (73°C) is, however, evident in an uncovered part of a concrete lined channel (Fig. 2D). The walls of the channel are lined with calcite, with the best-developed precipitates forming a vertical sheet that is 36 cm high (Fig. 2E). Precipitates below the water level have a rusty-red surface, whereas those above are reddish-cream (Fig. 2F and G). The sheet, <1 cm thick at its base, widens upwards before passing into a ledge at the air/water interface that is up to 10 cm wide. The uppermost surface, above the water level, is covered with round to ovate, laminated, stromatolitic columns that are 0·5 to 1 cm wide and up to 1 cm high (Fig. 2F and G).

Well-defined laminations, branching calcite crystals and aragonite characterize the deposits at Gongxiaoshe Spring (Fig. 3A and B). Although the deposits at Zhuyuan Spring are similar (Fig. 3C to H), they lack aragonite, but include laminated columns (Fig. 3C and D) and zones characterized by red, non-crystalline precipitates (Fig. 3H).

image

Figure 3. Thin section photomicrographs of precipitates from Gongxiaoshe (A) and (B) and Zhuyuan (C) to (H) hot springs. Blue = porosity. (A) Scan of thin section through precipitates from the upper ledge at Gongxiaoshe showing variations in laminae thicknesses. (B) Enlarged view from the central part of panel (A) showing alternating aragonite ‘A’ and calcite ‘C’ layers. Note the poorly defined branching in calcite crystals. (C) Scan of thin section through the upper part of precipitates from Zhuyuan (approximately in the position of letter ‘F’ on Fig. 2E) showing growth banding, outward splaying columns in the lower part and bases of vertical laminated columns on top. (D) Scan of thin section through the lower part of precipitates from Zhuyuan (lower part of Fig. 2E) showing growth banding and outward splaying columns. White letters ‘E’, ‘F’ and ‘H’ indicate the positions of panels (E), (F) and (G), and (H), respectively. (E) Outward radiating calcite crystals and well-defined growth banding. (F) and (G) Examples of irregularly branching calcite crystals, evident in more porous parts of precipitates. (H) Outermost laminae showing calcite (white) and red, Fe-rich, reticulate coatings that give the precipitates their distinctive red surface colour (Fig. 2E to G).

Download figure to PowerPoint

Climate

Climate records for Ruidianxiang are lacking because the village is located in a remote part of Yunnan Province. Tengchong County, which includes this area, is characterized by a subtropical highland climate with an annual mean temperature of 14·9°C and an annual average rainfall of 1480 mm (Jones & Peng, 2012, fig. 3). In December and January, the average temperature is between 1°C and 17°C, whereas temperatures between 14°C and 17°C characterize July, August and September. The hours of sunshine are at a minimum during June, July and August when rainfall reaches maximum levels.

Spring waters

Analyses of the water samples collected on 15 October 2011 show that the waters from Gongxiaoshe and Zhuyuan are similar (Table 1) in composition with high concentrations of Ca (21 to 25 mg l−1), Na (390 to 421 mg l−1), Cl (154 to 164 mg l−1) and HCO3 (938 to 983 mg l−1). Comparisons of the analyses for Gongxiaoshe obtained during this study with those reported by Zhang et al. (2008), however, reveal some significant differences, particularly with respect to the Ca, Mg and Na concentrations (Table 2). Zhang et al. (2008), for example, show that Mg > Ca, whereas the analyses based on the 2011 samples showed that Ca was significantly higher than Mg (Table 1). Although the reason(s) for these contrasts is (are) not clear, it may reflect: (i) the names of the springs with the Gongxiaoshe as identified by Zhang et al. (2008) being different from the Gongxiaoshe identified here; (ii) temporal changes in spring water chemistry; and/or (iii) analytical problems with the analyses of the water.

Table 1. Chemical analyses of waters from hot springs located in Ruidian geothermal area
SpringT (°C)pH (field)pH (lab)eH (mV)NaKCaMgCO3HCO3ClSO4FSiO2
  1. a

    Zhang et al. (1987) – data from their table 1 with springs numbered 23-1 to 23-6; spring names from Tong et al. (1986, their table 1).

  2. b

    Zhang et al. (2008) – data from their table 1, in mg/L.

  3. c

    This study – in mg/L.

23-1: (Gentlemen's Bath)a77·07·57·2 35034·022·23·30·077513932·07·0123
23-2: (Ladies Bath)a87·07·58·1 41045·04·54·236·389514031·27·0129
23-3: (Bath #105)a86·57·57·7 40042·023·63·40·091215716·08·5148
23-4: (Highway Maintenance Squad #23)a56·06·07·9 24628·57·53·224·65208014·64·0104
23-5: Dagoubiana64·0 8·1 37836·67·52·524·676813325·06·9126
23-6: Shapojiaoa55·07·57·5 40036·023·63·70·083716424·08·5140
Jiemingb79·38·2 −9615439·24·96·5 92718134·515·6160
Gongxiaosheb74·98·5 −2316542·94·96·318·093920034·517·3180
Gongxiaoshec73·57·58·2−2342144·021·03·8 98316428·05·0139
Zhuyuanc82·16·87·8+2039043·025·04·0 93815427·05·2128
Table 2. Stable isotopes values for water and calcite samples from Gongxiaoshe and Zhuyuan, Ruidian geothermal area
δ2D(water)δ18O(water) (VSMOW)SpringSampleδ18O(calcite) (VPDB)δ13C(calcite) (VPDB)
−83·1−10·75GongxiaosheA−20·52−1·74
−83·4−10·80B−21·04−1·90
C−21·19−1·92
−85·0−11·39ZhuyuanA−20·98−0·62
−85·0−11·35

Modelling of the waters by PHREEQC (Parkhurst & Appelo, 1999), using the water analyses for Gongxiaoshe (G) and Zhuyuan (Z), obtained in this study (Table 1), yielded Saturation Indices (SI) greater than 0 for aragonite (G = 0·7, Z = 0·2), calcite (G = 0·8, Z = 0·3), dolomite (G = 1·4, Z = 0·2), chalcedony (G = 0·4, Z = 0·3), quartz (G = 0·7, Z = 0·6) and talc (G = 5·2, Z = 1·7). Saturation indices less than 1 were obtained for anhydrite (G = −2·7, Z = −2·5), fluorite (G = −0·8, Z = −0·7), gypsum (G = −2·8, Z = 2·7), halite (G = −5·9, Z = −6·0) and sepiolite (G = −1·0, Z = −3·6). Such calculations suggest that there is the potential for the precipitation of aragonite, calcite, dolomite, chalcedony, quartz and talc from the waters in Gongxiaoshe and Zhuyuan springs. These two springs seem to differ only with respect to chrysotile with a SI of 1·3 for Gongxiaoshe but −3·2 for Zhuyuan.

Deposits at Gongxiaoshe Spring

Samples collected from the upper ledge at Gongxiaoshe are formed of thin, light cream (with a slight orange hue) layers (up to 3 mm thick) of calcite that alternate with thin (<1 mm) white layers of aragonite (Fig. 3A and B). In some parts of the succession, the calcite and aragonite layers are of sub-equal thickness whereas, in other parts, the calcite layers are thicker than the aragonite layers (Fig. 3A). Although samples from the lower ledge display the same basic architecture, they are generally softer (can be crushed between fingers) than samples from the upper ledge. The surface of the inner part of the lower ledge is green (Fig. 2B) because of the microbial mat that covers its surface. In contrast, buried surfaces are commonly light to dark brown.

Calcite

At Gongxiaoshe, calcite crystals that form the spring deposits include: (i) pseudodendrites (Fig. 4); (ii) dodecahedrons (Fig. 5A to C); and (iii) rhombohedrons (Fig. 5D to F). The pseudodendrites are rooted on aragonite substrates and grew upwards (Figs 4B, 5A and 5B). In contrast, the euhedral dodecahedrons and rhombohedra are randomly distributed around the pseudodendrites with no evidence of them being rooted to a substrate.

image

Figure 4. SEM photomicrographs of calcite pseudodendrites from Gongxiaoshe. (A) Sample from the upper ledge formed of alternating calcite ‘C’ and aragonite ‘A’ layers. (B) Calcite layer formed of upward expanding, V-shaped calcite crystals that show minimal evidence of branching. (C) Sample from another part of the upper ledge showing microscale variations in size and morphology of calcite crystals. Note the poorly developed branching in some of the larger crystals. (D) View of upper surface of calcite layer, parallel to bedding, showing the tops of individual calcite columns. Box labelled ‘E’ indicates the position of panel (E). (E) Top of calcite column showing a complex array of calcite crystals. Box labelled ‘H’ indicates the position of panel (H). (F) General view of a calcite column formed of numerous smaller crystals. (G) Enlarged view of the upper part of the column shown in panel (D), showing stacked crystals and poorly defined branching. (H) Edge of calcite column showing the inner calcite crystal coated by successive layers of calcite cement (‘C1’ to ‘C3’). (I) Top of the calcite column showing the central crystal coated by five layers of calcite cement (‘C1’ to ‘C5’). (J) Top of the calcite column showing a vase-like structure with an elevated outer rim and central part partly filled with loose calcite crystals. Box labelled ‘K’ indicates the position of panel (K). (K) and (L) Enlarged views of the calcite cement layer showing a hexagonal motif created by precipitation of calcite sub-crystals. (M) Outer part of top of the calcite column showing calcite dodecahedra and rhombohedra (arrows) sitting on top of different calcite cement layers (‘C1’ to ‘C3’). Box labelled ‘5A’ indicates the position of Fig. 5A.

Download figure to PowerPoint

image

Figure 5. SEM photomicrographs of euhedral calcite crystals associated with calcite columns, Gongxiaoshe. (A) Incomplete dodecahedron. (B) Dodecahedron with partly developed crystal faces capping interior that appears to be formed of numerous sub-crystals. White letter ‘C’ indicates the crystal face shown in panel (C). (C) Enlarged view of partly formed crystal faces shown in panel (B). (D) Rhombohedral calcite crystal surrounded by large coccoid microbes. (E) Cluster of intergrown calcite rhombohedra. (F) Incompletely formed rhombohedral calcite.

Download figure to PowerPoint

In thin section (Fig. 3B) and at low magnifications on the SEM (Fig. 4A and B), the calcite pseudodendrites are of variable crystal size and generally lack obvious branching. At low magnifications, some appear as elongate, (sub)erect crystals with rounded edges (Fig. 4A), others have an upward expanding V-shaped morphology (Fig. 4B), whereas others appear to irregularly branch with smaller crystals arising from larger crystals (Fig. 4C). Vertical cross-sections through some crystals reveal nested V-shaped crystals that have a solid basal core from which two peripheral arms develop (Fig. 3C). These vase-shaped columns have circular to ovate transverse cross-sectional shapes, which are clearly evident on the bedding surfaces (Fig. 4D and E). These cross-sections also show that the size and shape of the columns were strongly influenced by competition with their neighbours for growth space (Fig. 4D and E).

High magnification SEM imagery shows that the complex calcite columns are formed of: (i) numerous sub-crystals (Fig. 4F and G); (ii) layers of calcite that coat and progressively encase the formative sub-crystals (Figs. 4H to L); and (iii) euhedral dodecahedra and rhombohedra that are randomly distributed on the surfaces of the pseudodendrites (Figs 4M and 5A to F). It is difficult to determine the exact morphology of the sub-crystals in the columns because most surfaces are covered with thin veneers of calcite cement (Fig. 4H to K). These veneers are formed of up to five layers, with each one covering and, at the top, overlapping the previously formed layer (Fig. 4H and I). Each layer is constructed of very small, flat rhombic crystals (up to 0·5 μm long, 0·4 μm wide and <0·1 μm thick) that are systematically arranged in a hexagonal motif (Fig. 4K and L). These calcite veneers coat the calcite columns and hide all evidence of the branching pattern. Although branching is irregular and seemingly without pattern (Fig. 4), the manner in which new branches develop cannot be ascertained; accordingly, they are referred to as ‘pseudodendrites’.

The euhedral dodecahedra, up to 50 μm long, have 12 pentagonal-shaped faces (Figs 4M and 5A). Although some crystals are complete with well-defined crystal edges (Fig. 5A), others have no inter-facial edges and have only the central parts of their faces developed (Fig. 5B and C). The interiors of these incompletely formed crystals are constructed of sub-crystals that are difficult to delineate accurately because of their small size (<1 μm). The smooth, partly developed crystal faces, however, do not appear conformable with the interior structure of the crystal (Fig. 5B). Nevertheless, their morphological attributes are consistent on different faces of individual crystals (Fig. 5A) and between different crystals.

The euhedral rhombohedra, up to 10 μm long, occur as isolated crystals (Fig. 5D) or in clusters of intergrown crystals (Fig. 5E). Although most crystals have smooth crystal faces with sharply defined interfacial edges (Fig. 5D and E), others have incompletely formed edges and corners (Fig. 5F). In this sense, they are similar to the incompletely formed dodecahedra.

Aragonite

Most aragonite is concentrated into discrete layers that are formed of randomly orientated crystals that generally have their long axes lying parallel to bedding. Morphologically, the aragonite is divided into: (i) small crystals; (ii) large crystals; and (iii) crystal bundles (Fig. 6). These three types commonly occur in close association with one another (Fig. 6A).

image

Figure 6. SEM photomicrographs of aragonite crystals from Gongxiaoshe. (A) Mixture of small aragonite ‘SA’, large aragonite ‘LA’ and aragonite bundles ‘AB’ on the bedding surface, upper ledge. (B) Large aragonite crystal showing zigzag boundaries between neighbouring sub-crystals (arrows). Crystals lie parallel to the bedding surface. (C) and (D) Transverse cross-sections through large aragonite crystals showing hexagonal shape and constituent sub-crystals. (E) and (F) Small aragonite crystals encased by large aragonite crystals. (G) Bundle of aragonite crystals formed of nested cone-shaped arrays of divergent crystals. Box labelled ‘H’ indicates the position of panel (H). (H) Small, composite aragonite crystals in lower part of aragonite bundle with EPS between neighbouring crystals (arrows). (I) Bundles of aragonite surrounded by numerous calcite crystals. (J) Aragonite ‘A’ wheat sheaf encased by calcite crystal. (K) Fan-shaped array of aragonite ‘A’ crystals rooted on the surface of calcite crystal.

Download figure to PowerPoint

The small (up to 10 μm long and 1 μm wide) and large (50 μm long and 10 μm wide) aragonite crystals are morphology alike but differ in size (Fig. 6A to E). The small crystals occur individually or in small clusters, whereas the large crystals tend to be discrete. Each type of crystal has a hexagonal cross-section (Fig. 6C and D) and is formed of elongate sub-crystals that interlock with one another along serrated sutures that are clearly evident on the crystal faces (Fig. 6B). Many of the large aragonite crystals grew around and encased the smaller ones and now have ‘spikes’ protruding from their sides (Fig. 6E and F). The small aragonite crystals are not, however, being absorbed into the large crystals through recrystallization. The small and large aragonite crystals are morphologically akin to the aragonite crystals that Jones & Renaut (1996b, fig. 7) described from hot spring deposits at Chemurkeu in the Kenyan Rift Valley.

The bundles of aragonite crystals include fountain-like arrays (Fig. 6G and H), wheat sheaves (Fig. 6I) and isolated fan-shaped clusters (Fig. 6J) that are formed of small aragonite crystals. The fountain-like arrays, up to 150 μm long with a maximum diameter of 50 μm, are formed of a series of nested cones, with each cone being formed of aragonite needles that splay outwards (Fig. 6G and H). The wheat sheaf arrays, up to 50 μm long and 20 μm wide (Fig. 6I and J), named for their morphological similarity to wheat sheaves, are similar to those illustrated by Tracey et al. (1998, fig. 5) and Beck & Andreassen (2010, fig. 9), and dumbbells (González-Muñoz et al., 2010, fig. 6a). The fan-shaped clusters, typically rooted on the face of a calcite crystal (Fig. 6K), generally have a radius of <25 μm.

Collectively, the aragonite crystals are characterized by: (i) random intermixing of crystals of different sizes and morphologies (Fig. 6A); (ii) their biterminal morphology (Fig. 6); and (iii) only rare examples of crystals that nucleated on a pre-existing substrate (Fig. 6J). The aragonite crystals are surrounded by calcite (Fig. 6I), grew from the surfaces of calcite crystals (Fig. 6J) or are encased by calcite (Fig. 6K). Some of the aragonite bundles are partly coated with EPS (Fig. 6H) and some crystal surfaces are characterized by small dissolution pits (Fig. 6H).

Silicon–magnesium–iron precipitates

Many calcite crystals, aragonite crystals and microbes in the Gongxiaoshe deposits are covered with reticulate coatings, 0·1 to 5 μm thick, that have a distinctive net-like surface morphology with ridges outlining irregularly shaped openings (Fig. 7A to F). Many of these coatings are so thin that the EDX beam penetrates through them and provides analyses that largely reflect the composition of the underlying substrate. Repeated EDX analyses of the thickest coatings, however, show that they contain various combinations of Si, Mg and Fe (Fig. 8), elements that were not detected in the underlying calcite/aragonite substrates. If present, Fe is a minor component and far less common than Si and Mg. The Ca evident on most of the EDX analyses probably comes from the underlying substrate. Although the ratio of Si to Mg is variable (based on peak heights on EDX analyses), there is always more Si than Mg. Coatings with high Si seem to have sharper, better defined ridges than those coatings with lower Si and more Mg. Such analyses also indicate that the amount of Fe, which is highly variable over small areas, appears to have little effect on the morphology of the reticulate coating.

image

Figure 7. SEM photomicrographs of Si–Mg reticulate (SMR) and Si–Fe reticulate (SFR) coatings from Gongxiaoshe. (A) Calcite crystal ‘C’ covered with SMR with small open-mesh work and well-defined ridges. (B) Calcite crystal covered by SMR with open cells and low ridges. (C) Calcite crystal covered by thin SMR coating. ‘E1’ indicates the position of EDX analysis shown in Fig. 8A. (D) Filamentous microbe (note open lumen) encased by SMR coating. (E) Partly preserved filament covered with SMR resting on surface of calcite crystal coated with SMR. ‘E2’ indicates the position of EDX analysis shown in Fig. 8B. (F) Large mass formed of SFR. ‘E3’ indicates the position of EDX analysis shown in Fig. 8C. (G) Smooth EPS film overlying SMR coating. (H) Thin films of smooth EPS passing laterally into SMR coating. (I) Collapsed non-mineralized filament resting on top of SMR.

Download figure to PowerPoint

image

Figure 8. EDX analyses for various reticulate coatings from Gongxiaoshe. Positions of analyses shown in panels (A), (B) and (C) are shown on Fig. 7C, E and F, respectively.

Download figure to PowerPoint

Some substrates are covered with reticulate coatings and smooth EPS films that are typically ca 1 μm thick and locally characterized by small holes that probably developed through desiccation (Fig. 7H and I). Although EDX analyses of the smooth films are difficult because they are so thin, repeated analyses of many different films show that they commonly contain traces of Si and Mg, elements that are not detected in the underlying substrates. The relation between the reticulate coatings and the smooth EPS is complex and variable. In some areas, the reticulate coating grades laterally into smooth EPS with no apparent break (Fig. 7G and H). Elsewhere, smooth films and strands of EPS rest on top of the reticulate coating (Fig. 7I). Locally, the reticulate coatings are cros-sed by smooth, partly collapsed, non-mineralized filamentous microbes that are consistently 0·5 to 0·75 μm wide, have open lumens (where visible), and slight constrictions that may indicate the presence of septa. Throughout the samples, there are rare examples of filamentous microbes that have been covered with Si–Mg reticulate coatings (Fig. 7E). Such coatings are morphologically and compositionally the same as the coatings that cover the calcite and aragonite crystals.

Extracellular exopolymeric substances and microbes

At Gongxiaoshe, surfaces on the lower and upper ledges are locally covered with green microbial mats (Fig. 2B and C), up to 5 mm thick, which are formed mainly of large coccoid microbes, 6 × 4 μm in size (Fig. 9A), that: (i) are covered with a very thin layer of EPS (Fig. 9B); (ii) commonly have a granulose exterior ornamentation (Fig. 9C); and (iii) reproduced by binary fission (Fig. 9C). The surfaces of buried laminae, which are commonly medium to dark brown, are covered with the same large microbe. These latter microbes, however, are typically collapsed, flattened and locally desiccated (Fig. 9D). Although most are not mineralized, cross-sections through some specimens revealed partly calcified interiors (Fig. 9E). A variety of non-calcified filamentous and coccoid microbes are commonly embedded in the EPS that cover the large coccoid microbes (Fig. 9F to H). The large coccoid microbes, the small-diameter filamentous and coccoid microbes, and EPS are probably part of a complex biofilm assemblage.

image

Figure 9. SEM photomicrographs of microbes and EPS associated with precipitates from Gongxiaoshe. (A) Green surface covered with numerous large, ovate coccoid microbes covered with a thin, smooth EPS layer. (B) Coccoid microbe covered with a thin EPS layer. (C) Coccoid microbe with granular surface ornamentation showing binary fission. (D) Large coccoid microbes on a brown surface flattened due to partial desiccation and covered with a thin mucus layer. (E) Cross-section through a large coccoid microbe showing the interior formed of organic material and calcite (based on EDX). (F) Collapsed and partly desiccated small, round coccoid microbes. (G) to (I) Non-calcified microbes on surfaces of calcite crystals with open lumen evident in panel (G) and possible septa (arrows) in panel (I). (J) and (K) Smooth, non-calcified EPS covering substrate and spanning gaps between grains. (L) Oblique view of smooth non-calcified EPS coating – note the non-calcified filaments on the surface. (M) Surface of non-calcified EPS with numerous, non-calcified filamentous microbes on surface (arrows). (N) Cross-section [from edge shown in panel (L)] showing a layered structure of non-calcified EPS and non-calcified filamentous microbes on the surface.

Download figure to PowerPoint

When viewed under a fluorescent microscope, many of the microbes collected from the microbial mats on the upper ledge are highlighted by red autofluorescence with the paired cells being readily apparent (Fig. 10A). The red fluorescence of the large coccoid microbes indicates that they are probably cyanobacteria, which are tentatively identified as Synechococcus. Staining with Acridine Orange shows that the mats also contain small coccoid and filamentous microbes that occur among the clusters of large coccoid microbes (Fig. 10B).

image

Figure 10. Fluorescence photomicrographs of green microbial mat from Gongxiaoshe. (A) Red autofluorescence of microbes in untreated sample. (B) Fluorescence of microbial mat stained with Acridine Orange. Green highlights viable cells, whereas red shows dead cells.

Download figure to PowerPoint

Throughout the samples from the lower and upper ledges at Gongxiaoshe, the constituent calcite and aragonite crystals are commonly covered and almost completely wrapped in smooth EPS films that contain scant evidence of their formative microbes. These thin (<1 μm thick) films are commonly desiccated, broken and locally modified into thin filament-like strands (Fig. 9J and K). In some areas, there are thicker EPS layers, 1·5 to 2·0 μm thick, which are internally laminated with laminae ca 0·1 μm thick (Fig. 9L to N). Although many of these EPS films contain numerous collapsed, septate filamentous microbes (ca 1 μm wide), there is no evidence of mineralization and repeated EDX analyses failed to detect any Si, Mg, Fe or Al.

Isotopes

Three samples of calcite from Gongxiaoshe Spring yielded δ18O(calcite) values of −20·5‰, −1·8‰ and −21·0‰ VPDB and δ13C(calcite) values of −1·90‰, −21·2‰ and −1·92‰ VPDB, respectively. Two water samples, collected on 15 October 2011, yielded δ18O(water) values of −10·75‰ and −10·80‰ VSMOW and δ2H(water) values of −83·1‰ and −83·4‰ (Table 2).

Deposits from Zhuyuan Spring

The most obvious feature of the deposits from Zhuyuan Spring is their distinct red surface colouration (Fig. 2E and F) that contrasts sharply with their beige to light grey interior laminae. The vertical sheets that coat the channel wall (Fig. 2D and E) are formed of: (i) an inner zone, below water level, that is formed of laminated columns (up to 2 cm long and 0·5 cm in diameter) that have their long axes at 75 to 90° to the channel wall (Fig. 3C and D); (ii) an upper zone, above water level, that is formed of small, laminated columns (up to 1 cm high and 0·5 cm in diameter) that have a tan to light orange surface colour that contrasts sharply with the red colour of the surface that is below water (Fig. 2E and F); and (iii) an outer, Fe-rich laminated band that extends from the uppermost layer down the underwater face of the precipitates and unconformably masks the laminated interior (Fig. 3D and H).

All of the deposits at Zhuyuan are formed of various combinations of calcite pseudodendrites (Figs 3E to G, 11 and 12), ACC (Fig. 13), accessory minerals that include barite and gypsum (Fig. 14) and Si–Mg–Fe precipitates (Fig. 15) that are intimately associated with microbes and EPS (Figs 10, 12 and 13).

image

Figure 11. SEM photomicrographs of calcite crystals from Zhuyuan. (A) General view of the sample showing a vertical cross-section (left side) and a surface that was in contact with water (right side). In cross-section, note the horizontal laminae encased by laminae formed of outward radiating calcite pseudodendrites. Boxes labelled ‘B’ and ‘C’ indicate the positions of panels (B) and (C), respectively. (B) Radiating pseudodendrites from the outer band shown in panel (A). (C) Surface of precipitates showing spatial variations in packing of calcite crystals. Box labelled ‘D’ indicates the position of panel (D). (D) Surface of precipitates with group of euhedral calcite crystals with EPS between crystals. Box labelled ‘E’ indicates the position of panel (E). (E) Euhedral dodecahedron with collapsed, non-mineralized microbes on crystal faces. White letter ‘F’ indicates the position of panel (F). (F) Collapsed, non-mineralized microbes on crystal surface. (G) Group of small calcite crystals arranged in a trigonal pattern and partly covered by biofilm. (H) Calcite crystal with smooth interfacial edges but incomplete crystal faces. Box labelled ‘I’ indicates the location of panel (I). (I) Calcite crystal face formed of numerous sub-crystals with a distinct cross-hatched pattern. (J) Calcite crystal with smooth interfacial edges, but incomplete crystal faces. Box labelled ‘K’ indicates the location of panel (K). (K) Smooth interfacial edges with successive growth layers.

Download figure to PowerPoint

image

Figure 12. SEM photomicrographs of calcite crystals from Zhuyuan. (A) Surface covered with aligned calcite crystals. Boxes labelled ‘B’, ‘F’ and ‘G’ indicate the locations of panels (B), (F) and (G), respectively. (B) Composite calcite crystal with a collapsed filamentous microbe on the left side. Boxes labelled ‘C’, ‘D’ and ‘E’ indicate the locations of panels (C), (D) and (E), respectively. (C) and (D) Surface of a calcite crystal formed of numerous sub-crystals arranged in a cross-hatched pattern. (E) Surface of interfacial edge showing successive growth layers. (F) Trilete calcite crystal buried beneath EPS.

Download figure to PowerPoint

image

Figure 13. SEM photomicrographs of ACC found in deposits from Zhuyuan. (A) Surface of calcite crystal covered with thin EPS film, collapsed non-mineralized filaments and scattered ACC. (B) Surface of calcite crystal completely covered with ACC and partly overlain by collapsed non-mineralized filamentous microbes. (C) Complex network of EPS and filamentous microbes partly coated with clusters of ACC nanoparticles. Box labelled ‘D’ indicates the position of panel (D). (D) Clusters of ACC nanoparticles formed in and around EPS strands. (E) ACC nanoparticles beneath and on top of EPS. (F) Surface of calcite crystal covered with ACC nanoparticles, collapsed, non-mineralized filamentous microbes and filamentous microbe coated with beaded calcite. (G) Rhombic calcite crystal covered with ACC and non-mineralized filamentous microbes. The black dot is in the same location as the black dots shown in panels (H) and (I). (H) and (I) Corner of rhomb showing faces formed of numerous ACC nanoparticles.

Download figure to PowerPoint

image

Figure 14. SEM photomicrographs of accessory minerals associated with EPS and calcite deposits from Zhuyuan. (A) and (B) Plate-shaped barite ‘B’ associated with Si–Mg reticulate coating ‘RC’ partly filling a cavity in groundmass formed of calcite ‘Ca’ crystals; RC is under and over the barite crystals. (C) Cluster of gypsum crystals in cavity.

Download figure to PowerPoint

image

Figure 15. SEM photomicrographs of reticulate coatings (A) to (H) and Fe-rich precipitates (I) to (L) in deposits from Zhuyuan. White numbers ‘E4’ to ‘E8’ (on black circles) indicate positions of EDX analyses shown in Fig. 16. (A) Si–Mg reticulate coating on the surface of calcite crystal. Box labelled ‘B’ indicates the position of panel (B). (B) Si–Mg (minor Fe) reticulate coating with well-defined platy morphology. (C) and (D) Filament-like structures coated with Si–Mg (minor Fe) reticulate coating. (E) Small diameter filamentous microbes and substrate coated with Si–Mg reticulate coatings. Box labelled ‘F’ indicates the position of panel (F). (F) Transverse cross-section through small-diameter filament with walls formed of Si–Mg (minor Fe) reticulate coating. (G) Cross-section through deposit showing alternation between calcite ‘Ca’ layers and Si–Fe reticulate coatings. (H) Cavity lined with Si–Mg–Fe reticulate coating (‘E5’ – Fig. 16B) that is overlain by granular Fe–Si–Mg (‘E6’ – Fig. 16C) precipitates. (I) Calcite substrate (‘E7’ – Fig. 16D) covered with granular Fe–Si (‘E8’ – Fig. 16E) precipitates. (J), (K) and (L) Microbes and EPS partly covered with granular Fe–Si precipitates.

Download figure to PowerPoint

Calcite pseudodendrites

Most of the calcite laminae in the Zhuyuan deposits are formed of branching columns that are up to 1 mm long but typically <0·5 mm long (Figs 3E to G, 11A and B). Throughout the deposit, there are also thin laminae that are formed of solid, featureless calcite that contrasts sharply with the porous laminae that are formed of the branching calcite columns.

Each branching column typically has numerous branches that arise from the uppermost side of the main stem (Figs 3F, 3G and 11B). Although some of the main stems are straight (Fig. 3F), others have an irregular profile (Fig. 3G) that seems to vary from dendrite to dendrite. Numerous levels of branching that seem to follow no particular pattern characterize most of the pseudodendrites (Figs 3F, 3G and 11B). The frequency and scale of branching are highly variable, even between neighbouring columns (Fig. 11B). Determining the internal structure and branching style of these dendrites is difficult because, as with the dendrites in the Gongxiaoshe deposits, their exteriors are coated with calcite cement (Fig. 11B).

The outer growth surface, which is perpendicular to the long axes of the pseudodendrites, is characterized by small crystals that are isolated, in small clusters in vaguely defined rows (Fig. 11A, C and D) or, more rarely, arranged in a trigonal motif (Fig. 11G). The crystals, evident on these surfaces, are the accretionary tips of the dendrite branches. These dodecahedral (Fig. 11E and F) and rhombohedral (Figs 11H to J and 12A to F) crystals commonly display complex internal structures.

Many of the rhombohedra are characterized by incompletely formed crystal faces but smooth, partly developed bevelled interfacial edges (Figs 11H to K and 12A to E). The crystal faces are formed of numerous elongate sub-crystals, typically up to 200 nm long but less than 20 nm wide, which are arranged in a cross-hatch pattern (Figs 11I and 12B to D). In contrast, the smooth interfacial edges are formed of thin (<20 nm thick), incompletely formed layers (Figs 11J, 11K and 12E). There is commonly a sharply defined discontinuity between the cross-hatched structure of the crystal faces and the smooth interfacial surfaces. On some crystal faces, there appears to be an intermediate structural level that is highlighted by the better development of some sub-crystals (Fig. 12C and D). There are also rare examples of trilete crystals that have smooth-surfaced branches that appear to be the same as the smooth, interfacial surfaces that are evident on some of the more complete crystals (Fig. 12F).

Amorphous calcium carbonate

In some parts of the samples from Zhuyuan, there are nanoparticles, up to 200 nm in diameter, which are scattered across the surface of large calcite crystals (Fig. 13A), intimately associated with EPS (Fig. 13B to F), or completely covering the surfaces of calcite crystals (Fig. 13G to I). Some of the larger particles appear to be formed of numerous smaller nanoparticles that have fused together (Fig. 13B and D to F). These nanoparticles are considered to be ACC because they do not display any recognizable crystal form and there is no evidence of crystal edges (cf. Jones & Peng, 2012). The ACC is always associated with microbes and/or EPS, with the nanoparticles being found beneath and on top of the EPS (Fig. 13B and D to F). Nevertheless, there are only rare examples of filamentous microbes that are coated with ACC (Fig. 13F).

Some of the larger euhedral calcite crystals are coated with rounded to sub-rounded nanoparticles of ACC that are <200 nm long (Fig. 13G to I). It is easy to envisage a situation whereby these nanoparticles would merge together to form a continuous growth zone around the crystal.

Accessory minerals

In some of the small (<1 mm long) cavities that are intimately associated with the microbial mats, there are clusters of barite (Fig. 14A and B) or gypsum (Fig. 14C) crystals. The barite crystals, identified by their crystal shape and EDX signatures, are up to 5 μm long and typically have their flat faces perpendicular to the substrate (Fig. 14B). In most cavities, these crystals sit on top on reticulate Si–Mg coatings. Although the reticulate coatings are not found on the large, flat faces of the crystals, they are locally present on the edges (Fig. 14B).

The gypsum crystals, identified by their shape and EDX signatures, are up to 1 μm long and typically form small clusters that hang from cavity roofs (Fig. 14C). Gypsum is less common than barite.

Silicon–magnesium–iron precipitates

The calcite crystals and some of the microbes are coated with precipitates that contain Si, Mg and/or Fe (Fig. 15). Many layers within the deposits are covered with reticulate Si–Mg coatings (Fig. 15A to F) that are the same as those found in the Gongxiaoshe deposits. Structures akin to filamentous microbes, evident in some parts of these precipitates (Fig. 15C to F), are characterized by an open lumen, generally <0·5 μm in diameter, which is surrounded by the reticulate coating that is up to 0·5 μm thick (Fig. 15E and F). The reticulate coating masks all other features of the microbes. Some of the reticulate coatings contain more Fe than those in the Gongxiaoshe deposits. In general, as the amount of Fe increases, so the amount of Mg decreases. In some areas, calcite laminae (up to 1 mm thick) are separated from each other by thin (up to 0·1 mm) laminae formed of the Si–Fe reticulate coatings (Fig. 15G).

In the outermost laminated red band that forms the surface of the precipitates below water level (Figs 2F and 3H), there are small (<1 cm), irregular-shaped cavities of unknown origin that are lined with Si–Fe reticulate coatings and partly filled with granular Si–Fe precipitates (Fig. 15H and I). In these situations, the Fe content of the granular Si–Fe precipitates is significantly higher than the Fe content of the reticulate coating (Figs 15H, 16B and 16C), whereas Fe was not detected in the underlying calcite substrate (Figs 15I, 16D and 16E). The granular Si–Fe precipitates, characterized by agglomerations of sub-rounded to rounded grains that are <200 nm in diameter (Fig. 15J to L), are typically associated with EPS and filamentous microbes (Fig. 15J and K). In some areas, the granular precipitates pass laterally into the reticulate coatings (Fig. 15L).

image

Figure 16. EDX analyses of calcite and associated precipitates from Zhuyuan. Precise locations of analyses E4 to E8 are shown on Fig. 15. Note the variance in the amounts of Si, Mg and Fe.

Download figure to PowerPoint

Extracellular exopolymeric substances and microbes

Copious amounts of EPS were evident throughout the samples examined on the SEM, even though the formative organisms were rarely apparent (Fig. 12B, E and F). The EPS are apparent as thin sheets, partly desiccated sheets and strands that probably formed as a result of desiccation. Although distinct but collapsed filamentous microbes are locally apparent (Figs 11F and 13), they are not always associated with the EPS. Small (<200 nm) calcite grains are locally associated with some of the EPS and there are rare examples of filaments that have been coated with such calcite (Fig. 3A to F).

Throughout all of the samples examined, only rare examples of mineralized microbes were found. Such microbes, identified on the basis of their long, tube-like structures with central openings (typically ca 0·5 μm in diameter), are largely mineralized with reticulate Mg–Si coatings (Fig. 15C to F). The identity of those microbes remains questionable because there are insufficient diagnostic physical features.

Isotopes

One sample of calcite yielded values of −21·0‰ for δ18O(calcite) and −0·6‰ for δ13C(calcite) VPDB. Two water samples yielded δ18O(water) values of −11·39‰ and −11·35‰ VSMOW, and δ2H(water) values of −85·0‰ (Table 2).

Interpretations

  1. Top of page
  2. Abstract
  3. Introduction
  4. Geological setting
  5. Methodology
  6. Terminology
  7. Results
  8. Interpretations
  9. Discussion
  10. Conclusions
  11. Acknowledgements
  12. References

Interpretation of the precipitates at Gongxiaoshe and Zhyuan must include assessment of: (i) the factors that controlled calcite as opposed to aragonite precipitation; (ii) the growth of the pseudodendritic calcite columns; (iii) the origin of the unattached aragonite crystals and dodecahedral and rhombohedral calcite crystals; (iv) the ACC that is commonly associated with the EPS and some of the filamentous microbes; (v) the development of the non-crystalline Si–Mg–Fe precipitates; and (vi) isotopic signatures of the spring waters and calcite.

Calcite versus aragonite precipitation

The association of calcite and aragonite in the Gongxiaoshe deposits immediately raises the possibility that the calcite formed through aragonite inversion. For the precipitates from Gongxiaoshe, this possibility is considered unlikely because: (i) there is little evidence of chemical corrosion of the aragonite (Fig. 6A to F) apart from those crystals covered by EPS (Fig. 6H); (ii) the aragonite and calcite crystals that occur side by side are pristine with no evidence of inversion (Fig. 6I); (iii) the aragonite crystals encased by calcite crystals show no evidence of alteration (Fig. 6J); (vi) there are rare examples of aragonite clusters that nucleated on the surfaces of calcite crystals (Fig. 6K); and (v) there is a systematic alternation between calcite and aragonite laminae (Fig. 4A and B). Thus, the calcite and aragonite are both considered primary precipitates.

The alternation between primary calcite and aragonite precipitation, known from many different environmental settings, has been variously attributed to temperature (Moore, 1956), humidity levels (Pobéguin, 1955, 1957), sulphate poisoning (Siegel, 1965; Rowling, 2004), Mg poisoning (Fischbeck & Müller, 1971; Seemann, 1985; Hill & Forti, 1997; Rowling, 2004), high Mg, Fe and Mn concentrations (Bosák et al., 2002), high Sr concentrations (Urbani, 1997), and saturation levels and precipitation rates (Rowling, 2004). In caves, this has also been linked to the CO2 content of the cave air (Cabrol, 1978; Hill & Forti, 1997) and substrate type (Craig et al., 1984). Kitano (1962b, fig. 1) suggested experimentally that the precipitation of calcite and aragonite was controlled largely by water temperature with aragonite being precipitated from water >40°C and calcite when the water temperature was <40°C. On the basis of a survey of spring deposits from various parts of the world, Folk (1994) argued that:

  • aragonite will be precipitated if the water temperature is >40 to 45°C, regardless of water composition;
  • aragonite will be precipitated if the molar Mg:Ca ratio is >1:1, regardless of water temperature; and
  • calcite is precipitated from waters that are Ca-rich and cooler than 40°C.

Contrary to these maxims, Folk (1994) noted that: (i) aragonite forms at sites of extreme CO2 degassing even if the water temperature and molar Mg:Ca favour calcite precipitation; and (ii) calcite can form, even in hot waters, if ion delivery is retarded by high viscosity or microbial biofilms (cf. Buczynski & Chafetz, 1991). The relation proposed by Folk (1994) are not, however, universally applicable. Experimental work has shown, for example, that the temperature–calcite/aragonite polymorphic relation fails if CO2 is bubbled through the parent solution (Kitano, 1962b) or if various ions are added to the parent solution (Kitano, 1962a; Kitano et al., 1962). The calcite that formed from spring waters with a temperature >90°C in New Zealand (Jones et al., 1996), and intercalated calcite and aragonite precipitates formed from spring waters with a temperature of 85 to 100°C in the Chemurkeu hot springs in the Kenya Rift Valley (Renaut & Jones, 1997) clearly show that temperature cannot be the main control on the precipitation of calcite and aragonite in spring systems.

In the context of the spring deposits at Gongxiaoshe and Zhuyuan, the underlying controls over calcite and aragonite precipitation must allow: (i) precipitation of aragonite at Gongxiaoshe where the water temperature is 51 to 73°C (depending on location in the pool) but no precipitation of aragonite at Zhuyuan where the water temperature is 84°C; (ii) cyclic precipitation of calcite and aragonite at Gongxiaoshe; and (iii) the free growth of the aragonite crystals, given that very few crystals nucleated on pre-existing substrates. Unfortunately, there are no long- or short-term records of physical and chemical changes in these spring waters. Conversations with the local people, who rely on the hot water from these springs, suggest that there have not been any significant changes in water temperature for as long as they can remember. It is difficult to identify precisely the parameter that controlled the precipitation of aragonite as opposed to calcite because most of the precipitation took place in sub-micron domains in the biofilms that are impossible to characterize in terms of their physicochemical parameters. It seems possible, however, that saturation levels may be the ultimate control.

Growth of pseudodendrites

When viewed at low magnification, the calcite columns that dominate the deposits at Gongxiaoshe and Zhuyuan springs are deceptively simple in appearance, being formed of main stems with a limited number of branches (Figs 3B, E to G, 4A to C and 11A and B). This simple architectural appearance is due largely to the layers of calcite cement that coated the surfaces of the branches and hid the myriad of small crystals that form each of the branches (Fig. 4D to M). Fortuitous sections through some of the columns show that each branch is formed of small crystals that were progressively stacked on top of each other (Fig. 4F and G) and surfaces that were in contact with the water clearly show the individual crystals that are forming at the growth tips of the dendrite branches (Fig. 11C to E). Although some of the constituent crystals in individual branches have a consistent orientation, others appear to have been added in a more haphazard fashion.

Three-dimensional calcite dendrites are known from many different spring systems in the Kenya Rift Valley (Jones & Renaut, 1995), New Zealand (Jones et al., 2000), Iceland (Jones et al., 2005) and Canada (Turner & Jones, 2005; Jones & Renaut, 2008; Rainey & Jones, 2009). These calcite dendrites are collectively characterized by their architectural complexities with multiple levels of branching that develop in many different ways. The development of dendrites in the spring deposits found in Kenya (Jones & Renaut, 1995) and New Zealand (Jones et al., 2000) resulted from crystal splitting. The dendrites in the Gongxiaoshe and Zhuyuan deposits and the Clinton travertine (Jones & Renaut, 2008) followed a different style of development because they grew as multitudes of small calcite crystals were progressively added to the growth tips of the branches. Like the dendrites from Gongxiaoshe and Zhuyuan, the dendrites in the Clinton travertine are characterized by incompletely formed calcite crystals (Jones & Renaut, 2008, figs 5, 6, and 8), crystals with incomplete faces that are separated from the crystal interiors by a distinct discontinuity (Jones & Renaut, 2008, fig. 7B to F) and late-stage calcite cements that commonly mask the formative structures of the dendrites (Jones & Renaut, 2008).

Unattached euhedral aragonite and calcite crystals

Throughout the deposits at Gongxiaoshe, there are numerous unattached randomly distributed and orientated aragonite crystals and euhedral dodecahedral and rhombohedral calcite crystals (Figs 5 and 6). Unattached euhedral calcite dodecahedra and rhombohedra are also found in the Zhuyuan deposits. These crystals grew in a medium that allowed growth in all directions, which is feasible in EPS or in the water column. The lack of any aragonite or calcite crystals on the filter papers (examined on SEM) that had been used when collecting the water samples suggests that these crystals were not present in the water column when sampled. This, in turn, indicates that the crystals may have grown in the EPS. Weed (1889), Buczynski & Chafetz (1991), Folk (1994) and Bonny & Jones (2003), for example, all drew attention to biterminal and polyterminal composite crystals that grow in the soft, mucilaginous materials that are an integral part of all biofilms.

Amorphous calcium carbonate

All of the ACC detected in the samples from the Ruidian hot springs was intimately associated with EPS and/or filamentous microbes (Fig. 13). Notably, ACC was absent in areas that lacked EPS and/or filamentous microbes. Jones & Peng (2012) pointed out that the nanoparticles that form ACC are morphologically similar to the nannobacteria that have been widely reported from many different settings (Folk, 1993; Folk & Rasbury, 2007). Nevertheless, these authors argued that ACC can form abiotically by rapid precipitation from supersaturated fluids and that the constituent nanoparticles need not be indicative of calcified nannobacteria. Irrespective of the mechanism of formation, the question arises as to what happens to the ACC following precipitation in the EPS and around the filamentous microbes. Meldrum & Cölfen (2008) suggested that ACC will either: (i) act as a source of CaCO3 for subsequent dissolution and recrystallization; or (ii) fuse together to form a single crystal in what they termed a ‘nonclassical crystallization reaction’. Such reactions are thought to involve shrinkage because the crystalline phases generally have a higher density than the non-crystalline phase and also release the hydrated water that is locked in the ACC.

In the Zhuyuan deposits, there is petrographic evidence that supports the notion of fusion whereby the individual nanoparticles merge together and become part of a larger crystal (Figs 11K, 12E, and 13G to I). In addition, some of the textures evident in the interiors of rhombohedra (Figs 11I, 12C and D) could also be ascribed to partial fusion of ACC. Although not proven by the textures found in these deposits, the notion that some crystals may grow through the fusion of ACC in a non-classical crystallization scheme (cf. Meldrum & Cölfen, 2008) must be given serious consideration.

Non-crystalline silicon–magnesium–iron precipitates

The non-crystalline precipitates that contain varying quantities of Si, Mg and Fe form thin coatings that are characterized by their reticulate (Figs 7 and 15A to F) or granular (Fig. 15H to L) appearance. Reticulate coatings, characterized by their mesh-like structure and variable elemental contents (for example, Si, Mg, Al, Fe, Ca, Na, Cl and Mn), are known from hot spring deposits in the African Rift Valley (Casanova & Renaut, 1987; Casanova, 1994; Jones & Renaut, 1996a,c), hot spring deposits in New Zealand (McKenzie et al., 2001; Jones et al., 2003), in scale that lines pipes in the geothermal systems of Iceland (Gunnlaugsson & Einarsson, 1989; Kristmannsdottir et al., 1989; Sverrisdottir et al., 1992), submarine hot springs in Iceland (Geptner et al., 2002) and in speleothems found in wave-cut notches (Jones, 2010) and caves (Léveillé et al., 2000a,b; Polyak & Güven, 2000, 2004). Their precise mineralogy is difficult to establish because they form such thin coatings that it is generally impossible to extract them for XRD analysis and most appear non-crystalline with microscale variations in element concentrations. Léveillé et al. (2000a,b) suggested, for example, that Mg–Si-rich poorly crystalline precipitate found in microbialites on walls of Hawaiian caves might be kerolite (Mg3Si4O10[OH]2nH2O), whereas Jones et al. (2003) suggested that reticulate coatings associated with opal-A precipitates in one of the Tokaanu (New Zealand) geysers may be amorphous to poorly crystalline smectite, saponite or sepiolite. The reticulate coatings also resemble the honeycombed texture evident in some smectites and stratified chlorite-smectite (Welton, 1984).

The Fe-rich precipitates in the Zhuyuan precipitates that have a granular appearance are superficially similar to the hydrous ferric oxides (HFO) associated with bacterial deposits found around Orange Spring in the northern part of the Waiotapu geothermal area, New Zealand (Jones & Renaut, 2006; fig. 8). Those precipitates, however, developed in hot, acidic spring waters.

Microbes have been implicated in the development of reticulate coatings (Casanova, 1986, 1994; Casanova & Renaut, 1987; Jones & Renaut, 1996c). Such coatings are, for example, morphologically akin to fabrics found in stromatolitic layers in the Kopara of Polynesian atolls (Défarge et al., 1994, fig. 4), organic layers in lake deposits on Kiritimati in the Pacific Ocean (Trichet et al., 2001, fig. 6), and honeycomb aggregates of nontronite (Fe-rich octahedral smectite) that developed on a bacterial EPS template found in deep-sea sediments near Japan (Ueshima & Tazaki, 2001, fig. 2). Trichet et al. (2001) suggested that the honeycombed structure in their organic layers developed through the reorganization of the polysaccharidic and glycoproteic fibrous constituents of microbial sheaths. Microbialites from Satonda Crater Lake in Indonesia are formed, in part, of aragonite layers that alternate with amorphous Mg–Si layers (Arp et al., 2003, fig. 2B, D and E). Although Arp et al. (2003) did not illustrate the microstructure of the Mg–Si layers, these authors argued that CO2 released by decomposition of the EPS in entombed biofilms caused a decrease in pH that led to dissolution of Mg-calcite and simultaneous re-precipitation of Si to form the Si–Mg phase. Assigning a microbial origin to the reticulate coatings is commonly problematical because they are typically devoid of identifiable microbes (Jones & Renaut, 1995). Even in examples with identifiable microbes, it is impossible to determine whether they played a formative role in the development of the coatings, were trapped beneath the coating, or simply colonized the surfaces of the coatings following their formation (Jones & Renaut, 1995). Although the reticulate structure of these coatings is their definitive feature, its genesis is also one of the most difficult aspects to explain.

From an interpretive perspective, the critical features of the Si–Mg–Fe precipitates in the Gongxiaoshe and Zhuyuan deposits are as follows.

  • The close relation between the microbes and the reticulate and granular Si–Mg–Fe precipitates are highlighted by: (i) thin films of smooth, non-mineralized EPS that pass laterally into reticulate coatings (Fig. 7G and H) or granular Fe-rich precipitates (Fig. 15J); (ii) filamentous microbes coated with these precipitates (Figs 7D, 7E and 15C to F); (iii) strands of EPS or collapsed, non-mineralized filamentous microbes that lie on top of the reticulate coatings (Fig. 7I) or are embedded in the granular Fe-rich precipitates (Fig. 15K); and (iv) transitions from granular Fe-rich precipitates into reticulate coatings that both seem to have an underlying and common EPS framework (Fig. 15L).
  • The micron-scale spatial variability in the concentrations of Si, Mg and Fe (as judged from peak heights on EDX spectra), the general trend of decreasing Mg with increasing Fe and the apparent increase in the definition of the ridges in the reticulate coatings as the Si content increases.
  • Textural relations that show: (i) that development of the reticulate and granular Si–Mg–Fe coatings always post-dated precipitation of the aragonite or calcite; and (ii) the cyclic alternation between calcite precipitation and reticulate Si–Fe reticulate coatings (Fig. 15G).

The intimate association of non-mineralized EPS with the reticulate coatings shows that the reticulate morphology cannot be attributed to dehydration of the EPS that took place following sample collection and/or preparation, as has been suggested by Paterson (1982, 1995), Richards & Turner (1984) and McKenzie et al. (2001). The available evidence strongly indicates that microbial EPS play a formative role in the development of both the reticulate and the granular Si–Mg–Fe precipitates. The paucity of preserved microbes in the EPS means, however, that it is impossible to identify the organisms that generated the EPS. It seems, however, that the EPS absorbed the Si, Mg and Fe with the reticulate morphology progressively developing as more cations were absorbed. Although the reason for the formation of the granular as opposed to reticulate coatings remains debatable, it may be related to high Fe concentrations. It is possible, however, that these granular precipitates may be an Fe-rich variety of ACC.

Isotopes

The δ18O(water) (−10·75 to −11·39‰ VSMOW) and δ2H(water) (−83·1 to −85·0‰) values for the spring waters from Gongxiaoshe and Zhuyuan springs are directly comparable to the δ18O(water) (−10·99‰) and δ2H(water) (−76·8‰) values that Zhou et al. (2009, table 1, site #60) obtained and are within the range of δ18O(water) (−7·2 to −10·8‰) and δ2H(water) (−63·5 to 85·9‰) values that Zhang et al. (1987) used in their assessment of geothermal waters of the Rehai and Ruidian geothermal areas. According to Zhang et al. (1987) and Zhou et al. (2009), these values indicate that the waters in the Ruidian springs are largely of meteoric origin.

Calculations based on the δ18O(calcite) and δ18O(water) values, using the equations of Anderson & Arthur (1983), Hays & Grossman (1991), Kim & O'Neil (1997) and O'Neil et al. (1969), yielded temperatures of 71 to 77°C for Gongxiaoshe and 69 to 72°C for Zhuyuan (Table 3). The calculated temperatures for Gongxiaoshe are the same as the water temperature (73°C) measured above the vent, but are higher than the water temperature (50 to 55°C) measured around the margins of the pool where most of the samples were collected. The calculated temperatures for Zhuyuan are generally lower than the water temperature (82°C) that was measured on 16 June 2011 and on 15 October 2011. Nevertheless, comparisons of the calculated and measured temperatures suggest that the calcite is nearly in isotopic equilibrium with the spring waters.

Table 3. Temperatures (°C) calculated from average oxygen isotopes from calcite samples. δ18O (calcite) and δ18O(water) expressed on SMOW scale
 Calcite δ18OWater δ18OAnderson & Arthur (1983)Hays & Grossman (1991)Kim & O'Neil (1997)O'Neil et al. (1969)Measured water T (°C)
Gongxiaoshe9·26−10·757273717773
Zhuyuan9·26−11·356969677284

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Geological setting
  5. Methodology
  6. Terminology
  7. Results
  8. Interpretations
  9. Discussion
  10. Conclusions
  11. Acknowledgements
  12. References

Any model used to explain the evolution of the deposits found in the Gongxiaoshe (Fig. 2A to C) and Zhuyuan (Fig. 2D to G) springs, which are only 130 m apart with similar water chemistries, must recognize that these young deposits formed from hot spring waters in the confines of small spring pools. This model must also explain: (i) the compositional and crystallographic diversity of the deposits; (ii) the precipitation of ACC; (iii) aragonite being present in the Gongxiaoshe deposits but not in the Zhuyuan deposits; (iv) the formation of the Si–Mg–Fe reticulate coatings; (v) red, Fe-rich precipitates being common in the Zhuyuan deposits but poorly developed in the Gongxiaoshe deposits; and (vi) the near isotopic equilibrium between the calcite and the spring waters.

The precipitation of different polymorphs and/or crystal morphologies of CaCO3 has been attributed to various abiotic and biotic controls (Tourney & Ngwenya, 2009; Peng et al., 2010). Experimental studies have, for example, demonstrated that the abiotic factors include water temperature (Kitano, 1962b; Kitano et al., 1962), the presence of divalent ions, including Mg (Ries et al., 1998), pH levels (Tai & Chen, 1998), the rate and order at which reactants are added (Kitamura, 2001; Somani et al., 2006), the presence of various non-microbial organic molecules (Sondi & Matijevíc, 2001; Becker et al., 2003; Jimenez-Lopez et al., 2003; Lakshminarayanan et al., 2003; Tong et al., 2004; Sondi & Salopek-Sondi, 2005) and many other factors (Meldrum, 2003; Meldrum & Cölfen, 2008). Equally, experimental procedures have also shown that various biotic processes may be responsible for the precipitation of the different CaCO3 polymorphs and variations in the crystal morphology of those polymorphs (Meldrum, 2003; Meldrum & Cölfen, 2008; Tourney & Ngwenya, 2009). This category includes the ‘biologically induced’ processes whereby microbes directly influence precipitation and ‘biologically influenced’ processes, as used by Dupraz et al. (2009) and Decho (2010), whereby precipitation takes place passively as a result of the interaction of the extracellular organics and the geochemical environment. Many of the biologically influenced processes operate in the biofilms that facilitate microbial attachment to substrates and buffer them from the extracellular environment (Decho & Lopez, 1993; Decho, 2010). The highly hydrated EPS is particularly important in this respect (Arp et al., 1999; Kawaguchi & Decho, 2002; Braissant et al., 2003, 2007). This hydrogel, which is a three-dimensional polymeric network held in water, encompasses numerous micro-domains where various ions/molecules can reach high concentrations and thereby influence chemical processes that operate on a micron-scale (Kawaguchi & Decho, 2002; Decho, 2010).

At Gongxiaoshe, viable green biofilms are readily apparent on the substrates (Fig. 2B and C) and microbes and EPS are clearly evident in most samples (Fig. 9). At Zhuyuan, green biofilms are present on the surfaces above water level (Fig. 2D) but are not obvious on the subsurface precipitates (Fig. 2E to G). Nevertheless, SEM examination of those samples clearly shows the presence of microbes and copious amounts of EPS (Figs 11D to G, 12B, E and F and 13A to F). There is, however, an interesting dichotomy with respect to the microbes and the EPS. Where numerous microbes are evident, there appear to be little associated EPS (Fig. 9A to D). In contrast, much of the EPS lack any evidence of the formative microbes and extensive searching was required to find small areas with scattered non-mineralized microbes (Fig. 9F to I, L and M). The taxonomic affinity of the microbes found in these biofilms remains debatable because they cannot be matched with any extant taxa. Stone (2011), in describing microbiological research being undertaken on these springs, noted that the microbiota that form these mats have largely defied culturing and eluded identification.

The deposits at Gongxiaoshe and Zhuyuan springs may have grown through abiotic processes that are linked solely to the chemistry of the spring waters and/or biotic processes that operated in the microcosms of the biofilms that covered the substrates. Deriving a purely physicochemical model linked solely to the spring waters is difficult, given that it must explain: (i) the compositional variability of the different precipitates that formed; (ii) the variations in the crystal morphologies of the calcite and aragonite crystals; and (iii) the contrasts in the mineralogy between the two springs. With a purely abiotic model, it is also difficult to explain many of the micron-scale mineralogical and crystallographic variations. In contrast, a biologically influenced model with precipitation mediated largely in the micro-domains of the EPS can explain most of the features evident in the deposits from these two springs. Specifically, this model would allow: (i) the gradual growth of the complex pseudodentrites through the gradual addition of small crystals at the growth tips of the branches; (ii) the precipitation of ACC; (iii) the growth of both calcite and aragonite at Gongxiaoxhe but only calcite at Zhuyuan; (iv) the unrestricted growth of the euhedral calcite dodecahedra and rhombohedra; (v) the precipitation of calcite cements around the earlier formed pseudodendrites; (vi) the minor amounts of barite and gypsum precipitates found in the Zhuyuan deposits; and possibly (vii) the development of the Si–Mg–Fe precipitates.

The growth of natural and experimentally produced dendrite crystals has been attributed to abnormal crystal growth (Strickland-Constable, 1968) that may be related to highly supersaturated solutions (Hille et al., 1958; Cabrera & Coleman, 1963), temperature (Strickland-Constable, 1968), the presence of impurities (Buckley, 1951; Hill & Wanklyn, 1968; Doherty, 1980) or rapid supercooling of the parent solution (Goss, 1963). Jones & Renaut (1995, fig. 14), based on their analysis of calcite spring deposits from Kenya, argued that crystal morphology became more complicated as the disequilibrium factor in the parent fluid increased. According to the Jones & Renaut (1995) scheme, skeletal crystals formed under low disequilibrium conditions, dendrite crystals formed under intermediate disequilibrium conditions, and wheat sheaves, fans, and spherulites formed under the highest disequilibrium conditions. Oaki & Imai (2004, fig. 1) produced a very similar diagram based on experiments wherein crystals were grown in gels. These authors also argued that the different crystal morphologies reflected the driving force that included factors such as supersaturation and supercooling.

The fact that experimentally grown dendrites, irrespective of their cause, generally form very quickly (Jones & Kahle, 1986) commonly leads to the notion that natural calcite dendrites may also be the product of rapid bursts of precipitation that are associated with short-lived physicochemical conditions. Nevertheless, Jones et al. (2005) and Jones & Renaut (2008) demonstrated that some natural calcite dendrites found in spring deposits developed over long periods of time and commonly show evidence of seasonal growth spurts. Calcite dendrites from spring deposits in Lýsuhóll (Iceland) and Clinton (western Canada), for example, are formed of aggregates of crystals that are stacked on top of each other and therefore are similar to those found in the Gongxiaoshe and Zhuyuan springs. Incompletely formed calcite crystals with missing edges and/or crystal faces and calcite crystals formed of numerous sub-crystals are common in the Gongxiaoshe and Zhuyuan deposits (Figs 5B, E and F, 11H to K and 12). Similar crystals also characterize the spring deposits found at Lýsuhóll and Clinton. In the Chinese examples, many of these crystals appear to have grown in EPS. Such features are also commonly seen in crystals that have been experimentally grown in the presence of various gels, polymers or bacterial cells (Yang et al., 2003, figs 1 and 2b; Kulak et al., 2007, figs 2 to 5; Tourney & Ngwenya, 2009, figs 3 to 5).

The adoption of a biotic model to explain the precipitates in Gongxiaoshe and Zhuyuan springs provides a model whereby the diversity of precipitates and crystal forms that characterize those deposits can be explained. Nevertheless, it remains difficult to pinpoint the exact conditions that led to many of these precipitates and the different crystal forms. This is perhaps to be expected, given that much of this would have taken place in the micro-domains in the EPS where physicochemical conditions vary on micron and sub-micron scales. Perhaps the most puzzling aspect of these deposits are the Si–Mg–Fe precipitates and, in particular, the development of the reticulate coatings. There is evidence that their development did not coincide with the calcite/aragonite precipitation, as shown by: (i) reticulate coatings that commonly enveloped the calcite and aragonite crystals (Figs 7C, 15A and 15B); and (ii) locally formed thin laminae that separated layers of calcite (Fig. 15G). These patterns indicate that temporal changes in the composition of the spring water and/or behaviour of the spring may have been responsible. As in other settings where reticulate coatings are found, the lack of long-term records and a clear understanding of how the coatings formed preclude a definitive answer.

Particularly notable aspects of the precipitates from Gongxiaoshe and Zhuyuan are the incompletely formed calcite crystals and the ACC that is commonly associated with the EPS. Most of the incompletely formed calcite crystals seem to be growing and developing in ways that are inconsistent with the classical model of crystal growth. These unusual growth patterns are accentuated by the fact that some crystal faces seem to form through the merger of ACC (Figs 11K, 12E and 13G to I). Such growth, which is akin to what Meldrum & Cölfen (2008) termed ‘nonclassical crystallization’, may be a proxy for crystals that grew in the hydrogel of EPS.

Advancing a biologically influenced model for the development of the deposits found at Gongxiaoshe and Zhuyuan is possible because the spatial relations between the various precipitates and the EPS are still apparent. It seems highly probable, however, that evidence of the EPS and their formative microbes will be lost with time because they do not seem to be prone to calcification. Thus, it is easy to envisage a situation whereby there will be little or no evidence of the biofilms that mediated development of the precipitates. In stark contrast, opal-A precipitates that form in spring systems typically contain numerous well-preserved microbes along with mineralized EPS. For the carbonate spring deposits, identification of the biological input will probably have to rely on proxies that have yet to be fully identified. These may include features such as compositional diversity, dendrites formed of myriads of small crystals and/or the presence of incompletely formed crystals.

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Geological setting
  5. Methodology
  6. Terminology
  7. Results
  8. Interpretations
  9. Discussion
  10. Conclusions
  11. Acknowledgements
  12. References

The lithologically complex deposits from Gongxiaoshe and Zhuyuan, precipitated from hot spring waters (73 to 84°C), are formed of various combinations of: (i) calcite pseudodendrites that evolved as small crystals stacked one on top of the other; (ii) amorphous calcium carbonate (ACC); (iii) calcite dodecahedra and rhombohedra that are typically incompletely grown and unattached; (iv) crystallographically diverse aragonite with most crystals lying parallel to bedding and unattached to any substrate; (v) reticulate coatings that contain variable concentrations of Si and Mg with minor Fe; (vi) granular coatings that contain variable concentrations of Si and Fe with minor Mg; and (vii) minor amounts of barite and gypsum. Some substrates are covered with numerous large coccoid microbes, whereas others are covered with copious amounts of extracellular exopolymeric substances (EPS) that, in some areas, contain numerous filamentous microbes.

Available evidence indicates that the development of these precipitates was controlled by a biologically influenced precipitation model, rather than an abiotic model linked solely to temporal variations in the water chemistry. The biologically influenced model, with precipitation being controlled at the micron-scale in the micro-domains of the EPS hydrogel, provides a viable explanation for the microscale variations in the different precipitates and the variance in crystal morphologies. It appears, however, that both the microbes and the EPS had very low preservation potential. This means that evidence of the microbes would be rapidly lost and that older deposits of this type may contain little or no evidence of their biotic origin.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Geological setting
  5. Methodology
  6. Terminology
  7. Results
  8. Interpretations
  9. Discussion
  10. Conclusions
  11. Acknowledgements
  12. References

Financial support for this research came from the Natural Sciences and Engineering Council of Canada (to Jones), the National Natural Science Foundation of China (no. 41172309 to Peng) and the National Basic Research Program of China (no. 2012CB417300 to Peng). We are greatly indebted to Shun Chen and Hengchao Xu who collected water samples in October, 2011, George Braybrook who took the SEM images used in this paper, R. Drimmie (Isotope Tracer Technologies Inc., Ontario, Canada) who did the stable isotope analyses, and Dr. Giovanna Della Porta, Dr. Martyn Pedley and an anonymous journal reviewer who kindly provided thought-provoking reviews of the original manuscript.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Geological setting
  5. Methodology
  6. Terminology
  7. Results
  8. Interpretations
  9. Discussion
  10. Conclusions
  11. Acknowledgements
  12. References
  • Addadi, L., Raz, S. and Weiner, S. (2003) Taking advantage of disorder: amorphous calcium carbonate and its role in biomineralization. Adv. Mater., 15, 959970.
  • Allen, C.C., Albert, F.G., Chafetz, H.S., Combie, J., Graham, C.R., Kieft, T.L., Kivett, S.J., McKay, D.S., Steele, A., Taunton, A.E., Taylor, M.R. and Thomas-Keprta, K.L. (2000) Microscopic physical biomarkers in carbonate hot springs: implications in the search for life on Mars. Icarus, 147, 4967.
  • Anderson, T.F. and Arthur, M.A. (1983) Stable isotopes of oxygen and carbon and their application to sedimentologic and palaeoenvironmental problems. In: Stable Isotopes in Sedimentary Geology (Eds M.A. Arthur, T.F. Anderson, I.R. Kaplan, J. Veizer and L.S. Land), Short Course Notes, pp. 1.11.151. Society of Economic Paleontologists and Mineralogists, Tulsa, Oklahoma.
  • Arp, G., Thiel, V., Reimer, A., Michaelis, W. and Reitner, J. (1999) Biofilm exopolymers control microbialite formation at thermal springs discharging into the alkaline Pyramid Lake, Nevada, U.S.A. Sed. Geol., 126, 159176.
  • Arp, G., Reimer, A. and Reitner, J. (2003) Microbialite formation in seawater of increased alkalinity, satonda crater lake, Indonesia. J. Sed. Res., 73, 105127.
  • Beck, R. and Andreassen, J.-P. (2010) The onset of spherulitic growth in crystallization of calcium carbonate. J. Cryst. Growth, 312, 22262238.
  • Becker, A., Becker, W., Marxen, J.C. and Epple, M. (2003) In-vitro crystallization of calcium carbonate in the presence of biological additives – comparison of the ammonium carbonate method with double-diffusion techniques. Z. Anorg. Allg. Chem., 629, 23052311.
  • Binkley, K.L., Wilkinson, B.H. and Owen, R.M. (1980) Vadose beachrock cementation along a southeastern Michigan marl lake. J. Sed. Petrol., 50, 953962.
  • Blank, C.E., Cady, S.L. and Pace, N.R. (2002) Microbial composition of near-boiling silica-depositing thermal springs throughout Yellowstone National Park. Appl. Environ. Microbiol., 68, 51235135.
  • Bonny, S.M. and Jones, B. (2003) Microbes and mineral precipitation, Miette Hot Springs, Jasper National Park, Alberta, Canada. Can. J. Earth Sci., 40, 14831500.
  • Bosák, P., Bella, P., Cílek, V., Ford, D.C., Hercman, H., Kadlec, J., Osborne, A. and Pruner, P. (2002) Ochtiná aragonite cave (western Carpathians, Slovakia): morphology, mineralogy of the fill and genesis. Geol. Carpath., 53, 399410.
  • Braissant, O., Cailleau, G., Dupraz, C. and Verrecchia, E.P. (2003) Bacterially induced mineralization of calcium carbonate in terrestrial environments: the role of exopolysaccharides and amino acids. J. Sed. Res., 73, 485490.
  • Braissant, O., Decho, A.W., Dupraz, C., Glunk, C., Prezekop, K.M. and Visscher, P.T. (2007) Exopolymeric substances of sulphate-reducing bacteria: interactions with calcium at alkaine pH and implication for formation of carbonate minerals. Geobiology, 5, 401411.
  • Buckley, H.E. (1951) Crystal Growth. John Wiley and Sons Inc., London, 359 pp.
  • Buczynski, C. and Chafetz, H.C. (1991) Habit of bacterially induced precipitates of calcium carbonate and the influence of medium viscosity on mineralogy. J. Sed. Petrol., 61, 221233.
  • Cabrera, N. and Coleman, R.V. (1963) Theory of crystal growth from the vapor. In: The Art and Science of Growing Crystals (Ed. J.J. Gilman), pp. 328. Wiley, New York.
  • Cabrol, P. (1978) Contribution à l'étude du concrétionnement carbonaté des grottes du Sud de la France, morphologie, génèse, diagénèse. University of Montpellier, xxxxxx, 275 pp.
  • Cady, S.L. and Farmer, J.D. (1996) Fossilization processes in siliceous thermal springs: trends in preservation along thermal gradients. In: Evolution of Hydrothermal Ecosystems on Earth (and Mars?). Ciba Foundation Symposium (Eds G.R. Bock and J.A. Goode), pp. 150173. Wiley, Chichester, UK.
  • Campbell, K.A., Rodgers, K.A., Brotheridge, J.M.A. and Browne, P.R.L. (2002) An unusual modern silica-carbonate sinter from Pavlova spring, Ngatamariki, New Zealand. Sedimentology, 49, 835854.
  • Casanova, J. (1986) East Africa rift stromatolites. Geol. Soc. Spec. Pub., 25, 201210.
  • Casanova, J. (1994) Stromatolites from the East Africa Rift: a synopsis. In: Phanerozoic Stromatolites (Eds J. Bertrand-Safarti and C. Monty), pp. 193226. Kluwer, Dordrecht, The Netherlands.
  • Casanova, J. and Renaut, R.W. (Eds)(1987) Les stromatolites du milieu hydrothermal. Centres de Recherches Exploration – Production Elf, Aquitaine, 484490 pp.
  • Chafetz, H.S. and Folk, R.L. (1984) Travertines: depositional morphology and the bacterially constructed constituents. J. Sed. Petrol., 54, 289316.
  • Chafetz, H.S. and Guidry, S.A. (1999) Bacterial shrubs, crystal shrubs, and ray-crystal shrubs: bacterial vs. abiotic precipitation. Sed. Geol., 126, 5774.
  • Chafetz, H.S. and Guidry, S.A. (2003) Deposition and diagenesis of Mammoth Hot Springs travertine, Yellowstone National Park, Wyoming, USA. Can. J. Earth Sci., 40, 15151529.
  • Chafetz, H.S., Wilkinson, B.H. and Love, K.M. (1985) Morphology and composition of non-marine carbonate cements in near-surface settings. In: Carbonate Cements (Eds N. Schneidermann and P.M. Harris), pp. 337347. SEPM Spec. Publ., 36
  • Craig, K.D., Horton, P.D. and Reams, M.W. (1984) Clastic and solutional boundaries as nucleation surfaces for aragonite in speleothems. Bull. Natl Speleol. Soc., 46, 1517.
  • Decho, A.W. (1990) Microbial exopolymer secretions in ocean environments: their role(s) in food webs and marine processes. Oceanogr. Mar. Biol. Annu. Rev., 28, 73153.
  • Decho, A.W. (2000) Microbial biofilms in intertidal systems: an overview. Cont. Shelf Res., 20, 12571273.
  • Decho, A.W. (2010) Overview of biopolymer-induced mineralization: what goes on in biofilms? Ecol. Eng., 36, 137144.
  • Decho, A.W. and Lopez, G.R. (1993) Exopolymer microenvironments of microbial flora: multiple and interactive effects on trophic relationships. Limnol. Oceanogr., 38, 16331645.
  • Défarge, C., Trichet, J., Maurin, A. and Hucher, M. (1994) Kopara in Polynesian atolls: early stages of formation of calcareous stromatolites. Sed. Geol., 89, 923.
  • Doherty, R.D. (1980) Dendritic growth crystal growth. Sci. Solid State, 16, 485520.
  • Du, J., Liu, C., Fu, B., Ninomia, Y., Zhang, Y., Wang, C., Wang, H. and Sun, Z. (2005) Variations of geothermometry and chemical-isotope compositions of hot spring fluids in the Rehai geothermal field, southwestern China. J. Volcanol. Geoth. Res., 142, 243261.
  • Dupraz, C., Visscher, P.T., Baumgartner, L.K. and Reid, R.P. (2004) Microbe–mineral interactions: early carbonate precipitation in a hypersaline lake (Eleuthera Island, Bahamas). Sedimentology, 51, 745765.
  • Dupraz, C., Reid, R.P., Braissant, O., Decho, A.W., Norman, R.S. and Visscher, P.T. (2009) Processes of carbonate precipitation in modern microbial mats. Earth-Sci. Rev., 96, 141162.
  • Fischbeck, R. and Müller, G. (1971) Monohydrocalcite, hydromagnesite, nesqueehonite, dolomite, aragonite, and calcite in speleothems of the Frankische Schweiz, Western Germany. Contrib. Mineral. Petrol., 33, 8792.
  • Folk, R.L. (1993) SEM imaging of bacteria and nannobacteria in carbonate sediments and rocks. J. Sed. Petrol., 63, 990999.
  • Folk, R.L. (1994) Interaction between bacteria, nannobacteria, and mineral precipitation in hot springs of central Italy. Geog. Phys. Quatern., 48, 233246.
  • Folk, R.L. and Rasbury, E.T. (2007) Nanostructure of palygorsite/sepiolite in Texas caliche: possible bacterial origin. Carbonates Evaporites, 22, 113122.
  • Fouke, B.W. (2011) Hot-spring systems geobiology: abiotic and biotic influences on treavertine formation at Mammoth Hot Springs, Yellowstone National Park, USA. Sediementology, 58, 170219.
  • Fouke, B.W., Farmer, J.D., Des Marais, D.J., Pratt, L., Sturchio, N.C., Burns, P.C. and Discipulo, M.K. (2000) Depositional facies and aqueous-solid geochemistry of travertine-depositing hot springs (Angel Terrace, Mammoth Hot Springs, Yellowstone National Park, USA). J. Sed. Res., 70, 565585.
  • Geptner, A., Kristmannsdóttir, H., Kristjansson, J. and Marteinsson, V. (2002) Biogenic saponite from an active submarine hot spring, Iceland. Clay Clay Mineral., 50, 174185.
  • Given, R.K. and Wilkinson, B.H. (1985) Kinetic control of morphology, composition and mineralogy of abiotic sedimentary carbonates. J. Sed. Petrol., 55, 109119.
  • González-Muñoz, M.T., Rodriguez-Navarro, C., Martínez-Ruiz, F., Arias, J.M., Merroun, M.L. and Rodrizguez-Gallego, M. (2010) Bacterial biomineralization: new insights from Myxococcus-induced mineral precipitation. In: Tufas and Speleothems: Unravelling the Microbial and Physical Controls (Eds H.M. Pedley and M. Rogerson), pp. 3150. Geol. Soc. London Spec. Publ., 336, London.
  • Goss, A.J. (1963) Low melting point elements. In: The Art and Science of Growing Crystals (Ed. J.J. Gilman), pp. 314342. Wiley, New York.
  • Guidry, S.A. and Chafetz, H.S. (2003) Depositional facies and diagenetic alteration in a relict siliceous hot-spring accumulation: examples from Yellowstone National Park, USA. J. Sed. Res., 73, 806823.
  • Gunnlaugsson, E. and Einarsson, A. (1989) Magnesium-silicate scaling in a mixture of geothermal water and deaerated fresh water in a district heating system. Geothermics, 18, 113120.
  • Guo, L. and Riding, R. (1998) Hot-spring travertine facies and sequences, Late Pleistocene, Rapolano Terme, Italy. Sedimentology, 45, 163180.
  • Guo, L. and Riding, R. (1999) Rapid facies changes in Holocene fissure ridge hot spring travertines, Rapolano Terme, Italy. Sedimentology, 46, 11451158.
  • Hays, P.D. and Grossman, E.L. (1991) Oxygen isotopes in meteroic calcite cements as indicators of continental palaeoclimate. Geology, 19, 441444.
  • Hetzer, A., Morgan, H.W., McDonald, I.R. and Daughney, C.J. (2007) Microbial life in Champagne Pool, a geothermal spring in Waiotapu, New Zealand. Extremophiles, 11, 605614.
  • Hill, C.A. and Forti, P. (1997) Cave Minerals of the World. National Speleological Society, Inc., Huntsville, Alabama.
  • Hill, G.J. and Wanklyn, B.M. (1968) Fluxed-melt growth of nickel oxide crystals and their electrical properties. J. Cryst. Growth, 3, 475479.
  • Hille, M., Rau, H. and Schlipf, J. (1958) Concerning the crystallographic orientation of salt dendrites growth and perfection of crystals. In: Growth and Perfection of Crystals (Eds R.H. Doremus, B.W. Roberts and D. Turnbull), pp. 475479. Wiley, New York.
  • Inagaki, F., Motomura, Y., Doi, K., Taguchi, S., Izawa, E., Lowe, D.R. and Ogata, S. (2001) Silicified microbial community at Steep Cone hot spring, Yellowstone National Park. Microbes Environ., 16, 125130.
  • Jimenez-Lopez, C., Rodriguez-Navarro, A., Dominguez-Vera, J.M. and Garcia-Ruiz, J.M. (2003) Influence of lysozyme on the precipitation of calcium carbonate: a kinetic and morphologic study. Geochim. Cosmochim. Acta, 67, 16671676.
  • Jones, B. (2010) Speleothems in a wave-cut notch, Cayman Brac, British West Indies: the integrated product of subaerial precipitation, dissolution, and microbes. Sed. Geol., 232, 1534.
  • Jones, B. and Kahle, C.F. (1986) Dendritic calcite crystals formed by calcification of algal filaments in a vadose environment. J. Sed. Petrol., 56, 217227.
  • Jones, B. and Kahle, C.F. (1993) Morphology, relationship, and origin of fiber and dendrite calcite crystals. J. Sed. Petrol., 63, 10181031.
  • Jones, B. and Peng, X. (2012) Amorphous calcium carbonate associated with biofilms in hot spring deposits. Sed. Geol., 269-270, 5868.
  • Jones, B. and Renaut, R.W. (1995) Noncrystallographic dendrites from hot-spring deposits at Lake Bogoria, Kenya. J. Sed. Res., 65, 154169.
  • Jones, B. and Renaut, R.W. (1996a) Morphology and growth of aragonite crystals in hot-spring travertines at Lake Bogoria, Kenya Rift Valley. Sedimentology, 43, 323340.
  • Jones, B. and Renaut, R.W. (1996b) Skeletal crystals of calcite and trona from hot-spring deposits in Kenya and New Zealand. J. Sed. Res., 66, 265274.
  • Jones, B. and Renaut, R.W. (1996c) Influence of thermophilic bacteria on calcite and silica precipitation in hot springs with water temperatures above 90°C: evidence from Kenya and New Zealand. Can. J. Earth Sci., 33, 7283.
  • Jones, B. and Renaut, R.W. (1998) Origin of platy calcite crystals in hot-spring deposits in the Kenya Rift Valley. J. Sed. Res., 68, 913927.
  • Jones, B. and Renaut, R.W. (2006) Selective mineralization of microbes in Fe-rich precipitates (jarosite, hydrous ferric oxides) from acid hot springs in the Waiotapu geothermal area, North Island, New Zealand. Sed. Geol., 194, 7798.
  • Jones, B. and Renaut, R.W. (2008) Cyclic development of large, complex calcite dendrite crystals in the Clinton travertine, Interior British Columbia, Canada. Sed. Geol., 203, 1735.
  • Jones, B., Renaut, R.W. and Rosen, M.R. (1996) High-temperature (> 90°C) calcite precipitation at Waikite Hot Springs, North Island, New Zealand. J. Geol. Soc. London, 153, 481496.
  • Jones, B., Renaut, R.W. and Rosen, M.R. (1997) Biogenicity of silica precipitation around geysers and hot-spring vents, North Island, New Zealand. J. Sed. Res., 67, 88104.
  • Jones, B., Renaut, R.W. and Rosen, M.R. (2000) Trigonal dendritic calcite crystals forming from hot spring waters at Waikite, North Island, New Zealand. J. Sed. Res., 70, 586603.
  • Jones, B., Renaut, R.W. and Rosen, M.R. (2003) Silicified microbes in a geyser mound: the enigma of low-temperature cyanobacteria in a high-temperature setting. Palaios, 18, 87109.
  • Jones, B., Konhauser, K.O., Renaut, R.W. and Wheeler, R. (2004) Microbial silicification in Iodine Pool, Waimangu Geothermal area, North Island, New Zealand: implications for recognition and identification of ancient silicified microbes. J. Geol. Soc. London, 161, 983993.
  • Jones, B., Renaut, R.W., Owen, R.B. and Torfason, H. (2005) Growth patterns and implications of complex dendrites in calcite travertines from Lýsuhóll, Snæfellsnes, Iceland. Sedimentology, 52, 12771301.
  • Kawaguchi, T. and Decho, A.W. (2002) A laboratory investigation of cyanobacterial extracellular polymeric secretions (EPS) in influencing CaCO3 polymorphism. J. Cryst. Growth, 240, 230235.
  • Kearey, P. and Wei, H. (1993) Geothermal fields of China. J. Volcanol. Geoth. Res., 56, 415428.
  • Keith, H.D. and Padden, F.J. (1964) Spherulitic crystallization from the melt. I. Fractionation and impurity segregation and their influence on crystalline morphology. J. Appl. Phys., 35, 12701285.
  • Kim, S.-T. and O'Neil, J.R. (1997) Equilibrium and nonequilibrium oxygen isotope effects in synthetic carbonates. Geochim. Cosmochim. Acta, 61, 34613475.
  • Kitamura, M. (2001) Crystallization and transformation mechanism of calcium carbonate polymorphs and the effect of magnesium ion. J. Colloid Interf. Sci., 236, 318327.
  • Kitano, Y. (1962a) The behavior of various inorganic ions in the separation of calcium carbonate from a bicarbonate solution. Bull. Chem. Soc. Jpn., 35, 19731980.
  • Kitano, Y. (1962b) A study of the polymorphic formation of calcium carbonate in thermal springs with an emphasis of the effect of temperature. J. Earth Sci. Nagoya Univ., 35, 19801985.
  • Kitano, Y., Park, K. and Hood, D. (1962) Pure aragonite synthesis. J. Geophys. Res., 67, 48734874.
  • Kristmannsdottir, H., Olafsson, M. and Thorhallsson, S. (1989) Magnesium silicate scaling in district heating systems in Iceland. Geothermics, 18, 191198.
  • Kulak, A.N., Iddon, P., Li, Y., Armes, S.P., Cölfen, H., Paris, O., Wilson, R.M. and Meldrum, F.C. (2007) Continuous structural evolution of calcium carbonate particles: a unifying model of copolymer-mediated crystallization. J. Am. Chem. Soc., 12, 37293736.
  • Lakshminarayanan, R., Valiyaveettil, S. and Loy, G.L. (2003) Selective nucleation of calcium carbonate polymorphs: role of surface functionalization and poly(vinyl alcohol) additive. Cryst. Growth Des., 3, 953958.
  • Léveillé, R.J., Fyfe, W.S. and Longstaffe, F.J. (2000a) Geomicrobiology of carbonate–silicate microbialites from Hawaiian basaltic sea caves. Chem. Geol., 169, 339355.
  • Léveillé, R.J., Fyfe, W.S. and Longstaffe, F.J. (2000b) Unusual secondary Ca-Mg-carbonate-kerolite deposits in basaltic caves, Kauai, Hawaii. J. Geol., 108, 613621.
  • Liao, Z. and Guo, G. (1986) Geology of the Tengchong geothermal field and surrounding area, west Yunnan, China. Geothermics, 15, 339345.
  • Lofgren, G. (1974) An experimental study of plagioclase crystal morphology: isothermal crystallization. Am. J. Sci., 274, 243273.
  • McKenzie, E.J., Brown, K.L., Cady, S.L. and Campbell, K.A. (2001) Trace metal chemistry and silicification of microorganisms in geothermal sinter, Taupo Volcanic Zone, New Zealand. Geothermics, 30, 483502.
  • Meldrum, F.C. (2003) Calcium carbonate in biomineralisation and biomimetic chemistry. Int. Mater. Rev., 48, 187224.
  • Meldrum, F.C. and Cölfen, H. (2008) Controlling mineral morphologies and structures in biological and synthetic systems. Chem. Rev., 108, 43324432.
  • Moore, G.W. (1956) Aragonite speleothems as indicators of paleotemperature. Am. J. Sci., 254, 746753.
  • Neu, T.R. (1996) Significance of bacterial surface-active compounds in interaction of bacteria with interfaces. Microbiol. Rev., 60, 151166.
  • Neu, T.R. and Lawrence, J.R. (2009) Extracellular polymeric substances in microbial biofilms. In: Microbial Glycobiology. Structures, Relevance and Applications (Eds O. Holst, P.J. Brennan and M. von Itzstein), pp. 735758. Elsevier, Amsterdam.
  • Oaki, Y. and Imai, Y. (2004) Experimental demonstration for the morphological evolution of crystals grown in gel media. Cryst. Growth Des., 3, 711716.
  • Obst, M., Dynes, J.J., Lawrence, J.R., Swerhone, G.D.W., Benzerara, K., Karunakaran, C., Kaznatcheev, K., Tyliszczak, T. and Hitchcock, A.P. (2009) Precipitation of amorphous CaCO3 (aragonite-like) by cyanobacteria: A STXM study of the influence of EPS on the nucleation process. Geochim. Cosmochim. Acta, 73, 41804198.
  • O'Neil, J., Clayton, R. and Mayeda, T. (1969) Oxygen isotope fractionation in divalent metal carbonates. J. Chem. Phys., 51, 55475558.
  • Özkul, M., Varol, B. and Alçiçek, M.C. (2002) Depositional environments and petrography of Denizli travertines. Bull. Mineral Res. Explor., 125, 1329.
  • Parkhurst, D.L. and Appelo, C.A.J. (1999) User's guide to PHREEQC (version 2) – a computer program for speciation, reaction-path, 1D-transport, and inverse geochemical calculations. United States Geological Survey, Water Resource Investigations Report, 99-4259, 312 pp.
  • Paterson, M.S. (1982) The determination of hydroxyl by infrared absorption in quartz, silicate glasses and similar materials. Bull. Minér., 105, 2029.
  • Paterson, D.M. (1995) Biogenic structure of early sediment fabric visualized by low-temperature scanning electron microscopy. J. Geol. Soc. London, 152, 131140.
  • Pedley, H.M. and Rogerson, M. (2010) In vitro investigations of the impact of different temperature and flow velocity conditions on tufa microfabric. In: Tufas and Speleothems: Unravelling the Microbial and Physical Controls (Eds H.M. Pedley and M. Rogerson), Geol. Soc. London Spec. Publ., 336, 193210.
  • Peng, X., Zhou, H., Wu, J., Jiang, L., Tang, S., Yao, H. and Chen, G. (2007) Biomineralization of phototrophic microbes in silica-enriched hot springs in South China. Chin. Sci. Bull., 52, 367379.
  • Peng, X., Zhou, H., Yao, H., Li, J. and Wu, Z. (2010) Ultrastructural evidence for a novel accumulation of Ca in a microbial mat from a slight acidic hot spring. Acta Geol. Sinica, 84, 624631.
  • Pentecost, A., Bayari, S. and Yesertener, C. (1997) Phototropic microorganisms of the Pamukkle travertine, Turkey: their distribution and influence on travertine deposition. Geomicrobiol J., 14, 269283.
  • Pobéguin, T. (1955) Sur le concrétions calcaires observées dans la grotte de Moulis (Ariége). C. R. Hebd Seances Acad. Sci., 241, 17911793.
  • Pobéguin, T. (1957) Reproduction expérimentale de concrétions de carbonate de calcium. Ann. Spéléol., 12, 512.
  • Polyak, V.J. and Güven, N. (2000) Authigenesis of trioctahedral smectite in Magnesium-rich carbonate speleothems in Carlsbad Cavern and other caves in the Guadalupe Mountains, New Mexico. Clay Clay Mineral., 48, 317321.
  • Polyak, V.J. and Güven, N. (2004) Silicates in carbonate speleothems, Guadaloupe Mountains, New Mexico, U.S.A. In: Studies of Cave Sediments Physical and Chemical Records of Paleoclimate (Eds I.D. Sasowsky and J.E. Mylroice), pp. 303311. Kluwer Academic, New York.
  • Rainey, D.K. and Jones, B. (2009) Abiotic versus biotic controls on the development of the Fairmont hot Springs carbonate deposit, British Columbia, Canada. Sedimentology, 56, 18321857.
  • Renaut, R.W. and Jones, B. (1997) Controls on aragonite and calcite precipitation in hot spring travertines at Chemurkeu, Lake Bogoria, Kenya. Can. J. Earth Sci., 34, 801814.
  • Reysenbach, A.L., Ehringer, M. and Hershberger, K.L. (2000) Microbial diversity at 83°C in Calcite Springs, Yellowstone National Park: another environment where Aquificales and “Korarchaeota” coexist. Extremophiles, 4, 6167.
  • Richards, R.S. and Turner, R.J. (1984) A comparative study of techniques for the examination of biofilms by scanning electron microscopy. Water Res., 18, 767773.
  • Ries, J.B., Delgardo, G., Ramos-Cormenzana, A. and Delgardo, R. (1998) Biomineralization of carbonates by Holomonas eurihalina in solid and liquid media with different salinities: crystal formation sequence. Res. Microbiol., 149, 277287.
  • Rosenberg, E. (1989) Biofilms on water-soluble substrates. In: Structure and Function of Biofilms (Eds W.G. Characklis and P.A. Wilderer), pp. 5972. Wiley, Chichester, UK.
  • Rowling, J. (2004) Studies on aragonite and its occurrence in caves, including New South Wales caves. J. Proc. Roy. Soc. NSW, 137, 123149.
  • Sandberg, P. (1985) Aragonite cements and their occurrence in ancient limestones. In: Carbonate Cements (Eds N. Schneidermann and P.M. Harris), SEPM Spec. Publ., 36, 3357.
  • Seemann, R. (1985) Hydromagnesit und Degleitmineralien aus den Frauenmauer – Langstein - Höhlensystem, Hochschwab, Steiermark. Mitt. Mineral. Petrogr. Abtei. Landesmuseum Joann, 53, 1121.
  • Siegel, F.R. (1965) Aspects of calcium carbonate deposition in Great Onyx Cave, Kentucky. Sedimentology, 4, 285299.
  • Somani, R.S., Patel, K.S., Mehta, A.R. and Jasra, R.V. (2006) Examination of the polymorphs and particle size of calcium carbonate precipitated using still effluent (e.g. CaCl2+NaCl solution) of soda ash manufacturing process. Ind. Eng. Chem. Res., 45, 52235230.
  • Sondi, I. and Matijevíc, E. (2001) Homogeneous precipitation of calcium carbonate by enzyme catalyzed reaction. J. Colloid Interf. Sci., 238, 208214.
  • Sondi, I. and Salopek-Sondi, B. (2005) Influence of the primary structure of enzymes on the formation of CaCO3 polymorphs: a comparison of plant (Canavalia ensiformis) and bacterial (Bacillus pasteurii) ureases. Langmuir, 21, 88768882.
  • Stone, R. (2011) Biological dark matter exerts irresistible pull in Yunnan. Science, 333, 1209.
  • Strickland-Constable, R.F. (1968) Kinetics and Mechanism of Crystallization. Academic Press, London, New York.
  • Sverrisdottir, G., Kristmannsdottir, H. and Olafsson, M. (1992) Magnesium silicate scales in geothermal utilization. In: Water-Rock Interaction (Eds Y.K. Kharaka and A.S. Maest), pp. 14311434, Balkema, Rotterdam.
  • Tai, C.Y. and Chen, F. (1998) Polymorphism of CaCO3 precipitated in a constant-composition environment. AIChE J., 44, 17901798.
  • Tong, H., Ma, W., Wang, L., Wan, P., Hu, J. and Cao, L. (2004) Control over the crystal phase, shape, size and aggregation of calcium carbonate via L-aspartic acid inducing process. Biomaterials, 25, 39233929.
  • Tong, W., You, M., Liu, S.J. and Zhang, M. (1986) Hydrothermal systems in Tengchong, west Yunnan, China and their evolution during the past 400 years. Geothermics, 15, 331337.
  • Tourney, J. and Ngwenya, B.T. (2009) Bacterial extracellular polymeric substances (EPS) mediate CaCO3 morphology and polymorphism. Chem. Geol., 262, 138146.
  • Towe, K.M. (1967) Echinoderm calcite: single crystal or polycrystalline aggregate. Science, 157, 10481050.
  • Tracey, S.L., Williams, D.A. and Jennings, H.M. (1998) The growth of calcite spherulites from solution II. Kinetics of formation. J. Cryst. Growth, 193, 382388.
  • Trichet, J., Défarge, C., Tribble, J., Tribble, G. and Sansone, F. (2001) Christmas Island lagoonal lakes, models for the deposition of carbonate-evaporite-organic laminated sediments. Sed. Geol., 140, 177189.
  • Turner, E.C. and Jones, B. (2005) Microscopic calcite dendrites in cold-water tufa: implications for nucleation of micrite and cement. Sedimentology, 52, 10431066.
  • Ueshima, M. and Tazaki, K. (2001) Possible role of microbial polysaccharides in nontronite formation. Clay Clay Mineral., 49, 292299.
  • Urbani, F. (1997) Venezuelan cave minerals: a short overview. In: Proceedings of 12th International Congress of Speleology (Ed P.-Y. Jeannin), 7: Physical Speleology, pp. 243246. International Union of Speleology and Swiss Speleological Society.
  • Weed, W.H. (1889) On the formation of siliceous sinter by the vegetation of thermal springs. Am. J. Sci., 37, 351359.
  • Welton, J.E. (1984) SEM Petrology Atlas. American Association of Petroleum Geologists, Tulsa, OK.
  • Westall, F., Steele, A., Toporski, J., Walsh, M., Allen, C., Guidry, S.A., McKay, D., Gibson, E.K. and Chafetz, H.S. (2000) Polymeric substances and biofilms as biomarkers in terrestrial materials: implications for extraterrestrial samples. J. Geophys. Res., 105, 24.51124.527.
  • Yang, D., Qi, L. and Ma, J. (2003) Well-defined star-shaped calcite crystals formed in agarose gels. Chem. Commun., 10, 11801181.
  • Zhang, Z., Liu, S. and Zhai, F. (1987) Geochemistry of thermal waters in the Tengchong volcanic area, west Yunnan Province, China. Geothermics, 16, 169179.
  • Zhang, G., Liu, C., Liu, H., Jin, Z., Han, G. and Li, L. (2008) Geochemistry of the Rehai and Ruidan geothermal waters, Yunnan Province, China. Geothermics, 37, 7383.
  • Zhou, X., Fang, B., Zhou, H., Li, J. and Wang, Y. (2009) Isotopes of deuterium and oxygen-18 in thermal groundwater in China. Environ. Geol., 57, 18071814.
  • Zhu, M. and Tong, W. (1987) Surface hydrothermal minerals and their distribution in the Tengchong geothermal area, China. Geothermics, 16, 181195.