1

Epipelic diatoms from an extreme acid environment: Beowulf Spring, Yellowstone National Park, U.S.A. William O. Hobbs*1, Alexander P. Wolfe1, William P. Inskeep2, Larry Amskold1 & Kurt O. Konhauser1 1Department of Earth & Atmospheric Sciences University of Alberta Edmonton, AB T6G 2E3 Canada 2Department of Land Resources & Environmental Sciences Montana State University Bozeman, MT 59717 USA *author for correspondence (whobbs@ualberta.ca)

2

Abstract 1

A floristic study is presented of the diatom community from Beowulf Spring, 2

Yellowstone National Park (YNP), U.S.A., a strongly acidic (pH<3) and thermophilous 3

(temperature >50oC) environment with potentially toxic concentrations of dissolved 4

metals. A sampling transect between regions of active hydrous ferric oxide (HFO) 5

precipitation and mats of periphytic red algae (Cyandiaceae) revealed that diatoms are 6

present mainly in the regions of the spring characterized by the highest temperatures and 7

brown arsenic-rich HFO mats. The diatom flora is dominated by a deme within the 8

Nitzschia thermalis complex, in association with N. ovalis, Pinnularia acoricola, and 9

Eunotia exigua. Abiotic silica precipitates were observed on the surfaces of most diatom 10

cells in uncleaned material. Cells of Cyanidiaceae also participate in active silica 11

deposition. These observations not only highlight the tolerance of hot-spring diatoms to 12

extreme environments with respect to temperature, pH, and arsenic (As) concentration, 13

but also illustrate the clear zonation between epipelic algal communities within a hot 14

spring. 15

16

3

Introduction 16

Although diatoms have been shown to respond to temperature in culture (Patrick, 1977; 17

Suzuki & Takahashi, 1995), it remains uncertain whether they are directly thermophilous 18

in an eco-physiological sense, or indirectly influenced by environmental conditions that 19

covary with climate (Anderson, 2001; Wolfe, 2002). The bulk of previous studies on 20

diatoms from hot spring environments has focused solely on temperature (Hustedt, 1937-21

1938; Stockner, 1967; Brock & Brock, 1966), with some attention to the influences of 22

low ambient pH and elevated concentrations of potentially toxic elements, such as arsenic 23

(As) and copper (Cu) (Hargreaves et al., 1975; Whitton & Diaz, 1981). 24

25

In this paper, we present an account of the diatom flora from the Beowulf Spring in the 26

Norris Geyser Basin, Yellowstone National Park (YNP). This epipelic community 27

inhabits As-rich hydrous ferric oxide (HFO) mats that coat the sandy substratum of the 28

spring proximal to the vent (Figure 1). Peripheral to the HFO zone, a green algal mat 29

composed of cyanidiaceous Rhodophyta contains few, if any, diatoms. Light & scanning 30

electron microscopic (SEM) investigations of these communities are presented from both 31

uncleaned and oxidized materials. Water chemistry implies that the diatoms are 32

exploiting an ecological niche that may be toxic to other algae. Contrary to Villeneuve & 33

Pienitz’s (1998) suggestion that there is no single diatom assemblage typical of thermal 34

springs, our study suggests a strong floristic similarity among highly acidic hot springs 35

globally (Hustedt, 1937-1938; Carter, 1972; Hargreaves et al., 1975; Whitton & Diaz, 36

1981; Cassie & Cooper, 1989; Watanabe & Asai, 1995; DeNicola, 2000; Jordan, 2001), 37

4

implying that indeed there exists a distinct but biogeographically-disjunct flora in these 38

extreme environments. 39

40

Methods 41

Beowulf Hot Spring (44°43’53” N, 110°42’41” W) lies within the Norris Geyser Basin, 42

Yellowstone National Park, Wyoming, U.S.A. (YNP Thermal Inventory #NHSP35)) 43

(Fig. 1A). Surface sediment collections were obtained from both the central portion of the 44

spring, which is coated by HFO, and peripheral green algal mats (Fig. 1B). Samples were 45

either cleaned using 30% H2O2, or mounted uncleaned following dilution of the slurry in 46

deionized water. Valves were mounted in Naphrax and observed at 1000x under oil 47

immersion with differential interference contrast optics (Leica DMLB). For scanning 48

electron microscopy (SEM), slurry was evaporated onto stubs at room temperature in a 49

desiccator, prior to sputter-coating with Au, and examination in a JEOL-6301F field-50

emission SEM. Archives of these samples are kept at 4oC at the University of Alberta. 51

Our results confirms that it is clearly advantageous to examine both cleaned and 52

uncleaned material (Stockner 1967). 53

54

Beowulf is considered an acid-sulfate-chloride spring, where the spring discharges 55

extremely acidic (pH<3) and hot (>50˚C) water, with mM concentrations of Na+, K+, Cl–, 56

SO42- and Si (Figure 1; Table 1; Inskeep et al. 2004). Water samples were obtained by 57

syringe and filtered through 0.45 µm in-line nylon syringe filters. Samples were further 58

acidified with TraceMetal-grade HNO3 and kept refrigerated until analysis by ion 59

chromatography. Temperature and pH were measured in the field using a digital 60

5

thermometer and an Orion Ross (8165BN) pH meter calibrated between 2 and 4, 61

respectively. 62

63

Results and Discussion 64

Water chemistry 65

The Beowulf Spring has an ionic strength of 0.02 M, which is comparable to a marine – 66

brackish estuarine salinity classification (Juggins, 1992). Aqueous chemistry from 67

Beowulf Spring has been characterized on multiple occasions and exhibits very little 68

range in concentrations (±10% or less for most constituents; Inskeep et al., 2004). The 69

water chemistry reveals the presence of potentially toxic metals to algal growth, namely 70

Cu and As (Table 1). In particular, As concentrations reach 20 µM, much higher than 71

minimum concentrations required to inhibit growth of diatoms in culture (Saunders & 72

Riedel, 1998). Similarly, Beowulf Cu concentrations are two to three orders of magnitude 73

greater than those used by Saunders & Riedel (1998). Thus, natural concentrations of As 74

and Cu in the Beowulf Spring approach or exceed those from the most highly impacted 75

acid-mine drainages (Whitton & Diaz, 1981; Verb & Vis, 2000). 76

77

While it is not our intention to speculate as to whether there is active uptake of As by 78

diatoms in Beowulf Spring, As uptake is likely occurring among other microorganisms in 79

YNP hot springs, namely bacteria (Langner et al., 2001; Inskeep et al., 2004). 80

Metabolism of As by marine algae is hypothesized to occur because of the molecular 81

similarity between phosphorous and As (Edmonds & Francesconi, 1998). General 82

pathways of metals metabolism by algae include sequestration, complexation with 83

6

exuded dissolved organic carbon, and redox transformations during uptake (Saunders & 84

Riedel, 1998). Thus, a range of biogeochemical processes exist to allow not only the 85

survival, but indeed the success, of algae inhabiting these extreme environments. 86

87

The pH of Beowulf Spring at the time of sampling was 2.26, and the temperature was 88

67oC. A number of diatom communities have been documented at pH of <3, although 89

species richness decreases rapidly below a pH of 4.5 (DeNicola, 2000). The temperature 90

of Beowulf Spring exceeds that of most other springs that support diatoms (Brock & 91

Brock, 1966; Stockner, 1967; Hargreaves et al., 1975; Whitton & Diaz, 1981), with rare 92

exceptions (Mann & Schlichting, 1967; Compère & Delmotte, 1986). 93

94

Green mats 95

Samples of the green Cyanidiaceae mat lack diatoms (Fig. 1A-B), with the exception of 96

occasional frustules that presumably originate from transport. However, significant 97

amorphous silica precipitation occurs in association with Cyanidiaceae cells (Fig. 1C-D). 98

Silica biomineralization under highly acidic conditions is not well-documented, with the 99

exception of Asada & Tazaki’s (2001) study of Cyanidium caldarium in a Japanese 100

spring at pH<2. They proposed passive models including nucleation of silica on cell 101

walls, polymerization of silicic acid, and adhesion of colloidal silica gel. These models, 102

however, contradict general observations that spontaneous silica polymerization and 103

sinter formation typically occur at circumneutral pH (Konhauser et al. 2004), and that 104

excess H+ inhibits silica polymerization (Fournier, 1985). Nonetheless, the end product, 105

smooth cyst-like features that frequently coalesce between adjacent cyanidiacean cells 106

7

(Asada & Tazaki, 2001), is identical to our observations from the Beowulf material (Fig. 107

1C-D). 108

109

Brown mats 110

The HFO mat contained abundant diatoms that occur as solitary epipelic forms. All 111

observed specimens are biraphid taxa, hence capable of motile displacement within the 112

mat. Several specimens preserved intact chloroplasts when viewed in LM. SEM revealed 113

diatom surfaces both completely free of secondary silica precipitates and with clean 114

areolae (Fig. 1E & F), as well as thoroughly Si-encrusted specimens (Fig. 1G & H). In 115

the latter instances, inorganic Si precipitation has formed a continuous botryoidal coating, 116

comprised of coalescent individual papillae of approximately 100 nm diameter. We 117

hypothesize that living diatoms have cleaner valve surfaces than dead ones, and perhaps 118

that inorganic silica precipitation may at times overwhelm living cells. Inskeep et al. 119

(2004) have provided a complete characterization of the brown mat, revealing a variety of 120

SiO2 crystalline polymorphs and microbial communities encrusted with As-rich HFO. It 121

is unclear to us at this time what aspect of the HFO mat contributes to the success of the 122

diatoms living there, in sharp contrast to their absence in the green mat. Given that 123

Cyanidiaceae are virtually absent in the HFO mats, it is also possible that resource 124

competition plays a role in delimiting these algae. 125

126

Floristic description 127

The epipelic diatom community in Beowulf Spring consisted of five species (Table 1). 128

The dominant diatom taxon was Nitzschia cf. thermalis var. minor Hilse (56%), followed 129

8

in decreasing order of relative abundance by Pinnularia acoricola Hustedt (21%), 130

Eunotia exigua (de Brébisson ex Kützing) Rabenhorst (15%), Nitzschia ovalis Arnott 131

(7%) and Encyonema minutum (Hilse ex Rabenhorst) D.G. Mann (<1%). We discuss the 132

four dominant diatoms further below, and provide LM and SEM micrographs of each. 133

We do not consider E. minutum further because its abundance is insufficient to ascertain 134

with confidence that it was truly living in the spring. 135

136

Nitzschia cf. thermalis var. minor (Figs. 1E, 2F-H & 3A-E, Table 2) 137

Length: 15 – 28 µm 138

Width: 2-4 µm 139

Striae / 10 µm: 23 - 27 140

Fibulae / 10 µm: 8 - 11 141

We have identified specimens from the Beowulf Spring that appear to conform with N. 142

thermalis var. minor Hilse according to sketches by Grunow in Van Heurck (1880-1885; 143

Plate 59, Fig. 22) of the isotype material collected by Rabenhorst (no. 1266). Lange-144

Bertalot (1978) considered Nitzschia thermalis var. minor sensu Hustedt (Simonsen 1987, 145

Plate 346, Fig. 9-11) to actually represent a small specimen of N. umbonata (Ehrenberg) 146

Lange-Bertalot. He then proposed a new taxon, N. homburgiensis Lange-Bertalot, for the 147

N. thermalis complex, based on fibula morphology and the central constriction of the cell 148

margin. N. umbonata was combined from Ehrenberg’s basionym Navicula umbonata and 149

is a synonym for Nitzschia thermalis sensu Grunow (Lange-Bertalot, 1978). Of the two 150

species N. umbonata and N. homburgiensis, the Beowulf specimens more closely 151

conform to the latter (Table 2). Both our specimens and N. homburgiensis are linear with 152

9

apiculate ends, transapical striae composed of fine areolae, and a constricted central 153

margin (Figs. 3A-E). REM of N. homburgiensis and the isotype material of N. thermalis 154

var. minor Hilse reveal that the fibulae are irregular, widely spaced, and can extend to the 155

center of the transapical axis (Plate 39 of Lange-Bertalot, 1978). Comparatively, the 156

fibulae from Beowulf specimens range in thickness from 0.5 to 1.0 µm along the apical 157

axis (Fig. 2C) and are confined to the margin of the valve. In addition the raphe of our 158

specimens is more eccentric than that of N. homburgiensis (Fig. 3B). Cassie (1989) 159

illustrated N. umbonata from a shallow thermal pool (Tokannu, New Zealand) with 160

similar fibulae and raphe structures to Beowulf specimens, but associated with 161

significantly larger valves (36-45 µm). Morphometric comparisons of N. cf. thermalis 162

var. minor, N. hombergiensis, N. umbonata, N. steynii Cholnoky emend. (Schoeman & 163

Archibald, 1988), and N. rimosa Manguin (Bourelly & Manguin, 1952) indicate that each 164

of these taxa has a distinctive morphology (Table 2). 165

166

Further questions arise in considering the ecologies of these diatoms. Lange-Bertalot 167

(1978) associates N. homburgiensis with circumneutral, unpolluted freshwaters, including 168

the Arctic, whereas N. umbonata inhabits both polluted localities and, moreover, hot 169

springs. From a strictly ecological standpoint, specimens from Beowulf have far closer 170

affinity to Lange-Berlatot’s (1978) N. umbonata. Because the combination thermalis var. 171

minor predates all of the potentially equivalent taxa (and/or closely-allied forms), we 172

have identified the Beowulf specimens as Nitzschia cf. thermalis var. minor Hilse. At this 173

point, and pending further analysis, we feel that proposing a new name would only 174

further confuse nomenclatural issues within the ‘N. thermalis’ complex. There appear to 175

10

be no similar specimens held in the collections of either the Canadian Museum of Nature 176

or the Philadelphia Academy of Natural Sciences (P.B. Hamilton & M.B. Edlund, pers. 177

comm.). 178

179

Nonetheless, it is clear that diatoms of the N. umbonata-thermalis cluster occur in a 180

number of locations in Europe and N. America that have extreme water chemistries. For 181

example, Whitton & Diaz (1981) recorded N. thermalis at acid mine drainage sites with 182

pH between 3 and 4 in the U.K. and Belgium. Stockner (1967) noted the presence of N. 183

thermalis in alkaline hot springs (>35oC) situated in Mount Rainier, Washington, USA, 184

but not in similar springs from the Upper Geyser Basin, YNP, suggesting that this species 185

is not necessarily acidobiontic. It should be noted that N. thermalis is larger and has no 186

central interruption of the fibulae, in contrast to our specimens and N. thermalis var. 187

minor sensu Hilse (Van Heurck, 1880-1885), though both nominate and varietal forms 188

seem to occur in similar environments. Weed (1889) found specimens of Denticula 189

thermalis Kütz. in hot springs of the Pelican Creek area in YNP, which he may have 190

confused with any number of Nitzschia. From our extensive literature survey, it appears 191

that Nitzschia cf. thermalis var. minor is not widespread in hot springs (Copeland, 1936; 192

Stockner, 1967; Whitton & Diaz, 1981; Villeneuve & Pienitz, 1998; DeNicola, 2000). 193

This may suggest that N. cf. thermalis var. minor may be restricted to only the hottest 194

springs of high ionic strength and greatest acidity. 195

196

Pinnularia acoricola Hustedt (Figs. 2K-M & 3G-I) 197

Length: 11-17.5 µm 198

11

Width: 2.5-3.5 µm 199

Striae / 10 µm: 13-17 200

P. acoricola has a lanceolate to slightly oval valve morphology, with broadly rounded 201

apices. The striae are strongly convergent at the poles and radial in the central area. The 202

central area is either rounded or interrupts striae extending to the valve margin. The 203

external distal ends of the raphe are hooked in the opposite direction to the proximal 204

ends. DeNicola (2000) discusses the misidentification of P. acoricola as P. obscura, and 205

similarities to P. chapmaniana. However, neither of these species can be considered 206

synonyms to P. acoricola. Our specimens, examples of which are shown both in LM 207

(Figs. 2K-M) and SEM (Figs. 3G-I), are consistent with detailed descriptions and light 208

micrographs offered by Carter (1972), Krammer & Lange-Bertalot (1986) and Watanabe 209

& Asai (1995). 210

211

P. acoricola has been observed in a number of acidic waters globally (Hustedt, 1937-212

1938; Carter, 1972; Hargreaves et al., 1975; Whitton & Diaz, 1981; Cassie & Cooper, 213

1989; Watanabe & Asai, 1995; DeNicola, 2000; Jordan, 2001). From the literature, it 214

seems clear that this species is endemic to highly acidic environments, namely hot 215

springs and acid mine drainage. The pH of waters in which it has been observed ranges 216

from <1 to 4, and temperatures up to 50oC. Whitton & Diaz (1981) present one of the few 217

studies which considers the direct influence of dissolved heavy metals on the success of 218

diatoms in acidic waters. The authors use 59 records of P. acoricola in an attempt to 219

elucidate any relationships between pH and Cu and Zn, they present upper limits for this 220

diatom of 10 mg Cu l-1 (0.16 mM) and 100 mg Zn l-1 (1.53 mM). Comparatively, aqueous 221

12

concentrations of Cu and Zn in Beowulf Spring are four orders of magnitude lower 222

(Table 1). It is also noted that P. acoricola has been documented in close association with 223

Cyanidium caldarium elsewhere, in both hot springs (Whitton & Diaz, 1981) and 224

endolithic habitats (Hernández-Chavarría & Sittenfeld, 2006). 225

226

Eunotia exigua (de Brébisson ex. Kützing) Rabenhorst 1864 (Figs. 2A-E & 3J-L) 227

Length: 8-31 µm 228

Width: 3-5 µm 229

Striae / 10 µm: 16-25 230

Valves are crescent-shaped and isopolar, with ends capitate to rostrate, the ventral margin 231

is slightly concave. Striae are punctate and transapical. Raphe is short, located on the 232

ventral margin of the mantle, and occasionally migrate to the valve face distally with a 233

curved terminal fissure (Fig. 3J), where it terminates internally in a helictoglossa (Fig. 234

3L). A single rimoportula occurs at one pole of each valve. There is considerable 235

variability in the morphology of Eunotia exigua, which has led to the reference of the 236

species as a complex or ‘E. exigua sensu lato’ (Krammer & Lange-Bertalot, 1991; 237

DeNicola, 2000). 238

239

Eunotia exigua is truly a cosmopolitan species, present in lakes, ponds, bogs and hot 240

springs from tropical to arctic regions (Scherer, 1981; Whitton & Diaz, 1981; Round, 241

1991; DeNicola, 2000; Joynt & Wolfe, 2001). E. exigua is an acidobiontic taxa occurring 242

almost exclusively in acidic environments, and having a pH optimum <5 (Charles, 1985). 243

Whitton & Diaz (1981) observed a pH range of 2-4 in 44 samples from hot springs and 244

13

acid mine drainage. Furthermore, concentrations of Cu and Zn >100 mg l-1 did not 245

impede the diatom’s viability. 246

247

Nitzschia ovalis Arnott (Figs 2I-J & 3F) 248

Length: 12-20.5 µm 249

Width: 3.5-5 µm 250

Fibulae / 10 µm: 12-14 251

Valves are elliptical with rounded apices. Striae are fine and generally unresolvable in 252

light microscopy (Figs. 2I-J) and require SEM to reveal their morphology (Fig. 3F). 253

There is no central area or nodule, and fibulae are regularly spaced on the valve margin 254

with no central interruption. The raphe structure is peripheral and not visible in LM. 255

There is potential for misidentification of N. ovalis as N. communis Rabenhorst, and vice 256

versa (DeNicola, 2000). The morphology of N. communis is slightly narrower, and striae 257

are more evident under LM. Nitzschia ovalis also shares features with N. pusilla Grunow 258

emend. Lange-Bertalot and N. aurariae Cholnoky, both of which are narrower and, in the 259

case of N. pusilla, exhibits slightly rostrate apices (Krammer & Lange-Bertalot, 1988; 260

DeNicola, 2000). The Beowulf specimens of N. ovalis are consistent with both type 261

material in Van Heurck (1880-1885) and European specimens described and illustrated 262

by Krammer & Lange-Bertalot (1988). 263

264

Very few records of N. ovalis exist for freshwater environments despite its original 265

description as a continental species. However, there are a number of marine accounts for 266

this diatom, and it has been referred to as an estuarine benthic taxon (Saks, 1982). 267

14

Laboratory work showed that optimal growth occurred in waters of high ionic strength 268

(i.e., 0.03 M) and the cultures tolerated temperatures up to 36oC (Saks, 1982). Its 269

presence in Beowulf Spring most likely relates to the high ionic strength of the water, 270

coupled to broad tolerances with respect to pH. Other reported occurrences of N. ovalis in 271

freshwaters seem to be restricted to acid mine drainage in the U.K. (Hargreaves et al., 272

1975). 273

274

Conclusions 275

The extreme conditions of Beowulf Spring reduce algal diversity to taxa with tolerance of 276

low pH, high ionic strength, and elevated concentrations of aqueous metals that are 277

inhibitory or toxic to other groups. Clear zonation between cyanidiacean (green) and 278

diatom (brown) mats within the Beowulf Spring show that the habitat is further 279

subdivided spatially (Figure 1). The possibility of competitive exclusion between the 280

habitats occupied by either group cannot be dismissed, all the more as both groups appear 281

to metabolize silica actively. Epipelic diatoms from the brown mats in Beowulf Spring, 282

and their close affinities with other hot-spring assemblages, raise several interesting 283

questions concerning the ecophysiology and biogeography of diatoms. It appears that the 284

algae in the Beowulf system are not merely surviving extreme environmental conditions 285

beyond the tolerances of other forms, but potentially deriving benefits from certain 286

aspects of this habitat. Low diatom diversity and hence reduced inter-specific 287

competition, coupled to the absence of predators, are obvious factors that may contribute 288

to the success of this community. The potential metabolism of toxic metals by these 289

diatoms must also be considered, and this merits further investigation. The apparent 290

15

active silica metabolism of Cyanidiaceae may provide considerable phylogenetic insight, 291

given the relationship between diatoms and red algae (Keeling et al. 2004). In terms of 292

biogeography, two of the diatoms, Pinnularia acoricola and Nitzschia cf. thermalis var. 293

minor, appear almost exclusively associated with hot-springs, whereas another, N. ovalis, 294

is halophilous, and a fourth, Eunotia exigua, is an acidobiont that inhabits acid-mine 295

drainages and naturally-acidic environments with equal facility. It remains a mystery how 296

such disjunct associations have emerged. Together, our observations provide strong 297

incentives to more thoroughly catalog the diatom floras of YNP springs. 298

299

Acknowledgments 300

We thank the National Park Service (Department of the Interior) for permission to 301

conduct research in Yellowstone National Park. We also thank Stefan Lalonde for 302

sampling and laboratory support. Drs. Paul Hamilton and Mark Edlund graciously 303

confirmed the lack of Nitzschia cf. thermalis var. minor in the herbariums of the 304

Canadian Museum of Nature and the Philadelphia Academy of Natural Sciences, 305

respectively. Research was supported by the Natural Sciences and Engineering Research 306

Council of Canada (NSERC) and the National Science Foundation (NSF Microbial 307

Observatory). This paper is dedicated to Dr. Eugene Stoermer for his enormous 308

contribution to diatom research. 309

310

16

References 310

ANDERSON, N.J. (2001): Diatoms, temperature and climatic change. – Eur. J. Phycol. 311

35: 307-314. 312

313

ASADA, R. & K. TAZAKI (2001): Silica biomineralization of unicellular microbes 314

under strongly acidic conditions. – Can. Mineralogist 39: 1-16. 315

316

BROCK, T.D. & M.L. BROCK (1966): Temperature optima for algal development in 317

Yellowstone and Iceland hot springs. – Nature 209: 733-734 318

319

BOURELLY, P. & E. MANGUIN, (1952) Algues d’eau douce de la Guadaloupe et 320

dépendences. – Centre National de la Recherche Scientifique: 1-281. Soc. d’edition 321

d’Enseignements Sup.,Paris. 322

323

CARTER, J.R. (1972): Some observations on the diatom Pinnularia acoricola Hustedt. – 324

Microscopy 32: 162-165 325

326

CASSIE, V. (1989): A contribution to the study of New Zealand diatoms. - Bibliotheca 327

Diatomologica, Band 17: 1-266. J Cramer, Berlin. 328

329

CASSIE, V. & R.C. COOPER (1989): Algae of New Zealand thermal areas. - 330

Bibliotheca Phycolologica Band 78: 1-261. J. Cramer, Berlin. 331

332

17

CHARLES, D. F. (1985): Relationships between surface sediment diatom assemblages 333

and lakewater characteristics in Adirondack lakes. - Ecology 66: 994–1011. 334

335

COMPÈRE, P. & A. DELMOTTE (1986): Diatoms in two hot springs in Zambia 336

(Central Africa). – In: F.E. ROUND. Proceedings of the Ninth International Diatom 337

Symposium: 29-40. Biopress Ltd., Bristol. 338

339

COPELAND, J.J. (1936): Yellowstone thermal Myxophyceae. – Ann. N.Y. Acad. Sci. 340

36: 1-232. 341

342

DENICOLA, D. (2000): A review of diatoms found in highly acidic environments. – 343

Hydrobiol. 433: 111-122. 344

345

EDMONDS, J.S. & K.A. FRANCESCONI (1998) Arsenic metabolism in aquatic 346

ecosystems. – In: W.J. LANGSTON & M.J. BEBIANNO (eds.) Metal Metabolism in 347

Aquatic Environments: 159-183. Chapman and Hall, London. 348

349

FOURNIER, R.O. (1985): The behaviour of silica in hydrothermal solutions. – In: B.R. 350

BERGER & P.M. BETHKE (eds.) Geology and Geochemistry of Epithermal Systems: 351

45-61. Reviews in Economic Geology Vol. 2. 352

353

HARGREAVES, J.W., E.J.H. LLOYD & B.A. WHITTON. (1975): Chemistry and 354

vegetation of highly acidic streams. – Freshwat. Biol. 5: 563-576. 355

18

356

HERNÁNDEZ-CHAVARRÍA, F. & A. SITTENFELD. (2006): Preliminary report on the 357

extreme endolithic microbial consortium of ‘Pailas Frías’, ‘Rincón de la Vieja’ 358

Volcano, Costa Rica. – Phycol. Res. 54: 104-107. 359

360

HUSTEDT, F. (1937-1938): Systematische und Untersuchungen über den 361

Diatomeenflora von Java, Bali und Sumatra. – Arch. Hydrobiol., Suppl. 15 & 16. 362

363

INSKEEP, W.P., R.E. MACUR, G. HARRISON, B.C. BOSTICK & S.FENDORF 364

(2004): Biomineralization of As(V)-hydrous ferric oxyhydroxide in microbial mats 365

of an acid-sulfate-chloride geothermal spring, Yellowstone National Park. – Geochim. 366

Cosmochim. Acta. 68: 3141-3155. 367

368

JORDAN, R.W. (2001): Taxonomy, morphology and distribution of two Pinnularia 369

species from acidic lakes and rivers in Yamagata and Miyagi Prefectures, Northeast 370

Japan. – In: R.JAHN, J.P. KOCIOLEK, A. WITKOWSKI & P.COMPÈRE. Studies on 371

Diatoms: Lange-Bertalot Frestschrift: 279-302. A.R.G. Gantner Verlag K.G. 372

373

JOYNT III, E.H. & A.P. WOLFE (2001): Paleoenvironmental inference models from 374

sediment diatom assemblages in Baffin Island lakes (Nunavut, Canada) and 375

reconstruction of summer water temperature. – Can. J. Fish Aquat. Sci. 58: 1222-1243. 376

377

19

JUGGINS, S (1992): Diatoms in the Thames Estuary, England: ecology, paleoecology, 378

and salinity transfer function. - Bibliotheca Diatomologica, Band 25: 1-216. J. Cramer, 379

Berlin. 380

381

KEELING, P. J., J.M. ARCHIBALD, N.M. FAST, & J.D. PALMER (2004): Comment 382

on ‘The evolution of modern eukaryotic phytoplankton’. Science 306: 2191b. 383

384

KONHAUSER, K.O., B. JONES, V.R. PHOENIX, G. FERRIS & R.W. RENAUT 385

(2004). The microbial role in hot spring silicification. Ambio 33: 552-558. 386

387

KRAMMER, K. & H. LANGE-BERTALOT (1986): Bacillariophyceae 1. Teil: 388

Bacillariophyceae. – In: H. ETTL, J. GERLOFF, H. HEYNIG & D. MOLLENHAUER 389

(eds). Süßwasserflora Von Mitteleuropa 2. Fischer, Stuttgart. 390

391

KRAMMER, K. & H. LANGE-BERTALOT (1988): Bacillariophyceae 2. Teil: 392

Bacillariaceae, Epithemiaceae, Surirellaceae. – In: H. ETTL, J. GERLOFF, H. HEYNIG 393

& D. MOLLENHAUER (eds). Süßwasserflora Von Mitteleuropa 2. Fischer, Stuttgart. 394

395

KRAMMER, K. & H. LANGE-BERTALOT (1991): Bacillariophyceae 3. Teil: 396

Centrales, Fragilariaceae, Eunotiaceae. – In: H. ETTL, J. GERLOFF, H. HEYNIG & D. 397

MOLLENHAUER (eds). Süßwasserflora Von Mitteleuropa 2. Fischer, Stuttgart. 398

399

20

LANGE-BERTALOT, H. (1978): Zur systematic, taxonomie und ökologie des 400

abwasserspezifish wichtigen formenkreises um “Nitzschia thermalis”. – Nova Hedwigia 401

30: 635-652. 402

403

LANGNER, H.W., C.R. JACKSON, T.R. MCDERMOTT & W.P. INSKEEP (2001): 404

Rapid oxidation of arsenite in a hot spring ecosystem, Yellowstone National Park. – 405

Environ. Sci. Technol. 35: 3302-3309. 406

407

MANN, J.E. & H.E. SCHLICHTING (1967): Benthic algae of selected thermal springs 408

in Yellowstone National Park. – Trans. Amer. Microsc. Soc. 86: 2-9. 409

410

PATRICK, R. (1977): Ecology of freshwater diatoms. – In:. Werner, D (ed).: The 411

Biology of Diatoms: 284-332. Botanical Monographs Vol.13. Blackwell, UK. 412

413

ROUND, F.E. (1991): Epilithic diatoms in acid water streams flowing into the resevoir 414

Llyn Brianne. - Diatom Res. 6: 137-145. 415

416

SAUNDERS, J.G. & G.F. RIEDEL (1998): Metal accumulation and impacts in 417

phytoplankton. – In: W.J. LANGSTON & M.J. BEBIANNO (eds). Metal Metabolism in 418

Aquatic Environments: 59-76. Chapman and Hall, London. 419

420

SAKS, N.M. (1982): Temperature, salinty and ultraviolet irradiation effects on the 421

growth of strains of Nitzschia ovalis. – Mar. Biol. 68: 175-179. 422

21

423

SCHOEMAN, F.R. & R.E.M. ARCHIBALD (1988): Taxonomic notes n the diatoms 424

(Bacillariophyceae) of the Gross Barmen thermal springs in South West Africs/Namibia. 425

– S. Afr. J. Bot. 54: 221-256. 426

427

SHERER, R.P. (1981): Freshwater diatom assemblages and ecology/palaeoecology of the 428

Okefenokee swamp/marsh complex, Southern Georgia, USA. - Diatom Res. 3: 129-157. 429

430

SIMONSEN, R. (1987): Atlas and catalogue of the diatom types of Friedrich Hustedt, 431

Vol: 2. - J. Cramer, Stuttgart. 432

433

STOCKNER, J.G. (1967): Observations of thermophilic algal communities in Mount 434

Rainier and Yellowstone National Parks. – Limnol. Ocean. 12: 13-17 435

436

SUZUKI, Y. & M. TAKAHASHI (1995): Growth responses of several diatom species 437

isolated from various environments to temperature. - J. Phycol. 31: 880-888. 438

439

VAN HEURCK, H. (1880-1885): Synopsis des diatomées de Belgique. – Ducaju & Cie., 440

Anvers. 441

442

VERB, R.G. & M.L. VIS (2000): Comparison of benthic diatom assemblages from 443

streams draining abandoned and reclaimed coal mines and nonimpacted sites. - J. N. Am. 444

Benth. Soc. 19: 274-288 445

22

446

VILLENEUVE, V. & R. PIENITZ (1998): Composition Diatomifère de quatre sources 447

thermals au Canada, en Islande et au Japon. - Diatom Res. 13: 149-175. 448

449

WATANABE, T. & K. ASAI (1995): Pinnularia acoricola Hustedt var. acoricola as an 450

environmental frontier species occurred in inorganic acid waters with 1.1 – 2.0 in pH 451

value. – Diatom 10: 9-11. 452

453

WEED, W.H. (1889): The diatom marshes and diatom beds of the Yellowstone National 454

Park. – Bot. Gaz. 14: 117-120. 455

456

WHITTON, B.A. & B.M. DIAZ (1981): Influence of environmental factors on 457

photosynthetic species composition in highly acidic waters. Verh. Internat. Verein. 458

Limnol. 21: 1459-1465. 459

460

WOLFE, A.P. (2002): Climate modulates the acidity of arctic lakes on millennial 461

timescales. Geology 30: 215-218. 462

23

Table 1: Water chemistry of Beowulf Hot Spring, Yellowstone National Park. All

concentrations are reported in µM, except where indicated.

Parameter Concentration

Na 11136 *Si 4690 K 1225 Ca 125 Al 124 As 19.6 Mg 8.2 Fe 6.2 *P 4.5 Zn 0.9 Mn 0.6 Cu 0.1 anions *Cl- 13770 *SO4

2- 1510 *F- 147 *NO3

- 20.0 other pH 2.26 Temp (oC) 67 Ionic strength (M) 0.02 * data from Inskeep et al. (2004)

24

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. and

similar species.

Nitzschia cf. thermalis var. minor a

N. hombergiensis b N. umbonata b N. steynii c N. rimosa d

Length (µm) 15-28 32-52 22-125 27-54.5 60 Width (µm) 2-4 5-6 6-9 4-5 7 Striae / 10µm 23-27 34-40 24-30 24-26 25-28

Fibulae / 10µm) 8-10 9-15 7-10 10-12 7-9 Fibulae notes 0.5-1 µm thick

on valve margin; irregularly spaced

Irregular & widely spaced; can extend to centre of valve

Short, narrow, regularly spaced; convergence of 2-3 straie

Prominent & evenly spaced

Robust

Central area constricted yes yes slight variable no

Habitat Thermal springs Circumneutral; unpolluted waters

Thermal springs; eutrophic waters; acid mine drainage

Thermal springs

Thermal springs

a. 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)

25

Figure captions 463 464 Fig. 1. A: Beowulf Spring, Norris Geyser Basin. Brown areas are hydrous ferric oxides 465

(HFO) and green areas represent cover by rhodophytes including Cyanidium. B: Close-up 466

of the transition between the HFO mat and Cyanidiaceae, with penny for scale. C: SEM 467

photograph of silica casts produced by coccoid Cyanidiaceae cells. D: Close-up of cells 468

in (C) showing remnants of soft-bodied cells within the silica cast (cc). E-H: SEM 469

micrographs of diatoms from the HFO mat. (E) Nitzschia cf. thermalis var. minor with 470

clean areolae devoid of precipitate. (F) Nitzschia ovalis (No) and Pinnularia acoricola 471

(Pa), also with clean areolae; the cast of a Cyanidiaceae cell is also visible below N. 472

ovalis. (G) A cell of N. ovalis encrusted in a diagenetic silica precipitate that completely 473

sheathes the cell and blocks areolae. (H) A Pinnularia acoricola cell with diagenetic 474

silica papillae on the apical cell surface; areolae are visible beneath. Scale bars are 10 475

µm, except (H) where it is 1 µm. 476

477

Fig. 2. Light micrographs of Beowulf Spring diatoms from cleaned samples (1000x 478

magnification). A-E: Eunotia exigua. (A-D) Valve views showing variable morphology. 479

(E) Girdle view showing terminal raphe nodules at the apices (arrows). F-H: Nitzschia cf. 480

thermalis var. minor. (F-G) Valve view; central constriction of valve margin and 481

interruption of raphe and fibulae; fine striae (H) girdle view. I-J: Nitzschia ovalis valve 482

view; dense fibulae and striae not resolvable under LM. K-M: Pinnularia acoricola (K-483

L) valve view. (M) girdle view. Scale bar is 10 µm. 484

485

26

Fig. 3. SEM photographs of diatoms from cleaned samples of Beowulf Spring. A-E: 486

Nitzschia cf. thermalis v. minor. (B) Shows the external view of the central nodule (cn) 487

with the eccentric raphe. (C) An internal view showing the fibulae on the valve margin 488

with the interior of central nodule (cn) visible. (D) Apiculate distal end of the valve 489

showing fibulae and distal raphe end terminating in a helictoglossa (h). (E) Apiculate end 490

with eccentric raphe F: Nitzschia ovalis cell. G-I: Pinnularia acoricola. (G) Hypovalve 491

view showing the strongly convergent striae at the poles and radial pattern in the central 492

area. (H) Girdle view (I) distal raphe end deflected; striae appear as groups of individual 493

puncta in costa-like areas. J-L: Eunotia exigua. (J) Valve view with the curved raphe 494

terminus on the valve face. (K) Raphe structure on the valve mantle with helictoglossa 495

evident. (L) Close-up of the internal raphe terminus showing the helictoglossa (h) and 496

rimoportulae (r). White scale bars are 1 µm, except A, G, H, J, & K where it is 10 µm. 497

B

A C

D

E

F G H

cc

Pa

No

Figure 1

a b c d e f f’ g g’ h h’

i i’ j j’ k k’ l l’ m m’

Figure 2

A B C

F

G H I

h

r

J

D E

LK

h

Figure 3

cn

cn