Organic Matter in Soils

SOIL ORGANIC matter is a product of its environment.  Climate more than anything else determines the amount present. Climate, temperature, and rainfall together affect the rate of growth of the vegetation and consequently the amount of residues entering the soil. They affect also the activities of the micro-organisms that utilize the vegetation in the soil, and consequently the rate at which decomposition of the residues proceeds.

Some years ago H. Jenny, working at the Missouri Agricultural Experiment Station, showed that there is a clear relationship between mean annual temperature and the amount of organic nitrogen in the grassland soils of the Mississippi Valley or of the Great Plains.  The lower: the mean annual temperature, the higher the nitrogen and organic matter content of the soil and the wider the ratio between nitrogen and carbon. Similarly, when soils of similar origin along an annual isotherm are examined, the organic matter and total nitrogen content are found to rise with increasing rainfall and humidity.

R. H. Fuller and L. C. Wheeting more recently examined the factor of rainfall in certain prairie soils in western Washington that are relatively uniform in all climatic features except precipitation. Samples were taken from areas on which the annual rainfall varied from 16 to 120 inches. While there was generally a direct relationship between precipitation and organic matter content, there was also clear indication of a narrowing of the carbon-nitrogen ratio with a decrease in mean annual rainfall. In other words, the nitrogen content of the organic matter is low under high rainfall and increases with less rain.

Because of the operation of climate on the renewal and decomposition of organic matter, there is for every soil a stable equilibrium value under virgin conditions that becomes unstable as soon as the land is brought under cultivation.

Ordinarily, cultivation lowers the content of organic matter by speeding up microbial processes and reducing the amount of residues of vegetation that enter the soil. Rarely is the new level higher than the old.  For each land use the new equilibrium value will be slowly attained, but, by his choice of rotations and management practices, a farmer can influence the new level at which a balance is reached between the annual additions from the vegetation on the land and that used by the soil micro-organisms. That equilibrium is generally much lower than that attained under virgin conditions. The decline is not abrupt. It is not necessarily a matter for great concern. It must be accepted as inevitably accompanying the use of the land.

There are indications that in the Northern States at least the establishment of a new equilibrium level may take 70 to 100 years; it is probably true, therefore, that much of our farm land is only now approaching stability in content of organic matter. East of the Alleghenies, where the soils have been longer cultivated, and in the South, where both temperature and humidity are such that the rate of microbial utilization of the organic matter is higher, it is likely that the new equilibrium has already been attained.

Information on the effects of cropping practices on the changing level of organic matter in the soil is slowly accumulating. In Missouri, for instance, Dr. Jenny found a reduction in organic matter and nitrogen content of about one-third in 60 years of cultivation. In Pennsylvania, J.W. White and others compared the organic matter in plots that had been in a 4-year rotation of corn, oats, wheat, and mixed hay for 72 years with the organic matter in adjacent plots that had been continuously in grass for 72 years. The whole area previously had been farmed for at least 30 years. The differential that developed in the 72 years was substantial. The unfertilized grassland was found to have a content of organic matter 61 percent above that of the unfertilized cropped plots.  Applications of complete commercial fertilizer or manure and lime to the grass caused only a slightly higher equilibrium level to be reached.  Similar treatments applied to the cropped plots, however, did reduce the differential to 41 and 23 percent, respectively. The carbon-nitrogen ratios of the cultivated plots and the grass plots were quite similar; where any difference occurred those of the cultivated plots were wider.

The cropping system affects not only the final equilibrium value but also the rate of decline to this value. Under conditions at Manhattan, Kans., a 16-year rotation (including alfalfa for 4 years), a 3-year rotation (corn, soybeans, wheat), and continuous wheat were accompanied by similar losses of nitrogen and carbon. The average annual losses of carbon from the soil under those rotations were respectively 0.59, 0.68, and 0.56 percent. Continuous alfalfa, on the other hand, increased the supply of soil nitrogen and organic matter at the rate of 0.71 and 0.43 percent per year, respectively. The depletion during the corn years in the rotations was at a rate two to three times as great as under continuous wheat. In a 30-year period the loss in organic matter on Webster and Clarion soils in Iowa was greater under continuous corn or a corn-oats rotation than in longer rotations. The decreases in organic matter and nitrogen where manure had been applied consistently were much less, even than where all crop residues were returned.

Under more arid conditions microbial activities may sometimes be impeded. Comparisons were made of the changes in organic matter within a 22-year period at several places in Kansas. The losses were relatively low (about 0.5 percent a year), under systems of continuous grain, or alternate grain and fallow, but were appreciably larger when row crops were grown, or row crops and fallow were alternated.

Again the experience was that manure offset in part the depletion of soil carbon and nitrogen. Twelve tons an acre every third year cut losses by 50 percent or more, without concurrently increasing crop yields.

Irrigation adds another factor, in that the moisture status of the soil is completely changed. This might be expected to speed up microbial processes. Studies in Utah over 20 years revealed great differences in the changes in organic matter in response to cropping systems. Where alfalfa or beets were grown continuously, the latter with a manure application equal to 30 tons an acre annually, organic matter increased about 0.7 ton an acre yearly. At the other extreme, under continuous fallow or continuous oats, the annual loss was 1.6-1.7 tons of organic matter, which was halved in an alternate oats-fallow system. Sugar beets without manure caused an annual loss of 0.5 ton, but the application of 10 tons of manure annually reduced the loss only to 0.2 ton a year.

The progress made in the elucidation of the chemical nature of the organic matter of the soil is distressingly slow.  [In my opinion, that's because the productive power of organic matter comes from its influence on microbial life, both quantitative and qualitative. The actual elemental composition of organic matter is far less variant than its capacity to hold moisture, affect porosity and enhance the essential microbial balance. -ASC]

It is now generally accepted that the organic matter of the soil has its origin in the vegetation on the land. It consists only partly of residues of plant constituents that are less available or ‘are modified by the soil microflora. A part originates in cell-substance synthesized by the micro-organisms, or residues therefrom. Indeed, this microbially derived fraction forms a substantial portion of the soil organic matter. Attempts have been made to distinguish between decomposing plant residues, the chemistry of which can easily be related to that of the parent plant material, and fully decomposed residues, or humus, the chemistry of which is obscure.

Studies of the changes undergone by plant materials in decomposition have indicated that lignin, a constituent of the mature cell walls of plants, is relatively resistant to attack, and tends to accumulate, although somewhat modified by loss of methoxyl groups. There is, therefore, a lignin-derived fraction of soil organic matter. Certain of the reactions of soil organic matter, such as its behavior with chlorine, are in accordance with this view.

The changes that the lignin undergoes in decomposition are such that some of the characteristic chemical reactions that normal lignin undergoes do not take place. S. Gottlieb and S. B. Hendricks at Beltsville have shown that these changes must be substantial. They subjected a muck soil and soil extracts that presumably contained a lignin-derived fraction to certain rather vigorous chemical treatments that when applied to native plant lignin gave easily recognizable products.  It is a well-established procedure in organic chemistry to attempt to break up an unknown compound into smaller pieces that are simpler and can be identified. They did not obtain from the muck soil or the extracts the products that they expected to find had lignin been present, and they were forced to conclude therefore that “the material derived from plant lignin in the soil is drastically altered in the kind and position of the peripheral groupings on the aromatic rings.”

There may be interactions between the inorganic and organic components of soil that result in modification of the properties of both. Various European and Russian soil scientists believe that extremely stable inorganic-organic complexes are formed, particularly in chernozem soils, and that it is combination with the inorganic colloid that confers on the organic fraction its peculiar stability. The extreme in this viewpoint is probably represented by F. Yu Gel’tser, who defines as humus only that fraction of the organic colloid that is capable of forming a stable complex with inorganic colloids. The stable organic-inorganic complexes are said to play an important role in determining soil structure. Furthermore, the organic fraction concerned is not believed to be of plant origin, but to be formed by and from the micro-organisms accomplishing decomposition of the plant residues.

German workers have attempted to distinguish both chemically and functionally between “nutrient humus” (Nährhumus) and “stable humus” (Dauerhumus). Although the former is held to be the substrate upon which the soil microflora develops, and from which some plant nutrients are freed, the latter is regarded as being agriculturally the most vital fraction of the soil organic matter because of its influence on the physiochemical properties of the soil. Paradoxically enough, accumulation of “stable humus” is held to occur primarily under conditions of high biological activity and in the presence of calcium and montmorillonitic clays. Chemically the distinction is based on resistance to treatment with acetyl bromide or a mixture of acetic anhydride, acetic acid, and sulphuric acid.

These various theories indicate that there is much yet to be done to clarify the problems of the nature of soil organic matter. Their solution should permit the scientific management of soil organic matter to achieve the maintenance of fertility and conservation of soil.

Perhaps because the nature of soil organic matter is incompletely understood, exaggerated claims are sometimes made of its role. The use of composts derived from plant materials, with or without the addition of special amendments, with properties that verge on the magical, has been strongly advocated as a means of restoring the depleted supplies of organic matter in the soil.

It has even been asserted that crops grown in soil in which composts have been incorporated are more nutritious and less susceptible to disease. Such advocacy is usually coupled with an attack on commercial fertilizers, which are said to have been responsible for soil depletion, erosion, deficiencies of minor. elements or vitamins, and most other agricultural ills. No sound scientific evidence is marshalled in support of such claims. Largely they amount to a criticism of the practice of supplying a portion of the nutrient needs of the crop by application of mineral—that is, commercial—fertilizers that do not at the same time have a beneficial effect on the physical condition of the soil. It is reasonable to inquire, however, whether there is evidence of the presence in composts or decomposing plant materials of any substances that directly affect the growth of plants other than by meeting the nutrient needs or providing a beneficient physico-chemical environment.

Traces of auxins (substances that affect the rate of growth of some plant tissues) have been found in soil. These probably come from applications of barnyard manure, originating apparently in the urine. Thiamine also has been found in both manure and soil, but it has not been established whether it is present as free thiamine or is in the microbial soil population. Plant-growth regulating substances have been found in animal products such as dried blood, bonemeal, or meat. Some persons have strongly advocated that animal products or animal residues always-be added to composts, but there is no good evidence that they have value beyond the nitrogen and other nutrients that they contain.

In any event growth substances of the class of the auxins or heteroauxins have not been shown to exert any over-all growth stimulation of plants or to increase their yield or fruitfulness. They may, however, stimulate root growth of cuttings.

It cannot yet be said, then, that there is good evidence that composts or decomposed residues are characteristically endowed with substances that cause exceptionally good crop growth or quality.

It has often been noted that the incorporation of crop residues into soil is accompanied by an improvement in soil structure and an increase in the number and size of the stable soil aggregates. The duration of such improvement may be quite limited. Sod crops appear to be especially valuable in causing the development of an improved physical condition. The grass-derived prairie and chernozem soils developed a notably granular structure that for a period remains stable under cultivation, but that deteriorates slowly in intertilled crops.

There are probably several reasons why clay particles may be caused to stick together to form small aggregates that exhibit some stability when wet or when in water, but it is now certain that directly or indirectly one form of aggregation results from the activities of the soil organisms. To obtain this effect it is necessary that decomposible material be added to the soil. Crop residues affect soil aggregation only as they are decomposed. Soluble energy sources, such as glucose or aqueous plant extracts, have no direct effect on aggregation, but when they decompose they cause as marked an improvement in soil structure as may follow the incorporation of whole plant materials.

The implication of this fact is that it is not any residual fraction of the plant material that is responsible for the effects observed. It is now presumed that the micro-organic cell substance itself or products derived therefrom are largely responsible. The presence of ramifying fungal threads may temporarily serve to bind particles together, but such a mechanical effect might be expected to be of relatively short duration.  The rate of formation of stable aggregates has been shown to be highest during the period when microbial activity is greatest, shortly after incorporation of food material. The effect of the organisms is due to products formed by and from them. Many soil bacteria, and particularly some of the aerobic cellulose-decomposers, produce gummy substances, which, it is suggested, may have a cementing action. Such substances would not necessarily be unavailable to other organisms; hence their effect might also be of short duration. There is, of course, the possibility that such gummy products might form some complex with the inorganic colloid that would be more stable.

An alternative opinion is that the cementing substances are the breakdown products of bacteria that develop on and utilize the fungal mycelium that usually appears soon after plant residues are added to soil.

A grass sod has long been recognized as particularly effective in causing the formation of a granular soil structure. This is claimed to be due to the activity of bacteria in the rhizosphere (the zone immediately surrounding the roots and rootlets), utilizing root excretions or sloughed off cellular material. Species of Pseudomonas have been said to be particularly numerous, and products of break-down of certain of these species have been shown to exercise strong cementing action. The activity of rhizosphere bacteria on the fibrous ramifying root system of grasses is, accordingly, held to be responsible for the steady increase in soil aggregation that takes place when land is put in sod. Through the constant death and renewal of rootlets, such a crop also supplies energy material other than in the form of root excretions. Upon subsequent plowing there may be additional aggregation at the expense of the fungal tissues that then develop on the grass residues. The great difference between grass and other crops in extent of aggregation caused is, therefore, accounted for by the much denser and more extensive root system of the former and the fact that under sod excessively aerobic conditions do not prevail.

Other studies have been made of the microbial population of the soil near the roots and the rhizosphere, with somewhat different objectives. The flora of the rhizosphere is much larger numerically, and more active physiologically, than that of the adjacent soil in the same horizon.  The presence of the rhizosphere flora must surely affect the development of the plant; conversely, the nature of the plant influences the character of the microflora. Especially may this be true in the case of perennial plants. The incorporation of organic residues may produce substantial changes in the micropopulation of the general soil mass, but may have little effect on the microflora associated with crop roots.

Studies of the nutrient requirements of organisms isolated from the root region have shown that more types dependent on growth factors or amino-acid nitrogen are to be found therein than in soil away from roots. The view that microbial activity in the rhizosphere is maintained chiefly on sloughed off root cells is inadequate. Direct evidence has been obtained of the excretion from roots of soluble substances that are utilized in the rhizosphere and are responsible for the activity of the flora therein.  This may be a mechanism that affects the attack on plants by root pathogens. Resistant and susceptible varieties of a crop may differ in the nature of the root excretions. M. I. Timonin in Canada has studied particularly wilt-susceptible and wilt-resistant varieties of flax in this connection, and has adduced evidence of the excretion of minute quantities of hydrocyanic acid by roots of the latter.

Virgin genetic soil types have characteristic and distinctive microbial populations that reflect the influence of the factors involved in their development. When these soils are brought under cultivation, modifications may be impressed upon the flora. The changes that result are sometimes great if the environment is much altered by the new land use, as, for example, when poorly drained peat soils are drained, or forest podzols cleared and plowed to a depth that mixes the shallow upper horizons. The responses that follow different management practices subsequently applied may be much smaller and more difficult to detect. As a measure of these, little reliance is placed upon bacterial and fungal counts because it is appreciated that only a small fraction of the population is represented on agar plates, and, moreover, that the slow-growing forms, which probably constitute the basal autochthonous (native) flora, will not be included.

However, quantitative procedures have been much improved by the studies of N. James and M. L. Sutherland, of Winnipeg, Manitoba, who have examined statistically the reproducibility of the results and have made recommendations for modifications that increase the accuracy of plate counts. In field studies in Manitoba they demonstrated an apparent relationship between numbers of bacteria and soil moisture content. Evidence pointing to a similar conclusion had previously been obtained in counts of organisms in grassland soil, and, inasmuch as the moisture films between soil particles are the actual loci of the active bacteria, it is not surprising that such a relationship should be found.

Each horizon in the soil profile develops a characteristic microflora.  Organisms from one horizon introduced into another do not necessarily establish themselves in the new environment. Each population may be likened to a team of compatible organisms, the activities of which fit together so precisely that the available energy is utilized most efficiently.  A certain measure of population stability comes from the presence of inhibitory or antibiotic substances produced by some of the established microbial inhabitants. Introduced organisms would be likely to maintain themselves only if they accomplish as well or better some step in the sequence of transformations by which energy material is utilized and are unaffected by any inhibitory substances that may be present. This has implications when serious erosion has occurred since the newly exposed surface soil does not have, and may not readily acquire, a micropopulation similar to that of the uneroded surface soil.

The nitrogen cycle in soil, although well known in all its general aspects, still is incomplete in many details. It is predominantly, if not entirely, a biological cycle, but the populations or organisms responsible for the various steps differ greatly in degrees of specialization. The application of physiochemical methods to the study of the biochemistry of some of the chief transformations has been quite productive. The availability of the stable isotope of nitrogen, N¹⁵, is now permitting a greater refinement of such studies. The latter has so far been applied mainly to questions relating to nitrogen fixation by Rhizobium (bacteria which cause nodules to form on the roots of legumes and which fix nitrogen only when inside the nodules) and Azotobacter (bacteria which are free living in the soil and independent of plants). The technique will be of equal usefulness in attacking problems relating to immobilization or release of nitrogen in decomposition. Absolute values can now be obtained for recovery of nitrogen by a crop from some previous crop residues incorporated in the soil.

Study of the behavior of a large number of plant residues in soil in Australia has led to the conclusion that it is highly improbable that any material with a total nitrogen content of less than 1.5 percent will give a positive return of nitrogen in one season, and that only when the nitrogen exceeds 2.5 percent is a large early release secured coincident with the demand of a crop planted shortly after incorporation. In terms of the carbon-nitrogen ratios, these limits may be expressed approximately as 27—1 and 16-1, respectively.

Nitrate is the form of nitrogen ordinarily used by plants. The decomposition processes in soil result in the liberation of nitrogen as ammonia, and therefore the final step of conversion of this ammonia to nitrate, which is known as nitrification, is an important process. This change has been little studied in the last decade except in India where a number of purely chemical reactions said to be able to bring this change about have been under investigation. The activity of microbial nitrifying systems has, however, not been disproved, and there is no good reason for abandoning the long-established theories about the part played by this group of bacteria in soil. It must be admitted that the conditions under which the appearance of nitrate in soils occurs cannot be entirely reconciled with those under which the classical nitrifying bacteria carry out the oxidation of ammonia in pure culture in the laboratory.

Nitrogen has recently been supplied to irrigated crops by dissolving ammonia gas in the irrigation water. The subsequent oxidation of the ammonia to nitrate has been found to proceed only to the nitrite stage in certain alkaline desert soils in Arizona. The ammonia in solution is not toxic to the bacteria that should complete the oxidation even at a concentration as high as 300 parts per million; but if the soil is more alkaline than pH-7.7 the nitrite is not transformed to nitrate, and the crop cannot benefit properly from this unusual nitrogen fertilization.

The subject of nitrogen fixation has continued to attract many investigations with diverse interests and objectives. Some of the most refined techniques that have been devised for studying the physiology of bacteria have been applied to this problem, which is a challenging one because nitrogen-fixing organisms can accomplish readily what the chemist can accomplish only under extremely high pressures and temperatures.  It is easier to work with Azotobacter than with the rhizobia, which fix nitrogen only when this bacteria is in the nodules of a living leguminous plant. The first step in the process is believed to be the combining of nitrogen from the atmosphere with some component of an enzyme system, called the azotase system, in the organism. Very fundamental investigations on the characteristics of this system have resulted in a unification of the subject of nitrogen fixation that did not appear probable a few years ago. They make it highly likely that the biochemical mechanism of fixation is identical in Azotobacter and in nodulated legumes.

Attempts have also been made to find out the nature of the chemical steps involved in the fixation process and a plausible theory, partly supported by experimental evidence, has been proposed by A. I. Virtanen of Finland. He identified some organic compounds which under certain circumstances pass out or are excreted from the roots of leguminous plants into the soil or sand in which they are growing. As a result he suggested that hydroxylamine was first formed and that this then combined with oxalacetic acid produced by the plant. This next is converted to aspartic acid, an amino acid that could be used in protein building. Others have maintained that ammonia is a key intermediate.  Experiments by P. W. Wilson and others at Madison, Wis., in which Azotobacter was grown in the presence of various nitrogen compounds in a nitrogen atmosphere enriched with the heavy isotope, N¹⁵, showed that ammonia or compounds readily converted to ammonia are used by the bacteria to the exclusion of the nitrogen of the atmosphere.

Further studies of the distribution of species of Azotobacter in soils have indicated that this organism is found in all parts of the world but that its local distribution is erratic and limited primarily by the pH of the soil. Azotobacter rarely are found in soils, the pH of which is <6.0.  The acid-tolerant species A. indicum, originally isolated from certain Indian soils, does not occur generally in acid soils. That an environment of pH 6.0 is apparently limiting in distribution is strong presumptive evidence for the view that Azotobacter in soil are dependent on the fixation process, because this same pH value has been proved to be the limit below which fixation does not occur. When supplied with combined nitrogen, these organisms can develop under considerably more acid conditions.

Excretion of nitrogenous compounds from the nodules of leguminous plants, if it is of general occurrence, would be important in farming practice. It has many times been demonstrated that nonlegumes growing together with legumes in such a manner that their root systems are intertwined seem to derive benefit from the association. The growth of the nonlegume is increased and sometimes also its protein content is higher, just as would be the case if additional nitrogen had been made available to it. This effect of associated growth may not be solely or even usually due to excretion of nitrogen from the nodules because the latter has not been proved to be a general occurrence. A.I. Virtanen in Finland, under the environmental conditions there, obtained evidence that led him to think the phenomenon a common one, but workers in places as diverse as Scotland, Australia, Washington, D. C., Wisconsin, and Kentucky have only in a few instances obtained any evidence of excretion. In these instances the amounts concerned have invariably been small, and the experiments have not been reproducible.

The explanation of the beneficial effects of associated growth on the nonlegume may lie instead in loss of nodules from the legume, a process that is apparently normal but which is accentuated by clipping the tops of the legume as would occur in grazing or mowing of pastures. Nodule tissue is exceptionally high in nitrogen (about 6-10 percent), and would decompose rapidly with the liberation of available inorganic nitrogen that might be used by the nonlegume.

Strains of the various species of Rhizobium that differ in effectiveness have long been recognized. Some are almost wholly ineffective in benefiting the host legume. One of the features of “ineffective” strains now established is that the nodules they produce are smaller and, therefore, contain less active bacterial tissue. Another is that they remain on the roots for a shorter period before deterioration and degeneration occur. They may be no less efficient in fixation per unit mass of nodular tissue, but the shorter size and shorter duration of attachment together account for the smaller contribution made to the nitrogen economy of the legume.

In choosing commercial cultures for inoculation of leguminous crops it is common practice now to select and mix several effective strains isolated from widely planted varieties of the crops in question. There is a tendency, however, to imply that effectiveness or ineffectiveness is wholly a property of the organism. In the efficient operation of the fixation process the plant has its part, and there is some evidence to indicate that ability to be nodulated and to participate in this unique relationship may also be a plant genetic character.

The concept of “cross inoculation” groups of plants that can be nodulated by one species of Rhizobium is not now interpreted as rigidly as at one time was the case. Many individual exceptions in which plants of one group have been nodulated by an organism isolated from a plant in another group have been observed at various times, but the practical convenience of the grouping in the preparation of cultures for use by farmers is not impaired. It may, however, be desirable to introduce additional requirements into the selection of strains of rhizobia for incorporation into commercial legume cultures if the newly recognized factor of competition between strains is to be taken into account. It has been shown that effectiveness in nitrogen fixation and dominance in competition between strains are independent. Where two strains are present together in the zone surrounding the root of the legume the strain having the higher initial growth rate may suppress the development of the weaker strain and, therefore, be responsible for almost all the nodules. The object, therefore, will be to select for commercial cultures strains that are effective in fixation in the desired varieties of the crop in question and that are also vigorous in competition with other strains capable of nodulating the same plant.

Commercial legume inoculants are now quite generally dependable.  The most common form consists of a heavy suspension of organisms absorbed by peat, packaged in such a manner that moisture losses subsequently are minimized. There is a possibility, however, that lyophilized cultures (dried under vacuum at low temperature), that maintain viability for long periods and that would be less bulky, may be developed to the state of practicability. Some changes in procedure followed in seed inoculation may be desirable when seed disinfectants are used. Certain of these compounds are not compatible with rhizobia; others do not reduce nodulation if a heavier rate of inoculation is practiced and if the seed is planted shortly after treatment. An alternative which has been found effective when planting treated pea seed in Washington is to use bulk inoculant which is placed in the furrow through the fertilizer attachment of the grain drill.

Claims have been made at various times by Russian workers that very substantial yield increases in a variety of crops have followed inoculation of the soil with Azotobacter species, and such inoculants have apparently been prepared and used in a large scale. Trials of a similar character in this country in greenhouses and on small plots, where adequately replicated, have yielded predominantly negative results, although in one or two isolated cases significant and otherwise unaccountable yield increases have been obtained.

THE AUTHOR
A. G. Norman was professor of soils at the Towa State College until the fall of 1946, when he began research work for the Chemical Corps of the Army. He has interested himself particularly in studies of the decomposition of plant materials, and the chemistry of soil organic matter. Before 1937 he was biochemist at the Rothamsted Experimental Station, England.