survival of field-grown rhizobia over the dry summer period in Western Australia

So8 Biol. Biochem.Vol. 5, pp. 415-423. PergamonPress1973. Printedin Great Britain.

SURVIVAL OF FIELD-GROWN RHIZOBIA OVER THE DRY SUMMER PERIOD IN WESTERN AUSTRALIA D.

Department

L.

CHATEL*

and C. A.

PARKER

of Soil Science and Plant Nutrition, Institute of Agriculture, University of Western Australia, Nedlands, Western Australia, 6009 (Accepted

17 October 1972)

Summary-The survival over summer of field-grown root nodule bacteria was studied in the field and the laboratory during the course of an inv~tigation into a nodulation problem of annual clovers. Dry field soils containing R~izobiu~ trifo~i~and R. lupini were subjected to a range of temperatures in the laboratory, the bacteria surviving 6 h exposure to temperatures as high as 80°C. Soil temperatures during summer were recorded at different depths. Populations of rhizobia were estimated at these depths from the end of the growing season (October) to early autumn (April) in plots which had carried dense swards of subterranean clover and serradella. High populations of R. lupini were maintained in the serradella plots throughout the summer. Populations of R. tri’& in the subterranean clover plots were initially much lower. and declined with both time and depth. The problem known as ‘second-year clover mortality’ is primarily due to low numbers of clover rhizobia in the soil at the end of the growing season. This situation is aggravated over the long hot dry summer, when there is a further decline in numbers.

INTRODUCTION MOST of the work on rhizobiai survival has been oriented towards determining the factors involved in survival of rhizobia inoculated onto seed (Vincent, 1958, 1962; Vincent et al., 1962; Date, 1968). Studies on survival in soil have been largely laboratory based (Bowen and Kennedy, 1959; Vyas and Prasad, 1960; Sanderson, 1962; Marshall, 1964; Wilkins, 1967; Chowdhury et al., 1968). Marshall et al. (1963) suggested that poor nodulation in the year following successful establishment of inoculated clover and medics in some sandy soils in Western Australia was due to the failure of their nodule bacteria to withstand the dry heat of summer. They attributed the good nodulation of second-year lupin and serradelia stands to superior resistance of the nodule bacteria to dry heat. Sanderson (1962) showed that a dry soil from the Western Australian wheatbelt contained viable nodule bacteria for subterranean clover after heating at 80°C for 6 h, and in one case up to 8 h. In a series of Laboratory experiments Marshall (1964) found that R. Iupini survived at higher soil temperatures than either R. tr$olii or R. m&&i and that certain clay types, commonly found in the heavier textured nonproblem soils, afforded some protection to the fast growing species against the higher temperatures. This paper reports investigations into the survival and dry-heat tolerance of field-grown clover and lupin root-nodule bacteria in a soil type known to exhibit severe second-year clover mortality. The work has been described in full by Chatel(l967) and reported briefly by Chatel and Parker (1964) and Chatel et al. (1968).

* Present address: Department

of Agriculture, Jarrah Road, South Perth, Western Australia, 6151. 415

416

D. L. CHATEL AND C. A. PARKER MATERIALS AND METHODS

Survival over summer-autumn period E.~~eri~~e~tta~ site and design. The soil was a grey sand, at least 30 cm deep, of pH 6.3 (1: 5, soil: water), overlying pisolitic gravel, 32 km west of Carnamah. It has been described by Marshall et al. (1963), and corresponds with sample No. 5, Marshall (1964). The plots were in an area cleared of natural vegetation in 1954 and cropped to wheat in 1955 and 1959. Copper-zinc superphosphate at the rate of 168 kg/ha had been applied with each crop. The area was free of Trifofium spp., and R. trifolii, if present, would have been in very low numbers, since piant-dilution estimates failed to indicate their presence in the soil. The following three rhizobial treatments were included in duplicate plots, in a randomized layout, each 183 x 183 cm with 61 cm pathways:

(1) R. trifolii strain TAl inoculated onto subterranean clover (Trifolium subterraneum L. cv. Dwaiganup). Strain TAl has been used for a number of years as a component of the commercial peat inoculants used in Australia. (2) R. trifbZiistrain WU9.5, also inoculated onto Dwalganup sub. clover, was isolated in 1962, after heating in dry soil at 90°C for 6 h (Parker, unpublished). (3) R. lupim’strain WU425 inoculated onto serradella (Omithopus sativus Brot.). WU425 effectively nodulates lupins and serradella; neither of which exhibits second-year mortality (Parker, 1962). Serradella was chosen because of its small seed size which enables it to be grown in conveniently sized test tubes for plant-dilution estimates of soil population and its growth habit which gives a dense sward similar to that of subterranean clover. Seed inoculation and sowing. The seed was heavily sown (112 kg viable seed/ha) and heavily inoculated (lo5 viable cells/seed, based on plate counts) with rhizobial suspensions in 30% (w/v) gum arabic and 10% (w/v) sucrose. The plots were sown within 6 h of seed inoculation on 22 May 1963, into moist soil that had been hand-cultivated after top-dressing with plain superphosphate at 336 kg/ha. Sampling. On each sampling occasion three pits were excavated in each of the six plots, exposing a perpendicular face in each pit. A cube of soil was recovered from each face by inserting horizontally a 3 ‘8 cm square stainless steel tube, with sharpened edges. Four samples, O-3-8 cm, 3.8-7.6 cm, 7-6-11.4 cm and I1 ~4-15.2 cm were taken from a vertical surface of each pit. The three samples from each depth were bulked, mixed and placed in polythene bags. Subsamples from each bag were placed in tins for soil-moisture determinations. Estimation of populations. Soil samples were dried in a forced air-drying cabinet at room temperature and passed through a 2-O mm sieve to remove root material and pebbles and organic debris. Populations of rhizobia were estimated using a plant-dilution method (Date and Vincent, 1962); but nutrient sand was used in the place of nutrient agar. Duplicate I -0 ml aiiquots of each tenfold soil-water dilution were used to inoculate sterile host seedlings in duplicate 15 *2 x 2.5 cm test tubes. The tubes were placed in racks and immersed to the soil level in a refrigerated water bath (Asher et al., 1965) at 21 & l”C, in a glasshouse fitted with evaporative coolers. After a suitable growth period (4-5 weeks) ‘fertile tubes’ (containing nodulated plants) and ‘nonfertile tubes’ (containing non-nodulated plants) were recorded. Analyses of variance were done on the mean number of fertile tubes (mean fertile level) following Munch-Petersen and Boundy (1961, 1963 and 1964) who used dilution methods for comparing populations of rumen bacteria grown in different culture media, and Sanderson (1962) and Chowdhury (1965) who examined populations of rhizobia in soil.

SURVIVAL OF RHIZOBIA

417

Estimates of population densities as most probable numbers (MPN) were calculated from the ‘mean fertile level’ using the table of Stevens given by Fisher and Yates (1953) and employed by Date and Vincent (1962). Soii temperatures

Directly heated bead-type thermistors (Stantel Type FS2) water-bath calibrated, were buried in a ‘bare’ area adjacent to the plots. Two thermistors were buried in dry soil (11 November 1963) at each of the following depths : l-3, 5 * 1, 10 *2, 15 - 2 and 30 - 5 cm. Each thermistor was connected to a multiswitch wheatstone-bridge circuit, and its resistance was measured by null-point indication using a galvanometer. Tolerance offie/d grown root-nodule bacteria to dry-heat Soil sampling and treatment. Dry soils were sampled under senescent plants (O-12.7 cm) in a series of similar undisturbed plots adjacent to those already described. Each sample was passed through a 2.0 mm sieve. Replicate plot samples were bulked after thorough mixing, superphosphate added (equivalent to 224 kg/ha) and spread, 5 cm deep, on muslin-lined wire trays before heat treatment. The three soil samples (TAl, WU95 and WU425) were each subjected to four temperatures; 50, 60, 70 and 80°C for 6 h. A room temperature control of each was retained. Pot preparut~on, planting and ~lar~est~~lg. Half a kilogram of the treated soil was added to each pot (454 ml polythene) and lightly packed. Surface sterilized seed (30 subcIover in each of the TAl and WU95 pots and 60 serradella in each of the WU425 pots) were sown immediately before watering to field capacity with sterile water. Four pots of each soil from each temperature treatment were prepared, The nodulation pattern of each seedling was recorded as follows:

Tap-root nodulation-Nodules present on the upper 5 -0 cm of tap-root. Lateral root noduIation-NoduIes present on lateral roots up to 2.5 cm aIong laterals growing from the upper 3 98 cm of tap-root. Lateral nodulation was recorded only on plants without tap-root nodules. RESULTS

Counting procedure

We used Date and Vincent’s (1962) plant dilution method after comparing it with the drop viable count method (Miles and Misra, 1938). Thus a water suspension of TAl was estimated to contain 3.2 x 10’ cells/ml by drop count, and 5 ~0 x lo9 by the plant dilution method. This agreement was considered satisfactory. Soil temperatures

Soil and air temperatures were recorded at hourly intervals over a 24 h cloudless period in the middle of a 3-day heat-wave when the soil was dry. The maximum air temperature (shade) on each day was 39°C. Figure 1 shows the effect of time, depth and air temperature on the temperature of dry soil under bare ground. The maximum tem~ratures reached were: 59°C at 1.3 cm, 47°C at 5.1 cm, 39°C at IO-2 cm, 35°C at 15.2 cm and 31°C at 30.5 cm. Of additional readings taken on other occasions the highest soil temperature recorded was 62°C at 1.3 cm at 2.0 p.m. on 26 December 1963, when the shade temperature was 41°C.

418

D. L. CHATEL

AND C. A. PARKER

FIG. I. Soil tem~ratures under bare ground. Two readings were taken every hour at each depth over a 24 h period (6-O a.m. 18 Feb, to 6.0 a.m. 19 Feb, 1964). Key: (0) 1.3 cm; (0) 5.1 cm; (0) 10.2ctn; (m) 15.2cm; (&I 30.5cm; (x) Air (2 m above ground).

Survival over the summer-autumn period

The results are presented as most probable numbers in Fig. 2 and summaries of the statistical analyses in Table 1. Although samples were not taken at regular monthly intervals it is convenient to plot them as such. Because of the inability to detect any fertile tubes for TAl at the lower depths and later times (i.e. TAl populations were uncountable on these occasions) an analysis of variance was made on the data for the first two depths and three times only. There was a significant strain effect (P < O-01) R. ~~~~~~ being superior to both R. tr~~lii strains, and a significant depth x strain interaction (P < O-01). The populations of WU9.5 were higher (P < O*OOI) than those of TAX at the first sampling. The pattern of population decline with time did not differ significantly between strains of R. trifolii when all depths were combined. However, there was a significant (P < 0.05) correlation between decreasing numbers of WU95 in the surface layer and time (r = 0.85). No such decline was evident for WU9.5 at the remaining depths or for WU425. A comparison between R. trifolii (WU95) and R. Iupini (WU425) over all depths and sampling occasions indicated the superiority of WU425 over WU95 (P < O*OOl)in terms of both colonization and survival. The depth effect was highly significant (P < O-001), with both species showing a similar decline in numbers with depth. This decline was linear and deviations from linearity were not significant. There was a significant depth x strain x time interaction with a tendency for WU425 popuIations to increase with time, whereas those of WU95 showed a reverse trend.

R.trifolii TAl 76.

Rhrifolii wu 95

Rhini wu 425

0- 3,acm

s4. 3. 2-9

1.

C! 8

0

1 3,a-7dtm

s *

7.6-11.4cm

n

r-l

,1,2.3,4.5,6.7,

,1,2,3.4,5.6,7. fAMPut4G

,1,2,3.4,5.6.7,

OCCASION

FIG. 2. Soil populations of R~iz~bi~ ~rifo~ii(TAX and WU95) and R. iupini under dry swards of host-plants. The histograms show population densities (as mean fertile levels from piantdilution counts) at each of four soil depths sampled at approximately monthIy intervals over the summer-autumn period in the year of establishment. Sampling times were: 1 = 14 Oct. 1963; 2 = 12 Nov. 1963; 3 = 12 Dec. 1963; 4 = 7 Jan. 1964; 5 = 4 Feb. 1964; 6 = 3 Mar. 1964; 7 = 31 Mar. 1964. TABLE1. SIGNIFICANCEOF MAIN EF’FECTS ANDTH!.?IR INTERACTIONS Analysis* Source Strain (S) Time (T) TxS Depth (0) DxS DxT DxSxT

1

2

** NS NS NS

*** NS ** *** NS NS **

E **

* Analysis 1 = 3 strains (TAl, WU95, WU425) x 2 depths (O-3.8 cm, 3-S-7.6 cm) x 3 times. Analysis 2 = 2 strains (WU95, WU425) x 4 depths x 7 times. ** Significant at 0.1 ‘A level. *** Significant at 0.01 0/olevel. NS-not significant.

419

420

D. L. CHATEL

AND C. A. PARKER

Tolerance offield grown root-nodule bacteria to dry-heat

With soil from the TAI plots there was a steady decline in tap-root nodulation with increasing temperature (Table 2). Little can be said about the other two soils, apart from suggesting that initial population densities were too high for any responses to be detected with this technique. TABLE2. THE EFFECTOF 6 h EXPOSURESTO

HIGH TEMPERATURES SURVIVAL OF RHIZOBIA IN AIR-DRY SOIL*

Strain of Rhizobium

in soil TAl wu95 WU425

Heat treatment

ON

(% nodulated)

-

Nil?

50°C

60°C

70°C

80°C

29* 100 100

27 100 100

14 100 100

14 100 100

8 100 100

* Figures represent the per cent of host plants sown in the heat-treated soil that nodulated on the tap-root. Four replicate pots were bulked per nodulation assessment. t MPN of viable cells/g soil before treatment: TAl = ~5; WU95 = 3.4 x 103; WU425 = 1.0 x 10s.

It is difficult to compare estimates of the populations of the root-nodule bacteria (footnote to Table 2) with those from the main series of plots sampled at comparable times, since single estimates using duplicate tubes involve very wide confidence limits. The MPN estimates suggest that there was very little difference between the two series of plots. DISCUSSION

Heavy inoculation together with a heavy seeding rate gave full nodulation and dense swards. Marked differences in colonization and survival are evident, R. lupini being superior to R. trifolii and R. trifolii strain WU95 superior to strain TAl. These results are in accord with the field observation that lupins (Lupinus spp.) and serradella (Ornithopus spp.) nodulated by R. lupini, are not subject to second year nodulation problems, contrary to the experience with Trifolium spp. in these problem soils. Survival over the summer-autumn period

A surprising feature of these investigations was the very low populations in the subterranean clover treatments, particularly those of TAI at the commencement of sampling in early October (Fig. 2). The soil was very dry at this stage (less than 1.0 per cent soil moisture). Although Marshall (1964) demonstrated a fall in TAl numbers after drying a soil at 3O”C, this fall was very low compared with that obtained when the same soil was subjected to high temperatures (Marshall, 1964). There is also the possibility that the soil temperature may have reached lethal levels before and during the drying cycle when the soil was still moist and therefore when the bacteria were more vulnerable. Bowen and Kennedy (1959) have shown that R. trifolii was susceptible to temperatures as low as 40°C in moist soil.

SURVIVAL

OF RHIZOBIA

421

Because of the many occasions on which TAl numbers were too low to be counted, very little can be said of this strain other than that it was very much less abundant than WU95. Although there was a decline in TAl numbers, particularly from December, which may have been an effect of temperature, it is certain that high summer temperatures only aggravated an already serious situation. WU95 followed similar trends to TAI in that the numbers fell with both depth and time. The demonstration of the superiority of WU95 over TAl is important since it gave promise of finding differences within the species R. trifolii. The persistence of rhizobia in the surface layers of the soil at the end of summer is interesting. It is a common field observation that many of the nodulated plants in problem soils have their first nodule about 7.5-10.0 cm down the tap-root. Therefore, rhizobia, although present near the surface, may not always cause nodulation there. A similar finding with medics has been noted by Brockwell and Whalley (1970). The possible role of early moisture stress in this phenomenon has been considered by Hamdi (1970). The inverse relationship between the soil population of WU9.5 and time for the surface layer only is what would be expected in a situation where temperature gradients affect survival. TAl populations may have declined at similar rates. HeIy et ai. (1957) also demonstrated a decline in the soil population of R. trifolii over summer. The summer climate in the region where they worked differs markedly from that prevailing in the wheatbelt of Western Australia. In the former the summer period is characteristically wetter and cooler. R. ~~~~~j(WU425) showed a very different pattern of survival than either of the R. t~i~Zii strains. The numbers at all times were very much higher, and there was a significant increase with time, particularly near the surface. This increase may have been due to renewed activity resulting from a summer and an autumn rain (33 mm in February and 15 mm in March), but it is difficult to know if the increase was due to multiplication of soil rhizobia or to the release of rhizobia from nodule tissue, or both. It is noteworthy that there was no similar increase in the pop~ations of R. tr~~?~~at the same time. The temperatures recorded here do not differ greatly from those of Kininmonth (1962) who obtained maxima of 65~5°C at 2.5 cm and 47°C at 10 cm in a bare yellow sand at Perth from daily readings over a 5 year period. Our results do not provide information about the effect of repeated exposures to lower temperatures. Tolerame to dry-heat

Thompson and Vincent (1967) have used the nodulation of host seedlings in soil cores as an indirect measure of rhizobial population. However, nodulation in such cores is likely to be more a function of the distribution of viable cells than of absolute numbers. The technique used in this experiment differed from theirs in that the sample was mixed in order to distribute the rhizobia and is based on the assumption that the fast penetrating tap-root of a germinating seedling is susceptible to nodulation for a relatively short period. The resultant nodulation therefore indicates that one or more viable rhizobial cells were in close proximity to the pathway of the growing root. Although nodulation in the WU95 and WU425 treatments was optimal over all temperatures this does not imply that temperatures had no effect on the rhizobia. Indeed it is likely, in view of the work of Marshall (1964) that considerable death occurred, particularly at the higher temperatures. Nevertheless, survival was enough to ensure maximum tap-root nodulation; it appears that the technique used is practicable only with soils of very low initial populations or where distribution of the bacteria is irregular. There was no observable influence of temperature on the TAl population at 50°C but there was a marked drop in nodulated plants from 50 to 60°C and a further drop from 70

422

D. L. CHATEL AND C. A. PARKER

to 80°C. These results are not in agreement with those of ~arshali (1944) who showed no survival of TAI after 5 h at 7O”C, even though his soil had a much higher initial population. In this experiment nodulation did occur in the Carnamah soil after exposure to 80°C for 6 h, proving that R. trifolii cells had survived this temperature. This is in agreement with the finding of Sanderson (1962) that field-grown R. trifolii cells (strain unidentified) survived a soil temperature of 80°C for 8 h. Concluding convnents

These experiments show that the poor second-year nodulation of clovers in problem soils in Western Australia is indicative of low numbers of rhizobia available in the soil for nodulation of seedling clover in the second year. The experiments clearly demonstrate that populations under dense swards of clover were critically low before summer started, in contrast to the much higher populations of R. Eupiniunder serradella. It is considered that high soil temperature and desiccation during the summer are aggravating factors which lower populations already less than optimal in number. While strains more tolerant to heat and desiccation (Gillberg, 1968, 1969; Delin, 1969) would be useful, we maintain that improved soil colonization during the growing season is a prerequisite to ensuring adequate numbers for the second year. Acknowledgements-The authors gratefuhy acknowledge the technical assistance of Miss P. L. GROVE,and financial support from the Rural Credits Development Fund of the Reserve Bank of Australia and the State Wheat Industry Research Committee of Western Australia. We would also like to acknowledge the generosity of Mr W. D. GRIERSON who provided both land and assistance, and thank Mr A. E. OAKLEYfor the design of the equipment used to measure soil temperature. We wish to thank Dr D. GOODALLfor advice on the statistical treatment of the data. REFEREFXES ASHERC. J., OZANNEP. G. and L~NERAGANJ. F. (1965) A method of controlling the ionic environment of plant roots. Soil Sci. 100, 149-156. BOWEN G. D. and KENNEDYMARGARETM. (1959) Effect of high soil temperatures on Rhizobium spp. Qd. J. ugric. Sci. 16, 177-197. BROCKWELLJ. and WHALLEYR. D. B. (1970) Studies on seed pelleting as an aid to legume seed inoculation2. Survival of Rhizobium mrliloti applied to medic seed sown into dry soil. Aust. J. exp. Agric. Anim. Hush. 10,455-459. CHATELD. L. (1967) Studies on the Ecology of Rhizobium in Certain Sandy Soils in Western Australia. Ph.D. Thesis, University of Western Australia. CHATELD. L., GREENWOODR. M. and PARKER C. A. (t968) Saprophytic competence as an important character in the selection of Rh~~ob~umfor insulation. Transuct~onsof the ninth rnternutional Congress of Soif Science, Adelaide, Vol. 2, pp. 65-73. CHATELD. L. and PARKERC. A. (1964) Microbiolo~cal aspects of clover mortality in certain sandy soils in Western Australia: Population changes and the production of antiobiotics in situ in problem soils. Proceedings of the Australian Plant Nutrition Confereplce, Perth, CSIRO A(d) 10. CHATELD. L. and PARKERC. A. (1972) Inhibition of rhizobia by toxic soil-water extracts. Soil Biol. Biochem. 4, 289-294. CHOWDHURYM. S. (1965) The Growth and Persistence of Rhizobium lupini Schroeter in Sandy Soil. Ph.D. Thesis, University of Western Australia. CHOWDHURYM. S., MARSHALLK. C. and PARKER C. A. (1968) Growth rates of Rhizobium trifolii and Rhizobium lupini in sterilized soils. Aust. J. agric. Res. 19, 919-925. DATE R. A. (1968) Rhizobial survival on the inoculated legume seed, Transactions ofthe Ninth International Congress of Soil Science, Adelaide, Vol. 2, pp. 75-83. DATE R. A. and VINCENTJ. M. (1962) Determination of the number of root-nodule bacteria in the presence of other organisms. Aust. J. exp. Agric. Anim. Husb. 2, 5-7. DELIN S. (1969) Isolation of thermoduric strains of Rh~zob~um.Lantbr Hfigsk. Annir. 35,29-34. FISHERR. A. and YATESF. (1953) Stutistjeui Tables for Agricu~ti~ra~,3io~ogiea~,and other Research Workers. 4th Ed., p. 6 and Table VII12. Oliver 81 Boyd, London. GILLBERGB. 0. (1968) Heat resistance in Rhjzob~lirn. Arch. Mikrobiol. 62, 328-335.

SURVIVAL OF RHIZOBIA

423

GILLBERGB. 0. (1969) Heat resistance and pigmented variants of Rhizobium. Nuture, Lond. 222, 574. HAMDIY. A. (1970) Soil-water tension and the movement of rhizobia. Sail Biol. Biochem. 3, 121-126. RELY F. W., BERGER~EN F. J. and BROCKWELLJ. (1957) Microbial antagonism in the rhizosphere as a factor in the failure of inoculation of subterranean clover. Amt. J. ugric. Res. 8, 24-44. KININMO~H W. R. (1962) Soil temperature measurements. Perth, W. A. Mimeo, Rep. Bweau of ,%&teurology Australia. MARSHALLK. C. (1964) Survival of root-nodule bacteria in dry soils exposed to high temperatures. Amt. J. agric. Res. X5, 273-281. MARSHALLK. C., MULCAHYM. J. and CHOWDRURY M. S. (1963) Second-year clover mortality in Western Australia-a microbiological problem. J. Amt. Inst. agric. Sci. 29, 160-164. MILES A. A. and MISRA S. S. (1938) The estimation of the bactericidal power of the blood. J. Nyg., Cat& 38,732-748, MUNCH-PETERSEN E. and BOUNDY C. A. P. (1961) Use of rumen material in culture media for rumen bacteria. Amt. J. agric. Res. 12, 960-964. MUNCH-PETERSEN E. and BOUNDYC. A. P. (1963) Bacterial content in samples from different sites in the rumen of sheep and cows as determined in two culture media. Appl. Microbial. 11, 190-195. MUNCH-PETERSENE. and BOUNDYC. A. P. (1964) Culture media for rumen bacteria: An appraisal. i%/. Bakt. Abt. I, 191, 512-524. PARKERC. A. (1962) Light lands in Western Australia-3. MicrobioIogical problems in the establishment of legumes on light land. J. agric. W. Amt. (4th series) 3, 713-716. SANDERSON I. J. V. (1962) A Stady of Some Aspects of the Heat-Resistame of Rhizobium. Honours Thesis, University of Western Australia. THOMPSONJ. A. and VINCENTJ. M. (1967) Methods of detection and estimation of rhizobia in soil. PI. Soil 26, 72-94. VINCENTJ. M. (1958) Survival of the root-nodule bacteria. In Nutrition of the Legumes (E. G. Hallsworth, Ed.) pp. 108-123. Butterworths, London. VINCENTJ. M. (1962) Australian studies of the root-nodule bacteria. A review. Prof. Lim. Sue. N.S. W. 87, S-38. VINCENTJ. M., THOMPSON J. A. and DONOVANKATHLEEN0. (1962) Death of root-nodme bacteria on drying. Aust. J. agric. Res. 13, 258-270. VYASS. R. and PRA~ADN. (1960) Investigations on the failure of peas in ‘Goradu’ soils of Gujarat. Proc. Indian Acnd. Sci. B51,242-248 (see Marshall, 1964). WILKINSJEAN(1967) The effects of high temperatures on certain root-nodule bacteria. Aust. J. agric Res. 18, 299-304.

S.B.B.5/4-c