time-mama; 4mm . 1 mm: Av ‘ ,. u 1:31;... . : .5. .. 8.739.: 3:57....) V flatter: :4 i 2 v.- »A..i.agi rt... i!.u:xl§.\t.i . u ‘ 3..) o... Fall-‘5‘!!- Ié... l)‘ V I.) 3863!. 1.13;): .1 «lilti; $21.41;. ‘55 :»...».....§:..!. thllix‘... it! 1 ..)...¥!.L?¢ r: Villxrl: ~12 . 3.23.5.-.» tailllalvnkl) {ISIS-t .‘tluixecl. till? 1:531. .31. «5.31.3.1: r a .. .Ealxtilt In a..$....3);...l.x..; RSITY LIBRARIES \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\|\\\\\ \H \\\\\\\\| MICH' \ \\ This is to certify that the dissertation entitled Investigations into Turfgrass Black Layer presented by William Lee Berndt has been accepted towards fulfillment of the requirements for Ph - D - degree in We . 12/ 24W, Major pr essor Date 27 June, 1990 MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 EIBMRY Michigan State ‘ ‘ University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE mm JAN 0 8 1997 c:\cIrc‘\dateduoi pump. 1 INVESTIGATIONS INTO TURFGRASS BLACK LAYER by William Lee Berndt A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Botany and Plant Pathology 1990 4 I, / '_\‘ ‘ ‘ J’V‘/*K7‘ a ABSTRACT INVESTIGATIONS INTO TURFGRASS BLACK LAYER BY William Lee Berndt Turfgrass managers worldwide have noticed the appearance of a black layer in the soil profile of some highly maintained turf areas. It was hypothesized that black layer was an accumulation of metal 8-2 precipitates produced in response to low redox potential. Sulfide was thought to form when sulfate-reducing bacteria respired using 804-2 (or S) as a terminal e' acceptor. Low redox was probably the result of excess soil moisture or high rates of soil respiration. Chemical reaction between 8'2 and +2 resulted in formation of the soil metals such as Fe precipitate, whiCh settled on organic debris and in fine pore spaces. Thus, research was conducted to determine if 8'2 was involved in producing black layer and what factors might contribute to 5'2 based black layer production. Additionally, control of black layer was sought. The influence of S, S source, N source, soil moisture, organic matter and soil bacteria was investigated. Both lab and field studies were conducted. It was found that 8 addition to soils in vitro resulted in intense 8'2 accumulation when coupled with excess soil moisture and abundant William Lee Berndt organic matter. This resulted in intense blackening of the soil. Excess moisture, S and organic matter as lactate contributed to low redox. Nitrogen addition retarded S"2 production, but N03" was considerably more effective. The effect of NO3' was to keep redox potential elevated quelling the activities of sulfate-reducers. Sulfur addition at high levels to turf soils 1 situ also resulted in intense 5'2 accumulation when coupled with excess moisture. Intense blackening was also evident. Nitrate added in situ was not very effective in reducing 8-2 accumulation except near the soil surface. More NO3' than was applied was probably needed. Sulfate or S reducing bacteria were shown to be responsible for S cycling in existing black layer soil. This was documented using 358 as a tracer along with Moo4'2, a specific inhibitor of sulfate-reducers. Rates of reduction were also calculated to vary between 1 and 7 nM S cm—3 black layer soil day-1. It was concluded that black layer was an accumulation of 8-2 formed in response to low redox. Sulfur appeared to be the key element involved in black layer formation as long as organic matter was abundant and soil moisture was in excess. It was concluded that the best control of black layer centered around preventing 804:"2 or S reduction from occurring. Thus, withholding supplemental S and organic matter and adding fertilizer N03” routinely to keep soil redox elevated was suggested as a preventative control. To John T. Lynch, Jr., who inspired me when the going was rough. ACKNOWLEDGMENTS I would like to acknowledge my friend and advisor Dr. J.M. Vargas, Jr., for his help and encouragement. Without Joe Vargas I would not be where I am at this time. I would also like to acknowledge my guidance committee Drs. P.E. Rieke, B.E. Branham, and E.S. Beneke. I would especially acknowledge Dr. J.M. Tiedje, for being my role model. I must give a heartfelt thanks to my parents Bill and Phyllis, and to my brother Dave, for their love and support in my graduate school adventures. Without their help none of this thesis would have been possible. I would also thank other members of my family for putting up with me the last four years. Additionally, I would like to thank Ringer, Inc., the Michigan Turfgrass Foundation, and the Golf Course Superintendents Assn. of America for financial support throughout the course of this research project, and also the Department of Botany and Plant Pathology and the Pesticide Research Center. Finally, I would thank Joey for her love and devotion during the last few years, and for sticking by me when the easy way out was to quit. ii TABLE OF CONTENTS LISTOFTABLESOOOOOOOOO...OOO0....OOOOOOOOOOOOOOOOOOOOOOOO. LIST OF FIGURES................................ ............ CHAPTER I LITERATURE REVIEW AND OBJECTIVES................ What is Black Layer? .................................. Why Does Black Layer Form?.. ...... .................... The Particle Size Hypothesis ........... . .............. Other Thoughts on Black Layer......................... Dissimilatory Sulfate Reduction ........ ............... The Basics.... ........................................ Redox Potential ............................. . ......... Soil Reaction.. ....................................... Electron Donors.. ............... ...................... Electron Acceptors ...................... . ............. Fate of Sulfide.......... ........... .................. Control of Sulfate Reduction .......................... Overview and Objectives............................... References Cited........................... ......... .. CHAPTER II QUALITATIVE BLACK LAYER RESEARCH.......... ..... Abstract.. .......... . ............ ..................... Materials and Methods...... ........................... Experiments with Sulfur.......................... Experiments with Inhibitors.... .................. Experiments with Hydrogen Sulfide........... ..... Results and Discussion..... ................. . ......... Experiments with Sulfur.......................... iii page vi viii 1 10 11 ll 13 14 15 16 18 23 23 25 25 27 27 29 29 Experiments with Inhibitors...... ..... .... Experiments with Hydrogen Sulfide.......... conCIuSionSOOOOOOOOO......OOOOOOOOOOO......OOOOOOOOO References CitedOO......0..........OOOOOOOOOOOOOO0.0 CHAPTER III INFLUENCE OF SULFUR AND LACTATE ON IN VITRO SULFIDE ACCUMULATION AND REDOX POTENTIAL IN FLOODED SANDOO.........OOOOOOOOO00.0.0000... Abstract............................................ materials andMethOdSOOOO......OOOOOOOOOOO.......... Experiments with Sulfur and Lactate............ Results and DiscuSSj-ODOOOOOOOOOOOOOOI...00.0.0000... Experiments with Sulfur and Lactate............ Conclusions.. ........................... . ............. References Cited.‘ ......... 0.00.00.00.00.......OOOOO CHAPTER IV INFLUENCE OF SULFUR, AMMONIUM, AND NITRATE ON IN VITRO SULFIDE ACCUMULATION AND REDOX POTENTIAL ................. . .............. . ..... Abstract............. ......... ........................ Material and Methods..... ...... ... ................. Results and Discussion ................................ Conclusions...... ............................... . ..... Acknowledgements ...................................... References Cited......... ..... ........................ CHAPTER V A BLACK LAYER FIELD EXPERIMENT.. ..... ...... ..... Abstract..... ..... . ...... . ............................ Materials and Methods ........................ . ........ 1987 Research. ................................... 1988 Research... ............. . ................... iv 31 32 33 34 37 37 38 38 40 4O 45 46 48 48 50 52 64 64 65 68 68 69 69 71 1989 Research.. .................................. Results and Discussion ................. ..... 1987 Research.. ....................... . 1988 ResearCh. O O O O O O O O O O O O O O O O O O O O O O O O O 1989 ResearChOOOOOOOOOOOOOO ..... .00.... Conclusions ........................................... References Cited............................ CHAPTER VI EXPERIMENTS WITH RADIOACTIVE SULFUR.. AbstractOOOOOOOOOOOOO0.000000000000000000000 Materials and Methods ......... ... ........... . ......... Sulfide Distillation Apparatus ........ . .......... Sulfur Cycling in Black Layer............. Time Course Studies with a Reactor Vessel. Reduction of Elemental S.................. Results and Discussion......... ............. . ......... Sulfur Cycling in Black Layer............. Time Course Studies with a Reactor Vessel. Reduction of Elemental S ................... conCIuSionSOOOOO.....OOOOOOOOO....OOOOOOOOOOOOO References Cited............. .............. . .......... CHAPTER VII SUMMARY AND CONCLUSIONS ...... ......... ........ APPENDIXOOOOOO......OOOOOOOOOOOOOOOOO......OOOOOOOOOOOOOOOO Appendix I Measurement and Detection of Sulfide and Sulfate.......... ............... .......... References Cited ............................ .......... 72 73 73 78 81 96 97 99 99 100 100 101 103 105 106 106 111 113 116 117 120 122 123 126 table CHAPTER III Table Table CHAPTER IV Table CHAPTER V Table Table Table Table Table 1. LIST OF TABLES page Influence of S and lactate on 8'2 production and redox parameters in flooded Lake Michigan dune sand........................... 41 Trend analysis summary for lactate treatments across 8 levels.............................. 43 Linear regression statements applied to acid volatile S"2 accumulation as influenced by S, + - NH4,OI'NO3 coco. 0000000000000 0000000000000. 55 Influence of S and irrigation on black layer formation in 'Penncross’ creeping bentgrass for 1987.. ............. . ..... ................ 74 Influence of S and irrigation on visual quality of 'Penncross' creeping bentgrass for 1987.. ..... .............................. 75 Analysis of variance summary for total acid volatile sulfide occurring in a ’Penncross' creeping bentgrass green..................... 84 Trend analysis summary for total acid volatile sulfide (AVS).................. ..... 88 Analysis of variance summary for soil pH as influenced by sulfur level, nitrogen source and soil depth. .............................. 92 vi Table 6. Trend analysis summary for soil pH as influenced by sulfur applications averaged over fertilizer sources ...... ......... ....... 93 Table 7. Trend analysis summary of pH as influenced by sulfur, fertilizer source and soil depth.. 94 Table 8. Mean values for visual quality, algae and wilt ratings on plots of ’Penncross' creeping bentgrass turf afflicted with black layer ........................................ 95 CHAPTER VI Table 1. Summary of 358 cycling in intact black layer micro-cores as influenced by Na2M004 and Table 2. Influence of added 3 and N03“ on 3‘2 accumulation and redox potential in flooded sand ......................................... 115 vii LIST OF FIGURES figure page CHAPTER IV Figure 1. Influence of S, NH4+, and N03” on accumulation of AVS in flooded soil from a ’Penncross’ creeping bentgrass golf green..................................... 54 Figure 2. Influence of S, NH4+, and N03- on accumulation of H28 in flooded soil from a ’Penncross’ creeping bentgrass golf green..................................... 59 Figure 3. Influence of S, NH4+, and N03“ on redox potential as pE + pH in flooded soil from a ’Penncross' creeping bentgrass golf green..................................... 61 Figure 4. Influence of S, NH4+, and N03- on pH of flooded soil from a ’Penncross’ creeping bentgrass golf green...................... 63 CHAPTER V Figure 1. Influence of S on 8-2 accumulation in ’Penncross’ creeping bentgrass............ 80 Figure 2. Influence of S and N on 8-2 accumulation in 'Penncross' creeping bentgrass for 1989...................................... 83 Figure 3. Interactive effects between 8 level and soil depth ............. . .................. 86 Figure 4. Influence of S and N on pH(KCl) of plot viii CHAPTER VI Figure 1. soil in ’Penncross’ creeping bentgrass for 1989 ................................... ... 90 Separation of H2355 and AV358 fractions in black layer soil using a 8'2 trapping trainOOOOOO... ....... ......OOOOOOOOOOOOOOO 109 ix CHAPTER I LITERATURE REVIEW AND OBJECTIVES Introduction. Turfgrass managers and scientists from around the world have speculated on the nature and origin of black layer with much disagreement as to the exact cause. Even though various black layer symptoms have been reported on putting greens sinCe the early 1900’s little research data was generated to explain this controversial phenomenon (4, 24). The data published thus far consisted of results from preliminary studies which were published mainly in the popular literature or in abstracts to appease a demand for answers (5, 7, 17, 18, 23, 29, 30). Some data focus on what the black layer is while some focus on why it forms. The first part of this literature review will report on the these publishings. A second portion will focus on the 2 reduction, which was central to this authors 2 biology of SO4- hypothesis. As the scientific literature regarding SO4- reduction is voluminous, only the aspects related to black layer formation and control were considered. Finally, the hypothesis and objectives which set the stage for this authors research will be presented and explained. What is The Black Layer? The Sulfate Reduction Hypothesis. There have been reports which acknowledge the relationship between SO4-2 reduction and black layer formation (5, 7, 14, 16, 22, 29, 30). Results of an investigation in 1974 supported the hypothesis that black layer was produced by the activities of sulfate-reducing bacteria such as Desulfovibrio (22). This finding was reported to be corroborated by researchers at Michigan State University in 1987 (5). In their initial reports these investigators hypothesized that black layer was composed of accumulations of metal sulfides (i.e., FeS) brought about as a result of SO4-2 reduction. They postulated that FeS accumulated in pore space and on organic debris in response to low redox potential induced by either waterlogging, microbial respiration, or a combination of both. Further, addition of supplemental S was thought to aggravate the black layer condition. In glasshouse experiments with sand, 8, and crude cultures of sulfate-reducing bacteria researchers reported that black layer formed in as little as 14 days when soils were water-logged (5). Where no S was added no black layer formed even after 33 days. Addition of S at 48 kg ha'1 readily produced black layer. Higher levels of S intensified it. Further, addition of fertilizer N03- reportedly prevented the formation of black layer. Smith (30) also postulated that the black layer was ultimately produced from dissimilatory 804-2 reduction involving sulfate-reducing bacteria and water-logged conditions. Some of the physical factors postulated to enhance black layer development were heavy rainfall, perching of water tables, ponding, and the development of impervious layers in the soil profile. Smith further postulated that excess water was important in initiating black layer but the key reaction element was probably S. In a pot culture study which investigated the effects of N source and S on the development of black layer in bentgrass turf, urea was associated with most rapid black layer development possibly because urea N was in a form most readily available to sulfate-reducing bacteria (30). However, the condition developed with all tested N sources including ammonium nitrate. There was insufficient data to comment on the role of S. Rankin (29) stated that the U.S. razza-matazz that accompanied the discussion on black layer required a more moderate approach. He claimed that the condition which resulted in black layer was simply one of lack of oxygen in the rootzone and that restricted aeration, caused by several factors such as compaction, creation of layers, poorly managed irrigation and run-off from surrounding areas, was ultimately responsible. Thus, any practice which water-logged the rootzone or rapidly removed the oxygen may result in conditions which produce black layer. Further, conditions leading to anaerobiosis might develop in only a few hours. When Rankin analyzed black layer deposits on a tensiometer placed in an Australian black layer soil it was reported that the black deposits were composed of FeS and MnS. The research +2 showed a 10 fold increase in Mn concentration on tensiometers placed in water-logged soils with a less dramatic increase in Fe+2. It was concluded black layer was predominantly MnS. Further, in the absence of adequate S and organic matter black layer appeared blue-grey. Rankin proposed that an effective cure must tackle the cause and not the symptoms, and the cause was reported to be waterlogging. Correct diagnosis of the reason for waterlogging should provide the most effective solutions. Why does Black Layer Form? Tee Algee Hypothesis. A relationship between the occurrence of surface algae (i.e., Cyanobacteria) and black layer formation was also reported (11, 17). In preliminary microbial screenings of black layer in midwestern golf greens Hodges noticed that anaerobic bacteria were in high titer while aerobic bacterial activity was diminished. Further, surface algae seemed to be prolific in many cases of black layer. Hodges hypothesized that after algal colonization of surface soil, mucilage produced by the algae would diffuse downward to plug the profile, hinder water infiltration, and cause black layer. In laboratory experiments which sought to reproduce black layer Hodges (17, 18) examined the ability of the algae Oscillatoria to colonize columns of silica sand or calcareous sand. After 10 weeks incubation water infiltration rates on silica sand inoculated with algae were 54% of the control while infiltration rates for calcareous sand inoculated with algae were 4.8% of the control. Further, removal of the surface algal growth did not regenerate water infiltration for either sand. Thus, Hodges concluded algal polysaccharides must have contributed to plugging the profile, hindering water infiltration. However, after 10 weeks no black layer had developed in either sand even though water infiltration was reduced by the presence of algae. Black layer developed only after subsequent inoculation of the column with mixed cultures of anaerobic bacteria (i.e., sulfate-reducing bacteria). Twelve to 16 weeks after the subsequent inoculation well developed black zones were visible, especially in the silica sand. Hodges concluded that both algae and sulfate-reducing bacteria (i.e., Desulfovibrio), and five to six months incubation were necessary for black layer to form in his studies. It was suggested (11,19) that the algae probably physically modify the soil environment making it conducive for the proliferation of anaerobic bacteria while also providing a bacterial growth substrate. Couch (11) and Hodges (19) recommended monitoring water infiltration levels, using frequent aerification, reconstructing drainage and controlling surface algae with mancozeb to prevent black layer. They further speculated that sand type, water quality and quantity and mowing practices may help to dictate black layer dynamics, and that the superintendent may have little control over the development of black layer as he is locked into the requirements of sand green culture. The Particle Size Hypothesis. Burpee (7) recognized two distinct types of black layer in southern Ontario golf greens. In a type one layer the bottom was not distinguishable and the layer penetrated into the profile. In a type two layer a horizontal black banding with distinct margins several centimeters thick was apparent. Burpee postulated that some unique physical property must develop in type two layers that allows for proliferation of algae or anaerobic bacteria. He proposed that the origin of such black layers may be traced to the movement of silt and clay into distinct strata within the soil profile. Further, the existence of aberrations in particle size distribution (PSD) within the profile may lead to development of poorly drained, anaerobic zones. The hypothesis that PSD in type two black layer is different than PSD above or below black layer was tested in cores of root-zone mixes possessing type two layers. Burpee reported that all greens tested had a greater percentage of coarse sand and coarse sand plus very coarse sand below black layer than in black layer. Also, the mean distribution of coarse sand plus medium sand did not exceed 50% in or below black layer. Percent distribution of very fine and fine sand were higher in black layer than below black layer but the percent distributions for these particles were not different between black layer and layers below black layer. Results also suggested that the distribution of silt and clay were not different in black layer as opposed to below black layer, but the mean distribution of silt plus clay exceeded 10% in each of the zones tested. Lastly, percent organic matter was higher in black layer than below black layer. Burpee concluded that particle size and distributions were not homogeneous between black layers and zones below black layer, particularly with respect to the very coarse and coarse fractions. Also, distributions of fines exceeded recommended levels while coarse fractions were less than recommended. The research suggested that layers of different PSD exist in golf greens in southern Ontario and these layers may play a role in black layer development. Other Thoughts pp Black Layer. Opinions Apepp Sleek Leyep. Lubin (23) made the following observations regarding black-layer like problems in turf soils on the west coast: the metal ion concentrations were higher in black layer than elsewhere in the sample profile, and it was at the black layer that root growth stopped. Additionally, P appeared to be deficient in a number of black layer soil samples and in tissue samples from plants growing in affected soils. He reported that a number of golf courses in the southwest applied mono-ammonium phosphate on a regular basis to control black layer damage associated with metal ion build-up. Lubin indicated P addition functioned in controlling black layer by forming metal phosphate precipitates of low solubility and by stimulating root growth. Lubins conclusion was that application of soluble P was necessary for managing black layer. Goss (16) reported that soils in the coastal northwest commonly develop anaerobic conditions during the winter unless they are properly managed. He postulated that the black layer resulted from total neglect of one or several management practices. He acknowledged that sulfate reduction was probably involved and that algae might play a role. However, Goss also wrote that all involved in the black layer controversy should be embarrassed by forgetting the fundamental concepts of managing turf. Gockel (14) wrote that the answers for black layer could be found in the literature of agricultural drainage. She acknowledged the reduction of sulfate under anaerobic conditions and subsequent conversion to metal sulfides produce the black layer, but waterlogging was the real problem. She further wrote that obstructions to normal water movement will result in waterlogging. Thus, she postulated that silt, clay, and organic debris migration through the sand interstices may assist black layer formation. Once the obstructions are rectified, and soil moisture levels are under control, black layer should be little problem. Beard (4) suggested that increased fungicide use, use of 100% sand in the rootzone and intense topdressing with 100% sand are recent trends which may play a role in black layer formation. He wrote that excessive use of fungicides causes significant shifts in microbe populations which may lead to serious soil environment problems. Beard also suggested inconsistent irrigation regime and percolation problems involving sand over an impermeable clayey soil may be a factor. He pointed out, however, that serious excess-soil-water problems may occur in humid to high rainfall climates regardless of irrigation practices. Dissimilatory Sulfate Reduction. Tee Basics. Dissimilatory sulfate reduction is bacterial respiration using 804'2 as the terminal electron acceptor (TEA). Sulfate reduction involves specific groups of strict anaerobes and occurs only in the absence of oxygen. Genera included as sulfate- reducing bacteria are Desulfovibrio, Desulfotomaculum, Desulfomonasl Desulfonema, Desulfobacter egg Qeeulfoearcina (31). Desulfovibrio sp. are non-sporulating, heterotrophic, motile, spiral shaped organisms while Desulfotomaculum sp. are sporeforming, motile, heterotrophic rods (31). As a group, these organisms occupy many niches including soils, lake sediments, marine sediments, rumen and human intestinal tracts, and also tolerate a wide range of environmental pressures (26). Chemically, sulfate reduction is a 6 electron enzymatic reduction involving sulfate adenyltransferase, APS reductase and sulfite reductase. Sulfite reductase (MW 670,000) is among the most complex enzyme cascades known (32). It catalyzes the reaction: 3 nadph + 3 H+ + 503’2 < ---------- > 3 nadp+ + 5'2 + 3 H20 and is repressed by cysteine and cystine. However, the mechanism of 5'2 formation is not fully resolved (1). Sulfide formation may involve 3 hypothetical pathways: a) direct reduction of 803-2 with no free S being formed b) initial formation of 8203-2 or c) initial formation of 8306-2 (1). Some investigators believe that 10 8203-2 and 5306"2 are simply products of side reactions of dissimilatory 804-2 reduction while others argue that they are true intermediates (1). Sulfate reduction is influenced by redox potential, pH, organic matter input, sulfur source, temperature and pressure (1, 9, 15, 20, 33) among other less obvious factors. As much has 2 reduction, in this paper the topics of been published on SO4- redox potential, pH, electron donors and acceptors, environmental influences, and the fate and control of end products of 804:—2 reduction will be briefly discussed. Redox Potential. Connell and Patrick (9) reported that in an examination of flooded rice (pree spp.) paddy soils with closely controlled redox potentials little or no 8'2 accumulated with a redox potential above -150 mV. They also reported that the more negative the Eh was, the more complete the 804"2 reduction was. They concluded that there was little doubt redox potential was a controlling factor in 804'2 reduction. However, Vainshtein (35) detected sulfidic compounds from Indian Ocean sediments where Eh = +500 mV. He concluded that redox probably affects the sulfate- reducer type more than the actual process, but at the same time cites little doubt that redox potential is one of the prime ecological factors influencing the rate of 804-2 reduction, especially at boundaries dividing oxidized and reduced zones. Vainshtein demonstrated an upper limit of +110 mV for active 8"2 production in an apparatus which established Eh gradients using active H2. He also demonstrated that different species of 11 sulfate-reducers have differing upper-Eh limits. Thus, it 2 appears that 804' reduction may take place over a wide range of redox potentials depending on the specific sulfate-reducer type. Sell Reaction. Connell and Patrick (9) also found a pronounced effect of pH on 804-2 reduction. They demonstrated in studies involving soils adjusted to wide ranges of pH values that little reduction occurred outside the pH range 6.5 to 8.6. Although this concept is generally accepted, waterlogging of soils does tend to shift the pH of both acid and alkaline soils toward the neutral point as a result of chemical reactions involving iron and manganese. This tends to bring most water-logged soils into the pH range suitable for $04!.2 reduction (9, 10). Electron Donors. The rate of 804-2 reduction is clearly dependent on the nature and amount of organic C substrate (15). Goldhaber and Kaplan (15) conclude that in ocean sediments the rate of 804-2 reduction is dependent on: a) the total organic C preserved in sediment and b) the state of complexing of the organic C and it's availability for biogenic degradation. These parameters are influenced by the environment of deposition and rate of sedimentation. Goldhaber and Kaplan also cite that by changing the nature of the electron donor, the rate of reduction by a constant bacterial population may vary by as much as 50 fold. Postgate (27) demonstrated that the range of organic molecules 2 utilized in $04- reduction is limited to a few short chained carboxylic acids (i.e., lactate). Smith (31) reported that -.'—""—",_-....-—_- m.--‘ -. -1” _ 12 sulfate-reducers from freshwater lake sediments may use a variety of carboxylic acids as electron donors including acetate, succinate and butyrate. His studies also indicate that amino acids and propionate, as well as molecular hydrogen can also serve as donors. Badziong, Thauer and Zeikus (2) report that growth of sulfate-reducers in an acetate medium consumed 4.2 mol of molecular hydrogen per mol of 804"2 reduced to 8'2. Further, acetate was not involved directly but was required for biosynthetic purposes. Dicker and Smith (12) report that competition studies in salt marsh sediments indicated that the preferred order of substrate utilization for sulfate- reducers was lactate, then acetate followed by ethanol. The turnover of these compounds in salt marsh sediments was rapid suggesting that these low molecular weight compounds play a significant role in anaerobic metabolism. Goldhaber and Kaplan (15) maintain that if Postgate’s theory holds, then sulfate-reducers must rely upon a complex community of fermentative bacteria to supply such organic carbon. Indeed, Utsalo and Maier (33) demonstrated that in cellulose-based microcosms, sulfate-reducers rely on cellulose digesters (i.e., Cellulomonas) and lactic acid fermenters (i.e., Enterobacter) for their supply of organic C. Without the consortium including all three physiological types, 304-2 was not reduced. 2 The importance of SO47 reduction to C metabolism is supported by studies which report that SO4'2 reduction catalyzed up to 50% of C mineralization in marine sediment model systems (31). Since the rate of 804-2 reduction displays saturation kinetics at l3 2 greater than 10 mM, this implies that concentrations of 804' electron donor availability is the rate dependent factor in many sediments. Electron Acgentore. While SO4-2 is not the most energetically favorable electron acceptor for sulfate-reducing bacteria, it is the species most frequently encountered in nature (31). The 2 importance of 804' as an electron acceptor is related to the fact that it may be the dominant oxygen containing ion in some anaerobic soil habitats (15). 2 occurs near sediment surfaces Generally, accumulation of SO4- and decreases with increasing depth (31). Cappenberg (8) reported that in freshwater lake sediments maximum numbers of sulfate-reducers were found near the sediment surface and declined with depth. Thus, rates of 804’2 reduction are expected to be significantly greater near sediment surfaces wherever 504-2 pool concentrations and numbers of sulfate-reducers are high. Jorgensen (21) reports that in marine sediments, anaerobic C 2 mineralization is dominated by SO4- reduction near upper 2 sediments where high concentrations of 804' exist. Ramm and Bella (28) examined extracts from algal mats from tidal flat surface sediments and reported SO4-2 concentrations to be as high as 2,150 mg. per liter algal extract. However, drawing conclusions about reduction rates from pool sizes may be in error (31). vanGemerden (34) demonstrated that significant rates of S04”2 reduction in mixed cultures of sulfate-reducers can occur when concentrations are low if rapid turnover times are evident. 14 Cappenberg (8) found evidence of large numbers of sulfate-reducing bacteria and sulfidic compounds in freshwater 2 was detectable. lake sediments when no SO4— The significance of other electron acceptors to sulfate-reducers in nature has yet to be demonstrated (31). However, Postgate (27) and more recently Biebl and Pfennig (6) demonstrated the growth of several strains of sulfate-reducers with elemental S. Postgate concluded that colloidal S permitted slow growth or slow H2 absorption when a resting cell suspension of Q. desulfuricans was used. Biebl and Pfennig also showed that there was slow but definite growth of Q. glgee using S. They further demonstrated that strains which grew well with S did not sporulate and contained no desulfovidrin. Much less is known about $203-2 and 5406-2 reduction by sulfate-reducing bacteria (3). Postgate (27) was the first to report that these compounds could serve as electron acceptors. He quantitatively demonstrated that resting cell suspensions of Q. vulgaris could reduce either using H2 as a reductant. This type of reduction 2 probably constituted a tributary of SO3- reduction (3). Fate 9: Sulfide. The fate of reduction end products involves separation into water soluble, acid soluble and acid insoluble fractions. Acid soluble fractions include metal mono-sulfides; pyrite is acid insoluble. Hydrogen sulfide is water soluble. Connell and Patrick (10) report that in a normal submerged rice paddy soil which is well supplied with Fe+2 , most of the free H28 formed is removed from the solution as precipitated FeS (a metal mono-sulfide). With time, pyrite may be formed. However, if 15 there is not enough active Fe+2 free H28 may be evolved. Free H28 is a respiratory feedback inhibitor acting at cytochrome C + C3. It can accumulate in amounts which are toxic to rice plants (10). 2 reduction from the turnover of tracer Measurements of SO4- amounts of 358 amended in the surface peats of a Cape Cod salt marsh demonstrate that pyrite, and not metal mono-sulfides, is the major end product (20). Very little, at best 30%, of the resulting 35$ label ends up in acid-volatile pools (i.e., FeS). Howarth (20) demonstrated that in a salt-peat marsh, pyrite can +2 form in less than a day without Fe mono-sulfides as intermediates. He also suggests that if the label in the pyrite 2 fraction is not measured, rates of 804' reduction could be grossly underestimated. Control e; Sulfate Reduction. Connell and Patrick (10) investigated the effects of added N03” and 02 on 804'2 reduction in highly reduced rice paddy soils. They report that the presence of N03” inhibited the production of 8'2 in the paddy situation. Sulfide was not formed until the N03” pool was depleted (approximately 3-4 days). They concluded that the effect of added N03" was to delay the initiation of SO4-2 reduction, probably by providing a preferred electron acceptor which would yield more ATP for metabolism. They also reported that oxidation of 8'2 in a water-logged soil occurred rapidly in the presence of air. Fifteen minutes exposure to 02 decreased the sulfide 8’2 from 120 ppm to 60 ppm, and after 8 hours all the 8’2 had disappeared. They concluded that the disappearance was m --.- -. _ . ... e._-.. _ .. 16 very likely due to chemical oxidation of'S'2 to S. Connell and Patrick also report that addition of Fe203 was effective in precipitating free H28. Engler and Patrick (13) examined the effects of various chemical oxidants on inhibition of 8‘2 production in rice paddy soils. They report that KNO3 and MnO2 were effective delaying S"2 accumulation. Further, ferricitrophosphate and FePO4 were also found to be effective in delaying 804-2 reduction. The less soluble compounds such as FePO4 were less effective in delaying 8'2 production but persisted longer. In each experiment 8'2 production did not ensue until redox potential fell below -100 mV. Their work suggested that redox potential is a controlling 2 factor for $04- reduction, and that oxidants such as N03- act by retarding redox potential depression. Overview and Objectives. The objective of the research described in this thesis was to elucidate the nature of black layer, and to find an acceptable measure of control for black layer which could be utilized by turfgrass practitioners. The authors hypothesis related black 2 reduction. This meant that the black in layer formation to SO4- black layer was hypothesized to originate from depositions of acid volatile 8'2 (AVS) of bacterial origin. The AVS was thought to accumulate in pore spaces and on organic debris in response to low redox potential. Why it accumulated where it did was not addressed. It was likely that soil layering due to migration of 17 finer particles, accumulation of organic debris (such as algae) in a layer, thatch buried with topdressing, or many other reasons may play a role. For our experiments control of black layer centered around the use of NO3' fertilizer which was shown in previous research to inhibit 8'2 production by maintaining a relatively high redox potential. Specifically, the objective of Chapter II was to describe a series of qualitative experiments conducted to determine if addition of S would result in formation of black layer, and if it could be controlled. The objective of Chapter III was to describe research conducted to quantitatively determine the influence of added S and organic matter as lactate on 8'2 production. This was a point in time experiment. The objective of Chapter IV was to describe the effect of S, NH4+, and N03' on the time course of appearance of 8'2 production, and redox potential. The objective of Chapter V was to evaluate the influence of N source and S application level on 8'2 production in situ. This was a point in time experiment originally begun as a qualitative study to determine the influence of added S and irrigation frequency on black layer formation. Finally, the objective of Chapter VI was to use radio-isotope tracers to establish rates of SO4-2 reduction in intact sediment cores. Further, a modified reactor vessel was utilized to determine the time course of appearance of H28, and the influence of N03- addition on such production. Lastly, radioactive elemental S was used to ascertain whether 8 reduction was occurring. 18 REFERENCES CITED Alexander, M. 1977. Introduction to soil microbiology. John Wiley and Sons, N.Y. Badziong, W., R.K. Thauer, and G. Zeikus. 1978. Isolation and characterization of Desulfovibrio growing on hydrogen plus sulfate as the sole energy source. Arch. Microbiol. 116: 41-49. Barrett, E., and M. Clark. 1987. Tetrathionate reduction and production of hydrogen sulfide from thiosulfate. Microbiological Reviews 51: 192-205. Beard, J.B. 1988. Why has black layer increased? Grounds Maintenance 22: 10. Berndt, W.L., and J.M. Vargas et a1. 1987. Black layer formation in highly maintained turfgrasses. Golf Course Mgt. 55: 106-110. Biebl, H., and N. Pfennig. 1977. Growth of sulfate-reducing bacteria with sulfur as electron acceptor. Arch. Microbiol 112: 115-117. Burpee, L.L., and A. Anderson. 1987. The cause of black layer in golf greens: an alternative hypothesis. Greenmaster 23: 24. 10. 11. 12. 13. -I I m ‘IPI; t;.-‘_. at _ “_ -_ 19 Cappenberg, T. 1974. Interrelations between sulfate-reducing bacteria and methane-producing bacteria in bottom deposits of a freshwater lake. Antonie vanLeeuwenhoek 40: 285-295. Connell, W., and W. Patrick. 1968. Sulfate reduction in soil: effects of redox potential and pH. Science 159: 86-87. Connell, W., and W. Patrick. 1969. Reduction of sulfate to sulfide in waterlogged soils. Soil Sci. Soc. Am. Proc. 33: 711-715. Couch, H.B. 1987. Black layer: anaerobiosis is the condition but sulfur is not the cause. The Bull Sheet 41: 6-9. Dicker, H., and D. Smith. 1985. Metabolism of low molecular weight organic compounds by sulfate-reducing bacteria in a Delaware salt marsh. Microb. Ecol. 11: 317-335. Engler, R., and W. Patrick. 1973. Sulfate reduction and sulfide oxidation in flooded soil as effected by chemical oxidants. Soil Sci. Soc. A. Proc. 37: 685-688. 14. 15. 16. 17. 18. 19. 20. 21. 22. 20 Gockel, J. 1987. Black layer: looking for light in all the right places. Golf Course Mgt. 55: 26-32. Goldhaber, M., and I. Kaplan. 1975. Controls and consequences of sulfate reduction rates in recent marine sediments. Soil Science 119: 42-55. Goss, R. L. 1987. Managing anaerobic soils. Northwest Turfgrass Topics 30: 7. Hodges, C.F. 1987. Blue-green algae and black layer: part I. Landscape Mgt. 26: 38-44. Hodges, C.F. 1987. Blue-green algae and black layer: part II. Landscape Mgt. 26: 30-31. Hodges, C.F. 1989. Another look at black layer. Golf Course Mgt. 57: 54-58. Howarth, R. 1978. Pyrite: its rapid formation in a salt marsh and its importance in ecosystem metabolism. Science 203: 49-51. Jorgensen, B., and T. Fenchel. 1974. The sulfur cycle of a marine sediment model system. Mar. Biol. 24: 189-201. Kozelnicky, G.M. 1987. Turf talk from the old Koz. Georgia Turfgrass News 20: 3. 23. 24. 25. 26. 27. 28. 29. 30. 21 Lubin, T. 1987. Black layer: a western view. Divot News 25: 26-42. Norr, O.J. 1939. Excess rain is turf menace. Golfdom 13: 12-14. Norr, O.J. 1949. Physical soil factors cause big tufr loss in 1949. Golfdom 23: 70-75. Paul, E.A. 1980. Soil microbiology and biochemistry. University of Sasketchewan Printing, Saskatoon Sask. Postgate, J. 1951. The reduction of sulfur compounds by Desulfovibrio desulfuricans. J. Gen. Microbiol. 5: 725-738. Ramm, A., and D. Bella. 1974. Sulfide production in anaerobic microcosms. Limnology and Oceanography 19: 110-118. Rankin, P.C. 1988. When the black layer hit the fan. Turf Craft Aust. 6: 18-20. Smith, J.D. 1988. Black plug layer on Sasketchewan golf courses. Greenmaster 24: 6-21. 31. 32. 33. 34. 35. 22 Smith, R. 1981. Sulfate reduction in the sediments of a eutropic lake. Dissertation for the degree of Ph.D. Michigan State Univ. East Lansing, MI. Smith, E. et al. 1983. Principles of biochemistry. McGraw-Hill, NY. Ustalo, S., and S. Maier. 1986. Growth patterns and hydrogen sulfide production by mixed cultures in a cellulose based medium. J. Gen. Microbiol. 32: 491-498. vanGemerden, H. 1967. On the bacterial sulfur cycle of inland waters. Ph.D. Thesis Leiden, Holland. Vainshtein, M., and G. Gogotova. 1987. Effect of redox potential of the medium on sulfide production by sulfate-reducing bacteria. Mikrobiologiya (translated) 56: 31-35. CHAPTER II QUALITATIVE BLACK LAYER RESEARCH Abstract. Black layer formation in turfgrass soils was hypothesized to result from accumulations of metal sulfides 2 reduction in response to low redox potential. produced by 804' Qualitative greenhouse research was initiated to determine whether black layer could be experimentally reproduced and whether S addition or moisture content would influence development. In subsequent studies the effect of possible controls was determined as was the influence of free H28 on 'Penncross' creeping bentgrass. Addition of S at 48 kg 8 ha-1 resulted in production of black layer when soil moisture was excessive. Increasing the level of S to 244 kg ha“1 intensified the condition. Without 8 addition minimal black layer formed. Addition of 0.1 M M004-2 or NO3' solution prevented black layer formation. It was also shown that H28 at 1000 ug ml'l killed turfgrass. It was concluded that S was the key ingredient in black layer formation and that S reduction was responsible. Soil moisture was considered very important. Further, H28 could realistically be involved in the turfgrass decline. Finally, adequate NO3' fertilization of turf was proposed as a black layer control. Black layer has been called the number one malady on bentgrass greens and tees today (12). Black layer was reported to begin with a noticeable drop in percolation followed by the formation 23 24 of a black, offensive smelling layer somewhere in the upper soil profile (12). Affected turf shows symptoms of bronzing, often followed by a decline, thinning or outright loss (12). Many hypotheses regarding black layer formation have been proposed (2, 11). Initially, these ranged from the algae hypothesis (11) to blatant mismanagement (9). Work in our laboratory suggested that black layer formation was related to dissimilatory SO4-2 2 reduction. In this respiratory process soil bacteria use 804' as the terminal electron acceptor in place of O2 (1). It happens only where 02 is absent. The reaction produces H28 as an end product, which is very reactive with divalent soil metals such as Fe+2 (1). Upon reaction between H28 and soil metals, black metal 8’2 precipitates such as FeS are formed. These precipitates flocculate on organic debris and in soil pore space giving the profile a blackened appearance . It was suggested that an accumulation of such precipitates physically composed the black layer, probably in response to low redox potential. If not enough active metal is available in the system to bind H28 as e.g. FeS, free H28 may be evolved (7). Hydrogen sulfide is well known as a respiratory poison inhibiting electron transport (16). It has also been shown to interfere with the oxidative power of rice (pree spp.) roots (16). Thus, it was thought that H28 could be at least partially involved in the observed turfgrass decline. Addition of N03” was reported to keep the redox potential of rice paddy soils high enough to bypass 8'2 formation and thus alleviate rice suffocation disease (5). The effect of the N03- 25 was to act as an alternate electron acceptor as 02 became depleted (15). The objective of this research was to qualitatively determine whether black layer formation could be related to SO4-2 reduction and S addition, and whether H28 could produce a turfgrass decline. The use of fertilizer N03“ as a black layer control was also evaluated. MATERIALS AND METHODS Experiments ylpp Sulfur. This experiment was conducted in a glasshouse facility. The objective was to determine whether 8 addition would contribute to black layer formation, and whether soil moisture would exert an influence. Two liter plastic buckets having 3 small bottom drainage holes (i.e., 6 mm) were used in the study. Buckets were 19 cm high by 15 cm diameter at the top. Bottoms were lined with 3 cm of pea gravel. Buckets were then packed with 1 L of either washed Lake Michigan dune sand (99% > 0.1 mm and 84.5% > 0.25 mm, 0.0125% om), or dune sand plus peat as an 90:10 mix. A 2 cm layer of washed mortar sand (81% > 0.1 mm and 35.1% > 0.25 mm) was then placed over the dune sand in each unit. This was done to create a perched water table. Bulk density of the dune sand was 1.7 g cm-3. Packed buckets were uniformly wetted then treated with 200 mls of 0.01 M lactate amended with flowable 52% S. Sulfur was applied to the units at levels of 0.0, 48.8 or 244 kg S ha-1, and Fe+2 was applied as FeSO4 at 6.2 kg ha'l. Additionally, N as (NH4)ZSO4 was applied at 48.8 kg N ha'l. Fifty mls of crude lactate enriched cultures of sulfate-reducing bacteria was used 26 to inoculate each bucket.’ Bacteria were obtained from a wet area of Baker Woods in East Lansing. One half of the treated units of each sand type (i.e., sand or sand + peat) were then immersed in tap water contained in an aluminum pan. This was done to waterlog the units via capillarity. Pans were filled daily to keep the moisture content constant. The other buckets were arranged so that free drainage would occur and units would not be water-logged. A hundred mls of water were added periodically to keep the soil moist. Bucket units were fitted with lids to prevent evaporation and to reduce O2 diffusion. Units were then incubated for 33 days at ambient temperature. There were 4 replications of each treatment. After 33 days each bucket was disassembled and scored for the presence or absence of any black layering. In additional experiments buckets were prepared in much the same way except that re-precipitated colloidal S was used instead of flowable S. Two thirds of the sand was placed in the buckets prior to treatment with S. Then S was applied uniformly to the sand surface at levels of 96 or 488 kg ha'l. Remaining sand was then placed over the S to create a sandwich effect. Units were treated with 50 mls of sulfate-reducer inoculum, 200 mls 0.01 M lactate, and 48.8 kg ha"1 N, then uniformly wetted. The lids were placed and the units were water-logged as described. No peat was added to the sand in this portion of the study and all units were flooded for 33 days. After 33 days the buckets were disassembled and scored for the presence or absence of any black layering. 27 Experiments ylph Inhibitors. These experiments used sand buckets identical to the ones previously described. Both flowable and colloidal S were used. The objective was to determine whether black layer could be prevented or reversed by chemical addition. No (NH4)ZSO4 N was applied, but N as either Ca(NO3)2, KNO3 or NaNO3 was added as 200 mls of 0.1 M solution. In addition, other experimental units also received 200 mls of 0.1 M Cl' bleach, NH4MoO4, or NaN3. Chemicals were added at the time of 8 treatment or after 21 days when formation of black layer was evident. One set was left as a check. Experimental units were then inoculated with sulfate-reducers and lactate, fitted with lids and water-logged for 21 days. At the end of 21 days units treated with the oxidants were disassembled and scored for the presence or absence of any black layering. After 21 days, units in which black layer was active were treated with the oxidants and observations were made immediately. Experiments yith Hydrogen Sulfide. This portion of the experiment sought to determine whether H25 would be toxic to turfgrass. Cores were constructed from 10.8 cm diameter x 15 cm PVC. The PVC had been previously cured at 60 C for 48 hrs to drive off ethylene. Plexiglas was glued to the bottom of each core so that no leakage would occur. One 6 mm hole was drilled 2.5 cm from the bottom of the core and one hole was drilled 2.5 cm from the top. Plastic nipples were glued into each hole for attachment of tygon tubing. The holes served as solution ports. Covers for each core were constructed from a 5 cm ring of PVC to which plexiglas had been glued. 28 Each core was filled to within 2.5 cm of the top with equal weights of sterile 3, 4, and 5 mm Pyrex glass beads. The glass beads served as a soil matrix which would allow for movement of solution through the core and would allow fpr ample root generation. Next, a 10.8 cm diameter plug of ’Penncross' creeping bentgrass (Agrostis palustris Huds.) consisting of only turf and thatch (approximately 1.5 cm thick) was transplanted into each core. Cores were fertilized with 20-20-20 at a rate of 48.8 kg N ha"1 and incubated in a growth chamber for two weeks. Water was added as needed to prevent wilting. At the end of two weeks each core was attached from the bottom port to a peristaltic pump connected to a 4 L anoxic water reservoir. Each reservoir was subsequently connected to a pryogallol trap which minimized resovoir exposure to 02 as water was withdrawn by the pump. For each run of the experiment one resovoir contained only freshly boiled water while another contained water amended with H28 at 1000 ug ml'l. Water was amended by sparging with gaseous H28 and periodically checking concentration by the method of Cord-Ruswich (6). The pH of the fresh water was adjusted to correspond to the pH of the 8'2 amended water (i.e., pH = 3). Cores were then capped with the lids and sealed with tape, and the pump switched on at a speed equivalent to pumping 0.6 L day-1. Capping prevented undue atmospheric exposure to the core during the pumping but still allowed light to reach the turf. Cores were incubated in a fume hood under a plant gro lite (G.E.) at ambient temperature. Solution was circulated through each core for 7 days. As the cores became filled the solution excess drained from the upper 29 port. In this way the transplanted turfgrass was constantly exposed to only water or water plus H28 with minimal atmospheric 02 exposure. Also, since glass was the soil matrix no interfering ions (e.g., Fe+2) were available to bind the H28 as a precipitate and render it inert. RESULTS AND DISCUSSION Experiments with Sulfur. Where no 8 was applied no blackening occurred regardless of soil type or moisture status, even after 33 days. However, a slight darkening of both the soils did occur where water was in excess. The darkening appeared to be greater in the peat mix but the visual difference was minimal. Where soil was freely drained no color change appeared. Where S was applied at 48.8 kg ha"1 to freely drained soil, again no black layer formed regardless of soil type. A distinct blackening of the entire profile did take place when S was added at 48.8 kg ha"1 and soil moisture was in excess. This was considered a type I layer (4). No difference in black layer formation between sand or sand + peat was observed. When level of S increased to 244 kg he”1 the blackening became more intense in all replicates. Again, no differences between soils was evident. Sulfur applied at 244 kgha"1 to freely drained soil produced some blackening as "pockets" in 75% of the replicates. This blackening occurred near the bottom of the bucket and did not encompass the entire profile. The blackened soil reacted positively to 5’2 spot testing (8). Soil also lost the blackening upon exposure to atmosphere with time. Units 30 possessing black layer also had a distinct foul odor. Thus, blackening was considered an accumulation of 8'2 in response to 2 reduction. water-logging, probably produced by bacterial 804- The blackening of saturated sand was more intense in the mortar sand layer compared to the dune sand profile. This was attributed to the mortar sand having a wider range of particle diameters with a greater percentage of fine particles compared to the dune sand, hence smaller pore spaces (10). This implied that perched water tables created by textural interfaces may be areas of the profile most conducive to formation of intense black layering. Alternatively, it could be that a majority of the added 8 was effectively retained by the finer soil impeding its movement into the profile. However, this topic was not addressed in the research. It appeared that capping the bucket and flooding the soil reduced 02 diffusion (10) to the point where S or 804-2 reduction readily occurred. Adding large quantities of 8 probably also helped to reduce available 02 in response to S binding 02 upon oxidation, as evidenced by pocket black layer formation where S was applied at 244 kg ha'l. Lactate addition also undoubtedly stimulated microbial respiration contributing to 02 consumption. It also appeared that S was limiting for intense black layer formation in the experimental sand, as evidenced by lack of black layer where no 8 was applied and where lactate and excess water was added. Algae also appeared on the surface of all experimental units except those not receiving S. Algal invasion was secondary to black layer formation. Evidence of black layer appeared in 4-5 31 days while the algae proliferation did not appear to begin until near day 15. No algae growth was noticed below the soil surface. Algae were probably indigenous to the sand but some were undoubtedly added with the sulfate-reducer inoculum. In this experiment black layer formation was attributed to presence of S and induced anaerobiosis through waterlogging, not algal proliferation as suggested by Hodges (11). When sandwiched units were examined black layer had readily formed. In these units however, the black layer was observed to be in a distinct banding as opposed to coloring the whole profile. Thus, in this experiment a type II layer was produced (4). This would tend to suggest that the colloidal S was relatively immobile in the soil compared to the flowable S. The 488 kg level of S produced a darker banding that did the 96 kg ha'1 level. Experiments with Inhibitors. The compounds Ca(NO3)2, KNO3, NaNO3, NaMoO4, Cl' bleach, and NaN3 all prevented black layer formation when added concurrently with the S treatment. In addition the bleach and N03' also oxidized existing black layer at the experimental concentration (i.e., 0.1 M) but only when exposed to atmospheric 02° The effect of the N03- was to act as an alternate electron acceptor in the absence of 02. This in effect probably kept the redox potential from falling below a critical level necessary for 8’2 formation, but this was not demonstrated. The effect of the NaMoO4 was to act as a lcompetitive inhibitor of sulfate-reducing bacteria. This meant the Km value (i.e., Michaelis constant) was increased by addition of the M004 but Vmax was similar (13). The configuration of the 32 2 and had been 14004 is stereochemically similar to 804' demonstrated to inhibit ATP-sulfurylase, the first enzyme in the 804'"2 reduction pathway (15). The effect of the NaN3 was to act as a sterilizing agent killing the sulfate-reducers. Addition of azide proved that black layer formation was biological in nature while molybdate suggested 804"2 reduction was responsible. In a practical aspect these results imply addition of fertilizer N03- could be a successful control measure for preventing black layer, but probably not for reversing the condition unless coupled with core aerification where atmospheric 02 is not limiting. Although N is routinely applied to golf greens, in this modern age many turf managers have strayed from using N03- to protect the groundwater and environment. Slow release organic fertilizers have largely replaced the N03' for golf course use. These materials are acceptable if the soil is aerobic enough to permit nitrification, but if 02 status in soil is diminished, nitrification becomes a significant 02 sink further enhancing the depression of redox potential, possibly contributing to black layering. Experiments yith Hydrogen Sulfide. After 7 days incubation in the presence of H28, ’Penncross’ creeping bentgrass was killed. The turfgrass turned from deep green to mottled straw brown in only 3 days and to straw brown in 7. Turf aerial tissue was visibly stunted after only 12 hours exposure. Exposure to only water had no observable deleterious effect. Root tissue of the turf exposed to H28 was visibly stunted and necrotic compared to 33 the untreated turf. In addition, the thatch and roots exposed to H2S took on a blackened appearance characteristic of black layering. The effect of the H28 was to act as a respiratory toxin blocking electron transport at cytochrome C + C3 by binding to divalent metals such as Fe+2 (1). CONCLUSIONS This qualitative research has demonstrated a relationship between the addition of S to soil and formation of black layering, especially when soil moisture was in excess. The black layering was shown to be prevented by addition of fertilizer NO3'. The biological activity of black layer formation as 6'2 production was demonstrated using M004, a specific inhibitor of sulfate-reducing bacteria, and with NaN3, a universal toxin. The black layer reacted positively to spot testing, confirming the presence of 3'2 in the experimentally produced black layer. Finally, it was shown that H28 was toxic to creeping bentgrass. From these results we may conclude that black layering is biological in nature and involves the anaerobic chemistry of S. Further, it is possible that H28 may be involved in the decline of turfgrass which frequently accompanies black layer. Finally, an acceptable control measure in the field may be as simple as applying adequate fertilizer NO3-. 1) 2) 3) 4) 5) 6) 34 REFERENCES CITED Atlas, R.M., and R. Bartha. 1981. Microbial ecology: fundamentals and applications. Addison-Wesly, Reading, MA. Berndt, W.L., J.M. Vargas, Jr., A.R. Detweiler, P.E. Rieke, and B.E. Branham. 1987. Black layer formation in highly maintained turfgrass soils. Golf Course Mgt. 55: 106-112. Biebl, H., and N. Pfennig. 1977. Growth of sulfate-reducing bacteria with sulfur as the electron acceptor. Arch. Microbiol. 112: 115-117. Burpee, L.L., and A. Anderson. 1987. The cause of black layer in golf greens: an alternative hypothesis. Greenmaster 23: 24. Connell, W., and W. Patrick. 1969. Reduction of sulfate to sulfide in waterlogged soils. Soil Sci. Soc. Am. Proc. 33: 711-715. Cord-Ruwisch, R. 1985. A quick method for the determination of dissolved and precipitated sulfides in cultures of sulfate-reducing bacteria. Journal of Microbiological Methods 4: 33-36. 7) 8) 9) 10) 11) 12) 13) 14) 15) Engler, R., and W. Patrick. 1973. Sulfate reduction and sulfide oxidation in flooded soil as affected by chemical oxidants. Soil Sci. Soc. Am. Proc- 37: 685-688. Feigl, F. 1972. Spot tests in inorganic analysis. Elsevier, Amsterdam. Goss, R.L. 1987. Managing anaerobic soils. Northwest Turfgrass Topics 30: 7. Hillel, D. 1982. Introduction to soil physics. Academic Press, New York. Hodges, C.F. 1987. Blue-green algae and black layer: Part I. Landscape Mgt. 26: 38—44. Scott, J. 1986. The black plague. Golf Course Mgt. 54: 58-64. Segel, I.H. 1976. Biochemical calculations. John Wiley and Sons, New York. Smith, E. 1983. Principles of biochemistry. McGraw-Hill, New York. Smith, R. 1981. Sulfate reduction in the sediments of a eutropic lake. Dissertation for the degree of Ph.D. Michigan State Univ. East Lansing, MI. 36 16) Vamos, R., and E. Koves. 1971. Role of the light in the prevention of the poisoning action of hydrogen sulfide in the rice plant. J. Applied Ecology : 519-525. CHAPTER III INFLUENCE OF SULFUR AND LACTATE ON IN VITRO SULFIDE ACCUMULATION AND REDOX POTENTIAL IN FLOODED SAND Abstract. Qualitative research into turf black layer suggested it was an accumulation of 8’2. Other researchers reported that 8’2 production in rice paddy soil was dependent on the concentration of organic C present, and the redox potential of the surroundings. The purpose of this study was to determine how 8'2 production and pE + pH in flooded sand of low organic matter content was affected by adding different levels of S and C3HSO3Na. Sulfide production in response to S was dependent on concentration of lactate and vice versa. Results indicated that S was limiting for 8'2 accumulation, not organic C. Results also indicated that only when S or S + lactate were added did pE + pH become significantly depressed. It was concluded that soil S, organic matter status, and pH + pH dictated [S72] hence black layer potential. Black layer appearing in turfgrass soils was suggested to result from accumulation of metal sulfides (2, 3, 9, 10). Rankin (9) reported incidence of FeS and MnS in the analysis of a New Zealand black layer deposit. He concluded that the black deposits were the end result of reactions between S, organic matter and waterlogging. Smith (10) reported the occurrence of a sulfidic precipitate resulting from anaerobic bacterial activity in Saskatchewan turfgrass soils after waterlogging. He concluded that S seemed to be the key element in this black layer formation 37 38 and suggested that other factors leading to the development of the black layer were heavy rainfall, waterlogging and perching of water tables. Hodges (5, 6, 7) hypothesized that algae may induce a perched water table in high sand content soils which receive frequent, close mowing. He also reported a relationship between Desulfovibrio spp., and Qecilatoria or Nostoc spp. in forming a black precipitate in high sand content cultures in vitro. Only when both algae and sulfate—reducers were present did the black precipitate quickly form. He reported that organic substances produced by the algae probably helped to form the perched water table, generate anaerobic microsites, and supply growth substrate conducive to activity of sulfate-reducers. The objective of this research was to test the hypothesis that addition of S and C3H503Na to low organic matter content flooded sand would increase the [8'2], and decrease the redox potential, thereby enhancing the black layer formation potential. MATERIALS AND METHODS Experiments ylpp Sulfur and Lactate Two hundred grams washed Lake Michigan dune sand (99% > 0.1 mm and 84.5% > 0.25 mm, 0.0125% cm) placed in 125 ml glass serum bottles (Wheaton) was treated with flowable 52% S at rates of 0.0, 73.2 and 146.4 kg ha-l. Experimental units were then flooded so that no headspace existed with either tap water, or C3H503Na solution at either 112 or 1,120 mg kg'l. Sulfate content of the tap water was 30 ug ml'l. Units were then sealed with butyl rubber stoppers, crimped with aluminum, and then 39 incubated for 21 da in the dark at 30 C. Experimental design was completely randomized design and the treatments were arranged factorially with three replications. At the end of 21 days solution from each unit was sampled first for free H25 and 804-2. Two mls of sample solution were obtained by displacing headspace solution with 02 free N2 using 50c syringes. The syringe to receive the solution was equipped with a 0.22 um syringe filter to exclude particulates. In this way headspace solution from each experimental unit was sampled from completely filled vessles without contact from atmospheric 02 or interference from sand debris. To measure H28, 0.4 ml aliquots of sample drawn from the receptor syringe with a 1 cc tuburculin syringe were injected into 4 mls moving HCl:CuSO4 reagent (4) in a 10 cc cuvette. Absorbance was then determined at 480 nm with a Spectronic 20 (Milton-Roy). Reagent without sample served as a blank. The A between blanks and water at 480 nm was zero. A calibration curve made with dilutions of NaZS served as a S"2 concentration reference. Sulfate analyses were performed with remaining sample using a low pressure liquid anion chromatograph (Dionex). Next, the crimp and butyl stoppers were removed under a constant stream of N2 and the pE + pH of headspace solution was determined. Hydrogen ion concentration was measured with a IAg/AgCl combination electrode (Fischer Scientific) and pE determined using a Pt redox electrode with a calomel reference (8). A pH/mV microprocessor (Orion model 811) was used in conjunction with both electrodes. Lastly, 2 ml de-aired water was added to each bottle, the butyl stoppers were replaced and 40 re-crimped, and 2 ml 37% HCl was injected into each unit using another syringe as a vent. This step solublized any acid volatile 8’2, which was subsequently measured colorimetrically in HCl:CuSO4 reagent in the same fashion as H28. RESULTS AND DISCUSSION Experiments with S epg Organic Matter After 21 days the soil in experimental units was observed to be in a darkened state. Visually, it appeared that the soil was darker where S and lactate were added. When treatment effects were analyzed it was found that there were significant statistical interactions for the 8'2 measurements. Thus, response to S was dependent on level of lactate and vice versa. The effect of the S was to provide an abundance of electron acceptors to be reduced while the lactate provided necessary reducing equivalents. Adding S at 73.2 kg ha.1 increased the [H28] to a detectable level, and dramatically increased the acid volatile 8-2 (AVS) concentration compared to a no lactate, no S treatment ( see Table 1.). Increasing the level of S to 146.4 kg ha“1 did not influence [H28] but did cause [AVS] to increase slightly, although not significantly. Adding lactate at any level without S did not result in significant H28 or AVS accumulation compared to the no lactate, no S treatment. Adding S at 73.2 kg ha.1 with lactate at 112 or 1,120 mg kg-1 caused no increase in [H28] but did increase [AVS] compared to adding only 8 at 73.2 kg ha'l. Increasing S to 146.4 kg ha"1 Table 1. flooded Lake Michigan dune sand. -1 kg solution. flowable formulation at 0.0, 73.2 or 146.4 kg ha- incubated in the dark for 21 da at 30 C. 41 Influence of S and lactate on S- production and redox parameters in Lactate was added at 0, 112 or 1,112 mg C3H503Na Tap water was used to flood the sand. 1 Sulfur was added as a 52% Experimental units were Means were the averages of 3 reps. Treatment Sulfur Compoundsz Redox Parametersy Lactate 5 H2 5 AVS so 4'2 pE pE+pHx pH mg/kg kg/ha mg/kg 0.0 0.0 0.0 c* 5.1 e 19.1 e -0.5 a . 7.4 a 73.2 4.2 c 46.7 d 27.4 cd -2.2 bc . 7.0 cd 146.4 3.8 c 60.2 cd 33.6 ab -2.1 bc . 6.9 d 112 0.0 0.0 c 5.1 e 3.7 f -0.8 a 6.5 7.3 b 73.2 5.2 c 74.8 be 28.8 bc -2.2 bc . b 6.9 e 146.4 18.3 b 65.5 c 34.4 a -2.3 be 4.4 be 6.7 f 1,120 0.0 0.7 c 6.1 e 1.1 f -0.8 a 6.2 7.0 cd 73.2 2.7 c 86.7 ab 22.5 de -l.8 b 5.0 6.8 f 146.4 31.2 a 99.7 a 34.6 a -2.6 c 4.0 c 6.6 g LSD 05 = 9.1 17.7 7.5 0.6 0.6 0.05 2 mg S kg.l soil solution as free H28, acid volatile S—2 (AVS), or 504 Y pH measured with a Ag/AgCl combination electrode. y pE = Eh/59.2 measured with Pt electrode and calomel reference. x Denotes addition of pE and pH values. * Means with similar letters within columns are not different by LSD @ P = 0.05. 42 with 112 mg kg'1 of lactate resulted in significant jump in [H28] but increased the [AVS] only slightly. Increasing S to 146.4 kg ha'1 with 1,120 mg kg'1 lactate increased both [H28] and [AVS] significantly. Significant linear trends (see Table 2.) across S levels were detected for H28 appearance where lactate was added. Increasing lactate to 1,120 mg kg'1 resulted in the detection of a significant quadratic response. Significant linear responses were detected for all 3 lactate levels regarding AVS. Where lactate was added at 112 mg kg'l, and when it increased to 1,120 mg kg’1 significant quadratic responses were detected. It appeared that free H28 generation in response to S was limited by availability of S, and somewhere between the 73.2 and 146.4 kg ha'1 S levels lay a critical S application value above which free H28 was readily generated provided organic matter was abundent. If that threshold S application level was reached, increasing organic matter increased the [H28]. If the threshold was not reached additional organic matter did not cause an increase in [H28]. For AVS formation in response to S application, it again appeared that S was limiting. For a given level of organic matter, if sufficient S was present AVS readily accumulated. Without 8 application little AVS accumulated even though organic matter was abundent. However, increasing the level of applied S from 73.2 to 146.4 failed to produce a significant increase in [AVS] concentration. Thus, once S was added the organic matter content dictated the S-2 intensity. If S was not added then organic matter content did not matter. 43 Table 2. Trend analysis summary for lactate treatments across 3 levels. Lactate 1 solution. Sulfur was added as a was added as C3HSO3Na at 0, 112 or 1,112 mg kg- 52% flowable at 0.0, 73.2 or 146.4 kg ha-l. Values for SSQ were calculated with treatment means. Means were the values of three replications. Sulfur Compoundsz Redox Parametersy Source df H28 AVS 304"2 p8 pE+pHx pH Reps 2 S Level 2 0.0 Linear l 1 ns 43 ** 38 ** 31 ** 52 ** 375 ** Quadratic l 1 ns 4 ns 1 ns 13 ** l7 ** 45 ** 112 Linear 1 18 ** 52 ** 168 ** 27 ** 52 ** S40 ** Quadratic l 1 ns 29 ** 23 ** 7 * 9 ** 20 ** 1,112 Linear 1 50 ** 124 ** 200 ** 40 ** 57 ** 240 ** Quadratic l 13 ** 22 ** 5 * 1 ns 1 ns 1 ns Error 18 *, ** Denotes significance at P = 0.05 and P = 0.01 respectively. 2 2 2 Sulfur as free H28, as acid volatile S— (AVS), or as sulfate-S (804- ). y pH measured with a Ag/AgCl combination electrode. y pE = Eh/59.2 measured with a Pt electrode and calomel reference. x Denotes addition of pE and pH values. 44 No statistical interactions were detected regarding pE + pH. Adding lactate without 8 at 112 mg kg"1 tended to depress pE + pH, and increasing lactate concentration to 1,120 mg kg"1 tended to increase the depression (see Table 1.). However, the change in pE + pH between lactate levels without S was subtle and was not statistically significant. Adding S at 73.2 kg ha'1 without lactate depressed the pE + pH significantly compared to the no S control, and compared to the lactate only treatments. However, increasing the level of S without lactate did not result in a greater pE + pH reduction. The only additional reduction in pE + pH due to increasing level of S occurred at the 1,120 mg kg-1 lactate level. Organic matter at high concentrations might be enough to depress the pE + pH to such a level that intense 8'2 production occurs but this was not accomplished in the present study. Addition of S at near reccommended rates did depress pE + pH to the point where reduction was iminent. Reduction of pE + pH with lactate was assumed to involve increasing bacterial respiration rates so that 02 was eventually consumed. This assumption ws not unreasonable due to the fact that our experimental units were sealed, and 02 diffusion was stopped. Occurrence of lactate 1 situ is probably slight but occurrence of sloughed root mass in turfgrass soils should accomplish the same end, while also providing a source of S. Reduction of pE + pH with S probably involved the oxidation of S to 804'2 consuming 02 in the process: S "I" 3/2 02 + H20 < ----- > H2804 45 This would also account for the observed pH reduction (see Table 1.). As S application to lower soil pH is a common cultural practice in turfgrass management, this too may impact the aeration status of treated soil, especially during rainy periods, to the point where 304'2 reduction occurs. Thus, addition of lactate and S led to conditions environmentally conducive to 8’2 accumulation, low redox potential. CONCLUSIONS The results of this study suggested that S was limiting for 8'2 production in our experimental soil. Addition of lactate increased the 8'2 production only where [S] was sufficient; where interactive effects were operative more 8"2 was produced. Compared to lactate, addition of S also appeared to reduce redox potential to a greater extent. However, the combined effects of S and lactate produced the lowest redox. Thus, where S and soluble organic debris are plentiful, and 02 diffusion or availability is retarded for whatever reason, potential of black layer 8'2 produCtion is enhanced. l) 2) 3) 4) 5) 6) 7) 8) REFERENCES CITED Atlas, R.M., and R. Bartha. 1981. Microbial ecology: fundamentals and applications. Addison-Wesly, Reading, MA. Berndt, W.L., J.M. Vargas, Jr., A.R. Detweiler, P.E. Rieke, and B.E. Branham. 1987. Black layer formation in highly maintained turfgrass soils. Golf Course Mgt. 55: 106-112. Berndt, W.L., and J.M. Vargas, Jr. 1989. Sulfur, organic matter and black layer: Part I. Golf Course Mgt. 57: 80-84. Cord-Ruwisch, R. 1985. A quick method for the determination of dissolved and precipitated sulfides in cultures of sulfate-reducing bacteria. Journal of Microbiological Methods 4: 33-36. Hodges, C.F. 1987. Bluegreen algae and black layer: part I. Landscape Management 26: 38-44. Hodges, C.F. 1987. Bluegreen algae and black layer: part II. Landscape Management 26: 30-31. Hodges, C.F. 1989. Another look at black layer. Golf Course Mgt. 57: 54-58. Lindsay, W.L. 1979. Chemical equilibria in soils. John Wiley and Sons, New York. 9). 10) 11) 47 Rankin, P.C. 1988. When the black layer hit the fan. Turf Craft Aust. 6: 18-20. Smith, J.D. 1988. Black plug layer on Saskatchewan golf courses. Greenmaster 8: 6-21. Steel, R.G.D., and J.H. Torrie. 1980. Principles and procedures of statistics. McGraw-Hill, New York. CHAPTER IV INFLUENCE OF SULFUR, AMMONIUM AND NITRATE ON IN VITRO SULFIDE ACCUMULATION AND REDOX POTENTIAL Abstract. Previous research has suggested that black layer in turfgrass soils was an accumulation of 8’2, and that 8 addition positively influenced 8'2 formation. Fertilizer N03- was also previously shown to prevent 8'2 accumulation presumably by keeping redox potential elevated. Whether other N sources would prevent 8‘2 accumulation was unknown. Thus, the influence of S, NH4+, and N03" on 5'2 accumulation and redox potential in flooded soil from a 'Penncross' creeping bentgrass green was studied in the laboratory. Addition of flowable S to flooded soil at levels of 48 kg ha"1 caused an increase in free H28 and in acid volatile 5"2 compared to other treatments. Concurrent addition of N as NH4+ at 48 kg ha"1 retarded acid volatile S'2 accumulation but did not prevent formation of free H28. Nitrate was found to be considerably more effective at retarding 8"2 production, as no H28 was detected where NO3' was added, and AVS production was delayed by 6 days compared to other treatments. Nitrate was also shown to prevent redox potential depression where NH4+ did not. It was concluded that S contributed to and N retarded 8'2 formation. It was also concluded that a critical pE + pH was necessary for 8’2 formation, and that sufficient N037 from fertilization may help to prevent 8’2 production in turfgrass soils by maintaining a pE + pH above the critical point. 48 49 Black layer formation in highly maintained turfgrass soils has been related to an accumulation of S"2 in the profile ( 1, 9, 11). A major effort in black layer research was to find acceptable methods of black layer control which could be used by turfgrass practitioners. Core aerification, reconstruction of afflicted areas, installation of additional drainage and fungicide application have been used to combat black layer, all with limited success at best. Sulfide formation in flooded rice (Qxyxe sp) paddy soils was shown to be dependent on redox potential ( 2, 4, 5, 7, 14). Connell and Patrick (2) reported little 8'2 production in rice paddy soils with a redox potential above -150 mV. Likewise, Engler and Patrick (4) reported that 804"2 reduction and 8'2 oxidation in rice paddys occurred in the vicinity of -100 mV. However, Vainsthein (14) reported that a $04-2 reduction process and sulfidic compounds (pyrite) were detected in Indian Ocean sediments with redox potential near +500 mV. In experiments with an apparatus to modify redox with active H2, Vainsthein reported active SO4-2 reduction by Desulfovibrio baculatus at +120 mV. He reported that redox status probably affected SO4-2 reducer type more than the actual process. Thus, depending on the type of SO4-2 reducers present, 8'2 production may take place over a wide range of redox potentials. The effects of various chemical oxidants to inhibit 8'2 formation in flooded rice paddy soils were also studied (4, 7). It was found that if redox potential was kept high, 8'2 production was delayed and suffocation disease of rice was prevented. Results further indicated that oxidants which were 50 soluble, such as N03” and ferriocitrophosphate, had the greatest effect in maintaining high redox potentials, hence in preventing 8'2 production. Less soluble oxidants such as MnOz and FePO4 were not as effective but persisted longer. The objective of this research was to determine the effect of added S, NH4+ and N03' on the redox potential and 8’2 formation tendencies of flooded turfgrass soil. MATERIALS AND METHODS Thirty grams of sand (88.3% sand, 5.0% silt and 6.7% clay) from a 'Penncross’ creeping bentgrass (Agrostis palustris Huds.) golf green located at the Hancock Turfgrass Research Center in East Lansing was added to 30 ml glass anaerobe tubes (Bellco Glass). Soil was collected to a depth of 15 cm, then air dried, homogenized, and sieved to 500 um. Root tissue retained on the sieve was ground to 20 mesh with a Wiley mill and added back to the soil and mixed. Soil material not passing 500 um was discarded. The composite soil had a C:N of 11.4:1. Most probable number estimates (12) of sulfate-reducing bacteria in the composite soil incubated in a flooded state for 2 weeks were 1.7 x 106 SRBs per gram of soil. Where 8 was added to this soil at 48 kg S ha"1 the SRB numbers increased to 2.2 x 106. Tubes containing soil were then flooded with tap water so that the soil was entirely under water, and flowable S (52%) was added at a level of 48.8 kg S he’l. Sulfate content of the tap water was approximately 30 ug ml'l. Flooded tubes to which no 8 was added were left as controls. Tubes were not shaken or inverted 51 to mix the soil and S. Mixing was averted so that S would come to rest on the flooded soil surface simulating field application. One third of the S treated tubes also received NH4+-N from NH4Cl 1 while another third in solution at a level of 48.8 kg N ha’ received NO3'-N from Ca(NO3)2 also at 48.8 kg ha'l. There were three replications of each treatment (i.e., no S, 8 only, S + NH4+ or S + N03”). Sufficient tubes of each treatment were prepared to enable daily analysis of eaCh treatment for 18 days including day 0. After treatment with S and N, additional tap water was added to each tube so that no headspace existed. Each tube was stoppered with butyl rubber and crimped with aluminum seals to halt 02 diffusion and initiate anaerobiosis. Microbial respiration was assumed to initiate the anaerobiosis. Three tubes from each treatment were analyzed daily for accumulation of free H23, AVS, and redox potential as pE + pH (6). Measurement of free H28 using the method of Cord-Ruswich (3) was performed first, by withdrawing 2-3 mls of soil solution from closed tubes with a syringe. To do this a de-aired 5 ml syringe was inserted through the stopper so that the needle opening was under water. Next another syringe containing N2 was inserted and the N2 was injected to displace tube water, which entered the first syringe through a 0.2 um filter. In this way the filtered soil solution was sampled without interference from atmospheric 02' Stoppers were then removed under a stream of 02-free N2 and pH and Eh were determined. Soil solution reaction was measured with a Ag/AgCl combination electrode and Eh of surface soil measured with a Pt electrode with calomel reference. A salt bridge was necessary to measure Eh. An Orion 811 —_— 52 microprocessor served both electrodes. Lastly, AVS was measured by replacing withdrawn water with de-aired distilled water, re-stoppering then subsequently injecting 2 ml 37% HCl. The acid solubilized AVS which was then measured in the same way as H28. RESULTS AND DISCUSSION Soil treated with S produced more acid volatile 8'2 (AVS) after 18 days than a control soil (see Fig. 1). A final 8'2 concentration of 885 mg kg"1 for soil treated with 8, compared to a concentration of 165 for the control, represented a 8"2 increase of 536%. At the end of the experiment the AVS in tubes where S was added was visible as intense black banding beginning at the soil/water interface. Banding extended from the interface to 4-5 mm below the soil surface. No banding appeared below 5 mm. Soil past that depth had a gleyed appearance characteristic of reduced soils. Likewise, in control units no banding occurred, but the gleyed appearance was evident. Upon closer examination of the soil in control tubes, and in soil in S treated tubes below the banding, distinct semi-micro areas of intense blackening were also observed. These were probably formed in response to proximity to organic debris. Concurrent addition of N as either NH4+ or N03- with S retarded the production of AVS. This response was expected with addition of NO3' as it is a well established alternate electron acceptor (2, 12). The standard free energy of reduction of N03- was reported to be -53.6 kcal mol'1 H2 consumed, while that of SO4-2 or S was -9.1 and -6.7 respectively (12). Thus more 53 Figure 1. Influence of S, NH4+, and N03- on accumulation of AVS in flooded soil from a ’Penncross’ creeping bentgrass golf green. Sulfide production was enhanced by addition of S while concurrent addition of N as either NH4+ or N03- retarded it. Both S and N were applied at 48 kg ha'l. Means were the averages of 3 observations per day over 18 days. 54 Ne Or w w .v N O FLPL t L 933 m2: I—i I» \I- L Ir 0 \\ a x V m. noom D. A . .o. m. E u seems we do ..ooe m S a a [com m 21.02 + m I T. y. 214:2 + m I 5.1 m I!- room ( m OZ I 55 Table 1. Linear regression statements applied to acid volatile 8'2 accumulation as influenced by S, NH4+, or NO3'. Regressions were calculated by the method of least squares. Regression Parameters Treatment R2 Slope Intercept No S 0.93** 12.41 -44.18 S 0.97** 69.25 -367.42 S + NH4-N 0.97** 40.27 -193.61 S + NO3-N 0.96** 20.67 -208.78 *,** Denotes significance at P = 0.05 and P = 0.01 respectively. Strength of association. Ordinate intercept. 56 metabolic energy for biomass would be derived from N03- metabolism compared to metabolism on SO4-2 or S. In other words SO4-2 or S reduction (hence AVS formation) should have been "by-passed" in the presence of NO3' until the available N03” pool was depleted. After that 804"2 or S reduction should proceed unimpaired. Indeed, AVS formation in the tubes treated with N03- was delayed by 6 days compared to tubes treated with only S. After that time AVS formation proceeded at a relatively quick pace. Likewise, Engler and Patrick (4) and Connell and Patrick (2) reported that addition of N03" to flooded rice paddy soil delayed 804-2 reduction but did not decrease the rate of 8'2 formation once it commenced. The response to NH4+ was more intriguing. It was initially assumed that the highly reduced NH4+ would help to generate more AVS than S alone by providing the preferred N source for sulfate-reducing bacteria (8). Probably what happened was a portion of the NH4+ oxidized to NO3- in the first few days of the experiment effectively increasing the N03' concentration (10). Although N03- content was not measured in this study this conclusion seemed reasonable since highly aerobic tap water was used to flood the tubes, and nitrifying bacteria should be present in the soil. Regression analysis of the AVS accumulation curves (see Table 1.) indicated the production of AVS to be highly linear (i.e., R2 > 0.9 for all curves) with day 4 as the origin. The slope of the S treatment was the steepest followed by the NH4+ curve, the N03- curve and finally the control. 57 Free H28 formation was stimulated by addition of S or S + NH4+ (see Fig. 2). It was not detected where S was not added, or where N03“ was added concurrently. Free H28 was first detected on day 11, some 7 days after the formation of AVS was noticed. Once production began the rates of accumulation were similar for those treatments. Linear regression analysis indicated a slope of 15(x) for both the S and S + NH4+ treatments, and a slope of 0 for remaining treatments. This result implied that potential for turf plant loss due to stress imposed from the presence of H28 was heightened if S was in abundance or if ammonical fertilizers were applied with 8. Results also implied that S and not organic matter was limiting for production of free H28. This would concur with results generated in Chapter III where it was determined that a critical level of S was still necessary to generate free H28 when organic matter was not limiting.. Depression of redox potential as pE + pH was nearly linear for the first 3 days of the study (see Fig. 3). The pE + pH declined from near 16 to near 7 in 18 days_regardless of treatment. The soil treated with only 8 appeared have a slightly lower pE + pH initially (i.e., day 3) but the effect was short lived as no differences between the S, no 8, or S + NH4+ treatments were evident at day 5. A possible explanation for this initial difference may have been due to the oxidation of S to 804"2 consuming O2 in the process. Where N03“ was added with 8, either concurrently or from NH4+ oxidation, sufficient 02 was probably present to offset the effect. Where no 8 was added no scavenging of the 02 occurred. No other differences were detected between these treatments during the rest of the course of the study. 58 Figure 2. Influence of S, NH4+, and N03“ of accumulation of H28 in flooded soil from a 'Penncross' creeping bentgrass golf green. Free H28 production was enhanced by addition of S or S + NH4+ while addition of N03" prevented it. Also, no free H28 was detected where S was not added. Both 8 and N were applied at levels of 48 kg ha-l. Means were the averages of 3 observations per day for 18 days. 59 we m: r: 233 m2: me E m c a m o L _ a I F I O WOZ I m Ill r 21.22 + m I low 21.02 + mi... . .9. row row no? 60 Figure 3. Influence of S, NH4+, and N03' on redox potential as pE + pH in flooded soil from a 'Penncross’ creeping bentgrass golf green. Redox potential was calculated with a Pt redox electrode and Ag/AgCl pH electrode. Only N03- prevented the depression of redox which appeared to control 8’2 production. Means were the averages of 3 observations per day for 18 days. 61 $.63 22: we 3 Na N; or m _ e L _ e L _ so n 3.68 a do i: ©_ Hd+3d 62 Figure 4. Influence of S, NH4+, and N03_ on pH of flooded soil from a ’Penncross’ creeping bentgrass golf green. Soil reaction was measured with a Ag/AgCl combination electrode. Means were the averages of 3 observations per day for 18 days. 63 $33 22: mr © _ ewe Ne O_. m w .v N O .—I b b b LI b _ F — l— I—I IF P hi — L) + m m OZ I m Ill.- ncco n .55 zlfz + m I mam mmco n E: 21.02 + m I ,. .1 ms. to 850m rm to u 3&8 2 So 7. lm'l.‘ l .v III). ‘HI N. ’l‘lflfilurnfl '/ / (Hd) uonooea HOS 64 Only the N03” treatment significantly retarded pE + pH depression. The effect of the NO3_ was evident early on and lasted several days. At day 8 however, the effect of added N037 waned, and pE + pH dropped to 7.5 by day 12. By day 18 no differences between any treatments were apparent. The decrease in pE + pH to near 9 at day 10 for the S + N03" treatment coincided with the appearance of AVS for that treatment. Thus it appeared that a pE + pH of near 10 or above was sufficient to prevent sulfide formation. For treatments where pE + pH values were less than that, 8‘2 formation was imminent. Likewise, for the other treatments lowering of pE + pH to near 9 on or around day 4 also coincided with production of 8'2. Thus, it appeared that redox was a controlling factor in the production of 8'2. CONCLUSIONS This research has demonstrated that S intensified 8-2 production and N retarded it. Nitrate was shown to prevent 8'2 production by acting as an alternate TEA while also keeping redox above a critical level. The effect of NO3' was short lived and 8'2 production was probably initiated only after the pool of N03- was converted to N2. Thus, it appeared that N03" was a potential control for 8'2 based black layer. ACKNOWLEDGMENTS The authors would like to acknowledge B.G. Ellis for the use of the Orion microprocessor. 1) 2) 3) 4) 5) 6) 65 REFERENCES CITED Berndt, W.L., J.M. Vargas, Jr., A.R.Detweiler, P.E. Rieke, and B.E. Branham. 1987. Black layer formation in highly maintained turfgrass soils. Golf Course Mgt. 55: 106-112. Connell, W., and W. Patrick. 1968. Sulfate reduction in soil: effects of redox potential and pH. Science 159: 86-87. Cord-Ruwisch, R. 1985. A quick method for the determination of dissolved and precipitated sulfides in cultures of sulfate-reducing bacteria. Journal of Microbiological Methods 4: 33-36. Engler, R., and W. Patrick. 1973. Sulfate reduction and sulfide oxidation in flooded soil as affected by chemical oxidants. Soil Sci. Soc. Am. Proc. 37: 685-688. Goldhaber, M., and I. Kaplan. 1975. Controls and consequences of sulfate reduction rates in recent marine sediments. Soil Science 119: 42-55. Lindsay, W.L. 1979. Chemical equilibria in soils. John Wiley and Sons, New York. 7) 8) 9) 10) 11) 12) 13) 66 Ponnamperuma, F.N., W.L. Yuan, and M.T. Hung. 1965. Manganese dioxide as a remedy for a physiological disease of rice associated with reduction of soil. Nature 207: 1103-1104. Postgate, J. 1951. The reduction of sulfur compounds by Desulfovibrio desulfuricans. J. Gen. Microbiol. 5: 725-738. Rankin, P.C. 1988. When the black layer hit the fan. Turf Craft Aust. 6: 18-20. Reddy, K.R., and W. H. Patrick, Jr. 1975. Effect of alternate aerobic and anaerobic conditions on redox potential, organic matter decomposition and nitrogen loss in a flooded soil. Soil Biol. Biochem. 7: 87-94. Smith, J.D. 1988. Black plug layer on Sasketchewan golf courses. Greenmaster 24: 6-21. Smith, R. 1981. Sulfate reduction in the sediments of a eutropic lake. Dissertation for the degree of Ph.D. Michigan State Univ. East Lansing, MI. Steel, R.G.D., and J.H. Torrie. 1980. Principles and procedures of statistics. McGraw-Hill, New York. 67 14) Vainshtein, M., and G. Gogotova. 1987. Effect of redox potential of the medium on sulfide production by sulfate-reducing bacteria. Mikrobiologiya (Translated) 56: 31-35. CHAPTER V A BLACK LAYER FIELD EXPERIMENT Abstract. Laboratory research has shown a relationship between S addition and S'2 formation in flooded turfgrass soils. Nitrate was shown to prevent 8‘2 formation in these soils by keeping redox potential elevated. Whether S addition would stimulate 8'2 formation lp elpp, or whether N037 would act as a control lp elpp was unknown. Thus, influence of S level and N source on S'2 formation in a ’Penncross' creeping bentgrass golf green was investigated. Sulfur was applied monthly to plots at levels of 0, 48 and 244 kg ha—l. Nitrogen as either NO3--N or organic-N was applied weekly to plots at a level of 12 kg ha'l. Plots were saturated via irrigation on a daily basis. Plots were sampled at the end of the season at depths of 2.5, 5.0 and 10.0 cm to determine the total [S72] and pH of the soil. Differences in [872] between the control plots and those receiving 48 kg S ha.1 were slight at all depths. However, where S was applied at 244 kg ha—1 routinely, significant S"2 accumulated compared to other plots except at 10 cm. Nitrate reduced 8'2 accumulation near the surface of plots receiving heavy S applications. It was concluded that if enough S and water was applied significant S-2 accumulation resulted. Nitrate application was also somewhat effective in preventing accumulation of S"2 l situ, but levels higher than those used in the study were probably needed. 68 69 Black layer formation in turfgrass soil was reported to involve production of S72 in response to low redox potential (2, 7, 9). Previous black layer 8"2 research (i.e., Chapter I, II, III, and IV) suggested that S'2 accumulation was controlled by redox, [S], concentration of organic C, and presence or absence of N03”. This research was initiated under laboratory conditions. Whether S and N would affect black'layer 8'2 formation lp elpp was unknown. Thus, the following research was conducted to determine the effects of 8 level and N source on black layer 8'2 production in ’Penncross' creeping bentgrass (Aqroetis paluetrie Huds.). MATERIALS AND METHODS 1987 Research. This experiment was begun in 1987 at the Robert Hancock Turfgrass Research Center in East Lansing. The objective was to determine whether 8 application would intensify black layer 1 situ. The study was conducted on ’Penncross’ creeping bentgrass (Agrostis palustris Huds.) mowed approximately 3 times weekly at 0.6 cm. The area was previously maintained in a manner suitable for a research golf green. Fungicides and fertilizers were applied as needed; irrigation was supplied to prevent wilt. The experimental area was not topdressed or cultivated routinely but an occasional core aerification had been previously performed. This was documented by the existence of tine holes remaining throughout the green. 70 A composite sample of soil from the 0-15 cm depth excluding thatch was 0.07% N and 0.8%C. Total N excluding N037 was determined by the micro-kjeldahl method and C was calculated by loss on ignition. Most probable number estimates (9) of a sample of the composite soil incubated in a flooded state for 2 weeks was 1.7 x 106 sulfate-reducing bacteria per gram soil. Sulfate content of the irrigation water was between 20 and 30 ug ml-1 (10). In 1987 three irrigation blocks within the green were utilized for the study. In one block irrigation was supplied at wilt at approximately 0.1 to 0.3 cm per irrigation. In a second block, irrigation was applied daily at roughly 0.3 cm per day. The third block received a saturating irrigation each day, rain or shine. For this treatment irrigation was applied daily for 30 minutes. Soil moisture ranged from near 50% immediately after watering to slightly less than 10% by weight prior to the beginning of the next cycle. This fluctuation was relatively constant except during rain. In addition plots were syringed once or twice per day during hot weather when evaporation was greatest. Within each irrigation block plots were established. The fertilizer and fungicide application schedule (performed by the maintenance crew) remained unchanged. Plot size was 2.23 m2 (4’ x 6’). Treatments consisted of flowable 52% S applied every three to four weeks at levels of 0, 48, and 244 kg S ha'l applied with FeSO4 at 0, 6 or 30 kg ha'l. Total S application for the 71 season was 288 or 1,464 kg S ha'l. In addition several control plots and extra dummy plots were established. Experimental design was completely randomized with 3 replications. Plots were visually inspected for quality throughout the season. In August, 1987 samples from each plot were visually inspected for the presence or absence of black layer. Sulfur and irrigation treatments were continued until the irrigation system was winterized in October. leee Research. In 1988, plots in the saturated block were the only ones to continue to receive treatment. This was because turf had been lost in the dry block where high levels of S, and S plus high levels of FeSO4 were applied. Turf was also beginning to thin in the high 8 level plots which received 0.3 cm water per day. No turf was lost in the saturated block. Further, the 30 kg FeSO4 treatment combinations were discontinued as this level of iron seemed unreasonable and did not visually appear to contribute to black layer intensity. In August, 1988 plot samples were harvested to attempt a quantitative black layer 8'2 analysis using the method of Cord-Ruwisch (4). Duplicate 20 g samples from the 5, 10 or 15 cm depth were placed into 30 ml test tubes under a stream of N2. Tubes containing soil were flooded with anoxic water then stoppered with butyl rubber and crimped with aluminum seals. Two mls of 37% HCl was injected into each tube, then tubes were agitated for several minutes by shaking. Samples of the acidified soil solution were filtered to pass 0.2 um then measured for 8'2 concentration (4). The aim here was to document relative differences between treatments if such existed. Thus, 72 dilution effects were not accounted for. Again, 8 and irrigation treatments were continued until the system was winterized in October. 1989 Research. In 1989, plots in the saturated block were continued but the 6 kg level of FeSO4 was discontinued. This was because no effect on black layer production attributable to FeSO4 was observed. This change gave an experimental design which was now completely randomized with 6 replications of each S treatment instead of three. Each S plot (i.e., whole plots) was also split between treatments of inorganic-N or organic-N (i.e., sub-plots). Specifically, N037 or organic-N was applied weekly at levels of 12 kg N ha'l. Nitrogen treatment sources were Ca(NO3)2 and sewage sludge. Occasionally bio-organic-N replaced sewage sludge and KNO3 replaced Ca(NO3)2 so that an overabundance of Ca+2 was not applied and so that K+ inputs would be similar. The organic sources of N were alternated so that a variety of organic components and organisms would be added. Nitrogen totals for 1989 were 216 kg N ha’l. Whole plot size was still 2.23 m2 while sub-plot size was 1.12 m2. The plots were inspected monthly for visual quality. In August, 1989 the plots were again sampled for 8'2 concentration. However, instead of weighing the soil (which probably subjected any black layer s”2 to significant oxidation) a sampling tool was constructed. The tool, manufactured from a standard cup cutter and stainless steel sampling syringe, allowed uniform volume samples to be taken at depths of 2.5, 5.0 and 10.0 cm with minimal exposure to ambient atmosphere. Duplicate samples from 73 each plot at each depth were placed into glass tubes which were flooded with known amounts of freshly boiled distilled water cooled to room temperature under N2. As the amount of water added to each tube was measured (hence the volume of soil added was known) dilution effects could be estimated. Tubes were stoppered with butyl rubber so that no headspace existed then crimped with aluminum seals. Two mls of 37% HCl was then injected into each tube, which was agitated for several minutes. Sample solution was then filtered to pass 0.2 um and measured for 8'2 concentration using the method of Cord-Ruwisch (4). In addition, the pH(KCl) for samples from each plot at each depth (plus thatch) was determined. Soil reaction of a 2 parts 0.01 M KCl to one part soil was measured with a Ag/AgCl combination electrode. The KCl was used to eliminate any junction potential (6). Visual quality and other ’surface’ data was analyzed as a CRD split-plot with 6 reps. For measurements involving depth the data was analyzed as a CRD split-split plot. Polynomial trend analysis was also performed on S"2 and PH(KC1) means (10). RESULTS AND DISCUSSION 1231 Research. It appeared that level of irrigation was the most important factor in producing black layer in 1987. Where irrigation was applied at the wilt point no black layer was observed at any point in the plots (see Table 1.). Significant burn and turf loss did appear in some plots (see Table 2.). Where 0.3 cm of water was applied daily, sporadic pockets of 74 Table 1. Influence of S and irrigation on black layer formation in ’Penncross’ creeping bentgrass for 1987. Sulfur was applied as a 52% flowable formulation. Irrigation was applied at wilt at 0.1 to 0.3 cm, daily at 0.3 cm and for 30 minutes daily to saturate the plots. Presence or absence of black layer is denoted by a plus or a minus. Treatment Wilt 0.3 gm Saturated -1 _ 48 kg S ha - + 244 kg 5 ha‘1 - - + 6 kg Fe ha.1 - - + 30 kg Fe ha‘l - - + 48 kg s + 6 kg Fe ha"1 - - + 244 kg S + 6 kg Fe ha'1 - +/- + 48 kg S + 30 kg Fe ha'l - — + 244 kg s + 30 kg Fe ha'l - +/- + control - - +/- 75 Table 2. Influence of S and irrigation on visual quality of ’Penncross’ creeping bentgrass for 1987. Sulfur was applied as a 52% flowable formulation. Irrigation was applied at wilt at 0.1 to 0.3 cm, 0.3 cm daily and for 30 minutes to saturate the plots. Higher numbers denote better turf. Scale was 9 = superior turf while 1 = dead turf. Treatment Wilt 0.3 gm Saturated 48 kg 5 ha’1 7.5 7.0 6.0 244 kg 8 ha"1 3.3 7.3 6.2 6 kg Fe ha"1 7.4 7.6 6.7 30.kg Fe be“1 7.7 7.5 6.5 48 kg s + 6 kg Fe ha‘1 7.8 7.6 6.3 244 kg s + 6 kg Fe ha'1 7.4 7.6 6.9 48 kg s + 30 kg Fe ha“1 3.6 7.0 5.2 244 kg s + 30 kg Fe ha"1 4.5 6.8 5.4 control 7.6 7.7 6.5 118130.05: 0.8 0.8 1.8 76 black layer appeared but only in the plots treated routinely with 244 kg S ha'1 plus iron. Where turf was saturated on a daily basis black layer appeared in all plots regardless of treatment. Black layer was a type 1 layer (3) which was visually most intense near 5 cm. No black layer appeared in the thatch layer nor much below 15-20 cm. Black layer appeared just under the thatch/soil interface, and had a distinct upper margin. It was uniform throughout the entire plot area which received irrigation. In addition, the black layer spread outward from the plot area several yards in all directions. Subsequently, adjacent plot areas, hence other researchers experiments were probably affected by black layer to some degree. The movement was presumed due to lateral water flow. No visual differences in black layer intensity were observed between treated plots in the saturated block except the control. Black layer in one of the control plots was variable and existed as coalesced pockets further down in the profile. As the black layer existed in the 0.3 cm daily irrigated plots receiving high levels of S plus iron, and black layer was variable in the check plots receiving saturating irrigation, it appeared that supplemental 8 could have been at least a contributing factor to black layer development. Spot testing (5) confirmed the presence of metal 8’2 in the experimental black layer. Soil without black layer did not react to the spot test. If exposed to atmosphere for several hours the black layer disappeared and soil regained a more characteristic earthy appearance. Black layer was not present immediately adjacent to any aerification tine holes remaining in the green. 77 The boundary dividing black layer and no layer here was a distinct gradient with out a distinct margin. If acidified with HCl black layer also disappeared. In addition the black layer had the distinct aroma of rotting eggs (i.e., reduced S compounds). Thus, it was concluded that adequate water and S, whether supplemental or naturally occurring, contributed to the formation and accumulation of metal 8'2 manifested as a black layer. The formation of S"2 is biological in nature (1). Most probable number estimates of sulfate-reducing bacteria (9) were between 2.3 x 102 and 3.3 x 103 SRBs per gram of black layer soil. Thus, it was further concluded that 804'2 (or S) reduction was ultimately responsible for formation of the S'2 in the experimental black layer. Turf growing on black layer was healthy and alive. Visual quality of the turf was, however, affected by irrigation and S (see Table 2.). Sulfur applied to plots irrigated at wilt at 244 kg ha'l, or at both 48 and 244 kg ha'1 plus iron at 30 kg ha-1 contributed to loss of turf by late August. Excessive irrigation also depressed overall turf quality, compared to turf receiving favorable irrigation. The effect of the irrigation was to restrict 02 diffusion creating anaerobic or anoxic conditions. This condition probably restricted oxidative processes in plant roots, hence the decline in quality. Additionally, the generation of sulfides which are toxic to plants may have impacted quality but this was not proven. 78 12§§ Research. In 1988 it was found that S application to turf soil receiving saturating irrigation on a daily basis contributed to increases in black layer [S72] (see Fig. 1.). However, S totals for the 244 and 48 kg S ha'1 plots by the end of 1988 were 2,928 and 576 kg S ha'l respectively. This translates into roughly 60 and 12 pounds S applied per 1000 square feet in two seasons respectively, very excessive rates to the practitioner. The differences were not readily apparent visually, but more $72 existed where 244 kg S ha'1 was applied routinely compared to where 48 kg was applied except at the 15 cm depth where no differences occurred. More S"2 also existed at the 5 cm depth where 48 kg S ha"1 was applied compared to control plots. Differences in S'2 by depth may be due to differences in soil organic matter at those depths. Measurements were not taken but visual observation suggested there was less rooting (hence sloughed organic debris) at 15 cm than at 10 cm. Further, there appeared to be less rooting at 10 cm compared to 5 cm. Previous research (Chapter III) suggested that an abundance of organic matter was necessary to manifest intense S'2 accumulation. If one was to assume that sloughed root tissue was the e' donor for $04—2 reduction, and an abundance was necessary to result in active accumulation of metal S72, less accumulation would be expected where organic debris, and natural soil S was limiting. Also, as S was shown to be limiting for S"2 accumulation (in Chapter III), this may also suggest that S was relatively immobile in soil, coming to rest within the surface 10 cm. Movement of applied S was not documented in this research. Fig. 79 1. Influence of S on S72 accumulation in 'Penncross’ creeping bentgrass. Sulfur was applied as a flowable 52% carrier at levels of 0.0, 48 or 244 kg 8 ha'1 per application. Measurements were made at depths of 5, 10, and 15 cm. Means were the averages of 3 replications averaged over Fe+2 treatments. 80 mime _ 16m 3 rm 9: mm_ mt: men we: I m $5 I..- xomro «I... mm 608 mu _ mm 81 Above 5 cm there was an abundance of rooting (and thatch) but very little black layer. Perhaps this was due to the presence of adequate 02 near the soil surface. Oxygen diffusion into the plots was not measured but it was assumed that saturating the plots effectively contributed to anaerobic conditions in the profile, exemplified by the presence of S72. 1989 Research. When [S72] was again measured in 1989 the effect of high levels of S was very apparent (see Fig. 2.). When averaged over soil depth, more 8'"2 accumulated where plots were routinely treated with S at 244 kg ha'1 compared to other treatments. No differences between the control plots and those treated routinely with S at 48 kg ha"1 were observed. Differences attributable to N source were not detected by AOV (see Table 3.) but the treatment with the lowest overall [S72] ‘was the control treated with N037 while the highest was the high level of S combined with organic N (see Fig. 2.). No S'2 was visible in the thatch layer in any plot. When averaged across treatments more 8’2 accumulated at 5 cm compared to either 2.5 cm or 10 cm. No differences between any treatments were detected at 10.0 cm but the high S treatments had the highest [S72] (see Fig. 2.). When statistically analyzed it was found that only S and soil depth significantly influenced [S72] (see Table 3.). No effect attributable to N source was detected. There was however, an interactive effect between S level and soil depth. Production of black layer s”2 in response to S depended on depth in soil (see Fig. 3.), probably because of differences in [S] and concentration of organic debris, and O2 diffusion. Visual Fig. 82 2. Influence of S and N on S'2 accumulation in ’Penncross’ creeping bentgrass for 1989. Sulfur was applied as a flowable 52% carrier at levels of 0.0, 48 or 244 kg S ha-l per application. Nitrogen was applied as N037 or organic-N at 12 kg N ha'l. Duplicate measurements were made at depths of 2.5, 5, and 10 cm. Means were the averages of 6 replications. 83 01.02 0:0 001.50 00:39 00/1 DON Com 00m Cots 00m CON cm: 0 7. I. -LIII .IHI. I. e _ LIII- _ L LI I. L _ -iILI II_I !. -ILIIIILfiI IV —\ .II 90 20 I - .02 .16 ..l. . ... 190+ mo: «La swansfimo 7:1 , .02 + m V; I 7 1 OLD IT m me TI» \\\.>X\\O\‘\\ 4 filo—...! % . aoz + m xm : VII \x\\ r w... r .\\x \\ T0... was \ \ ..\ r \\\\\ \W“\ T W I \\\ \\\\\\ IQ]- \I/ I >\\*/ A \l/ - w / / I , / I II /\ / // / // is I / T /> //x r I. IN! I + 84 Table 3. Analysis of variance summary for total acid volatile sulfide occurring in a ’Penncross' creeping bentgrass green. Experimental design was a CRD split-split plot with six replications. Source df Mean Square F Sulfur 2 473,398 34 ** Error 15 13,842 N Source 1 40,717 2 ns S x N 2 20,479 1 ns Error 15 17,278 Soil Depth 2 545,878 45 ** S x D 4 74,407 6 ** N x D 2 17,431 1 ns S x N x D 4 11,680 1 ns Error 60 11,983 *, ** Significant at P = 0.05 and 0.01 respectively. 85 Fig 3. Interactive effects between S level and soil depth. A change in the magnitude of direction of response to S at 10 cm is noticable. 86 0'91 00 003%: 20.300 omN CON one Om: 00 O _ H _ F L Q I. IIIIIIIIITII 00c I‘ll III II III I I\c\\\I WOON T \\\\I IIII4 \\ \\-.I I .\ \\\ room 1 .\\\\\\..I\\\ \\\\ f. I.) \\\\\ woos . xxx 4 I \\\\ mm H 8.609 noon . \ s I. 000 ,. Eo or 4L. 1 I E0 on fit 100.” 1 E0 0N I + (LEM 5w) spams 10101 87 observation showed that there appeared to be more root material generated nearer the surface compared to the 10 cm depth, which seems reasonable. It would seem reasonable, too, that applied S that did not oxidize to 804-2 and leach from the profile would settle in the upper few centimeters, as it has been implied that S is immobile in soils (12). However, soil 8 content was not measured due to time constraints. Also, 02 diffusion was not measured but due to the amount of irrigation applied it would seem reasonable that more 02 would exist closer to the surface. Results of trend analysis (see Table 4.) showed significant linear response to applied S at both 2.5 and 5.0 cm but not at 10 cm. Additionally, a significant quadratic response was observed at 5 cm. When the pH(KCl) of plot soil was measured it was found that a significant reduction resulted from heavy S application (see Fig 4.). When averaged over N treatments and depth, soil treated with S routinely at the 244 kg ha"1 level had significantly lower pH than other treatments, and soil in plots treated at 48 kg 8 ha-1 had a lower pH than control plot soil. However, by this time the S totals for the treatments were 864 and 4,392 kg S ha-1. This translated into 18 or 90 pounds per 1000 ftz. Thus, along with pH results reported in Chapter IV it would appear that S does have the capacity to change soil pH, but only when applied at heavy rates or in confined soil. Also, the fact that the greatest pH depression occurred near the surface in the plots receiving high levels of S indicated that S was probably not very mobile. 88 Table 4. Trend analysis summary for total acid volatile sulfide (AVS). Values of F for analyzing 8'2 accumulation in response to applied S levels at soil depths of 2.5 cm, 5 cm and 10 cm. averaged over N sources were obtained from means of six replications. Source df Mean Square F Replication Sulfur 2.5 cm Linear 1 291,722 21 ** Quadratic 1 50,245 4 ns 5.0 Cm Linear 1 697,686 50 ** Quadratic 1 175,232 13 ** 10.0 cm Linear 1 15,914 1 ns Quadratic 1 16,381 1 ns Error 15 13,842 *, ** Significant at P = 0.05 and 0.01 respectively. Fig. 89 4. Influence of S and N on pH(KCl) of plot soil in ’Penncross’ creeping bentgrass for 1989. Sulfur was applied as a flowable 52% carrier at levels of 0.0, 48 or 244 kg S ha'l per application. Nitrogen was applied as N03- or organic-N at 12 kg N ha'l. Duplicate measurements were made at depths of 2.5, 5, and 10 cm. Means were the averages of 6 replications. 90 A318 zQBEm gem x w m I _ r 90 x0 OIIO I ON! I. . .02 v6 .1- Imel II mto+mxr «La wwrl I noz + m i To 4 Ir No n 8.0ng 90 + m xm >II> 1:! % n I HII. + x xix III OZ m m wmrl mm 7 J I II ImI , 2.x... .2 q T \ IIMWII \I/ I \ \ // I O I IoI W HI __4\ f/ TIAVII I _ // a I I.\I\< > \\x I I \vIv\\ I- I. m I. H-.\.I.\..I-IIII.\ I I I I I0 91 When statistically analyzed only S and soil depth were significant influences on soil pH. Generally speaking, soil pH decreased with soil depth except where high levels of S were applied. In those plots pH was more depressed higher in the profile, probably due to the effect of excess S. However, significant interactions were detected, including a three-way interaction between S, N source and soil depth (see Table 5.). This indicated that soil pH was dependent on the level of S applied, the N source, and depth of soil. Trend analysis (see Tables 6 and 7) detected significant linear responses in pH change due to S but only in thatch or at the 2.5 cm depth. No quadratic effects were detected. Visual quality ratings, and also algae and wilt ratings are presented in Table 8. There did not appear to be any difference in turf quality due to S application. However, early in the year there was a distinct response to S application. The difference faded later in the season. Turf treated with the organic-N was of better quality considering greeness. Turf treated with organic-N was also prone to having algae blooms, especially where S was applied. It may have been that additional algae were carried into the plots in the organic material. However, algal cells would be present in the soil naturally, and thus may have been stimulated by something in the organic matter or more likely by the S as a result of pH changes. Turf treated with high levels of S was prone to wilt, especially where N03” was applied. This was assumed to be the result of a salt effect produced interactively by the S and N03“. 92 Table 5. Analysis of variance summary for soil pH as influenced by sulfur level, nitrogen source and soil depth. Experimental design was a CRD split-split plot with six replications. Source df Mean Square F Sulfur 2 8.66 19.7 ** Error 15 0.44 N Source 1 0.04 0.6 ns S x N 2 0.12 1.7 ns Error 15 0.07 Soil Depth 3 1.31 43.7 ** S x D 6 0.73 24.3 ** N x D 3 0.13 4.3 ** S x N x D 6 0.10 3.3 ** Error 90 0.03 *, ** Significant at P = 0.05 and 0.01 respectively. 93 Table 6. Trend analysis summary for soil pH as influenced by sulfur applications averaged over fertilizer sources. Source df Mean Square F Sulfur 2 Thatch Linear 1 8.93 20 ** Quadratic 1 0.58 1 ns 2.5 cm Linear 1 7.53 17 ** Quadratic 1 1.84 4 ns 5.0 cm Linear 1 1.88 4 ns Quadratic 1 0.42 1 ns 10.0 cm Linear 1 0.51 1 ns Quadratic 1 0.01 1 ns Error 15 0.44 *, ** Significant a t P = 0.05 and 0.01 respectively. 94 Table 7. Trend analysis summary of pH as influenced by sulfur application, fertilizer source and soil depth. Source df Mean Square F Sulfur Organic Thatch Linear 6.57 ** Quadratic 0.31 ns 2.5 cm Linear 2.88 * Quadratic 0.41 1 ns 5.0 cm Linear 1.04 2 ns Quadratic 0.12 1 ns 10.0 cm Linear 0.48 1 ns Quadratic 0.04 1 ns Nitrate Thatch Linear 2.82 * Quadratic 0.24 1 ns 2.5 cm Linear_ 4.69 ** Quadratic 1.56 ns 5.0 cm Linear 0.91 2 ns Quadratic 0.32 1 ns 10.0 cm Linear 0.10 1 ns Quadratic 0.04 1 ns Error 15 0.44 *, ** Significant at P = 0.05 and 0.01 respectively. 95 Table 8. Mean values for visual quality, algae and wilt ratings on ~plots of ’Penncross’ creeping bentgrass turf afflicted with black layer. High ratings indicated superior quality turf, or more severe occurrrence of algae or wilt. Means were the averages of 8 ratings for visual quality, 2 ratings for algae and 3 for wilt. Sulfur Levelz N Sourcey VQR Algae Wilt Check Organic 7.5 3.4 1.2 Nitrate 7.2 1.6 1.1 48.8 kg ha"1 Organic 7.6 6.3 1.0 Nitrate 7.0 1.7 1.0 244 kg ha"1 Organic 8.0 6.1 2.8 Nitrate 7.6 1.8 4.0 LSD P = 0.05 0.4 1.9 0.9 Sulfur was applied at three week intervals as a flowable 52% S carrier. Y Organic N was derived from sewage sludge or bio-organic while nitrate N was from the Ca+2 or K+ salt. Carriers were applied at 12 kg ha.1 weekly. 96 CONCLUSIONS This field study has demonstrated that in situ black layer created in our research was an accumulation of S'z. It was assumed to result from the activities of sulfate-reducing bacteria. Water was shown to be a very important component of S-2 black layer formation, as was S. It appeared that black layer intensity was governed by such factors as S level, concentration and type of N, and depth of soil. Response to soil depth probably resulted from differences in content of soil organic debris. Addition of NO3' was somewhat effective in reducing black layer [8'2] but levels greater than those used in the 1989 research were probably needed. 97 REFERENCES CITED Atlas, R.M., and R. Bartha. 1981. Microbial ecology: fundamentals and applications. Addison-Wesley, Reading, MA. Berndt, W.L., J.M. Vargas, Jr., A.R. Detweiler, P.E. Rieke, and B.E. Branham. 1987. Black layer formation in highly maintained turfgrass soils. Golf Course Mgt. 55: 106-112 Burpee, L.L., and A. Anderson. 1987. The cause of black layer in golf greens: an alternative hypothesis. Greenmaster 23: 24. Cord-Ruswich, R. 1985. A quick method for the determination of dissolved and precipitated sulfides in cultures of sulfate-reducing bacteria. Journal of Microbiological Methods 4; 33-36. Feigl, F. 1972. Spot tests in inorganic analysis. Elsevier, Amsterdam. O’Conner, G.A. 1979. Soil Chemistry. John Wiler and Sons, New York. Rankin, P.C. 1988. When the black layer hit the fan. Turf Craft Aust. 6: 18-20. 10. 11. 12. 98 Smith, J.D. 1988. Black layer plug on Saskatchewan golf courses. Greenmaster 24: 6-21. Smith, R. 1981. Sulfate reduction in the sediments of a eutropic lake. Dissertation for the degree of Ph.D. Michigan State Univ., East Lansing, MI. Steel, R.G.D., and J.H. Torrie. 1980. Principles and procedures of statistics: a biometrical approach. McGraw-Hill, New York. Tabatabi, M.A. 1974. Determination of sulfate in water samples. Sulfur Inst. J. 10: 11-13. Tisdale, S.L. 1985. Soil fertility and fertilizers. Macmillen, New York. CHAPTER VI EXPERIMENTS WITH RADIOACTIVE SULFUR Abstract. Previous research suggested black layer in turfgrass soils was associated with accumulations of 5-2. A convenient way to study 8'2 formation was reported to be with 35S as a tracer. This technique was used to examine in-situ black layer from a ’Penncross’ creeping bentgrass green. Injection of tracer as 35804 directly into intact black layer soil micro-cores resulted in production of H235S and acid volatile 358-2. Injection of tracer into black layer soil placed in a reactor vessel resulted in S04.2 being reduced at a rate of 1.5 nMols S cm"3 soil day—1. Addition of N03” at 48 kg N ha"1 reduced the rate by 7 fold. Elemental 35S was reduced at a rate of 6.5 nMols S cm'3 soil day-1. This research suggested that active bacterial S04"2 reduction occurred in black layer and that elemental S reduction probably also occurred. It also supported the hypothesis that S was involved in black layer formation. Further, it suggested that fertilizer N03” would reduce the rate at which black layer S"2 forms. Black layer formation in turfgrass soils has been reported to result from dissimilatory S04:2 reduction (2, 5, 11, 13). 2 reduction Radioactive 358 has been used to study in-situ SO4- processes, mostly in marine or marsh sediments (6, 8, 14, 16). Rates of SO4-2 reduction in sea-water and in bottom sediments of the Black Sea were estimated by Sorokin (16) using 35SO4'2. Reported rates ranged from 2.3 to 0.03 mg S'2 produced L-1 99 100 sediment/water day-1. Smith (14) reported that addition of both 35S04"2 and 35S to profundal surface sediments of a freshwater lake resulted in production of H2358 with no lag time. Rates of 35SO4”2 reduction were reported to vary between 0.9 umols $04-2 reduced L"1 hr"1 at the in-situ [804-2] and 19.1 umols SO4-2 reduced L'l hr'1 at a 1 mM [504-2]. Rate of elemental 358 reduction was reported to be 8.8 umols S L'1 day-1. Smith (14) and Jorgensen (8) also reported using 35804-2 in research involving model sediment reactor systems. The objective of this 2 and elemental 35S to determine research was to use 35804- whether S cycling occurs in black layer. MATERIALS AND METHODS Sulfide Distillation Apparatus. A S"2 distillation and collection apparatus similar to the one described by Jorgensen (8) and Smith (14) was constructed. The apparatus consisted of a 200 ml screw top canning jar (Ball Glass) as a reaction vessel. The vessel was connected from the lid by glass line to a series of 20 ml scintillation vials containing 3 (or 5) mls of 2% CdCl2 (i.e., 8'2 traps). Vials were arranged so that the vessel atmosphere containing volatile S"2 could be continuously purged with 02 free N2 exhausted through each of the traps in sequence. The exhausted H2355 (or acidified AV3SS) would then react with the CdC12 according to the following reaction: H2358 + Cdc12 < ------ > Cd35s + 2 HCl pK = -27 (10). 101 The resulting Cd358 produced in each trap would then be quantified directly by liquid scintillation counting (14). There were 4 vials in the trapping train, which eventually exhausted into a 125 m1 erlenmeyer flask containing 100 mls of 2% CdClz. Sample from the exhaust trap would then also be counted. The jar lid was also attached to a Hungate type (7) gassing system equipped with a pressure regulator. The N2 gassing line passed through a flask of boiled water amended with cysteine as reducing agent. This was done to humidify the gas and minimize vessel evaporation. The gas inlet port on the inside of the vessel lid was also equipped with a teflon needle arranged so that incoming gas would first bubble into vessel solution before being exhausted. In addition, the jar lid was equipped with a stoppered port to facilitate entry into the vessel by needle. Sulfur Cycling in glggk Layer. The objective of this experiment was to determine whether 35804-2 would be cycled into H235S directly in an existing black layer. Intact micro-cores of black layer soil from the 10 cm depth of a ’Penncross’ creeping bentgrass (Agrostis palustris Huds.) green were collected for the experiment in December, 1989. Moisture content of the intact black layer soil was 10%. An air dried composite sample of the soil from the 0-15 cm depth in that green was 0.07% N and 0.8% C. The $04-2 concentration in air dried sediment core soil was between 22 and 45 ug 804!"2 cm”3 soil (17) and was variable from core to core. Most probable number estimates (1) indicated an approximate concentration of between 3.3 x 103 and 7 x 105 sulfate-reducing bacteria per gram of black layer soil. The soil 102 was a modified sand (88.3% sand, 5.0% silt, and 6.7% clay). Thatch was 5 cm thick. The green was located at the Robert Hancock Turfgrass Research Center in east Lansing. Micro-cores were constructed from 10 cc syringe cylinders (Becton-Dickenson) impaled horizontally into the black layer. Syringe cylinders containing 10 cc of soil were capped with rubber stoppers and transported to the laboratory for storage. Micro-cores were stored in an anaerobe jar with an atmosphere of 90% N2 and 10% H2 at ambient temperature in the dark until time for use. Black layer was visible in each core but appeared variable from core to core. Triplicate micro-cores were then injected with 1/2 ml of an active carrier-free Na235SO4 solution diluted in 10'3 M NaZSO4. Further, the label was also added as 10'3 NaZSO4 in 10% NaZMoO4, or in 10% NaN3. One half ml of each solution contained 0.3 uCi of 35804"2 (i.e., 11,000 dps or Bq). Cores were then incubated in an anaerobe jar at 23 C in the dark for 2 days. After incubation micro-core soil was placed in the reaction vessel. The vessel lid was attached, and the vessel was purged of air with N2 for 1 min. Next, 20 mls of freshly boiled distilled water cooled to room temperature under N2 was added to the vessel through the stoppered port in the lid, and the N2 flow rate was adjusted to near 100 mls min-1. In this fashion any volatile H235S which had formed was distilled from the soil solution in the vessel and exhausted into the Cd traps for quantification. After 30 minutes of distillation traps were changed and 2 mls of anoxic 37% HCl was injected into the vessel. This liberated any acid volatile 35S"2 (AV3SS) which was subsequently distilled for 103 30 min. For each trap 15 mls of LSC cocktail for aqueous solutions (RPI, Mt Prospect, IL) was added directly and the amount of radioactivity distilled into each trap was then determined by counting in a Beckman LS 8100 scintillation counter. Samples were corrected for quenching by the H# method. Time Course Studies with a Reactor Vessel. The objective of this experiment was to document the time course of appearance of H2358 and AV3SS. A subsequent objective was to determine the fate of the added 35S label. Twenty cc of black layer soil was added to the reactor vessel. This soil was identical to that used in the preceding section. The lid was attached then the vessel was purged of air with N2 for 1 min. Thirty mls of freshly boiled distilled, deionized water cooled to room temperature under N2 was then added to the vessel by syringe. The N2 flow rate through the trapping train was then adjusted to near 100 mls min-1. Sediment in the vessel was allowed to equilibrate under the moving N2 atmosphere for 3 days. This was ' done to allow for any anaerobic microbial growth which may have been limited by moisture or exposure to ambient atmosphere. After the equilibration period fresh CdClZ traps were attached to the trapping train, then 0.3 uCi of carrier free Na235804 in deionized water was injected into the vessel. The gas flow rate was maintained at near 100 mls min-1. Just prior to the injection of the label 5 mls of vessel soil solution was 2 concentration withdrawn by syringe for determination of S04- (17). Traps were changed each hour for the first three, then again at hours 6, 12, and finally at hour 24. At hour 24, 3 mls of 10% NaN3 and 2 mls of 37% HCl were then injected into the 104 vessel to halt bacterial activity and to liberate AVS respectively. The vessel was purged until radioactivity in the exhaust decreased to near background. For each trap 15 mls of LSC cocktail was then added directly and the amount of radioactivity distilled into each trap was determined by counting in a Beckman LS 8100 scintillation counter. Samples were corrected for quenching by the H# method. In a subsequent experiment NO3' as KNO3 was added concurrently with the 35804-2 at a level of 48 kg N ha-l, then analyses were performed in an identical fashion. Rates of 804'2 reduction were calculated according to the following equation: ([804'21)(a)(1.06) / (A)(V)(t) = nM s cm’3 day‘1 where [S0472] is the sulfate concentration in nM cm'3, a is the radioactivity of H28 plus AVS, 1.06 is an isotope fractionation factor (used by both Jorgensen and Sorokin), A is the original amount of radioactivity added, V is the sediment volume in cm'3, and t is the time of incubation in days (8). After the trapping of volatile 355 was complete the vessel soil was extracted with several portions of fresh water until radioactivity was near background. Soil was oven dried at 60 C for 24-48 hrs then re-extracted with several portions of fresh benzene again until radioactivity was near background. After oven drying a second time the soil was digested in alkaline NaOBr (18) and again extracted with water. Portions of all extracts (5 mls) were filtered to 0.2 um, then counted in a Beckman LS 8100 liquid scintillation counter. Fifteen mls of LSC mix for aqueous samples was added to water samples while 10 mls of LSC for 105 non-aqueous samples was added to the benzene. The extractions were performed to roughly determine how much of the added label remained as 35504-2, how much oxidized to 358, and how much was incorporated into an organic fraction. Reduction 2: Elemental S. The objective of this experiment was to determine whether elemental S was reduced directly. Elemental 35S (1 uCi mg'l) was dissolved in benzene giving a solution with specific activity of roughly 13 uCi ml'l. This was determined by liquid scintillation counting of 0.1 ml solution. An aliquot of this solution, 10x larger by volume than that counted, was transferred to a 125 ml serum bottle (Wheaton) and allowed to evaporate, leaving behind a residue of 35S. Eighteen cm“3 of black layer soil from the 10 cm depth of a ’Penncross’ creeping bentgrass golf green was then added to the bottle. This black layer soil was collected in May, 1990 and was 14% moisture. Next, 50 mls of freshly boiled distilled water cooled to room temperature under N2 and amended with 3 mls of 2% cysteine was added. The bottle was thoroughly purged of air with N2 then stoppered with butyl rubber and crimped with aluminum. The bottle was then attached to a modified trapping train similar to the one previously described. In this trapping train the bottle became the reaction vessel, and was attached to the train and Hungate apparatus with syringe needles and tygon tubing through the stopper. Fresh S'2 traps were attached and the N2 flow rate was adjusted to near 100 mls min-1. Traps were changed each hour for the first three, then at hours 6, 18 and 24. At hour 24, 3 mls of 10% NaN3 and 2 mls of 37% HCl were added to the bottle to 106 halt bacterial activity and release AVS respectively. Bottle exhaust was then trapped until radioactivity was near background. Radioactivity in traps was determined as previously described. In a supportive experiment, 200 g of soil from a ’Penncross’ creeping bentgrass golf green contained in 125 ml serum bottles (Wheaton) was treated with cold S from either NaZSO4 or from 52% flowable S at a level of 48 kg S ha'l. Experimental units were then water-logged with tap water so that no headspace existed, stoppered with butyl rubber and crimped with aluminum. Units were then incubated for 21 days at 30 C in the dark. After 21 days units were sampled for concentration of free H28 and AVS with the method of Cord-Ruswich (4). RESULTS AND DISCUSSION Sulfur Cycling 13 gleek Leyer. Active 35S cycling was evident when 35504:"2 was injected into existing black layer (see Table 1.). After 2 days approximately 4.4% of the recovered label existed as free H2358 while 95.6% of the recovered label existed as AV358. The trapped 35S'2 represented nearly one third of the added label. It was assumed that the e- donor was sloughed turfgrass root mass or other organic debris. The S'2 fractions appeared clearly separated by the water/acid distillation procedure (see Fig. 1.). Jorgensen (8) reported it was previously suggested (Hartman) that sulfides were too unstable to be separated in this way. However, Kaplan et a1. (9), Smith (14), and Jorgensen (8) used this type of method to make such distinctions. 107 Table 1. Summary of 358 cycling in intact black layer micro-cores as influenced by NaZMoO4 and NaN3. One half ml carrier-free Na35504 (0.6 uCi ml'l) plus 10% NaN3 or 10% NazMoO4, was injected into 10 cc intact sediment cores collected directly from a black layer in a ’Penncross' creeping bentgrass golf green. After 48 hrs volatile H2358 from each soil core was distilled for 30 minutes into 2% CdClz, then counted as Cd3ss. Next, AV358 was acidified with HCl then distilled for 30 minutes into CdCl2 and again counted. Cores were injected in triplicate. Sulfide Fraction Treatment uCi 35-S Recovered Free H28 10’3 M 35304“2 4.40 x 10‘3 b 10’3 M 35504-2 in 10% NaN3 0.10 x 10’3 c 10'3 M 3550,,“2 in 10% Moo4 0.10 x 10'3 c AVS 10‘3 M 35304‘2 94.70 x 10‘3 a 10'3 M 35304'2 in 10% NaN3 0.10 x 10"3 c 10'3 M 35304'2 in 10% Moo4 0.20 x 10"3 c Values followed by similar letters were not different by DMRT P=0.05. 108 Figure 1. Separation of H235S and AV358 fractions in black layer soil using a S-2 trapping train. Radioactive 35804”2 was injected directly into 10 cc black layer micro-cores harvested from a ’Penncross’ creeping bentgrass golf green. Accumulated H2358 was distilled from the sample in water. Acidification with HCl volatilized AV358 which was subsequently distilled. The S'2 trapping train consisted of a reactor vessel which contained the sample and 4 8'2 traps hooked together in sequence and to the reactor via glass line. 109 mmmEDZ mime Lu W N e O _ V _ N LVII ...! lllllll I.’ 1111111 w T I 1. /k I HI I II I I II INF II II wow I I .3 I 08 Ir m? .1. I m I OI. TmN (c—OL x) paJeroea-z—SSC 13” 110 When NazMoO4 was concurrently injected with 35SO4-2 , cycling was drastically diminished although some did occurs (see Table 1.). Free H235S and AV35S were collected in approximately equal concentrations which represented only about 0.03% of the added label each. The effect of the MOO4-2 was to act as a competitive inhibitor of sulfate-reducing bacteria (14). Molybdate was reported to be stereochemically similar to 804'2 , inhibiting ATP-sulfurylase, the first enzyme in the sulfate-reducing pathway (14). As the inhibition is competitive in nature (14), this means reaction velocity (Vm ) in the presence or absence of 2 ax MoO4-2 is similar at a high SO4- concentration, but the saturation constant (Km) is much less in the absence of Moo);-2 (12). Thus, the ability of MoO4'2 to effectively inhibit the activities of sulfate-reducing bacteria would be dependent on the SO4-2 concentration as well as on the concentration of MoO4"2 (14). Smith (14) reported 100% inhibition of SO4-2 reduction in profundal lake sediments with addition of NazMoO4 at only 0.2 mM. ' Perhaps in our experiment the S04"2 concentration in the spiked core was great enough (i.e., > 500 nMols added per core) to facilitate some limited reduction in the presence of MoO4'2. It may have also been that some reduction occurred prior to the MoO4'2 effect, or that the label reached some bacteria while the inhibitor did not. When NaN3 was concurrently injected, S04"2 reduction was again drastically curtailed (see Table 1.). The effect of the NaN3 was to act as a biocide killing sulfate-reducing bacteria, along with other organisms. Azide combines with the oxidized heme iron of cytochrome a and a3 preventing their reduction during electron 111 transport (15). As some cycling of 35S did occur it was probable that SO4-2 reduction occurred prior to the effect of the poison. Another possibility was again that the poison did not physically reach some sulfate-reducers while some 35804'2 did. These questions were, however, not addressed in the research. 2 reduction was an These results of this study prove that SO4- active, bacterial process in the experimental black layer. Time Course Studies yiih e Reactor Vessel. Labeled HZS was detected with no apparent lag time. The H2358 accrued in a near linear fashion for 24 hours (i.e., R2 = 0.99 Y’= 7.5 x 10'5 + 6.6 x 10'4 (x)). This represented 4.9% of the added label. Acid volatile 353'2 recovered after 24 hours represented 8.8% of the added label. The rate of S04-2 reduction was calculated to be 1.5 nM S cm'3 soil day-1. This translated into a mean residence time (12) for the SO4'2 pool (i.e., 4,160 nMols) of 137 days. Residence half life of the labeled SO4-2 was calculated to be 4 days. The calculated rates of 804"2 reduction for this research were similar to those reported independently by Berner, and Kaplan in the upper 3 m of the Santa Barbra basin (in California), by Sorokin in the surface sediments of the Black Sea, and by Hartman and Nielsen in the surface sediments of the Bay of Keil (8). Our experimentally calculated rates were less than rates reported for the sediments of a eutrophic lake (14) or for the sediment of a model reactor system (8). When N03” was added concurrently with the label at a level of 48 kg N ha'1 S"2 production was curtailed. Free H235S again accrued nearly linearly (i.e., R2 = 0.80 Y'= 1.7 x 10"4 + 1.8 x 10"5 (x) for 24 hours. This however represented only 0.2% of the 112 added label. Acid volatile 35S'2 recovered after 24 hours represented only 0.9% of the added 35$04.2. The rate of 3 soil day-1, reduction was calculated to be 0.2 nM S cm- considerably less than where no NO3' was added. This translated into a mean residence time for the S04-2 pool (i.e., 7,940 nMols) of 1,726 days. Residence half life of the labeled SO4-2 was calculated to be 54 days. The effect of the added N03” was to act as an alternate 2 electron acceptor, which prevented S04” reduction from occurring by also elevating redox (3, 14). Reduction of N03“ was reported to Proceed prior to reduction of 804'2 on thermodynamic grounds. The free energy liberated from reduction of N03- to N2 was -53 kcal mol'1 H2 while that from reduction of S0,"2 to H28 was only -9.1 kcal mol'1 H2 (14). Thus, the reduction of N03" is much more spontaneous and would yield more usable energy. However, the effect of the added N037 would be expected to be fairly short lived. Once the N03- pool is depleted through NO3' reduction to N2, 504"2 reduction should proceed unimpaired. Results of Chapter IV suggest that addition of N03“ at 48 kg N ha'1 to similar soil under flooded conditions prevented depression of redox for 9 days, after which redox potential declined and 8'2 production ensued. Thus, frequent NO3' addition might be necessary to combat black layer 5'2 production 1 situ. When soils from the study were extracted with water 72% of the added label was recovered where N03- was not added, and 95% of the label was recovered where NO3' was applied. For the benzene extract 0.8% of the added label was recovered from both soils. For the NaOBr digest extract the values were 3.3% and 2.9% 113 respectively. These results indicated that a large percentage of the label remained in the water phase probably as 35804-2. Additionally, smaller percentages mineralized into elemental 35$ and immobilized probably as biomass. Label recovery totals were 89.8% where no NO3' was added and 99.8% where N03” was added. Remainder of the label was unaccounted for but a distinct 2 possibility was that where active SO4- reduction took place a fraction of the label may have mineralized into-acid insoluble pyrite (FeSZ) or other acid-insoluble oxidation sates. This was not demonstrated in the current research but Howarth (6) showed that pyrite is a major end product in the surface peat of a Cape Cod salt marsh. He further reported that pyrite can form in a day or less without iron mono-sulfides as intermediates. Thus, if the pyrite fraction of the 358 was ignored the rate of S04"2 reduction would be grossly underestimated. This would in turn magnify the observed differences in rates of reduction where N03- was and was not added. Reduction e: Elemental Si Labeled H2355 was detected with no apparent lag time. The H235S accrued in a nearly linear fashion for the 24 hour period (i.e., R2 = 0.95 Y’= 0.29 + 0.02(x)). This represented 6.4% of the added label. The AV35S recovered after 24 hours represented 25.2% of the added 358. The rate of 3 elemental S reduction was calculated to be 6.5 nM S cm- soil day-1. This was very similar to the rates of 35S reduction reported by Smith (14). This translated into a mean residence time for the S pool (i.e., 6,250 nMols) of 54 days. Residence half life of the added label was estimated to be 2 days. In the supportive experiment it was found that slightly more 114 than 4x more total 8’2 accrued where soils were treated with flowable 52% S compared to where soils were treated with S04"2 as a source of S (see Table 2.). This would agree with projected results considering the stoichiometry of SO4'2 vs S reduction: 504"2 + 8e- + 8 H+ < ----- > 5’2 log K = 20.7 s + 2e" + 2 H+ < ----- > 5‘2 log K = 4.9 In other words for a given amount of reducing equivalents about 4 times more S72 could theoretically be produced from the reduction of elemental S compared to the reduction of S0472. Thus, from measurement of total 8'2, and from observation of reduction of 358, it appeared that direct reduction of elemental S was occurring. Definitive proof that direct reduction was happening was not collected. To absolutely prove this it would need to be demonstrated that oxidation of 35S to 35SO4"2 prior to reduction did not occur. Additionally, it is difficult to accurately access S reduction due to its insolubility in the aqueous phase (14). However, addition of the reducing agent cysteine to the system at the onset of the study should have helped to prevent any 358 oxidation. Also, the fact that nearly 4 times more S-2 was produced where S was added (as opposed to SO4'2) in the supportive study, together with the results of the 35S reduction study was strong evidence that direct S reduction occurred. It was also very interesting that the observed rate of 35S reduction was some 4 times greater that the measured rate of 35SO4'2 reduction in our study with turfgrass soil. Smith (14) reported that the rate of elemental S reduction was less that 40% 2 2 of the observed rate of 804- reduction at the in situ SO4- concentration in profundal surface sediments, and thus, S 115 -2 Table 2. Influence of added S and N03- on S accumulation and redox potential in flooded sand. Sulfur was added as either 52% flowable s or as 304‘2-5 at 48.8 kg S ha-1. Nitrogen was added as NO3— at 48.8 kg N ha-l. Z Treatments Sulfides Redox Parameters H28 AVS pH Y pE x pH+pE Flowable S no N03’ 8.0 a* 53.2 a 7.5 b -1.2 b 6.3 b Flowable S + N03- 2.8 b 1.6 b 7.2 c 2.7 a 9.9 a Sulfate S no N03‘ 5.2 ab 7.8 b 7.8 a -2.1 b 6.2 b LSD 05 = 4.4 16.3 0.2 0.6 0.6 -2 -1 . . z mg S kg 8011 solution. y pH measured with Ag/AgCl combination electrode. x pE = Eh (millivolts)/59.2. * Means followed by similar letters were not differenct by LSD @ P = 0.05. 116 reduction by sulfate-reducers may or may not be of significant magnitude to influence total reduction rates. However, Smith also reported that the two activities may actually involve totally different populations of organisms, since in addition to Desulfuromonas, photosynthetic bacteria and cyanobacteria were shown to reduce elemental S in the dark. Thus, different populations of S or SO4-2 reducers may exist in turfgrass soils. Alternatively, it may be that the total populations of sulfate-reducers in our experimental soil may have differed greatly from point to point, reflected in our sampling. CONCLUSIONS This research has demonstrated that active S cycling of both SO4-2 and elemental S occurs in black layer soils. The cycling was shown to be bacterial in nature. Further, it was shown that addition of N03“ at traditional levels had an antagonistic 2 influence on rates of 804- (or S) reduction. The research results have also lended support to the hypothesis that black 2 layer formation involves bacterial 504- or S reduction. 1) 2) 3) 4) 5) 6) 117 REFERENCES CITED Alexander, M. 1982. Most probable number method for microbial populations. p. 815-820. In A.L. Page (ed.) Methods of soil analysis part II. American Society of Agronomy, Madison, WI. Berndt, W.L., and J.M. Vargas, Jr., et a1. 1987. Black layer formation in highly maintained turfgrass. Golf Course Mgt. 55: 106-110. Connell, W., and W. Patrick. 1969. Reduction of sulfate to sulfide in waterlogged soils. Soil Sci. Soc. Am. Proc. 33: 711-715. Cord-Ruwisch, R. 1985. A quick method for the determination of dissolved and precipitated sulfides in cultures of sulfate-reducing bacteria. Journal of Microbiological Methods 4: 33-36. Hodges, C.F. 1989. Another look at black layer. Golf Course Mgt. 57: 54-58. Horwath, R. 1978. Pyrite: its rapid formation in a salt marsh and its importance in ecosystem metabolism. Science 203: 49-51. 7) 8) 9) 10) 11) 12) 13) 14) 118 Hungate, R.E. 1950. The anaerobic mesophilic cellulolytic bacteria. Bacteriol. Rev. 14: 1. Jorgensen, BB., and T. Fenchel. 1974. The sulfur cycle of a marine sediment model system. Mar. Biol. 24: 189-201. Kaplan, I.R., K.O. Emery, and S.C. Rittenberg. 1963. The distribution and isotopic abundance of sulfur in recent marine sediments of southern California. Geochim. cosmochim. Acta 27: 297-231. Lindsay, W.L. 1979. Chemical equilibria in soils. John Wiley and Sons, New York. Rankin, P.C. 1988. When the black layer hit the fan. Turf Craft Aust. 6: 18-20. Segel, I.H. 1976. Biochemical calculations. Second edition. John Wiley and Sons, New York. Smith, J.D. 1988. Black plug layer on Saskatchewan golf courses. Greenmaster 24: 6-21. Smith, R. 1981. Sulfate reduction in the sediments of a eutropic lake. Dissertation for the degree of Ph. D. Michigan State University, East Lansing, MI. W 15) 16) 17) 18) 119 Smith, E., et al. 1983. Principles of biochemistry. McGraw-Hill, New York. Sorokin, Y.I. 1962. Experimental investigation of bacterial sulfate reduction in the Black Sea using 35S. Mikrobiologiya 31: 402-410. [Transl. Microbiology 31: 329-335 (1962)]. Tabatabai, M.A. 1974. Determination of sulfate in water samples. Sulfur Institute Journal 10: 11-13. Tabatabai, M.A., and J.M. Bremner. 1970. An alkaline oxidation method for determination of total sulfur in soils. Soil Sci. Soc. Am. Proc. 34: 62-65. CHAPTER VII SUMMARY AND CONCLUSIONS Laboratory and field research was used to gain an insight into the nature of 8'2 based turf black layer. A variety of experiments were performed, most with the intent of illustrating some of the basic principles governing biological 8'2 production. Why black layer formed where it did was not addressed in the research, as it would seem logical that each individual black layer should have its own unique set of particulars. Nor was the assumption made that each and every black layer was s”2 based. Again it would seem logical that other reasons for the manifestation of a blackened or banded soil profile may exist. However, of the black layers occurring in golf greens observed by this author in the past four years, most appeared to possess a similar characteristic, the accumulation of S-Z. This was evidenced by black layer being associated with a foul odor, wet conditions, a blackened appearance which disappeared with time when exposed to air, and positive reactions to a spot test specific for S’z. Therefore, specific inputs which were thought to affect S"2 production, hence black layer formation, were examined. Conclusions based upon those examinations are presented below: 1) The concentration of soil S, soil moisture, soil N03”, and soil organic matter were important black layer 8’2 determinants. Both adequate soil 8 and excess irrigation or water-logging were needed to generate black layer 8"2 120 2) 3) 4) 121 accumulation. Sulfur was considered the key ingredient. Addition of organic matter intensified the condition while NO3' retarded the accumulation. Redox potential dictated whether S"2 would form regardless of inputs such as S. A critical pE + pH of between 9 and 10 was found to be necessary to generate S'Z. Nitrate was effective at retarding black layer s"2 formation because it kept redox potential elevated, while also acting as an alternate e' acceptor. The activities of sulfate-reducing (or S-reducing) bacteria were central to black layer S"2 production. By using 2 carrier-free 35804” as a tracer it was determined that the rate of 804—2 reduction in black layer was 1.5 nM S cm-3 soil day-1. Nitrate reduced the rate by 7 fold. The rate of elemental S reduction, calculated by using elemental 35S was 6.5 nM S cm-3 soil day-1. Thus, both SO4'2 reduction and elemental 8 reduction occurred in existing black layer. Control of black layer 1 situ centerd around prevention rather than curative control. Prevention lies in keeping aeration status as high as possible by whatever means necessary. This means using core aeration, and application of nitrates to elevate redox. This also means adjusting drainage, both surface and sub-surface, to be as efficient as possible. Finally, it also means that irrigation should be applied very judiciously. APPENDIX 122 APPENDIX I MEASUREMENT AND DETECTION OF SULFIDE AND SULFATE Measurement 9: Sulfide i3 Solution Copper sulfate reagent was prepared by amending anoxic distilled water with CuSO4 to 798 mg kg-l, and with 37% HCl to 1,800 mg kg"1 (1). A stock solution of NaZS:9H20 was then prepared from a washed crystal dissolved to a known concentration. A 1:1 dilution series was prepared from the stock solution. Solution transfers were made by syringe into tubes flushed with 02 free N2 and stoppered with butyl rubber (Bellco Glass). Five mls of moving reagent in 10 ml cuvettes was injected with duplicate samples of each dilution measuring absorbance at 360 nm. The reaction upon injection was as follows: CuSO4 + NaZS < ----- > CuS + NaZSO4 The procedure was repeated at wavelengths of 480, 620 and 760 nm. Reagent was moved with a magnetic spin bar. A cuvette containing only reagent served as a blank. TheLSA between blanks and water at each wavelength was used as a correction factor if necessary. Sulfide concentration curves were then prepared by plotting absorbance vs 5'2 concentration, and determining linear regression by least squares. If soil solution extracts were to be analyzed the sample was filtered with a 0.2 um syringe filter to remove particulates. Dissolved H28 was measured in soil solution without modification of the soil. Acidification of the soil to pH 0 released acid 123 I , .mm—J’WI 1” my VW-r—“F'H'v PW, ATP} I.—» 4 i- —- - 124 volatile S'Z. Precise corrections between sample absorbance in H20 and in reagent were necessary to correct for acid extractable color (i.e., tannins etc.) Spot Detection Procedure _e; Sulfide Spot detection solution was prepared by dissolving 4.0 g KI with 1.27 g 12 and 3.0 g NaN3 in 100 ml distilled H20 (2). Five to six g of sample soil was then added to 10 mls or more of solution on a 50 mm or larger watchglass. The solution was non-reactive except in the presence of S"2 which acted catalytically to drive production of NaI according to the reaction: 2 NaN3 + 12 <----> 2 NaI + 3 N2 Observing the evolution of N2 as fine bubbles from the solution was considered a positive reaction for the presence of S'z. Meaeurement e: Sulfate ie Solution Barium chloride-gelatin reagent (3) was prepared by dissolving 0.6 g Difco gelatin in 200 mls warm (ca. 70 C) deionized distilled water. The solution was allowed to stand 16 hours at 4 C and was then brought to room temperature. Two grams reagent BaC12:2HZO was then added and swirled until dissolved. This reagent was stored at 4 C and brought to room temperature before use. Also, solutions of 0.5 N HCl and a standard SO4-2 solution were needed. The SO4'2 standard was made from 5.434 g KZSO4 per liter. One ml of this solution contained 1 mg SO4'2. 125 Before analysis, water samples were filtered to pass 0.2 um to remove particulate. A standard curve to reference S04.2 was prepared by diluting the standard to between 10 and 100 ug S04!-2 ml'l. Then one ml of each dilution was diluted to 20 mls with S04.2 free water. Two mls 0.5 N HCl and one ml of the barium "seed" was then added. The resulting solution was swirled and allowed to stand for 30 min. After 30 min the Solution was again swirled and the resulting turbidity measured at 420 nm. The turbidity resulted from the formation of BaSO4 according to the following reaction: BaClz + K2804 < ----- > BaSO4 + 2 KCl Solution with no addition of 804'2 was used as a blank. The sulfate concentration curve was then prepared by plotting 2 absorbance vs S04” concentration then determining linear regression by least squares. 1) 2) 3) 126 REFERENCES CITED Cord-Ruwisch, R. 1985. A quick method for the determination of dissolved and precipitated sulfides in cultures of sulfate-reducing bacteria. Journal of Microbiological Methods 4: 33-36. Feigl, F. 1972. Spot tests in inorganic analysis. Elsevier, Amsterdam. Tabatabi, M.A. 1974. Determination of sulfate in water samples. Sulfur Institute Journal 10: 11-13. I III’ LIBRQRIES M M 649 III ”I 31 ZQBOIIUI'S’JHOB’B) I!) (If! I) I)" . .....I. ..._.,L.....:.__ s. s ... I _ ......L . ...: ..52...:1.14.5.1: