some RELATIONSHIPS BETWEEN PHYSIOLOGY, MORPHOLOGY, ' AND DISTRIBUTION III : ‘SULFUR-OXIDIZING BACTERIA Bisserfiatéon for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY DOUGLAS EDGAR CALDWELL ‘ 1974 ‘ .vfiwflfik my. HUAG & SUNS' _ 800K BINDER‘I‘ INC.‘ LIBRARY amosns "'1“... l . IICIIOII z:- ‘ L , BINDING BY ’ ' I'M .r" ABSTRACT SOME RELATIONSHIPS BETWEEN PHYSIOLOGY, MORPHOLOGY, AND DISTRIBUTION IN SULFUR-OXIDIZING BACTERIA BY Douglas Edgar Caldwell Twelve sulfide-containing aquatic environments were examined using phase microscopy, electron microscopy, gradient enrichments, plate counts and physicochemical determinations. In anaerobic hypolimnia, flagellate or gas-vacuolate photosynthetic bacteria were found. As light decreased and sulfide increased (with depth), there was an increase in the surface to volume ratio of the cells, a transition from internal to peripheral chromaphores, a decrease in the carotenoid content of the cells, an increase in bacteriochlorophyll d, and a decrease in oxygen tolerance. These changes were due to variations in community structure rather than the phenotypic adapta- tion of individual populations. The correlation of these transitions with gradients of sulfide, oxygen, and light can be explained on the basis of light requirements and sulfide and oxygen toxicity. .Bacterial populations occurred in sequential layers within hypolimnia. In Burke Lake the microstratification exhibited the following sequence (from the thermocline to the sediments): Thiospirillum, Chromatium, Thiocystis, and Chlorochramatium. In Douglas Edgar Caldwell Wintergreen Lake Thiopedia, Thiocystis, and CZathrochZoris occurred in sequential layers while Prosthecochloris was randomly distributed ‘within the layer of CZathrochZoris. The first series of layers was slightly turbid and red, contained purple sulfur bacteria and is referred to as community A. The next series of layers appeared green, contained green sulfur bacteria and is referred to as community B. The lowest layer occurred within 1 m of the sediment, was turbid and white, contained colorless bacteria, and is referred to as community C. The bacterial populations found within this layer have not been reported previously. The two ponds studied contained only community A. In littoral zones and in a sulfur spring, the sulfur-oxidizing bacteria were attached to surfaces either by a holdfast or by mucilage which encased microcolonies. Although both photosynthetic and heterotrophic bacteria were found, the chemoautotrophic thiobacilli were not. In situ, thiosulfate gradient enrichments selected fluorescent pseudomonads rather than thiobacilli and resulted in a three-fold increase in growth-rate as the thiosulfate concentration increased from 55 ug/l to 1 g/l S Isolate TBT - H and Pseudbmonas 203". aeruginosa were found to produce fluorescent sulfide-binding exudates which resulted in formation of globules of elemental sulfur from sulfide. This reaction did not consume oxygen. Previous reports on the fluorescent properties of P. aeruginosa exudates were confirmed and additional observations were made on the effect of pH on fluorescence. From a pH of 2.0 to 10.0 the exudate had an absorption peak at 405 nm. From a pH of 6.0 to 12.9 the Douglas Edgar Caldwell exudate had a fluorescence peak at 460 nm (excitation, 400 nm) which shifted to 430 nm at a pH of 1.0. Fluorescence at 460 nm was most intense-at a pH of 7.0. Lake samples (pH 7.0) were found to fluoresce at 440 nm although the intensity was nearly three orders of magnitude below that of the culture exudate. SOME RELATIONSHIPS BETWEEN PHYSIOLOGY, MORPHOLOGY, AND DISTRIBUTION IN SULFUR-OXIDIZING BACTERIA BY Douglas Edgar Caldwell A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Microbiology and Public Health 1974 "We should not be limited by what we believe is possible." Robert Ornstein ii ACKNOWLEDGEMENTS I am grateful to Drs. P. Hirsch, G. Sanders, and J. M. Tiedje, who have all profoundly affected my development as a graduate student. I thank Mary Firestone for assistance in counting the hypo- limnetic bacteria, Richard Strayer for oxygen analyses, Judy Lobato for help with the sulfide assays, Dr. M. J. Klug for the use of his epifluorescent photomicroscope, Dr. P. Filner for the use of his spectrofluorometer, Stuart Pankratz for assistance in electron microscopy, and Stjepko Golubic for pointing out the layers of Thiocystis in the soil surrounding the sulfur spring. I also thank Drs. G. Lauff, M. Klug, and R, Wetzel for making the facilities of the Kellogg Biological Station available. I am grateful to my wife, Sarah, for two of the electron micrographs, for help in the preparation of the thesis, and for her encouragement in times of failure and disappointment. I am also grateful for financial assistance from the Environ- mental ProtectioneAgency, Training Grant No. 5P3—2P—264, and the National Institutes of Health, Training Grant No. GM-01911-06. Special thanks is given to Dr. N. Pfennig and Dr. P. Hirsch for their comments on portions of Chapters IV and V and to my father, Edgar Caldwell, who manufactured the gradient plates and assisted in their design and development. iii TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . vi LIST OF FIGURES. . . . . . . . . . . . . . . . . . . . . . . . vii INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . 1 CHAPTER I. AN IN SITU GRADIENT ENRICHMENT METHOD FOR THE ISOLATION AND STUDY OF FREE-LIVING BACTERIA O O O O O O O C I O O O O O O O O O O O 3 BACKGROUND . . . . . . . . . . . . . . . . . . . 3 MATERIALS AND METHODS. . . . . . . . . . . . . . 5 RESULTS. . . . . . . . . . . . . . . . . . . . . 14 DISCUSSION . . . . . . . . . . . . . . . . . . . 20 LITERATURE CITED . . . . . . . . . . . . . . . . 21 CHAPTER II. ISOLATION OF A HETEROTROPHIC SULFUR-OXIDIZING BACTERIUM USING GRADIENT ENRICHMENTS . . . . . . 23 BACKGROUND . . . . . . . . . . . . . . . . . . . 23 MATERIALS AND METHODS. . . . . . . . . . . . . . 24 Gradient enrichment culture . . . . . . . 24 Media . . . . . . . . . . . . . . . . . . 25 Sulfide oxidizing activity. . . . . . . . 26 Habitats. . . . . . . . . . . . . . . . . 26 RESULTS. . . . . . . . . . . . . . . . . . . . . 27 DISCUSSION . . . . . . . . . . . . . . . . . . . 34 LITERATURE CITED . . . . . . . . . . . . . . . . 35 CHAPTER III. FLUORESCENT SULFURPOXIDIZING EXUDATES FROM PSEUDOMONAS AERUGINOSA O O O O O O I O O O C O O 36 BACKGROUND . . . . . . . . . . . . . . . . . . . 36 MATERIALS AND METHODS. . . . . . . . . . . . . . 37 RESULTS AND DISCUSSION . . . . . . . . . . . . . 39 LITERATURE CITED . . . . . . . . . . . . . . . . 49 iv CHAPTER IV. CHAPTER V. THE CELLULAR MORPHOLOGY OF SULFUR-OXIDIZING BACTERIAL COMMUNITIES. . . . . . . . . . . . BACKGROUND . . . . . . . . . . . . . . . . . . MATERIALS AND METHODS. . . . . . . . . . . . . Light microscopy. . Electron microscopy . . . . . . . . . . RESULTS AND DISCUSSION . . . . . . . . . . . Pelodictyon cLathnatifbrme. . . . . . . Prosthecochloris-AncaZochloris. . . . . Clathmchloris hypoZinmica. . . Chlorochromatium aggregatum . . . . . . Type 1 procaryotes (T1) . . . . . . . . Thiothrflx . . . . . . . . . . . . . . . Thiocystis, Chromatium, Thiospirillum, and Thiopedfia . . . . . . . . . . . . Type 2 procaryotes (T2) . . . . . . . Type 3 procaryotes (T3) . . . . . . . . Type 4 procaryotes (T4) . . . . . . . . Type 5 procaryotes (T5) . . . . . . . . Type 6 procaryotes (T6) . . . . . . . . LITERATURE CITED . . . . . . . . . . . . . . . THE DISTRIBUTION OF SULFUR-OXIDIZING BACTERIA. BACKGROUND . . . . . . . . . . . . . . . . MATERIALS AND METHODS. . . . . . . . . . . . . Media . . . . . . . . . . . . . . . . . Sampling. . . . . . . . . . . . . . . . Bacterial enumeration . . . . . . . . . Enumeration of morphological types. . . Chemical assays . . . . . . . . . . . . Habitats. . . . . . . . . . . . . . . . RESULTS AND DISCUSSION . . . . . . . . . . The distribution of sulfur-oxidizing bacteria in hypolimnia. . . . . . . . The distribution of sulfur-oxidizing bacteria in the littoral zone . . . The distribution of sulfur-oxidizing bacteria in a sulfur-spring . . . . . A comparison of the sulfur spring, lit- toral zone, and hypolimnion . . . . . CONCLUSIONS. . . . . . . . . . . . . . . . . . LITERATURE CITED . . . . . . . . . . . . . . . Page 51 51 52 52 53 54 55 58 61 61 65 70 71 78 78 81 84 84 87 91 91 92 92 92 93 94 95 96 96 96 114 116 116 120 122 LIST OF TABLES Table Page CHAPTER I 1 Fraction of saturating concentration . . . . . . 10 2 Microorganisms that have been successfully grown in the gradient plate. . . . . . . . . . . 17 CHAPTER II 1 Growth, extracellular pigment production, and extracellular deposition of elemental sulfur by eight isolates and eight known genera of bacteria . . . . . . . . . . . . . . . 33 CHAPTER V 1 Location and characteristics of the aquatic habitats studied . . . . . . . . . . . . . . . . 97 2 Concentration of HS-, So, thiobacilli, and total procaryotes in 13 water samples from lakes in southwestern Michigan. See text for explanation. . . . . . . . . . . . . . . . . . . 98 vi Figure LIST OF FIGURES Page CHAPTER I Illustration of the diffusion gradient plate showing the cover, window, diffusion plate, and base (from top to bottom). The path of the solu- tion with which the gradient is equilibrated (which form the two-dimensional gradient) through the agarose layer within the diffusion plate is shown by the thin arrows . . . . . . . . . . . . . . 7 (a) A diagrammatic representation of the distri- bution pattern of colonies formed by E. coli growing in a two-dimensional gradient of glucose vs. bicarbonate. Note that the bicarbonate gradient provides an optimal pH range and the substrate gradient also has an optimum. (b) The pattern of microcolonies formed by Hyphomicnobium, Rhodbmicrobium, and Thiapedia growing in a two- dimensional gradient of NaZS vs. CH3NH2-HC1 (Diagram). (c) Response of the planktonic microflora of a forest pond to a gradient of acetate vs. mud. The same plate is shown in both A and B, but printed with different light intensities. (Left), large colonies enhanced. (Right), small colonies enhanced. . . . . 15 Concentration of 14C-acetate as a function of distance for a one-dimensional gradient of 14C-acet ate 0 o o o o o o o o o o o o o o o o o o o o 19 95% confidence interval for the concentration of 14C-acetate (based on 6 duplicate experiments) as a function of distance for a one-dimensional gradient of 14C-acetate. . . . . . . . . . . . . . . l9 Concentration of 14C-acetate as a function of dis- tance in a two-dimensional gradient of acetate vs. 14C-acetate. Also shown is the concentration of 14C-acetate as a function of distance for a one- dimensional gradient of 14C-acetate on a 1n and linear scale . . . . . . . . . . . . . . . . . . . . l9 vii Figure Page Color photograph of a two-dimensional gradient of safranin vs. crystal violet . . . . . . . . . . . 19 CHAPTER II Oikography, the in situ response of a littoral bacterial community to a one-dimensional gradient of 3203:. Decomposing plant debris was homogenized and used to inoculate an in situ enrichment culture. The gradient cultures were equilibrated (ll/5/72 - 11/17/72) with Buttercup Spring (pH 6.9, 8 C). The number of cells in the microcolonies which formed (Figure 2) was used to determine the growth rate which was plotted versus the concentra- tion of thiosulfate. The range of thiosulfate con- centration resulting in the formation of elemental sulfur falls between the two dashed lines. The organism (TBT-H) was identified as a fluorescent pseudomonad. . . . . . . . . . . . . . . . . . . . . 29 CHAPTER III Emission spectrum of Pseudbmonas aeruginosa exudate and a lake sample at an excitation of 400 nm. The exudate was diluted with 3 volumes of buffer and scanned at a multiplier setting of 0.03 (460 nm peak). The lake sample was undiluted and scanned at a setting of 0.001 (440 nm peak). . . 41 The wavelength of the absorption peak (A ), fluorescence peak (I ), and the fluorescence intensity (0 ) vs. PH for Pseudomonas aeruginosa exudates . . . . . . . . . . . . . . . . . . . . . . 43 Volume change following addition of H5 to phosphate buffer (0 ) and to Pseudanonas aeruginosa exudate in the phosphate buffer (C ). Formation of 8° globules ( A ) by later incubation mixture is shown as change in 0.D.340 per min . . . . . . . . . . . . 46 Sulfide consumption.by Pseudbmonas aeruginosa exudate in phosphate buffer ( C ) and by buffered mediumalone(.).................48 CHAPTER IV Paladictyon clathratifbrme (bars represent 1 um) . . 57 Prosthecochloris (bars represent 1 pm) . . . . . . . 60 viii Figure 10 11 Page Clathrochloris hypolimnica (bars represent 1 pm) . . 63 Chlorochromatium aggregatum. . . . . . . . . . . . . 67 The Type 1 procaryotes (Tl) (bar represents 1 um). . 69 The microenvironment at the surface of decomposing plant material taken from the littoral zone of Burke Lake (electron micrograph of a thin section). A microcolony of Thiocystis fills the upper portion of the figure. The large inclusions (2,3) are sulfur granules. The microcolony is embedded in mucilage (1) which serves to anchor the colony to the plant wall (6). A transverse section through a Thiothrix filament shows the sheath (5) and peripheral membranes (4). Also seen are organisms with a scalloped cell wall and nucleophilic granules (7) which are characteristic of thiobacilli. Bar equals 1 pm. Micrograph by Sarah Caldwell . . . 73 Bacterial attachment to the surface of decomposing plant material taken from Burke Lake (electron micrograph of thin section). (A holdfast character- istic of the genus Thiothrix is shown (1). The debris at the base (2) generally contains iron hydroxide (soluble in dilute acid) and as a result is unusually electron dense. A bacterial micro- colony is also shown which attaches to the plant surface using the mucilage which encases the colony (4) and which contains centrally located membrane stacks (5). Nitrocystis also contains structures resembling carboxysomes (6) and is coccus shaped. Thiocystis is also present in the section (3). Micrograph by Sarah Caldwell . . ... . . . . . . . . 75 Thiqpedia and the Type 2 procaryotes (T2) (bar represents 1 um) . . . . . . . . . . . . . . . . . . 77 The Type 3 procaryotes (T3) (bar represents 1 um). . 80 The Type 4 procaryotes (T4) (bar represents 1 pm). . 83 The Type 5 and 6 procaryotes (T5 and T6) (bar represents 1 pm) . . . . . . . . . . . . . . . . 86 CHAPTER V Some physical, chemical and biological parameters of Wintergreen Lake on 7/4/71, and Burke Lake on 8/23/72 as a function of depth. The organisms shown are representatives of the A, B, and C com- munities which occur below the thermocline in sequential layers. . . . . . . . . . . . . . . . . . 101 ix Figure Page 2 The community structure of Wintergreen Lake on 7/4/71, and Burke Lake on 8/23/72. . . . . . . . . . 103 3 The fraction of the bacterial community in Wintergreen Lake composed of photosynthetic bacteria (A ), Thiopadia sp. (0 ) and Thiopedia sp. containing intracellular sulfur granules (O ) versus depth. Data for graph A were obtained on 7/4/71, and for graph B on 6/26/71 . . . 106 4 8.5 m, Community A. Cells exceed 2 m in diameter, have either gas vacuoles or flagella (but not both) and are purple sulfur bacteria with intra- cytOplasmic vesicles or membranes. Shown are Thiospimlllum (1), C'hromatiwn (2), and Thiocystis (3). . . . . . . . . . . . . . . . . . . . . . . . . 108 5 9.0 m, Community B. Cells are less than 2 pm in diameter, have either gas vacuoles or flagella (but not both) and are green sulfur bacteria with peripheral vesicles. Shown is Chlorochromatium aggregatum (4) . . . . . . . . . . . . . . . . . . . 110 6 9.5 m, Community C. Cells are less than 2 um in diameter, have gas vacuoles and are colorless bacteria. Shown are morphological types T3(5), T5(6), and T6(7) . . . . . . . . . . . . . . . . . . 110 7 3.0 m, Community A. Cells exceed 2 um in diameter, have either gas vacuoles or flagella (but not both) and are purple sulfur bacteria. Shown is Thiopedia (l) . . . . . . . . . . . . . . . 110 8 4.3 m, Community B. Cells are less than 2 pm in diameter, have either gas vacuoles or flagella (but not both) and are green sulfur bacteria. Shown is Chzthrochloris. . . . . . . . . . . . . . . 110 9 5.7 m, Community C. Cells are less than 2 pm in diameter, have gas vacuoles and are colorless bacteria. Shown are the morphological types T4(4), T5(3) and T6(2) . . . . . . . . . . . . . . . 110 10 The relative position of sulfur-oxidizing bacterial communities within the HS-, 02, and substrata x washout hypervolume (29). Each of the three ranges (1,2,3) shown represents a range of physicochemical conditions affecting bacterial competition. . . . . . . . . . . . . . . . . . . . . 119 INTRODUCTION The last review article on the ecology of the sulfur-oxidizing bacteria, "Studies on the Sulphur Bacteria", was written in 1925 by Bass-Backing. Since that time, very little has been added to our knowledge of the sulfur-oxidizing bacteria which participate in decomposition processes. Bass-Becking referred to this community of organisms as the "sulphuretum" community. He believed it was restricted to the surface of black muds, although it has since been found in anaerobic hypolimnia and sulfur springs. The difficulty in studying the ecology of the sulfur bacteria or any group of bacteria was long ago recognized by both Bass- Becking (1925) and van Niel (1955), who stated: "Pure culture of an organism gives us certainty about a great many facts. It fails, however, to account for the particular place of the organism in an inorganic cycle, inasmuch as its occurrence is influenced and restricted by competition." Baas-Becking "It is therefore clear that we cannot draw sound con- clusions concerning the natural role of microbes from experience gained with elective cultures unless we restrict our inferences to apply only to rigorously comparable conditions." van Niel The bacteriologist was thus limited to chemical determinations, light microscopy, and plate counts in his study of bacterial ecology, despite the advances that occurred in physiology, biochemistry, and l 2 molecular genetics. In the present study, two new tools, the electron microscope and diffusion gradients, were used to minimize the diffi- culties cited by van Niel and Baas-Becking. Chapters I and II concern the development and application of gradient methodology to enrich for sulfur-oxidizing bacteria in situ. Chapter III concerns laboratory studies of sulfide oxidation by Pseudbmonas aeruginosa, a fluorescent pseudomonad similar to those obtained using in Situ thiosulfate gradients. Chapter IV concerns the direct observation of sulfur-oxidizing communities and Chapter V correlates the results from the preceding chapters with distribu- tional information. CHAPTER I AN IN SITU GRADIENT ENRICHMENT METHOD FOR THE ISOLATION AND STUDY OF FREE-LIVING BACTERIA BACKGROUND Natural habitats often abound with unknown bacteria. The use of direct observation techniques has revealed a remarkable diversity of form, size, and arrangement (7,9,10,12). The fraction of unknown bacteria which have been isolated is small compared with the total observed. This may be due, in part, to a shortage of effective techniques for isolating and studying the ecology of organisms from natural habitats. The methods of Winogradsky (l6), Beijerinck (2), and van Niel (14) have strongly influenced the design of enrichment procedures. In general, an attempt is made to reproduce the natural microenviron— ment after observing and measuring as many parameters as possible. Success normally depends on the precision of environmental measure- ments, though in some cases even an accurate measurement is insufficient. For example, conditions measured at the peak of blooms are often suboptimal since optimal conditions precede the stationary phase of population growth. Because of the changing environmental conditions in enrichments, the growth of more tolerant bacteria may lead to the suppression of more fastidious forms. Therefore, one 4 usually obtains tolerant organisms, while isolation of those more susceptible to stress remains accidental. The observation of nutrient concentration gradients in natural habitats has led to suggestions that these be used for enrichment procedures. The Winogradsky columns (16), Perfil'ev capillaries (10), or the use of redox gradients created by decaying algae (8) have demonstrated the success of the gradient approach. The in situ gradient enrichment method described in this chapter was thus devised for isolating free-living bacteria and studying their response to changing environmental conditions. The concept of exposing organisms to a continuous gradient of test substances was originally used by Beijerinck (1). However, the extension of the gradient concept to more than one dimension, the replacement of time-dependent gradients with steady-state gradients, and the equili- bration of these gradients with natural physicochemical conditions has resulted in a new tool for ecological studies. Although many ecologists prefer to study the distribution and activity of organisms along natural gradients (3,13,15), the relationr ship between environmental variation and microbial response is difficult to demonstrate. An example of a natural gradient is the distribution of bacteria in the anaerobic zone of eutrophic lakes. Although the distribution of these bacteria has been quantitatively related to gradients of light quality, redox potential, H28 concentra- tion, and other factors (6,11), it has been impossible to determine the relative importance of each since all vary together as functions of a single dimension, in this case depth. 5 Artificial gradients allow the effects of environmental factors to be observed independently or in a limited number of combinations. The number of test variables that can be used is limited to the number of dimensions available. The gradient plate is a two- dimensional test surface in which a maximum of two artificial concentration gradients can be established while other environmental factors remain in equilibrium with in situ conditions. Thus, the effect of all combinations of concentration of two compounds on the test population can be evaluated. To do the same experiment with test tubes would require a forbidding number of tubes. MATERIALS AND METHODS The diffusion gradient plates used for laboratory enrichment studies were originally developed by Caldwell and Hirsch (4). These were later modified (5) and used for quantitative and in situ enrichment studies. The procedure below applies to the modified diffusion gradient plate (DGP-300, Uni-Lab Products, 419 E. La Selle, Royal Oak, MI 48073). The gradient plate consisted of four units, the cover, window, diffusion plate, and based These were assembled as shown in Figure l and held in place by six stainless steel bolts. The diffusion plate contained a square depression 25 mm on each side as well as two rectangular slots. After the slots were covered with masking tape, both the slots and the square depression were filled with molten agarose (1.0%, 60-100 C) using a 5 m1 hypodermic syringe. The amount of agar in the square depression was adjusted until the surface of the agar (viewed by reflection of a light source across Figure 1. Illustration of the diffusion gradient plate showing the cover, window, diffusion plate, and base (from top to bottom). The path of the solution with which the gradient is equilibrated is shown by the dark arrows. The path of the two test substances (which form the two-dimensional gradient) through the agarose layer within the diffusion plate is shown by the thin arrows. Figure l 8 the room) was slightly depressed (allowing space for the addition of the inoculum.later). At all times during this procedure the agarose remained molten so that the agar in the slots and square depression was united by a continuous agar bridge through the slits on two sides of the square depression. Bubbles of air sometimes formed and these were removed using a syringe. The diffusion plate was passed over a burner to warm the plastic before and during this procedure. After the dif- fusion plate was filled with agar and all air bubbles were removed, the agar was allowed to cool. Two saturated solutions of the test substances were added to the base plate. Crystals of the test substance were then added in addi- tion to the saturated solutions, the diffusion gradient plate was assembled, inverted, and a flow of 10 ml/h was begun. The flow'was across the plate toward the two slits which supply the diffusion plate with the test substances. After 24-48 h a steady-state gradient was established. The plate was then dismantled and the inoculum added in agarose (1.02, 40 C). The depression was either filled until the surface of the agarose was even with the plastic surface (as viewed by reflected light) or filled to excess and a glass slide used to shape the agar. The diffusion plate was then returned to the gradient plate assembly and incubated. All components of the gradient plate were sterilized under 15 psi of steam for 20 minutes except the diffusion plate, which was boiled, exposed to ultraviolet light or treated with ethylene oxide. 9 The most common problem in using the gradient plate was the rupture of the agarose bridge between the plate and the reservoir. This may have been caused by a pressure difference between the reservoirs and diffusion plate or by improperly pouring the agarose initially. The problem was avoided by using a short out-flow tube maintained at the same height as the gradient assembly. The equili- bration solution was pumped in and allowed to flow out. Concentrations were determined directly from Table l, which summarizes the numerical solution of the gradient (5). The table has an estimated error of less than 5%; however, variations in the thickness of the plastics used to construct the plates result in an error of;i 502. The values listed in the table represent the fraction of the saturating concentration of the test substances at various locations in the diffusion plate. The fraction found in the table is multiplied by the solubility of the test substance and added to the concentration of the test substance in the equilibration solution to obtain the concentration (g/liter) at any location in the diffusion plate. The first two digits of the table value represent the coefficient. The remaining digits represent the negative exponent. The table is valid only if the concentration of test substance in the equilibration solution is small (less than 0.1%) compared to the saturating concentration and if the flow rate is several orders of magnitude above the diffusion rate (10 m1/h). For a table value of 312, a solubility of 100 g/liter and a concentration of 1.0 x 10-6 3] liter in the equilibration solution, the concentration would be 3.1 x 10-2 multiplied by 100 g/liter plus 1.0 x 10-6 or 3.1 g/liter. 10 new 50H «ma 0H0 000 05¢ 00m mam 00H and 0.N new 50H mmH 0H0 000 one man waw 00H and 0.H new RAH NNH 000 000 00¢ 0mm wqu 0nd mNH 0.H nmu 50H NNH 000 000 000 mam 0NN 00H QNH ¢.H NAN mmH “Ha was 0mm 00¢ wwN mom me mad N.H an find 000 050 00¢ 0mm 0mm 00H 00H 0mm 0.H nma NHH 005 00m 000 mam mHN mmH man 005 0.0 NNH new 0H0 00¢ mam mNN 00H mNH 000 000 0.0 000 00m 0H0 00m mam me wad 000 mum 0H0 «.0 0H0 waw mam wmfi wHH 0mm 00m 000 00a mHN N.0 I win En 5.0 «.0 0.0 06 N.m 0.0 «.0 0.0 Ev summon Aufio>uomou ouwmommo madam dogmnmwfiv mo owvo Baum afiv moamumwn 0H0 000 05¢ mom 0mm away 00H mNH 00H 00H 0.N 0H0 000 050 0cm 0mm 00H 00H mNH 00H 00H 0.H 000 000 000 0mm 00m 00H 00H mad 00H 0H00 0.H mqm 000 00¢ mum 0mm 05H mma mHH 0Hmm 0Hmm ¢.H 00m 0mm 00¢ mam oHN 00H 0NH 0Hmm 0H50 0Hm0 N.H 0N0 00¢ 0mm 0mm 00H 00H mad 0H~0 odes odes 0.H 00m 00¢ 0mm mHN 00H 0NH 0H0w 0Hmn 0H¢0 0H~0 0.0 mmq mam 0mm 00H mNH 0H00 0H00 0Hhm 0H0m 0Hme 0.0 00m mHN mmH mHH 0H~w 0HHO 0Hm¢ 0Hmm 0H¢m 0Hmm «.0 mmH mad 0me ounm came 0HHm 0He~ 0HOH 0HNH 0HNH N.0 0.m N.m w.~ EN 0.N min NJ” 0.0 «.0 0.0 ABS Eamon Auwo>uomou ouamoano madam noqmsm0fiv no 0000 aoum aav ooamuqu sowuouuaooooo waHumHOuom mo defiuomum A «Home 11 emu emu emu eue eee eem eeu eeu emu emu e.u emu emu emu eue mes eee eeu emu emu ems e.u emu emu mum eee eue men eeu euu euu eoa e.u euu euu eee eue eee mum . emu eeu euu eee e.u emu sud eeu eee mos euu euu eeu euu eeu u.u emu eeu eee eoe eem eeu eeu emu eee eee o.u euu nee eee mus eon euu emu euu emu . eue e.o eee eue eee mum emu euu euu eee . eue eee e.o nae eue eon euu eeu euu eue eee eue eon e.o mom euu emu euu eoe eue eue eon euu eeu u.o e.eu u.eu u.eH m.eu a.mu a.mu a.mu u.uu , m.uu .e.uu ...Aaae eeuwm AHHO>HOm0H mafimommo Gum—HQ dOHmdmmHv. NO ONE EOHM ~55 fluflwumHn eue eee eem eeu emu emu uum uee.. , uee use e.u eue eee eem eeu emu emu uum uee uee use e.u eee eue eem eeu euu euu use use uee umn e.u. eue eee eum emu eeu euu uee use . ,uee uue e.u eee eoe emu euu eeu euu uuu uee use ueu u.u ems eem eeu eeu emu use uee use uem ueu o.u eue eon euu emu euu ueu uue use ueu uuu e.o eum emu eeu euu uee uue use uum umu ueu e.o euu eeu euu uue uee uue use uuu ueus . uuu e.o euu uau uue uue umu uuu ueu uuu .emu eee u.o e.uu u.uu u.ou m.eu e.e e.e u.m u.e m.e e.u uaae enema AuHoPuomou ouamomao madam aonomevMo owoosowu 880 Snowman Aeosaueaooe u mueeu 12 ama u0m um0 ume u0m u0u uma uma mmm m00 0.u ama u0m um0 ume u0m u0u uma uma mmm mm0 0.a ama umm ~50 uwe uem umu uma uma mum m00 0.a aua umw um0 ume uum umu una uua mum mu0 e.a aaa ~00 umm uae u0m uau una uaa mow mum u.a unm ~05 u0m u0m u0u uma uea mum mom mom 0.a uom umm uue u0m uuu u0a uua mam mmm mue 0.0 uu0 ume uum umu uua uua mum mm0 mme mum 0.0 uue u0m uau u0a uua mmm m00 mme mam muu e.0 uau uma uaa mam mmm mue mom muu m0a -,maa u.0 a.mu oau e.uu u.uu eéu eéu eéu e.ou - u.ou u.uu -Aaae flame Auao>homou ouamoamo woman noamsmuav mo ammo Scum sav moamuman mme mmm mmu mma mma eem e00 eme emm emu 0.u mme mmm mmu mma mma eea e00 eme emm emu 0.a mme mem mmu mma mma eam e00 ewe eem eeu 0.a mme mum mmu mua mua e00 eu0 eme eNm emu e.a mae mom mau mma maa e0u enm eae emu eau u.a m0m m0u mma mma e00 em0 e00 e0m e0u ema 0.a m0m muu m0a maa eam e00 eue e0m euu ema 0.0 mmu mna mua e50 eu0 eme eum emu eua eua 0.0 m0a maa eum emn eme eam euu e0a .eaa mum e.0 eam emm eue e0m euu e0a eaa mow mum IIII mae u.0 e.0a 0.ma 0.0a m.ma 0.5a 0.5a a.na n.0a m.0a . . 0.0a . A530 sumac Auao>uomou muamomno woman Goamswmav moqmwuo sown aav monouman Aconcauaoov a mamas 13 cos sse sen suu . ses o.u cos sse sum suu ses e.s cos see sum seu ses e.s sme see see seu sus e.s see sum smu ssu . ses u.s ssm see seu ses sms e.s seu smu . ses ses , sss e.o sus ses ses sss use e.o sss sss umm -.ueu uee e.o uee use uee uum I- ueu u.o e.eu o.eu e.eu u.eu . e.mu . .Aaae semen Asao>somou ouamommo woman aoamawwa0.mo mmuo aouw say uodduman. Avoanauaoov a oanmw 14 RESULTS To demonstrate a differential growth pattern, E. coli Kr12 was grown in a glucose vs. bicarbonate gradient (Figure 2a). Large colonies appeared where growth conditions were most favorable, while other cells of the inoculum remained static throughout the rest of the plate. Thus, there was an optimum growth range for both the pH and substrate. The inoculum (10 cells/ml) was purposely kept small to avoid cross-feeding effects between colonies. A mixed community consisting of Hyphomicrobium sp. B-522, Thiopedfia sp., and Rhodbmicrobium vannielii was also used as an inoculum, with gradients of NaZS and CH3NH2°HC1. Continuous illumi- nation at an intensity of 30 ft-c was provided for the photosynthetic organisms. Each organism formed a characteristic pattern of colonies despite the random distribution of the inoculum (Figure 2b). Resting cells of R. vannielii were almost spherical, while those observed on the edges of large, fast-growing colonies were spindle shaped. These morphological changes were also observed in pure cultures. The effect of two factors on the natural surface community of a forest pond (Michigan State University Experimental Forest, Augusta, Michigan) was observed qualitatively. Pond surface water served as the inoculum while pond mud supplied naturally occurring growth factors. Acetate and pond mud were placed in the troughs. The plate was illuminated with 12 ft-c at a temperature of 24 C and the other parameters were as described above. After 81 h, many different colony types had developed and were densely distributed over the plate. While the large colonies consisted of bacteria with terminal 15 HYPHOMICROBIUM - ’ ‘ GLUCOSE -.' ' . .\ NH2CH3 > > THIOPEDIA i/Ly: A" _. /. . Q...\ ‘ A ‘nHoooumnosmu u."c°3 ".23 diagrammatic representation of the distribu- tion pattern of colonies formed by E. calI' owing in a The pattern of microcolonies formed by III-pha- two-dimensiona Ionalgndiemofglucosevs .bicar nate. Note "III-,rubium Rhodomitrahium. and T/Iltlpttliu growtng In a that the bicarbonate gradient provide: an optimal pH wo-dimensional gradient of Na; 5 \5 .HC HINH~ H-C.l range and the substrate gradient also has an optimum. (D Response of the planktonic microflora of a forest tpond to a gradient of acetate VS. mud. Th same Mplaten shown in both A and B but printed with different light intensities. (Left). large colonies en- hanced.(llight). small colonies enhanced. Figure 2 16 endospores, many of the smaller ones were isolated and identified as Bacillus sp. The background of small colonies could be eliminated by selecting a long exposure time and a less sensitive paper during the photographic printing process (Figure 2c [left]). Smaller colonies (Figure 2c [right]) appeared mainly in the area of high mud extract and high acetate concentration. The biological utility of gradient plates depends on the successful cultivation of naturally occurring microorganisms. Table 2 shows a variety of microorganisms that have been successfully cultivated using two-dimensional gradients. The cultivation of Haematococcus, which has delicate cell walls, suggests that many naturally occurring microorganisms may be able to tolerate the physical stress of being heated and embedded in agarose. Empirical data were collected by sampling gradient plates which contained 14C-acetate by using 100 capillaries to withdraw plugs from the agarose layer (Figure 3). These empirical data were summarized in the form of a regression equation relating concentration and the distance from the reservoir containing the test compound (r2-0.997) (Figure 4). A theoretical model was developed by Dr. Sung Ho Lai (5) to predict the gradient pattern at steadyhstate for any test compound used in the plate. The theoretical equation fell.within the 95 percent confidence interval of the empirical regression equation. When a gradient of nonradioactive phosphate or acetate was imposed at a right angle to a gradient of radioactive acetate, the empirical results remained within the 95 percent confidence interval 17 Table 2 Microorganisms that have been successfully grown in the gradient plate Test compounds lst dimension 2nd dimension Incubation site Bacteria: _ Eacherichia sp. acetate .a. laboratory Thiopedfia sp. CHBNH2°HC1 Nazs laboratory Hyphomicrobium sp. CHBNHz-HCl NaZS laboratory Rhodbmicrobium sp. CH3NH20HC1 NaZS laboratory Thiothrix sp. Na2C03 NaZS spring Algae: Padiastrum sp. NaHCO3 KZHPO4 laboratory Sbenedasmus sp. NaHCO3 KZHPO4 laboratory Haematococcus sp . NaHCO3 KZHP04 laboratory Cbsmarium sp. NaHCO3 KZHPO4 laboratory Fungus: Alternaria sp. KZHPO4 NH4N03 spring of the regression expression for acetate alone (Figure 5). This showed that gradients can be perpendicularly superimposed without affecting their individual gradient patterns (Figure 6). Thus, it is possible to calculate the concentration of test sub- stance at any location in the agarose layer. All that is required is a knowledge of the physical dimensions of the plate, the solubility, and the in situ concentration of the test substance(s). It should also be noted that the gradient pattern is not a function of the coefficient of diffusion for the test substance, the nature of the diffusion medium, or the flow rate. However, the flow 18 Figure 3. Concentration of 14C—acetate as a function of distance for a one-dimensional gradient of 14C-acetate. Figure 4. 95% confidence interval for the concentration of l4C-acetate (based on 6 duplicate experiments) as a function of distance for a one-dimensional gradient of 14C-acetate. Figure 5. Concentration of 14C-acetate as a function of dis- tance in a two-dimensional gradient 2f acetate vs. 14C-acetate. Also shown is the concentration of 1 C-acetate as a function of distance for a one-dimensional gradient of l4C-acetate on a 1n and linear scale. Figure 6. Color photograph of a two-dimensional gradient of safranin vs. crystal violet. l9 um ° . ‘3'" \ Isa-teraiaaasev ”mm-mun; Figures 3, 4, 5 and 6 20 rate must exceed the diffusion rate by several orders of magnitude, and the concentration of test compounds in the flowing solution must be less than 0.1% of their solubility. Although the time required to reach steady-state does depend on the diffusion coefficient, a gradient of most monomolecular substances reaches steady state in less than two days. DISCUSSION The selection of a two-dimensional plate rather than a three- dimensional cube (capable of including three variables at once) was made becauseof the difficulty of equilibrating a cube with in situ physicochemical conditions. A thin plate allows the rapid equilibra- tion of steady-state gradients with aquatic environments. The use of multi-dimensional gradients results in a logarithmic increase in the amount of information that can be obtained in a single experiment. Interaction effects, which occur if the response to one environmental variable is a function of another variable, are revealed when microorganisms are exposed to a two-dimensional gradient. The steady-state nature of the gradients allows the study and culti- vation of organisms which compete most effectively when low concen- trations of substrate are continually resupplied. Steady-state conditions also permit the use of longer incubation periods increasing the differential response of slow-growing microorganisms to test substances. There are several limitations to the method. Even using arti- ficial gradients, it is difficult to restrict the test variables to one per dimension due to the nature of most test substances. For 21 example, if NaZS is used as a gradient it will be accompanied byr‘ gradients of pH, osmolarity, and redox. This effect may be discounted in many cases since each of these factors would normally be determined by sulfide concentration in natural environments just as in the artificial gradient plates. There can be exceptions and, for that reason, one should always consider this limitation. In some cases the test compound will be precipitated by some chemical factor in the environment. In this case the precipitating compound can be used in the reservoir in place of the original test substance. The fact that the original substance precipitates in the environment eliminates it as a factor at concentrations above its solubility. It is obvious that gradients cannot be extended below the environmental concentration for in situ studies (5). For this reason, it is only possible to use increases in concentration as test stimuli. LITERATURE CITED 1. Beijerinck, M. 1889. Auxanography, a method useful in micro- biological research, involving diffusion in gelatin. Archives Neerlandaises des Sciences Exactes et Naturelles, Haarlem 23:367-372. 2. Beijerinck, M. W. 1901. Uber oligonitrophile Mikroben Zentralbl. Bakteriol. Parasitenk. Infektionskr. Hyg. Abt. 7(2): 561-582. 3. Brock, T. D. 1967. Relationship between primary productivity and standing crop along a hot spring thermal gradient. Ecology 48:566-571. 4. Caldwell, D. E. and Hirsch, P. 1973. Growth of microorganisms in two-dimensional steady-state diffusion gradients. Can. J. Microbiol. 19:53-58. 10. ll. 12. 14. 15. 16. 22 Caldwell, D. E., S. H. Lai, and J. M. Tiedje. 1973. A two- dimensional steady-state diffusion gradient for ecological studies. In Modern Methods in the Study of Microbial Ecology. Edited by T. Rosswall. Bull. Ecol. Res. Comm. (Stockholm). Vol. 17 pp. 151-158. Genovese, S. 1963. The distribution of H28 in the Lake of Faro (Messing) with particular regard to the presence of "red water". 123 Marine Microbiology. Edited by C. H. Oppenheimer. Charles C. Thomas, Springfield, IL pp. 194-204. Hirsch, P., and Pankratz, H. S. 1971. Study of bacterial populations in natural environments by use of submerged electron microscope grids. Z. Allg. Mikrobiol. 10:589-605. La Riviera, J. W. M. 1965. Enrichment of colorless sulfur bacteria. Symp. G6ttingen 1964; Anreicherungskultur und Mutanten-Auslese. Zentralbl. Bakteriol. Parasitenk. Infektionskr. Hyg. Abt. Supplementh. l(l):l7-27. Nikitin, D. J., and Kuznetsov, S. J. 1967. Electron micro- scope study of the microflora of water. Mikrobiologiya 36:938-941. Perfil'ev, B. V., and Gabe, D. R. 1961. Capillary methods of investigating microorganisms. Izd. Akad. Nauk SSSR. Transl. Oliver and Boyd. 1969 pp. 321-331. Sorokin, Y. I. 1968. Primary production and microbiological processes in Lake Gek-Gel. Mikrobiologiya 37:345-354 (in Russian). Staley, J. T. 1968. Prosthecomicrobium and Ancalomicrobium, new prosthecate fresh water bacteria. J. Bacteriol. 95:1921-1942. Terborgh, J. 1971. Distribution on environmental gradients: Theory and a preliminary interpretation of distributional patterns in the avifauna of the cordillera vicabamba, Peru. Ecol. 52:23-40. van Niel, C. G. 1931. On the morphology and physiology of the purple and green sulphur bacteria. Arch. Mikrobiol. 3:1-112. Walker, B. H. 1971. Relationships between derived vegetational gradients and measured environmental variables in Saskatchewan wetlands. Ecol. 52:85-95. Winogradsky, S. 1888. Beitrage zur Morphologie und Physiologie der Bakterien. Pflanzenforschung I:l-120. CHAPTER II ISOLATION OF A HETEROTROPHIC SULFUR-OXIDIZING BACTERIUM USING GRADIENT ENRICHMENTS BACKGROUND Sulfide, one of the products of bacterial decomposition, is slowly oxidized inorganically by 02 to So, 8203-, or $04- depending on the reaction conditions (5). Chen and Morris (5) identified a pH dependent induction period for HS- oxidation of from 0.2 to 6.0 h before abiologic 0 uptake begins. Thus HS- and 02 can apparently 2 exist simultaneously. In the littoral zone, HS- and 0 may be supplied 2 together by decomposition processes and wave action. Thus reduced sulfur compounds can be available for aerobic bacterial oxidation despite simultaneous inorganic oxidation. There are several groups of bacteria which might be involved in biologic HS- oxidation. These include the Chromatiaceae, Chloro- biaceae, colorless sulfur bacteria, and heterotrophic sulfur- oxidizing bacteria. Our knowledge of the role of these bacteria in the transformation of inorganic sulfur compounds in nature is based primarily upon selective enrichment as described by van Niel (9) and also upon the subsequent study of pure cultures. However, the elegance and convenience of this approach have sometimes overshadowed the fact that minor environmental variations between the enrichment (elective) culture and the natural habitat result in a completely 23 24 different bacterial flora. As stated by van Niel: "It is therefore clear that we cannot draw sound con- clusions concerning the natural role of microbes from the experience gained with elective cultures unless we restrict our inferences to apply only to rigorously comparable conditions." For this reason, an in situ enrichment method (3,4) was used to investigate sulfur transformations which occur in the littoral zone of the lake described here. This procedure insured that the conditions of enrichment were comparable to natural conditions. Direct observations of decomposing plant surfaces also served to confirm the conclusion that the organisms studied using gradients were actually members of the natural community of sulfur-oxidizing bacteria. MATERIALS AND METHODS Gradient enrichment culture Gradient enrichment culture was carried out as described by Caldwell et al. (4) using diffusion gradient plates (Uni-Lab Products, 419 E. La Salle, Royal Oak, MI 48073 U.S.A.). The growth rate was determined from the number of cells per colony and the incubation period. Concentrations were determined using the numerical solution of the diffusion gradient. The inoculum was prepared by slicing decomposing plant materials (washed in sterile distilled water) into amall sections and grinding with a ground glass homOgenizer. The plant materials were taken from the littoral zone of Burke Lake and included decomposing aspen leaves and grasses. These were observed using phase and electron microscopy. They contained Thiothrlx and 25 Thiocystis (Chapter IV, Figures 6 and 7) as well as unidentified rod- shaped organisms. Precautions to preserve sterility were taken throughout this procedure. The diffusion gradient plate was incu- bated with the base up on a plastic coated test tube rack. The cover was removed to expose the agarose directly to the water currents. uses Thiobacillus Broth (Difco Laboratories, Detroit, MI U.S.A.) was used with 1.52 Nobel Agar (Difco) for plate counts of thiobacilli (pH 4.8). To determine the presence or absence of thiobacilli in 82038 enrichments the S, R, and F thiobacillus media recommended by Aaronson (l) were used in addition to the ThiObacillus Broth. The T8 (thiosulfate) medium for maximum pigment production of fluorescent pseudomonads is a modification of the media of Frank and DeMoss (6) and Sands and Rovira (8). It contained: 22 glycerol (w/v), 12 Na 0 -7 H O, 1% (w/v) case amino acids, 0.01 M 2323 2 K HPO 0.01 M.MgC12, and 1.5% Nobel Agar (for solid media). In 2 4’ some cases antibiotics were added, including: 45 mg/liter Penicillin G, 45 mg/liter novobiocin (Upjohn, Albamycin), 75 mg/liter cyclo- heximide (Upjohn, Acti—Dione), and 5 mg/l chloramphenicol. Fluorescent pseudomonads were detected by laying decomposing aspen leaves on the TS (thiosulfate) agar medium supplemented with antibiotics as suggested by Sands and Rovira (8). This resulted in the formation of microscopic fluorescent colonies in 24 to 48 h. The number of colonies per unit area of leaf surface was determined by counting fluorescent colonies using epifluorescence microscopy. By homogenizing a known leaf area and diluting the homogenate, it 26 was possible to estimate the number of viable cells per colony. Sulfide oxidizing activity A substrate solution of 0.5 M Na28-9 H20 in 0.5 M KZHPO4 was adjusted to pH 7.6 using HCl. This was added to filtered (0.45 um filter, Millipore Corp., Bedford, MA U.S.A.) bacterial exudates to a final concentration of 0.1 mM HS-. In active samples, a colloidal precipitate of elemental sulfur formed in 0.1 to 5 min. The pH remained at 7.6 during the reaction. Sulfur was determined after collecting the So precipitate on a 0.45 um Millipore solvinert filter, washing five times with distilled water, drying, and dissolv- ing overnight in petroleum ether on~a wrist action shaker. Sulfur was then determined using the procedure of Bartlett and Skoog (2). The presence or absence of So forming activity was determined fdr 8 isolates and 8 known genera: Pseudbmonas aeruginosa, Micrococcus sp., Proteus vulgaris, Eacherichia coli, sarcina lutea, Corynebacterium sp., Bacillus megatarium, and Enterobacter aerogenes. Each organism was streaked on solid TS Media containing 0, l, 5, 10, and 202 8203-7 H20. After a 48 h incubation period a crystal of Na28°9 H20 was placed near the streak. A precipitate of SO in the form of a white ring around the crystal was considered a positive test for So forming activity. Habitats Burke Lake is a spring fed lake 9.7 m in depth. Some of the physical, chemical and biological characteristics have been described 27 in Chapter IV (Table l). The gradients were equilibrated in Buttercup Spring (pH 7.5) which feeds Burke Lake. RESULTS 3-bicarbonate were used to enrich for sulfur-oxidizing bacteria found on decompos- Gradients of HS", 3203:, and HS- versus NaHCO ing plant surfaces. Over a period of several weeks, no colony development occurred in HS- gradients because a constant supply of HS- could not be maintained due to the formation of a precipitate at higher HS_ concentrations. However, the one-dimensional 8203= gradient (Figure l) permitted the formation of So encrusted colonies of rod-shaped bacterial cells. Thiothrfir also formed colonies within the agar and on the agar surface. The gradient conditions under which Thiothrix deposited sulfur intracellularly coincided with the conditions resulting in sulfur deposition by the rod-shaped organisms. This range was 0.067 to 13.5 g 8203. per liter at a temperature of 8 C. The maximum growth rate of 19 h per generation was at a $203- concentration of 1-5 g/liter (Figure l), which also corresponded to the maximum concentration for So deposition. The lowest growth rate occurred at 55 ug/liter, the lowest concentration tested, and no So was deposited. A S0 encrusted colony of rod-shaped organisms is shown in Figure 2. The S0 was not deposited intercellularly or intracellularly but contiguous to the colony surface and in the agarose adjacent to the colony. The rod-shaped, So depositing organism was subcultured every two weeks in situ within S 03- gradients for a period of six weeks. 2 This resulted in a lawn of microcolonies. From this enrichment the I 28 Figure l. Oikography, the in situ response of a littoral bacterial community to a one-dimensional gradient of 8203'. Decomposing plant debris was homogenized and used to inoculate an in situ enrichment culture. The gradient cultures were equili- brated (ll/5/72 - ll/l7/72) with Buttercup Spring (pH 6.9, 8 C). The number of cells in the microcolonies which formed (Figure 2) was used to determine the growth rate which was plotted versus the concentration of thiosulfate. The range of thiosulfate concen- tration resulting in the formation of elemental sulfur falls between the two dashed lines. The organism (TBT-H) was identi- fied as a fluorescent pseudomonad. nouns PER Geuenmou x to" 29 66 III generation I l I l 1-59Illtor. / I l l 19 Moon-ration l I ll 11 o s 10 ggcouceurnartou (In Ina/III») Figure l 30 Figure 2. A S0 encrusted colony from the gradient described in Figure 1. a. Colony of fluorescent pseudomonads in the 8203’ gradient (2). Elemental sulfur was deposited at the colony surface (1) and in the agar surrounding the colony (3). Bar represents 100 um. b. Higher magnification of the colony. The crust of 8° globules (4) encasing the cells (5) can be seen. Bar represents 10 um (phase micrograph). c. The colony shown above has been crushed to show that the refractile crust encasing the cells is composed of indi- vidual S° globules. The characteristic diffraction pattern of 8° globules can also be seen (6). Bar represents 10 um (phase micrograph). 31 Figure 2 32 rod-shaped organism was isolated in pure culture using the TS medium. The organism was identified as a fluorescent pseudomonad. The isolate, TBT-H, utilized glucose, lactose, mannitol, and citrate as carbon sources. It was nonfermentative, Gram negative, motile, catalase positive, microaerophilic, and hydrolized gelatin. Sulfide was not produced and no growth occurred on thiobacillus media. The organism excreted a number of fluorescent pigments when grown in the TS medium and in the TS medium when 803. or 804- were substituted for $203-. No pigments were produced when inorganic sulfur compounds were deleted from the medium although growth still occurred. When the isolate was grown in liquid TS medium (pH 7.6), the addition of HS- to cell free exudates (1 mM) resulted in the formation of 0.49 i 0.13 mM (95% confidence interval) of elemental sulfur (final pH 7.6). No S0 was formed when the HS- was added to uninoculated medium or to boiled (20 min) exudates. As an additional control, eight heterotrophic organisms isolated from the littoral zone, which did not produce extracellular fluorescent pigments, together with eight pure cultures of known genera were tested for So forming activity (Table 1). Only the exudates from the fluorescent isolate and Pseudbmonas aeruginosa (both of which produced fluorescent pigments under these conditions) were active. Both of these organisms failed to produce either the pigment or activity on plates which lacked $203.. No S0 was formed on uninoculated control plates containing only the TS medium. All attempts to isolate Thiothrix in the TS medium, from either natural samples or gradient enrichments, yielded pure cultures of 33 I I I I I I I I I I + + + + + moxumouoo suuooeosoeam I I I I I I I I I + + + + Esmsouomos wxumwoem I I I I I I I I I + + + + .om sawsouooeoamsob I I I I I I I I + + + oouxa 6:0on I I I I I I I I I + + + + wuoo oweomumeomm I I I I I I I I I I + + + + + masseuse assesses I I I I I I I I I I + + + + + .em oxooooosowe I + + + I + + + + I + + + + omoswmxsoo moxosowaosm I I I I I I I I I + + + + .s. I I I I I I I I + + + m I I I I I I I + + z I I I I I I I + + .s I I I I I I I I + + + m I I I I I I I I I + + + + s I I I I I I I I I + + + + o I + + + I + + + + I + + + + + m "amaadwmm 0u 0a m a 0 0u 0a m a 0 cu 0a m a 0 "mummaomoana N “amaze amuaoaoao mo doau .Imoauoovouml. nuBoso tamomow Hoanaaoomsuxm uaosmaa soaoaaooouuxm Ilutllmrulu 0 mm. o m oz N mauouomn 00 shadow cocoa uswao 0am mouoaooa unwao he Humans amuooaoao mo soauamomov umaoaaoomuuxo 0am .ooauoovonm Housman Hoasaaoomuuxo .nusouw .a sassy 34 Sphaerotilus. The use of a variety of liquid media also failed to yield a culture of Thiobacillus from the enrichments. The presence of Thiothrix at the leaf surface was confirmed by direct observation of the holdfast, sheath, cellular filaments, and intracellular sulfur globules which are characteristic of this genus. The problem of detecting fluorescent pseudomonads, which have no distinguishing morphological features, was overcome by laying plant leaves on the TS agar medium supplemented with antibiotics as suggested by Sands and Rovira (8). This resulted in the formation of microscopic fluorescent colonies in 24 to 48 h. The number of colonies per unit area of leaf surface was determined by counting fluorescent colonies using epifluorescence microscopy. By homogenizing a known leaf area and doing dilution plate counts, it was possible to also estimate the number of viable cells per colony. We found the method most effective with decaying aspen leaves. There were approximately 200 colonies/mm2 and 10 cells/colony. These values serve to demonstrate the presence of fluorescent pseudomonads in these aquatic environ- ments but do not represent an extensive quantitative survey. DISCUSSION The presence of fluorescent pseudomonads in HS- containing waters among autotrophic sulfur-oxidizing bacteria is anomalous due to the toxicity of HS-, which binds the ferric iron of cytochrome oxidase and prevents the reoxidation of cytochromes, thus preventing aerobic respiration. In addition, So forming activity as well as the three- fold increase in growth rate caused by thiosulfate gradients are incon- sistent with the role of pseudomonads as oxidative heterotrophs. 35 .Further studies to confirm the excretion of So forming com- pounds by fluorescent pseudomonads as well as studies on the effect of reduced sulfur compounds on metabolism are thus required. It is possible that fluorescent pseudomonads occupy the ecological niche previously attributed to the thiobacilli since thiosulfate gradients selected pseudomonads in situ while inorganic thiosulfate labora- tory enrichments commonly select thiobacilli. LITERATURE CITED Aaronson, S. 1970. Experimental Microbial Ecology. Academic Press, New York. p. 115. Bartlett, J. K., and Skoog, D. A. 1954. Colorimetric determi- nation of elemental sulfur in hydrocarbons. Anal. Chem. 26:1008-1011. Caldwell, D. E., and Hirsch, P. 1973. Growth of microorganisms in two-dimensional steady-state diffusion gradients. Can. J. Microbiol. 19:53-58. Caldwell, D. E., Lai, S. H., and Tiedje, J. M; 1973. A two- dimensional steady-state diffusion gradient for ecological studies. In Modern Methods in the Study of Microbial Ecology. Edited by T. Rosswall. Bull. Ecol. Res. Comm. (Stockholm). Vol. 17 pp. 151-158. Chen, K. Y., and Morris, J. C. 1972. Kinetics of oxidation of aqueous sulfide by 02. Environ. Sci. Tech. 6:529—537. Frank, L. H., and DeMoss, R. D. 1959. On the biosynthesis of pyocyanin. J. Bacteriol. 77:776-782. Rovira, A. D., and Ridge, E. H. 1973. .MOdarn.Methods in the ‘Study ofUMicrobial Ecology. Edited by T. Rosswall. Bull. Ecol. Res. Comm. (Stockholm). Vol. 17 pp. 329-335. Sands, D. C., and Rovira, A. D. 1970. Isolation of fluorescent pseudomonads with a selective medium. Appl. Microbiol. 20:513-514. van Niel, C. B. 1955. Natural selection in the microbial world. J. Gen. Microbiol. '13:201-217. CHAPTER III FLUORESCENT SULFURrOXIDIZING EXUDATES FROM PSEUDQMONASIAERUGINUSM BACKGROUND Pseudomonas aeruginosa excretes several water soluble pigments. One of these, pyocyanin, has been isolated and studied previously (2,6,7,9). The fluorescent pigments have not been isolated (5,13,15) but have been studied spectrophotometrically and spectrophoto- fluorometrically (5,11,16). The effect of culture conditions on fluorescence has also been investigated (8,13). Despite the possi- bility that more than one fluorescent compound may be excreted (5,13), the fluorescent exudates have been referred to as "fluorescein" (12) or "pyoverdine" (5). Sulfate, sulfite (13), or thiosulfate (Chapter II) is required for the production of both pyocyanin and the fluorescent pigments, despite the fact that these pigments are not known to contain S (13). Inorganic sulfur compounds also induce the formation of a protein- carbohydrate(l) slime (13). Although pyocyanin is known to serve as an H donor for nitrite reductase (4), no chemical or biological activity has been found for the fluorescent pigments. The results of preliminary experiments, discussed in Chapter II, suggested that fluorescent exudates from pseudomonads might be involved in the formation of So globules from HS-. This would partially explain the 36 37 occurrence of fluorescent pseudomonads among sulfur-oxidizing bacteria in nature (Chapter V) as well as the induction of fluorescent pigments by inorganic sulfur compounds. For these reasons, the effect of fluorescent exudates from P. aeruginosa on the formation of So from.HS- was investigated. This organism is well known as are the spectral characteristics of its exudates. The spectral characteristics of the exudates could thus be studied and compared to previous reports (5,10,11).. MATERIALS AND METHODS An inoculum of Paeudomonas aeruginosa was taken from an agar slant of the TS medium (Chapter II), transferred to 50 m1 of TS medium in a 250-300 ml Erlenmeyer flask and incubated at 25 C on a rotary shaker at 200-250 rpm. After 20-24 h the cultures became blue-green and 1 ml was subcultured to a 1,000 ml Erlenmeyer flask containing 250 ml of the TS medium with MgCl2 and glycerol deleted. In 15-20 h the culture appeared slightly yellow and turbid. The walls of the flask were encrusted with cells and the medium fluoresced blue. The liquid was decanted and centrifuged at 8,000 g for 20 min (4 C) in a refrigerated centrifuge (Sorvol Inc., Nbrwalk, CO). The fluid was decanted and filtered (0.45 um membrane filter, HAWG-O47-00, Millipore Corp., Bedford, MA) at 4 C. The filtrate was then sterile filtered (0.45 pm Nalgene Filter Unit, 120-0045, Nalge Sybron Corp., Rochester, NY) and stored at 4 C before use. An Aminco-Bowman spectrofluorometer and recorder were used to obtain excitation and emission spectra. Three volumes of 0.2 M 38 phosphate buffer (pH 7.6) were added to exudates and the pH adjusted before spectra were recorded. Consumption of HS,- by exudate was determined in a reaction mixture containing an equal volume of 0.2 M phosphate buffer (pH 7.6) and exudate. Sulfide was added to a final concentration of 1 mM and the sample incubated in a sealed reaction vessel; the final pH was 7.6. The incubation was ended by acidifying the sample and flushing with N for 10 min through a trap of ZnAcetate. Sulfide 2 was assayed by a modification of the method of Johnson and Nishita (8) as described by Roy and Trudinger (l4). Resazurin was used to titrate the sample to an orange end-point (pH 3.8) before flushing with N2. In a preliminary experiment, the addition of H2804 to a final concentration of 1.7 M resulted in a recovery of 0.95 i 0.138 mM HS- from a 1 mM solution containing phosphate buffer alone. However, when HS_ was added to the buffered TS medium, the addition of 112304 to 1.7 M resulted in recovery of only 0.06 i 0.03b mM as”. If the buffered medium was instead titrated to a resazurin end-point (pH 3.3) with 0.2 M H 504, then 0.92 i 0.108 mM as" were recovered. 2 Thus the pH indicator was used in all subsequent HS- determinations. Sulfide dependent gas uptake or production by buffered exudates was measured using a Warburg respirometer. Exudates were buffered by adding an equal volume of 0.2 M phosphate buffer (pH 7.6). Samples 895% confidence interval from 5 duplicates. b95% confidence interval from 3 duplicates. 39 (50 ml) were then added to 100 ml Warburg flasks. Sulfide was added from the side arm of the flasks to a final concentration of 1.0 mM. The pH was determined after all experiments and found to be 7.6. Elemental sulfur was determined as described in Chapter II. RESULTS AND DISCUSSION Some of the fluorescent characteristics of exudates from Paeudbmonas aeruginosa are shown in Figures 1 and 2. From a pH of 2.0 to 10.0 the exudate had an absorption peak at 405 nm. From a pH of 6.0 to 12.9 the exudates had a fluorescence peak at 460 nm (excitation, 400 nm). However, below a pH of 6.0 the fluorescence peak shifted to 430 nm at a pH of 1.0. Fluorescence at 460 nm was most intense at a pH of 7.0. The emission spectrum (excitation, 400 nm) of a.water sample taken from a bed of Chara within the littoral zone of Burke Lake, which had been filtered (0.45 um membrane filter) and adjusted to a pH of 7.0, is also shown in Figure l. The relative intensity of fluorescence of the culture exudates was 840 times as great as the fluorescence of the lake sample. The lake sample had a fluorescence maximum at 440 nm. Since the bed of Chara contained large numbers of fluorescent pseudomonads (Chapter V), there is a possibility that naturally occurring fluorescent pigments were partially responsible. The addition of sulfide (1 mM, pH 7.6) to a 24 h culture of .P. aeruginosa resulted in the formation of refractile globules 0.1 to 5.0 pm in diameter, which produced a diffraction pattern of colored concentric rings in a light microscope and adsorbed to the 40 Figure 1. Emission spectrum of Pseudbmonas aeruginosa exudate and a lake sample at an excitation of 400 nm. The exudate was diluted with 3 volumes of buffer and scanned at a multiplier set- ting of 0.03 (460 nm peak). The lake sample was undiluted and scanned at a setting of 0.001 (440 nm peak). Relative lntogolty u- o 000 200 41 460mm Figure l 42 Figure 2. The wavelength of the absorption peak ( b ), fluorescence peak (I ), and the fluorescence intensity (0 ) vs. pH for Pseudomonas aeruginosa exudates. 43 3.2.3:. Figure 2 44 cells. The globules were produced only upon addition of sulfide and were soluble in CClA, petroleum ether, or pyridine but not in water or 0.1 M HCl. This was considered presumptive evidence that the globules were composed of elemental sulfur. Although the addition of as‘ to phosphate buffer (pH) resulted in the release of H S, the addition of HS- to the exudates produced 2 no H28 odor. When the reaction was carried out in the Warburg respirometer (Figure 3), an initial volume increase occurred in the buffer flask but not in the exudate flask. Thus the volume changes corresponded to the qualitative difference in the presence of H28 odor. After an induction period of 50 min the volume of the buffer flask decreased at a constant rate. This was probably due to the consumption of 02 resulting from abiologic HS- oxidation (3). Oxygen uptake apparently did not occur in the case of the exudate, since the vessel remained at a constant volume after 50 min. Thus, the inorganic oxidation of HS- by 02 appears to have been prevented by the exudate. In addition, HS- must have been rapidly bound upon addition to the exudate. If it had not, the initial pH decrease would have immediately released H S as it did with buffer alone. 2 The rate of globule formation was followed spectrophotometrically as the change in OD340 per min (Figure 3). Sulfur globule formation occurred during the initial period of apparent binding (Figure 3); no globules were formed in vessels without exudate. The consumption of HS- by buffered exudate was compared to the consumption of HS- by buffered medium (Figure 4). The concentration of HS- remained constant in the buffer but decreased to 0.5 mM in 45 Figure 3. Volume change following addition of HS to phosphate buffer (0 ) and to Pseudomonas aeruginosa exudate in phosphate buffer (C ). Formation of So globules (A ) by later incubation mixture is shown as change in 0.D.340 per min. 46 4. 5.: son DC 2. 0. Figure 3 47 Figure 4. Sulfide consumption by Pseudomonas aeruginosa exudate in phosphate buffer (C ) and by buffered medium alone (I ). 48 Figure 49 the presence of exudate. After 3 min both the formation of So globules and the uptake of sulfide ended. Thus, when HS- was added to buffered exudates there was a decrease in the level of HS- (Figure 4), So was formed (Figure 3), and no measurable 02 uptake occurred (Figure 3). Since the only HS- oxidation product which does not require 02 is So, it appears likely that the exudates of Pseudbmonas aeruginosa result in the formation of So from HS-. LITERATURE CITED 1. Bartell, P. F., Orr, T. E., and Chudio, B. 1970. Purifica- tion and chemical composition of the protective slime antigen of Pseudbmonas aeruginosa. Infect. Imm. 2:543- 548. 2. Blackwood, A. C., and Heish, A. C. 1957. Pyocyanine formation from labelled substrates by.PseudOm0nas aeruginosa. Can. J. Microbiol. 3:165-169. 3. Chen, K. Y., and Morris, J. C. 1972. Kinetics of oxidation of aqueous sulfide by 02. Environ. Sci. Tech. 6:529- 537. 4. DeLay, J. 1964. Pseuddmonas and related genera. Ann. Rev. Microbiol. 18:17-46. 5. Elliot, R. P. 1958. Some properties of pyoverdine, the water- soluble fluorescent pigment of the pseudomonads. Appl. Microbiol. 6:241-246. 6. Frank, L. H., and DeMoss, R. D. 1959. On the biosynthesis of pyocyanin. J. Bacteriol. 77:776-782. 7. Friedheim, E., and Michaelis, L. 1931. Potentiometric study of pyocyanine. J. Biol. Chem. 91:355-368. 8. Garibaldi, J. A. 1967. Media for the enhancement of fluorescent pigment production by Pseudbmonas species. J. Bacteriol. 94:1296-1299. 9. Ingram, J. M., and Blackwood, A. C. 1962. Studies on the biosynthesis of pyocyanine. Can. J. Microbiol. 8:49-56. 10. Johnson, C. M., and Nishita, H. 1952. Microestimation of sulfur in plant materials, soils and irrigation waters. Analyt. ll. 12. l3. 14. 15. 16. 50 Kraft, A. A., and Ayres, J. C. 1961. Production of fluorescence on packaged chicken. Appl. Microbiol. 9:549-553. Lehmann, K. B., Neumann, R. 0., and Breed, R. S. 1931. Detenwinative Bacteriology. English translation, 7th German edition. J. E. Stechert and Co., New York. Palumbo, S. A. 1972. Role of iron and sulfur in pigment and slime production by Pseudbmonas aeruginosa. J. Bact. 111:430—436. Roy, A. B., and Trudinger, P. A. 1970. The Biochemistry of Inorganic COmpounds of'Squur. Cambridge University Press. Turfitt, G. E. 1937. XXV. Bacteriological and biochemical relationships in the pyocyaneus—fluorescens group II. Investigations on the green fluorescent pigment. Biochem. J. 31:212-218. Wasserman, A. E. 1965. Absorption and fluorescence of water- soluble pigments produced by four species of Pseudomonas. Appl. Microbiol. 13:175-180. CHAPTER IV THE CELLULAR MORPHOLOGY OF SULFUR- OXIDIZING BACTERIAL COMMUNITIES BACKGROUND The range of in situ conditions leading to the numerical dominance of a few species is often narrow and definable (6,7). As a result, the morphology of the dominant bacteria seen under selective in situ conditions may be remarkably uniform (20,28). In contrast, transfer to laboratory media may result in pleomorphism because of unusual and variable culture conditions in which the organism would not normally be able to successfully compete (27,29). There are many instances in which the natural microflora cannot be successfully or reproducibly cultivated using artificial conditions (7,19,28,29). In particular, the gas vacuolate sulfur bacteria of the hypolimnion are difficult to cultivate while transient Rhodbspirillaceae grow readily (29,32). Despite these difficulties, differential and selective media are commonly used for the enumera- tion of aquatic bacteria (32). Direct observation, however, allows one to observe all of the dominant morphological types in a community, provides information on the population density and biomass of each type (5), and allows one to observe physical relationships among the components of the community. For these reasons, direct observation 51 52 was used here preferentially for studying the bacteria of the anaerobic hypolimnion, the littoral zone, and a sulfur spring. Previous in situ descriptions of these bacteria have been limited to the light microscope and photographs have often been of poor resolution and contrast. This has resulted in questionable identification of some of the bacteria and uncertainty of the existence of certain bacterial groups (1). In this study, the in situ morphology of common bacterial forms from the hypolimnia and littoral zones of two southern Michigan lakes has been documented by phase and electron microscopy. The morphology of isolated genera has been compared to that found in nature. MATERIALS AND METHODS Lake samples from the hypolimnion were obtained as described elsewhere (5) so as to minimize exposure of the samples to 02, light, and high temperature. After collection, samples were stored on ice and observed within 8 h. All photographs shown were taken of samples obtained from Wintergreen Lake, TlS:R9W:S8, Kalamazoo County, Michigan, and Burke Lake, T5N:R1W:S23, Clinton County, Michigan. The characteristics of these lakes have been described elsewhere (5). Light microscopy A Zeiss photomicroscope was used for phase microscopy. All cells were mounted on slides coated with a thin dried layer of NObel Agar (Difco, Detroit, MI). The coated slides were prepared by adding a thin layer of sterile melted agar (0.752) to a cleaned slide 53 and allowing the agar to dry in a laminar flow hood under ultraviolet light. A small drop of lake water (0.02 ml) was placed on the dry agar and cover slip was immediately pressed over the drop. The resulting film of water was absorbed by the agar which pressed the organisms against the coverslip in a monolayer. The deformation of the agar by the cells increased the contrast between the cells and the background. Many of the morphological characteristics described could not be clearly observed unless this procedure was used. Refractile inclusions which could be removed either by centri- fugation or by pressing the cover slip of prepared slides were considered to be gas vacuoles. The remaining inclusions which pro- duced a diffraction pattern of colored concentric rings in the phase microscope were considered to be sulfur granules. Electron microscopy One hundred milliliter lake samples were collected in suction flasks containing 100 m1 of glutaraldehyde fixative (6% in 0.1 M phosphate buffer at pH 7.2). The cells were collected on.a.Millipore filter (UHW PO4700 Millipore Corp., Bedford, MA) with a pore size of 0.22 um, and a thin layer of 12 Nobel Agar (Difco, Detroit, MI) was then added to the filter under a slight vacuum. The filter was placed in 3% glutaraldehyde fixative (in 0.1 M phosphate buffered at pH 7.2) overnight. The filter was then washed in 3 to 5 changes of 0.1 M phosphate buffer (pH 7.2) and the cells fixed overnight with 0804 according to the Ryter-Kellenberger procedure (22). The filters were subsequently treated for 2 h.with 0.52 uranyl acetate dissolved in the Ryter-Kellenberger buffer and dehydrated (5 min in 50%, 70%, 54 95%, 3 changes of 100% ethanol, 3 changes in propylene oxide). The filters were infiltrated with plastic by submerging them for 30 min in 2 parts propylene oxide and 1 part Epon 812 (Polysciences, Warrington, PA), 1 h in 1 part propylene oxide and 2 parts Epon 812, and finally in undiluted Epon 812 overnight. The specimens were polymerized for 36 h at 40 C and then for 24 h at 60 C. Sections were prepared using glass knives and an LKB Ultotome. The thin sections were post stained 10 min in 2% uranyl acetate (11) and 5 min in 2% lead citrate (3,41). The sections were observed using a Hitachi 11 electron microscope. To assist in identification of the thin—sections, glutaraldehyde fixed samples and l um thick sections were examined by light microscopy. Whole cell specimens were prepared by allowing the cells to clump after a brief exposure to oxygen (29) and transferring a clump to a formvar coated electron microscope grid. Excess moisture was drawn from the grid using a tissue. Cells at the edge of a clump were observed after shadowing at an angle of 15 degrees using a platinum—carbon pellet. RESULTS AND DISCUSSION The organisms were grouped according to genus in the case of previously described forms and according to morphological similarity in the case of unreported forms. The morphological groups were assigned a type number to facilitate their discussion. All of the organisms described are either sulfur-oxidizing bacteria themselves or associated with sulfur-oxidizing communities. 55 Since the samples were of natural communities made up of several populations, it is not possible to be absolutely certain that we have correctly related the organisms observed by electron microscopy (particularly for thin-sections) to the same organism observed by phase microscopy. In the subsequent descriptions where the morpho- logical relationship between organisms shown by both methods is not obvious, we have indicated the reasons for our assignment. In many cases the presence of chlorobium vesicles (a cortical array of photosynthetic vesicles), which occur only in the green bacteria, was used to identify members of this group and to distinguish them from the purple bacteria which contain either intracytoplasmic vesicles or lamellae (31). Pe Zodic tgion o la thra tiforme Pelodictyon, a green sulfur bacterium, was described by Szafler (39) and Lauterborn (25) as a rod-shaped organism which forms a net-like aggregate (Figure 1a). Closer examination has revealed that the rod-shaped cells (0.5 to 1.5 by 1.2 to 4 um) are not only attached to each other end to end but may form truncated prosthecae (Figure 1a,b,c). This characteristic is retained in enrichments (30) and appears more reliable for identification pur- poses than the net-like configuration commonly observed for the aggregate in nature. Thin sections reveal the cortical array of vesicles that extend into the truncated prosthecae (30) (Figure 1b). The gas vacuoles are usually arranged parallel to the longitudinal axis of the cell (Figure 1b,c). The cell wall is bounded by a layer 0 of striations with an amplitude of 150 A (Figure 1d). Strains 1831 56 Figure 1. PeZodictyon clathratifbrme (bars represent 1 pm) The net-like configuration, gas vacuoles, and truncated prosthecae characteristic of the genus Pelodictyon (phase micrograph). A longitudinal cross section showing the cortical vesicles (2), longitudinal gas vacuoles (3), truncated prosthecae (4) and surface striations (5) (electron micrograph). A shadowed preparation showing the above features and the characteristic longitudinal gas vacuoles (l) (electron micrograph). An oblique section showing the surface striations (6) which have an amplitude of 150 A (electron micrograph). 57 Figure 1 58 and 2730 isolated by Pfennig and Cohen-Bazire (30) also possess these striations. Pfennig's strain 2730 possessed two layers of"' striations while strain 1831 had only one, as did the strain found in the hypolimnion. Pfennig and Cohen-Bazire also found that under certain culture conditions gas vacuoles are not formed, although we have always found gas vacuoles present in situ. Pelodictyon was found in both lakes but was most conspicuous in Burke Lake and layered above the other green bacteria. ProsthecochZoris-Ancalochloris Prosthecochloris was first described and isolated by Gorlenko (14,15). It is a green bacterium with prosthecae (Figure 2a,b,c) and chlorobium vesicles (Figure 2a,c). In a second isolation attempt Gorlenko (16) followed the pro- cedure outlined by Pfennig (29) for the enrichment of Pelodiotyon. In this case a gas vacuolate organism was enriched and named Ancalochloris. The difference between this organism and Prostheco- ‘ chloris was the presence of gas vacuoles and longer prosthecae. However, in this report the prosthecate, gas vacuolate, green sulfur bacteria are referred to as Prosthecochloris. All of the prosthecate green sulfur bacteria found in the hypolimnia possessed gas vacuoles. The size of the free-living form of Prosthecochloris (0.5 to 1.0 um in diameter) is about the same as the isolates of Gorlenko (0.5 to 0.7 pm by 1.0 to 1.2 um); however, the prosthecae are larger (0.2 by 0.7 pm as compared to 0.10 by 0.17 um for the isolate). 59 Figure 2. Prosthecochloris (bars represent 1 pm). a. Shadowed preparation of Prosthecochloris. The cells are prosthecate (l) and contain gas vacuoles as well as intra- cellular granules (electron micrograph). b. Phase micrograph of Prosthecochloris with gas vacuoles removed by centrifugation. The cells are clearly pros- thecate (2). No sulfur granules are present. c. A thin section of two cells which shows all of the features above plus the presence of a cortical array of vesicles (3). Both the cytoplasm and vesicles extend into the prosthecae. The granules do not have the refractile appearance of sulfur. 60 Figure 2 61 Prosthecochloris accounted for 5 to 20% of the total community of green sulfur bacteria in Wintergreen Lake and layered below the purple sulfur bacteria. It was present in Burke Lake although in lower numbers. 6' Zathroch Zomls hypo Zimnica CLatrochZoris has been described by Geitler (13) as a unicellular green sulfur bacterium which deposits sulfur internally. The identi- fying characteristic is the trellis-like aggregation of gas vacuolate cells (Figure 3a). Each of the cells was found to contain storage granules although sulfur granules were not apparent (Figure 3b,c). The pigment in samples containing primarily CZathrochZoris was found to be bacteriochlorophyll d (chlorobium chlorophyll 650). The observed purity ratio (optical density 650/optical density 500) in methanol saturated with hydrogen sulfide was 23.5, which compared favorably to a value of 24.7 for the chromatically purified prepara- tions of Kaplan and Sieberman (21) obtained from pure cultures of Chlorobium sp. CZathrochZoris comprised 80—902 of the bacterial community in the 4.0 m zone of Wintergreen Lake; it layered immediately below the purple bacteria. ChZorochromatium_gggregatum Although recognition of the genus ChlorochromatiuM’has been questioned (37), the fine structure of coenobia taken from the hypolimnion supports its validity. The coenobium consisted of a central colorless cell with peripheral green cells which gave the 62 Figure 3. CZathrochZoris hypolimnica (bars represent 1 um). a. Phase micrograph of a typical aggregate. Gas vacuoles cause the high refractility of the cells. b. Phase micrograph after the gas vacuoles were removed by centrifugation. c. A shadowed preparation showing intracellular granules (1), gas vacuoles (2) and the trellis-like arrangement for which the aggregate was named (electron micrograph). Figure 3 64 aggregate a barrel-like appearance (Figure 4a,b). The photosyn- thetic symbiont contained bacteriochlorophyll d (chlorobium chloro- phyll 650) and chlorobium vesicles (Figure 4c,d). The heterotrophic symbiont lacked photosynthetic vesicles (Figure 4c) and its outer cell wall consisted of a layer of cups which covered the surface (Figure 4c,e). This structural feature has been observed in Lampropedia (8) and Thiopedfia (Hirsch, unpublished data), which are metabolically dissimilar but which both form plates of cells. This suggests that the cups may play a role in maintaining the shape of the coenobium. Possibly, the heterotrophic symbiont of Chloro- chrawatium has evolved an adhesive surface structure which is specific for its symbiont. Its fusiform shape and one-dimensional division plane would prevent cohesion and the formation of cell plates as in Thiopedfia and Lampropedia. Without the flagellated heterotrophic symbiont, the photosynthetic symbiont would lack a mechanism for maintaining its position in the water column. The symbiosis may thus be obligatory in the hypolimnion. The envelope of the photosynthetic symbiont consisted of a membrane interrupted periodically by mesosomal invaginations (Figure 4c) and a cell wall composed of four electron dense layers (Figure 4d). The membrane often adhered to the vesicles and separated from the cell wall (Figure 4c). A few of the aggregates observed possessed brown instead of green symbionts. These appear to fit the description of the Pelochromatium roseum consortium (25) which was abandoned in the seventh edition of Bergey's Manual (4). However, this genus is 65 still cited due to the common occurrence of the aggregate in hypo- limnia (17,35) and will be reinstated in the eighth edition of Bergey’s Manual (Pfennig, personal communication). No differences other than pigmentation were noted between the two types. Chlorochromatium was observed in both of the lakes studied, but it was most dense during the winter stratification of Burke Lake and gave the water a bright green color. The high density of Chlorochromatium in samples used for electron microscopy made identification of thin-sections of this organism simple. Type lgprocaryotesg(T1) Tl consists of an aggregation of elliptical cells 1.2 to 1.5 pm by 1.5 to 2.5 pm (Figure 5b,c). The cells are joined at the base forming a rosette. Each cell has an optically dense area near its outer tip which is most clearly seen when the gas vacuoles are removed (Figure So). This dense area contains both lamellae and gas vacuoles but not photosynthetic vesicles (Figure 5a,b). These lamellae suggest that T1 may be either a blue-green alga or a purple bacterium since both of these groups may possess photosyn- thetic lamellae. However, polar stacks of membranes, like those shown in Figure 5a, occur only in the budding purple bacteria (31). Buds observed on the cell surface (Figure 5b) were also seen in thin-sections. Protruding from the surface of each cell are 10 to 50 filaments approximately 100 A in diameter and 3 um in length. T1 was distributed among the purple and the green bacteria but was never found in great abundance. 66 Figure 4. Chlorochromatium aggregatum. The aggregation after being mounted on an agar slide (bar represents 1 pm). Both the colorless heterotroph (1) and the outer photosynthetic cells can be seen (2) (phase micrograph). The aggregation as it appears in a wet mount before being mounted on an agar slide (bar represents 1 pm). Only the green photosynthetic cells are visible (phase micrograph). A cross section through the transverse axis of Chloro- chramatium aggregatum showing the peripheral vesicles of the photosynthetic symbionts (4) and the heterotrophic cell (5) with external surface structures (6) (bar repre- sents 1 pm). Enlargement of the peripheral vesicles (7), cell wall (8) and membrane (9). The membrane often separates from the wall and adheres to the vesicles (bar represents 0.1 um). An oblique section of the central heterotroph showing the external surface structures which appear cup—shaped (10) (bar represents 1 pm). 67 68 Figure 5. The Type 1 procaryotes (T1) (bar represents 1 pm). a. A thin section of T1 showing the polar stack of lamellae and gas vacuoles. b. Phase micrograph of the aggregate (3) after centrifugation to remove gas vacuoles. The cytoplasm toward the outside of the aggregation is dense and well defined (4). c. A shadowed preparation showing the filaments which pro- trude from the surface of each cell (1). The upper portion of each cell contains a cluster of gas vacuoles (2) (electron micrograph). 69 Figure 5 70 The assumption that Figure 5a,b,c is of the same organism is based on the increased density of the cytoplasm surrounding the gas vacuoles at the ends of the cells (Figure 5b), which suggests that the photosynthetic apparatus surrounds the vacuoles as shown in Figure 5a. The size of the section also correlates with the size of the organism in light micrographs. Thiothrix It has been suggested by several authors that the bacterial genera Cladothrix (2,33,38), Sphaerotilus, and Thiothrix (36) may represent morphological variations of a single genus, Sphaerotilus. Sphaerotilus consists of a chain of rod-shaped cells enclosed in a sheath with a holdfast attaching the chain to the substratum (24). While in the sheath, the cells become flagellated (lophotrichously) and are then released (36). Members of the genus Ciadothrix‘were originally distinguished from Sphaenotilus by dichotomous branching (9). However, it has been subsequently shown that branching is a morphological variation of Sphaerotilus induced by low concentration (<0.lZ) of peptone (34). At high concentrations (>0.252), cells occur randomly within a mucilaginous mass (12). Changes in the concentration of phosphate buffer have little effect on filamentous growth but levels about 0.05 M inhibit growth (12). It has been suggested that Thiothrix, like Cladothrix, is simply a morphological variation of Sphaerotilus (18). Thiothrix was originally distinguished from Sphaerotilus on the basis of the gliding motility of conidia, intracellular deposition of sulfur, formation of a true trichome and obligate chemoautotrophic metabolism 71 (18,36). Although Skerman has shown that Sphaerotilus is capable of depositing sulfur internally (in the presence of sulfide), he still maintained the differentiation between genera on the basis of true trichomes and gliding motility (the autotrophic metabolism of Thiothrflr has never been shown directly since no pure cultures have been isolated) (33). However, the gliding motility of conidia observed by Winogradsky in 1888 (18) has not yet been confirmed despite the efforts of Keil (23) and Pringsheim (33). The forms of Thiothrix observed in the littoral zone (Figures 6 and 7) on decaying plants deposited elemental sulfur, contained peripheral lamellae, were 2 to 5 um in diameter, and attached to the plants using a holdfast (Figure 7). However, the lamellae were not observed by Drawert and Metzner-Kuster (10). Thiocystis, Chroma 7%szL Thiospirillwn, and Thiopedia These organisms were prevalent in many samples but are well ‘known and for that reason are only briefly mentioned. Chromatium and Thiocystis generally deposited sulfur. Thiocystis attached to decomposing plant surfaces in the littoral zone (Figure 6) using its mucilage as a holdfast. Figure 8a shows a plate of Thiopedia cells containing sulfur granules. Contrary to most keys, Thiopedfla (Figure 8a,b) and Thiospirillum were usually found without sulfur granules when growing in the hypolimnion in the presence of measurable sulfide. 72 Figure 6. The microenvironment at the surface of decomposing plant material taken from the littoral zone of Burke Lake (electron micrograph of a thin section). A microcolony of Thiocystis fills the upper portion of the figure. The large inclusions (2,3) are sulfur granules. The microcolony is embedded in mucilage (l) which serves to anchor the colony to the plant wall (6). A transverse section through a Thiothrix filament shows the sheath (5) and peripheral membranes (4). Also seen are organisms with a scalloped cell wall and nucleophilic granules (7) which are characteristic of thiobacilli., Bar equals 1 um. Micrograph by Sarah Caldwell. 73 Figure 6 74 .Hamaeamo Assam kn admuwouofiz .AMV coauoom onu a“ unwound omHm mH hegemoovx& .eommam moooou me new on moaomzxonumo wowanawmou monouuouum m:HMu:oo omHm owemmooauwz .Amv mxumum mamunama emumooa kHHmuuooo mofimusoo nofiss was Adv haoaoo one mommooo scams owmawooa ecu moan: oommuom unman one ou manoeuum sagas aaonm omHm ma hooaooouofia Hafiuouomn < .omamv souuooao %Hamvooa= ma uadwou o no new Aefioo manage cw oanoaomv ovaxouvhn Goya mawmucoo haamuoomw ANV some use on mfiunoe may .AHV macaw ma afieeuow£& modem may mo owumfiuouomumso ummmwao: < .Asowuoom menu mo ammuwouofia souuooamv mxmq oxuom Boum coxMu Hmwuouma nomad wswmoqaoooe mo moowuom map on unmanomuum Hmfiuouomm .n ouowwm 75 Figure 7 76 Figure 8. Thiopedia and the Type 2 procaryotes (T2) (bar repre- sents l um). a. Thiopedia with gas vacuoles removed by centrifugation revealing several sulfur granules (1). b. Plate of Thiopedia cells with gas vacuoles (phase micro- graph). c. T2 as it appears with phase microscopy (2). Figure 78 Typengrocaryotes (T2) The size of the cells of type 2 procaryotes is 1.2 to 1.5 um by 4 to 12 um (Figure 8c). The curved fusiform characteristic (Figure 8c) is generally present although the degree of curvature may vary (in some cases the cells are simply fusiform). These organisms do not contain gas vacuoles or sulfur granules. This was the only large, curved, and fusiform organism found in the hypo- limnion. As in the case of ThiospiriZZum and Prosthecochloris, T2 does not appear to be colored when viewed by phase or bright field microscopy. T2 has a distribution pattern which is identical to that of the green bacteria though it was often only 1% of the community. Type 3 procaryotes (T3) The individual cells of T3 occur in plates of 2,4,8,16, or 32 cells (Figure 9a,b,c). The arrangement appears to be maintained by a colorless mucilage in which refractile granules are occasionally deposited. The cells are separated by a clear region (Figure 9a). No surface structures that might poseibly be involved in maintaining the plate arrangement have been observed. The cells always contain gas vacuoles. Reproduction is by binary fission which occurs when the ends of the curved cells touch. The cells are 0.5 to 0.6 um by 2.0 to 6.0 um in size. There is a strong resemblance between T3 and the gas—vacuolated strains of Microcyclus aquaticus described by Van Ert and Staley (40). HOVever, T3 is from the anaerobic zone, occurs in plates, and forms complete circles (Figure 9a), each cell being of uniform thickness. The gas vacuolated forms of.Microcchus 79 Figure 9. The Type 3 procaryotes (T3) (bar represents 1 pm). a. A shadowed preparation showing gas vacuoles (l) and binary fission (2). Plates of 2, 4, 8 or 16 cells are typical. Although they may be broken into two cell units, they rarely occur unicellularly (electron micrograph). b. T3 before removal of gas vacuoles (arrow) with cell plates disrupted (phase micrograph). c. T3 after the removal of gas vacuoles by centrifugation (phase micrograph). 80 Figure 9 81 aquaticus are aerobic, unicellular, and vibriod (40). T3 is most abundant immediately above the sediments (20 to 40% of the community) although it appears in low numbers throughout the hypolimnion. Type 4 procaryotes (T4) The type 4 procaryotes are distinguished by their mode of division and the shape of their coenobia. The fusiform cells con— tain gas vacuoles and occur in a pallisade arrangement within a colorless mucilage. The structure of the coenobia, as in the case of T3, is very distinct from that of Lampropedia (8) and Thiopedia (Hirsch, unpublished data) and does not require contact between cells (Figure 10a,b). T4 always appears very transparent and could not be resolved unless phase microscopy and agar slides were used. Division occurs after the cells have elongated to approximately 9.0 um (Figure 10b). An apparent thickening at the center of each cell occurs in coenobia (Figure 10c). Thin sections of samples con- taining T4 frequently reveal structures like the one shown in Figure 10d. Based on phase microscopic examination of the sample, T4 is the organism which most likely could produce this structure. The thickening may be a result of an unusual division process resulting in the formation of a cellular extension after fission (Figure lOa,d). T4 is distributed throughout the hypolimnion but is most abundant at the water-sediment interface reaching a density of 20% of the total community. 82 ' Figure 10. The Type 4 procaryotes (T4) (bar represents 1 pm). a. Note the palisade arrangement, gas vacuoles (1) and cellular extension (2) (electron micrograph of a shadowed preparation). b. T4 after the division process (gas vacuoles removed by cen- trifugation, phase micrograph). c. T4 during the division process. Note the thickening (3) at the center of the plate (gas vacuoles removed by centri- fugation, phase micrograph). d. A thin section showing the site of possible division (4). t y_ . . . . _, ‘ . . .a 'r £55.. ,"~- .- A A ‘A . ' t ; 2." .fl‘l‘ V Figure 10 84 Type 5 procaryotes (T5) T5 is fusiform in shape (Figure lla,b,c) with two clusters of gas vacuoles on either side of the division site (Figure llb). Division occurs by binary fission. The cells are 0.5 to 0.8 pm by 3.0 to 15 pm in length. The poles of the cells are square (Figure 11c). The cells may be either straight or curved. The curvature is often confined to the ends of the cell (Figure 11a). These organisms are confined to the sediment-water interface reaching a density of 50-60% of the total community. Type 6 procaryotes (T6) T6 consists of filamentous cells (1.0 to 1.5 um by 10 to 100 pm) with irregularly spaced cross walls (Figure 11c,d,e) which are photon dense (Figure lld,e) and electron transparent (Figure 11c). The wall of the filament is covered.by longitudinal striations (Figure 11c) which are occasionally visible in the light micro- scope. Each cell contains gas vacuoles. T6 is found near the sediment and reaches a density of 5% of the bacterial community. Most of the purple and green bacteria described above have been reported previously (29), although their in situ morphology has never been observed using the electron microscope. Only in the case of Chromatium and Thiocystis did the description of pure cultures correlate exactly with morphology of these genera in situ. The gas- vacuolate achlorophyllous bacteria found at the sediment-water interface have never been isolated or reported previously. 85 Figure 11. The Type 5 and 6 procaryotes (T5 and T6) (bar represents 1 pm). a. T5 with gas vacuoles removed by centrifugation. The cells may be curved at the tips or rod shaped (phase micrograph). b. T4 with gas vacuoles which are separated into two clusters (1,2) by the division site (phase micrograph). c. T5 and T6 showing the square shape and dense cytoplasm (3) of the ends of T5 and the longitudinal striations (4) of T6 (electron micrograph of a shadowed preparation). d. T6 with gas vacuoles removed (5) by centrifugation (phase micrograph). e. T6 with gas vacuoles (6) (phase micrograph). 86 Figure 11 10. ll. 87 LITERATURE CITED Anagnostidis, K., and Overbeck, J. 1966. Methane oxidizing and hypolimnetic sulfur-oxidizing bacteria. Studies of the ecological biokinetics of aquatic microorganisms. Beri. Deutsch. Botan. Ges. 73:163-174. Bahr, H. 1953. Zur Biologie von Sphaerotilus natans Kutzing, Schweiz. Z. Hydrol. 15:285-301. Berlin, J. D., and Ramsey, J. C. 1970. Electron microscopy of the developing cotton fibers. IIE_28th Proceedings of the electron microscopy society of America. Edited by C. J. Arceneaux. Claitor's Publishing Div., Baton Rouge, LA pp. 128-129. Breed, R. S., Murray, E. G. D., and Smith, N. R. 1957. Bergey's Manual of’Deterwinative Bacteriology. Williams and Wilkins Co., Baltimore. Caldwell, D. E., and Tiedje, J. M. 1974. The structure of anaerobic bacterial communities in the hypolimnion of several Michigan lakes. Can. J. Microbiol. (in press). Caldwell, D. E., Lai, S. H., and Tiedje, J. M. 1973. A.two- dimensional steady-state diffusion gradient for ecological studies. In Modern Methods in the Study of Microbial Ebology. Edited by T. Rosswall. Bull. Ecol. Res. Comm. (Stockholm). Vol. 17 pp. 151-158. Caldwell, D. E., and Hirsch, P. 1973. Growth of microorganisms in two-dimensional steady-state diffusion gradients. Can. J. Microbiol. 19:53-58. Chapman, J. A., Murray, R. G. E., and Salton, M. R. J. 1963. The surface anatomy of Lampropedfia hyalina. Proc. Roy. Soc. London, Ser. B. 158:498-513. Cohn, F. 1875. Untersuchungen fiber bacterien II. Beitr. Biol. Planz. 1:141-207. Drawert, H., and Metzner—Kuster, I. 1958. Fluorescenz und electron mikroskopische untersuchungen un Beggiatoa alba und Thiothrix nivea. Archiv. fur Mikrobiologie 31: 422-434. Fahmy, A. 1967. An extemporaneous lead citrate stain for electron microscopy. In_25th Proceedings of the electron microscopy society of America. Edited by C. J. Arceneaux. Claitor's Publishing Div., Baton Rouge, LA pp. 148-149. 12. l3. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 88 Gaudy, E., and Wolfe, R. S. 1961. Factors affecting filamen- tous growth of Sphaerotilus natans. Appl. Microbiol. 9:580—584. Geitler. 1925. Pascher, Die Suswasserflora Deutschlands. Osterr u. Schweiz. Jena. 12:457. Gorlenko, V. M., and Zhilina, T. W. 1968. Study of the ultra- structure of green sulfur bacterium, strain SKs413. Mikrobiologiya. 37:1052—1056. Gorlenko, V. M. 1970. A new phototrophic green sulphur bacterium - Prosthecochloris aestuarii nov. gen. nov. spec. Z. Allg. Mikrobiol. 10:147-149. Gorlenko, V. M., and Lebedeva, E. V. 1971. New green sulfur bacteria with appendages. Mikrobiol. 40:1035-1039. Gorlenko, W. M., and Kusnezow, S. I. 1972. The photosynthe- sizing bacteria of lake Kononjer. Arch. Hydrobiol. 70:1-13. Harold, R., and Stanier, R. Y. 1955. The genera Leucothrix and Thiothrix. Bact. Rev. 19:49-64. Hirsch, P. 1972. New methods for observation and isolation of unusual or little known water bacteria. Z. Allg. Mikrobiol. 12:203-218. Hirsch, P., and Pankratz, S. H. 1970. Study of bacterial populations in natural environments by use of submerged electron microscope grids. Z. Allg. Mikrobiol. 10:589-605. Kaplan, I. R., and Sieberman, H. 1959. Spectroscopy of bacterial chlorophylls separated by paper and cellulose column chromatography. Arch. Biochem. Biophys. 80:114-124. Kellenberger, E., and Ryter, S. A. 1958. Electron microscope study of DNA-containing plasms. J. Biophys. Biochem. Cytol. 4:671—678. Kiel, F. 1912. Beitrage zur physiologic der farblogen schwefel bacterien. Beitr. Biol. Pflanz. 11:335-372. Kutzing, F. T. 1833. Sphaerotilus natans eine neve susswasseralge. Linnaea. 8:385. Lauterborn, R. 1913. Zur kenntniseinger Sapropelischer Schizomyceten. Allg. Bot. Z. 19:97-100. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 89 Lauterborn, R. 1915. The saprophytic portion of the biosphere: a contribution to the biology of decaying mud in natural waters. Verh. Naturhist. Mediz. Ver. Heidelburg NF. 13:395-481. Nikitin, D. I. 1973. Direct electron.microscopic techniques for the observation of microorganisms in soil. In; Modern Methods in the Study of Microbial Ecology. Edited by T. Rosswall. Bull. Ecol. Res. Comm. (Stockholm). Vol. 17 p 92. Olah, J., Hajdu, L., and Elekes, K. 1972. Electron microscopic investigation of natural bacterial populations in the water and sediment of Lake Balaton and Lake Belso. Tihanny 39:123-129. Pfennig, N. 1967. Photosynthetic bacteria. Ann. Rev. Microbiol. 21:258-324. Pfennig, N., and Cohen-Bazire, G. 1967. Some properties of the green bacterium Pelodictyon clathratiforwe. Arch. Mikrobiol. 59:226-236. Pfennig, N., and Truper, H. G. 1973. The Rhodospirillales. IE Handbook ofHMicrobiology. Vol. 1. Edited by A. I. Laskin and H. H. Lechevalier. CRC Press, Cleveland, OH pp. 17-27. Pratt, D. C., and Gorham, E. 1970. Occurrence of Athiorhodaceae in woodland, swamp, and pond soils. Ecology 52:346-349. Pringsheim, E. C. 1949. The filamentous bacteria Sphaerotilus, Leptothrix, Cladothrix, and their relation to iron and manganese. Trans. Roy. Soc. (London) Series B, 233: 453-482. Pringsheim, E. G. 1949. The relationship between bacteria and the myxophyceae. Bact. Rev. 13:47-48. Schegg, V. E. 1971. Production and destruction in the tropho- genic zone. Schweiz. Z. Hydrolog. 33:425-532. Skerman, V. B. D., Dementjeva, C., and Carey, B. J. 1957. Intracellular deposition of sulfur by Sphaerotilus natans. J. Bact. 73:504-512. Skerman, V. B. D. 1967. A Guide to the Identification of the Genera of Bacteria. Second Ed. Williams and Wilkins, Baltimore, MD p. 102. 38. 39. 40. 41. 90 Stokes, J. L. 1954. Studies on the filamentous sheathed iron bacterium Sphaerotilus natans. J. Bact. 67:278-291. Szafer, W. 1910. Zur Kenntnis der Schwefelflora in der Umgebung von Krakau. Anzeiger der Akad. der Wissenschaften in Krakau. Math. Klasse B. pp. 161-162. Van Ert, M., and Staley, J. T. 1971. Gas vacuolated strains of Microcyclus aquaticus. J. Bacteriol. 108:236-240. Watson, M. L. 1958. Staining of tissue sections for electron microscopy with heavy metals. J. Biophys. Biochem. Cytol. 4:475-478. CHAPTER V THE DISTRIBUTION OF SULFUR-OXIDIZING BACTERIA BACKGROUND In anaerobic hypolimnia, sulfide may be produced and oxidized biologically in the absence of oxygen. However, in the upper region of an anaerobic hypolimnion, in the littoral zone, and in sulfur springs, sulfide and oxygen containing waters can be mixed. When this occurs, sulfide oxidation proceeds both biologically and abio- logically. Abiologically sulfide is oxidized to sulfur, thiosulfate, and/or sulfate depending on the reaction conditions (7). Biologically sulfide may be oxidized by the Chromatiaceae, Chlorobiaceae,color1ess sulfur bacteria and heterotrophic sulfur—oxidizing bacteria. In this chapter the distribution of these organisms in sulfide containing environments is discussed and interpreted using the information from preceding chapters on physiology, morphology, and in situ gradient enrichment. Although fluorescent pseudomonads and thiobacilli were enumerated using plating media, the failure of cultural methods to support many photosynthetic bacteria and Thiothrix sp. required the use of direct microscopy. The distinctive morphology of sulfur-oxidizing bacteria has allowed several previous studies on the distribution of these bacteria as a function of physicochemical factors (2,3,8,11,12,13). 91 92 The present study improves on these by providing more detailed information on morphological variation in freshly collected samples, by collection of samples from smaller depth increments (0.1 m for our study of hypolimnia), by attempting to correlate structural information with determinative environmental parameters, and by including heterotrophic sulfur-oxidizing bacteria. MATERIALS AND METHODS seas. Thiobacillus Broth (Difco Laboratories, Detroit, MI U.S.A.) was used with 1.5% NObel Agar (Difco) for plate counts of thio- bacilli (pH 4.8). Plates were observed after 48 h, 7 days, and 30 days. The TS (thiosulfate) medium (pH 7.5) was used to induce pigment production by fluorescent pseudomonads. It is a modification of the media of Frank and DeMoss (10), Sands and Rovira (21), and Rovira and Ridge (20). It contained: 2% glycerol (w/v), 1% Na 8 O °7H20, 2 2 3 1% case amino acids, 0.01 M KZHPO 0.01 M MgClz, and 1.5% Nobel 49 agar (for solid media). In some cases antibiotics were added, including: 45 mg/liter Penicillin G, 45 mg/liter novobiocin (Upjohn, Albamycin), 75 mg/liter cycloheximide (Upjohn, Acti-Dione), 5 mg/liter chloramphenicol. Sampling For hypolimnetic bacterial counts and chemical determinations, the following sample collection procedure was used. Amber serum bottles (37 m1) filled with nitrogen and sealed with silicone rubber 93 septa were used as collection bottles. The bottles were evacuated at the sampling site by introducing a 27 gauge hypodermic needle which was connected to a vacuum source. The sample was drawn from the lake through a latex hose which was connected to a second hypo— dermic needle. To initiate collection this needle was inserted through one wall of an evacuated rubber tube allowing the sampling hose to be flushed with several volumes of sample. The evacuated hose was then pressed against the septum of the collection bottle (which was still connected to the vacuum source) and the needle was inserted through both the second wall of the vacuum hose and the septum of the evacuated bottle. Once the bottle had filled, the vacuum and collection needles were removed. At the end of the col- lection period the samples were placed in a bucket of ice water. Samples were taken at 0.1 m increments of increasing depth. The sampling hose was connected to a rigid sample inlet and a lead weight which was lowered by a brass chain to the desired depth. Sampling was done from 3 to 5 am to avoid surface disturbances and direct sunlight. For less intensive samplings, a Van Dorn sampler was used. Samples from the littoral zone and springs were collected using sterile 50 ml syringes to replace the sampling hose. Bacterial enumeration For estimating fluorescent pseudomonads in samples taken either from springs or the littoral zone, syringes were used to collect and dispense aliquots on sterile, 0.45 um black, membrane filters (HABG, Millipore, Bedford, MA U.S.A.) and these were incubated on agar media. Thiobacilli were estimated directly on agar spread plates. 94 In the case of plant materials, samples were collected in sterile, 500 m1, widedmouthed bottles. The plant materials were ground in a Waring Blender (previously rinsed with 70% ethanol) and then homogenized (using a sterile ground-glass homogenizer). The dry weight of plant material per milliliter was determined by drying 5 m1 aliquots in aluminum containers at 110 C. Bacterial numbers I were then determined as described above. 1-1 Direct counts of the total number of bacterial cells found in the littoral zone were obtained by drying a 1.0 ul aliquot from each sample on a formvar coated 300 mesh electron microscope grid. Each grid was counted using a transmission electron microscope. The number of cells per milliliter was calculated from the area of the drop, the area of open spaces within each grid, and the number of cells per space. Enumeration of morphological types Within 5 h after collection the samples were centrifuged and the pellet transferred to agar coated slides. These slides were examined by phase microscopy and photographed. One slide was pre- pared for each sample and ten microscopic fields were photographed for each slide. Each field contained 10 to 500 cells with an average of approximately 100 cells per field. The population data are reported for each morphological type as percent of total cells observed; this has been referred to as congruity by Terborgh (24). Sulfur granules in Thiopedia cells were easily observed by microscopy after removing the gas vacuoles by pressing on the cover slip or by 95 centrifuging. The use of photography was necessary for more accurate size determination and allowed the data to be recorded immediately after sampling before morphological and numerical changes could occur. Chemical assayg Sulfide was determined by the method of Pachmayr as described by Brock et al. (4). Serum bottles containing 2 ml of 0.2% zinc acetate in 0.2% acetic acid were used to collect samples. One tenth milliliter of phenylenediamine reagent was added to 10 m1 of sample. The tubes were mixed, 0.1 m1 of ferric reagent was added, and the tubes again mixed. The absorbance was read at 670 nm against the appropriate blank using a Beckman DB-G spectrophotometer. A standard curve using NaZS-9H20 was prepared for each assay. Known quantities of NaZS added to hypolimnetic samples were accurately determined using this assay. Oxygen was determined by a modification of the Winkler method as described by Mackereth (17). Chlorophyll was determined on the pellet of cells obtained from centrifuging 37 ml of lake sample. The pellet was lysed by the addition of 5 ml of H S saturated acetone (15). The extract was then 2 recentrifuged or passed through an acetone insoluble membrane filter with a pore size of 0.22 n (Millipore UHW PO4700, Bedford, MA) to remove particulate debris. Spectra from 400 to 700 nm.were obtained using a Beckman DB-G scanning spectrophotometer. The concentration of bacteriochlorophyll was calculated using the method of Stanier and Smith (23). 96 Elemental sulfur was determined as described in Chapter II. Habitats Numerous lakes and ponds located within a 50 mile radius in southern Michigan were surveyed in this study as well as a sulfur spring in the Florida Keyes ("Flowing well", John Pennekamp Coral Reef State Park). The location and characteristics of these lakes are shown in Tables 1 and 2. The reported data on community structure are from samplings done on 7/4/71 for Wintergreen Lake and 8/23/72 for Burke Lake. Similar studies were done on 6/26/71 and 8/21/72 for Wintergreen Lake and 7/16/72, 8/1/72 and 9/6/72 for Burke Lake. Qualitative observations were also made from 1969 to 1974. RESULTS AND DISCUSSION The distribution of sulfur-oxidizing bacteria in hypolimnia Three distinct microbial communities were recognized in the hypolimnia of Wintergreen and Burke Lakes. The upper layer was slightly turbid and red, contained purple sulfur bacteria and is referred to as community A. The next lower layer appeared green, contained green sulfur bacteria and is referred to as community B. The lowest layer occurred within 1 m of the sediments, was turbid and white, contained colorless bacteria, and is referred to as community C. The two ponds studied contained only community A. Microstratification of populations of bacteria occurred within the upper two communities of photosynthetic bacteria as predicted by van Niel (26) from studies on enrichment cultures. In Burke 97 .Aeomav uoohoum ho>uom axed .soauo>uom:oo mo usoauuomoa sowfinowz ecu use .ANomHV hufimum>fiaa oumum damaged: .uooaeoao>on monsomom mo unoauumeon one an eoaaeeom muons mNmuzaeuzma souseao oeeouuoso mosses e.HH «e.s uses usher nNmH3OHmumHH ooumemamx II II II m.H~ Aoavefiav moxmg money Nmmuzamumaw ooumamamm II II II II xoouo muoowoe NNmuzmmumHH oonmamamx II xoouo 0.a No.0 econ mm Nmmuzmmumaa ooumEmme II xoouo 0.a mo.o vsom ummuom mmuzmmumafi ooumemamx oanouomom II N.¢ m.HH oxen x059 wmuzmmumaa oonmEMHmM canouomem II m.o m.nH axed moouwuouaaz mumuzamuzay kuumm ownouom II II N.H axed monousmu ommuzmmuzah huuom II II II ¢.H mom meson 3OHMAZmH meson II II II II mwsfiuqm sexes» 3nmquH auumm oanoumm II II o.Ho~ oxen Hana mamuznmnzHH henna oHnouomsm wawuam e.s H.OH mama meammmo meofiumooa mussoo dowaaaaonam mouoom d8 .nuaov .mm: .mou< mamz canamuwoou amwwnowz Hausa asaaxmz eofieaum mumUHnms owumavm who mo mofiumfiuouomumno was QOHumooA .H manna 98 oav Hv so H.H Aucmawvom uponmv vaom mm ooo.m av eNN e.s Aooomuoo usoaeooov oeoe em oe Hv on o.o mwoaunm sexes» ooe.w Hv on o.o mom mouse ooe.e Hv we o.o mom xenon cod Hv III o.o 03mg Haze oww Hv III o.o oxmq monou3mq 0H Hv III o.o xoouo oumaws< oeo Hv on ~.o axed goon co av mm «.0 use; smouwuousfiz om Hv sea N.H moan; mouse OOH Hv Hm o.~ magma money oomsm Hv hmm o.m oxmq oxunm mloa x mouomuoooud annoy mIoH x “Hawomnofisa om H\ma Imm H\wa madamm .aowumnmamxo How uxou mom .amwwsofiz auoumoenuaom nH moxma scum moaaaow Hausa ma ea monomumooum Houou new .HHHHUdAOfinu .om . mm mo dofiumuuaooaoo .N canny 99 Lake the microstratification of photosynthetic bacteria exhibited ‘ the following sequence (from the thermocline to the sediments): Thiospirillum, Chromatium, Thiocystis, Pelodictyon and Chlorochromatium. In Wintergreen Lake Thiopedia, Thiocystis, and Clathrochloris occurred in sequential layers while Prosthecochloris was randomly distributed within the layer of Clathrochloris. The distributional pattern of the three major communities shown in Figures 1 and 2 corresponds to the ecotone model of Terborgh (24), which is a description of biological distribution along physical gradients. In biological systems possessing this characteristic the spread of populations is blocked by habitat discontinuities in contrast to the gradient and competition models in which distribu- tion is controlled either directly by physical characteristics varying continuously and in parallel with the gradient or by competi- tive exclusion (24). This implies that the distribution of hypo- S, O limnetic bacteria may not be simply controlled by H or light 2 2 as previously suggested (18,26), since they each varied continuously, but rather by a function, F(H2S, 0 light, x...), which is discon- 2, tinuous through the hypolimnion. From Figures 1 and 2 it can be seen that a massive change from a purple sulfur flora to a green sulfur flora occurred in less than 0.1 m (Figure 1), while sulfide, oxygen, and light varied continuously. The data shown in Figures 1 and 2 as well as the data from four similar studies suggest that the anaerobic bacteria from the hypo- limnia of these lakes possess distinct characteristics held in common as well as unique characteristics which can be used to distinguish the three major communities. 100 Figure 1. Some physical, chemical and biological parameters of Wintergreen Lake on 7/4/71, and Burke Lake on 8/23/72 as a function of depth. The organisms shown are representatives of the A, B, and C communities which occur below the thermocline in sequential layers. Meters From Surface 101 Wintergreen lake 2 s o 2 O c «“9"». “five; ,6 3 / 4 o 5 i ‘I ' I l I l " I’d? m.’ I '0 0 II 1.0 '0 20 total total . " Total 5 . s Burke la e .......m~;“...e«-o~.. 9 ‘ ' R.,/I lug/l ' ' toceI " tom ' total Figure l to”, Cllld I 'mglI l 102 Figure 2. The community structure of Wintergreen Lake on 7/4/71, and Burke Lake on 8/23/72. lotus m. Surface 2- 3 4 5 ‘ "' Mg L= é ? ="' ; == 3 5 E. -".= = T : = ? —- 2. E ? E — . , - . — O - — . ~= ‘ g = g =' ' é :- : i I z -' Percent of total Count: 103 Figure 2 1-’ “00.04.00..." to... mflrd- |I|I|4II-I.I-II..m....m| u _ - - - - - e I - o - - 104 The organisms possessed gas vacuoles or flagella (with the possible exception of T2 since flagellation was not determined). In all cases, samples taken from the hypolimnia clumped and migrated in an irregular fashion (bioconvection patterns) as observed by Pfennig (18). The clumping of the cells could be reversed by adding NaZS to a concentration of 1 g/l. Since the organisms were capable of reversible (upward and downward) migrations, possessed either flagella or gas vacuoles (but not both), and occurred in discrete layers, the assumption that gas vacuoles are used to regulate bouyancy and hence position in the water column is reasonable. However, diurnal migrations did not occur in the hypolimnion as they do in the epilimnion (this did not include the movements of inver- tebrates [Stentor sp.] which were observed feeding on purple bacteria). The slow rate of movement of Chromatium’and Thiospirillum, less than 10 mm/h (25), makes the possibility of daily bacterial migra- tions unlikely. The layers shown in Figures 2-9 were stable from day to day and week to week as evidenced by the qualitative observa- tion of samples throughout a 24 h period and by the distributions shown in Figure 3 for samples collected one week apart. Over periods of weeks, the layers of bacteria expanded and contracted,but the relative position of the three major communities described below remained constant. The density of each community varied accordingly and reached a maximum of 108 cells/ml when the layers of green bacteria were compressed during the winter months in Burke Lake. The three communities could be separated on the basis of density by centrifuging a concentrated sample at 7,000 g for 10 min. The gas 105 Figure 3. The fraction of the bacterial community in Wintergreen Lake composed of photosynthetic bacteria (A ), Thiopedia sp. (0 ) and Thiopedia sp. containing intracellular sulfur granules ((3 ) versus depth. Data for graph A were obtained on 7/4/71, and for graph B on 6/26/71. FRACTION 0F GOIIUNI‘I’Y 106 Figure 3 0.. 107 Plate 1 Burke Lake hypolimnetic water samples (phase micrographs). All cells were centrifuged to remove gas vacuoles. Bars represent 1.0 pm. Figure 4. 8.5 m, Community A. Cells exceed 2 um in diameter, have either gas vacuoles or flagella (but not both) and are purple sulfur bacteria with intracytoplasmic vesicles or membranes. Shown are Thiospirillum (1), Chromatium (2), and Thiocystis (3). Figure 5. 9.0 m, Community B. Cells are less than 2 pm in diameter, have either gas vacuoles or flagella (but not both) and are green sulfur bacteria with peripheral vesicles. Shown is Chlorochromatium aggregatum (4). Figure 6. 9.5 m, Community C. Cells are less than 2 um in diameter, have gas vacuoles and are colorless bacteria. Shown are morphological types T3(5), T5(6), and T6(7). 108 V’ \U: Li )1! 5y ‘ TI“ :I; I fir '11,, 0 109 Plate 2. Wintergreen Lake hypolimnetic water samples (phase micro- graphs). All cells were centrifuged to remove gas vacuoles. Bars represent 1.0 pm. Figure 7. 3.0 m, Community A. Cells exceed 2 um in diameter, have either gas vacuoles or flagella (but not both) and are purple sulfur bacteria. Shown is Thiopedia (1). Figure 8. 4.3 m, Community B. Cells are less than 2 pm in diameter, have either gas vacuoles or flagella (but not both) and are green sulfur bacteria. Shown is Clathrochloris. Figure 9. 5.7 m, Community C. Cells are less than 2 pm in diameter, have gas vacuoles and are colorless bacteria. Shown are the morphological types T4(4), T5(3) and T6(2). 110 Plate 2 lll vacuoles collapsed and the organisms formed a pellet with green (comunity B), white (community C) and red (community A) layers occurring sequentially from top to bottom. The occurrence of the first community below the thermocline (community A) has been thoroughly documented by Genovese (11), although he did not describe the organisms which caused the presence of the fired water" which he observed. These organisms were 2 um in diameter or larger, purple sulfur bacteria, and possessed internal vesicles or lamellae. Although the A community may have been depressed well below the thermocline it was always the first community of anaerobic bacteria. In Burke Lake (Figures 2 and 4) this layer was numerically dominated by Thiospirillun (8.5 m, 50% of the population), Chromatium (8.6 m, 4%), and Thiocystis (8.7 to 9.0 m, 40%). In Wintergreen Lake (Figures 2 and 7), Thiopedfla (2.6 to 3.5 m, 90%) was the dominant organism, while in Cassidy Lake both Thiopedia and Thiocystis were present. The fact that all of the organisms present in community A are anaerobes, have unusually low surface to volume ratios (compared to other bacteria), possess internal chromaphores, and are the prevalent organisms in shallow ponds, suggests that 02 tolerance is a factor which gives the A community an advantage at high light intensities where traces of oxygen occur. Oxygen tolerance has also been demonstrated using pure cultures of Chromatium (14). In contrast, the characteristics of members of community B (described below) make them more susceptible to the oxidation of bacteriochlorophyll. 112 Genovese (11) observed that the A camaunity occurred where quantities of HS- ranged from 0.78 mg/l to 49.5 mg/l while Vetter (27) found Thiopedia within a HS- range of 0.5-0.9 mg/l. In our study the A community occurred in habitats containing 0-2 mg/l HS-. Schegg (22) found that 0 concentrations of less than 1 mg/l were 2 required for the formation of the layer of Thiopedia which occurs in the Rotsee while Vetter found that Thiopedia occurred if 02 concentrations were less than 2.5 mg/l (28). Genovese (11) states that no detectable 02 was present in the "red layer" of Lake Faro. we have similarly found that the A community occurred where there was less than 1 mg/l 02. In Wintergreen Lake the morphology of Thiopedia changed drama- tically as a function of depth (Figure 3). At 3.0 m the bacterial community consisted of 80 to 90% Thiopedfla sp. and all cells were virtually free of intracellular sulfur (So). However, at 3.5 m the photosynthetic portion of the community increased while the numbers of Thiopedia declined and the intracellular accumulations of sulfur began to appear until at 4 m 50% of the Thiopedia cells contained sulfur granules. At this depth 4-8 mg/l of sulfide was present and no oxygen was detected. Below this depth 80-100% of the cells of Thiopedia observed contained sulfur granules. At 5.0 m a second peak of Thiopedia occurred. This peak corresponded to a decrease in sulfide to 2—3 mg/l. Although this peak was much smaller than the one at 3 m, it was reproducible (Figure 3A,B). Possibly, sulfide toxicity is also an important factor in the natural distribution of Thiopedia along with oxygen and light. This effect 113 of sulfide toxicity has been demonstrated previously in enrich- ments of purple and green sulfur bacteria (26). Thus, the accumula- tion of internal granules of sulfur in not a reliable taxonomic character but rather may be a response to excess sulfide. The B community was contiguous with communities A and C. B community organisms were 2 pm or smaller, contained peripheral (chloro- bium) vesicles, were green sulfur bacteria, contained bacteriochloro- phyll d and did not deposit So internally. Chlorochramatium (8.8-9.0 m, 10% of the community as aggregates with 15-25 cells per aggregate) was the dominant representative of the B community in Burke Lake (Figures 2 and 5), while Clathrochloris (3.6-5.0 m, 95%) was dominant in Wintergreen Lake (Figures 2 and 8). The small cell size and peripheral location of vesicles characteristic of the B community could explain its apparently greater sensitivity to oxygen. On the other hand, the peripheral location of vesicles (in the case of Prosthecochloris the vesicles are also located in the prosthecae) and the high surface to volume ratio could result in more efficient light utilization. This would give community B a competitive advantage over community A in deeper zones with lower light intensities. Thus, community A occurs in shallow ponds and the upper hypolimnion where light is more abundant, traces of oxygen are present (less than 1 mg/l) and sulfide is low (less than 2 mg/l) while community B occurs in the lower portion of the hypolimnion where light is limiting, oxygen is undetectable, and sulfide is high (greater than 2 mg/l). It should be recognized that the range of concentrations of sulfide, oxygen and light in which each community can be found is 114 important only as a preliminary observation. In the hypolimnion the relative effect of each factor cannot be determined since light, oxygen and sulfide vary simultaneously. The actual relationship between light, oxygen, sulfide, growth and competition could be determined by using artificial gradients (5,6) which hold all but one (per dimension) of the independent variables constant. The C community was contiguous with community B and the sedi- ments. Cells were 2 um in diameter or smaller, contained gas vacuoles, were colorless and did not deposite internal So. This community of bacteria (Figures 2, 6 and 9) has been overlooked previously because of its close proximity to the sediments. It occurs in a layer generally 0.1 to 0.7 m thick immediately above the sediments. The most notable characteristics are the absence of detectable concentrations of bacteriochlorophyll (except, in some cases, chlorophylls derived from sedimenting cells from upper layers) and the presence of gas vacuoles. The distribution of sulfur-oxidizing_ bacteria in the littoral zone In a study of Quarry Spring (6/21/74) which feeds Burke Lake, the number of fluorescent pseudomonads was determined at three loca- tions between the source of the spring and the lake. The spring feeds a small pool containing Chara, flows across the gravel floor of the quarry, and passes into a second pool (also containing Chara) before entering the lake. The flow rate was 2 l/sec. Between the first and second pool, the sand and gravel on the bed of the stream was covered by macroscopic masses of periphytic diatoms (pennate and centric). 115 The spring and its stream were located entirely within the quarry, thus reducing the possibility of contamination by fecal and soil bacteria. The spring was sampled at its source, at the outlet of the first pool, and at the outlet of the second pool. Samples of diatoms between the pools and Chara from the second pool were also taken. At the mouth of the spring (9 C) there were 8.7 i 1.73 fluorescent colony forming units per milliliter. At the outlet of the first pool (17 C) there were 86.5 i 9.5’a and at the outlet of the second pool (18 C) there were 102 i 17.81 Virtually all of the colonies formed were fluorescent although the medimn was differen- tial rather than selective (Chapter II, Table 1). There were 50,800 i 24’0008 fluorescent colonies formed per gram dry weight in the sample of periphytic diatoms and 34,200 : 19,0003 per gram dry weight of Chara. The number of non-fluorescent colonies found in both the planktonic and the plant samples was insignificant (<2%). This suggests that fluorescent pseudomonads were indigenous, numerically dominant, and associated with aquatic plants and micro- algae in this environment. In a recent survey of 17 pond and lake samples from southwest Greenland (9), Thiobacillus thioparus and Thiobacillus neapolitans were either absent or present in low numbers. In the study of thirteen samples shown in Table 2, the same result was obtained despite the presence of sulfide and elemental sulfur in many of the samples. 895% confidence interval based on 5 duplicate samples. 116 Electron microscopic observation of thin sections (Chapter IV, Figures 6 and 7) of littoral plant materials shows that heterotrophs, Thiothrix, and Thiocystis may be members of the littoral community. Thiothrér attached to plants using its holdfast and Thiocystis attached using its mucilage. The presence of fluorescent pseudo- monads as microcolonies on plant surfaces was noted in Chapter II. The distribution of sulfur—oxidizing bacteria in a sulfur-spring A sulfur spring ("Flowing Well" John Pennekamp Coral Reef State Park) was observed qualitatively from 5/11/73 to 5/14/73. The source of the spring was covered with filaments of Thiothrix. However, filamentous algae were also found in abundance and may have served as a source of organic substrates. Examination of the sandy soil surrounding the spring revealed three layers of photosynthetic organisms including a filamentous alga beneath which two layers of Thiocystis were present. Although Thiocystis occurred among particles of silica, the overlying algae may have provided organic substrates. A comparison of the sulfur spring, littoral zongi and hypolimnion Comparison of the sulfur spring, the littoral zone, and the hypolimnion revealed that aquatic habitats containing reduced sulfur compounds, oxygen, and organic substrates favored the proliferation of fluorescent pseudomonads, Thiothrix or Chromatiaceae. The structure of the community depended on the proportion and concentration of each of these substrates, the form of reduced sulfur, the availability of 117 a substratum for attachment and subsequent formation of a micro- environment modified by bacterial exudates. In the hypolimnion, each bacterial population appeared to maintain its position along continuous gradients of increasing sulfide and decreasing oxygen concentration with increasing depth.u Where sulfide and oxygen existed simultaneously, the large (greater than 2 pm) planktonic Chromatiaceae predominated to the exclusion of all other bacteria while at lower depths where oxygen was depleted the Chlorobiaceae were dominant. In the littoral zone, wave action constantly replenished the supply of oxygen despite the decomposition of plant debris. As a result, even sequestered crevices on the plant surfaces failed to support the growth of Chlorobiaceae and, instead, the oxygen tolerant Chromatiaceae predominated as in the upper regions of the hypolimnion and the sulfur spring. However, in addition to the Chromatiaceae several other groups of bacteria were present in the littoral zone, including Thiothrix and Pseudemonas. These organisms are character~ istic of aerobic sulfide containing environments. The niche of several bacterial groups is shown in Figure 10. The niches shown represent that portion of the habitat hypervolume (29) in which each group of organisms successfully competed. The niches are tentative and incomplete in that only three physicochemical factors, sulfide concentration, oxygen concentration, and the product of substratum availability and washout rate, are considered. However, the figure summarizes the results discussed above and also provides a conceptual perspective for future studies of sulfur-oxidizing bacteria. 118 Figure 10. The relative position of sulfur-oxidizing bacterial communities within the HS‘, 02, and substrata x washout hypervolume (29). Each of the three ranges (1,2,3) shown represents a range of physicochemical conditions affecting bacterial competition. Range 1 represents the potential conditions found in anaerobic hypolimnia where no 02 is present and HS varies. There is no substratum and no washout. The B and C communities are found here. Range 2 represents the potential conditions found in the upper region of anaerobic hypolimnia where 02 and HS vary but HS does not reach the high level sometimes found in sulfur springs or lower hypolimnia. There is no substratum and little washout. The A community is found here. Range 3 represents the potential conditions found in_sulfur springs, the littoral zone, and small ponds where HS varies widely, as do the washout and availability of substrata; how- ever, 02 is higher here than in the anaerobic hypolimnia. Fluorescent pseudomonads, Thiothrix sp. and Chromatiaceae are found here. 119 substrata x washout ’ g - Figure 10 120 From qualitative observations, it appears that in stagnant waters (i.e., hypolimnia), the availability of substrata is not an ' important factor. However, in flowing waters substrata permit colonization by periphytic bacteria which would otherwise be washed from the system despite the existence of a favorable chemical environment. Qualitative observations show that periphytic sulfur- oxidizing bacteria occur in the littoral zone and springs while planktonic bacteria occur in the hypolimnion. For this reason the product of substratum availability and washout rate was included as a physicochemical factor although it was not measured nor the units defined. CONCLUSIONS In all of the sulfide containing waters observed, light was available and photosynthetic bacteria were present. In hypolimnia, flagellate or gasdvacuolate bacteria were found. As light decreased and sulfide increased (with depth), there was an increase in the surface to volume ratio, a transition from internal to peripheral vesicles, a decrease in the carotenoid content of the cells, and an increase in bacteriochlorophyll d. These changes in bacteria ecotype were due to changes in community structure rather than the phenotypic adaptation of individual populations. The correlation of these bio- chemical and morphological transitions with physicochemical gradients of sulfide, oxygen, and light is adequately explained on the basis of light requirements as well as sulfide and oxygen toxicity. 121 In the littoral zone and the sulfur spring, the sulfur-oxidizing bacteria were attached to surfaces either by a holdfast or by mucilage which encased microcolonies. Only the oxygen tolerant or aerobic sulfur-oxidizing bacteria were found in the littoral zone and sulfur spring. Photosynthetic and heterotrophic organisms were found Inn: members of the genus Thiobacillus were not. In situ enrichments using thiosulfate gradients selected fluorescent pseudo- monads rather than thiobacilli and resulted in a three-fold increase in growth rate. Isolate TBT-H and.PseudOmonas aeruginosa were found to excrete a fluorescent sulfide binding compound which resulted in the formation of sulfur (So) globules. However, the metabolic role of sulfide in fluorescent pseudomonads was not determined. The statements of Vishniac (28) and Rittenburg (19) suggesting the importance of heterotrophic and/or mixotrophic bacteria in sulfur oxidation are supported by the results obtained here. Although there have been numerous studies on the mixotrophic capabilities of autotrophic thiobacilli (28,19), the present study indicates that the mixotrophic capabilities of fluorescent pseudomonads should also be investigated. The excretion of a sulfide binding or oxidizing compound(s) would allow aerobic bacterial respiration in sulfide containing waters. Despite the extracellular oxidation of sulfide, energy could still be available for metabolism since the subsequent intracellular oxidation of sulfite or elemental sulfur to sulfate can result in the production of ADP from AMP (16). Thus sulfide might be detoxified by incomplete oxidation extracellularly and then further oxidized intracellularly. This would explain the excretion 122 of a sulfide binding compound(s), the presence of elemental sulfur extracolonially but not intracolonially or intracellularly, and a three-fold increase in growth rate caused by gradients of thiosulfate. However, further physiological studies, both in the laboratory and in situ, will be required before the role of fluorescent pseudomonads in the biogeochemistry of sulfur compounds can be elucidated. The concepts of habitat and niche as defined by Whittaker (29) rely on the gradient concept as discussed in Chapter I. The niche may be determined either by correlating physicochemical conditions with successful competition or by observing the distribution of organisms along environmental gradients. The distributional data summarized in Figure 10 were collected in this way. Figure 10 is thus a conceptual model of the niche hypervolume for various sulfur- oxidizing bacteria. However, using a soil inoculum and artificial two-dimensional gradients, the hypervolume could be determined experimentally. The application of gradient methods to correct and expand the results shown in Figure 10 may thus be the next step in the ecology of sulfur-oxidizing bacteria. LITERATURE CITED 1. Aaronson, A. 1970. Experimental Microbial Ecology. Academic Press, New York p. 115. 2. Anagnostidis, K., and Overbeck. J. 1966. Methane oxidizing and hypolimnetic sulfur oxidizing bacteria. Studies of the ecological biokinetics of aquatic microorganisms. Beri. Deutsch. Botan. Geselt. 73:163-174. 3. Bass-Becking, L. G. M. 1925. Studies on the sulfur bacteria. Ann. Bot. 39:613-651. 10. ll. 12. l3. 14. 15. 123 Brock, T. D., Brock, M. L., Bott, T. L., and Edwards, M. R. 1971. Microbial life at 90 C: the sulfur bacteria of Boulder Spring. J. Bacteriol. 107:303-314. Caldwell, D. E., Lai, S. H., and Tiedje, J. M. 1973. A two- dimensional steady-state diffusion gradient for ecological studies. 29. Modern Methods in the Study of Microbial Ecology. Edited by T. Rosswall. Vol. 17 pp. 151-158. Bull. Ecol. Res. Comm. (Stockholm). Caldwell, D. E., and Hirsch, P. 1973. Growth of microorganisms in two-dimensional steady-state diffusion gradients. Can. J. Microbiol. 19:53-58. Chen, K. Y., and Merris, J. C. 1972. Kinetics of oxidation of aqueous sulfide by 02. Environ. Sci. Tech. 6:529-537. Cooper, R. C. 1962. Photosynthetic bacteria in waste treatment. Deg. Ind. Microbiol. 4:95-103. Fjerdingstad, E., and Berg, 0. 1973. Some characteristic features of the bacterial decomposition in sediments from lakes and ponds in South-West Greenland (the Narssaq area). Arch. Hydrobiol. 72:313-332. Frank, L. H., and DeMoss, R. D. 1959. On the biosynthesis of pyocyanin. J. Bacteriol. 77:776-782. Genovese, S. 1963. The distribution of H28 in the Lake of Faro (Messina) with particular regard to the presence of "Red water". 13 Marine Microbiology. Edited by C. H. Oppenheimer. Charles C. Thomas, Publisher, Springfield, IL pp. 194-204. - Gorlenko, V. M., and Kusnezow, S. I. 1972. Photosynthesizing bacteria of Lake Kononjer. Arch. Hydrobiol. 70:1-13. Holm, H. W., and Venues, J. W. 1970. Occurrence of purple sulfur bacteria in a sewage treatment lagoon. Appl. Microbiol. 19:988-996. Hurlbert, R. E. 1967." Effect of oxygen on viability and substrate utilization in Chromatium. J. Bacteriol. 93:1346—1352. Kaplan, I. R., and Sieberman, H. 1959. Spectroscopy of bacterial chlorophylls separated by paper and cellulose column chromatography. Arch. Biochem. Biophys. 80:114-124. l6. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 124 Lyric, R. M., and Susuki, I. 1970. Enzymes involved in the metabolism of thiosulfate by Thiobacillus thioparus III. Properties of thiosulfate-oxidizing enzyme and proposed pathway of thiosulfate oxidation. Can. J. Biochem. 48:355-363. Mackereth, F. J. H. 1963. Some methods of water analysis for limnologists. Freshwater Biol. Sci. Publ. No. 21. Pfennig, N. 1967. Photosynthetic bacteria. Ann. Rev. Microbiol. 21:285-324. Rittenberg, S. C. 1972. The obligate autotroph - the demise of a concept. Ant. van Leeuwenhoek J. Microbial. Ser. 38:457-478. Rovira, A. D., and Ridge, E. H. 1973. Mcdern Methods in the Study ofUMicrobial Ecology. Edited by T. Rosswall. Bull. Ecol. Res. Comm. (Stockholm). Vol. 17 pp. 329-335. Sands, D. C., and Rovira, A. D. 1970. Isolation of fluorescent pseudomonads with a selective medium. Appl. Microbiol. 20:513-514. Schegg, E. 1971. Production and destruction in the tropho- genic zone. Schweiz. Z. Hydrologie. 33:425-532. Sanier, R. Y., and Smith, J. H. C. 1960. The chlorophylls of green bacteria. Biochim. Biophys. Acta 41:478-484. Terborgh, J. 1971. Distribution on environmental gradients. Theory and a preliminary interpretation of distributional patterns in the avifauna of the Cordillera Vicabamba, Peru. Ecology 52:23-40. Vaituzis, Z. and Doetsch, R. N. 1969. Motility tracks: technique for quantitative study of bacterial movement. Appl. Microbiol. 17:584-588. van Niel, C. B. 1955. Natural selection in the microbial world. J. Gen. Microbiol. 13:201-217. Vetter, H. 1937. Limnologische Untersuchungen uber das phyto- plankton und seine beziehungen zur ernahrung des zoo- planktons in schleinese bei langenargen am bodensee. Int. Rev. Hydrol. 34:499-561. Vishniac, W., and Santer, M. 1957. The thiobacilli. Bacterial. Rev. 21:195-213. Whittaker, R. H., Levin, S. H., and Root, R. B. 1973. Niche, habitat, and ecotope. Amer. Natur. 107:321-338. ERSIT MICIITIIIWINWVSINIWIWWI IHHIIIHYIIWIBIWIWES 3 119.“ (13012 4316