.1’ \‘.V .n. “mm ‘l . . n. - !“J‘ . THE-33:31:. | ,(‘q '. v 64'" V) 1 | , u" ' H! I L.'. h“. ‘ .L' ‘7' . . ~ , ’5 ' ‘ ‘ ‘ MN 7 . (Ag-5V \ 'l' ‘ ff; “ ‘ . ‘ ' mu J" ‘K J ‘ . | u I. n. 4. I [’5 1‘ n ~ ma IIIIIIIIIIZIIQIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 0396 3066 This is to certify that the thesis entitled SHORT-TERM MEASUREMENT OF SOIL DENITRIFICATION USING C2H2 ‘INHIBITION:- RESPOSNSE TO ANAEROBIOSIS AND THE EFFECT OF THE RHIZOSPHERE presented by MORGAN SCOTT SMITH has been accepted towards fulfillment of the requirements for Ph.D. degree in Soil Science flm 4r Major pqofessor Date 0-7639 O V q. Q - SHORT-TERM MEASUREMENT OF SOIL DENITRIFICATION USING CZHZ INHIBITION: RESPONSE TO ANAEROBIOSIS AND THE EFFECT OF THE RHIZOSPHERE BY Morgan Scott Smith A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Crop and Soil Sciences ABSTRACT SHORT-TERM MEASUREMENT OF SOIL DENITRIFICATION USING C2H2 INHIBITION: RESPONSE TO ANAEROBIOSIS AND THE EFFECT OF THE RHIZOSPHERE by MORGAN SCOTT SMITH Methodological limitations make it difficult or impossible to measure rates of soil denitrification or ratios of the two denitri- fication products, N20 and N2, under natural conditions. Practical methods of reducing denitrification loss in agricultural practice are not available. Acetylene inhibits the reduction of N20 to N2. This eliminates the methodological problem of high atmospheric N2 concentrations masking the denitrification products and permits measurement of low denitrification rates during a short-term analysis. The concentra- tion of CZHZ required for effective inhibition was shown to increase as soil N03- concentrations decrease. When low concentrations (approximately 0.1 pg/g soil) of NO3-, as carrier-free 13NO3-, were added to anaerobic soil slurries 0.1 atm C2H2 was required for inhi- bition. Denitrification rates of anaerobic slurries determined by the CZHZ inhibition method were compared to rates determined by a method 13 using the short-lived radioactive isotope of N, N. There were no consistent differences between the results with the two methods. Furthermore, direct 13N measurements of the ratio NZO/(NZO + N2) agreed with indirect measurements using C H2, that is, N 0 production 2 divided by rates with C 2 rate by soils without CZH It was con- 2H2' cluded that C2H2 inhibition is a valid, sensitive, and convenient 2 method of measuring denitrification rate. However, the dependence on N03 concentration and the slow diffusion of inhibitor and product gases requires that the method be cautiously used in poorly defined conditions or undisturbed soils. After the imposition of anaerobiosis on soil slurries, two dis- tinct phases of denitrification rate were observed. Phase I denitri- fication rate was always linear, was not inhibited by chlorampheni- col, was increased slightly or not at all by carbon amendments, and lasted for 1 to 3 hours after the onset of anaerobiosis. Phase I was attributed to the activity of pre-existing denitrifying enzymes in the soil. Results of phase I assays indicated that denitrifying en- zymes are present even in well aerated soils. Following phase I, enzyme synthesis was derepressed and denitrification rate increased. Chloramphenicol inhibited this increase. In soils without carbon amendment a second linear phase, phase II, was attained after 4 to 8 hours of anaerobic incubation. The linearity of this phase was attributed to full derepression of denitrifying activity by the indigenous population and to lack of significant growth of denitri- fiers. Carbon amendment eliminated or abbreviated the linearity of this phase and the rate continued to increase, apparently due to growth. Phase I but not phase II was increased by decreased aeration state of the soil in situ. Therefore, phase I may be more directly related to natural denitrification rates. The effect of roots on denitrification was studied with C2H2 inhibition methods. Anaerobic assays of soil slurries indicated that a greater supply of carbon increases the potential for denitrifica- tion in the rhizosphere. This was observed with greenhouse and fresh field soils. A split-plate experiment suggested that denitrifying activity decreases rapidly in the first few mm away from the root. A specific enrichment of denitrifiers relative to aerobes was observed in planted soils. Soils with and without intact plants were also assayed for denitrification rate. These soils were water-saturated and in an aerobic atmosphere, thus approximating natural denitrifying conditions. When soil NO - concentrations were high the results 3 conformed to the prevailing view that denitrification is enhanced in 3 rate was significantly lower in planted soils than in unplanted. the rhizosphere. Yet, at low NO concentrations the denitrification This is believed to be the result of competition for NO3-between plant uptake and denitrifiers. High N03- concentrations caused a significant increase in NZO/(NZO + N2) in these soils. To Susan and Hannah ii ACKNOWLEDGEMENTS The greatest contribution to my education at Michigan State has come from the knowledge, experience, and friendly cooperation of my fellow graduate students. I have been quite fortunate to work closely with such outstanding people as Mary Firestone, Mike Betlach, and Bill Caskey. Mary Firestone should be specifically mentioned for initiating our investigation of the CZHZ inhibition method and for directly participating in much of the research described in Chapter 2. I am indebted to Dr. James Tiedje, chairman of my guidance com- mittee, for establishing an excellent environment in which to conduct research and for allowing me to carry out this research independently. I thank the remaining members of my guidance committee; Drs. M. J. Klug, A. R. Wolcott, and G. A. Safir for their efforts. This work was supported in part by USDA Regional Research Project NE-39, and by a grant from the National Science Foundation. iii LIST OF TABLES TABLE OF CONTENTS LIST OF FIGURES . . . . . . . . . . . . . CHAPTER I. CHAPTER II. CHAPTER III. CHAPTER IV. APPENDIX A. APPENDIX B. INTRODUCTION: DIRECTING SOIL MICROBIOLOGY RESEARCH TOWARD MINIMIZING DENITRIFICATION Strategies which sacrifice yield . . Specific chemical inhibition of denitrification. . . . . . . . . Scheduling application of nitrogen to favor plant, rather than denitrifier, utilization of nitrate . . . . . . Scheduling irrigation to minimize denitrification. . . . . . . . . . Research objectives . . . . . . . LITERATURE CITED . . . . . . . . THE ACETYLENE INHIBITION METHOD FOR SHORT-TERM MEASUREMENT OF SOIL DENITRIFICATION AND ITS EVALUATION USING 13N. . . . . . MATERIALS AND METHODS . . . . . . . . RESULTS . . . . . . . . . . . . . DISCUSSION . . . . . . . . . . . . LITERATURE CITED . . . . . . . PHASES OF DENITRIFICATION FOLLOWING OXYGEN DEPLETION IN SOIL. . . . . . . . . . MATERIALS AND METHODS . . . . . . . . RESULTS 0 O O O O O O O O O O O 0 DISCUSSION . . . . . . . . . . . . LITERATURE CITED . . . . . . . . . . DENITRIFICATION IN THE RHIZOSPHERE . . . MATERIALS AND METHODS . . . . . . . . RESULTS . . . . . . . . . . . . . DISCUSSION . . . . . . . . . . . . LITERATURE CITED . . . . . . . . . . ENUMERATION OF SOIL DENITRIFIERS. . . . . A RIFAMPICIN RESISTANCE MARKER TO STUDY SOIL DENITRIFIERS. . . . . . . . . . iv Page vii 10 14 16 17 21 33 36 37 38 40 S6 62 64 65 72 9O 96 98 107 Table 10 11 12 13 14 15 16 LIST OF TABLES Characteristics of soils used. . . . . . . . Reduction of nitrous oxide by soil in the presence of various acetylene concentrations. . . . . . Recovery of added nitrate as nitrous oxide in the presence of various acetylene concentrations. . . Effect of various concentrations of acetylene on the ratio of denitrification products as determined by 13N 0 O I I O O O O O O O O O 0 Comparison of denitrification rates and ratios of products determined by 13N and C2H2 inhibition methods 0 O O C Q 0 O I C O O O O O I Effect of glucose additions on denitrification rate of Brookston soil . . . . . . . . . . Phase I denitrification rate of Spinks loamy sand preincubated aerobically at varying water contents. Generalized description of the phases of soil denitrification after the imposition of anaerobic conditions . . . . . . . . . . . . . . Portion of N gas evolved as N20 by anaerobic Brookston soil with and without plants. . . . . Denitrification rate of Miami soil sampled at various distances from a corn row. The assays were anaerobic. . . . . . . . . . . . . Number of denitrifiers in planted and unplanted 80118. O C O C C O O O O C O O O O O Denitrification rate of planted and unplanted soil assayed in aerobic gasbags. . . . . . . . . Portion of N gas evolved as N20 by soils in aerobic gasbags . . . . . . . . . . . . . . . Effect of N03” concentration on MPN of denitrifiers Comparison of MPNs obtained in low NO3-, low NO -, - 2 and high N02 0 O O O O I O O O I O O 0 Effect of initial aeration state of Hungate tubes on MPNs of denitrifiers. . . . . . . . . Page 18 22 24 25 26 46 53 57 76 80 84 88 89 102 103 104 Table 17 18 19 Comparison of two tests for the presence of denitrifiers in MPN tubes. Production of N20 by bacteria not commonly believed to be denitrifiers . Summary of results with rifampicin resistant denitrifiers . vi 0 Page 105 106 109 Figure 10 11 LIST OF FIGURES Nitrous oxide production by 4 CZH inhibited soils. Points are means of 3 rep icates. . . . Comparison of denitrification by the Miami soil under different experimental conditions. All soils were treated with CZHZ' Points are means of 3 replicates. . . . . . . . . . . . Two linear phases of denitrification in 3 C2H2 inhibited soils. Points are the means of 3 replicates . Effect of Chloramphenicol on denitrification rates in a soil and a sand. Points are the means of 3 rep- licates. Numbers in parentheses are rates (nmoles NZO-g dry soil 1omin ) with 95% confidence limits . . Derepression of denitrifying enzyme synthesis by pure cultures of 5 strains of denitrifiers. Denitrifiers were grown either aerobically or in one case anaero- bically, harvested, and then incubated anaerobically with NO - . . . . . . . . . . . . . . 3 Phase I and phase II denitrification by Miami soil sampled before and after irrigation. Points are the means of 3 replicates. Numbers in arentheses are rates (nmoles N20 g dry soil min ) with 95% con- fidence limits . . . . . . . . . . . Evolution of N20 by C2H2-inhibited Miami soil in an aerobic atmosphere. Water was added to half of the samples at the beginning of the incubation. There were 5 replicates per treatment. . . . . . . Split-plate apparatus used to determine spatial rela- tionship between roots and denitrifying activity. Note that roots are restricted to the outer chamber . . Saran gasbag used to measure denitrification rate of soils with and without intact plants . . . . . . . Denitrification by non-rhizosphere planted, unplanted, and rhizosphere Brookston soil assayed anaerobically. Soils were amended with N03 only after 5 hours of in- cubation. Points are means of 3 replicates. . . . . Effect of glucose and succinate amendment on the phase II denitrification rate of planted and unplanted Brook- ston soil. Rates have been normalized relative to the rate of unamended, planted soil. There were 3 repli- cates per treatment. . . . . . . vii Page 29 32 42 45 49 52 55 68 71 74 78 Figure Page 12 Denitrifying activity related to distance from oat roots. Spinks soil was taken from split-plates. Rates were determined 200 to 400 min after the impo- sition of anaerobiosis. There were 3 replicates per treatment. . . . . . . . . . . . . . . . . 82 13 Time course of N20 production by unplanted high N03— Brookston soil with C2H2 in aerobic gasbags. Points are means of 5 replicates . . . . . . . . . . . 86 viii CHAPTER I INTRODUCTION: DIRECTING SOIL MICROBIOLOGY RESEARCH TOWARD MINIMIZING DENITRIFICATION LOSSES Biological Abstracts cites 112 publications in 1976-77 related to denitrification. This intense activity has succeeded in defining the important parameters for denitrification, for example: aeration state, supply of electron donor, and supply of electron acceptor, and it has begun to reveal the biochemistry of denitrification. Yet the rewards of this research have been limited. We have not approached a level of understanding which permits quantitative correlation of denitrification rates with natural conditions. Reliable measurements of the magnitude of denitrification loss and the ratio of the two denitrification products, N20 and N2, are not available. Nor has this research provided practical strategies for reducing nitrogen lost through denitrification. This large allocation of research resources is not unjustified. Enough information is available to indicate that denitrifiers claim a significant fraction of the N applied to agricultural soils. Methodo- logical limitations (to be discussed later) and inherent soil varia- bility have resulted in a large range of values reported for denitri- fication loss, usually between 15% of applied fertilizer (Allison, 1955) and 70% (Rolston gt al., 1976). Hauck has recently reviewed much of this literature (personal communication) and tentatively concluded that the best estimate of average N loss from agricultural soils lies between 20 and 30%. There are reasons for the great interest in denitrification, 1 other than the efficient use of fertilizer. Of special concern has been the suggestion that N 0 from fertilizer catalyzes a significant 2 destruction of atmospheric ozone (Johnston, 1972; CAST, 1976) possibly resulting in serious environmental perturbations. It is my opinion that there are no valid estimates of the percentage of denitrification gases released as N 0 under natural conditions, therefore, N20 flux 2 can only be grossly approximated. Even given a greater understanding of atmospheric chemistry than currently exists (Crutzen, 1976) the magnitude of this hazard cannot presently be determined. Several recent reviews have examined the denitrification liter- ature in detail. A comprehensive review here would add very little. Instead I will acknowledge pertinent research at the beginning of each chapter and where appropriate. Among the reviews which can be recom- mended are: Focht (1978) on methodology, Payne (1973) on biochemistry and microbiology, Focht and Verstraete (1977) on biochemical ecology, and a general review by Delwiche and Bryan (1976). The major objective of this introductory chapter is to relate the direction of my research and the type of denitrification research being conducted elsewhere to denitrification loss in agricultural practice. Many laboratory studies are, perhaps casually, justified by the possibility of increasing the efficiency of nitrogen use. The validity of this justification is rarely examined critically. Is it economically feasible to manipulate agronomic practice to reduce denitrification? What information, which can be obtained by soil microbiologists, is most likely to yield tangible rewards? I will suggest four general approaches to minimizing denitrification, discuss what is known and what needs to be known about soil microbiology for the development of these approaches, and speculate on their proba- bility of success. Any pretention of agricultural economics is denied. These dis- cussions are in no way intended to be complete cost analyses; innumer- able assumptions and simplifications have been made. I will simply attempt to make suggestive comparisons of relative costs. The produc- tion and cost figures have been obtained from personal communication with Dr. M. L. Vitosh and from M.S.U. extension bulletins E—lllO, E— 802, and E-857. Two major factors are ignored in the following discussions. First, it is assumed that the environmental cost of increased N20 production is not significant. Yet, I do not wish to imply that this is necessarily true. Second, it should be emphasized that agronomic practice will ultimately be determined only by controlled field studies under conditions which approximate those of farming. Labora- tory studies can merely define the important factors and suggest possible strategies. Strategies which sacrifice yield. This is the most general approach to be discussed and includes numerous possibilities. Among these are rotations with low value legumes such as alfalfa which can increase soil organic nitrogen but do not provide the cash return of, for example, continuous corn cropping. In irrigated systems, water input (and soil moisture) could be reduced to less than that necessary for optimum growth. A simple approach is to reduce the application of nitrogen ferti- lizer. Although there are undoubtedly cases where more nitrogen is applied than necessary, the alternatives considered here are to apply nitrogen to approximate maximum yield or most profitable return, or to deliberately sacrifice yield to reduce nitrogen loss. There is good evidence that denitrification rate is related to soil NO - concentra- 3 tions. Apparent first-order kinetics for denitrification rate and N03 concentration are observed in soils provided that the supply of electron donor is not limiting and N03 concentrations are below about 40 ppm N in solution (Starr and Parlange, 1975; Stanford gt_§l., 1975). To examine the probability of success with these strategies, compare some costs involved in the production of irrigated corn for grain. This farming system is now fairly common in Michigan; its use is increasing. It was chosen for this and the following discussions because it involves a medium value crop, a relatively high N input, and offers a number of opportunities for controlling N transformations. A yield of 180 bu/A and a corn price of 2.50 $/bu is assumed. The crop value is then 450$/A. A yield reduction of 5% would cost 22.50 S/A. Approximately 200 1b N/A is recommended for irrigated corn. If all of this were applied as anhydrous ammonia (0.12 $/1b), the cost of N fertilizer would be 24 $/A. More commonly about 2/3 is applied as anhydrous ammonia and 1/3 is applied as N solution (0.24 $/1b) in the irrigation water. In this case fertilizer would cost 32 $/A. It will be assumed that Hauck's estimate of denitrifi- cation loss is accurate; approximately 25% of the fertilizer applied is denitrified. The denitrification cost is then about 8 $/A. It is concluded that if denitrification could be totally elimi- nated, which seems very unlikely, even a small yield reduction would cost much more, 22.50$, than would be gained in fertilizer cost, 8 $. For this strategy to become marginally acceptable, N cost must at least triple relative to corn price. In the U.S. agricultural system, this does not seem likely in the near future. The cost of N is only a small part of the total cost of crop production. Fertilizer N, even if used inefficiently, provides a large economic return. Researchers anticipating the application of their denitrification studies to farm operation should be aware of this. The same reasoning applies to another topic currently under extremely active investigation, N 2 fixation. Specific chemical inhibition of denitrification. Chemicals which specifically inhibit nitrification have been dis- covered; these include carbon disulfide and nitrapyrin (N-Serve). N- Serve is still being evaluated, but it appears that it will have some practical application, at least under certain conditions. This sug- gests that comparable inhibitors of denitrification might exist. (Though it is not clear that direct inhibition of denitrification is preferable to the indirect approach of inhibiting nitrification and reducing the substrate supply for denitrification.) If such an inhibitor is to be found, it seems most likely to arise from basic research on denitrification. Some of the data al- ready available indicate that a general inhibitor is less likely for denitrification than for nitrification. The taxonomic diversity of denitrifiers is apparently greater than that of nitrifiers. Payne (1973) lists 15 genera of denitrifiers, whereas Nitrosomonas and Nitrobacter are considered to be primarily responsible for soil nitrification and only Nitrosomonas need be affected to inhibit nitri- fication. (Schmidt's (1978) recent fluorescent antibody studies suggest, however, that nitrifiers may be considerably more diverse than commonly believed.) More significantly, the biochemistry and control mechanisms of denitrification are apparently more diverse. It has been observed that some denitrifiers require NO3 and anaerobiosis to derepress aynthesis of denitrifying enzymes, but others do not 3 (W. J. Payne, personal communication). Pseudomonas require NO perfectomarinus synthesizes all denitrifying enzymes simultaneously (Payne E£.§l" 1971). However, Micrococcus denitrificans synthesizes first NO3 , then N02- reductase (Lam and Nicholas, 1969). N02- reductases of several organisms have been shown to be heme proteins, but not the copper containing NO2 reductase of Achromobacter cycloclastes (Iwasaki and Matsubara, 1972). On the other hand, some degree of biochemical unity, necessary for the function of a general inhibitor, might be expected. For example, a component with an absorption maxima at 573 nm related to the binding of nitric oxide has been observed in all denitrifiers examined (W. J. Payne, personal communication; Rowe gt al., 1977). In this laboratory it has been demonstrated that N O is a freely diffu- 2 sible intermediate for essentially all soil denitrifiers (Firestone gt 31,, 1977 and unpublished data), contrary to some earlier assertions that the denitrification pathway is variable (for example; Stefanson, 1972). If a specific denitrification inhibitor were available, its use in soils might actually have unfavorable results. It is probably not feasible to inhibit the first step in denitrification, NO3 reduction to N02-, because it is also carried out in plant and microbial assimi- lation of NO3-, presumably by similar mechanisms (see Payne, 1972, for 3 the first unique reactions of denitrification were inhibited, the a comparison of assimilatory and dissimilatory NO reduction). If accumulation of toxic NO or NOZ- could occur. If a workable denitrification inhibitor were available, it is reasonable to assume that its cost would be comparable to that of N- Serve, about 3 $/A. The application costs of N-Serve are insignifi- cant since it is suited for injection with anhydrous ammonia. This might not be true of a denitrification inhibitor and application cost might be an additional 1 to 2 $/A. Comparing this to the previously calculated denitrification cost of 8 $/A and assuming no change in yield, at least a 50% inhibition of denitrification would be necessary for even a marginal profit. It is apparent that any strategy for reducing denitrification must have a very small cost to be profitable. The remaining two strategies to be considered have minimal cost and, in fact, merely involve a refinement of current recommended farm practice. They are also more directly related to the research described in this thesis. Scheduling nitrogen application to favor plant, rather than denitrifier, utilization of N01-. It seems obvious that plants and denitrifiers, particularly 3 O my knowledge this viewpoint has not been explicitly stated or criti- denitrifiers in the rhizosphere, compete for soil NO However, to cally examined. There is evidence (reviewed in Chapter 4) that poten— tial denitrification rates are increased in the rhizosphere and it is frequently concluded that plants stimulate denitrification (for exam- ple; Focht and Verstraete, 1977) by increasing the supply of carbona- ceous substrate and so the demand for electron acceptor. Yet merely by removing NO _ from soil plants would be expected to reduce denitri- 3 fication under some conditions. It might be possible to schedule the application of N to favor the plant in this competitive relationship. To devise such a strategy more information is required about the relative rates of denitrification in rhizosphere and non-rhizosphere soil. The effect of soil conditions and stage of plant growth on the rhizosphere activity and on the rate of plant uptake, and the depen- 3 trations must also be determined. Initial investigations of these dence of denitrification and plant uptake rates on soil NO concen- relationships are reported in Chapter 4. With further information of this kind it might be possible to predict, for example, that at X days after planting uptake is rapid yet the rhizosphere population has not become very active. At X + Y days plant uptake might begin to decline and the potential rhizosphere activity increase due to exudation and death of root cells. It could then be suggested that N be applied so that maximum N03- concentration occurs at time X and is greatly re- duced at time X + Y. The cost of this strategy is small. In fact, it may be zero since it is currently practiced to some extent. Split applications, applying a portion of the N before planting and a portion after plant emergence, have been found to increase the efficiency of fertilizer use and improve the chances of optimum plant growth. The cost of an additional trip across the field to fertilize is about 2 $/A, if done by the farm operator. On irrigated soils there is no significant 9 application cost since N solution can be added to the irrigation water. However, N solution is somewhat more expensive than anhydrous ammonia or solid N fertilizer. Although these methods are currently in use, it is doubtful that the schedules have been optimized for maximum plant growth or minimum denitrification. Computer models may eventually be available to schedule irrigation for farmers (Jackson gt 31, 1977). With sufficient knowledge about the relationship between plant growth and soil nitrogen transformations, scheduling of N applications could be incorporated into this service. Scheduling irrigation to minimize denitrification. It has been established that soil moisture, through its effect on soil aeration, is of primary importance in determining denitrifi- cation rate (Bremner & Shaw, 1958). Increased denitrification rates have been associated with application of irrigation water and rainfall (Ardakani E£.§l°’ 1977). This relationship suggests that in irrigated systems soil moisture might be better controlled to reduce denitrifi- cation. A reasonably wide range of soil moisture, from about 50 to 100% of field capacity, permits optimum plant growth. Therefore, there is some opportunity to manipulate soil moisture without reducing yields. The quantitative relationship between soil moisture and denitri- fication rate has not been determined. A steady-state approach to this question is insufficient; the dynamics of the response to changes in aeration state are also important. How soon after irrigation does denitrification rate begin to increase? How long do denitrifiers and their abilities to denitrify persist under drying conditions? 10 Our lack of understanding of the effect of soil moisture on denitrification is demonstrated by the following two strategies. They are opposite, yet both are reasonable based on current limited know- ledge. Frequent brief irrigation could be scheduled to keep soil moisture at a low, constant level. This might reduce denitrification by keeping the soil continuously well-aerated. Trickle irrigation might be the best way of accomplishing this. On the contrary, in- creasing the drying period between irrigations could greatly reduce the size and activity of the denitrifying community minimizing the response to the brief period of saturation during irrigation. It seems likely that these practices would actually increase efficiency of water use and so might decrease costs. An indication of the possible relative expense of altering irrigation practice can be obtained from the fuel and labor cost of 2 $/A inch of irrigation water. (In Michigan, irrigated corn requires about 8 inches of water per year.) Therefore, the cost of refined irrigation management would probably be much less than the denitrification cost of 8 $/A and could yield additional profits from increased efficiency of water use. Research objectives. Many of the questions posed in this chapter can be answered by laboratory research. I believe the following research objectives to be consistent with the need for lowcost methods of increasing the efficiency of N fertilizer use. 1. Develop new methods for the study of soil denitrification. Denitrification methodology has been reviewed in detail by Focht (1978). Current methods, briefly discussed in Chapter 2, are 11 insensitive or tedious or difficult to apply to undisturbed systems. A large part of my thesis research has been devoted to investigation and development of new methods. The reported inhibition by acetylene of N20 reduction to N2 (Balderston gt _l., 1976; Yoshinari and Knowles, 1976) was seen to be a potentially useful tool. Very sensitive analytical techniques are available for N20 measurement. This method also eliminates the problem of detecting low rates of N2 production against a very high atmo- spheric background of this gas. Chapter 2 considers the application of C2H2 inhibition methodology to soil denitrification studies. The number of soil denitrifiers is not directly related to soil denitrifying activity primarily because denitrifiers can respire either 02 or N oxides. Nevertheless, enumeration of soil denitrifiers is frequently of interest. Several currently used methods of enumer- ation have been evaluated and found to be generally unsatisfactory. Appendix A will present our investigations of possible modifications. It is difficult or impossible to study the in gitg physiology of denitrifiers or any other specific soil microorganism. A possible solution is the use of an antibiotic (rifampicin) resistance marker. This would permit quantitative recovery of the microorganism from soil. Preliminary investigations of this approach are presented in Appendix B. The use of the radioactive isotope 13N, despite its ten minute half-life, is an exciting new development in denitrification research. My role in the development and use of 13N methodology has been secon- dary to that of other members of our group (Mary and Richard Firestone and Michael Betlach), and so this subject will be only briefly dis- 12 cussed in Chapter 2. A complete report of this method is in prepara— tion (Tiedje_g£_al., 1978). 2. Investigate the denitrification response to the depletion of 02 from soil. During the investigation of the C2H2 inhibition method two linear phases of denitrification rate were observed after the imposition of anaerobiosis. My second major objective was to determine the cause of this pattern and relate it to the response to decreases in soil aer- ation, i.e., irrigation or rainfall, in nature. It was anticipated that this research might reveal: A. The absence or presence of denitrifying activity (functional denitrifying enzymes) in dry soils. B. The mechanisms involved in the denitrification response to re- duced aeration. Possible mechanisms include removal of 02 inhibition, derepression of synthesis of denitrifying enzymes, and growth of denitrifiers. C. The rate of response to increased moisture. D. The magnitude of the response. E. The times and conditions for significant denitrification loss from agricultural soils. 3. Investigate the effect of plants on denitrification. I have hypothesized that the competition between plants and 3 O investigations (reviewed briefly in Chapter 4) have not considered denitrifiers is of great importance to the fate of soil NO Earlier this aspect but have apparently sought only to observe a stimulatory l3 effect of roots on denitrification, sometimes leading to ambiguous results. Chapter 4 describes the application of newly developed methods to the following questions: A. Can the presence of plants reduce denitrification rate under some conditions? B. When can a stimulatory effect be expected? C. If the denitrification rate is increased or reduced, what is the mechanism? This introductory discussion should not be taken to imply that all denitrification research must be justified by agricultural neces— sity. Denitrification is a critical reaction in the Earth's nitrogen cycle, and so has relevance well beyond agricultural practice. lo. 11. 12. 13. 14 LITERATURE CITED Allison, F. E. 1955. The enigma of soil nitrogen balance sheets. Adv. Agron. 7:213-250. Ardakani, M. S., H. Fluhler, and A. D. McLaren. 1977. Rates of nitrate uptake with sudangrass and microbial reduction in a field. Soil Sci. Soc. Am. Journal 41:751-757. Balderston, W. L., B. F. Sherr, and W. J. Payne. 1976. Blockage by acetylene of nitrous oxide reduction in Pseudomonas perfectomarinus. Appl. Environ. Microbiol. 31(4):504-508. Bremner, J. M. and K. Shaw. 1958. Denitrification in soil. I. Factors affecting denitrification. Journal of Agricultural Science 51:40-51. Council for Agricultural Science and Technology. 1976. Effect of increased nitrogen fertilizer fixation on stratospheric ozone. Report 53. 33p. Iowa State University. Ames. Crutzen, P. J. 1976. Upper limits on atmospheric ozone reduc- tion following increased application of fixed nitrogen to soil. Geophys. Res. Lett. 3:169-172. Delwiche, C. D. and B. A. Bryan. 1976. Denitrification. Ann. Rev. Microbiol. 30:241-262. Firestone, M. K., M. S. Smith, R. B. Firestone, and J. M. Tiedje. 1977. Factors influencing the biological release of N20 during denitrification as determined by N methodology. Agronomy Abstracts. Focht, D. D. 1978. Methods for analysis of denitrificatin in soils. .In Nitrogen in the Environment. D. R. Nielsen and J. G. McDonald (eds). Academic Press. Focht, D. D. and W. Verstraete. 1977. "Biochemical ecology of nitrification and denitrification". In Advances in Microbial Ecology, v.1. M. Alexander (ed). Plenum Press. Iwasaki, H. and T. Matsubara. 1972. A nitrite reductase from Achromobacter cycloclastes. J. Biochem. 71:645-652. Jackson, R. D., S. B. Idso, R. J.Reginato, and W. L. Ehrler. 1977. Crop temperature reveals stress. Crops and Soils Magazine. 29(8):10-13. Johnston, H. 1972. Newly recognized vital nitrogen cycle. Proc. Nat. Acad. Sci. U.S.A. 69(9):2369-2372. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 15 Lam. Y. and D. J. D. Nicholas. 1969. Aerobic and anaerobic res- piration in Micrococcus denitrificans. Biochim. Biophys. Acta 172:450-461. Payne, W. J. 1973. Reduction of nitrogenous oxides by micro— organisms. Bact. Rev. 37(4):409-452. Payne, W. J., P. S. Riley, and C. D. Cox, Jr. 1971. Separate nitrite, nitric oxide, and nitrous oxide reducing fractions from Pseudomonas perfectomarinus. J. Bacteriol. 106:356-361. Rolston, D. E., M. Fried, and D. A. Goldhamer. 1976. Denitri- fication measured directly from nitrogen and nitrous oxide gas fluxes. Soil Sci. Soc. Am. Journal 40:259-266. Rowe, J. J., B. F. Sherr, W. J. Payne, and R. G. Eagon. 1977. A unique nitric oxide-binding complex formed by denitrifying Pseudomonas aeruginosa. Biochem. Biophys. Res. Comm. 77(1): 253-258. Schmidt, E. L. 1978. Nitrification by soil microorganisms. Abstracts of 144th Annual Meeting, American Association for the Advancement of Science. Stanford, G., R. A. Vander Pol, and S. Dzienia. 1975. Potential denitrification rates in relation to total and extractable soil carbon. Soil Sci. Soc. Am. Proc. 39:284-289. Starr, D. L. and J. Y. Parlange. 1975. Nonlinear denitrification kinetics with continuous flow in soil columns. Soil Sci. Soc. Am. Proc. 39:875-880. Stefanson, R. C. 1972. Soil denitrification in sealed soil- plant systems. I. Effect of plants, soil water content and soil organic matter content. Plant and Soil 33:117-127. Yoshinari, T. and R. Knowles. 1976. Acetylene inhibition of nitrous oxide reduction by denitrifying bacteria. Biochem. Biophys. Res. Comm. 69(3):705-710. CHAPTER II THE ACETYLENE INHIBITION METHOD FOR SHORT-TERM MEASUREMENT OF SOIL DENITRIFICATION AND ITS EVALUATION USING N The increased cost of fixed nitrogen and the possibility that soil-evolved N20 may contribute to atmospheric ozone depletion (McElroy, 1976) have caused renewed interest in the process of deni- trification. Despite a considerable mass of research on denitrifi- cation, reliable values for rates of N2 and N20 production in field soils are lacking. Due to limited sensitivity, previous methods have required extensive amendment of soils and/or long term incubation. Although these methods have elucidated the basic controlling factors, the dynamics of denitrification and the quantitative effects of en- vironmental or management parameters on natural soils are largely unknown. The inhibition of N20 reductase by acetylene in pure culture was reported by Yoshinari and Knowles (1976) and by Balderston, 35 a1 (1976). It is now widely accepted that N 0 is an obligatory, and 2 probably freely diffusible intermediate in the denitrification pathway (Payne, 1973; St. John and Hollocher, 1977). Therefore, one would expect the quantity of N 0 produced by C -inhibited microorganisms 2 2H2 to be a direct measure of the total gaseous N produced without inhi- bition. If N20 is the sole denitrification product, analysis is greatly simplified since N O, unlike N is a minor atmospheric con- 2 2’ stituent (approximately 300 ppb) and can be assayed by sensitive gas chromatographic detectors. The successful application of this method to soil denitrification studies has been reported by Yoshinari, e£_al 16 17 (1977). I have used gas chromatography and 13NO3- to evaluate the acetyl- ene inhibition technique in soils and have identified the conditions and soil types for which this method is valid and for which blockage of N20 reductase is complete. The radioactive isotope, 13N, provides an extremely sensitive assay, with excellent temporal resolution, which can be used without alteration of native N03- concentration. MATERIALS AND METHODS The soils used are described in Table l. Soils were near field capacity when collected and, without drying, were passed through a 5 mm sieve and stored in sealed plastic bags at 20 C until used. The storage period varied from 1 day to 6 months. Soil slurries were used in most of the C2H2 inhibition assays. This made it easier to amend the soils and simplified interpretation of the results by limiting the effects of diffusion. Soil, usually 75 g fresh weight, was placed in 125 m1 erlenmeyer flasks and the desired amendments, with 30 ml of H 0, were stirred in. All solute additions 2 were on the basis of weight per fresh weight of soil. The aerobic assays were conducted on soils, without added water, in 50 ml centri- fuge tubes. The vessels were sealed with a rubber stopper which was pierced by a glass tube capped with a serum stopper. Flasks were made anaerobic by twice evacuating and flushing with He or Ar. Acetylene was added as desired. Slurries were incubated on a rotary shaker at 250 rpm. In all gas chromatography experiments there were at least three replicates for all treatments. Samples of the headspace ( ;fO.5 ml) were periodically removed 18 .Ammmav umcawum mam vacuusm mo cosmos mzu >2 mwGHEuwumv mm3 mam Hwom w\ow: mo mafia: ca ma 0 manmufiamumcfiz « mamvaaqmz oma m.H «.0 ofiucmeemmm vcmw zamoq mxcfiam Lameaaam; sea ~.~ o.o awake emoH Assam “ems: umwuammwume mam o.mw m.o awake eon: mamaaumo Haoavmfiwum mma N.m 0.5 owaxh Smog cOUmJOOHm woman: «0 mpnmuwampmcwz owcmmuo N :a coHumowwfimmmfio eunuxwe mmfiumm .vow: wawom Lo mowuwfiuwuomumso .~ manme 19 with a Pressure-lock syringe (precision Sampling Corp., Baton Rouge). In most experiments a Carle model 8515 gas chromatograph (Carle Instru- ments Inc., Fullerton, Calif.) with a microthermistor detector was used to analyze gases. The sensitivity of this instrument was in- creased by addition of a 5X/10X operational amplifier. The lower limit of detection was about 100 ppm (v/v). A Porapak Q column (3 mm x 1.8 m) was used to separate CO N O, and C 2, 2 A switching valve 2H2' made it possible to direct the effluent of this column either to the detector or to a Molecular Sieve 5A column (3 mm x 1.8 m) to separate N2, 02, and NO. The column temperature was 450C and the carrier gas was He at a flow of 25 ml/min. Peak areas were determined with a computing integrator. A Perkin-Elmer model 900 gas chromatograph (Norwalk, Conn.) with a 63Ni electron capture detector (Rasmussen, et al., 1976) operated at 3000C was used for measuring N20 concentrations from ambient to 100 ppm. Excellent separation of N20 was achieved by operating the Pora- pak Q column at ambient temperatures. The carrier gas was 5% CH4 in Ar with a flow rate of 40 m1/min. The 13N was generated at the MSU Cyclotron by the reaction 160(p,a)13N using water as a target; the details of the production, characterization and detection of this isotope are described elsewhere (Tiedje, g£_al., 1978). In these experiments the 13N used was :_85% 13NO3- with minor quantities of 13NH4+ and 13N02-. In some experi- ments N114+ was removed prior to use by making the 13N solution alka- 3 + N02-) was added to the flasks containing soil slurries. In certain line and evacuating to dryness. Approximately 1 mCi of 13N (4 pg NO experiments 14N03- carrier was also added. To determine rates 20 of gas production, a gas flushing system similar in principal to the one of Gersberg gt a1. (1976) was used. The flask containing a soil slurry and incubated on a magnetic stirrer, was connected to a helium sparging system which continuously flushed gaseous products into the differential trapping system (Tiedje g£_al. 1978) which separated l3N20 from 13N2 to allow quantitation of each gas. Specific activity of the denitrification products was assumed equal to that of the reactants, which was determined by counting a subsample of the 13N solution and by extracting the soil following the incubation and measuring NO3 + N02- by Technicon Autoanalyzer standard methods. In all experiments, the concentration of N20 in solution was calculated from the measured headspace concentration and denitrifi- cation rates were corrected accordingly. In some slurries as much as 1/2 of the total N 0 produced remained in solution. It was verified 2 that published values of the Bunsen absorption coefficient approxi- mated the N O solubility in this system by adding N 2 O, in the concen- 2 tration range we normally encountered, to autoclaved soil slurries. The slurries were shaken, allowed to equilibrate, and the quantity of N20 remaining in the headspace was determined. At 200 C I obtained a coefficient of 0.74 which is slightly higher than the published value of 0.66 for pure water solutions. Assays for testing completeness of acetylene inhibition: Three approaches were used to determine the completeness of the C2112 inhibition in soil slurries. Anaerobic slurries of Brookston soil were preincubated for 48 hours with 0.5% added glucose to deplete 3 . atmosphere was replaced as before. Various quantities of C2112 were naturally occurring NO After this preincubation period the flask 21 added and allowed to mix for 25 min prior to injection of 0.25 ml of N20. N20 remaining in the headspace was periodically determined. The second approach evaluated the recovery of added NO3 —N as NZO-N. Unamended Brookston soil was preincubated for 48 hours as an anaerobic slurry. Eight ppm NO --N and 0.1% glucose were added after 3 injecting the desired quantities of C2H2. The quantity of N after 24 and 48 hours was determined. 20 present The third and most rigorous test of the effectiveness of CZH2 . . . . 13 . Inhibition was conducted w1th N methods. Brookston 5011 was pre- incubated anaerobically for 48 hours with 4 ppm NO --N and 0.5% glu- 3 cose. Preliminary experiments had shown that with this treatment all of the added N03- would be depleted and that N2 would be high. Acetylene was added to the flasks as desired and the O reducing activity contents mixed for l to 2 hours. Carrier-free 13N was then added to soil incubated on a rotary shaker. After approximately 15 min a 10 ml gas sample was removed by syringe and injected into the differential 13N gas trapping system to determine 13N O and 13N2 produced. Correc- 2 tion was made for differential solubility of the two gases. To some flasks 2 ppm 1[INO3_-N were added before the addition of 13N. RESULTS Completeness of acetylene inhibition of NqO reduction in soils: Reduction of N20 added to soil in the presence of varying C2H2 concentrations is shown in Table 2. The N20 concentration slowly decreased even in the flasks with 1 atm CZHZ' slow adsorption reactions rather than biological reduction. Whereas This could be due to very low concentrations of C2H2 did inhibit N20 reduction, 0.15 atm 22 .00H x awe ma up 0 z m mEau voumofivcfi um o z mum mmmonucmuma aw mosam> N N 1» Im mum muwafiq + .xmmac some 00 emuem owz modes: ~.HH « mo.u x ANWV AQWV Aomy «m. + n.0H 0H. + m.0H an. + w.oH o.H Aomc Aomv Anmv cm. + H.0H mm. + H.0H BC. + m.OH mH.o Se :8 3e mm. + o.¢ ca. + m.o NH. + 0.0H «0.0 GB Ce 2% NH. + m.m mm. + m.m «N. + c.0H Ho.o 2: $8 3% o OH. + ©.m +0 . + m.o o «wcwchEmu 0N2 mmHoEn AEumv musoc cm muse; m muse: N szua meu comumnsucH .mcofiumuucmucoo mamazumom mooHum> Lo oucmmmua msu cm Hwom xn omwxo mocha“: mo newuuammm .N mHQmH 23 CZHZ was required for maximum inhibition. The stoichiometric conversion of N03 to N20 in the presence of various concentrations of acetylene is shown in Table 3. The recovery of N20 was complete for all acetylene concentrations at 24 hours but by 48 hours N O was apparently further reduced in flasks containing 2 the lower concentrations of C2H2° one, I verified that significant quantities of N In this experiment and the previous 0 were not produced 2 by the N03- depleted soils with 1 atm CZH2 but without added NO3-. The effectiveness of the acetylene in inhibiting the reduction of 13 13 N20 to N2 is reported in Table 4. Because of the high specific activity of 13NO3-, the N03- concentration was extremely low. Under these conditions high acetylene concentrations (:_O.15 atm) were re- quired to obtain an effective block. When a small quantity (2 ppm) of 14NO3--N was added, a dramatic increase in N20 was noted. This 3 aids the inhibition of N20 reduction. This interpreta- tion is also supported by the other two experiments since high con- suggests NO 3 but not when N03- was high (first 24 hours of second experiment, centrations of acetylene were needed when NO concentrations were low (Table 3). Determination of soil denitrification rates: Table 5 provides a comparison of denitrification rates measured by the C2H2 inhibition method and the 13N method. The experiments necessarily differed in that the soil for the 13N assay was continous- 1y stripped of product gases while for the acetylene inhibition assay the flasks were sealed and incubated on a rotary shaker. In spite of these differences the results were reasonably similar for the two methods. Furthermore, the ratios of NZO/(total gaseous N) determined 24 .xmmaw non vmvvm m oz mmHoE: 0.Nq ”210 m0. x u Im mum mugs“; N m + 2 mm woum>oomu 2: oz 0000M mo unmoumm « Ambv Ava 0.0 + 0.0N m.m + 0.0m 0.H Abov ANov N.0 + m.ma m.N + #.0N mH.0 68 an: m.0 + 0.HN 0.0 + 0.m~ 00.0 Awhv «Amway q.m + n.0H +0.5 + H.mN H0.0 0 0 0 0N2 mmHOE: AEumv meso: wq muse: 0 N N a N m 0 oEWMIdOWDmnsocw .mcofiumuucmocou onwazuwow mDOwum> mo oocmwmua ecu c“ mvwxo mSOHBHc mm wumuufic vmvvm mo >pm>ooom .m magma 25 m oo.c 2.: oz and N + mo. Hm.o znumoz and N + o mm.o ma. mo.o mo. om.o Ho. mo.o o Aeumv ANZmH + 0N2m~v\o~2m~ Nmmoa .z zn vmcproqu mm mausvopa cowumuwwwuuwcmw mo ovumu may mm mcopkumom Lo mcofiumpucocou m30wum> mo uumwwm .c maamH 26 .mmmwnucouma a« mum mwmownouwmam mo ummco umumm muso: + .ocmHNUmom Bum H.o 0cm moz cmcvm on vm>fimomu I «H mumzuo umcwazumom Eum no.0 0cm lemoz3 and w Saws vowcmem mm3 Hwom COumxooum x $5 :5 o~.o 26 3-3 mxzwmm 2.0 $6 $4 2.~ 8L: mHmHHumU 0N.0 qm.0 mq.a 0n.0 AmINv scoumxooum 3.0 8.0 $6 3.0 :78 «CODmxooum A N N N IIIIII.0 z + sz\0 Zollllll. IHISHE . Hum . mmw meoECI N N N N . c . c a“: H I 0 2mH aw: H I 0 zmH muosmopm mo owumm dump acaumofiufiuuwcma Hfiom .mvosume sewuwnwccfi N:N0 0cm 2 >3 UmCNEumumn MN mauspoua 00 mewumu 0cm mommy GOMUmowwwuuwsmv mo comfiumano .m mHQMH 27 by the 13N method were similar to those determined by the acetylene inhibition method. For the latter method the ratio was determined by comparing N 0 produced in the presence of acetylene to N 0 produced by 2 2 the same soil in the absence of acetylene (Yoshinari, E£.§l° 1977). Both methods indicate that the total denitrification rate of the Brookston soil, as well as the proportion of N 0, increased soon after 2 the onset of anaerobic conditions. In the sealed vessels with N03- present, the reduction of headspace N 0 does not appear to be sig— 2 nificant since the ratio of gaseous products was similar in the sealed . 13Y and continuous flow ( N) assay systems. It was also observed that when the flasks without C2H2 were evacuated and flushed to remove accumulated N20, the rate of N20 production resumed at essentially the same rate (data not shown). Therefore, the rate of release of N20 was apparently independent of N 0 concentration in the headspace. 2 During the C inhibition assay of the Brookston soil, we also 2H2 monitored CO2 concentration in the headspace. There was no signifi- cant difference between the flasks with and without C2H2, suggesting that 0.03 atm C2H2 had little effect on soil respiration. In other experiments with up to 1 atm acetylene I have observed no significant effect of CZH2 on CO2 evolution within a 12 hour assay period. I have also observed that the rate of N 0 production was not affected by C2H 2 2 concentrations up to 1 atm. The CZHZ inhibition method was used to compare denitrification rates of four different soils over a longer time period. The soils were incubated as anaerobic slurries at 180 C with 8 ppm NO3 -N and 0.04 atm acetylene added. The time course of N20 production is plotted in Figure l. Shortly after the onset of anaerobic conditions, the 28 Figure 1. Nitrous oxide production by 4 C2H2 inhibited soils. Points are means of 3 replicates. 29 mpmwpemu so \ \ 00¢ Acwsv we?“ cowumnaucm com cow 0 2: u M com a ZN 0 a p m 8m m. 08 28 Figure l. Nitrous oxide production by 4 C2H2 inhibited soils. Points are means of 3 replicates. 29 Assay mews cowsmnsucm com cos com com ooH d OOH U m com m N Z 0 .m p .M 8m m. .....t \s ooe m? .P 7.3 \o \ \ \ \ \ s 30 rates increased for all soils until an approximately constant rate was attained. The soils differed in the time required to reach linearity and also in the ratio of the final linear rate to the first rate observed. The denitrification rates corresponded with mineralizable carbon. I have found that denitrification rates can easily be measured without carbon or N03- amendments. In some soils denitrification rates were stimulated by these amendments, however in other soils the rates were not increased by either or both additions. These effects have not been thoroughly investigated by me but other work in our laboratory, conducted primarily by Mary Firestone, has demonstrated that in anaerobic soil slurries denitrification rate is independent of NO3 concentration between about 10 and 1000 ppm N. Since the C2H2 inhibition method requires no substrate additions, we have been able to apply it to less disturbed soils. Figure 2 allows a comparison of the denitrification rates of an unamended anaerobic slurry, anaerobic soil aggregrates (no water added), and aerobic aggregates. The Miami soil, which had been sieved and kept at 20C for 5 months, was used in all cases. The reduced rate in the anaerobic aggregrates without added water suggests that substrate diffusion limited the denitrification rate. The soil had been stored moist (near field capacity) and so contained ample water for biolo- gical activity. Production of N20 could not be detected in the aerobic treatment using the microthermister detector (detection limit 8 nmoles - g-l). However, with the much more sensitive electron capture detector we were able to observe a linear increase in head- space N 0 which continued throughout the experiment (32 hours). This 2 Figure 2. 31 Comparison of denitrification by the Miami soil under different experimental conditions. All soils were treated with C2H2' Points are means of 3 rep- licates. 32 00¢ AcFEV wEPp cowumazucH DOM DON OOH -m.mm—OE: wv H umwoe ppm?» ownoemm ownoemmcm Bates sport 0 Ln 0 O r—I omfl I_1.Los Rap 5,02N salowu 33 rate was 2.7 x 10-4 nmoles - g - min—1 which is about 1000 x less than for the anaerobically incubated soils. Headspace gas analysis after 32 hours indicated that 02 had not decreased by more than 4%. DISCUSSION I believe that the C2H2 inhibition method will be extremely valuable in denitrification research since it allows determination of denitrification rates in unamended soils and within a short time period, thus minimizing changes due to the assay environment. My purpose was to evaluate this method by comparison with an independant method and to determine the assay conditions for acceptable results. I confirmed the finding of Yoshinari SE El' (1977) that acetylene blocks the reduction of N20 by the indigenous soil microflora. However, I found that this inhibition was enhanced in the presence of nitrate. Thus, when soil nitrate concentrations were low, e.g. less than a few ppm, higher acetylene concentrations were needed to obtain a suitable degree of inhibition. My findings indicate that 0.1 atm of acetylene should be adequate for low nitrate samples (or samples where the nitrate would be exhausted before the assay is terminated). The marked effect of nitrate on N20 reduction also has important implications for the question of what causes N20 production in soils. As shown in Table 4, as little as 2 ppm N03--N caused a shift from only 3% N20 to 91% N20. This finding and further experiments in our laboratory conducted by Mary Firestone (unpublished results) indicate that NO3- concentration is of primary importance in determining the percentage of denitrification gases released as N O. 2 The absence of a detectable effect of C2H2 on soil respiration 34 suggests that the inhibitor does not have confounding short-term influences on soil processes which could, directly or indirectly, alter denitrification rates. Acetylene is a biologically active com- pound, however, (for example; Brouzes and Knowles, 1971) and it may be necessary to adjust the concentration used to the system being studied. The similarity of rates measured by the 13N and the acetylene inhibition methods was very encouraging. It is particularly interes— ting that comparison of rate of N 0 production by inhibited and unin- 2 hibited soils provided a measurement of NZO/(NZO + N2) which corre- . 13 13 13 . lated well Wlth N20/( N20 + N2). This approach is apparently valid for the assay conditions described here, however caution is advised in its application to other systems. The acetylene inhibition method has several advantages over pre- viously used methods. The problem of lack of sensitivity due to the high atmospheric concentration of N is eliminated. When coupled with 2 electron capture detector analysis of N 0, it is extremely sensitive. 2 A significant advantage is that substrate additions and long term in- cubations are not required and only generally available equipment is necessary. Since C2H2 is water soluble and, as a gas, is readily diffusable in a porous matrix, the method should be applicable to undisturbed systems-~soil columns or cores, or perhaps field studies. I have made preliminary attempts to conduct C2H2 inhibition assays on soil cores with mixed results, probably because diffusion of product and/or inhibitor was limiting for our chamber design. Other potential prob- lems are the effect of nitrate concentration on the inhibition and the oxidation of C2H2 which can occur under aerobic conditions. Though I 35 feel the method still holds promise for work with undisturbed soils, simple designs may not suffice and any procedure developed will have to be thoroughly tested to verify its reliability. I observed that denitrification rates increased within a few hours after the imposition of anaerobic conditions. Soil assays without added electron donor generally yielded bilinear plots of accumulated N20 vs. time (Figure 1). The rapid temporal changes in denitrifying activity which occur after the imposition of anaerobiosis in the laboratory should be analogous to the processes which occur when a soil or a portion of the microsites becomes anaerobic in nature. Therefore, these temporal patterns have been examined in detail. The results of these studies are presented in the next chapter. The strong effect of 02 on the indigenous denitrifying enzymes was indicated by the 1000 fold difference in denitrification rates of soil aggregrates under aerobic vs. anaerobic atmospheres. The aerobic rate was equivalent to 2 kg - ha furrow slice.1 - 100 days-1. Deni- trification obviously proceeds at a very slow rate in well aerated soils. However, whenever oxygen becomes limiting at a microsite, the indigenous denitrifying enzymes should immediately begin reducing nitrate. Because of this strong 0 effect significant amounts of 2 nitrate could be rapidly lost from soils following irrigation or rainfall. This scenario would predict that denitrification occurs in periodic bursts, in response to changes in 0 status, against a back- 2 ground of very slow, yet continuous denitrification. 10. 11. 36 LITERATURE CITED Balderston, W.L., B. Sherr, and W.J. Payne. 1976. Blockage by acetylene of nitrous oxide reduction in Pseudomonas perfectomarinus. Appl. Environ. Microbiol. 31:504-508. Brouzes, R. and R. Knowles. 1971. Inhibition of growth of Clostridium pasteurianum by acetylene: Implications for nitro- gen fixation assay. Can. J. Micro. 17:1483-1489. Burford, J.R. and J.M. Bremner. 1975. Relationships between the denitrification capacities of soils and total, water-soluble and readily decomposable soil organic matter. Soil Biol. Biochem. 7:389-394. Gersberg, R., K. Krohn, N. Peck, and C.R. Goldman. 1976. Denitrification studies with N-labeled nitrate. Science. 192:1229-1231. McElroy, M.B., J.W. Elkins, S.C. Wofsky, and Y.L. Yung. 1976. Sources and sinks for atmospheric N20. Rev. Geophy. Space Phys. 14:143-150. Payne, W.J. 1973. Reduction of nitrogenous oxides by micro- organisms. Bacteriol. Rev. 37:409-452. Rasmussen, R.A., J. Krasnec and D. Pierotti. 1976. N 0 analysis in the atmosphere via electron capture-gas chromatography. Geophys. Res. Letters 3:615—618. St. John, R.T., and T.C. Hollocher. 1977. Nitrogen 15 tracer studies on the pathway of denitrification in Pseudomonas aeruginosa. J. Biol. Chem. 252:212-218. Tiedje, J.M. 35 El. 1978. (manuscript in preparation). Yoshinari, T. and R. Knowles. 1976. Acetylene inhibition of nitrous oxide reduction by denitrifying bacteria. Biochem. Biophys. Res. Comm. 69:705-710. Yoshinari, T., R. Hynes, and R. Knowles. 1977. Acetylene inhibition of nitrous oxide reduction and measurement of deni- trification and nitrogen fixation in soil. Soil. Biol. Biochem. 9:177-183. CHAPTER III PHASES OF DENITRIFICATION FOLLOWING OXYGEN DEPLETION IN SOIL It is generally believed that denitrification occurs in agricul- tural soils predominantly under saturated or near-saturated conditions (Focht and Verstraete, 1977). Soil denitrification rates have been shown to increase with added water or reduced aeration (Bremner and Shaw, 1958; Ardakani 33 al., 1977; Pilot and Patrick, 1972). This is consistent with known physiology of denitrifiers; 02 both inhibits denitrifying enzyme activity and represses synthesis of new denitri- fying enzymes (Payne, 1973). There are reports suggesting biological production of nitrogen gases in well-aerated soils (Starr gt al., 1974; Broadbent and Clark, 1965; Bremner & Blackmer, 1978), though the significance of nitrogen loss under these conditions has not been determined. The close relationship between aeration state and deni- trification rate indicates that it is important to characterize the short-term soil response to reduced aeration. Until now, methodo- logical limitations have precluded studying the dynamics of the soil response, that is, how rapidly and to what extent denitrification rates change following the onset of anaerobic or partially anaerobic conditions. Nor has the biological component of this response, the short-term reaction of soil denitrifiers to imposed anaerobiosis, been examined. In this chapter I address these questions which are impor- tant to the basic understanding, and to the potential control and prediction of denitrification. 37 38 MATERIALS AND METHODS Soils used were Brookston loam, Spinks loamy sand, Miami sandy loam, and Carlisle muck; their characteristics have been presented previously (chapter 2). The maximum water holding capacity, as deter— mined by saturating soil in a filter paper funnel, was 43 ml/100 g for the Brookston and 36 m1/100 g for the Spinks and Miami. After col- lection the soils were stored at 2 to 4 C without drying. However, for many experiments fresh samples were collected from the same sites immediately before experimentation. All soils were at or below field capacity when collected. The validity of the acetylene inhibition method for measuring denitrification rates has been established previously (chapter 2). The methods used here were similar. Briefly: soil to be assayed was made into a slurry and incubated anaerobically in closed flasks. Acetylene (0.1 atm) was added to inhibit the reduction of N 0 to N . Nitrate 2 2 was also added to the soil suspensions, 10 ppm NO --N (soil fresh wt 3 basis) for short-term assays and repeated additions of 200 ppm for longer experiments. Rapid attainment of anaerobiosis for phase I assays was insured by magnetically stirring the suspensions while evacuating and flushing the flasks; incubation was under a He atmo- sphere. There were at least 3 replicates of all treatments. To determine denitrification rates in aerobic atmospheres, soils were passed through a 4 mm sieve to break up large clods and remove debris. This treatment had little effect on smaller, stable aggre- gates. Thirty five g of soil was gently shaken with 0.1 atm C2H2 in 39 a stoppered 50 m1 centrifuge tube, then compacted by gently tapping the tube. The concentrations of N20 were measured by gas chromatography using 63Ni electron capture and microthermistor detectors as previ- ously described. When Chloramphenicol (Calbiochem; La Jolla, Cal.) was used to in- hibit protein synthesis, it was added to the suspension at 0.27 mg/g soil. These experiments were done with a mixture of 50 g washed silica sand and 5 g Brookston soil as well as the Brookston soil alone. The sand mixture was incubated anaerobically with succinate and N03- several days before the experiment to develop denitrifying populations. Pure cultures of denitrifiers were grown aerobically to log phase in nutrient broth (Difco; Detroit, Mi.). They were harvested on 0.45 p Millipore filters and washed with three volumes of phosphate buffer (pH 7.0, 0.05 M). The filters with the cells were placed in flasks of nitrate broth (Difco). The flasks were then made anaerobic and assayed in the presence of acetylene in the same manner as the soil suspensions. Some cultures were also assayed in nitrate broth with 0.5 g/l Na thioglycollate, a reducing agent. The flasks with thioglycollate were evacuated and flushed with He 16 hours before innoculation. The denitrifiers had been isolated from soils and were fully characterized during earlier work in this laboratory (T. N. Gamble, The Commonality of Numerically Dominant Denitrifier Strains Isolated from Various Habitats, M.S. Thesis, 1976). They were: Pseudomonas fluorescens biotype II 72, P, fluorescens biotype IV 206, P, stutzeri 224, P. aureofaciens 59, and Flavobacterium sp. 175. 40 Numbers refer to our strain designation. Most probable numbers of denitrifiers were determined with a microtiter system (Rowe E£.El°’ 1977). Disappearance of nitrate and nitrite from nitrate broth was determined with diphenylamine. RESULTS Two linear phases of denitrification were observed following the imposition of anaerobiosis on aerobic soil (Figure 3). Acetylene was used in these experiments so N 0 production actually represents total 2 denitrification. Phase I lasted from approximately 15 min, the time for N20 to reach easily measured concentrations, to between 1 and 3 hours. Previous work had shown that this initial period is followed by an increase in rate until a second linear phase, phase II, is established. Some phase II data from this earlier work is included in Figure 3 to demonstrate the biphasic nature of denitrification rates. The duration of both phases was quite variable among soils. However, the pattern was observed consistently with both fresh and stored (moist) soil samples. In some experiments attempts were made to eliminate 0 more 2 rapidly and completely. These included deaerating the solution and flask before adding the soil, varying the duration of evacuating and flushing the flasks, and agitating the suspensions while evacuating. These procedures did not alter the bilinear pattern indicating that residual 02 or delayed establishment of equilibrium was not respon- sible for the observed phases. 41 Figure 3. Two linear phases of denitrification in 3 C2H2 inhibited soils. Points are the means of 3 replicates. 42 cm Acrev we?» cowumnzocH 0v mxcwgm couwaOLm H wmufiomamo wCHvaon uwums answme «0 New w on ‘ m 0 I a S N z ~.o 0 m I U; a Du p S d Du a a _ I ma 56 51% of maximum water holding capacity and the same soil adjusted to 94% of maximum water holding capacity (approximately saturation) just prior to the assay were compared for pattern of N20 evolution. A very low rate of N20 evolution, 1.1 x lo-lamoles N20 - m1 headspace.1 omin-l, was observed from the drier soil and initially from the moistened soil. Approximately six hours after wetting, the rate began to in- crease rapidly. Similar results were observed with several other soil samples. The lag period ranged from 4 to 7 hours. The 02/N2 ratio in the headspace, as determined by gas chromatographic analysis was reduced by less than 5% during these experiments. DISCUSSION It is concluded that the following sequence of events occurs when a soil or soil microsite becomes anaerobic: first, the 0 inhibition 2 of native denitrifying enzymes is removed, which results in an initial linear phase of denitrification (phase 1). After a lag period of at least one hour newly synthesized denitrifying enzymes become func- tional and the rate increases (if permitted by substrate supply). At 4 to 8 hours all denitrifiers are fully derepressed so each cell has attained its maximum capacity to denitrify (phase 11). At this point denitrification rate can increase only by an increase in number of cells; significant growth will occur only when the supply of electron donor is large. These events and their relationship to the phases observed are summarized in Table 8. It should be noted that phases 11a and 11b were referred to as phases II and III in an earlier report (Tiedje 35 al., 1978). This designation has been changed to avoid the 57 Acopuwo maanHm>m nwflnv mumwmwuuwcow mo cu30uo wowmmmwocH wuficfiwmbcH wlq nHH zusouw ucmofimwcwwm Aconumo o: .vwmmouawumv zaasw magmafim>m 30Hv zuHCSEEoo msocmwmc:H ucmumcou mUHc«mmwcH wlq mHH mwmmSucxm mamucm mo scammouomuwb m>wuo< mewmwmuocH mic Mia cofiufimcmuw mmexncm wcfiumwxmlmua mo mufi>wuo< ucmumcou mIH «\H H vow uumum Amazonv wumu mo Hm>u0uGH mafia :oHumcmHaxm mofiumwuwuowwmzo vmufiamumcmu mmmcm .mcofiuwvcoo canouwmcm mo cofiuamoqu mnu umumm comumowmwwuwcwo Hwom mo mommza mcu mo cowuafinommc vwnwamumcmu .m manme 58 implication that these are always distinct phases. The evidence for this explanation of the phases can be summarized as follows: first, various methods of rapidly removing 0 had no 2 effect, indicating that residual 02 could not account for the lower initial rate. Chloramphenicol reduced phase II denitrification rates but not phase 1, suggesting that the increase was due to enzyme synthesis. The small initial stimulation by glucose relative to subsequent stimulation demonstrates that the capacity to utilize electron donor increases after phase 1. Pure cultures of soil deni- trifiers exhibited a lag time consistent with the duration of phase I before demonstrating derepression of denitrifying enzymes. Payne and Riley (1969) observed a slightly shorter lag time, 40 min, with the marine denitrifier, Pseudomonas perfectomarinus. This could be due to a difference in the physiology of this organism or might be related to a difference in habitat; it would be interesting to compare the phase pattern of soils with that of sediments, which are more or less continuously anaerobic. Phase I and phase II rates measure completely different factors in soil denitrification. Phase II rate corresponds to what is gener- ally called denitrification potential in the literature (for ex.; Balasubramanian and Kanehiro, 1976). My results indicate that this rate is a function of the number of soil denitrifiers and the lim- itations imposed on them by the energy and electron acceptor supply, pH, and temperature. All of these factors might be altered by im- position of the assay conditions. I have shown that addition of an energy source has a major effect on the rate and the pattern in phase II. It is also probable that moistening of a completely dry soil has 59 effects similar to adding an energy source. Denitrification potential has been shown to correlate very well with mineralizable carbon (Bremner and Shaw, 1958). In fact, a measurement of mineralizable C probably provides about as much information about soil denitrification as does denitrification potential. Phase I, in contrast to phase II or denitrification potential, is sensitive to the aeration state of the native soil. Phase I is essentially an enzyme assay and reflects the immediate biological effect of changes in soil moisture and aer- ation--extremely important factors for soil denitrification. There- fore, phase I rate appears to provide more information about in ElEB denitrifying activity than does denitrification potential and should be a more useful approach to the study of soil denitrification. The importance of derepression of denitrifying enzyme synthesis in nature, as well as in laboratory incubations, is indicated by the increase in phase I rate of soil during field irrigation. Also when water was added to aerobic soils a pattern was observed similar to that in anaerobic incubation, that is, a low initial linear rate was followed by an increase in denitrification rate. It is tempting to attribute this pattern to the same causes in aerobic and anaerobic incubations. Although derepression is undoubtedly involved, other factors must be considered in the aerobic experiments and in field soils. The time required for the generation of anaerobic microsites accounts for the longer initial phase in the aerobic experiments. The difference between anaerobic phase I and phase II was usually less than an order of magnitude, while the rate increase subsequent to wetting of aerobic soils was significantly greater. This appears to be due to removal of 02 inhibition of existing denitrifying enzymes 60 which would be expected to result in much greater rate increases than would derepression of enzyme synthesis. The results of this research demonstrate the value of working with simplified soil systems, in this case stirred anaerobic suspen- sions. I have been able to isolate and identify the biological effects of reduced soil aeration by controlling physical variables, particularly 0 and substrate diffusion. 2 My data provide some information about when and under what conditions denitrification could be expected to occur. I have demon- strated that denitrifying enzymes are present even in very dry soils. Since denitrifying enzymes are not constitutive (Payne, 1973) this implies that either denitrification can occur in anaerobic microsites of well aerated soils or that denitrifying enzymes may persist in functional form in the presence of O I have also observed evolution 2. of N20 from the Miami sandy loam at 51% of maximum water holding cap- acity in an aerobic atmosphere. If it is assumed that the rates observed in our assays are similar to field rates, then a gross approximation of field N loss can be made. This is done by assuming that the net flux of N20 from the soil in the tubes, i.e., the rate of accumulation in the headspace divided by the soil surface area, is equal to the flux of N from the soil in the field. Thus, the rate of N loss from the drier Miami sample (Figure 7) would be low, 69 mg N oha-l-day-l. Some well-aerated soils I assayed aerobically evolved N20 at significantly greater rates, up to 6 g N-ha-loday-l. These "aerobic" rates are too small to be significant in the N economy of agricultural soils. Yet it is interesting that N20 evolution was observed from virtually all of the well-aerated soils examined. I 61 have not yet demonstrated what fraction of this N20 results from biological denitrification; several other mechanisms are possible. Bremner and Blackmer (1978) have suggested that nitrification causes low rates of N20 production in aerobic soils, presumably by a process observed earlier in Nitrosomonas cultures (Yoshida and Alexander,l970). Non—biological reactions of NO2 also might contribute to N 20 produc— tion. It is also possible that the soil samples were super-saturated with N20. Denitrifying conditions in the field well before sampling might have caused accumulations of N 0 within aggregates which slowly 2 equilibrate with the atmosphere. The relative contribution of each of these mechanisms will be difficult to determine. Preliminary ex- periments were not revealing due in part to variability within and among soil samples, and in part to the difficulty of experimentally isolating the various mechanisms of N 0 production. 2 In summary, nitrogen gases can be produced more or less contin- uously by soils. Several mechanisms could contribute to this, but only denitrification in very wet soils is likely to result in large N losses. Subsequent to a reduction in soil aeration, denitrification rate is greatly increased, but only after a lag period of several hours. Anaerobic or partially anaerobic conditions, established by respiration and reduced oxygen diffusion rate, eliminate 0 inhibition 2 then derepress the synthesis of denitrifying enzymes. Most nitrogen loss would occur during brief periods beginning a few hours after irrigation or a rainfall and lasting until NO - is depleted or the 3 soil water content decreases. 10. ll. 12. LITERATURE CITED Ardakani, M. S., H. Fluehler, and A. D. McLaren. 1977. Rates of nitrate uptake with sudangrass and microbial reduction in a field. Soil Sci. Soc. Am. J. 41, 751-757. Balasubramanian, V., and Y. Kanehiro. 1976. Denitrification potential and pattern of gaseous N loss in tropical Hawaiian soils. Trop. Agric. 53, 293-303. Bremner, J. M., and A. M. Blackmer. 1978. Nitrous oxide: Emis— sion from soils during nitrification of fertilizer nitrogen. Science 199(4326):295-296. Bremner, J. M. and K. Shaw. 1958. Denitrification in soil. II. Factors affecting denitrification. J. Agric. Sci. 51, 40-52. Braodbent, F. E. and F. E. Clark. 1965. Denitrification. In Soil Nitrogen. W. V. Bartholomew and F. E. Clark (eds.). American Society of Agronomy. Focht, D. D. and W. Verstraete. 1977. Biochemical ecology of nitrification and denitrification. In Advances in Microbial Ecology. M. Alexander (ed.). Plenum Press. Payne, W. J. 1973. Reduction of nitrogenous oxides by micro- organisms. Bacteriol. Rev. 37:409-452. Pilot, L., and W. H. Patrick, Jr. 1972. Nitrate reduction in soils: Effect of soil moisture tension. Soil Sci. 114, 312—316. Payne, W. J. and P. S. Riley. 1969. Suppression by nitrate of enzymatic reduction of nitric oxide. Proc. Soc. Exptl. Biol. and Med. 132(1): 258-260. Rowe, R., R. Todd and J. Waide. 1977. Microtechnique for most- probable-number analysis. Appl. En. Micro. 33(3):675-680. Smith, M. S., M. K. Firestone and J. M. Tiedje. 1978. The acety- lene inhibition method for short-tegm measurement of soil denitri- fication and its evaluation using N. Soil Sci. Soc. Am. J. (in press). Starr, J. L., F. E. Broadbent and D. R. Nielsen. 1974. Nitrogen transformations during continuous leaching. Soil Sci. Soc. Am. Proc. 38, 283-289. 62 13. 14. 63 Tiedje, J. M., M. K. Firestone, M. S. Smith, M. R. Betlach and R. B. Firestone. 1978. Shor -term measurement of denitrifi- cation rates in soils using N and acetylene inhibition methods. ‘13 Proceedings International Symposium on Microbial Ecology, Springer-Verlag (in press). Yoshida, T. and M. Alexander. 1970. Nitrous oxide formation by Nitrosomonas europaea and heterotrophic microorganisms. Soil Sci. Soc. Am. Proc. 34:880-882. CHAPTER IV DENITRIFICATION IN THE RHIZOSPHERE It is generally believed that increased available organic matter in the rhizosphere stimulates microbial activity, reduces 0 concen- 2 trations, and thereby increases soil denitrification. This belief is . based on numerous studies. In some of these (Woldendorp, 1962; Brar, 1972; Volz t 1., 1976) denitrification has been measured indirectly, usually by NO -disappearance. These methods are generally insensitive 3 - + and may be confounded by non-assimilatory reduction of N03 to NH4 Stanford t 1., 1975b), plant uptake of N03-, or respiratory re- duction of N03- only to N02-. has been required in other studies (Bailey, 1976; Brar, 1972; Garcia, Perturbation of the plant-soil system 1975). Removal of the shoot from the roots, the roots from the soil, or any alteration in the soil matrix would be expected to cause drastic changes in nutrient supply, soil metabolism, nitrogen sinks, and gas exchange. Only Stefanson (1970, 1972a, 1972b) has directly measured denitrification products from soil with intact plants. He did long- and N 0 in elaborate sealed chambers. A consis- term ana1y31s of N2 2 tent plant stimulation of denitrification was observed only under certain conditions: water contents near field capacity and N applied as N03“ rather than NH4+. The complexity of this equipment made replication difficult and limited manipulation of plant and soil variables. It can be concluded from previous work that the rhizosphere is potentially a very important site for denitrification. Yet none of the methods previously used appeared promising for further investi- 64 65 gation of this relationship. This chapter presents the results of several new approaches to the study of denitrification in the rhizo- sphere. MATERIALS AND METHODS Soil treatments: The Miami sandy loam, Brookston loam, and Spinks loamy sand have been characterized in chapter 2. Corn, oats and orchard grass were grown in these soils either in the greenhouse or a growth chamber. In most experiments plants were grown in 125 g soil in styrofoam cups (9 02) with a small hole in the bottom for drainage. Equal water and fertilizer was applied to soils with and without plants. Soils were treated with 1 mg K HPO4/10 g soil at the time of 2 planting. The amount of N added varied from one experiment to another. Treatments referred to as low N03- received no N amendment. Most high NO treatments were leached with 100 ppm KNO 3 -N at planting and then 3 at monthly intervals for the grass, weekly for the corn. However, Brookston high NO3 soils received 1 mg KNO3-N/10 g soil at planting and were leached with N03- solution only before the aerobic assays. After some experiments, soil NO --N was determined with an Orion N0 - 3 3 electrode, following extraction with a 0.01 M CaSO 1 mM AgSO 4’ 4 solution. In most experiments I compared planted and unplanted soil, but in some cases I made a gross separation of rhizosphere soil from the non- rhizosphere soil in the planted cups. This was done by shaking the roots lightly to remove non-rhizosphere soil, then vigorously agi- tating to collect rhizosphere soil. 66 Fresh soil samples were 0 to 15 cm cores (5 cm diameter) collec- ted from the Miami soil planted to corn. Subsamples were immediately assayed as anaerobic slurries for denitrification rate. Split-plate experiments: This apparatus was based on a soil chamber designed for the study of endomycorrhizal fungi (Hattingh E£.fll" 1973). My modification consisted of a small (5.4 cm diameter) plastic petri dish glued in the center of a larger (8.8 cm) one (Figure 8). A portion of both sides of the inner dish was cut away and covered by nylon mesh cloth (30 um openings). The chamber was filled with Spinks soil, and both dishes were tightly covered. A germinated seed was planted through a hole in the side of the outer plate and the chamber was placed on its end allowing the roots to grow over the outer surface of the nylon mesh but not enter the inner dish. After one month the chambers were opened and those with roots in the center were discarded. Soil samples were carefully removed from the outside plate (root zone) and from the inside chamber at 5 mm incre- ments from the root zone. Anaerobic assays: The method of assaying denitrification anaerobic- ally has previously been described in detail (chapter 2). Briefly: soils were made into a slurry and the appropriate substrates added. Serum bottles (25 ml) were used for the split-plate experiments because of the small amount of soil available. In all other experi- ments 125 m1 Erlenmeyer flasks were used. The incubation vessels were twice evacuated for 5 min and flushed with He to achieve anaerobiosis. C H2 (0.1 atm) was added to inhibit reduction of N O to N A micro- 2 2 2° thermistor detector was employed for N20 analysis. 67 Figure 8. Split-plate apparatus used to determine spatial relationship between roots and denitrifying activity. Note that roots are (restricted to the outer chamber. 68 69 Aerobic assays: Corn (2 plants per cup) was grown for 9 to 13 days and orchardgrass (4 plants per cup) for approximately 4 months in styrofoam cups. Three days before the assay the soil was briefly 3 for high NO3 . When C2H2 was used, the cups were placed in a dessi— leached with a KNO solution; 2 ppm-N for low N0 3 treatments, 100 ppm cator containing 3% CZHZ for 12 hours before assaying for denitrifi- cation. At the beginning of the assay the soil was again briefly 3 solution. C2H2 was bubbled through the leaching solution before the assay. Cups were allowed to leached with 80 ml of the appropriate N0 drain in the dessicator for 1/2 to 1 hour. Thus the soil water content was reasonably uniform, approximating the maximum water holding capacity. A glass tube (7 cm x 5 mm i.d.) packed with Asca- rite was taped to the side of each cup. This kept the CO concen- 2 tration low, but not below ambient, during the analysis and reduced interference with N20 analysis. Cups, with plants and soil still intact were then placed in 30 x 15 cm Saran gasbags (Anspec, Ann Arbor, Mi.) which had an inflated volume of approximately 1 1 (Figure 9). The bags are manufactured with a tube, sealed by a serum stopper, which allowed introduction and sampling of gases. I modified the bags, cutting open the bottom, so the plant and cup could be placed inside. The bottom was resealed with a screw clamp made of wood and rubber. Enough 02H2 was immediately injected into the bag to give approximately a 5% concentration by volume. The volume of the gas space in the bag was determined by injecting 10 m1 He then measuring the He concentration after 90 min with a microthermistor detector. In some experiments the rate of gas loss from the bag was measured by following the concentration change of introduced He or Ne. A 63Ni .1lll..vl. [ill l. Figure 9. 70 Saran gasbag used to measure denitrification rate of soils with and without intact plants. 71 72 electron capture detector was used for N 0 analysis in these experi- 2 ments (chapter 2). Enumeration of soil bacteria: Twenty g soil samples were blended for l min in 180 ml 0.85% NaCl with 1 drop of Tween 80. Dilutions were plated on nutrient agar (Difco, Detroit, Mi.) and grown at 300C. The counts on these plates after 3 days were considered to be a measure of total aerobes. Denitrifiers were enumerated by a most probable number procedure with 10 fold dilutions and 5 tubes per dilution. I used Hungate tubes (initially aerobic) containing 3.5 mM KNO in nutrient broth (Difco). 3 - and NO - was determined with After 7 days, disappearance of N03 2 diphenylamine (Appendix A). RESULTS Anaerobic assays: The initial denitrification rate of soil which adheres to corn roots (rhizosphere) was greater than non-rhizosphere soil in the same cups (Figure 10). Soil from unplanted cups had the was 3 depleted, first in the rhizosphere soil, then in the planted soil. lowest rate. After this brief initial period endogenous soil NO There was no indication of N03- depletion from the unplanted soil. Only after 5 hours of anaerobic incubation were the soils amended with N03- (1 ml of 0.1 M KNO3). After the addition of N03- the initial order of denitrification rates was quickly reestablished (Figure 80). i.e., rhizosphere greater than planted greater than unplanted. This result implies that competing NO3 sinks, presumably plant uptake, may reduce rhizosphere denitrification even though the potential for 73 Figure 10. Denitrification by non-rhizosphere planted, unplanted, and rhiZQSphere Brookston soil assayed anaerobically. Soils were amended with NO3 only after 5 hours of incubation. Points are means of 3 replicates. 74 Any mewu eoepmnaocfi \\ . of 4 mm U m w a S N . 820 m. D. x J opocemonwculao: M .cmucmPe . mv m. 4‘ . oz+ .MV 1 ms cmucmFQc: memeamo~wce i ooH 75 denitrification is greater in the rhizosphere. As was the case in previous work (chapter 3) denitrification rates became approximately linear (phase II) after an initial period (phase I). I have suggested that the ratio of N20 produced in the absence of C2H2 to N20 produced with C2H2 is a valid approximation of the ratio NZO/(N2 + N20). This approach was used to compare products from unplanted and planted (corn) soils (Table 9). In this, and all of the following experiments, planted soil refers to all of the soil in the cups with plants. The percentage of N O was at all times higher 2 in the unplanted soil. In the planted soil N20 reached a maximum concentration at about 10 h, then remained constant in the flasks with C2H2 but declined to 0 at 18 hours in the flasks with no C2H2' In the unplanted soil with no CZHZ’ N20 reached a maximum at 35 h then slowly declined. As in the previous experiment, total N20 production was higher (both with and without C2H2) from the planted soil than the unplanted. Addition of glucose to soils just before assaying increased the denitrification rate of both planted (corn) and unplanted Brookston soil (Figure 11). All soils were also amended with 10 ppm NO3--N (soil fresh wt. basis). Succinate slightly increased the rate of unplanted soil but not of planted soil. Reduced stimulation or in- hibition was observed at high concentrations of added carbon. Sig- nificantly, amendment with both carbon sources reduced the difference between planted and unplanted soils. This indicates that at least part of the rate increase in planted soil may be due to supply of available carbon. However, both soils are apparently limited by the supply of electron donor under the assay conditions used, since 76 .szo :uHB mum» uo>o Nzwo unocufi3 cowuoawOHQ oNz mo oumm « om.o c on oq.o o NN mq.o oo.o cg we.o om.o om mq.o ~q.o o mm.o om.o N musos voocmfiacz vmucwam mafia coaumnaocH .AoNz + sz\oNz .mucmam usocuma vow nuaz Hwom coumeOpm ownooomcm >9 Omz mm bo>ao>w mww 2 mo cofiuuom no manmh Figure 11. 77 Effect of glucose and succinate amendment on the phase II denitrification rate of planted and unplanted Brookston soil. Rates have been normalized relative to the rateaof unamended, planted soil. There 5 were 3 replicates per treatment. .0' 78 Aawom sow z\3v sauna moatomssm a m.H o.H m.o - d 4 elil mumewoosm+umpempecs mumcwoozm+wmuempa mmoospo+umucmpaea mmoo:_m+cmoempa om owfi [10$ paiueld ‘papuaweun lo 319i lo % 79 glucose amendment increased the rate of planted as well as unplanted soil. The relationship between the presence of plants and anaerobic denitrification rate was examined with fresh soil samples. The two experiments shown (Table 10) were conducted at different times and with soil from different plots so no valid comparison can be made between the experiments. In experiment 1 soil between corn plants in a row was compared to soil from a border strip which had received the same fertilizer and water treatment. Soil was taken at the row and at various positions between rows in experiment 2. Samples were amended with 10 ppm NO3--N and phase II rate determined. In both experi- ments, soil closest to the corn plants denitrified most rapidly during the anaerobic assay. The spatial relationship between roots and soil denitrifying activity can be defined more precisely with the split-plate technique. Figure 12 shows the denitrification rate 200 to 400 min after the imposition of anaerobiosis. Nitrate (10 ppm) was added to all. The soil from the root zone (the outer chamber) denitrified more than twice as fast as any of the samples from the inner chamber. A gradual decrease in potential rate was observed with distance from the roots. The correlation coefficient of rate with distance for the samples from the inner chamber was -0.54, significant at the .05 level. Because of the design of the chamber, soil moisture was also negatively cor- related with distance from the root. Although the difference was small, 17.5% gravimetric water content in the root zone vs. 14.7% at the center, this may have contributed to the observed differences in denitrification rate independently of a plant effect. 80 mo. n.l um+ an NH “.49 om.o N “.00 ms.o N “.mm o N cm.“ OHS m ion.“ ooN o a N . c E. om my a. mo 08 s- a sums s o z N NNuoH e womb sou some ucmEHuoaxm cowumommwuuwcmc mocmumwa .ofinoummcm moo: mzmmmm one .30» Chou m Boom mmocmumfip m:0«um> um vaoEmm Hfiom New“: «0 mumu cofiumofimauuwcwa nos «News Figure 12. 81 Denitrifying activity related to distance from oat roots. Spinks soil was taken from split-plates. Rates were determined 200 to 400 min after the imposition of anaerobiosis. There were 3 replicates per treatment. 7'150 ” -1 200 ’ omin 0:9 dry soil 0 O I pmoles N2 50 L 82 1 I I root zone 10 15 Distance from roots (mm) 20 25 83 The number of isolatable aerobes and denitrifiers from several soils is shown in Table 11. These soils had been treated in the same manner as the soils used in the aerobic assays, they were kept in cups in a growth chamber. In 4 of 5 cases both aerobes and denitrifiers were more numerous in the soils with plants, although these differ- ences were generally not statistically significant because of the large error inherent in MPN procedures (about 1 order of magnitude). The ratio of denitrifiers to aerobes was, in every case, higher in the planted soil. This implies that the presence of the plants specific- ally enriched for denitrifiers relative to aerobes. The time course of N20 production by saturated Brookston soil, with high N03- and no plants is shown in Figure 13. After saturating the soil with water,the cups were placed in the Saran gasbags and 5% C2H2 was added. A lag period of about 4 hours was observed, consis- tent with earlier results (chapter 3). I have attributed this lag to the time required to consume soil 0 and to derepress the synthesis of 2 denitrifying enzymes. Between 8 and 11 hours the maximum rate of N20 evolution was attained. The 02 concentration in the headspace of the bags was only decreased by about 3% after 24 hours. The concen- tration of 20 ml Neon injected into the bags decreased by about 0.5%/hour, suggesting that leakage of evolved gases from the bags was slow. It was also observed that the denitrification rates of replicates of a treatment did not fit a normal distribution. Therefore the non- parametric Mann-Whitney test was used to determine statistical dif- ferences between treatments (Snedecor and Cochran, 1967). 84 Table 11: Number of denitrifiers in planted and unplanted soils. Soil Plant Aerobes Denitrifiers Denitrifiers Aerobes -1og number-g-l- Brookst9n* Corn 7.5 6.1 .033 (10“ N03) None 6.9 5.1 .015 Brookston Corn 7.7 6.5 .077 (high N03) None 7.6 4.5 .001 Brookston Grass 8.0 6.6 .037 (high N03) None 7.5 5.1 .004 Miami _ Grass 7.7 5.4 .006 (high N03) None 7.9 5.5 .005 Miami _ Corn 8.2 6.2 .011 (low N03) None 7.1 4.6 .005 Refer to text for complete description of NO3 treatments . 85 Figure 13. Time course of N20 production by unplanted 3 Brookston soil with C2H2 in aerobic gasbags. Points are means of 5 high N0 replicates. w nmoles NZO-nfl-1 86 400 800 Time after saturating 1000 (min) 1200 87 Generally experiments were conducted only for an 8 to 10 hour period after saturation of the soils. This time was sufficient for differences between the treatments to develop yet leakage of N20 from the bag was insignificant, and O depletion and accumulation of pro- 2 duct gases in the headspace was minimal. The effects of N03- concentration and plants on denitrification rate of two soils is shown in Table 12. With high NO 3 soils denitrified faster than the unplanted. Yet in the low N03- treatments the planted soils denitrified at the slowest rates. This the planted suggests that plant uptake can compete with denitrifiers for N03 . The high N03- treatments denitrified faster than the low N03- treatments, except for the unplanted Brookston. This negative correlation of NO 3 concentration and denitrification rate was repeated in separate experiments (data not shown) with unplanted Brookston. The ratio of N20 produced in bags without C2H2 to N20 produced in bags with CZH2 is an approximation of the fraction of total gas evolved as N20, that is, NZO/(N2 + N20). The fractions obtained for the Brookston and Miami soils are shown in Table 13. Most apparent are the large range of values, from 0 to k of the gas is released as N20. The ratios for the Brookston were always higher than the com- parable treatment of the Miami. The concentration of NO — also had a 3 consistent effect. At low N03- a smaller fraction of the gas was N20. After about 8 hours, there was no net production of N20 by the low NO3- Miami soils without C2H2. At this time, the same soils blocked with C2112 continued to evolve N20 at an undiminished rate. The presence or absence of plants was not consistently related to the ratio. 88 .ucmeummuu moz mo cofiumfiuommv mumHano wow axon mom § .umou zoau«:3Iccmz mnu up aw>wa Nm ecu um ucmwowwfiv sauchNMNcwfim uoa mum umuuma wamw emu sows mmumm .muso: w HON mumu owmum>< + .2 mm .mfimmn cowusaom Hfiom « mmmuwvumnouo smcaanmsme no: mow.o som.sN swam aoumxoonm m.~ N.~ mma.a amm.o 304 mmmuwvumnouo 0.0m o.mN mme.N umN.ON ems: News: c.m q.~ qu.H an.o 304 ouoo N.mN~ ~.wo~ nmm.c mmm.m awe: acumxooum IIIII.EQQIIIIII. Id CHE.H wan.o~z meoECI « I I woucmfimc: woman—e boucmamsa voucmam acmEWmouH ucmam a I oz w Hwom wousmmoz moz +wumu coHumonwwuficoa .mwwn mmw ownouom cm vo>mmmm mawom wmucmaac: new vmmcmHe mo sump coaumowmauuwcoo .Nfi wanme 89 . mquEumwhu umuum musoc m wow u c003umn .N m oz mo 0040m004axo MOM uxou now + .wooMHMumm 0H03 m440m o nu43 mums um>0 0 unonn43 some I o o 304 mmmumvumnouo so.o No.0 ewfiz New“: no.0 44.o 304 cuoo mq.o mm.o sw4: commxooum voucm4eca voucm4m ucmEumouH Imoz ocm4ml44om IANz + osz\oNz .mumnmmw 0440umm :4 m440m >4 0 N 2 mm 00>40>m mom 2 m0 coaunoa .ms mHBme 90 DISCUSSION I have used several new methods to demonstrate that roots may increase denitrification rates. One approach was to determine the denitrification rate of soil slurries under totally anaerobic condi— tions. A rate increase was observed in soils from planted pots. The rate was also greater in rhizosphere soil than in non-rhizosphere soil from the same pots. Fresh field soils sampled near corn rows denitri- fied faster than soil at a distance from the row. The split-plate experiments revealed the spatial distribution of potential denitri- fying activity relative to roots. The activity decreased very rapidly in the first few mm from the roots then declined slowly with increas- ing distance. Because I added N03- to the soil in these experiments, the differences observed can be attributed primarily to variation in the number of denitrifiers and the supply of energy. This approach allows these important variables to be isolated and their effect determined independently of diffusion limitations, aeration state, and N03- limitation. However, because direct effects of the latter set of variables have been eliminated these results cannot be directly related to field denitrification rates. The number of isolatable denitrifiers was generally greater in planted than in unplanted soils. Yet this increase does not necessar- ily imply an increase in denitrification since denitrifiers are facul- tative aerobes. Their number can increase through respiration of 02 as well as of N oxides. I also observed an increase in the ratio of denitrifiers to plate-count aerobes in planted soils, which is better evidence that denitrifying activity is enhanced. The relative 91 enrichment of a specific group of microorganisms should in general be better related to microbial activity than absolute numbers. However, it is possible that another mechanism could be responsible for the relative increase in number of denitrifiers in planted soils. For example, the substrate range of the pseudomonads (Bergey's Manual, 1974) corresponds fairly well to the compounds exuded by plant roots (Rovira, 1965). The results with intact plants and soil in the aerobic gas bags are clearly best related to field denitrification rates. The physical conditions in the soil, the spatial relationship between roots and denitrifiers, and the sources and sinks of substrates are probably not significantly different from field soils under denitrifying conditions. 0f previous methods, only Stefanson's (1970) satisfied these condi- tions. The method presented here requires a less complex incubation chamber and allows greater replication and manipulation of variables. Although all of these experiments have been conducted with saturated soils, it should be possible to study the effect of soil water content by slightly modifying the procedures. The slow diffusion rate of C2H2 in soils must be considered in the design of these experiments, however. Long-term assays may increase the statistical uniformity of results and eliminate the problem of integrating rates to approximate total annual denitrification loss but they are time-consuming and greatly reduce the number of experiments possible. I have previously presented evidence suggesting that most denitrification occurs during brief periods immediately after soil wetting (chapter 3). An approximation of denitrification rate in the field can be made 92 by assuming that the flux of N 0 from the soil in the gasbags is equal 2 to the flux of N gases from the same soil in the field. The highest 2 3 Brookston with orchardgrass between 6% and 8 hours after saturation. rate I observed was 36 nmoles N O.bag-1-min-l, from the high NO This is extrapolated to 4 kg N-ha-l-day-l. Therefore, large quanti- ties of N may be denitrified during a brief period after irrigation or a rainstorm. When NO3- concentrations were high, my results conformed to the prevailing opinion that roots increase denitrification rate. This was true of the Miami soil with orchardgrass, the Brookston soil with corn, and the Brookston soil with orchardgrass. Although I observed higher rates with orchardgrass than with corn, this may be due simply to the longer growth period of the orchardgrass. These methods will make it possible to examine differences between plant species or varieties. In one anaerobic assay not presented here, there were no significant differences between the anaerobic phase II (potential) rates of Miami soil from plots of corn, orchardgrass, or alfalfa. Stefanson (1972a) observed a consistent increase in denitrifi- cation by planted soils only near field capacity. However, plant growth was greatly reduced by long-term exposure to saturated or near- saturated conditions. My short-term assays demonstrate that plants may also stimulate denitrification during the brief but critical period after soil is saturated by rain or irrigation. The results provide some indications about the mechanism by which roots may increase denitrification rates. First, an increase in number of denitrifiers was observed in planted soils. Second, although both planted and unplanted soils were shown to be carbon-limited under 93 totally anaerobic conditions when N03- was non-limiting, carbon amend- ment decreased the difference in rate between them. This suggests that part of the difference is due to a lower supply of energy source in the unplanted soils. The planted soils denitrified no more than 4 times as fast as unplanted soils in the anaerobic assays but 4 to 28 times as fast in the aerobic gas bags. This is indirect evidence for the involvement of a third factor, reduced 02 concentration in the rhizosphere of intact soils due to increased oxidation of available organic matter. All of these mechanisms have been considered previ- ously (Stefanson, 1972a; Woldendorp, 1962). I would also suggest 3 to the rhizosphere in the transpiration stream increases the supply of electron acceptor to the another possibility, that mass flow of NO denitrifiers. In spite of these considerations, I believe that denitrification will actually be reduced by the presence of roots under some condi— tions. This is supported by the rapid depletion of NO - by the rhizo- 3 was available to the 3 denitrifiers. Concentrations of N03- were consistently lower in the planted soils. Most convincing are the very low denitrification rates sphere soil in the anaerobic assay; less NO observed in the aerobic assays with the low N0 -, planted soils. More N03- may have been depleted from these soils by enhanced denitrifica- tion prior to the assay. It seems more likely that plant uptake was the important N03- sink. Therefore, competition between plants and denitrifiers is believed to be of great importance in determining the fate of soil NO3-. The long-term result of this competition is pro- bably best determined by long-term studies, preferably with 15N, rather than the 8 to 10 hour experiments I have used. It is 94 interesting in this regard that when Stefanson (1972b) amended soil with NH +-N, rather than NO --N long—term denitrification losses were 4 3 in some cases significantly greater from soils without plants. Nitrate concentration was thus seen to interact in a complex manner with the presence of roots. When roots were present, presum— ably meaning that energy supply was not the primary limitation, in- creased NO3- increased the denitrification rate. This relationship has been observed by others (Starr and Parlange, 1975; Stanford gt al., 1975b). When N03- was not limiting, the plants caused a con- sistent rate increase. In the absence of roots the rate was limited 3 tration could be established. High N03- did cause a reproducible by carbon and no simple relationship between rate and N0 concen- decrease in the evolution of N gas from the unplanted Brookston soil. It is possible that in this instance the reduction of N03- to N02- was able to satisfy much of the small requirement for electron acceptor. This hypothesis is consistent with observations that pure cultures of some denitrifiers accumulate NO - before producing N gas and that NO - 2 3 inhibits reduction of NO to N20 and N2 (Payne and Riley, 1969). The concentration of N03- did affect the percentage of gas evolved as N20, as determined by a comparison of uninhibited and CZHZ inhibited soils in the aerobic gasbags. In every case, a greater 3 with our earlier results (chapter 2 and Firestone E£.§l°’ 1977). It percentage appeared as N O in the high NO treatments. This agrees 2 is believed that N03- or N02- may directly inhibit the reduction of N20 to N2 or alternatively, that the reduction of NO3-to N02- and N20 competes for electrons with the reduction of N20. I expected that the presence of plants might also be related to 95 the ratio of NZO/(NZ + N20). It was believed that the increased energy supply in planted soils would increase the demand for electron acceptor and drive the denitrification reactions more towards com- pletion. This was, in fact, observed in the anaerobic experiments and in the high NO -Brookston soil assayed aerobically. However, the 3 ratio was slightly lower for the unplanted soils in the other aerobic experiments. The large variation I have observed in the ratio of NZO/(N2 + N20) is noteworthy. From half to none of the gas was evolved as N20. This emphasizes that current estimates of N O flux to the atmosphere 2 must be considered tentative and subject to large errors. The ratio of N20 to N2 is a critical value in determining the effect of in- creased fertilizer use on stratospheric ozone stability (Johnston, 1977). In summary, the total effect of the rhizosphere on denitrifica- 3 ditions and stage of plant growth. The N03- concentration and the presence of roots interact; simple independent relationships between tion depends on the concentration of NO and probably on soil con- these variables and the denitrification rate or the ratio of NZO/(N2 + N20) should not be expected. The potential for denitri- 3 centrations this potential will be reflected in the actual denitri- fication is clearly greater in the rhizosphere. At high NO con- 3 is limiting plant uptake will compete with denitrifiers for N03 , thus, denitrification can actually fication rate. However, when NO be reduced in the rhizosphere. 10. ll. 12. 13. LITERATURE CITED Bailey, L. D. 1976. Effects of temperature and root on denitri- fication in a soil. Can. J. Soil Sci. 56:79-87. Brar, S. S. 1972. Influence of roots on denitrification. Plant and Soil 17(2):267-270. Buchanan, R. E. and N. E. Gibbons. 1974. Bergey's Manual of Determinative Bacteriology. Williams and Wilkins Co. Baltimore. Firestone, M. K., M. S. Smith, R. B. Firestone, and J. M. Tiedje. 1977. Factors influencing biologigal release of N20 during denitrification as determined by N methodology. Agronomy Abstracts. Garcia, J. L. 1975. Effet rhizosphere du riz sur la denitri- fication. Soil Biol. Biochem. 7:139-141. Hattingh. M. J., L. B Gray, and J. W. Gerdemann. 1973. Uptake and translocation of P—labeled phosphate to onion roots by endomycorrhizal fungi. Soil Science 116:383-387. Johnston, H. S. 1977. Analysis of the independent variables in the perturbation of stratospheric ozone by nitrogen fertilizer. J. Geophysical Research 82:1767-1772. Payne, W. J. and P. S. Riley. 1969. Suppression by nitrate of enzymatic reduction of nitric oxide. Proc. Soc. Exp. Biol. Med. 132:258-260. Rovira, A. D. 1965. Plant root exudates and their influence upon soil microorganisms. In Ecology of Soil-Borne Plant Pathogens. K. F. Baker and W. C. Snyder (eds). University of California Press. Snedcor, G. W. and W. G. Cochran. 1967. Statistical Methods. pp. 128-131. Iowa State University Press. Stanford, G., R. A. Vander P01, and S. Dzienia. 1975a. Potential denitrification rates in relation to total extractable soil carbon. Soil Sci. Soc. Am. Proc. 39:284-289. Stanford, G., J. 0. Legg, S. Dzienia, E. C. Simpson. 1975b. Denitrification and associated nitrogen transformations in soils. Soil Science 120(2):147-152. Starr, J. L. and J. Y. Parlange. 1975. Non-linear denitrifi- cation kinetics with continuous flow in soils. Soil Sci. Soc. 96 14. 15. 16. l7. 18. 97 Stefanson, R. C. and D. J. Greenland. 1970. Measurement of nitrogen and nitrous oxide evolution from soil-plant systems using sealed growth chambers. Soil Science 109(3):203-206. Stefanson, R. C. 1972a. Soil denitrification in sealed soil-plant systems. I. Effect of plants, soil water content and soil organic matter content. Plant and Soil 33:113-127. Stefanson, R. C. 1972b. Soil denitrification in sealed soil- plant systems. 11. Effect of soil water content and form of applied nitrogen. Plant and Soil 37:129-140. Volz, M. G., M. S. Ardakani, R. K. Schulz, L. H. Stolzy, and A. D. McLaren. 1976. Soil nitrate loss during irrigation: Enhancement by plant roots. Agronomy J. 68:621-627. Woldendorp, J.W. 1962. The quantitative influence of the rhizo- sphere on denitrification. Plant and Soil l7(2):267-270. APPENDICES APPENDIX A ENUMERATION OF SOIL DENITRIFIERS Several published methods of enumerating soil bacteria and original modifications of these methods have been evaluated. This is a summary of research conducted by Bill and Nancy Caskey and myself, to be published in full at a later date. Although numbers of denitrifiers cannot be directly related to soil denitrification rate (because denitrifiers can also grow by respiration of 02) enumeration of denitrifiers is often of interest as related to other measurements of soil denitrification. Difficulty in our laboratory with commonly used methods led to the initiation of this study. The most widely used method of determining numbers of soil denitrifiers is probably that of Focht and Joseph (1973). They used an MPN procedure with 9.9 mM NO3 in nutrient broth. We compared MPNs determined in this manner to MPNs determined in nutrient broth 3 concentration increased the recovery of denitrifiers. It is be- but with 3.5 mM NO3-. Table 14 shows that decreasing the NO 3 more completely reduced by denitrifiers and observation of their lieved that the smaller quantities of N0 were more rapidly and activity was less affected by competition or antagonism by anaer— obes. Volz (1977) has suggested that N02 is a more specific electron acceptor for denitrifiers than is NO3 and should therefore be used in MPN procedures for these microorganisms. We encountered numerous difficulties wtih this approach, due in part to the apparent produc- tion of inhibitory products when N02 is autoclaved in the closed 98 99 Hungate tubes we used. An additional problem with Volz's method is that the high concentrations of N02 which he used (Table 15). In the 2 tubes after autoclaving. In this case, 3.5 mM N02- and 3.5 mM N03- gave MPNs which were not consistently or significantly different. experiments shown NO was sterilized separately and added to the The MPNs were considerably reduced in 7.2 mM NO2 which was the lowest concentration used by Volz. Table 16 shows the MPNs obtained when soil dilutions were inoculated into initially aerobic, but closed Hungate tubes, com- pared to truly anaerobic tubes. MPNs were significantly reduced in the anaerobic tubes. Thioglycollate (0.5%) was used as a reducing agent in these experiments. This concentration did not have a significant inhibitory effect on the growth of denitrifiers in pure culture. Furthermore, the same effect was observed in anaerobic tubes using titanium citrate as a reductant and in tubes, without a reductant, made anaerobic by flushing and evacuating with He. This result suggests that many soil denitrifiers are not able to grow when switched abruptly from aerobic to anaerobic conditions, presum- ably because no mechanism of energy generation is available for the synthesis of denitrifying enzymes. Denitrifiers which are already partially derepressed for denitrification would, of course, be able to continue growth when switched to total anaerobiosis. MPNs obtained using tryptic soy broth were not significantly different from those determined with nutrient broth. Tryptic soy broth has been said to be the equivalent of soil extract media in the recovery of maximum numbers of aerobes from soil (Martin, 1975). Patriquin and Knowles (1974) have suggested that the presence 100 of N20 be used a test for the presence of denitrifiers rather than the disappearance of N03_ and N02“. Using this approach, we found that some tubes from which all of the N03- and N02- had disappeared did not contain detectable amounts of N20. This problem could be 2H2 to prevent the reduction of N20 to N2. Under these conditions, significantly higher MPNs were determined eliminated by adding C by N20 detection than by absence of N03- and N02- (Table 17). That is, many more tubes showed production of N20 than complete reduc- tion of NO - and NO -. However, the validity of this approach was 3 2 made questionable by the observation that NO3--respiring organisms such as E, coli, not normally associated with denitrification, produce significant amounts of N20. This reaction appears to be 2 incubation period of lor 2 weeks (Table 18). Because the signifi- too slow, however, to completely reduce all N0 during an.MPNuaFr cance of N20 production by these bacteria in soil is unknown, N20 production was rejected as a means of detecting the presence of deni- trifiers. Evolution of N20 by E. ggli appears to be dependent on enzyme activity. Autoclaved cell suspensions do not evolve N20 from NO3-or NOZ- (Table 18). N20 is produced only near the end of log phase growth and production is terminated by the addition of arsenite, suggesting that chemical decomposition of accumulated N02- is not responsible (data not shown). Recovery of denitrifiers was further evaluated by adding a rifampicin resistant (see Appendix B) strain of Pseudomonas fluorescens to soil. After 15 min, recovery from the soil was deter- mined by inoculating soil dilutions to MPN tubes containing 3.5 mM 101 NO3-, nutrient broth, and 50 ug/ml rifampicin. Dilutions of the inoculum were also plated directly onto rifampicin—containing agar. The number recovered from the soil was not significantly different from the number determined to be in the inoculum. We conclude that 3.5 mM N03- or N02- with initially aerobic conditions gives excellent recovery of soil denitrifiers and is the best available method for enumeration of them. LITERATURE CITED 1. Focht, D. D. and H. Joseph. 1973. An improved method for the enu- meration of denitrifying bacteria. Soil Sci. Soc. Am. Proc. 37:698-699. 2. Martin, J. K. 1975. Comparison of agar media for counts of viable soil bacteria. Soil Biol. Biochem. 7:401-402. 3. Patriquin, D. G. and R. Knowles. 1974. Denitrifying bacteria in some shallow-water sediments: enumeration and gas production. Can. J. Microbiol. 20;1037-1041. 4. Volz, M. G. 1977. Denitrifying bacteria can be enumerated in nitrite broth. Soil Sci. Soc. Am. J. 41:549-551. .0 0m mm3 musumuweewu .0640 cofiumnsocH « 102 qo.m mw.q um>ocoo 4m.m qm.m wasnucoux mm.~ oq.m coumxooum oo.m qm.m a Hemwz em.m mw.m o 4Em4z mw.m Nu.q m 45mwz mo.q wN.o < 4Em42 I4 w.mp04m4uuwcoo w04I Amsms «so .Amsme NV ze o.s :5 m.m 440m Luoun ucowuusc cw mumuuwz .mpm4w4uuwcmv mo 2e: co cowmmumcmocoo moz m0 uowuum .«4 mHAmH 103 .o On 00 mco4umnsoc4 men s .cmoun ucmwuusc 04 44< « Nm.m wo.m - MN.s -Noz :2 N.N mq.s mw.q em.o 00.4 -Noz ze m.m mw.s mc.m me.s Nm.s Imoz 2e m.m IH «.mumfimwuuwcov w04I m < um>ocoo wemwz GOumxooum 00mmxo0um Enabmz an 440m . N N m I oz cwfis was .I oz 304 . oz 304 :4 0004mu40 mzmz mo somfiumaeou .m4 04409 104 .0 0m om mxmb n owum40004 .Imoz SE m.m .ucmuusvmu mm mum44oom4wofinu Nm.o « oo.m mw.q um>0coo cm.m Nm.q commxooum m~.m mo.m 4804: I4 440m w.mu04wwuu4cmv w04I 0440p0wcw 0440u0m 440m NHHmfiowcH s44m404c4 .mu04w4uuwcwv m0 mzmz no woman mumwcsz L0 swoon :04umuom 4mwuwcw m0 uomwmm .04 044mb 105 .A>\>V Eon Om z4ouwsfix0uaem 04844 cowuomuoa a .mzmp m umumm vamp monsu .Luoun Imoz ZE o.o « mm.m m4.© cw.m 4 m©.N muonncmux 44.m CODmxooum m4.< 4504: I w.muwwwwuu4con w04I acmz INOZ bum Imoz 440m no mocmwoue uo mocmummoommwo k.mmnsu 23: Ca muwwuwuuwcmn m0 0000mmuo 0:0 ooh momma 03o wo comwumesoo .NH 044mb 106 Table 18. Production of N 0 by bacteria not commonly be ieved to be denitrifiers. Additions to nutrient broth %N recovered as N 0 after 7 days -# * N03 0 N02 0 autoclaved soil + N02 0 autoclaved E, coli + N02 0 _, coli 0 .E. coli + N03" (6.7)+ 8.4 E, coli + N02" 13.2 _, typhimurium + NO; (6.5) 13.2 .5. aerogenes + N03- (6.4) 3.3 # 3 3.5 mM NO ' and NO ' * Detection limit about 1 ppm (v/v) f final pH of medium. APPENDIX B A RIFAMPICIN RESISTANCE MARKER TO STUDY SOIL DENITRIFIERS I have begun to investigate the feasibility of using a rifampi- cin resistance marker to study the ecology of soil denitrifiers. The presumed advantages of this approach are that the marker would allow selective recovery of a specific denitrifier from soil. Rifampicin is suited for this application because it has little or no clinical usefulness and rifampicin resistance has generally been found to be a stable trait. Rifampicin inhibits DNA-dependent RNA synthesis in prokaryotes. (Wehrli and Staehelin, 1971; Weller and Saettler, 1978.) Naturally occurring resistant strains have been isolated from * pure cultures of the soil denitrifiers, Pseudomonas fluorescens and .P. aureofaciens? With 50 ug/ml rifampicin (Rifampin B, Calbiochem, 8 San Diego, Ca.) the frequency of resistance was 1/1.8 x 10 , in agreement with published values for other bacterial genera. Resis- tance was determined to be a stable trait by repeated transfers of the resistant strains in rifampicin-free media, followed by deter- mination of frequency of resistance. After 10 transfers the number of cells capable of growth on rifampicin plates was 70 to 100% of the number on rifampicin-free plates, i.e., no significant difference. The background of rifampicin resistant bacteria in soil was low; less than 102/g under anaerobic conditions: less than 103/g under aerobic conditions. Furthermore, these naturally occurring resistant organisms grew very slowly in the presence of rifampicin and could almost always be distinguished from the selected strain on the basis of colony size and morphology. It was necessary to include 50 ug/ml cycloheximide in the media to prevent the growth of soil fungi. 107 108 A preliminary experiment suggested that this technique would indeed be useful. A rifampicin resistant strain of P, fluorescens was added to soil at a density of 104 cells/g. The soil was satu- rated with water and after 6 days dilutions of the soil were plated on N03- agar containing rifampicin and cycloheximide. The plates were incubated anaerobically. P, fluorescens had increased to a density of 3.3 x 104/g of soil. Another experiment involving the use of a rifampicin resistant denitrifier has been described in Appendix A. These experiments in- dicate that the use of a rifampicin resistance marker will make it possible to precisely determine small changes in population size of a specific microorganism under natural soil conditions. Some results of these preliminary experiments are summarized in the table on the following page. LITERATURE CITED 1. Wehrli, W. and M. Strehelin. 1971. Action of the rifamycins. Bacteriological Reviews 35(3):290-309. 2. Weller, D. M. and A. W. Saettler. 1978. Rifampin-resistant Xanthomonas phaseoli var. fuscans and Xanthomonas phaseoli: Tools for field study of bean blight bacteria. Phytopathology (in press). P, fluorescens strain number 72, P, aureofaciens strain number 59. T Anaerobic incubations in glove box with N03- agar. 109 Table 19. Summary of results with rifampicin resistant denitrifiers. Frequency of resistant organisms in culture of P. fluorescens Stability of resistance, frequency of resistance after 10 transfers Naturally occurring resistance, background in soil aerobes anaerobes + denitrifiers Estimated minimum detectable frequency of an introduced resistant denitrifier in soil 1/1.8 x 10 70 to 8 100% cells/g cells/g cells/g