Epipelic diatoms from an extreme acid environment - Earth and ...

Epipelic diatoms from an extreme acid environment: BeowulfSpring, Yellowstone National Park, U.S.A.William O. Hobbs* 1 , Alexander P. Wolfe 1 , William P. Inskeep 2 , Larry Amskold 1 & KurtO. Konhauser 11Department of Earth & Atmospheric SciencesUniversity of AlbertaEdmonton, AB T6G 2E3Canada2 Department of Land Resources & Environmental SciencesMontana State UniversityBozeman, MT 59717USA*author for correspondence (whobbs@ualberta.ca)1

161718192021222324IntroductionAlthough diatoms have been shown to respond to temperature in culture (Patrick, 1977;Suzuki & Takahashi, 1995), it remains uncertain whether they are directly thermophilousin an eco-physiological sense, or indirectly influenced by environmental conditions thatcovary with climate (Anderson, 2001; Wolfe, 2002). The bulk of previous studies ondiatoms from hot spring environments has focused solely on temperature (Hustedt, 1937-1938; Stockner, 1967; Brock & Brock, 1966), with some attention to the influences oflow ambient pH and elevated concentrations of potentially toxic elements, such as arsenic(As) and copper (Cu) (Hargreaves et al., 1975; Whitton & Diaz, 1981).25262728293031323334353637In this paper, we present an account of the diatom flora from the Beowulf Spring in theNorris Geyser Basin, Yellowstone National Park (YNP). This epipelic communityinhabits As-rich hydrous ferric oxide (HFO) mats that coat the sandy substratum of thespring proximal to the vent (Figure 1). Peripheral to the HFO zone, a green algal matcomposed of cyanidiaceous Rhodophyta contains few, if any, diatoms. Light & scanningelectron microscopic (SEM) investigations of these communities are presented from bothuncleaned and oxidized materials. Water chemistry implies that the diatoms areexploiting an ecological niche that may be toxic to other algae. Contrary to Villeneuve &Pienitz’s (1998) suggestion that there is no single diatom assemblage typical of thermalsprings, our study suggests a strong floristic similarity among highly acidic hot springsglobally (Hustedt, 1937-1938; Carter, 1972; Hargreaves et al., 1975; Whitton & Diaz,1981; Cassie & Cooper, 1989; Watanabe & Asai, 1995; DeNicola, 2000; Jordan, 2001),3

3839implying that indeed there exists a distinct but biogeographically-disjunct flora in theseextreme environments.4041424344454647484950515253MethodsBeowulf Hot Spring (44°43’53” N, 110°42’41” W) lies within the Norris Geyser Basin,Yellowstone National Park, Wyoming, U.S.A. (YNP Thermal Inventory #NHSP35))(Fig. 1A). Surface sediment collections were obtained from both the central portion of thespring, which is coated by HFO, and peripheral green algal mats (Fig. 1B). Samples wereeither cleaned using 30% H 2 O 2 , or mounted uncleaned following dilution of the slurry indeionized water. Valves were mounted in Naphrax and observed at 1000x under oilimmersion with differential interference contrast optics (Leica DMLB). For scanningelectron microscopy (SEM), slurry was evaporated onto stubs at room temperature in adesiccator, prior to sputter-coating with Au, and examination in a JEOL-6301F fieldemissionSEM. Archives of these samples are kept at 4 o C at the University of Alberta.Our results confirms that it is clearly advantageous to examine both cleaned anduncleaned material (Stockner 1967).54555657585960Beowulf is considered an acid-sulfate-chloride spring, where the spring dischargesextremely acidic (pH50˚C) water, with mM concentrations of Na + , K + , Cl – ,SO 2- 4 and Si (Figure 1; Table 1; Inskeep et al. 2004). Water samples were obtained bysyringe and filtered through 0.45 µm in-line nylon syringe filters. Samples were furtheracidified with TraceMetal-grade HNO 3 and kept refrigerated until analysis by ionchromatography. Temperature and pH were measured in the field using a digital4

6162thermometer and an Orion Ross (8165BN) pH meter calibrated between 2 and 4,respectively.6364656667686970717273747576Results and DiscussionWater chemistryThe Beowulf Spring has an ionic strength of 0.02 M, which is comparable to a marine –brackish estuarine salinity classification (Juggins, 1992). Aqueous chemistry fromBeowulf Spring has been characterized on multiple occasions and exhibits very littlerange in concentrations (±10% or less for most constituents; Inskeep et al., 2004). Thewater chemistry reveals the presence of potentially toxic metals to algal growth, namelyCu and As (Table 1). In particular, As concentrations reach 20 µM, much higher thanminimum concentrations required to inhibit growth of diatoms in culture (Saunders &Riedel, 1998). Similarly, Beowulf Cu concentrations are two to three orders of magnitudegreater than those used by Saunders & Riedel (1998). Thus, natural concentrations of Asand Cu in the Beowulf Spring approach or exceed those from the most highly impactedacid-mine drainages (Whitton & Diaz, 1981; Verb & Vis, 2000).77787980818283While it is not our intention to speculate as to whether there is active uptake of As bydiatoms in Beowulf Spring, As uptake is likely occurring among other microorganisms inYNP hot springs, namely bacteria (Langner et al., 2001; Inskeep et al., 2004).Metabolism of As by marine algae is hypothesized to occur because of the molecularsimilarity between phosphorous and As (Edmonds & Francesconi, 1998). Generalpathways of metals metabolism by algae include sequestration, complexation with5

848586exuded dissolved organic carbon, and redox transformations during uptake (Saunders &Riedel, 1998). Thus, a range of biogeochemical processes exist to allow not only thesurvival, but indeed the success, of algae inhabiting these extreme environments.87888990919293The pH of Beowulf Spring at the time of sampling was 2.26, and the temperature was67 o C. A number of diatom communities have been documented at pH of

107108(Asada & Tazaki, 2001), is identical to our observations from the Beowulf material (Fig.1C-D).109110111112113114115116117118119120121122123124125Brown matsThe HFO mat contained abundant diatoms that occur as solitary epipelic forms. Allobserved specimens are biraphid taxa, hence capable of motile displacement within themat. Several specimens preserved intact chloroplasts when viewed in LM. SEM revealeddiatom surfaces both completely free of secondary silica precipitates and with cleanareolae (Fig. 1E & F), as well as thoroughly Si-encrusted specimens (Fig. 1G & H). Inthe latter instances, inorganic Si precipitation has formed a continuous botryoidal coating,comprised of coalescent individual papillae of approximately 100 nm diameter. Wehypothesize that living diatoms have cleaner valve surfaces than dead ones, and perhapsthat inorganic silica precipitation may at times overwhelm living cells. Inskeep et al.(2004) have provided a complete characterization of the brown mat, revealing a variety ofSiO 2 crystalline polymorphs and microbial communities encrusted with As-rich HFO. Itis unclear to us at this time what aspect of the HFO mat contributes to the success of thediatoms living there, in sharp contrast to their absence in the green mat. Given thatCyanidiaceae are virtually absent in the HFO mats, it is also possible that resourcecompetition plays a role in delimiting these algae.126127128129Floristic descriptionThe epipelic diatom community in Beowulf Spring consisted of five species (Table 1).The dominant diatom taxon was Nitzschia cf. thermalis var. minor Hilse (56%), followed7

130131132133134135in decreasing order of relative abundance by Pinnularia acoricola Hustedt (21%),Eunotia exigua (de Brébisson ex Kützing) Rabenhorst (15%), Nitzschia ovalis Arnott(7%) and Encyonema minutum (Hilse ex Rabenhorst) D.G. Mann (

153154155156157158159160161162163164165apiculate ends, transapical striae composed of fine areolae, and a constricted centralmargin (Figs. 3A-E). REM of N. homburgiensis and the isotype material of N. thermalisvar. minor Hilse reveal that the fibulae are irregular, widely spaced, and can extend to thecenter of the transapical axis (Plate 39 of Lange-Bertalot, 1978). Comparatively, thefibulae from Beowulf specimens range in thickness from 0.5 to 1.0 µm along the apicalaxis (Fig. 2C) and are confined to the margin of the valve. In addition the raphe of ourspecimens is more eccentric than that of N. homburgiensis (Fig. 3B). Cassie (1989)illustrated N. umbonata from a shallow thermal pool (Tokannu, New Zealand) withsimilar fibulae and raphe structures to Beowulf specimens, but associated withsignificantly larger valves (36-45 µm). Morphometric comparisons of N. cf. thermalisvar. minor, N. hombergiensis, N. umbonata, N. steynii Cholnoky emend. (Schoeman &Archibald, 1988), and N. rimosa Manguin (Bourelly & Manguin, 1952) indicate that eachof these taxa has a distinctive morphology (Table 2).166167168169170171172173174175Further questions arise in considering the ecologies of these diatoms. Lange-Bertalot(1978) associates N. homburgiensis with circumneutral, unpolluted freshwaters, includingthe Arctic, whereas N. umbonata inhabits both polluted localities and, moreover, hotsprings. From a strictly ecological standpoint, specimens from Beowulf have far closeraffinity to Lange-Berlatot’s (1978) N. umbonata. Because the combination thermalis var.minor predates all of the potentially equivalent taxa (and/or closely-allied forms), wehave identified the Beowulf specimens as Nitzschia cf. thermalis var. minor Hilse. At thispoint, and pending further analysis, we feel that proposing a new name would onlyfurther confuse nomenclatural issues within the ‘N. thermalis’ complex. There appear to9

176177178be no similar specimens held in the collections of either the Canadian Museum of Natureor the Philadelphia Academy of Natural Sciences (P.B. Hamilton & M.B. Edlund, pers.comm.).179180181182183184185186187188189190191192193194195Nonetheless, it is clear that diatoms of the N. umbonata-thermalis cluster occur in anumber of locations in Europe and N. America that have extreme water chemistries. Forexample, Whitton & Diaz (1981) recorded N. thermalis at acid mine drainage sites withpH between 3 and 4 in the U.K. and Belgium. Stockner (1967) noted the presence of N.thermalis in alkaline hot springs (>35 o C) situated in Mount Rainier, Washington, USA,but not in similar springs from the Upper Geyser Basin, YNP, suggesting that this speciesis not necessarily acidobiontic. It should be noted that N. thermalis is larger and has nocentral interruption of the fibulae, in contrast to our specimens and N. thermalis var.minor sensu Hilse (Van Heurck, 1880-1885), though both nominate and varietal formsseem to occur in similar environments. Weed (1889) found specimens of Denticulathermalis Kütz. in hot springs of the Pelican Creek area in YNP, which he may haveconfused with any number of Nitzschia. From our extensive literature survey, it appearsthat Nitzschia cf. thermalis var. minor is not widespread in hot springs (Copeland, 1936;Stockner, 1967; Whitton & Diaz, 1981; Villeneuve & Pienitz, 1998; DeNicola, 2000).This may suggest that N. cf. thermalis var. minor may be restricted to only the hottestsprings of high ionic strength and greatest acidity.196197198Pinnularia acoricola Hustedt (Figs. 2K-M & 3G-I)Length: 11-17.5 µm10

199200201202203204205206207208209210Width: 2.5-3.5 µmStriae / 10 µm: 13-17P. acoricola has a lanceolate to slightly oval valve morphology, with broadly roundedapices. The striae are strongly convergent at the poles and radial in the central area. Thecentral area is either rounded or interrupts striae extending to the valve margin. Theexternal distal ends of the raphe are hooked in the opposite direction to the proximalends. DeNicola (2000) discusses the misidentification of P. acoricola as P. obscura, andsimilarities to P. chapmaniana. However, neither of these species can be consideredsynonyms to P. acoricola. Our specimens, examples of which are shown both in LM(Figs. 2K-M) and SEM (Figs. 3G-I), are consistent with detailed descriptions and lightmicrographs offered by Carter (1972), Krammer & Lange-Bertalot (1986) and Watanabe& Asai (1995).211212213214215216217218219220221P. acoricola has been observed in a number of acidic waters globally (Hustedt, 1937-1938; Carter, 1972; Hargreaves et al., 1975; Whitton & Diaz, 1981; Cassie & Cooper,1989; Watanabe & Asai, 1995; DeNicola, 2000; Jordan, 2001). From the literature, itseems clear that this species is endemic to highly acidic environments, namely hotsprings and acid mine drainage. The pH of waters in which it has been observed rangesfrom

222223224225concentrations of Cu and Zn in Beowulf Spring are four orders of magnitude lower(Table 1). It is also noted that P. acoricola has been documented in close association withCyanidium caldarium elsewhere, in both hot springs (Whitton & Diaz, 1981) andendolithic habitats (Hernández-Chavarría & Sittenfeld, 2006).226227228229230231232233234235236237238Eunotia exigua (de Brébisson ex. Kützing) Rabenhorst 1864 (Figs. 2A-E & 3J-L)Length: 8-31 µmWidth: 3-5 µmStriae / 10 µm: 16-25Valves are crescent-shaped and isopolar, with ends capitate to rostrate, the ventral marginis slightly concave. Striae are punctate and transapical. Raphe is short, located on theventral margin of the mantle, and occasionally migrate to the valve face distally with acurved terminal fissure (Fig. 3J), where it terminates internally in a helictoglossa (Fig.3L). A single rimoportula occurs at one pole of each valve. There is considerablevariability in the morphology of Eunotia exigua, which has led to the reference of thespecies as a complex or ‘E. exigua sensu lato’ (Krammer & Lange-Bertalot, 1991;DeNicola, 2000).239240241242243244Eunotia exigua is truly a cosmopolitan species, present in lakes, ponds, bogs and hotsprings from tropical to arctic regions (Scherer, 1981; Whitton & Diaz, 1981; Round,1991; DeNicola, 2000; Joynt & Wolfe, 2001). E. exigua is an acidobiontic taxa occurringalmost exclusively in acidic environments, and having a pH optimum

245246acid mine drainage. Furthermore, concentrations of Cu and Zn >100 mg l -1 did notimpede the diatom’s viability.247248249250251252253254255256257258259260261262263Nitzschia ovalis Arnott (Figs 2I-J & 3F)Length: 12-20.5 µmWidth: 3.5-5 µmFibulae / 10 µm: 12-14Valves are elliptical with rounded apices. Striae are fine and generally unresolvable inlight microscopy (Figs. 2I-J) and require SEM to reveal their morphology (Fig. 3F).There is no central area or nodule, and fibulae are regularly spaced on the valve marginwith no central interruption. The raphe structure is peripheral and not visible in LM.There is potential for misidentification of N. ovalis as N. communis Rabenhorst, and viceversa (DeNicola, 2000). The morphology of N. communis is slightly narrower, and striaeare more evident under LM. Nitzschia ovalis also shares features with N. pusilla Grunowemend. Lange-Bertalot and N. aurariae Cholnoky, both of which are narrower and, in thecase of N. pusilla, exhibits slightly rostrate apices (Krammer & Lange-Bertalot, 1988;DeNicola, 2000). The Beowulf specimens of N. ovalis are consistent with both typematerial in Van Heurck (1880-1885) and European specimens described and illustratedby Krammer & Lange-Bertalot (1988).264265266267Very few records of N. ovalis exist for freshwater environments despite its originaldescription as a continental species. However, there are a number of marine accounts forthis diatom, and it has been referred to as an estuarine benthic taxon (Saks, 1982).13

268269270271272273Laboratory work showed that optimal growth occurred in waters of high ionic strength(i.e., 0.03 M) and the cultures tolerated temperatures up to 36 o C (Saks, 1982). Itspresence in Beowulf Spring most likely relates to the high ionic strength of the water,coupled to broad tolerances with respect to pH. Other reported occurrences of N. ovalis infreshwaters seem to be restricted to acid mine drainage in the U.K. (Hargreaves et al.,1975).274275276277278279280281282283284285286287288289290ConclusionsThe extreme conditions of Beowulf Spring reduce algal diversity to taxa with tolerance oflow pH, high ionic strength, and elevated concentrations of aqueous metals that areinhibitory or toxic to other groups. Clear zonation between cyanidiacean (green) anddiatom (brown) mats within the Beowulf Spring show that the habitat is furthersubdivided spatially (Figure 1). The possibility of competitive exclusion between thehabitats occupied by either group cannot be dismissed, all the more as both groups appearto metabolize silica actively. Epipelic diatoms from the brown mats in Beowulf Spring,and their close affinities with other hot-spring assemblages, raise several interestingquestions concerning the ecophysiology and biogeography of diatoms. It appears that thealgae in the Beowulf system are not merely surviving extreme environmental conditionsbeyond the tolerances of other forms, but potentially deriving benefits from certainaspects of this habitat. Low diatom diversity and hence reduced inter-specificcompetition, coupled to the absence of predators, are obvious factors that may contributeto the success of this community. The potential metabolism of toxic metals by thesediatoms must also be considered, and this merits further investigation. The apparent14

291292293294295296297298active silica metabolism of Cyanidiaceae may provide considerable phylogenetic insight,given the relationship between diatoms and red algae (Keeling et al. 2004). In terms ofbiogeography, two of the diatoms, Pinnularia acoricola and Nitzschia cf. thermalis var.minor, appear almost exclusively associated with hot-springs, whereas another, N. ovalis,is halophilous, and a fourth, Eunotia exigua, is an acidobiont that inhabits acid-minedrainages and naturally-acidic environments with equal facility. It remains a mystery howsuch disjunct associations have emerged. Together, our observations provide strongincentives to more thoroughly catalog the diatom floras of YNP springs.299300301302303304305306307308309AcknowledgmentsWe thank the National Park Service (Department of the Interior) for permission toconduct research in Yellowstone National Park. We also thank Stefan Lalonde forsampling and laboratory support. Drs. Paul Hamilton and Mark Edlund graciouslyconfirmed the lack of Nitzschia cf. thermalis var. minor in the herbariums of theCanadian Museum of Nature and the Philadelphia Academy of Natural Sciences,respectively. Research was supported by the Natural Sciences and Engineering ResearchCouncil of Canada (NSERC) and the National Science Foundation (NSF MicrobialObservatory). This paper is dedicated to Dr. Eugene Stoermer for his enormouscontribution to diatom research.31015

310311312ReferencesANDERSON, N.J. (2001): Diatoms, temperature and climatic change. – Eur. J. Phycol.35: 307-314.313314315ASADA, R. & K. TAZAKI (2001): Silica biomineralization of unicellular microbesunder strongly acidic conditions. – Can. Mineralogist 39: 1-16.316317318BROCK, T.D. & M.L. BROCK (1966): Temperature optima for algal development inYellowstone and Iceland hot springs. – Nature 209: 733-734319320321322BOURELLY, P. & E. MANGUIN, (1952) Algues d’eau douce de la Guadaloupe etdépendences. – Centre National de la Recherche Scientifique: 1-281. Soc. d’editiond’Enseignements Sup.,Paris.323324325CARTER, J.R. (1972): Some observations on the diatom Pinnularia acoricola Hustedt. –Microscopy 32: 162-165326327328CASSIE, V. (1989): A contribution to the study of New Zealand diatoms. - BibliothecaDiatomologica, Band 17: 1-266. J Cramer, Berlin.329330331CASSIE, V. & R.C. COOPER (1989): Algae of New Zealand thermal areas. -Bibliotheca Phycolologica Band 78: 1-261. J. Cramer, Berlin.33216

333334CHARLES, D. F. (1985): Relationships between surface sediment diatom assemblagesand lakewater characteristics in Adirondack lakes. - Ecology 66: 994–1011.335336337338COMPÈRE, P. & A. DELMOTTE (1986): Diatoms in two hot springs in Zambia(Central Africa). – In: F.E. ROUND. Proceedings of the Ninth International DiatomSymposium: 29-40. Biopress Ltd., Bristol.339340341COPELAND, J.J. (1936): Yellowstone thermal Myxophyceae. – Ann. N.Y. Acad. Sci.36: 1-232.342343344DENICOLA, D. (2000): A review of diatoms found in highly acidic environments. –Hydrobiol. 433: 111-122.345346347348EDMONDS, J.S. & K.A. FRANCESCONI (1998) Arsenic metabolism in aquaticecosystems. – In: W.J. LANGSTON & M.J. BEBIANNO (eds.) Metal Metabolism inAquatic Environments: 159-183. Chapman and Hall, London.349350351352FOURNIER, R.O. (1985): The behaviour of silica in hydrothermal solutions. – In: B.R.BERGER & P.M. BETHKE (eds.) Geology and Geochemistry of Epithermal Systems:45-61. Reviews in Economic Geology Vol. 2.353354355HARGREAVES, J.W., E.J.H. LLOYD & B.A. WHITTON. (1975): Chemistry andvegetation of highly acidic streams. – Freshwat. Biol. 5: 563-576.17

356357358359HERNÁNDEZ-CHAVARRÍA, F. & A. SITTENFELD. (2006): Preliminary report on theextreme endolithic microbial consortium of ‘Pailas Frías’, ‘Rincón de la Vieja’Volcano, Costa Rica. – Phycol. Res. 54: 104-107.360361362HUSTEDT, F. (1937-1938): Systematische und Untersuchungen über denDiatomeenflora von Java, Bali und Sumatra. – Arch. Hydrobiol., Suppl. 15 & 16.363364365366367INSKEEP, W.P., R.E. MACUR, G. HARRISON, B.C. BOSTICK & S.FENDORF(2004): Biomineralization of As(V)-hydrous ferric oxyhydroxide in microbial matsof an acid-sulfate-chloride geothermal spring, Yellowstone National Park. – Geochim.Cosmochim. Acta. 68: 3141-3155.368369370371372JORDAN, R.W. (2001): Taxonomy, morphology and distribution of two Pinnulariaspecies from acidic lakes and rivers in Yamagata and Miyagi Prefectures, NortheastJapan. – In: R.JAHN, J.P. KOCIOLEK, A. WITKOWSKI & P.COMPÈRE. Studies onDiatoms: Lange-Bertalot Frestschrift: 279-302. A.R.G. Gantner Verlag K.G.373374375376JOYNT III, E.H. & A.P. WOLFE (2001): Paleoenvironmental inference models fromsediment diatom assemblages in Baffin Island lakes (Nunavut, Canada) andreconstruction of summer water temperature. – Can. J. Fish Aquat. Sci. 58: 1222-1243.37718

378379380JUGGINS, S (1992): Diatoms in the Thames Estuary, England: ecology, paleoecology,and salinity transfer function. - Bibliotheca Diatomologica, Band 25: 1-216. J. Cramer,Berlin.381382383KEELING, P. J., J.M. ARCHIBALD, N.M. FAST, & J.D. PALMER (2004): Commenton ‘The evolution of modern eukaryotic phytoplankton’. Science 306: 2191b.384385386KONHAUSER, K.O., B. JONES, V.R. PHOENIX, G. FERRIS & R.W. RENAUT(2004). The microbial role in hot spring silicification. Ambio 33: 552-558.387388389390KRAMMER, K. & H. LANGE-BERTALOT (1986): Bacillariophyceae 1. Teil:Bacillariophyceae. – In: H. ETTL, J. GERLOFF, H. HEYNIG & D. MOLLENHAUER(eds). Süßwasserflora Von Mitteleuropa 2. Fischer, Stuttgart.391392393394KRAMMER, K. & H. LANGE-BERTALOT (1988): Bacillariophyceae 2. Teil:Bacillariaceae, Epithemiaceae, Surirellaceae. – In: H. ETTL, J. GERLOFF, H. HEYNIG& D. MOLLENHAUER (eds). Süßwasserflora Von Mitteleuropa 2. Fischer, Stuttgart.395396397398KRAMMER, K. & H. LANGE-BERTALOT (1991): Bacillariophyceae 3. Teil:Centrales, Fragilariaceae, Eunotiaceae. – In: H. ETTL, J. GERLOFF, H. HEYNIG & D.MOLLENHAUER (eds). Süßwasserflora Von Mitteleuropa 2. Fischer, Stuttgart.39919

400401402LANGE-BERTALOT, H. (1978): Zur systematic, taxonomie und ökologie desabwasserspezifish wichtigen formenkreises um “Nitzschia thermalis”. – Nova Hedwigia30: 635-652.403404405406LANGNER, H.W., C.R. JACKSON, T.R. MCDERMOTT & W.P. INSKEEP (2001):Rapid oxidation of arsenite in a hot spring ecosystem, Yellowstone National Park. –Environ. Sci. Technol. 35: 3302-3309.407408409MANN, J.E. & H.E. SCHLICHTING (1967): Benthic algae of selected thermal springsin Yellowstone National Park. – Trans. Amer. Microsc. Soc. 86: 2-9.410411412PATRICK, R. (1977): Ecology of freshwater diatoms. – In:. Werner, D (ed).: TheBiology of Diatoms: 284-332. Botanical Monographs Vol.13. Blackwell, UK.413414415ROUND, F.E. (1991): Epilithic diatoms in acid water streams flowing into the resevoirLlyn Brianne. - Diatom Res. 6: 137-145.416417418419SAUNDERS, J.G. & G.F. RIEDEL (1998): Metal accumulation and impacts inphytoplankton. – In: W.J. LANGSTON & M.J. BEBIANNO (eds). Metal Metabolism inAquatic Environments: 59-76. Chapman and Hall, London.420421422SAKS, N.M. (1982): Temperature, salinty and ultraviolet irradiation effects on thegrowth of strains of Nitzschia ovalis. – Mar. Biol. 68: 175-179.20

423424425426SCHOEMAN, F.R. & R.E.M. ARCHIBALD (1988): Taxonomic notes n the diatoms(Bacillariophyceae) of the Gross Barmen thermal springs in South West Africs/Namibia.– S. Afr. J. Bot. 54: 221-256.427428429SHERER, R.P. (1981): Freshwater diatom assemblages and ecology/palaeoecology of theOkefenokee swamp/marsh complex, Southern Georgia, USA. - Diatom Res. 3: 129-157.430431432SIMONSEN, R. (1987): Atlas and catalogue of the diatom types of Friedrich Hustedt,Vol: 2. - J. Cramer, Stuttgart.433434435STOCKNER, J.G. (1967): Observations of thermophilic algal communities in MountRainier and Yellowstone National Parks. – Limnol. Ocean. 12: 13-17436437438SUZUKI, Y. & M. TAKAHASHI (1995): Growth responses of several diatom speciesisolated from various environments to temperature. - J. Phycol. 31: 880-888.439440441VAN HEURCK, H. (1880-1885): Synopsis des diatomées de Belgique. – Ducaju & Cie.,Anvers.442443444445VERB, R.G. & M.L. VIS (2000): Comparison of benthic diatom assemblages fromstreams draining abandoned and reclaimed coal mines and nonimpacted sites. - J. N. Am.Benth. Soc. 19: 274-28821

446447448VILLENEUVE, V. & R. PIENITZ (1998): Composition Diatomifère de quatre sourcesthermals au Canada, en Islande et au Japon. - Diatom Res. 13: 149-175.449450451452WATANABE, T. & K. ASAI (1995): Pinnularia acoricola Hustedt var. acoricola as anenvironmental frontier species occurred in inorganic acid waters with 1.1 – 2.0 in pHvalue. – Diatom 10: 9-11.453454455WEED, W.H. (1889): The diatom marshes and diatom beds of the Yellowstone NationalPark. – Bot. Gaz. 14: 117-120.456457458459WHITTON, B.A. & B.M. DIAZ (1981): Influence of environmental factors onphotosynthetic species composition in highly acidic waters. Verh. Internat. Verein.Limnol. 21: 1459-1465.460461462WOLFE, A.P. (2002): Climate modulates the acidity of arctic lakes on millennialtimescales. Geology 30: 215-218.22

Table 1: Water chemistry of Beowulf Hot Spring, Yellowstone National Park. Allconcentrations are reported in µM, except where indicated.ParameterConcentrationNa 11136*Si 4690K 1225Ca 125Al 124As 19.6Mg 8.2Fe 6.2*P 4.5Zn 0.9Mn 0.6Cu 0.1anions*Cl - 137702-*SO 4 1510*F - 147-*NO 3 20.0otherpH 2.26Temp ( o C) 67Ionic strength (M) 0.02* data from Inskeep et al. (2004)23

Table 2: A comparison of the morphology, ultrastructure and ecology of Nitzschia cf.thermalis var. minor from Beowulf Spring, Yellowstone National Park, U.S.A. andsimilar species.Nitzschia cf. N. hombergiensis b N. umbonata b N. steynii c N. rimosa dminor athermalis var.Length (µm) 15-28 32-52 22-125 27-54.5 60Width (µm) 2-4 5-6 6-9 4-5 7Striae / 10µm 23-27 34-40 24-30 24-26 25-28Fibulae / 10µm) 8-10 9-15 7-10 10-12 7-9Fibulae notesCentral areaconstricted0.5-1 µm thickon valve margin;irregularlyspacedIrregular & widelyspaced; can extendto centre of valveShort, narrow,regularly spaced;convergence of 2-3straieProminent& evenlyspacedyes yes slight variable noHabitat Thermal springs Circumneutral;unpolluted watersThermal springs;eutrophic waters;acid mine drainagea. Description of specimens from Beowulf Spring, Yellowstone Nat. Park, U.S.A (this paper).b. Lange-Bertalot (1978) and Krammer & Lange-Bertalot (1988)c. Schoeman & Archibald (1988)d. Bourelly & Manguin (1952)ThermalspringsRobustThermalsprings24

463464465466467468469470471472473474475476Figure captionsFig. 1. A: Beowulf Spring, Norris Geyser Basin. Brown areas are hydrous ferric oxides(HFO) and green areas represent cover by rhodophytes including Cyanidium. B: Close-upof the transition between the HFO mat and Cyanidiaceae, with penny for scale. C: SEMphotograph of silica casts produced by coccoid Cyanidiaceae cells. D: Close-up of cellsin (C) showing remnants of soft-bodied cells within the silica cast (cc). E-H: SEMmicrographs of diatoms from the HFO mat. (E) Nitzschia cf. thermalis var. minor withclean areolae devoid of precipitate. (F) Nitzschia ovalis (No) and Pinnularia acoricola(Pa), also with clean areolae; the cast of a Cyanidiaceae cell is also visible below N.ovalis. (G) A cell of N. ovalis encrusted in a diagenetic silica precipitate that completelysheathes the cell and blocks areolae. (H) A Pinnularia acoricola cell with diageneticsilica papillae on the apical cell surface; areolae are visible beneath. Scale bars are 10µm, except (H) where it is 1 µm.477478479480481482483484Fig. 2. Light micrographs of Beowulf Spring diatoms from cleaned samples (1000xmagnification). A-E: Eunotia exigua. (A-D) Valve views showing variable morphology.(E) Girdle view showing terminal raphe nodules at the apices (arrows). F-H: Nitzschia cf.thermalis var. minor. (F-G) Valve view; central constriction of valve margin andinterruption of raphe and fibulae; fine striae (H) girdle view. I-J: Nitzschia ovalis valveview; dense fibulae and striae not resolvable under LM. K-M: Pinnularia acoricola (K-L) valve view. (M) girdle view. Scale bar is 10 µm.48525

486487488489490491492493494495496497Fig. 3. SEM photographs of diatoms from cleaned samples of Beowulf Spring. A-E:Nitzschia cf. thermalis v. minor. (B) Shows the external view of the central nodule (cn)with the eccentric raphe. (C) An internal view showing the fibulae on the valve marginwith the interior of central nodule (cn) visible. (D) Apiculate distal end of the valveshowing fibulae and distal raphe end terminating in a helictoglossa (h). (E) Apiculate endwith eccentric raphe F: Nitzschia ovalis cell. G-I: Pinnularia acoricola. (G) Hypovalveview showing the strongly convergent striae at the poles and radial pattern in the centralarea. (H) Girdle view (I) distal raphe end deflected; striae appear as groups of individualpuncta in costa-like areas. J-L: Eunotia exigua. (J) Valve view with the curved rapheterminus on the valve face. (K) Raphe structure on the valve mantle with helictoglossaevident. (L) Close-up of the internal raphe terminus showing the helictoglossa (h) andrimoportulae (r). White scale bars are 1 µm, except A, G, H, J, & K where it is 10 µm.26

ACDBccEF G HPaNoFigure 1

a b c d e f f’ g g’ h h’i i’ j j’ k k’ l l’ m m’Figure 2

cncnA B ChDEFG H IhrJKLFigure 3