i^T'*. D i v i s i o Agricultural Sciences l\ \; UNIVERSITY OF CALIFORN "0 c I ■ < 14* A * •) L * hili HI^BHBHIHHI^H HHHH^I HHBll CALIFORNIA AGRICULTURAL EXPERIMENT STATION BULLETIN 842 Xhe great importance of legumes in agriculture is that they add nitrogen to the soil and thereby save the costs of nitrogen fertilizer, both for the legume crop and for the associated nonlegume crop. Leg- umes obtain nitrogen from the air, which contains uncombined nitro- gen — nearly 80 per cent — along with oxygen and other gases. Very few plants can utilize atmospheric nitrogen in its free form. However, if root-nodule bacteria of an effective strain have formed nodules on the roots of a legume, atmospheric nitrogen may be fixed within the nodules and converted to a utilizable compound. The successful establishment of legumes, particularly in a pasture mix for grazing, depends on effective nodulation. This can be obtained by inoculating the seed with an appropriate strain of root-nodule bac- teria. Methods of inoculating seed and measures that help to avoid inoculation failure are given in boxes on the following pages. All of the recommendations given here are of critical importance. There is no economical way to inoculate a field after planting. Faulty inoculation usually results in failure or partial failure of the legume stand. From this cause alone, California growers waste many thousands of dollars' worth of range-legume seed every year and waste also the labor and the fertilizer used. The Authors A. A. Holland was Lecturer in Agronomy and Assistant Research Agronomist in the Experiment Station, Davis, at the time this work was done. He is now at Monash Uni- versity, Clayton, Victoria, Australia. J. E. Street is Range Improvement Specialist, Agricultural Extension Service, Davis. W. A. Williams is Professor of Agronomy and Agronomist in the Experiment Station, Davis. AUGUST 1969 An abstract appears on page 19 RANGE-LEGUME INOCULATION AND NITROGEN FIXATION BY ROOT-NODULE BACTERIA INTRODUCTION nitrogen is an essential constituent of the amino acids and proteins and is the key element in the structure of all living cells. Legumes are specialized plants because of their faculty for fixing atmospheric nitro- gen into compounds that can be utilized in plant metabolism. Actually, the legumes alone cannot accomplish this. The process takes place only in nitrogen-fixing nodules, which are formed on legume roots in sym- biotic association with root-nodule bac- teria of the genus Rhizobium. Both mem- bers of the symbiosis have their particular functions, and each step of the process in- volves a close interaction between them. In this association both members derive advantages. The rhizobia live within the root tissues, where they are protected from competitors and receive nutrients, such as carbohydrates. The legumes receive amino acids, which they build into pro- teins. However, neither the legumes nor the rhizobia have any nutrient advantage without the symbiosis. Together they can fix surprisingly large quantities of atmo- spheric nitrogen. Donald (1960) estimated that the world's annual income of biologically fixed nitro- gen, mainly from symbiotic sources, was of the order of 100 million tons. Russell (1950) estimated that well-nodulated leg- umes could fix up to 500 pounds of nitro- gen per acre per year — the equivalent of 2,352 pounds per acre of ammonium sul- fate. Erdman (1959) gave 106 pounds per acre as the average amount of nitrogen fixed in a legume-grass pasture in North America. Under California dryland conditions, effectively nodulated legumes are able to fix significant amounts of nitrogen, though the amounts do not approach Russell's estimate. A reasonable average for a good stand of range legumes might be at least 52 pounds per acre during one growing season — equivalent to 250 pounds per acre of ammonium sulfate. The cost of the inoculant is about 15 cents an acre. The following tabulation gives data from Cali- fornia. The amounts of atmospheric nitro- gen fixed in the clover roots were esti- mated by comparing the yields from well- nodulated, thrifty legume stands with those from unthrifty stands in the same area. The amount of nitrogen in purple vetch was compared with that in barley, a nonlegume. LEGUME, COUNTY, NITROGEN GAIN AND REFERENCE LB./ACRE Subterranean clover, Hum- boldt (Williams, Lenz, and Murphy, 1954) 45 Rose clover, Stanislaus (Williams, Love, and Berry, 1957) 50 Purple vetch, Santa Barbara (Williams, Ririe, et al., 1954) 53 Rose clover, Placer (Martin, Williams, and Johnson, 1957) 60 1 Submitted for publication September 29, 1964. Cover photo: Kondinin rose clover plants 105 days old. Vigorous plant at left grown from pellet-inoculated seed; small plant at right from uninoculated seed, planted at the same time and only a few feet away. Yolo County, March 1, 1967. 3] ROOT NODULES AND NITROGEN FIXATION Nodule development on a legume root The first stage in the symbiosis between rhizobia and legumes is the multiplication of rod-shaped rhizobia in the soil in the region of a legume root. The multipli- cation is initiated and stimulated by sub- stances secreted by the root and exuded from it. One of these exudates is trypto- phan (Rovira, 1959). The rhizobia convert tryptophan to indoleacetic acid (Kefford, Brockwell, and Zwar, 1960), which causes curling of certain root hairs — a prelude to the invasion of these root hairs by the rhi- zobia. This process occurs only at certain specialized infection sites — foci — on a young root (Nutman, 1948 and 1949; Dart and Pate, 1959). The infection foci remain available for only a limited time. As the root elongates, the new infection foci are progressively more distant from the seed. Many strains of rhizobia are able to multiply in the rhizosphere of any legume SEED INOCULATION 1 . Select a good commercial inoculant, pre- pared from a strain of root-nodule bacteria that is specific for the legume to be planted. Make sure the legume species is named on the label; no other inoculant will do. Mixtures of bacterial strains suitable for a great num- ber of different species and varieties of leg- umes are not satisfactory. 2. Check the freshness of the inoculant by the expiration date stamped on the con- tainer. The inoculant is a living culture of root-nodule bacteria, which are killed by dry- ing and also by high temperatures. Make sure this culture has been stored continuously un- der refrigeration. Poor storage conditions can cause nodulation failure by reducing the number of viable bacteria in the inoculant. 3. Use the inoculant before the expiration date. Do not carry it over to another year. The number of bacteria that survive will be too low for effective nodulation. 4. Use the inoculant at the rate recom- mended by the manufacturer and stamped on the container, or else at a higher rate — up to four times this amount (see paragraph 9, be- low). Never use a lower amount. 5. Mix inoculant with seed thoroughly. Follow closely the printed directions on the container for the wet-mix method, but do not use excess moisture, Make sure the inocu- lant sticks to all the seeds. 6. Hold inoculated seed in a cool, shady place, and plant it in a moist seedbed. Guard against drying, both before and after planting. Always remember that drying kills ' these bacteria quickly. 7. Plant as soon as possible. Reinoculate seed if it has been held longer that 24 hours after inoculating. The cost of inoculant is small in comparison with the costs of seed, ., fertilizer, soil preparation, and planting. Pel- let the seed (see next paragraph) if planting must be delayed. 8. Pellet the inoculant on the seed to pro- tect the bacteria from drying. Directions for pelleting are in the box on page 6. Always use pellet-inoculated seed if it must be sown in a ^ dry seedbed, or broadcast without a soil cover, or sown from the air. Even with pelleting the * protection is temporary, and the inoculant becomes valueless in three weeks under dry i conditions. 9. Always use pellet-inoculated seed where' 1 native legumes are abundant. This includes many areas of California. Pellet with four times the amount of inoculant recommended ^ by the manufacturer. The root-nodule bacteria already in the . soil in such areas often act as parasites on in- troduced legumes. These bacteria are ineffec- ' tive nitrogen fixers with introduced legumes, but they do invade the young roots in great numbers and thus prevent adequate nodu- lation of a new legume by bacteria of the ef- fective strain supplied in the inoculant. It re- 4 quires a large amount of inoculant to compete [4 and can initiate the preliminary stages of root-hair curling described above. How- ever, because of the specific relationship between a host legume and a compatible strain of rhizobia, the bacteria that cause curling of root hairs may belong to a strain that cannot nodulate the available legume. If so, these rhizobia do not enter the root hairs. The following reactions proceed only if the strain of Rhizobium is invasive, i.e., capable of stimulating nodule formation in a legume root. Rhizobia of an invasive strain produce an extracellular polysaccha- ride slime, which induces the legume to secrete polygalacturonase (Fahraeus and Ljunggren, 1959). The function of poly- galacturonase has not been fully deter- mined. It may act with indoleacetic acid to increase the plasticity of the cellulose wall of a root hair and enable it to invagi- nate (fold inward). The root hair responds to the presence of invasive rhizobia by folding inward at some point near its tip to form a very fine tube, the infection thread, which extends inside the root hair (Nutman, 1963). Concurrently with the formation of an infection thread, a specialized tetraploid cell forms in the adjacent root-cortex AND FIELD PROBLEMS with a large population of ineffective bacteria. The pellet concentrates the inoculant around the germinating seed and the emerging root, so that a nitrogen-fixing nodule is started im- mediately by the effective bacteria supplied in the inoculant. 10. Do not mix acid fertilizers with inocu- lated seed, and do not sow in contact with such fertilizers. The acidity may kill most or all of the root-nodule bacteria. 11. For the same reason, do not mix the seed with fertilizers that contain trace elements. 12. Check the acidity of the soil where in- oculated seed is to be sown. Nodulation is usually defective where the soil is more acid -than pH 5.2. 13. Do not use herbicides, fungicides, or any other pesticides when planting inoculated seed. Most of these poisons are highly toxic to root-nodule bacteria. „_ 14. Make sure the soil contains adequate amounts of plant nutrients. A legume suffer- ing from deficiency of any nutrient other than nitrogen cannot benefit fully from inocu- lation. To realize grazing potentials, the leg- umes must have an adequate supply of avail- able phosphorus and sulfur — the elements most commonly deficient on California range- land. Many pasture legumes — for example, subterranean clover — have poorer phospho- rus-foraging ability than have the grasses. 15. Do not use nitrogenous fertilizers when establishing legumes in a pasture mix. Well- nodulated legumes do not need any nitrogen from the soil, and nitrogenous fertilizers usu- ally give grasses a competitive advantage over the legumes. Moreover, nitrates and nitrites inhibit legume nodulation under most cir- cumstances. 16. Maintain adequate pasture manage- ment. In a grass-dominant pasture, direct every effort toward making the environment favorable for the legume rather than for the grasses. A good legume stand should have at least 20 plants per square foot. Grazing the pasture reduces competition from grasses and forbs. Reduce grazing pressure only while the legumes are flowering and seeding, because abundant production of legume seed is es- sential. However, pastures must be grazed very heavily in summer after the legume seed is ripe. 17. Dig a few seedlings after the legumes have produced three or four leaves. The type and pattern of nodulation can give useful information. a. Lack of nodulation usually indicates some fault in the inoculation or sowing tech- nique. Review the above instructions. b. A few large nodules on the crown or up- per root indicate early, effective nodulation — provided these nodules are pink inside. c. Many small, white nodules scattered over the entire root system indicate ineffective nodulation by root-nodule bacteria already in the soil (see paragraph 9, above). [5] SPECIAL PROCEDURE pelleting is an improved method of seed inoculation. Each pellet contains a legume seed, the inoculant in an adhesive, and a coat- ing of calcium carbonate. Both the adhesive and the coating material assist the survival of the inoculant. The following recommen- dations are based on experiments in Califor- nia. Inoculant. Use peat inoculant prepared specifically for the legume to be planted. Mixed inoculants prepared for a great variety of legumes are not satisfactory. The inoculant must be fresh and of good quality. Use four times the amount recommended on the con- tainer. Review paragraphs 1, 2, 3, and 9 in the preceding box. Adhesive. Use gum arabic ground to pass an eight-mesh screen — or finer. It must be with- out preservative. Do not dissolve more gum arabic than will be used in 36 hours. Without preservative, the gum arabic is decomposed rapidly by fungi and bacteria. These destroy the sticking properties of the gum and may make it toxic to the root-nodule bacteria. Coating. Ground calcium carbonate, CaCOs, is the most uniform and beneficial* of the many materials tested. It is marketed under the labels: lime, calcium carbonate, calcite, enamel whiting, or 280 whiting. Do_ not use quicklime; it is highly toxic to root- nodule bacteria. The coating material should be finely ground, so that it does not flake off from the finished pellet: At least 80 per cent' should pass through a 200-mesh screen. Quantities of pelleting materials needed vary with the size of the seed. For the ad-^ hesive, use 4 pounds of gum arabic powder to 1 gallon of water. This amount makes abou-;- 5 quarts of solution. • Vetch. For 100 pounds of seed, use 21/4 quarts of gum arabic solution and 30 pounds, of calcium carbonate. • Subterranean, rose, or crimson clover/ For 100 pounds of seed, use 5 quarts of gum arabic solution and 50 pounds of calcium carbonate. •Alfalfa or bur clover. For 100 pounds of seed, use 5 quarts of gum arabic solution and 40 pounds of calcium carbonate. tissue. As the infection thread elongates within the root hair and approaches this specialized cell, it stimulates cell division both in the tetraploid cell and in the neighboring diploid cells. The mass of cells so formed develops into a nodule. The in- fection thread branches and spreads throughout the nodule, in the central zone of the actively dividing tetraploid cells. Rod-shaped rhizobia invade the cytoplasm of the tetraploid cells by way of the infec- tion thread (Goodchild and Bergerson, 1966). They multiply further and change into bacteroids. The bacteroid is club-shaped. It is an incomplete growth stage in the sense that it has no cell wall but only a cytoplasmic membrane. Bacteroids are able to increase their numbers by cell division inside a nodule, but they never return to the rod shape. After the nodule has formed and if hemo- globin has developed in the nodular cells, nitrogen fixation may proceed. Nitrogen is fixed only in tissue containing both bac- teroids and hemoglobin. The presence of hemoglobin in a nodule is evidence of ef- fective nitrogen fixation. It is indicated by the pinkish color of the nodule tissue, seen most clearly when the nodule is sliced open. Nutman (1963) gives further details of nodule formation and the factors in- volved. Theory of symbiotic nitrogen fixation Bergersen and Briggs (1958) studied ma- ture, infected nodule cells by electron microscopy. They found that the cyto- plasm of these cells is filled by a complex system of membrane envelopes, which are formed from folds of the outermost proto- plasmic membrane of the host cytoplasm and enclose small groups of bacteroids. They suggested that these membranes play an important part in nitrogen fixation, be- [6 FOR SEED PELLETING Preparing the pellets 1. Use a cement mixer for large quantities of seed. Small lots may be pelleted by hand in a tub or a bucket, or on a smooth floor. 2. Calculate the appropriate quantities of gum arabic and water to use with the desired quantity of the particular seed to be pelleted. 3. In a separate container, dissolve gum ar- abic in water. There are two possible meth- ods: Either add the gum powder to the water slowly, while stirring vigorously, or else make a paste by adding half the water to the powder and then dilute with the remaining water. .Gum arabic dissolves in cold water if left over- night, or in hot water in about 30 minutes. Do not let the solution boil. 4. Cool the gum arabic solution. Just before pelleting, add the appropriate amount of inoeulant and stir to a smooth slurry. This mixture must not stand for more than 30 minutes. Some gum arabic is acid and will harm the bacteria unless the acid is neutral- ized by the calcium carbonate as soon as pos- sible. 5. Pour the seed into the mixer. Add the gum-inoculant mixture and rotate the mixer at high speed, for good tumbling action, until all the seeds are coated. 6. Without stopping the mixer, dump in the calcium carbonate all at once and let the mixer run until all the seeds are pelleted. 7. Do not clean the mixer between loads. After the whole job is done, clean the mixer by running a load of water and gravel through it. 8. Pellets are firmer if they are allowed to season for 24 hours. They then work better for drill-seeding. 9. Screen the pelleted seed to remove lumps; they are likely to clog the seeding equipment. If there is an excess of calcium carbonate pow- der, screen or winnow it out so as not to clog the seeding equipment. The proportions given above allow for an excess of calcium carbonate, because the stickiness of the ad- hesive may vary slighty in different lots. 10. Remove vigorous agitators from seeding equipment, to prevent injury to the pellet coating. cause nitrogen is fixed only inside the mem- branes. This research and the associated work of Appleby and Bergersen (1958), Bergersen (1958, 1960a, 1960&), and Ber- gersen and Wilson (1959) led to the formu- lation of a hypothesis to explain symbiotic nitrogen fixation. The process is summa- rized below. For more technical data see Vincent and Waters (1962). In symbiotic nitrogen fixation, five pri- mary reactions occur in the soluble cyto- plasm inside the membrane envelopes — not on the membrane surface, as Bergersen proposed initially. 1. Free nitrogen occurs in this cytoplasm. Here it is activated — that is, ionized. 2. The legume supplies carbohydrates and other carbon compounds, which are partially oxidized by action of the bacte- roids — releasing electrons, which are the energy source. 3. Hemoglobin, in solution in the cyto- plasm, combines loosely and reversibly with these freed electrons, thus providing an essential linkage for their transport. The electron-transport train begins with the bacteroids. The ultimate acceptor of the electrons is activated nitrogen, which is thereby reduced to ammonia. 4. Products of the partial oxidation of the carbon compounds then serve as ac- ceptors of ammonia in the process of amino acid production by the bacteroids. 5. Most of these amino acids are avail- able to the host legume for production of protein, nucleic acids, and other nitrog- enous plant products. Breakdown of nodules After some weeks or months of nitrogen- fixing activity, the host nodule breaks down. The bacteroids die and undergo lysis — a process of disintegration or dis- solution. Rod-shaped rhizobia — previously dormant inside the infection thread (see nodule development, above) — then be- [7] come active, invade the senescent nodular tissue, and multiply there but do not be- come bacteroids. When the nodule decays further, these new rods are released into the soil. They may start another cycle of infection on the same legume root or they may oversummer in the soil and start a new cycle in legumes in the next growing season. Neither the rod form of rhizobia nor the bacteroid form produces spores or odier specialized reproductive cells, as do many bacteria. Therefore, they are always in the vegetative state and are subject to rapid death on drying. If host legumes are not grown in the soil, the numbers of rhizobia decline gradually (Krasil'nikov, 1958; Ro- vira, 1961). SPECIFIC RELATIONSHIPS BETWEEN HOST LEGUMES AND RHIZOBIA The relationship of a species or strain of Rhizobium to certain legumes is fixed and usually specific. Invasibility Usually one species of Rhizobium can cause formation of nodules with many of the legumes in one or more genera. Thus, the agriculturally important legumes have been classified into seven so-called cross- inoculation groups — namely, the clover, alfalfa, pea and vetch, bean, lupine, soy- bean, and cowpea groups. Within each group of legumes there is commonly a mu- tual compatibility of the host plants with several related strains of one Rhizobium species. Moreover, rhizobia of this one species do not usually invade legumes out- side their particular group. The following list gives only the more important forage legumes in three of the cross-inoculation groups. l.The clover group, nodulated by Rhizobium trifolii Alsike clover, Trifolium hybridum Berseem clover, T. alexandrinum Crimson clover, T. incarnatum Ladino clover, T. repens giganteum Red clover, T. pratense Rose clover, T. hirtum Strawberry clover, T. fragiferum Subterranean clover, T. subterraneum White clover, T. repens 2. The alfalfa group, nodulated by Rhi- zobium meliloti Alfalfa, Medicago saliva Barrel medic, M. tribuloides Black medic, M. lupulina Bur clover, M. hispida Fenugreek, Trigonella foenumgraecum Hubam clover, Melilotus alba annua Sourclover, M. indica White sweetclover, M. alba Yellow sweetclover, M. officinalis 3. The pea and vetch group, nodulated by Rhizobium leguminosarum Common vetch, Vicia sativa Field pea and garden pea, Pisum sativum Hairy vetch, Vicia villosa Horsebean, V. faba Purple vetch, V. benghalensis Woollypod vetch, V. dasycarpa The classification into cross-inoculation groups is somewhat arbitrary, since rhizo- bial invasion can sometimes occur outside the accepted host groups and not every strain of the given species of Rhizobium invades all of the legumes in the group. There are a number of legume species outside the seven recognized groups. Each of these legumes is dependent on its own particular strain of Rhizobium for effective nodulation. The following species among these ungrouped legumes are important in California: Big trefoil, Lotus uliginosus Birdsfoot trefoil, prostrate, L. tenuis Birdsfoot trefoil, upright, L. corniculatus [8] a. H O N a ft: - o z < H Q w H U a w - o H - H O H 1 3 c«V 2 e 02 « C3^2 OT373T3T3T3 i^a" e a •a N^iCTtNiOOCO -•*£ i— i CO i— i CD >0 W5 ^ 't O « C "& CO .—1 rH ^ £ w g <» <=+_ las s NCOCOtDNCO>OtO .3 g'JJ u 0000>^(NO)00 b. ft. o 5 d 1 o bJD u d O o t-i C 13. hfi . O • d d .2 .2 "43 '4-5 +3 '2 1 K c *3> 1 -a 5 03 .5 "o8 lants with r ounty ounty a, County . . d Field Sta d Field Sta ►ldt County lants withoi 02 JOO g§§^ -%££ § aa a ^ OOogOOSo . co" oT oo t^" i-T oo . » M CO M IN N ^ 0> oWWWWWW o £tftftftftftf£ .2c?S o 21 i 1 M CO (N W (N h h 6 e cS c E 1 X! . bC 3 fl • O 09 a> . t-, bO o • d ■ .2 : "8 "E +-> • ■+-> 1 d '8 . c3 • • • • p W !§ -a . -1-3 • if If -tf p-3 8 -a ^2 : § § § ■ - • o o o £ d C3 e 1 02 2heck plants Sonoma Cou Hopland Fie Yolo County Humboldt C Humboldt C Humboldt C Jheck plants . (M~ rt^ CO -O i-T 00 . © M N CO ■* IC ■* o oWWWMWW o 5z ' p 5 a . £ . a i P. 3 p 5 £ a I .2 "o S 5 1 I I § a 5- fa "* ^ . >» 2 '« ^"f ^ass ¥ 4? SI TA I fit it ii 3$ 5? Fig. 1. Laboratory test of nine selected strains of Rhizobium trifolii on rose clover seedlings growing in nutrient solution under sterile conditions at controlled temperature. Test tubes at the two ends of the rack contain check plants without rhizobia: left, with added nitrogen; right, without nitrogen. Effectiveness of rhizobial strains As a further complication, each species of Rhizobium contains strains that differ in their ability to fix nitrogen in symbiosis with a given species or variety of legume. Also, a strain of rhizobia that fixes but little nitrogen with a certain legume may fix a large amount of nitrogen with an- other legume in the same cross-inoculation group. As an example, some strains of Rhizobium trifolii that nodulate both red clover and subterranean clover fix much more nitrogen with the red clover than with the subterranean clover. The reverse is true with some other strains of R. trifolii. Such variability in effectiveness of strains of rhizobia emphasizes the necessity for using a specific inoculant for each legume. Selection of strains of rhizobia Strains of rhizobia for use as inoculants are isolated from nodules on thrifty legumes in the field. Preferably, these strains should be selected from the general geographic region where the inoculants are to be used. Strains so selected are then compared with each other on seedlings grown aseptically under controlled conditions (fig. 1 and table 1). The best strains are tested further in the field. Resident rhizobia in a soil Root-nodule bacteria occur in the soil wherever legumes normally grow, but these resident rhizobia may not be effective nitrogen-fixers with a newly planted leg- ume. An abundant soil population of in- vasive resident rhizobia will nodulate young legume roots as soon as they start to grow. If the legume seed carries only a few rhizobia of the desirable strain these few are not able to compete with the large numbers of resident soil rhizobia, and in- effectively nodulated plants result. Com- petition by resident rhizobia appears to be common on California rangelands, because of the prevalence of native legumes and their associated rhizobia. The following authors have reported on competition be- tween strains of rhizobia: Vincent and Waters (1953); Harris (1954); Jenkins, Vincent, and Waters (1954); Purchase and Nutman (1957); Means, Johnson, and Erdman (1961); and Johnson and Means (1964). The following authors have re- ported on competition between rhizobia and other soil microorganisms: Hcly, Iter- [10 Table 2 EFFECTS OF INOCULATION ON SUBTERRANEAN CLOVER* Nodulation of 50 plants from each treatment! Yield! Inoculant bacteria On crown Elsewhere Nodules/ root system Nitrogen content§ Foliage wgt., oven dry no. /seed 9 X 10 2 no. of plants 43 no. of plants 50 50 7 av. no. 56 43 12 per cent 1.9 2.0 3.7 gm/plot || 3.3 4.2 21 X 10 3 27.3 * Before inoculation all the seeds were surface-sterilized with ethanol and hydrogen peroxide, drained, washed three times in sterile water, and dried under aseptic conditions. Plants were grown 120 days from seed (winter of 1963-64) in field plots 3 feet square, separated by buffer areas. Sutherlin soil series, Hopland Field Station, Mendo- cino County. t Ten plants selected at random from each of five replicate plots. t There were three 3-foot rows in a plot. Each treatment was replicated in five plots in a randomized block. The center row of each plot was harvested for the yield data. § Analysis of trifoliate leaves only. || Difference for testing significance at the 1 per cent level = 19.9 grams. Fig. 2. Subterranean clover plants 69 days after seeding. Left, no inoculant; plant ineffectively nodulated by resident rhizobia. Right, heavy application of a specific inoculant; plant effectively nodulated near crown. Sutherlin soil, Hopland Field Station, Mendocino County. vi