Numerical taxonomy of moderately halophilic Gram-negative bacteria from hypersaline soils

Journal of General Microbiology (1983), 129, 2649-2657. Printed in Great Britain 2649 Numerical Taxonomy of Moderately Halophilic Gram-negative Bacteria from Hypersaline Soils By E. QUESADA,' A. VENTOSA,'? F. RODRIGUEZ-VALERA,* L. MEGIAS3 AND A. RAMOS-CORMENZANA'* I Department of Microbiology, Faculty of Pharmacy, University of Granada, Granada, Spain Department of Microbiology, University of Alicante, Alicante, Spain Department of Anatomy, Faculty of Medicine, University of Granada, Granada, Spain (Received 15 November I982 ; revised 26 January 1983) ~ ~ ~ ~ ~ A total of 132 moderately halophilic bacteria were isolated from hypersaline soils with a C1- content between 2-36 and 12.72% (w/v) located near Alicante (S.E. Spain) and examined for 98 phenotypic characteristics including their response to cytological, physiological, biochemical and nutritional tests. They were submitted to a numerical analysis together with six reference strains using both simple matching (SsM) and Jaccard (S,) coefficients, and cluster analysis was carried out by the unweighted pair group method of association (UPGMA), single linkage and complete linkage. With the S, coefficient and UPGMA clustering, eight phenons were obtained at the 65% similarity level. From each phenon representative strains were chosen for the determination of DNA base composition and for electron microscopy. Bacteria belonging to phenons D, E, and F were assigned to the genus Alcaligenes. Phenon G included 27 strains assigned to Acinetobacter, but the high G + C composition (58.9 mol%) of a representative strain of this phenon suggests that it may represent a new taxon. Phenons A, B, and C were designated Flavobacterium and phenon H was Pseudomonas. The bacteria found in these environments are not related to those from hypersaline waters or normal soils. INTRODUCTION Moderately halophilic bacteria have been defined as those showing optimal growth between 3 and 15% (w/v) NaCl (Kushner, 1978). Although very few groups of moderate halophiles have previously been described (Larsen, 1962; Kushner, 1978), we recently showed that a variety of taxonomic groups had moderately halophilic representatives in hypersaline waters (Ventosa et al., 1982) and soils (Quesada et al., 1982). The lack of systematic studies of isolates from natural sources has resulted in a scarcity of taxonomic information concerning moderate halophiles. Most of the descriptions refer to organisms isolated from salted food (Buchanan & Gibbons, 1974) or found as accidental laboratory contaminants (Novitsky & Kushner, 1976). In hypersaline waters we found most of the genera that can be isolated from seawater : Vibrio was the most abundant, followed by Flavobacterium, Alcaligenes, Alteromonas and Chromobacterium (Ventosa et al., 1982). All these bacteria were moderate halophiles having, in addition to a high (5-10%, w/v) salt requirement for optimal growth, a minimal requirement of 2-5% (w/v) NaCl in most cases. In hypersaline soils, however, although many halotolerant Gram-positive bacteria were found, most of the moderate halophiles were Gram-negative rods. These also required a high (about 5-10%, w/v) salt concentration for optimal growth, but a lower minimal salt concentration (0-5-2 %, w/v) supported growth. These results probably reflect the ecological differences between water and soil environments, water being relatively homogeneous and constant whereas soil is heterogeneous and more affected by factors such as rainfall, which may eliminate over-specialized organisms (Quesada et al., 1982). f Present address : Department of Microbiology, Faculty of Pharmacy, University of Sevilla, Sevilla, Spain. 0022-1287/83/OOOl-0909 $02.00 0 1983 SGM 2650 E. QUESADA AND OTHERS The purpose of this work was to compare moderate halophiles from soils with those isolated from waters, and also to provide a more complete taxonomy of this ecological group. Using numerical methods, we have studied 132 randomly-chosen new isolates of moderate halophiles, together with some reference strains from culture collections. METHODS Zsolation and maintenance of strains. The micro-organisms were isolated from hypersaline soils near Alicante (S.E. Spain) which have been described previously in physical and chemical terms (Quesada et al., 1982). Samples were taken as before at regular intervals from January 1980 to April 1981 (12 samplings from five different sites). The C1- content of the soils (determined using AgNOJ was in the range 2.36 to 12.72%. Isolation medium contained (%, w/v): yeast extract (Difco), 1 ; proteose-peptone no. 3 (Difco), 0.5; glucose, 0.1 ; bacto-agar (Difco), 2. This medium was supplemented with a balanced mixture of sea salts giving final concentrations of 0.5, 5, 10, 20, or 25% (w/v). The salts and proportions used were as in evaporated seawater according to Subov (Quesada et al., 1982). One gram of each soil sample was suspended in 10 ml of a salt solution similar to those used for the culture media, tenfold dilutions were made (always keeping the same balanced salt concentrations) and 0.1 ml of each dilution was plated on the isolation medium of the same salt concentration. After incubation at 32 "C for 15 d in sealed plastic bags, three to five colonies per plate were randomly selected and subcultured on the same medium until pure. A total of 132 isolates of Gram-negative rods that were moderately halophilic, i.e. able to grow in media containing 2-20% (w/v) salts (the salt growth range determined as previously: Ventosa et al., 1982), were selected for this study. Reference strains. 'Chromobacterium marismortui' ATCC 17056, 'Pseudomonas halosaccharolytica' CCM 285 1 and Vibrio costicola NCMB 701 were included. Quotation marks indicate species not in the Approved Lists of Bacterial Names (Skerman et al., 1980). Reference strains from a previous work on hypersaline waters (Ventosa et al., 1982) were included: Alcaligenes sp. A-336, Alteromonas sp. A-387 and Flavobacterium sp. A-101. Maintenance medium. The strains were maintained on slants of medium suitable for moderate halophiles (MH medium) containing 10% (w/v) marine s?lts. The final composition of this medium was (%, w/v): NaCl, 8-1 ; MgC12, 0.7; MgS04, 0.96; CaCl,, 0.036; KCI, 0-2; NaHCO,, 0.006; NaBr, 0.0026 (Rodriguez-Valera et al., 1980) supplemented with 0-5 % (w/v) proteose-peptone no. 3 (Difco), 1 % (w/v) yeast extract (Difco) and 0.1 % (w/v) glucose. This was solidified with bacto-agar (Difco) 2.0% (w/v) when necessary. Characterization of isolates. The phenotypic characteristics (98) examined are given in Table 1. Most of the tests have been described elsewhere (Ventosa et al., 1982); details of the remainder are given below. Incubation was at 32 "C for up to 7 d in sealed containers; all media contained 10% (w/v) marine salts and had their pH adjusted to 7.2 with 1 M-KOH (Rodriguez-Valera et al., 1980). The ability to grow anaerobically was evaluated on plates of MH medium which were streaked and incubated in jars with the Gas Pak Anaerobic System (BBL). The appearance of growth was checked after 15 d incubation. The ability to grow in narrow Weinberg tubes of liquid MH medium containing glucose and phenol red (0.001 %, w/v) was determined by rapid inoculation into the boiled and cooled medium. The tubes were sealed with sterile agar plugs and also covered with liquid paraffin. Phosphatase activity was tested by adding 1 % (w/v) aqueous phenolphthalein diphosphate solution to MH medium. Electron microscopy. Representative strains from each phenon were grown to mid-exponential phase on surface MH plate cultures covered with liquid medium. Samples of the liquid cultures were negatively stained with a 2% (w/v) solution of phosphotungstic acid (pH 7.0) and examined in a Carl Zeiss EM-9s-2 transmission electron microscope. DNA composition. Exponential-phase cells of representative strains from each phenon were ruptured, and the DNA was purified using the method of Marmur (1961). The mol % G + C content was determined from the midpoint of the thermal denaturation profile (Marmur & Doty, 1962). Numerical analysis. Taxonomic characters were coded in a binary form of the presence/absence type. Strain similarities were estimated with both simple matching (SsM) (Sokal & Michener, 1958) and Jaccard (S,) (Jaccard, 1908) coefficients, and cluster analysis was carried out using the unweighted pair group method of association (UPGMA), single linkage and complete linkage (Sneath & Sokal, 1973). Test error was estimated in eight repeated strains (Sneath & Johnson, 1972). Cophenetic correlation was also evaluated (Sneath & Sokal, 1973). These computations were performed using the MINT program of Dr F. J. Rolf of the Department of Ecology and Evolution, State University of New York at Stony Brook, N.Y., U.S.A., using a Univac 1108 computer in the Computer Centre of the University of Granada, Spain. RESULTS Morphological and halophilic characteristics All the strains were Gram-negative rods and obligate aerobes. Conspicuous polyhydroxybutyrate inclusions were observed in the majority of strains. Taxonomy of bacteria from hypersaline soils 265 1 All 132 strains grew optimally in 5-10% (w/v) salts media and are therefore moderate halophiles (Kushner, 1978). Most required at least 2% (w/v) salts for growth (Table 1). The detailed results of all the physiological, biochemical and nutritional tests are summarized in Table 1. Numerical analysis The results of the numerical study of the characteristics of the strains grouped by means of the SJ coefficient and UPGMA clustering yielded the dendrogram shown in Fig. 1. Simple matching coefficient (SsM) and single linkage and complete linkage algorithms were also used, but cluster composition was not markedly affected by either the coefficient or the clustering method. Cophenetic values were 0.9726 and 0-8236 for the SJ and S,, coefficients, respectively. The estimated test error was less than 3.0%, which would not significantly affect the cluster analysis. The majority of the strains were grouped into eight phenons at a 65% similarity level. Eleven strains clustered separately, five of which were reference strains. Table 1 shows the features of the eight phenons, and Table 2 summarizes their differential characteristics. Phenon A. This group included nine strains which were non-motile rods and oxidase positive. They reduced nitrates but not nitrites. Neither proteolytic, lipolytic nor amylolytic activity was shown. The G + C content of strain F8-11, chosen as representative of this group, was 694 mol%. These strains could be included in the genus Flavobacterium (Buchanan & Gibbons, 1974). Phenon B. This phenon, containing only three strains, was very similar to phenon A (Table 1) but these strains reduced nitrites. Strain F8-6, taken as representative, had a G + C content of 61.8 molx. This phenon could also be placed in the genus Flavobacterium (Buchanan & Gibbons, 1974). Phenon C. This group included three strains which were non-motile short rods and oxidase positive. They reduced nitrate and nitrite and utilized a narrow range of compounds as sole source of carbon and energy. Strain G-4 was selected as representative of this group and had a G + C content of 65.9mol%. These strains may be placed in the genus Flavobacterium (Buchanan & Gibbons, 1974). Phenon D. This comprised 16 strains that clustered together because of their similar biochemical and nutritional characteristics ; however, some were assigned to the genus Flavobacterium and others to Alcaligenes. Strains G-23 and F8-10, representatives of each group, respectively, had G + C contents of 65.4 and 69.2 mol%. Reference strain Alcaligenes sp. A-336 appeared in this phenon. Phenon E. Twenty-four strains clustered in this group; all were straight, peritrichously flagellated rods. Most of the strains were oxidase and nitrate negative. None possessed proteolytic or amylolytic activities. Twelve of the strains were phosphatase positive, and this feature differentiated two subphenons related at different similarity levels. The two representative strains selected, G-30 (phosphatase positive) and F9-4 (phosphatase negative), had G + C contents of 66.7 and 67.4 mol%, respectively. Both subphenons can be assigned to Alcaligenes (Buchanan & Gibbons, 1974). Phenon F. The 40 strains included here were long, motile rods with peritrichous flagellation (Fig. 1). They were oxidase positive but not proteolytic nor lipolytic. Some (30%) of the strains hydrolysed urea. Strain F5-7 had a G + C content of 66.7 mol%. All members of this phenon were assigned to the genus Alcaligenes (Buchanan & Gibbons, 1974). Phenon G. This group comprised 27 strains that clustered at the 66% similarity level and constituted a very homogeneous group of predominantly short rods, occasionally with large curved cells and filaments (sometimes rather numerous). Most were non-motile, oxidase negative, H2S producers and reduced nitrates but not nitrites. All were phosphatase positive. 2652 E. QUESADA AND OTHERS Percentage similarity Phenon No. of strains/ 0 20 40 60 80 100 f I I I I I I1 I I 1 i I I I I I I I I ni I 11‘ I I 1 I I I 1 I I1 1 I I I A B C D E F G strain designation 9 3 3 16 Fll-7 24 40 Flauobacterium sp. A-101 27 ‘Chrotnobacterium , marismortui’ - / Alteromonas sp. A-387 -A - / F8-2 H4 F9- 1 G-l I -- \ - F 12-7 -- - __z - F8-4 -1 Vibrio costicola \ ‘Pseudomonas halosaccharolytica’ Fig. 1. Simplified dendrogram showing the clustering of strains into eight phenons based on the S, coefficient and unweighted average linkage clustering (UPGMA), for 132 moderately halophilic Gram-negative bacteria isolated from hypersaline soils and six reference strains. ' Taxonomy of bacteria from hypersaline soils 2653 Table 1. Frequencies of positive characters found in the eight phenons, expressed as a percentage of the total scored to each group for the given test Phenon . . . A 9 No. of strains . . . Morphology Long rods Coccobacillary Colony pigment Cream Yellowish Orange Motility Salts: growth at % (w/v) 0.5 2 15 20 25 30 PH 5 PH 8 PH 9 pH 10 5 "C 15 "C 45 "C Galactose Glucose Lactose Maltose Mannitol Sucrose Xylose pH: growth at Temperature : growth at Acid production from Oxidase H,S production Nitrate reduction to nitrite Nitrite reduction to gaseous compounds Casein hydrolysis Gelatin hydrolysis Starch hydrolysis Tween 80 hydrolysis Urea hydrolysis Methyl red Voges-Proskauer DNAase Phosphatase Utilization of organic compounds as sole source of carbon and energy: Carbohydrates Aesculin L-Arabinose D-Cellobiose D-Fructose D-Galactose Lactose Maltose D-Mannose L-Raffinose L-Rhamnose Salicin L-Sorbose 44 56 66 34 0 0 0 100 90 50 11 11 0 89 54 22 33 89 11 11 77 0 22 0 22 0 89 33 100 0 0 0 0 0 0 0 0 0 0 88 55 11 100 66 33 100 55 0 0 0 33 B 3 0 100 33 66 0 0 0 100 1 00 100 0 0 0 100 100 66 0 100 0 33 0 0 0 33 33 0 100 33 100 100 0 0 0 0 0 0 0 0 0 0 0 0 100 33 0 66 0 0 0 0 0 C 3 0 100 0 100 0 0 0 100 100 1 00 33 0 33 100 100 100 100 100 0 0 33 0 100 0 0 33 100 0 100 100 0 0 0 0 0 0 0 0 0 0 0 0 66 0 0 100 33 0 0 0 0 D 16 63 37 44 56 0 56 12 100 100 94 50 25 0 100 100 75 50 94 56 12 87 0 68 48 75 0 100 0 69 62 0 0 6 6 0 0 0 0 0 6 0 0 94 31 0 I00 50 0 12 0 0 E 24 96 4 88 12 0 96 0 100 96 96 83 40 0 100 96 88 50 100 50 75 50 0 25 4 16 12 20 76 20 0 0 0 0 0 0 4 4 0 54 16 80 20 100 100 0 92 92 0 54 20 4 F 40 100 0 78 22 0 95 10 100 100 1 00 93 93 47 100 93 88 10 100 25 93 93 2 83 10 88 22 80 80 65 2 0 0 2 0 30 0 0 0 0 78 98 90 100 98 10 90 90 2 15 85 0 G 27 4 96 67 7 26 22 85 96 100 96 89 89 41 100 100 96 70 100 48 4 7 0 0 7 0 4 4 89 93 0 4 26 0 0 0 0 0 4 100 100 93 1 00 100 100 100 100 100 93 89 96 37 H 4 100 0 0 100 0 100 50 100 75 50 0 0 0 100 100 25 0 100 0 75 100 0 75 0 0 75 100 50 100 25 0 0 100 100 0 0 0 0 0 0 0 0 100 0 0 100 75 0 0 0 0 2654 E. QUESADA AND OTHERS Table 1 (continued) Phenon . . . No. of strains . . . Starch Sucrose D-Trehalose D-x ylose Alcohols Adonitol Dulcitol DL-Glycerol meso-Inosi to1 D-Mannitol D-Sorbitol Acetate Benzoate Caprylate Citrate Formate Fumarate D-Gluconate Hippurate DL-Lactate DL-Maiate Oxalate Pyruvate Propionate Succinate D-Tartrate Carboxylic acids Utilization of amino acids as sole source of carbon, nitrogen and energy : L-Alanine L-Arginine L- Asparagine L-Aspartic acid L-Glutamic acid L-Histidine L-Isoleucine L-Leucine L-Lysine L-Ornithine L-Serine L-Tryptophan L-Valine A 9 55 77 77 88 88 0 88 88 44 88 88 0 0 88 100 100 100 66 100 100 0 100 88 100 11 100 88 100 100 100 88 33 33 11 11 100 0 0 B 3 0 100 66 100 0 0 66 66 33 100 100 0 0 100 100 100 100 66 100 100 66 100 100 100 66 100 66 100 100 100 0 66 100 0 0 100 33 33 C 3 0 66 100 33 0 0 100 0 0 66 66 0 0 100 100 100 66 0 66 66 0 66 66 100 0 100 66 100 100 100 0 0 0 0 66 100 0 100 D 16 0 100 100 31 0 0 87 31 44 44 100 6 0 94 100 100 75 56 94 100 100 100 94 87 19 100 94 100 94 100 81 56 81 56 81 100 6 25 E 24 32 92 100 60 0 0 96 64 50 54 96 28 0 100 100 100 100 66 88 100 8 100 1 00 100 16 100 100 100 100 100 96 96 80 80 96 100 24 0 F 40 0 98 95 85 2 2 90 0 35 32 95 95 2 93 95 95 93 90 98 98 10 90 98 98 5 95 95 98 93 95 15 20 25 75 57 98 27 2 G 27 4 100 100 100 44 33 100 96 74 89 96 0 18 100 100 100 100 15 15 15 11 100 100 100 89 100 100 67 100 100 100 44 52 67 33 100 7 7 H 4 100 25 25 25 0 0 75 25 0 0 75 0 0 100 100 100 100 25 100 100 0 50 100 100 0 50 25 100 50 100 0 0 0 25 25 25 25 0 All strains were Gram-negative, formed catalase, were strict aerobes, grew at 5 and 10% (w/v) total salts, grew at pH 6 and 7, grew at 25 and 32 "C, and grew on D-glucose as sole source of carbon and energy. None formed indol or grew on inulin or L-cysteine. They were very versatile nutritionally, and were able to utilize a very high number of compounds as sole carbon and energy sources. The strains in this phenon were able to grow at low salt concentrations (0.5 %, w/v) although for optimal growth they needed 5-10% (w/v) total salts. Strain F2-12 had a G + C content of 58.9 mol%. With the exception of G + C value these strains could be assigned to the genus Acinetobacter (Buchanan & Gibbons, 1974). Phenon H. This phenon was composed of four strains, which were straight rods motile by one or a few polar flagella. Thin filaments were common. They were oxidase positive, reduced nitrates and showed amylolytic and lipolytic activities. Few organic compounds were used as sole carbon and energy sources. The representative strain F12-1 had a G + C content of 58.7 mol%, These four strains can be included in the genus Pseudomonas (Buchanan 8z Gibbons, 1974). Taxonomy of bacteria from hypersaline soils 2655 Table 2. Diferential characteristics of the eight phenons of moderately halophilic Gram-negative bacteria from hypersaline soils Phenon ... A B C D E F G H No. of strains.. . 9 3 3 16 24 40 27 4 Morphology Long rods Coccobacillary Motility Flagella Growth at 0.5% (w/v) salt concentration Growth at 5 "C Oxidase H,S production Nitrate reduction to nitrite Nitrite reduction to gaseous compounds Starch hydrolysis Tween 80 hydrolysis Phosphatase Utilization of: DCellobiose Oxalate L-Histidine L-Leucine - - - 3-+ 8++ 3 1- +++ -++ - - - 1- - 1 2- 8- - 3+- - 10 23 + 1+ 9 23 38 6 + -/per. per. per. -/per. pol. 4 23 2 2- 8 12 4 19 - + 5 - 32 1+ - 19 32 24 2 6 1-26- - - - 1- 1-+ 1- - -+ -13-+- - 536+- + 243- - 13g6+- 13 10 14 - - +, all strains positive; -, all strains negative. Numbers indicate number of positive strains; underlined numbers indicate that the number represents 80% or more of the strains. per., peritrichous flagella; pol., polar flagella. DISCUSSION There are few published studies of bacteria from saline soils (Cervantes & Olivares, 1976; Mahmoud et al., 1978), and taxonomic descriptions of such organisms are scarce. Henis & Eren (1963), working on a saline soil near the Red Sea, concluded that more colonies developed on media without salts than on media with 10% (w/v) NaCl. These results differ from ours, probably because these authors used media with NaCl as sole salt, while other ions such as Mg2+ and K+ are also required for good growth of moderately halophilic bacteria (Kushner, 1978). As in our previous study on bacteria from hypersaline soils (Quesada et al., 1982), most of the population living in the soils studied were moderate halophiles as defined by Kushner (1978), since they grew optimally between 5 and 10% (w/v) salts. This was also the case for bacteria from hypersaline waters (Rodriguez-Valera et al., 1981 ; Ventosa et al., 1982). Nevertheless, many bacteria isolated from hypersaline soils were able to grow at low salt concentrations, (about 0.5%, w/v), while bacteria from hypersaline waters could not. This was a common character of the Gram-positive bacteria described by Quesada et al. (1982); however, 31 of the Gram-negative strains included in this study also showed tolerance for low salt concentrations. Although differences existed among the strains included in phenons A, B and C, all can be considered as representatives of the genus FZavobacterium, and the separation among them probably arose from their different capability to use organic compounds as sole carbon and energy sources. Within the genus Flavobacterium, the only moderately halophilic species described to date is F. halmephilum (Elazari-Volcani, 1940). The strains included in these phenons show some resemblance the description of this species, although they differ by their ability to produce acids from sugars, their proteolytic activity and their nutrient requirements. Besides, F. halmephilum has a G + C content of 45.6 mol % and therefore belongs to Section I of this genus, while our strains, with a G + C content in the range 61.8-69.4 mol%, are in Section 11. Among the marine bacteria belonging to the genus Flavobacterium and included in the Approved Lists of Bacterial Names (Skerman et al., 1980), F. oceanosedimentum is the only one that resembles the strains that appear in this study, and it differs from them only by the production of acids from some sugars, the ability to reduce nitrates, and salt response ; the 2656 E. QUESADA AND OTHERS G + C content of F. oceanosedimentum is also very similar (67.5 mol%) to that of the strains in phenons A, B and C. Holmes & Owen (1979, 1981) have proposed a redefinition of the genus Flavobacterium, restricting this genus to the species with low G + C content (31-40 mol %). Since it is clear that flavobacteria of high G + C content are easily isolated from a wide variety of environments (Hayes & Mapp, 1981), more studies are necessary to decide if they could be accommodated in the genus Empedobacter as proposed by McMeekin & Shewan (1978). The strains included in phenons E and F were assigned to the genus Alcaligenes. Phenon E was very similar to the description of the marine bacterium Afcaligenes cupidus (Baumann et al., 1972) except in the ability to reduce nitrate to nitrite and their different salt range for growth. However, phenon F was composed of a large group of bacteria with properties very different from those of previously described species of Alcaligenes. According to the G + C content, they were related to the marine bacterium A. pacificus (Baumann et al., 1972) but there were many physiological, biochemical and nutritional differences from this species. In our opinion, they could constitute a new taxon not previously described. However, more genetic and molecular data are necessary to support this suggestion. The 27 strains that clustered in phenon G had morphological, physiological, biochemical and nutritional characteristics of the genus Acinetobacter (Buchanan & Gibbons, 1974). However, the G + C content of the representative strain of this group was 58-9 mol%, whereas the range accepted for this genus is 42-47 mol % (Henriksen, 1973). Moderately halophilic bacteria have not been described and accepted as members of this genus. Onishi & Hidaka (1978) studied the properties of an amylase produced by a moderately halophilic Acinetobacter sp. but a description of this strain has not been published. It seems clear that the strains in phenon G represent a new group of micro-organisms that could be included in Acinetobacter, but the different G + C content could indicate that they belong to not only a different species but possibly a different genus. Studies are in progress to determine if this difference in the G + C content is also reflected by differences at the phenotypic level. Phenon H was included in the genus Pseudomonas, but these strains were not related to the moderate halophile P. halosaccharolytica used a marker in this study (Fig. 1). There were differences in nutritional versatility as well as in amylolytic, lipolytic and proteolytic activities. With respect to marine species of the genus Pseudomonas included in the Approved Lists (Skerman et al., 1980), they could be related to P. doudorofii (Baumann et al., 1972), the difference being their lipolytic and amylolytic activities, absent in this species. Comparison with previous studies (Ventosa et al., 1982) suggests that the bacterial populations in the saline terrestrial environments studied here show many differences from bacteria from hypersaline waters. In fact, the reference strains from hypersaline waters did not group with the strains from the terrestrial habitat. All the strains studied here had a strict aerobic metabolism, a predominant feature that we had already reported for bacteria from hypersaline soils: very few strains were facultatively anaerobic (Quesada et al., 1982). This point has been confirmed by an independent study carried out by Dr M. E. Rhodes-Roberts (personal communication). However, in hypersaline waters facultative anaerobes represent a major part of the population (Ventosa et al., 1982). In aqueous saline environments Vibrio spp. were the most abundant group of moderately halophilic bacteria, whereas this genus is very scarce in hypersaline soils, which are dominated by the genera Alcaligenes and Flavobacterium. Pseudarnonas is a very common inhabitant of normal soils (Alexander, 1961), but scarce in the hypersaline soils we studied. Nevertheless, the microflora of these hypersaline soils is much more similar taxonomically to the microflora of normal soils (Alexander, 1961) than it is to the microflora of hypersaline waters. This suggests that the general features of the environment may be more important than individual factors such as high salinity in determining the microflora of a particular habitat. The authors thank Dr A. Vila for her invaluable help with the computation and the Computer Centre of the University of Granada, Spain, for the use of its computer and facilities. We are grateful to Dr F. Ruiz-Berraquero for his advice during the course of this work, Dr C. Calvo for the G + C content determination, J. Muiioz for his aid in the preparation of the EM grids, and especially to Dr M. E. Rhodes-Roberts (Department of Botany and Taxonomy of bacteria from hypersaline soils 2657 Microbiology, University College of Wales, Aberystwyth, U.K.), who kindly confirmed some of the results of this work and gave valuable advice and discussion. E. Q. gratefully acknowledges the receipt of a fellowship from the Plan de Formacibn del Personal investigador. REFERENCES ALEXANDER, M. (1961). Soil Microbiology. Edited by J. Wiley. New York: Blackwell. BAUMANN, L., BAUMANN, P., MANDEL, M. & ALLEN, R. D. (1972). Taxonomy of aerobic marine eubacteria. Journal of Bacteriology 110, 402429. BUCHANAN, R. E. & GIBBONS, N. E. (editors) (1974). Bergey’s Manual of Determinative Bacteriology, 8th edn. Baltimore: Williams & Wilkins. CERVANTES, L. A. & OLIVARES, J. (1976). Estudio microbiol~jgico de un suelo salino ‘‘El Salitre”. Revista latinoamericana de microbiologia 18, 73-76. ELAZARI-VOLCANI, B. (1940). Studies on the microflora of the Dead Sea. Ph.D. thesis, Hebrew University, Jerusalem. HAYES, P. R. & MAPP, S. J. (1981). Taxonomic studies on high G + C % flavobacteria. In The Flavobacteria-Cytophaga Group, pp. 33-38. Edited by M. Reichenbach & 0. B. Weeks. Weinheim: Verlag Chemie. HENIS, Y. & EREN, J. (1963). Preliminary studieson the microflora of a highly saline soil. Canadian Journalof Microbiology 9, 903-904. HENRIKSEN, S. D. (1973). Moraxella, Acinetobactm and the Mimeae. Bacteriological Reviews 37, 522-561. HOLMES, B. & OWEN, R. J. (1979). Proposal that Flavobacterium breve be substituted as the type species of the genus in place of Flavobacterium aquatile and emended description of the genus Flavobacterium : status of the named species of Flavobacterium. International Journal of Systematic Bacteriology 29, 41 6 426. HOLMES, B. & OWEN, R. J. (1981). Emendation of the genus Flavobacterium and the status of the genus. Development after the 8th edition of Bergey’s manual. In The Flavobacterium-Cytophaga Group. Edited by M. Reichenbach & 0. B. Weeks. Weinheim : Verlag Chemie. JACCARD, P. (1908). Nouvelles recherches sur la distribution florale. Bulletin de la Sociktk vaudoise des sciences naturelles 44, 223-270. KUSHNER, D. J. (1978). Life in high salt and solute concentrations : halophilic bacteria. In Microbial Life in Extreme Environments, pp. 31 1-368. Edited by D. J. Kushner. London: Academic Press. LARSEN, H. (1962). Halophilism. In The Bacteria, vol. IV, pp. 297-336. Edited by 1. C. Gunsalus & R. Y. Stanier. London : Academic Press. MAHMOUD, S. A. Z., EL SAWY, M., ISHAC, Y. L. & EL SAFTY, M. M. (1978). The effects of salinity and alkalinity on the distribution and capacity of N2- fixation by Azotobacter in Egyptian soils. Environmental role of nitrogen-fixing blue-green algae and symbiotic bacteria. Ecological Bulletin 26, 99-109. MARMUR, J. (1961). A procedure for the isolation of deoxyribonucleic acid from microorganisms. Journal of Molecular Biology 3, 208-218. MARMUR, J. & DOTY, P. (1962). Determination of the base composition of deoxyribonucleic acid from its thermal denaturation temperature. Journal of Molecular Biology 5, 109-1 18. MCMEEKIN, T. A. & SHEWAN, J. M. (1978). Taxonomic strategies for Flavobacterium and related genera. Journal of Applied Bacteriology 45, 321-332. NOVITSKY, T. J. & KUSHNER, D. J. (1976). Planococcus halophilus sp. nov., a facultatively halophilic coccus. International Journal of Systematic Bacteriology 26, 53-57. ONISHI, H. & HIDAKA, 0. (1978). Purification and properties of amylase produced by a moderately halophilic Acinetobacter sp. Canadian Journal of Microbiology 24, 1017-1023. QUESADA, E., VENTOSA, A., RODRIGUEZ-VALERA. F. & RAMOS-CORMENZANA, A. (1 982). Types and properties of some bacteria isolated from hypersaline soils. Journal of Applied Bacteriology 53, 155-161. RODRIGUEZ-VALERA, F., RUIZ-BERRAQUERO, F. & RAMOS-CORMENZANA, A. (1980). Isolation of extremely halophilic bacteria able to grow in defined inorganic media with single carbon sources. Journal of General Microbiology 119, 535-538. RODRIGUEZ-VALERA, F., RUIZ-BERRAQUERO, F. & RAMOS-CORMENZANA, A. (198 1). Characteristics of the heterotrophic bacterial populations in hypersaline environments of different salt concentrations. Microbial Ecology 7, 235-243. SKERMAN, V. B. D., MCGOWAN, V. &SNEATH, P. H. A. (1980). Approved lists of bacterial names. International Journal of Systematic Bacteriology 30, 225- 420. SNEATH, P. H. A. & JOHNSON, R. (1972). The influence on numerical taxonomic similarities of errors in microbiological tests. Journal of General Microbiology 72, 373-392. SNEATH, P. H. A. & SOKAL, R. R. (1973). Numerical Taxonomy. The Principles and Practice of Numerical ClassZfication. San Francisco: W. H. Freeman. SOKAL, R. R. & MICHENER, C. D. (1 958). A statistical method for evaluating systematic relationships. University of Kansas Science Bulletin 38, 1409- 1438. VENTOSA, A., QUESADA, E., RODRIGUEZ-VALERA, F., RUIZ-BERRAQUERO, F. & RAMOS-CORMENZANA, A. (1982). Numerical taxonomy of moderately halophilic Gram-negative rods. Journal of General Microbiology 128, 1959-1969.