Soil Biol. Biochem. Vol. 25, No. 1, pp. 69-74, 1993 Printed in Great Britain. All rights reserved
0038-0717/93 $6.00 + 0.00 Copyright © 1993 Pergamon Press Ltd
*Present address: Department of Forestry, North Carolina State University, Box 8002, Raleigh, NC 27695, U.S.A.
†Author for correspondence.
ACCURACY OF MOST-PROBABLE-NUMBER ESTIMATES OF RHIZOBIA FOR TREE LEGUMES
DAN TURK* and HAROLD H. KEYSER†
NifTAL Project, University of Hawaii, 1000 Holomua Road, Paia, HI 96779, U.S.A. (Accepted 10 July 1992)
Summary-Rhizobia that nodulate tree legumes have rarely been enumerated by the most-probable number (MPN) plant infection assay. We compared MPN estimates of pure cultures of rhizobia with plate counts, using as hosts seedlings of 14 tree legume species grown separately in growth pouches or glass tubes. Reasonable agreement was obtained with 11 of the 14 tree species in one or both growth systems, with closer agreement in glass tubes than growth pouches for small-seeded species. Weekly assessment of MPN assays indicated that glass tubes and growth pouches should be scored for nodulation at least 7 and 5 weeks after inoculation. respectively.
Nitrogen-fixing leguminous trees are widely planted for fuelwood, fodder, construction material and biologically-fixed N for associated crops in agroforestry systems. In many tropical soils inadequate rhizobial populations limit NZ fixation (Singleton et al., 1992). In such soils, inoculation with appropriate rhizobia can improve yield, provided no other constraints limit growth.
Estimating rhizobial population density is a means of predicting whether or not a legume will respond to inoculation (Thies et al., 1991b). Although there is no direct way of measuring the density of indigenous rhizobial populations in soil (Brockwell, 1980), the most-probable-number (MPN) plant-infection assay (Vincent, 1970) provides an indirect estimate. MPN assays for enumerating rhizobia are also used in ecological studies, and for evaluating inoculant quality (Brockwell, 1980).
Although frequently used with grain and forage legumes, the MPN assay has seldom been used with tree legumes as hosts (Table 1). There is, therefore, a need to evaluate the accuracy of the technique when used with trees. Comparisons of pure cultures with plate counts have been recommended as a means of assessing the accuracy of estimates obtained from particular legume-rhizobium growth systems (Brockwell, 1963). Such comparisons have only rarely been made with tree legumes (Table 1).
The main objective of our experiments was to evaluate the accuracy of MPN assays using pure rhizobial cultures by comparing MPN estimates with plate counts. Other objectives were to determine an appropriate MPN growth system for each of the tree species used and to determine how long the assays needed to be maintained
before scoring for nodulation.
MATERIALS AND METHODS
Two growth systems were used in this study: plastic growth pouches (Northrup King Co.) and agar slants in 25 x 250 mm glass tubes. Acacia auriculiformis A. Cunn. ex Benth., Acacia mangium Willd., Acacia mearnsii De Wild., Leucaena diversifolia (Schlecht.) Benth., Paraserianthes falcataria (L.) Nielsen, Robinia pseudoacacia L., and Sesbania grandii lora Poir. were grown in pouches and on agar slants. Additional MPN assays were conducted in pouches with Albizia lebbeck (L.) Benth., Albizia saman (Jacq.) F. Muell., Calliandra calothyrsus Meissn., Enterolobium cyclocarpum Griseb., Flemingia macrophylla (Willd.) Merrill, Gliricidia sepium (Jacq.) Steud., Leucaena leucocephala (Lam.) de Wit, and S. sesban (L.) Merr. Seeds of all species were obtained from either the Nitrogen Fixing Tree Association (Box 680, Waimanalo, HI 96795, U.S.A.) or NifTAL Project seed collections.
Agar slants were prepared using 15 ml per tube of N-free nutrient solution modified from Singleton (1983) and 15 g agar 1-'. After autoclaving, the agar was allowed to cool in the tube slanted at an angle of 10° from the horizontal, producing an agar surface ca 10 cm long, beginning at the base of the tube. Growth pouches were prepared by filling each pouch with 50 ml of N-free nutrient solution, minus the agar. An additional 50 ml of N-free nutrient solution was added prior to inoculation. After inoculation pouches were provided with sterile deionized water as needed until harvest.
Seeds were appropriately scarified and surface sterilized. After imbibition, seeds were germinated on water-agar plates (Somasegaran and Hoben, 1985).
arabinose-gluconate medium (Sadowsky et al., 1987). Cultures were serially diluted using either 4- or 10-fold dilution ratios as outlined in Somasegaran and Hoben (1985). The diluent contained the salts found in yeast-extract mannitol broth (Vincent, 1970) with 0.01 % Tween 80 (Fisher Scientific Co.) added as a surfactant.
Seedlings were inoculated 7-12 days after planting, when root radicles had reached the bottom of the tubes or had begun to differentiate into secondary roots in pouches. Diluted rhizobial culture (1 ml) was applied directly to roots in pouches and to the lower portion of roots in tubes. An average of seven uninoculated growth units were included per MPN assay as checks for contamination within the system. Uninoculated controls received 1 ml of rhizobia-free diluent.
Plate counts were conducted at the time of inoculation using the Miles and Misra drop plate method (Somasegaran and Hoben, 1985). At least four replicate 30 pl aliquots were used from each of three dilution levels from the inoculation dilution series. Prior to data collection, plates were kept at 27°C for 7-10 days for strains of Bradyrhizobium Jordan, and for 3-5 days for strains of Rhizobium Frank.
Two to five days later, when radicles had reached 0.5-1.5 cm in length, the seed coats of uniform seedlings were removed to prevent seed coat hardening and to facilitate examination of tubes for nodulation. In tubes, seedlings were placed on the surface of the agar, with the root collar ca 6-7 cm from the bottom of the slant. One and two seedlings were planted in tubes and pouches respectively. After planting, racks of pouches and tubes were placed in a growth room receiving > 300 μE m-2 s-1 of photosynthetically-active radiation at plant height for 16 h day--1 from 1000 W high-pressure sodium lamps.
Prior to inoculation, growth units with poorly growing plants were eliminated. These included tubes with seedlings whose tap roots penetrated the surface of the agar, in accordance with the observation of Woomer et al. (1988b) that agar penetration by roots of L. leucocephala resulted in poor nodulation. Enough uniform plant growth units were retained for at least six dilution levels of four replicate growth units each.
Rhizobial cultures, inoculation, and plate counts
Rhizobial strains (Table 2) were grown for 6-10 days in either yeast-extract mannitol broth (Vincent, 1970) or
MPN estimates for tree legumes
> 25 mg seed-' (S. grandiora, A. lebbeck, Albizia saman, C. calothyrsus, G. sepium and L. leucocephala), all but two (A. lebbeck and Albizia saman) had PC: MPN ratios < 35. Of the seven species grown in tubes and pouches, five (A. auriculiformis, Acacia mangium, A. mearnsii, P. falcataria and S. grandi flora) had PC: MPN ratios > 200 in one of the two growth systems.
Five of 10 MPN estimates from nodulation scores taken weekly from 2 to 7 weeks after inoculation increased after the fourth week (Fig. 1). No nodules were formed in any of the 48 uninoculated control growth units used in these assays. For species grown in tubes and pouches, the average number of weeks to arrive at the final MPN estimate was 6.0 in tubes compared with 4.8 in pouches. In tubes, the MPN estimates of A. auriculiformis and A. mearnsii increased during the interval from 6 to 7 weeks after inoculation. In pouches, the MPN estimate of only one species, Acacia mangium, increased in the interval from 5 to 6 weeks and no species after 6 weeks.
With two exceptions scored at 4 weeks, growth units of MPN assays were scored for nodulation 5-7 weeks after inoculation. The MPN was determined using a computer program (Woomer et al., 1990), beginning with the lowest dilution step in which all growth units nodulated.
MPN assays in pouches and tubes containing A. auriculiformis, Acacia mangium, A. mearnsii and R. pseudoacacia, were scored weekly for nodulation from the second to the seventh week after inoculation. Weekly scores were also recorded for S. grandiflora and L. diversifolia, grown in pouches and tubes respectively.
Plate counts were within the corresponding 95% confidence interval of MPN estimates (Cochran, 1950) for 18 of 53 plate count-MPN comparisons (Table 3). In all but three plate count-MPN comparisons the MPN was lower than the plate count.
Rhizobial density of the original solutions ranged from 3.67 x 10' to 7.55 x 109 rhizobia ml-' as measured by the plate count method. In every MPN assay, nodules were found in only 5 of 384 uninoculated control growth units, with never more than one nodulated control unit per assay.
Plate count: MPN estimate (PC: MPN) ratios < 35 were obtained in either tubes or pouches for 11 of the 14 tree species (Table 3). In tubes, all species with seed weights < 25 mg seed-' (A. auriculiformis, Acacia mangium, A. mearnsii, L. diversifolia, P. falcataria and R. pseudoacacia) had average PC: MPN ratios < 26. In pouches, of six species with seed weight
Comparing MPN estimates with plate counts of pure rhizobial cultures indicates the degree of compliance of the MPN assay with its major underlying assumption, that the presence of a single rhizobium will result in nodule formation (Scott and Porter, 1986). The comparison provides an indication of a growth system's potential for accurate results using a particular species and set of growth conditions. This is the basis for recommending that pure culture comparisons of MPN estimates with plate counts accompany MPN assays conducted in soil for each growth system-species combination used
(Thompson and Vincent 1967; Scott and Porter, 1986; Singleton et al., 1991).
MPN estimates obtained with a variety of grain and forage legumes indicate that the MPN assay tends to underestimate plate counts, suggesting that one rhizobium is generally not sufficient to cause nodule formation. PC: MPN ratios < 6 have been obtained with many herbaceous legumes including Trifolium spp (Brockwell, 1963; Tuzimura and Watanabe, 1961), Medicago sativa (Weaver and Frederick, 1972; Scott and Porter, 1986), Cicer arietinum (Toomsan et al., 1984) and Glycine spp (Weaver and Frederick, 1972; Brockwell et al., 1975). The discrepancy of < 1.5 log units between MPN estimates and plate counts obtained with A. mearnsii, L. diversifolia and P. falcataria in tubes and with L. diversifolia, C. calothyrsus and S. sesban in pouches, indicates that reasonable agreement between MPN estimates and plate counts can also be obtained with a variety of tree legumes.
Our results with trees agreed with the recommendation of Vincent (1970) that agar slants are suitable for legumes with seed weight less than that of "vetch", ca 25 mg seed-1. Growth pouches were suitable for most larger-seeded tree species, but not for A. lebbeck and Albizia saman. Other growth systems, such as Leonard jars, should be evaluated for these species.
Growth pouches were not suitable growth systems for seven of the 14 tree species and tubes not suitable for S. grandiflora. These combinations had PC: MPN ratios > 100, indicating that over 100 rhizobia were required for nodule formation in these particular systems. Other authors have reported large discrepancies between plate counts and MPN estimates, notably Boonkerd and Weaver (1982), who reported PC: MPN ratios in pouches > 250 for Vigna unguiculata and Macroptilium atropurpureum. Boonkerd and Weaver (1982) suggested that the large ratios obtained were due to the inherent inability of these legumes to nodulate with a single rhizobial cell.
MPN estimates for tree legumes
However, large PC: MPN ratios from species grown in only one of the two growth systems used in these experiments imply that problems with nodulation are growth-system related. The range of PC: MPN ratios observed emphasizes the need to evaluate growth system-species combinations prior to performing MPN assays for experimental purposes.
Woomer et al. (1988b) noted the importance of placing the diluted inoculant directly on the roots of the plants to get good nodulation. They noticed that the roots of M. atropurpureum tended to accumulate at the bottom of the glass tube without growing extensively on the surface of the slanted agar where the inoculant was applied and found that estimating rhizobial populations with this species was less precise than for species with greater root proliferation. We observed that roots of the Acacia spp, in particular, did not grow extensively on the agar surface, with most nodulation occurring at the bottom of the tubes. The quantity of agar we used and the angle of the slant, were, therefore, designed to channel both roots and rhizobia directly to the bottom of the tube to ensure maximal contact between them. We found that 15 ml of agar was sufficient to support growth of tree seedlings for the duration of the assays.
Weekly assessment of MPN estimates (Fig.1) indicated the importance of keeping tubes for at least 7 weeks and pouches for at least 5 weeks after inoculation. Delayed nodule formation in tubes relative to pouches appeared to be related to longer retention of photosynthetic capability: seedlings in tubes remained healthy longer than plants in pouches, despite being smaller. These times for scoring nodulated growth units are longer than recommended for other legumes. For grain and forage legumes, 2-4 weeks have been recommended as adequate by Vincent (1970), Brockwell (1980) and Boonkerd and Weaver (1982). Reports of times to assessment of MPN assays with tree legumes (Table 1) indicate a range from 2 to 3 weeks to 11 weeks after inoculation, with nodulation in four of seven reports assessed at S 3 weeks. Our results suggest that scoring at < 5 weeks is likely to underestimate the true population density, at least for species other than L. leucocephala
In conclusion, the results from our experiments provide a basis for selecting appropriate growth systems for MPN assays with a range of tree species. They show that MPN assays can provide accurate estimates of rhizobial populations for tree legumes but point to the need for selecting growth systems that are tailored to the characteristics of the particular tree species being used. It is imperative that growth systems be tested for each tree species using pure cultures of rhizobia prior to conducting MPN assays with soils.
Acknowledgements-We thank Kathy MacGlashan, Patty Nakao and Joe Rourke for technical assistance, and Paul Singleton and Janice Thies for reviewing the manuscript.
This publication was made possible through support provided under cooperative agreements by the United States Agency for International Development, Office of Agriculture, Bureau of Research and Development. Conclusions in this paper and mention of product names do not constitute an endorsement by USAID. Journal series No. 3663 of the Hawaii Institute of Tropical Agriculture and Human Resources.
Boonkerd N. and Weaver R. W. (1982) Cowpea rhizobia: comparison of plant infection and plate counts. Soil Biology & Biochemistry 14, 305-307.
Brockwell J. (1963) Accuracy of a plant-infection technique for counting populations of Rhizobium trifolii. Applied Microbiology 2, 377-383.
Brockwell J. (1980) Experiments with crop and pasture legumes-principles and practice. In Methods for Evaluating Biological Nitrogen Fixation (F. J. Bergersen, Ed.), pp. 417-488. Wiley, Chichester.
Brockwell J., Diatloff A., Grassia A. and Robinson A. C. (1975) Use of wild soybean (Glycine ussuriensis Regel and Maack) as a test plant in dilution-nodulation frequency tests for counting Rhizobium japonicum. Soil Biology & Biochemistry 7, 305-311.
Cochran W. G. (1950) Estimation of bacterial densities by means of the "most-probable-number". Biometrics 6, 105-116.
Davis P. E. (1982) Desmanthus virgatus as a test plant for dilution-frequency infection tests using Leucaena leucocephala rhizobia. Soil Biology & Biochemistry 14, 313-314.
Sadowsky M. J., Tully R. E., Cregan P. B. and Keyser H. H. (1987) Genetic diversity in Bradyrhizobium japonicum serogroup 123 and its relation to genotype-specific nodulation of soybean. Applied and Environmental Microbiology 53, 2624-2630.
Sambrook J., Fritsch E. F. and Maniatis T. (1989) Molecular Cloning: A Laboratory Manual. 2nd Edn, Vol. 1. Cold Spring Harbor Laboratory Press, Cold Spring Harbor.
Sanginga N., Mulongoy K. and Ayanaba A. (1987) Evaluation of indigenous strains of Rhizobium for Leucaena leucocephala (Lam.) De Wit in Nigeria conditions. In Les Arbres Fixateurs d'Azote: 1'Amelioration Biologique de la Fertilite des Sols, pp. 416-436. Colloques et Seminaires. ORSTOM, Paris.
Sanginga N., Mulongoy K. and Ayanaba A. (1985) Effect of inoculation and mineral nutrients on nodulation and growth of Leucaena leucocephala. In Biological Nitrogen Fixation in Africa: Proceedings of the African Association for Biological Nitrogen Fixation (H. Ssali and S.O. Keya, Eds), pp. 419-427. Matianum Press Consultants, Nairobi.
Scott J. M. and Porter F. E. (1986) An analysis of the accuracy of a plant infection technique for counting rhizobia. Soil Biology & Biochemistry 18, 355-362.
Singleton P. W. (1983) A split-root growth system of evaluating components of the soybean-Rhizobium japonicum symbiosis. Crop Science 23, 259-262.
Singleton P. W. and Tavares J. W. (1986) Inoculation response of legumes in relation to the number and effectiveness of indigenous rhizobium populations. Applied and Environmental Microbiology 51, 1013-1018.
Singleton P. W. Bohlool B. B. and Nakao P. L. (1992) Legume response to rhizobial inoculation in the tropics: myths and realities. In Myths and Science of Soils in the Tropics (R. Lal and P. Sanchez, Eds), SSSA Special Publication No. 29, pp. 135-155. American Society of Agronomy, Madison.
Singleton P. W., Woomer P. L., Thies J. E., Nakao P. L. and Bohlool B. B. (1991) Protocols for Field and Greenhouse Experimentation in Legume BNF: Standardized
DAN TURK and HAROLD H. KEYSER
Network Trials to Develop Models Describing the Ecology and Performance of the Rhizobium-Legume Symbiosis. University of Hawaii NiFFAL Project, Paia.
Somasegaran P. and Hoben H. J. (1985) Methods in Legume -Rhizobium Technology. University of Hawaii NiFTAL Project, Paia.
Thies J. E., Singleton P. W. and Bohlool B. B. (1991a) Influence of the size of indigenous rhizobial populations on establishment and symbiotic performance of introduced rhizobia on field-grown legumes. Applied and Environmental Microbiology 57, 19-28.
Thies J. E., Singleton P. W. and Bohlool B. B. (1991b) Modeling symbiotic performance of introduced rhizobia in the field by use of indices of indigenous population size and nitrogen status of the soil. Applied and Environmental Microbiology 57, 29-37.
Thompson J. A. and Vincent J. M. (1967) Methods of detection and estimation of rhizobia in soil. Plant and Soil 26, 72-84.
Toomsan B., Rupela O. P., Mittal S., Dart P. J. and Clark K. W.(1984) Counting Cicer-Rhizobium using a plant infection technique. Soil Biology & Biochemistry 16, 503-507.
Tuzimura K. and Watanabe I. (1961) Estimation of number of root-nodule bacteria by a nodulation-dilution frequency method. Soil Science and Plant Nutrition 7, 61-65.
Vincent J. M. (1970) A Manual for the Practical Study of Root-Nodule Bacteria. Blackwell, Oxford.
Virginia R. A., Jenkins M. B. and Jarrell W. M. (1986) Depth of root symbiont occurrence in soil. Biology and Fertility of Soils 2, 127-130.
Weaver R. W. and Frederick L. R. (1972) A new technique for most-probable-number counts of Rhizobia. Plant and Soil 36, 219-222.
Woomer P., Bennett J. and Yost, R. (1990) Overcoming the inflexibility of most-probable-number procedures. Agronomy Journal 82, 349-353.
Woomer P., Singleton P. W. and Bohlool B. B. (1988a) Ecological indicators of native rhizobia in tropical soils. Applied and Environmental Microbiology 54, 1112-1116.
Woomer P., Singleton P. W. and Bohlool B. B. (1988b) Reliability of the most-probable-number technique for enumerating rhizobia in tropical soils. Applied and Environmental Microbiology 54, 1494-1497.