Vol. 46, No. 5APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Nov. 1983, p. 1207-12130099-2240/83/11 1207-07$02.00/0Copyright © 1983, American Society for Microbiology

Effect of Acidity on the Composition of an Indigenous SoilPopulation of Rhizobium trifolii Found in Nodules of Trifolium

subterraneum L.tMUKTAR H. DUGHRI AND PETER J. BOTTOMLEY*

Departments of Microbiology and Soil Science, Oregon State University, Corv allis, Oregon 97331-3804

Received 3 June 1983/Accepted 31 August 1983

Acidity affected which members of an indigenous soil population of Rhizobiulmtrifolii nodulated Trifolium subterraneum L. cv. Mt. Barker. In three experimentsinvolving plants grown either in mineral salts agar adjusted to pH 4.8 or 6.8 andinoculated with a soil suspension or grown directly in samples of unamended soil(pH 4.8) or soil amended with CaCO3 (pH 6.4), 121 of 151 isolates of R. trifoliiwere placed into four serogroups. Seventy-nine of these isolates were placed intotwo serogroups (6 and 36) whose nodulating ability was affected by the pH of theplant root environment. Representatives of serogroup 6 occupied the greatestpercentage of the nodules at the low pH in both mineral salts agar (77%) and inunlimed soil (47 and 57%). The same serogroup was a minor nodule occupant atthe higher pH in mineral salts agar (0%) and in limed soil (0 and 10%). In contrast,serogroup 36 was virtually absent in nodules formed at the low pH, whereas it wasthe dominant serogroup at the higher pH in both mineral salts agar (32%) and inlimed soil (35 and 49%). Despite the isolates from within each serogroup beingantigenically identical, separation of cellular proteins by sodium dodecyl sulfate-polyacrylamide gradient gel electrophoresis revealed four and six different geltypes within serogroups 6 and 36, respectively. Isolates represented by one or twogel types dominated the contribution of each serogroup to the nodule population.Further evidence for differences between isolates within each gel type were

revealed from measurements of symbiotic effectiveness.

Inhibition of the growth of tropical and tem-perate legume species by soil acidity is welldocumented. Legume nutrition per se can beinhibited by soil acidity-related factors such asexcessive levels of hydronium, manganese, andaluminum ions and deficiencies of calcium,phosphorus, and molybdenum (7, 19). Evidencehas been presented that many legumes are moresensitive to acidity when dependent on N2 fixa-tion than when grown on combined nitrogen (2).This has been explained as being due to thesensitivity to acid conditions of either nodulefunction (21, 22) or the initiation and develop-ment of nodulation (14, 18). It is not surprising,therefore, that amelioration of soil acidity withlime is often (though not always) beneficial tothe growth of many legumes (20). Indeed, evensmall amounts of lime drilled into acid soils withinoculated seeds of red (1) and subterranean (15)clovers can be as effective at stimulating nodula-tion as 50 times the amount of lime applied to thebulk soil.

t Oregon State University Agricultural Experiment StationTechnical Paper no. 6852.

Amending acid soils with lime has been shownto enhance the survival and multiplication ofRhizobium trifolii both in the absence (38) and inthe presence (8, 17, 31) of the host plant. Cir-cumstantial evidence has been presented thatsoil acidity can influence selective nodulationwithin a soil-borne population of R. trifolii. Itwas shown that a higher percentage of effectiveisolates was found in nodules on uninoculatedplants of white clover (Trifolium repens) sowninto limed soil than was found in unlimed soil(11). Studies performed with mineral salts agarunder bacteriologically controlled conditionsshowed that acidity influences the relative num-bers of nodules formed by an equal mixture of aneffective and an ineffective strain of R. trifolii. Incontrast to the field studies cited above (11), theeffective strain formed the majority of nodules atthe lower pH values, whereas the ineffectivestrain made up the majority at the higher pHvalues (10, 32). The objectives of this researchwere: (i) to utilize three complementary methodsof strain identification to delineate the composi-tion of the nodule occupants found on subclover(Trifolium subterraneum L.) cv. Mt. Barker,

1207

on March 31, 2020 by guest

http://aem.asm

.org/D

ownloaded from

1208 DUGHRI AND BOTTOMLEY

sown as uninoculated seed into an acidic soilharboring an indigenous population of R. trifolii,and (ii) to determine whether there was aninfluence of soil acidity upon the members of theindigenous population of R. trifolii that wouldnodulate subclover.

MATERIALS AND METHODS

Soil and site description. The soil used was a deep,well-drained alluvial silty clay loam of the Abiquaseries, a member of the family of fine, mixed, mesicCumulic Ultic Haploxerolls. The soil was under apermanent grass pasture and contained a small amountof subclover cv. Mt. Barker. The pasture had beensown with inoculated seed of this legume several yearsbefore our study. Three sampling areas (10 by 5 m)were selected randomly within the pasture, and theirpositions were coordinated to fixed positions on thefence line for future reference. From each area, eightsoil cores (0.8 kg of soil per core) were removed to thelower limit of the Al horizon (0.2 m) with a sterileauger. A distance of 1 m separated the location of eachcore. The eight soil samples from each area werebulked and thoroughly mixed and while moist (13%[wt/wt] water) were passed through a 2-mm sieve. Thespecific soil characteristics were as follows: pH 4.8 ±

0.1; organic matter, 6.2%; cation exchange capacity,29 cmol (NH4+) kg of soil-'; base saturation, 63%;extractable phosphorus, 14 mg kg of soil-', calcium,6.4 cmol kg-'; magnesium, 2.6 cmol kg-'.

Soil dilution experimental protocol. Seeds of T. sub-terraneum L. cv. Mt. Barker were surface sterilizedby standard procedures (36) and germinated on wateragar plates for 2 days at 30°C. Seedlings were trans-ferred aseptically to sterile test tubes (30 by 3.0 cm)containing 20 ml of a mineral salts medium solidifiedwith 15 g of Bacto-Agar per liter (Difco Laboratories,Detroit, Mich.). The pH of the medium was adjustedto either 4.8 or 6.8. The concentrations of the con-

stituents were (in grams per liter): CaCl2, 0.6;MgSO4 * 7H20, 0.2; NaCl, 0.1; ferric citrate, 0.05; and10 ml of a trace element mixture (3). Ferric citrate wasdissolved in a small quantity of HCI before addition tothe growth medium. The protocol for setting the pH ofthe mineral agar was as follows. K2HPO4 (0.5 g) andKH2PO4 (0.25 g) were dissolved in 100 ml of distilledwater, autoclaved, cooled, and added to the remainder(900 ml) of the sterilized growth medium. The final pHof the complete medium was checked and adjusted ifnecessary to pH 6.8 with small amounts of sterile HCI.To adjust mineral salts agar to the low pH, KH2PO4(final concentration, 0.75 g liter-') was used as thesole phosphate salt, autoclaved separately, and addedto the remainder of the sterile growth medium. Thefinal pH was adjusted to 4.8 with sterile HCI.

Fifty grams of one of the soil samples was dispensedinto 475 ml of sterile 0.15 M NaCl and successivelydiluted over the range 10-2 to 10-6. Portions of eachdilution (0.2 ml) were used to inoculate each of fourseedlings at the two pH values. The seedlings were

grown in a completely randomized design in a glasshouse where the temperature range was 18°C (night) to27°C (day) and natural daylight was supplemented with16 h of illumination provided by F48T12/D/RS fluores-cent lamps hanging 0.5 m above the growth tubes.

Irradiance at plant height averaged 80 W m-2. Steriledistilled water (3 ml) was added to each seedling 21days after inoculation. The experiment was terminatedafter 42 days, and the tubes were scored for nodula-tion. The most probable number of R. trifolii wasdetermined as described previously (36). The most-probable-number values for two determinations ondifferent soil samples gave the same result indepen-dent of the pH of the seedling agar (5 x 103 and 8 x 103R. trifolii per g of dry soil). Two plants were pickedfrom the 10-2 dilution series of each pH treatment, andall the nodules were recovered. Low numbers ofnodules were formed on plants in the 10-' dilutionseries, and therefore the 10-2 series was chosen foranalysis of nodule occupants.

Soil core experiments. Two experiments (1 and 2)were performed on two different composite soil sam-ples. The protocol for each experiment was as follows.The soil sample was mixed while moist, and half of thesoil received sufficient analytical grade fine CaCO3 toraise the pH to 6.4 as determined by lime requirementmeasurements. The other half of the soil sample wasunamended. The soil samples were brought to andmaintained at a water potential of 30 kPa (23.8 g ofwater per 100 g of dry soil) and equilibrated for 8weeks. The pH of the soil was measured in a 1:2 soil-water suspension at weekly intervals until it stabilizedat pH 6.4 ± 0.1. Approximately 1 kg of the CaCO3-amended soil was potted into each of two plastic pots,and another two pots were filled with identical quanti-ties of the unamended soil. Ten surface-sterilizedseeds of subclover cv. Mt. Barker were sown into eachpot, which was covered with a plastic petri dish lid.The pots were placed under the glass house conditionsdescribed above. After 10 days, the lids were re-moved, and the soil was amended with sterile water asrequired to maintain a water potential of 30 kPa. Theplants were thinned to one per pot 4 weeks aftersowing. Twelve weeks after sowing the plants wereremoved, the roots were washed off in distilled water,and as many nodules as possible were recovered fromthe tap root and lateral root systems of the two plantsfrom each of the soil treatments. Experiment 2 differedonly in that three pots of soil were used for eachtreatment so that a larger collection of nodules, andsubsequently isolates, could be obtained.

Isolation of R. trifolii. The nodules from each experi-ment were surface sterilized and squashed, and thecontents were streaked onto plates containing yeastextract-mannitol medium (36). Single colonies werepicked, restreaked, and transferred to agar slants andgrown at 30°C before storage. The symbiotic effective-ness of the isolates was determined by inoculating 1-week-old seedlings of subclover cv. Mt. Barker with 5x 108 cells of each isolate. Nitrate-fertilized plantswere used as controls and received 3 ml of a sterilesolution of 18 mM KNO3 at 7 and 21 days afterinoculation. The seedlings were arranged in a random-ized complete block design with four replicates of eachisolate. After 35 days of growth under the conditionsdescribed above, the shoots were removed, dried at60°C for 10 days, and weighed. Analysis of variancewas carried out on the data and the Duncan newmultiple range test was used to compare the symbioticeffectiveness of the isolates relative to the nitrate-grown plants.Methods of identification. (i) Growth media. Isolates

APPL. ENVIRON. MICROBIOL.

on March 31, 2020 by guest

http://aem.asm

.org/D

ownloaded from

EFFECT OF ACIDITY ON NODULE POPULATIONS 1209

were grown routinely in broth culture and in a minimalmedium (MGA) composed of the following (in gramsper liter): mannitol, 10; L-(+)-glutamic acid, monoso-dium salt, monohydrate, 1; K2HPO4, 0.5; MgSO4 '

7H20, 0.2; NaCl, 0.1; calcium pantothenate, 0.01;thiamine hydrochloride, 0.01; biotin, 3 x 10-4; and 10ml of a trace element mixture (3). The pH of eachmedium was adjusted to pH 6.8 before autoclaving.

(ii) Preparation of cell extracts. Each isolate wasgrown in 40 ml of MGA at 30°C to the late exponentialphase of growth (optical density at 560 nm, 0.8;equivalent to 6 x 108 cells per ml). Cultures wereharvested by centrifugation at 4°C for 20 min at 25,000x g. Cell pellets were washed three times in ice-cold10 mM Tris-hydrochloride, pH 7.6, with centrifugationsteps between washes. The final cell pellet was resus-pended in 0.5 ml of the same Tris buffer and disruptedby 10 cycles of rapid freezing in liquid nitrogen fol-lowed by thawing. Then, 0.5 ml of sample buffer (13)was added to each suspension and incubated at 30°Cfor 2 h, and the extracts were stored at -20°C beforeelectrophoresis.

(iii) Gel electrophoresis. The cell extracts werethawed, mixed, and centrifuged for 3 min in a Beck-man model B microcentrifuge. Ten- to twenty-microli-ter (30 to 60 ,ug of protein) samples of each supernatantwere subjected to slab gel electrophoresis in a sodiumdodecyl sulfate (SDS)-Tris-glycine buffer system asdescribed by Laemmli (13). A microslab system wasused in an apparatus similar to that described byMatsudaira and Burgess (16). Gel slabs (0.8 mm thick)were cast between two large microscope slides (10 by8 cm; A. H. Thomas Co., Philadelphia, Pa.). Theresolving gel (8 by 6 cm) was composed of a gradientfrom 9 to 15% (wt/vol) acrylamide. The stacking gel (8by 1.5 cm) was 5% (wt/vol) acrylamide. Gels wereelectrophoresed at 4°C for 60 min with an initialvoltage of 200 V and a current of 0.025 A. Aftercompletion of electrophoresis, the gels were stainedand destained by the rapid procedure of Matsudairaand Burgess (16). Gels were recorded on photographsand dried for storage.

(iv) Serological analysis. Four isolates of R. trifoliiwere used for raising antisera. The choice was basedupon preliminary information obtained regarding theirSDS-polyacrylamide gel electrophoresis (SDS-PAGE)protein profiles. The four isolates, designated 6, 16,27, and 36, were obtained from the subclover plantsgrown at pH 6.8 in the soil dilution experiment. Eachisolate was grown in MGA medium for 48 h at 30°C onan orbital shaker. The cells were harvested by centrif-ugation, washed, and recentrifuged three times in 0.15M physiologically buffered saline, pH 7.0 (NaCl, 8.5 g;KH2PO4, 20.4 g; Na2HPO4, 21.3 g liter-1). The finalpellets were suspended in 0.15 M NaCl to a concentra-tion of 2 x 109 cells per ml. These suspensions wereheated for 60 min at 95°C to denature flagellar (H)antigens. Each antigen preparation was injected intothree New Zealand white male rabbits. The immuniza-tion schedule involved injecting into the thigh muscle 2ml of an emulsion containing equal parts of Freundcomplete adjuvant (Difco) and the above bacterialsuspension. Four weeks later, a 1-ml booster injectionalso was administered intramuscularly. It containedequal parts of a freshly grown, washed, and heat-treated bacterial suspension and Freund incompleteadjuvant (Difco). Ten days after the booster injection,

blood samples were withdrawn from the marginal earvein and tested by the tube agglutination methodagainst the homologous antigen. All rabbits respondedwith acceptable titers (>1:320 dilution of sera) in theagglutination tests. A blood sample (50 to 70 ml) wascollected from each rabbit via cardiac puncture on twoseparate occasions. The second collection was 48 hafter the first. Blood was allowed to clot overnight atroom temperature. The serum was collected by cen-trifugation at 600 x g in a bench centrifuge and thenheated at 56°C for 30 min to inactivate complement.The sera were stored at -20°C.

(v) Tube agglutination. Antigenic preparations of theisolates of R. trifolii were as described above. Thefinal cell suspensions were made in 0.085 M NaCl.This negated the problem of autoagglutination whicharose in the case of a few isolates when suspended in0.15 M NaCl. Somatic agglutination was tested byusing 0.4 ml of antigen suspension mixed with 0.4 ml ofconsecutive twofold dilutions of the sera. The dilutionseries were initiated at 1:10 and were taken to 1:5,120.Tubes were incubated partially submerged in a waterbath at 52°C for 4 h. Titers were recorded as thegreatest serum dilution in which agglutination wasclearly visible.

(vi) Gel immunodiffusion. The gel immunodiffusionsystem was basically an Ouchterlony double diffusionagar plate method. The gels were prepared with 0.8%(wt/vol) agarose (type IV; Sigma Chemical Co., St.Louis, Mo.) in physiologically buffered saline contain-ing 0.025% (wt/vol) sodium azide as a preservative.The gel layer, 4 mm in depth, was contained in smallpetri dishes (50 by 9 mm, no. 1006; Falcon Plastics,Oxnard, Calif.). Suspensions of cells of each isolate ofR. trifolii grown in MGA medium were made inphysiologically buffered saline to a final concentrationof 8 to 10 mg of dry weight per ml. These cells weretested both as whole cells and as sonicated extracts.Sonication was performed with a Branson sonifier(model 200) fitted with a double step microtip. Thesonication protocol consisted of six separate 30-sbursts at 40-W output with 2 min of cooling betweenbursts.

Portions (42 ,ul) of nondiluted antisera were pipettedinto the center wells (4-mm diameter), and 85 RI of theantigens was placed into the outer wells (5-mm diame-ter). The plates were incubated at room temperature ina sealed humidified atmosphere. The development ofprecipitin lines was monitored over 2 to 3 days. Datawere recorded photographically by using dark-fieldindirect illumination.Chemicals. Acrylamide, bisacrylamide,

N,N,N',N',-tetramethylethylenediamine (TEMED),ammonium persulfate, SDS, Coomassie brilliant blueR250, and 3-mercaptoethanol were purchased fromBio-Rad Laboratories, Richmond, Calif. Tris, glycine,agarose, and standard proteins were purchased fromSigma.

RESULTSThe data presented in Table 1 summarize the

serological analyses performed by tube aggluti-nation on the nodule isolates taken from the soildilution experiment. Differences were seen inthe serogroup compositions represented at thetwo pH values. Although serogroup 6 dominated

VOL. 46, 1983

on March 31, 2020 by guest

http://aem.asm

.org/D

ownloaded from

1210 DUGHRI AND BOTTOMLEY

TABLE 1. Serological distribution of the noduleisolates obtained from the soil dilution experiments

No. of pH of the % of isolates of serogroup":seedlingIsolates agar" 6 16 27 6/27' 36 Others'

13 4.8 77 8 0 0 0 1522 6.8 0 18 23 23 32 4' Initial pH value obtained after autoclaving the

medium.h Serogroups are numbered in reference to the spe-

cific isolates to which antisera were raised.' Isolates in this group cross-reacted with antisera 6

and 27 in the tube agglutination reaction.CIsolates which did not react with any of the four

antisera available.

the nodule population at the low pH, where itaccounted for 77% of the total isolates, it wascompletely absent at pH 6.8. In contrast, sero-groups 27, 6/27, and 36 were only present at thehigher pH, making up 23, 23, and 32%, respec-tively, of the total isolates. Only serogroup 16was found at both pH values and was a minorcomponent in both cases.The impact of acidity upon the composition of

the nodule occupants was confirmed in twomore experiments where plants were growndirectly in larger quantities of the soil and wherethe acidity change was brought about by anapplication of CaCO3.

Data presented in Table 2 show that serogroup6 was the dominant serogroup present in thenodules formed in the unamended soil in bothexperiments, occupying 47 and 58% of the nod-ule population. In contrast, serogroup 36 occu-pied 0 and 8% of the nodules in the same soiltreatment. In the limed soil, the situation wasreversed, with serogroup 36 being the dominant(experiment 2), or one of the dominant (experi-ment 1), serogroups, occupying 49 and 35% ofthe nodules, whereas serogroup 6 occupied 0and 10% of the nodules. Serogroups 16 and 27were also represented in both of the soil coreexperiments. Although the numbers of isolates

TABLE 2. Serological distribution of the isolates ofR. trifolii obtained from subclover plants grown in

limed and unlimed soil

No. of Soil % of isolates of serogroup:Expt islts treat- __- _ment" 6 16 27 6/27 36 Others

1 19 Unlimed 58 5 26 0 0 1120 Limed 0 15 35 5 35 10

2 38 Unlimed 47 0 11 0 8 3439 Limed 10 3 13 0 49 25

"The pH values of the soils were determined beforesowing the seed. The pH values were 4.8 and 6.4 forthe unlimed and limed soils, respectively.

in these two groups were generally less thaneither serogroup 6 or 36, it was noticeable thatthe contribution of serogroup 27 to the nodulepopulation was not affected to any great extentby the pH variable. In all three experiments,which involved a total of 14 plants and 151isolates, only 30 isolates were unidentifiablewith the antisera at our disposal.

Since serogroups 6 and 36 were the represen-tatives of the indigenous population most pro-foundly affected in their nodulating ability by theacidity level, these isolates from the two soilexperiments were analyzed further by SDS-PAGE. Data presented in Fig. 1 show that sixand four different protein profiles (gel types)were found to encompass the 26 and 29 isolatesfrom serogroups 36 and 6, respectively. Al-though all the isolates from serogroup 36 reactedin a homologous manner in gel immune diffusionwith the respective parent antiserum and inde-pendent of their gel type (Fig. 2B) and only oneisolate from serogroup 6 showed no immunopre-cipitin reaction (Fig. 2A), the number of isolatesrepresented by each gel type did vary. Data inTable 3 summarize the percentages of the iso-lates from each serogroup that fell into each geltype. In the case of serogroup 6, isolates repre-sented by gel type a were found to be wellrepresented in both soil experiments and ac-counted for 82 and 44% of the representatives ofthat serogroup in experiments 1 and 2, respec-tively. In contrast, isolates with gel type d werenot represented in experiment 1 even though

Ps a b c d e f g h i

FIG. 1. Representative protein profiles of each geltype from serogroup 6 (lanes a through d) and sero-group 36 (lanes e through j) displayed on a 9 to 15%(wt/vol) linear gradient polyacrylamide gel slab. Pro-tein standards (ps) are, from top to bottom, bovineserum albumin, 66 kilodaltons; egg albumin. 45 kilo-daltons; trypsinogen. 24 kilodaltons; P3-lactoglobulin.18.4 kilodaltons; and lysozyme. 14.3 kilodaltons.

Appt-. ENVIRON. MICROBIOL.

on March 31, 2020 by guest

http://aem.asm

.org/D

ownloaded from

EFFECT OF ACIDITY ON NODULE POPULATIONS

FIG. 2. Gel immunodiffusion characteristics offourisolates representing each of the gel types of serogroup6 (A) and six isolates representing each of the gel typesof serogroup 36 (B) against their respective parentantiserum. The antisera are in the center wells. Theantigens are arranged in the outer wells and designatedby their identification number.

they were present at a level slightly greater thanisolates with gel type a in experiment 2. No geltype of serogroup 36 was common to bothexperiments; gel types e and i were dominant inexperiments 1 and 2, respectively.Data presented in Table 4 show the symbiotic

effectiveness of isolates from serogroup 6 as

they relate to gel type. Isolates from within thesame gel type showed various degrees of effec-

TABLE 3. Distribution of the isolates within the geltypes of serogroups 6 and 36

% of serogroup in gelSerogroup Gel type" type (no. of isolates)

Expt 1 Expt 2

6 a 82 (9) 44 (8)b 9 (1) 0c 9 (1) 0d 0 56 (10)

36 e 0 63.5 (12)f 0 5 (1)g 0 26.5 (5)h 0 5 (1)

71 (5) 0j29(2) 0

"Gel types are identical to those in Fig. 1.

TABLE 4. Symbiotic effectiveness of the isolates ofserogroup 6 as related to gel type

Total No. in effectivenessGl no. of category"~:type' isolates El E2 F,

a 17 4 8 5b 1 0 0 1c 1 1 0 0d 10 2 5 3

a Gel types are identical to those in Fig. 1.b Effectiveness categories are defined as follows. El

isolates produced plant yields (29 to 32 mg) not signifi-cantly less than nitrate-fertilized plants (P = 0.05)according to the Duncan new multiple range test. Themean dry weight of nitrate-grown plants was 32.0 +3.5 mg. E. isolates produced yields (24 to 28 mg) lessthan E1 isolates and greater than E3 isolates (P =0.05). E3 isolates produced yields (20 to 23 mg) lowerthan E. isolates (P = 0.05). The mean dry weight ofuninoculated plants was 13.0 ± 2.0 mg.

tiveness, with 76.5 and 80% of isolates in thedominant gel types a and d, respectively, beingin the two suboptimum effectiveness categories.Similar results were obtained for isolates withinthe different gel types of serogroup 36 (Table 5);75 and 80% of the isolates in the dominant geltypes e and i were of suboptimum effectiveness.Furthermore, the difference between the symbi-otic effectiveness of isolates identical by bothantigenic and protein profile methodologies indi-cates that the isolates within these two dominantserogroups are even more diverse than wasreadily apparent. A minimum of 8 and 12 differ-ent organisms must be present within sero-groups 6 and 36, respectively, to account for theresults of the combination of screening methods.

DISCUSSIONData are presented in this paper which show

the tremendous diversity within an indigenoussoil population of R. trifolii and that soil acidity

TABLE 5. Symbiotic effectiveness of the isolates ofserogroup 36 as related to gel type

Gel Total No. in effectivenessGypel no. of categorytype" isolates E.

e 12 3 7 2f 1 0 1 0g 5 0 4 1h 1 1 0 0

5 1 3 1j 2 0 1 1

"Gel types are identical to those in Fig. 1.^ Effectiveness categories are defined in Table 4.

footnote b.

VOL. 46, 1983 1211

on March 31, 2020 by guest

http://aem.asm

.org/D

ownloaded from

1212 DUGHRI AND BOTTOMLEY

can influence which members of that populationwill nodulate T. subterraneum L. cv. Mt. Bark-er. Of some interest is the fact that a preliminaryscreening of isolates from the soil dilution ex-periment by SDS-PAGE allowed us to selectfour distinctly different isolates for raising anti-bodies. Subsequent screening with those fourantisera allowed us to group 121 of the 151isolates from the three experiments into fivemajor serogroups. Such success could be con-sidered fortuitous; however, we believe this tobe an approach that researchers of the Rhizobia-ceae should consider before expensive and time-consuming serological screening of many un-known field isolates.The data provide evidence that although a

wide diversity of strains of R. trifolii can persistin the free-living state in this acidic soil, thecapacity of some members of the population tonodulate the host legume is enhanced or imped-ed depending on the acidity of the soil. Previousobservations have shown that both the speciesand the variety of a species of Trifolium canaffect the competitive nodulating success of mix-tures of different strains of R. trifolii (5, 9, 28, 29,33, 37). Reports on physicochemical parametersaffecting this selection are less numerous. Ef-fects of root temperature have been observedwith Glycine maxlRhizobium japonicum (30, 39)and with T. repens/Rhizobium trifolii (6). Effectsof acidity have only been seen previously in pH-adjusted mineral salts agar by using high celldensities (107 cells per ml) of a simple mixture oftwo strains of R. trifulii (10, 32).By using SDS-PAGE to obtain protein profiles

of the isolates from within the serogroups, weshowed that isolates within the two major sero-groups, 6 and 36, are heterogeneous despitetheir antigenic identity. This confirms the obser-vations of others, working with R. japoniclum,that serological identity does not necessarilyimply that isolates are unequivocally identical(4, 12, 23). Our observation that isolates repre-sented by a specific gel type from within a multi-gel type serogroup can dominate the contribu-tion of that serogroup to the nodule populationsuggests that studies of the population dynamicsof rhizobia, either in soils or in nodules, by usingserological means will not illustrate the varia-tions in population size of the strains within theserogroup. In the case of serogroup 6, isolateswith the common gel type a were found in thenodules from both soil core experiments, indi-cating that isolates with this gel type may trulyreflect a strain widely distributed within this soiland yet only highly competitive under acidicconditions. Obviously, such isolates are candi-dates for further studies on the mechanisms andfactors which affect competition.The variations among symbiotic effective-

nesses of isolates that were identical by antigen-ic determinations and by one-dimensional pro-tein profile patterns confirm other observations,which showed that the serological stability ofsubcultures of a strain and the serological identi-ty of different strains of R. trifolii, was often atvariance with differences among the symbioticeffectivenesses of the same strains (24, 34, 35).We wish to reemphasize that the testing ofunknown isolates for symbiotic effectiveness beconsidered an essential part of any battery ofcomplementary methods of strain identification.At this time we have no simple explanation for

the effect of acidity upon the change in domi-nance of serogroups 36 and 6. Pure culturesrepresenting serogroups 6 and 36 have beenshown to nodulate subclover plants growing onseedling agar at both pH 4.8 and 6.4, even withan inoculum size of <10 viable cells (unpub-lished data). Several hypotheses are in need oftesting, and for each there is supportive, albeitcircumstantial, evidence in the literature. (i)There is a differential proliferation of the mem-bers of the serogroups due to amelioration of soilacidity and this occurred before, or independentof, the host plant presence (38). (ii) Differentialproliferation is dependent on the presence of thehost (or nonhost) plant through changes in thequality or quantity or both of root exudates as aresult of changes in soil acidity (8, 11, 17, 31).(iii) The change in serogroup dominance is inde-pendent of proliferation and due to events asso-ciated directly with the site of infection andnodule initiation. Evidence for the latter possi-bility has come from research into the competi-tive dominance of strains of R. japonicum andRhizobilim phaseoli, which has led to skepticismof the theory of legume host-specific stimulationof rhizobia (25-27). Further research on thedynamics of the soil-borne and rhizosphere pop-ulations of these two acid-responsive serogroupswill allow us to address these possibilities.

ACKNOWLEDGMENTS

This research was supported by funds from U.S. Depart-ment of Agriculture-Science and Education Administration/Competitive Research Grants Office research agreement 59-2411-1-1-726-0 and the Oregon Agricultural ExperimentStation. M.H.D. gratefully acknowledges financial supportthrough a scholarship awarded by the El-fatah University ofthe People's Libyan Arab Jamahiriya.The technical assistance of Janice Fuquay is gratefully

acknowledged.LITERATURE CITED

1. Albrecht, W. A., and E. M. Poirot. 1930. Fractional neu-tralization of soil acidity for the establishment of clover. J.Am. Soc. Agron. 22:649-657.

2. Andrew, C. S. 1976. Effect of calcium, pH and nitrogen onthe growth and chemical composition of some tropical andtemperate pasture legumes. I. Nodulation and growth.Aust. J. Agric. Res. 27:611-623.

3. Evans, H. J. 1974. Symbiotic nitrogen fixation in legumenodules, p. 417-426. In T. C. Moore (ed.), Research

APPL. ENVIRON. MICROBIOL.

on March 31, 2020 by guest

http://aem.asm

.org/D

ownloaded from

EFFECT OF ACIDITY ON NODULE POPULATIONS 1213

experiences in plant physiology, a laboratory manual.Springer-Verlag, Berlin.

4. Gibson, A. H., W. F. Dudman, R. W. Weaver, J. C. Hor-ton, and I. C. Anderson. 1971. Variations within sero-group 123 of Rhizobilum japonicium. Plant Soil (Spec.Vol.):33-37.

5. Hagedorn, C., and B. A. Caldwell. 1981. Characterizationof diverse Rhizobium trifolii isolates. Soil Sci. Soc. Am. J.45:513-516.

6. Hardarson, G., and D. G. Jones. 1979. Effect of tempera-ture on competition amongst strains of Rhizcobilum trifoliifor nodulation of two white clover varieties. Ann. Appl.Biol. 22:229-236.

7. Jackson, W. A. 1967. Physiological effects of soil acidity,p. 43-124. In R. W. Pearson and F. Adams (ed.), Soilacidity and liming. American Society of Agronomy, Madi-son, Wis.

8. Jones, D. G. 1966. The contribution of white clover to amixed upland sward. II. Factors affecting the density andthe effectiveness of Rhizobiium trifolii. Plant Soil 24:250-259.

9. Jones, D. G., and G. Hardarson. 1979. Variation withinand between white clover varieties in their preference forstrains of Rhizobiium trifolii. Ann. Appl. Biol. 92:221-228.

10. Jones, D. G., and S. J. Morley. 1981. The effect of pH onhost plant 'preference' for strains of Rhizobilim trifoliiusing fluorescent ELISA for strain identification. Ann.Appl. Biol. 97:183-190.

11. Jones, D. G., J. M. M. Munro, R. Hughes, and W. E.Davies. 1964. The contribution of white clover to a mixedupland sward. I. The effect of Rhizobiirm inoculation onthe early development of white clover. Plant Soil 21:63-69.

12. Kowalski, M., G. E. Ham, L. R. Frederick, and I. C.Anderson. 1966. Relationship between strains of Rhizobi-urm japonic um and their bacteriophages from soil andnodules of field grown soybeans. Soil Sci. 118:221-228.

13. Laemmli, U. K. 1970. Cleavage of structural proteinsduring the assembly of the head of bacteriophage T4.Nature (London) 227:680-685.

14. Lie, T. A. 1969. The effect of low pH on different phasesof nodule formation in pea plants. Plant Soil 31:391-406.

15. Loneragan, J. F., D. Meyer, R. G. Fawcett, and A. J.Anderson. 1955. Lime pelleted clover seeds for nodulationon acid soils. J. Aust. Inst. Agric. Sci. 21:264-265.

16. Matsudaira, P. T., and D. R. Burgess. 1978. SDS-micros-lab linear gradient polyacrylamide gel electrophoresis.Anal. Biochem. 87:386-396.

17. Mulder, E. G., and W. L. Van Veen. 1960. Effect of pHand organic compounds on nitrogen fixation by red clo-ver. Plant Soil 13:91-113.

18. Munns, D. N. 1968. Nodulation of Medicago sativo insolution culture. I. Acid sensitive steps. Plant Soil 28:129-146.

19. Munns, D. N. 1977. Mineral nutrition and the legumesymbiosis, p. 353-391. In R. W. F. Hardy and A. H.Gibson (ed.), A treatise on dinitrogen fixation, section IV.Agronomy and ecology. John Wiley & Sons, Inc., NewYork.

20. Munns, D. N., and R. L. Fox. 1977. Comparative limerequirements of tropical and temperate legumes. PlantSoil 46:533-548.

21. Munns, D. N., R. L. Fox, and B. L. Koch. 1977. Influence

of lime on nitrogen fixation by tropical and temperatelegumes. Plant Soil 46:591-601.

22. Munns, D. N., H. H. Keyser, V. W. Fogle, J. S. Hohen-berg, T. L. Righetti, D. L. Lauter, M. G. Zaroug, K. L.Clarkin, and K. W. Whitacre. 1979. Tolerance of soilacidity in symbioses of mung bean with rhizobia. Agron.J. 71:256-260.

23. Noel, K. D., and W. J. Brill. 1980. Diversity and dynamicsof indigenous Rhizobuiisn japoniculn populations. Appl.Environ. Microbiol. 40:931-938.

24. Purchase, H. F., and J. M. Vincent. 1949. A detailed studyof the field distribution of strains of clover nodule bacte-ria. Proc. Linn. Soc. N.S.W. 74:227-236.

25. Reyes, V. G., and E. L. Schmidt. 1979. Population densi-ties of Rhizcobilurnjaponiciumn strain 123 estimated directlyin soil and rhizospheres. AppI. Environ. Microbiol.37:854-858.

26. Reyes, V. G., and E. L. Schmidt. 1981. Populations ofRhiZobiltnjaponicum associated with the surfaces of soil-grown roots. Plant Soil 61:71-80.

27. Robert, F. M., and E. L. Schmidt. 1983. Populationchanges and persistence of Rhizobiium phaseoli in soil andrhizospheres. Appl. Environ. Microbiol. 45:550-556.

28. Robinson, A. C. 1969. Host selection for effective Rhiz-o-biuim trifolii by red clover and subterranean clover in thefield. Aust. J. Agric. Res. 20:1053-1060.

29. Roughley, R. J., W. M. Blowes, and D. F. Herridge. 1976.Nodulation of Trifoliuon subterraneumn by introduced rhi-zobia in competition with naturalized strains. Soil Biol.Biochem. 8:403-407.

30. Roughley, R. J., E. S. P. Bromfield, E. L. Pulver, andJ. M. Day. 1980. Competition between species of Rhizobi-iin for nodulation of Glvc-ine max. Soil Biol. Biochem.12:467-470.

31. Rovira, A. D. 1961. Rhlizobiirn numbers in the rhizo-spheres of red clover and paspalum in relation to soiltreatment and the numbers of bacteria and fungi. Aust. J.Agric. Res. 12:77-83.

32. Russell, P. E., and D. G. Jones. 1975. Immunofluores-cence studies of selection of strains of R. trifolii by S184white clover (T. repens L.). Plant Soil 42:119-129.

33. Russell, P. E., and D. G. Jones. 1975. Variation in theselection of Rhizobiium trifolui by varieties of red andwhite clover. Soil Biol. Biochem. 7:15-18.

34. Vincent, J. M. 1944. Variation in the nitrogen-fixing prop-erty of Rhizobiumn trifolii. Nature (London) 153:496-497.

35. Vincent, J. M. 1945. Host specificity amongst root-nodulebacteria isolated from several clover species. J. Aust.Inst. Agric. Sci. 11:121-127.

36. Vincent, J. M. 1970. A manual for the practical study ofroot-nodule bacteria. IBP Handbook no. 15. BlackwellScientific Publications. Ltd.. Oxford. U.K.

37. Vincent, J. M., and L. M. Waters. 1953. The influence ofthe host on competition amongst clover root-nodule bac-teria. J. Gen. Microbiol. 9:357-370.

38. Vincent, J. M., and L. M. Waters. 1954. The root-nodulebacteria as factors in clover establishment in the redbasaltic soils of the Lismore district, New South Wales.II. Survival and success of inocula in laboratory trials.Aust. J. Agric. Res. 5:61-76.

39. Weber, D. F., and V. L. Miller. 1972. Effect of soiltemperature on Rhizobium jiaponicumn serogroup distribu-tion in soybean nodules. Agron. J. 64:796-798.

VOL. 46, 1983

on March 31, 2020 by guest

http://aem.asm

.org/D

ownloaded from