FOR MA TION OF CAL CITE BY CHEMOLITHOAUTOTROPHIC BAC TE RIA – A NEW HYPOTHESIS, BASED ON MICROCRYSTALLINE CAVE PISOIDS

FOR MA TION OF CAL CITE BY CHEMOLITHOAUTOTROPHIC

BAC TE RIA – A NEW HYPOTHESIS,

BASED ON MICROCRYSTALLINE CAVE PISOIDS

In sti tute of Geo log i cal Sci ences, Jagiellonian Uni ver sity, Oleandry 2a, 31-063 Kraków, Po land; e-mail: michal.gradzinski@uj.edu.pl

2

De part ment of Mi cro bi ol ogy, Ag ri cul tural Uni ver sity of Cra cow, Al. Mickiewicza 24/28, 31-120 Kraków, Po land; e-mail: mjchmiel@poczta.onet.pl

3

Fac ulty of Ge ol ogy, Geo phys ics and En vi ron men t Pro tec tion, AGH Uni ver sity of Sci ence and Tech nol ogy, Al. Mickiewicza 30, 30-059 Kraków, Po land; e-mail: motyka@agh.edu.pl

Gradziñski, M., Chmiel, M. J. & Motyka, J., 2012. For ma tion of cal cite by chemolithoautotrophic bac te ria – a new hy poth e sis, based on microcrystalline cave pisoids. Annales Societatis Geologorum Poloniae, 82: 361–369. Ab stract: A new mech a nism, stim u lat ing the pre cip i ta tion of cal cite, is pos tu lated. The supersaturation with re spect to car bon ate min er als is changed, as a re sult of CO2 con sump tion by chemolithoautotrophic, hy dro

gen-ox i diz ing bac te ria. This mech a nism con trols the growth of atyp i cal, microcrystalline cave pisoids in Perlová Cave, in Slovakia. The pisoids grow un der calm con di tions in rimstone pools, where they are bathed con tin u ously in stag nant wa ter. The wa ter is su per sat u rated, with re spect to cal cite and ar agon ite. The bac te ria in habit the outer parts of the pisoids, cov ered by biofilms. The biofilm in flu ences the sup ply of the Ca2+ ion, slows down the pre ci-p i ta tion rate, and fa vors cal cite ci-pre cici-p i ta tion over that of ar agon ite. The cal cite ini tially ci-pre cici-p i tates as bac te rial rep li cas, which fur ther act as seeds for the grow ing cal cite crys tals. This pro cess leads to the oblit er a tion of the pri mary, bac te rial fab rics. Since hy dro gen-ox i diz ing bac te ria oc cur in a wide spec trum of nat u ral hab i tats, the mech a nism of cal ci fi ca tion, pos tu lated above, also may op er ate in other en vi ron ments.

Key words: mi cro bial car bon ates, biomineralization, biofilm, speleothems, Carpathians. Manu script re ceived 20 November 2012, ac cepted 20 December 2012

IN TRO DUC TION

Bac te ria are ubiq ui tous or gan isms, ex ist ing al most ev -ery where, from the deep subsurface to the at mo sphere. They have the abil ity to stim u late the pre cip i ta tion of min er -als, both in side and out side their cells (see Ehrlich, 1996, 1999, for re view). The role of bac te ria in the pre cip i ta tion of car bon ate min er als has been dis cussed over the last hun dred years and it has been con firmed, both in na ture and the lab o -ra tory (Rid ing, 2000). Sev eral mech a nisms, driven by nonphotosynthetic bac te ria, lead to the pre cip i ta tion of car bon ate min er als. Most of them in volve heterotrophic bac te -ria (Castanier et al., 2000; Wright and Oren, 2005).

The precipitation of car bon ate min er als, un der the in flu ence of bac te ria, has been rec og nized in ma rine and ter -res trial en vi ron ments. In ter -res trial set tings, this pro cess is op er a tive in soils (Boquet et al., 1973; Mon ger et al., 1991; Braissant et al., 2003), tufas (Pedley, 2000), travertines (Re-naut and Jones, 2000), and caves (e.g., Jones, 2001, 2010, 2011b; Melim et al., 2001; Northup and Lavoie, 2001; Bar-ton and Northup, 2007; Blyth and Frisia, 2008; Baskar et al., 2011). Jones and Mac Don ald (1989) and Jones (2009) have

doc u mented microcrystalline lay ers in cave pisoids (called cave pearls) from Grand Cayman Is land that orig i nated un -der the in flu ence of mi crobes. The or i gin of the ma jor ity of the cave pisoids, which are com posed of sparry crys tals, has been at trib uted mainly to physicochemical pro cesses that are con trolled largely by the supersaturation lev els of the par ent wa ter, with re spect to cal cite or ar agon ite (e.g., Gra-dziñski and Radomski, 1967; Hill and Forti, 1997, p. 84–86; Nader, 2007; Melim and Spilde, 2011).

The pres ent ac count de scribes a mech a nism, by which chemolithoautotrophic, hy dro genox i diz ing bac te ria can in -flu ence the pre cip i ta tion of cal cite and in this way play a cri- tical role in the for ma tion of microcrystalline cave pisoids. As well, sim i lar, but as yet unrecognized, mech a nisms can op er ate in other en vi ron ments.

EN VI RON MEN TAL SET TING

Perlová Cave (in Slo vak Perlová jaskyòa) is lo cated in Slovakia, in the north ern part of the Great Fatra Moun tains (in Slo vak Velká Fatra), which form part of the West ern

Carpathians (Fig. 1). Its en trance is in Belanská Val ley, at an al ti tude of 910 m (19°06'06''E, 48°57'46''N). The cave is de vel oped in bed ded, Mid dle Tri as sic lime stone, be long ing to the Krížná unit, which was thrusted over a Me so zoic, autochthonous cover of the crys tal line core of the Great Fatra Moun tains (Mahe¾, 1968). The area above the cave is cov ered with a de cid u ous for est and the thick ness of the rocks above the cave is about 10 m.

The cave is 408 m long (Fig. 2; Holúbek and Kleskeò, 1993). Its in ter nal tem per a ture, ac cord ing to mea sure ments by the au thors, var ies be tween 5.1 °C and 6.8 °C. The wa ter is ponded in small, stepped rimstone pools (Fig. 3). The depth of the pools ranges from 2 cm to 6 cm, the larg est be -ing 1 × 1.2 m. Each pool con tains from about a dozen to sev eral hun dred pisoids. The wa ter is sup plied only dur ing the rainy sea son, by drip ping and mainly by spill ing over the rim from the higher pools to the lower ones. The wa ter in the pools is nearly stag nant. In tact, frag ile moonmilk rims tes tify that the wa ter is never strongly ag i tated. No pisoids are ce mented to the pool bot tom.

MA TE RI ALS AND METH ODS

Wa ter tem per a ture, pH, and spe cific elec tri cal con duc -tance (SEC) were mea sured in the field. The to tal al ka lin ity (as bi car bon ate HCO3) was de ter mined, us ing 0.05 mo lar HCl acid by Gran ti tra tion. Chlo ride (Cl) con tents were de -ter mined by the method of Mohr, us ing 0.01 mo lar AgNO3, while ni trate (NO3) con tents were de ter mined by the cap il -lary electrophoresis method, us ing 270 AH-T equip ment, a PerkinElmer prod uct. The concentrations of other com po -nents were de ter mined by in duc tively cou pled

plasma-ato-mic emis sion spec tros copy (ICP AES), us ing a Perkin-Elmer prod uct OP TIMA 7300DV. The DIC and equi lib ria were cal -cu lated for wa ter sam ples, us ing the pro gram PHREEQC (Park hurst and Appelo, 1999). The sat u ra tion in dex (SI) has been ap plied, as a mea sure of equi lib rium, ac cord ing to the for mula: SI = log (IAP/KT), where IAP is an ionic ac tiv ity prod uct for a given min eral, and KT is a sol u bil ity prod uct for that min eral.

Some pisoids were col lected asep ti cally, placed in auto- claved glass flasks, stored in a re frig er a tor and de liv ered to the lab o ra tory within 24 hours. For mi cro bi o log i cal anal y -sis, 10 g of each sam ple were cen tri fuged in phys i o log i cal

362

Fig. 1. Lo ca tion of Perlová Cave

Fig. 2. Map of Perlová Cave, af ter Holúbek and Kleskeò (1993), sim pli fied; big ar row in di cates cave en trance, ar rows in di -cate sam pling sites: 1. Pearl Pas sage (Perlová chodba) – pools 1, 4–6; 2. Par lia ment Cham ber (Parlament) – pools 8–10

Fig. 3. Stepped pools with pisoids, Pearl Pas sage, scale bar is 3 cm long. Pho to graph from Gradziñski (2001)

salt, shaken and later in cu bated at 20 °C and 35 °C from 1 to 21 days. The growth of mi cro-or gan isms was sys tem at i cally mon i tored. The following, mi cro bi o log i cal me dia were used for iso la tion: Beaf Ex tract – Nu tri ent Broth – Merck, Trypticase Soy Broth (Soy bean-Ca sein Di gest Me dium) – BioMerieux, Nu tri ent Agar – Merck, TSA (Trypcase Soy Agar) – BioMerieux, Soil Ex tract Agar (At las and Parks 1997), Iron Bac te ria Iso la tion Me dium (At las and Parks, 1997) and Actinomycetes Iso la tion Agar (At las and Parks, 1997). Mor phol ogy, Gram stain and bio chem i cal pro pri eties of the bac te ria were an a lyzed to iden tify the mi croor gan -isms. Spe cies iden ti fi ca tion was based on Bergey’s Man ual of De ter mi na tive Bac te ri ol ogy and Bergey’s Man ual of Sys -tem atic Bac te ri ol ogy (Holt, 1989, 1994). Since there are no stan dard, bio chem i cal tests for the ma jor ity of iso lated gen era, the bio chem i cal tests were in di vid u ally se lected, ac -cord ing to di ag nos tic man u als.

The pisoid in ter nal struc tures were stud ied un der a scan ning elec tron mi cro scope (SEM) JEOL 5410, cou pled with a microprobe (EDS) Voy ager 3100 (Noran prod uct). To pre vent the col lapse of the or ganic struc ture, some sam -ples were treated, us ing pro ce dures for bi o log i cal sam -ples, that is, im me di ately plunge-frozen in isopentane, cooled by liq uid ni tro gen and then lyophilized. Or ganic mat ter from other sam ples was re moved, us ing H2O2 prior to SEM ex -am i na tion. Stan dard thin sec tions were also made from the

pisoids. Their min er al ogy was de ter mined, us ing the XRD method and IR spec tros copy.

RE SULTS AND IN TER PRE TA TION

The wa ter is mainly of the Ca-HCO3 type (Ta ble 1). All wa ter sam ples were su per sat u rated, with re spect to cal cite, and many were also su per sat u rated, with re spect to ar agon ite. The pisoids are mostly flat tened spheres, up to 2 cm across. They are rel a tively soft and lack nu clei. Low-Mg, microcrystalline cal cite is their only autochthonous car bon ate phase. They have rough sur faces and mammilated lam i -na tion, with microstromatolitic struc tures (Fig. 4A). The lamination is vis i ble, ow ing to con cen tra tions of noncar bon -ate par ti cles, in cor po r-ated into the pisoid cor ti ces, which was con firmed by EDS (Fig. 5; see also Jones, 2009; Gra-dziñski et al., 2010).

The microbiological anal y ses re vealed var i ous strains of bac te ria that in habit the pisoids. Bac te ria, be long ing to a physiologically de fined hy dro gen-ox i diz ing (knallgas) group (Aragno and Schlegel, 1991), were iden ti fied in each sam ple stud ied (Ta ble 2). Species of Xanthobacter were the most com mon. Dinitrogenfix ing bac te ria, be long ing to the gen -era Arthrobacter, oc curred in each sam ple. No fungi were de tected.

Fig. 4. Internal struc ture of cave pisoids. A – Irregular lam i na tion of pisoid. B – EPS build ing al ve o larseptal frame work on pisod sur -face. C – Sur face of Xanthobacter col ony, grow ing in lab o ra tory. D – Bac te rial fab rics of pisoid. E – Rod-shaped, bacterial cells, partly cov ered by EPS. F – Cal cite rep li cas of bac te rial cells; note cir cu lar cross-sec tions of rep li cas (ar row). G – Nee dle and fil a men tous cal cite crys tals. H – Bac te rial cells with small cal cite par ti cles, the first step of rep lica for ma tion. I – Cal cite crust, grow ing on bac te rial surface. J – Over growth of small crys tals with cal cite, lead ing to for ma tion of larg est crys tal. A – thin sec tion, B–J un der SEM. Sam ples in B–E and G–J were plunge-frozen in isopentane, cooled by liq uid ni tro gen and then lyophilized; sam ple in F was treated with H2O2 to re move

Un der the SEM, the pisoid sur face re vealed a threedi -men sional, al ve o lar-septal biofilm (Fig. 4B), re sem bling that de scribed by Défarge et al. (1996) from mod ern stromatolites of the Pa cific re gion. The struc ture is built pre -dom i nantly of extracellular, poly meric sub stances (EPS). It closely re sem bles the sur face of a Xanthobacter col ony, which grew in lab o ra tory con di tions (Fig. 4C). Since bac-teria, be long ing to this ge nus, ex crete co pi ous amounts of slime (Wiegel, 1991; Braissant et al., 2003), they prob a bly play a ma jor role in pro duc ing the biofilm, cov er ing the pisoids studied.

The pisoids are built of cal cite crys tals of var i ous shapes and of bod ies, formed by or ganic mat ter, as sug gested by EDS anal y ses. The rod-like bod ies are ~0.5 µm wide and 0.8 to 3 µm long, whereas the glob u lar forms are <1 µm in di -am e ter. They ag glom er ate in clumps, cov ered with EPS (Fig. 4D, E). The di men sions and shape of the or ganic bod -ies de scribed, along with the pres ence of liv ing bac te ria in the sam ples stud ied, sug gest that, de spite their min ute di -men sions, the bod ies in ques tion rep re sent liv ing, bac te rial

cells. They oc cur only in the outer part of the pisoid cor tex, up to a few mil li me ters be low the sur face (Fig. 6). A sim i lar phe nom e non is also typ i cal of the ter res trial oncoids, de -scribed by Jones (2011a).

The largest crys tals, up to a few micrometres across, pre dom i nantly oc cur in the cen tral parts of pisoids, whereas small crys tals are dom i nant in the outer parts, close to the sur face of the pisoids. Al though the for mer are for the most part ir reg u larly shaped, some ex hibit faces and edges. The lat ter, up to 3 µm long, com monly show rounded edges and dis play cir cu lar cross-sec tions. Careful ex am i na tion un der the SEM did not re veal such crys tals, at tached to the sur face of the biofilms that cover the pisoids (Fig. 6A). This in di -cates that these crys tals were not trapped and bound by the sticky biofilm, which cov ers the sur face of the pisoids. Thus, they are an autochthonous com po nent, which orig i -nated within a pisoid. They are re mark ably sim i lar, both in shape and size, to the bac te rial cells, de scribed above, and never ex ceed sig nif i cantly their di men sions (Fig. 4F). This sim i lar ity sug gests that such crys tals are three-di men sional cal cite rep li cas of bac te rial cells. They formed by the crys -tal li za tion of cal cite around the liv ing cell or just af ter the death of the or gan ism (see Jones and Kahle, 1986).

The ob ser va tions un der the SEM re vealed sev eral gen -er a tions of cal cite crys tal for ma tion. Some bac te rial cells, al though still built of or ganic mat ter, are cov ered with min ute, ir reg u lar min eral par ti cles, 0.1 µm across, most prob a bly cal cite, and re flect an early step of cal cite rep lica for ma -tion (Fig. 4H). Later on, the crys tal lites co alesced (Fig. 4I) and, in con se quence, form a con tin u ous crust on the bac te rial surface. The rep li cas and bun dles of fi brous cal cite sub -se quently -served as the sub strate for the fur ther growth of cal cite crys tals. The biofilm macromolecules limited the growth of crys tals to fine, microcrystalline sizes (see Arp et

al., 1999). Dur ing the de com po si tion of the biofilm, fur ther

growth of crys tals is pos si ble.

Apart from the small, anhedral crys tals, sin gle, nee -dle-like crys tals and fil a men tous crys tals also oc cur in the outer parts of the pisoids (Fig. 4G). The lat ter are ~0.2 µm wide. They are curved and closely in ter twined, hence their length was dif fi cult to es ti mate; it prob a bly ex ceeds 10 µm. Sim i lar, fil a men tous crys tals are known from var i ous, con ti -nen tal car bon ates (see Jones and Kahle, 1993; Verrecchia and Verrecchia, 1994 for re view) and are re garded as

bio-364

Ta ble 1

Chem is try of pool wa ter from Perlová Cave Pool number t (°C) pH Eh (mV) TDS (mg/L) HCO3 mg/L) SO4 (mg/L) Cl (mg/L) NO3 (mg/L) Ca (mg/L) Mg (mg/L) Na (mg/L) K (mg/L) DIC (mmol/L) SI calcite SI aragonite P 1 5.9 8.56 423 332.5 215.1 14.06 2.59 16.14 77.88 5.76 <0.2 0.85 3.314 1.02 0.86 P 4 5.9 8.56 421 342.9 236.9 10.33 3.04 8.46 78.17 5.31 <0.2 0.65 3.608 1.05 0.89 P 5 5.9 8.50 423 342.4 230.2 12.71 2.88 12.40 77.85 5.54 <0.2 0.74 3.592 0.98 0.82 P 6 5.7 8.47 427 333.9 208.6 17.49 2.44 23.88 73.97 6.29 <0.2 0.90 3.274 0.89 0.73 P 8 5.6 8.43 422 354.5 238.0 15.49 2.94 12.25 75.66 9.48 <0.2 1.71 3.733 0.92 0.64 P 9§ 5.4 8.32 418 394.0 245.0 29.20 4.79 24.70 75.16 16.20 <0.2 0.74 3.835 0.95 0.79 P 10§ 5.5 8.39 423 389.6 260.7 16.64 2.55 18.56 74.41 15.95 <0.2 0.75 4.116 0.90 0.75 Note: Un less oth er wise stated, mean data from three sam pling trips; § – Mean data from two sam pling trips

Ta ble 2

Bac te rial as sem blage in pisoids from Perlová Cave

Bacteria pool number

1 4 5 6 9 10 Agromyces sp. + + + – – – Alcaligenes sp. – – – – + – Arthrobacter crystallopoietes – + + + – + Arthrobacter sp. + + + + + + Bacillus alcalophilus + – – – + + Bacillus azotoformans – – – + – + Bacillus badius – – – + – – Bacillus brevis – + – + – – Bacillus megaterium + + + + – – Pseudomonas carboxyhydrogena * + – + + – – Pseudomnas sp. – – + – + – Xanthobacter autotrophicus * – – – + + + Xanthobacter flavus * + + + – + – Xanthobacter sp. * + – – – + –

+ pres ence of given taxon; – ab sence of given taxon; * hy dro gen-ox i diz ing (knallgas) bac te rium

genic (Gradziñski et al., 1997; Loisy et al., 1999; CaÔaveras

et al., 2006; Bindschedler et al., 2010) or purely abiogenic

pre cip i tates (Borsato et al., 2000). Their or i gin is also as cribed to the precipitation of cal cite, due to a so lu tion–pre cur sor–solid mech a nism, in the pres ence of dis solved, or -ganic mat ter in a par ent so lu tion (Olszta et al., 2004; CaÔaveras et al., 2006).

Small, anhedral and fil a men tous crys tals were suc ces -sively over grown with cal cite (Fig. 4J). The pro cess led to com plete oblit er a tion of the pri mary, bac te rial fab rics of the pisoids (Fig. 6), as pre vi ously de scribed from tufa stromato- lites by Szulc and Smyk (1994) and from travertines by Guo and Rid ing (1994).

DIS CUS SION

The in ter nal struc tures of the pisoids stud ied show that they dif fered mark edly from most speleothems, dis play ing dis tinct, crys tal line fab rics, even those grow ing be neath the wa ter level (González et al., 1992; Frisia et al., 2000), in -clud ing typ i cal cave pisoids (Nader, 2007; Melim and Spilde, 2011). The dif fer ence arises, in spite of the fact that the pisoids grew in very sim i lar con di tions to other speleo-thems and are sup plied with wa ter of sim i lar chem is try. It im plies that the pisoid growth is gov erned by dif fer ent fac -tors than that of crys tal line speleothems. The pisoids stud ied bear a strong, struc tural re sem blance to mi cro bial car bon

-ates, which along with the oc cur rence of liv ing bac te ria within the pisoids, in di cates that their growth can be pro -moted microbially. It seems rel e vant to dis cuss how bac te ria can in flu ence the pro cess of cal ci fi ca tion and thus the for -ma tion of the pisoids.

The cal ci fi ca tion takes place around the bac te rial cells, so that the pro cess is of ex ter nal type (Rid ing, 1991), which may be driven solely by en vi ron men tal con di tions or by mi -cro bial phys i ol ogy. Rapid de gas sing of CO2 can be ex -cluded as the main fac tor, driv ing cal cite pre cip i ta tion, since the pisoids grow in sta ble con di tions, in a calmwa ter set -ting, com pletely bathed in stag nant pool wa ter. It sug gests that an other fac tor, such as bac te rial phys i ol ogy, may stim u -late cal cite pre cip i ta tion.

The se quence of crys tal growth, de scribed above, from a sin gle, min er al ized bac te rial cell to a more reg u lar, de vel oped crys tal, shows that cal ci fi ca tion de veloped pro gres sively from the min er al ized, bac te rial cells. Hence, it is sim -i lar -in style to cyanobacter-ial cal c-i f-i ca t-ion -in a low DIC-high Ca2+ hard-wa ter set ting (see Ta ble 1), where the photosynthetic ac tiv ity causes car bon re moval and cre ates a lo -cal shift in supersaturation (Arp et al., 2001, 2010; Shiraishi

et al., 2008). Bear ing in mind a spe cific cave en vi ron ment,

such an ac tiv ity should be ruled out. Thus, the hy poth e sis can be for mu lated that a cru cial role is played by chemoautolithotrophic, hy dro genox i diz ing bac te ria in the for ma -tion of cal cite. They ac tively take up CO2 from their sur -round ings, be cause it is their ma jor source of car bon (Ara-Fig. 5. Laminae, com posed of de tri tal grains within pisoid. A – pol ished thin sec tion un der SEM, chem i cal com po si tion of some grains is in di cated. B, C – EDS spec tra of alumnosilicate, de tri tal grains

gno and Schlegel,1991). Thus, they cause de ple tion of dis -solved CO2 in their pericellular re gion, which leads to rapid con ver sion of HCO3 to CO2. This pro cess re sults in al ka li za tion of the microenvironment, and thus re sults in the cre -ation of CO32- ions, which fi nally causes cal cite crys tal li za -tion (Buhmann and Dreybrodt, 1985).

Ac cord ing to the above hy poth e sis, the rate of the pro cess is con trolled by the ac tiv ity of hy dro genox i diz ing bac -te ria. They can live in a spelean en vi ron ment, where the sup ply of or ganic car bon is strongly lim ited, ow ing to a che- molithoautotrophic mode of life, de pend ing on in or ganic

sources of en ergy. They grow on CO2, gas eous ox y gen, and gas eous hy dro gen. Con sid er ing the ac ces si bil ity of the two first com po nents, the sup ply of gas eous hy dro gen seems to be of cru cial im por tance, as it oc curs in min ute amounts in most nat u ral en vi ron ments, in clud ing caves. In the case stu-died, it is most prob a bly a by-prod uct of co-oc cur ring, dini-trogen-fix ing bac te ria, be long ing to the ge nus Arthrobacter (see Smyk and Ettlinger, 1963; Jones and Keddie, 1991).

Biofilms in flu ence the sup ply of re ac tants, since they have dif fu sion-slow ing prop er ties (Decho, 2000). In the pisoids stud ied, the sticky biofilm slows down the transpor-tation of Ca2+, which in turn slows the pre cip i ta tion re ac -tion. This leads to the pre cip i ta tion of cal cite, and in hib its the for ma tion of ar agon ite, even though the macroenviron-ment is su per sat u rated with re spect to both minerals. A similar phe nom e non was ex per i men tally demonstrated by Buczynski and Chafetz (1991), where the higher vis cos ity of the me dium fa vored bac te ri ally in duced cal cite pre cip i ta -tion over that of ar agon ite.

The pro cess of cal ci fi ca tion, in duced by chemolithoautotrophic bac te ria, pos tu lated above and so far un rec og -nized, cor re sponds to the ‘dark CO2 fix a tion’, pro posed by Krumbein (1979) and Simkiss (1986). The hy dro genox i diz ing bac te ria are fre quent in a great va ri ety of nat u ral hab -i tats: so-ils, mod ern lake sed -i ments, hot-spr-ings and even sea wa ter (Aragno and Schlegel, 1991; Bae et al., 2001; Aguiar

et al., 2004). Authigenic car bon ate min er als of microcrysta-

lline type are formed in all of these en vi ron ments. Hence, the in flu ence of hy dro genox i diz ing bac te ria may also ex -plain the or i gin of other, not only spelean, microcrystalline car bon ates. Gradziñski (2003) also sug gested that this type of cal ci fi ca tion in flu ences the ox y gen sta ble iso to pic sig na -ture of cal cite.

None the less, the above hy poth e sis is to some ex tent spec u la tive. Firstly, it is based only on the clas sic de ter mi -na tion of mi crobes. Ac tu ally, it is known that only a small per cent age of mi crobes in sam ples from the cav ern en vi ron -ment can be cul ti vated and de ter mined (Northup and Lavoie, 2001). There fore, in the sam ples stud ied, other mi crobes also may have been pres ent and they could have in -flu enced cal cium car bon ate pre cip i ta tion, as well.

Sec ondly, the hy dro gen-ox i diz ing bac te ria, de ter mined in the pisoids, are only fac ul ta tive autotrophs that also can grow on or ganic me dia (Aragno and Schlegel, 1991). The pos si bil ity can not be ex cluded that in a way of life, other than chemolithoautotrophic, they might in duce the pre cip i ta tion of cal cium car bon ates. For in stance, bac te ria, be long -ing to the ge nus Xanthobacter, which are com mon in the pisoids stud ied, can uti lize cal cium ox a late and pro duce cal -cium car bon ate. Such a phe nom e non was rec og nized in a soil ex tract from Ivory Coast (Braissant et al., 2004). Such bac te ria are also known for their ca pa bil ity to stim u late the pre cip i ta tion of vaterite (Braissant et al., 2003).

Thirdly, there ex ists a great body of lit er a ture on the pre cip i ta tion of min er als within biofilms and mi cro bial mats (see Dupraz et al., 2009 for re view). Sev eral mech a nisms of car bon ate min eral pre cip i ta tion are known to oc cur with out the in ter ac tion of liv ing or gan isms (organomineralization

sensu Trichet and Défarge, 1995; bi o log i cally-in flu enced

min er al iza tion sensu Dupraz et al., 2009). It can not be ruled

366

Fig. 6. Con trast ing fab rics of dif fer ent parts of pisoid. A – out -er most part of pisoid, com posed of ir reg u lar clumps of glob u lar and rod-like bod ies, cov ered with EPS and nee dle-fi bre cal cite. B – well de vel oped nee dlefi bre cal cite and microcrystalline cal -cite ag gre gates in outer part of pisoid (around 3 mm be neath the sur face). C – ag gre gates of spiky cal cite crys tals aligned along their long axes and microcrystalline cal cite ag gre gates, cen tral part of pisoid. Sam ples were plunge-frozen in isopentane, cooled by liq uid ni tro gen and then lyophilized

out that one of these mech a nisms op er ates within the bio-film, cov er ing the pisoids in Perlová Cave. In such a case, it could have con trib uted to the crystallization of cal cium car -bon ate and, hence, the for ma tion of the pisoids stud ied.

Bear ing in mind these res er va tions, the pos tu lated in flu -ence of hy dro gen-ox i diz ing bac te ria on the pre cip i ta tion of cal cium car bon ate should be tested, us ing other meth ods. The pre cip i ta tion of cal cium car bon ate in cul tures of such bac te ria, con ducted in mon i tored lab o ra tory con di tions, could test the hy poth e sis, for mu lated in this pa per.

CON CLU SIONS

Chemolithoautotrophic, hy dro gen-ox i diz ing bac te ria can cause bi o log i cally in duced cal ci fi ca tion. The es sence of the pro cess is a shift in cal cite supersaturation, due to bio-genic CO2 con sump tion. Bac te rial biofilms, be cause of their dif fu sionslow ing prop er ties, in hibit the pre cip i ta tion of ar -agon ite and thus pro mote the pre cip i ta tion of cal cite. The pres ence of a biofilm limits the size of the grow ing cal cite crys tals. Thus, the bac te ria stim u late the for ma tion of mi crocrys tal line cave pisoids and in flu ence their in ter nal fab -rics. How ever, it must be em pha sized that this view is based ex clu sively on the clas sic de ter mi na tion of mi cro or gan isms. It should be sup ported ad di tion ally by mod ern, mo lec u lar meth ods of in ves ti ga tion.

Ac knowl edg ments

The pa per is an out growth of M. Gradziñski’s PhD the sis, su -per vised by the late Pro fes sor Andrzej Radomski. Pe ter Holúbek, Jaga Faber, Mariusz Czop, Renata Jach and Grzegorz Haczewski are thanked for their help. The study was fi nanced by KBN grant 6P04D01914. M. G. was sup ported by the Foun da tion for Pol ish Sci ence (J. KaŸmierczak Grant for Re search ers). An early ver sion of the manu script bene fited from the con struc tive com ments of Brian Jones. The au thors are in debted to re view ers Leslie A. Melim and Tadeusz Peryt, as well as to ed i tors Frank Simpson and Al fred Uchman for their help in im prov ing the manu script.

REF ER ENCES

Aguiar, P., Bever idge, T. J. & Reysenbach, A.-L., 2004. Sulfuri-hydrogenibium azorense, sp. nov., a thermophilic hy dro gen-ox i diz ing microaerophile from ter res trial hot springs in the Azores. In ter na tional Jour nal of Sys tem atic and Evo lu tion ary Mi cro bi ol ogy, 54: 33–39.

Aragno, M. & Schlegel, H. G., 1991. The mesophilic hy dro -gen-oxidizing (knallgas) bac te ria, In: Balows, A., Trüper, H. G., Dworkin, M., Harder, W. & Schleifer, K.-H. (eds), The Prokaryotes. A Handbook on the Biology of Bacteria: Eco-phys i ol ogy, Isolation, Identification, Applications, Vol ume III. Springer, New York, pp. 344–384.

Arp, G., Bisset, A., Brinkmann, N., Cousin, S., de Beer, D., Friedl, T., Mohr, K. I., Neu, T. R., Reimer, A., Shiraishi, F., Stackebrandt, E. & Zippel, B., 2010. Tufaform ing biofilms of Ger -man streams: mi cro or gan isms, exopolymers, hydrochemistry and cal ci fi ca tion. In: Pedley, H. M. & Rogerson M. (eds), Tufas and Speleothems, Un rav el ling the Mi cro bial and Phys i -cal Con trols. Geo log i -cal So ci ety Special Publication, 336:

83–118.

Arp, G., Reimer, A. & Reitner, J., 2001. Pho to syn the sis-in duced biofilm cal ci fi ca tion and cal cium con cen tra tions in Phane-rozoic oceans. Sci ence, 292: 1701–1704.

Arp, G., Thiel, V., Reimer, A., Mi chae lis, W. & Reitner, J., 1999. Biofilm exopolymers con trol microbialite for ma tion at ther mal springs dis charg ing into the al ka line Pyr a mid Lake, Ne -vada, USA. Sed i men tary Ge ol ogy, 126: 159–176.

At las, R. M. & Parks, L. C., 1997. Hand book of Mi cro bi o log i cal Me dia. CRC Press, Boca Raton, 1706 pp.

Bae, S., Kwak, K., Kim, S., Chung, S. & Igarashi, Y., 2001. Iso la -tion and char ac ter iza -tion of CO2-fix ing hy dro gen-ox i diz ing

ma rine bac te ria. Jour nal of Bio sci ence and Bio en gi neer ing, 91: 442–448.

Barton, H. A. & Northup, D. E., 2007. Geomicrobiology in cave en vi ron ments: past, cur rent and fu ture per spec tives. Jour nal of Cave and Karst Stud ies, 69: 163–178.

Baskar, S., Baskar, R. & Routh, J., 2011. Biogenic ev i dences of moonmilk de po si tion in the Mawmluh Cave, Meghalaya, In -dia. Geomicrobiology Jour nal, 28: 252–265.

Bindschedler, S., MilliÀre, L., Cailleau, G., Job, D. & Verrecchia, E. P., 2010. Calcitic nannofibres in soils and caves: a pu ta tive fun gal con tri bu tion to carbonatogenesis. In: Pedley, H. M. & Rogerson M. (eds), Tufas and Speleothems. Un rav el ling the Mi cro bial and Phys i cal Con trols., Geo log i cal So ci ety Special Publication, 336: 225–238.

Blyth, A. & Frisia, S., 2008. Mo lec u lar ev i dence for bac te rial me -di a tion of cal cite for ma tion in cold high-al ti tude caves. Geo-mi cro bi ol ogy Jour nal, 25: 101–111.

Borsato, A., Frisia, S., Jones, B. & van der Borg, K., 2000. Cal cite moonmilk: crys tal mor phol ogy and en vi ron ment of for ma tion in caves in the Ital ian Alps. Jour nal of Sed i men tary Re search, 70: 1179–1190.

Boquet, E., Boronat, A. & RamosCormenzana, A., 1973. Pro duc -tion of cal cite (cal cium car bon ate) crys tals by soil bac te ria is a gen eral phe nom e non. Na ture, 246: 527–529.

Braissant, O., Cailleau, G., Dupraz, C. & Verrecchia, E. P., 2003. Bac te ri ally in duced min er al iza tion of cal cium car bon ate in ter res trial en vi ron ments: The role of exopolysaccharides and amino-ac ids. Jour nal of Sed i men tary Re search, 73: 485–490. Braissant, O., Cailleau, G., Dupraz, C. & Verrecchia, E. P., 2004.

Bi o log i cally in duced min er al iza tion in the tree Milicia ex celsa (Moraceae): its causes and con se quences to the en vi ron -ment. Geobiology, 2: 59–66.

Buczynski, C. & Chafetz, H. S., 1991. Habit of bac te ri ally in duced pre cip i tates of cal cium car bon ate and the in flu ence of me dium vis cos ity on min er al ogy. Jour nal of Sed i men tary Pe -trol ogy, 61: 226–233.

Buhmann, D. & Dreybrodt, W. 1985. The ki net ics of cal cite dis so -lu tion and pre cip i ta tion in geo log i cally rel e vant sit u a tions of karst ar eas. 1. Open sys tem. Chem i cal Ge ol ogy, 48: 189–211. CaÔaveras, J. C., Cuezva, S., Sanchez-Moral, S., Lario, J., Laiz, L., Gon za lez, J. M. & Saiz-Jimenez, C., 2006. On the or i gin of fi ber cal cite crys tals in moonmilk de pos its. Naturwissen-schaften, 93: 27–32.

Castanier, S., Le Métayer-Levrel, G. & Perthuisot, J.-P., 2000. Bac te rial roles in the pre cip i ta tion of car bon ate min er als. In: Rid ing, R. E. & Awramik, S. M. (eds), Mi cro bial Sed i ments. Springer, Berlin, pp. 32–39.

Decho, A. D., 2000. Exopolymer microdomains as a struc tur ing agent for het er o ge ne ity within mi cro bial mates. In: Rid ing, R. E. & Awramik, S. M. (eds), Mi cro bial Sed i ments. Springer, Berlin, pp. 9–15.

Défarge, Ch., Trichet, J., Jaunet, A.-M., Rob ert, M., Tribble, J. & Sansone, F. J., 1996. Tex ture of mi cro bial sed i ments re vealed

by cryoscan ning elec tron mi cros copy. Jour nal of Sed i men -tary Pe trol ogy, 66: 935–948.

Dupraz, Ch., Reid, R. P., Braissant, O., Decho, A. W., Nor man, A. S. & Visscher, P. T., 2009. Pro cess of car bon ate pre cip i ta tion in mod ern mi cro bial mat. Earth-Sci ence Re views, 96: 141– 162.

Ehrlich, H. L., 1996. Geomicrobiology. Mar cel Dekker, New York, 719 pp.

Ehrlich, H. L., 1999. Mi crobes as geo log i cal agents: their role in min eral for ma tion. Geomicrobiology Jour nal, 16: 135–153. Frisia, S., Borsato, A., Fairchild, I. J. & McDermott, F., 2000. Cal

cite fab rics, growth mech a nisms, and environments of for ma -tion in speleothems from the Ital ian Alps and southwestern Ire land. Journal of Sed i mentary Research, 70: 1183–1196. González, L. A., Car pen ter, S. J. & Lohmann, K. C., 1992. In or

-ganic cal cite mor phol ogy: roles of fluid chem is try and fluid flow. Jour nal of Sed i men tary Pe trol ogy, 62: 382–399. Gradziñski, M., 2001. Role of bac te ria in the growth of cave

pearls. In: Pro ceed ings of the 13th In ter na tional Con gress of Spe le ol ogy, Bra si lia. Un ion In ter na tional de Spéléologie, Brazil. [4 pages in CD].

Gradziñski, M., 2003. Bac te rial in flu ence on speleothem ox y gen iso tope com po si tion: An ex am ple based on cave pisoids from Perlová Cave (Slovakia). Geologica Carpathica, 54: 199– 204.

Gradziñski, M., Chmiel, M. J., Lewandowska, A. & Michalska-Kasperkiewicz, B., 2010. Siliciclastic microstromatolites in a sand stone cave: Role of trap ping and bind ing of de tri tal par ti -cles in for ma tion of cave de pos its. Annales Societatis Geo-logorum Poloniae, 80: 303–314.

Gradziñski, M., Szulc, J. & Smyk, B., 1997. Mi cro bial agents of moonmilk cal ci fi ca tion. In: Jeannin, P.-Y. (ed.), Pro ceed ings of the 12th In ter na tional Con gress of Spe le ol ogy, Vol ume 1. In ter na tional Un ion of Spe le ol ogy, Basel, pp. 275–278. Gradziñski, R. & Radomski, A., 1967. Pisoliths from Cu ban caves.

Rocznik Polskiego Towarzystwa Geologicznego, 37: 243– 265.

Guo, L. & Rid ing, R., 1994. Or i gin and diagenesis of Qua ter nary shrub fa cies, Rapolane Terme, cen tral It aly. Sedimentology, 41: 499–520.

Hill, C. & Forti, P., 1997. Cave Min er als of the World. Na tional Speleological So ci ety, Huntsville, pp. 1–463.

Holúbek, P. & Kleskeò, J., 1993. Objavy v Perlovej jaskyni. Spra-vodaj Slovenskej Speleologickej Spoloènosti, 23: 20–22. [In Slo vak].

Holt, J. G., ed., 1989. Bergey’s Man ual of Sys tem atic Bac te ri ol -ogy. Vol ume 1–4. Wil liams & Wilkins, Bal ti more, 2648 pp. Holt, J. G., ed., 1994. Bergey’s Man ual of De ter mi na tive Bac te ri

-ol ogy. Wil liams & Wilkins, Bal ti more, 787 pp.

Jones, B., 2001. Mi cro bial ac tiv ity in caves A geo log i cal per spec -tive. Geomicrobiology Jour nal, 18: 345–357.

Jones, B., 2009. Cave pearls – the in te grated prod uct of abiogenic and biogenic pro cesses. Jour nal of Sed i men tary Re search, 79: 689–710.

Jones, B., 2010. Mi crobes in caves: agents of cal cite cor ro sion and pre cip i ta tion. In: Pedley, H. M. & Rogerson M. (eds), Tufas and Speleothems. Un rav el ling the Mi cro bial and Phys i cal Con trols., Geo log i cal So ci ety Special Publication, 336: 7–30. Jones, B., 2011a. Biogenicity of ter res trial oncoids formed in soil pock ets, Cayman Brac, Brit ish West In dies. Sed i men tary Ge -ol ogy, 236: 95–108.

Jones, B., 2011b. Sta lac tite growth me di ated by biofilms: Ex am ple from Nani Cave, Cayman Brac, Brit ish West In dies. Jour nal of Sed i men tary Re search, 81: 322–338.

Jones, B. & Kahle, C. F., 1986. Den dritic cal cite crys tals formed

by cal ci fi ca tion of al gal fil a ments in a vadose en vi ron ments. Jour nal of Sed i men tary Pe trol ogy, 56: 217–227.

Jones, B. & Kahle, C. F., 1993. Mor phol ogy, re la tion ship, and or i gin of fi ber and den drite cal cite crys tals. Jour nal of Sed i men -tary Pe trol ogy, 63: 1018–1031.

Jones, B. & Mac Don ald, R. W., 1989. Mi croor gan isms and crys -tal fab rics in cave pisoliths from Grand Cayman, Brit ish West In dies. Jour nal of Sed i men tary Pe trol ogy, 59: 387–396. Jones, D. & Keddie, R. M., 1991. The ge nus Arthrobacter. In:

Balows, A., Trüper, H .G., Dworkin, M., Harder, W., & Schleifer, K.-H. (eds), The Prokaryotes. A Handbook on the Bi ol ogy of Bac te ria: Ecophysiology, Iso la tion, Iden ti fi ca tion, Applications, Vol ume II. Springer, New York, pp. 1283–1299 Krumbein, W. E., 1979. Cal ci fi ca tion by bac te ria and al gae. In: Trudinger, P. A. & Swaine, D.J. (eds) Biogeochimical Cy -cling of Min eral-Form ing El e ments. Elsevier, Am ster dam, pp. 47–68.

Loisy, C., Verrecchia, E. P. & Dufour, P., 1999. Mi cro bial or i gin for pedogenic micrite as so ci ated with a car bon ate paleosoil (Cham pagne, France). Sed i men tary Ge ol ogy, 126: 193–204. Mahe¾, M., 1968. The Me so zoic. In: Mahe¾, M. & Buday, T.

(eds), Re gional Ge ol ogy of Czecho slo va kia, Part II, The West Carpathians. Geo log i cal Sur vey of Czecho slo va kia, Ac a de -mia, Praha, pp. 138–143.

Melim, L. A., Boston, P. J., Northup, D. E., Spilde, M. N. & Queen, J. M., 2001. Ev i dence for mi cro bial in volve ment in pool fin ger pre cip i ta tion, Hid den Cave, New Mex ico. Geo-mi cro bi ol ogy Jour nal, 18: 311–329.

Melim, L. A. & Spilde, M. N., 2011. Rapid growth and recrys-tallization of cave pearls in an un der ground mine. Jour nal of Sed i men tary Re search, 81: 775–786.

Mon ger, H. C., Daugherty, L. A., Lindemann, W. C. & Lid dell, C. M., 1991. Mi cro bial pre cip i ta tion of pedogenic cal cite. Ge ol -ogy, 19: 997–1000.

Nader, F. H., 2007. Petrographic and geo chem i cal study on cave pearls from Kanaan Cave (Leb a non). In ter na tional Jour nal of Spe le ol ogy, 36: 39–50.

Northup, D. & Lavoie, K. H., 2001. Geomicrobiology of caves: A re view. Geomicrobiology Jour nal, 18: 199–222.

Olszta, M. J., Gajjeraman, S., Kaufman, M. & Gower, L. B., 2004. Nanofibrous cal cite syn the sized via a so lu tionpre cur sorsolid mech a nism. Chem is try of Ma te ri als, 16: 23552362.

Park hurst, D. L. & Appelo, C. A. J., 1999. User’s guide to PHREEQC (ver sion 2) – A com puter pro gram for speciation, batch re ac tion, onedi men sional trans port, and in verse geo chem i cal cal cu la tion. U.S. Geo log i cal Sur vey Wa ter Re -source In ves ti ga tions Re port, 99-4259: 1–326.

Pedley, M., 2000. Am bi ent tem per a ture fresh wa ter mi cro bial tufas. In: Rid ing, R. E. & Awramik, S. M. (eds), Mi cro bial Sed i ments. Springer, Berlin, pp. 179–186.

Renaut, R. W. & Jones, B., 2000. Mi cro bial pre cip i tates around con ti nen tal hot springs and gey sers. In: Rid ing, R. E. & Awramik, S. M. (eds), Mi cro bial Sed i ments. Springer, Berlin, pp. 187–195.

Rid ing, R., 1991. Clas si fi ca tion of mi cro bial car bon ates. In: Rid -ing, R. (ed.), Cal car e ous Al gae and Stromatolites. Springer, Berlin, pp. 21–51.

Rid ing, R., 2000. Mi cro bial car bon ates: the geo log i cal re cord of cal ci fied bac te rial–al gal mats and biofilms. Sedimentology, 47: 179–214.

Shiraishi, F., Reimer, A., Bissett, A., de Beer, D. & Arp, G., 2008. Mi cro bial ef fects on biofilm cal ci fi ca tion, am bi ent wa ter che- mistry and sta ble iso tope re cords in a highly su per sat u rated set ting (Westerhöfer Bach, Ger many). Palaeo ge ogra phy, Palaeoclimatology, Palaeo ec ol ogy, 262: 91–106.

Simkiss, K., 1986. The pro cesses of biomineralization in lower plants and an i mals an over view. In: Lead beat er, B. S. C. & Rid ing, R. (eds), Biomineralization in Lower Plants and An i -mals. The Sys tem at ics As so ci a tion, Spe cial Vol ume, 30: 1937. Smyk, B. N. & Ettinger, L., 1963. Recherches sur quelque espÀces d’arthrobacter fixatrices d’azote isoles des roches karstiques alpines. Annales de l’Institut Pas teur, 105: 341–348. Szulc, J. & Smyk, B., 1994. Bac te ri ally con trolled cal ci fi ca tion of

fresh wa ter Schizotrix-stromatolites: an ex am ple from the Pie-niny Mts., South ern Po land. In: Bertrand-Sarfati, J. & Monty, C. (eds), Phanerozoic Stromatolites II. Kluwer, Dordrecht, pp. 31–51.

Trichet, J. & Défarge, C., 1995. Non-bi o log i cally sup ported

organomineralization. Bul le tin de l’Institute Océanographi-que, Numéro Spécial, 14: 203–236.

Verrecchia, E. P. & Verrecchia, K. E., 1994. Nee dle-fi ber cal cite: A crit i cal re view and a pro posed clas si fi ca tion. Jour nal of Sed i men tary Re search, A64: 650–664.

Wiegel, J., 1991, The ge nus Xanthobacter. In: Balows, A., Trüper, H. G., Dworkin, M., Harder, W. & Schleifer, K.-H. (eds), The Prokaryotes. A Handbook on the Bi ol ogy of Bac te ria: Eco-phys i ol ogy, Iso la tion, Iden ti fi ca tion, Applications, Vol ume III. Springer, New York, pp. 2366–2383.

Wright, D. T. & Oren, A., 2005. Nonphotosynthetic bac te ria and the for ma tion of car bon ates and evaporites through time. Geomi cro bi ol ogy Jour nal, 22: 27–53.