Click on author name to view affiliation information
Black layer (BL) has reduced the quality of many putting greens since the 1980s. Initially, the nature of BL was unknown. Research established that BL was sulfide (S2−) formed in response to low redox. Its formation was linked to dissimilative sulfate reduction using labeled sulfate (35SO4 2−). The objective of this study was to see if elemental sulfur (S0) reduced to S2−. When labeled sulfur (35S0) with a specific activity of 3.7 × 104 Bq·mg−1 was added to soil from a green with BL in a reaction vessel kept at a low redox potential, it reduced at a per-minute rate of 5.3 nmol·cm−3, resulting in accrual of labeled hydrogen sulfide (H2 35S) and acid-soluble sulfide (AS35S). Nearly 32% of the 35S0 reduced to labeled sulfide (35S2–) in 24 h. Adding S0 to greens with low redox may result in rapid formation of S2– and an accelerated rate of BL development. Avoiding this requires limiting the input of S0 or encouraging high soil redox through chemical or physical means such as fertilizing with nitrate (NO3 –) and aerifying. This is the first report implicating S0 reduction as a source of BL development in putting green soil.
Root-zone blackening in putting greens was termed black layer (BL) (Scott, 1986). Sometimes BL appears as a horizontal band of variable thickness, whereas other times the entire profile is blackened. A decline in turf quality usually accompanies BL. Because of its impacts on turf quality, BL can be a major problem of putting greens. It was referred to as the number one malady of creeping bentgrass (Agrostis palustris Huds.) putting greens in the late 1980s (Scott, 1986).
Black layer was determined to be an accumulation of sulfidic (S2–) precipitates such as iron sulfide (FeS) within the soil and plant root matrix (Berndt et al., 1987):
This was confirmed by spot testing BL from greens at 26 golf courses in Michigan and Ohio with a solution of azide (NaN3) and iodine (I2) (Berndt, 1990; Fiegl, 1972). Metal sulfides (MeS) such as FeS were present if bubbling was observed on introducing BL soil to solution. Bubbling occurred because MeS catalyzes production of NaI with release of N2 from solution:
Samples of BL at all 26 courses tested positive for the presence of MeS.
One source of sulfide (S2–) in putting green soil containing BL was dissimilative reduction of SO4 2– by sulfate-reducing bacteria (SRBs) in response to low redox (Berndt and Vargas, 1987, 2006):
This was established using labeled sulfate (35SO4 2–) (Berndt and Vargas, 2006). When 10−3 M 35SO4 2– with a specific activity (SA) of 1.554 × 105 Bq·mg−1 was injected into intact microcores taken from a BL in a ‘Penncross’ creeping bentgrass green, it reduced at a mean per-minute rate of 4.9 × 10−3 nmol·cm−3 (Berndt and Vargas, 2006). Labeled hydrogen sulfide (H2 35S) and labeled acid-soluble sulfide (AS35S) were produced. Adding either azide (N3–) or molybdate (MoO4 2–) with the label decreased the rate to 2 × 10−5 and 7 × 10−6 nmol·cm−3, respectively, which proved 35SO4 2– reduction was biological and SRBs were involved (Berndt and Vargas, 2006).
Although tracers proved bacterial reduction of 35SO4 2– occurred, other research suggested elemental sulfur (S0) was reduced to S2– (Berndt, 1990; Berndt and Vargas, 1992, 1996). For example, when anoxic sand was treated with S at 48 kg·ha−1 using S0 or SO4 2–, there was ≈4.7 times more S2– produced where S0 was applied (Table 1). Differences in S2– concentration occurred because eight electrons (e –) are needed to reduce a molecule of SO4 2– to S2–, whereas only two are needed to reduce a molecule of S0 to S2– (Lindsay, 1979):
This means four times more S2– results from reduction of S0 for a given level of reducing equivalents such as soil organic matter. This in turn implies that adding S0 to putting greens having conditions favorable for BL development, like low soil redox, may result in a rapid accumulation of sulfidic precipitates, hence an accelerated rate of BL formation.
Reduction of S0 occurs in many environments, and greater than 90 species of bacteria reduce S0 through enzymes like sulfur reductase (Hao and Ma, 2003; Hedderich et al., 1999; Stetter and Gaag, 1983). However, information on the reduction of S0 in soils supporting putting greens is needed, because this topic is absent in the turf literature. The objective of the research described in this article was to determine if 35S0 reduces to 35S2– in soil from a creeping bentgrass green affected by BL.
Soil with BL was collected from a ‘Penncross’ creeping bentgrass green at the Robert Hancock Turfgrass Research Center at Michigan State University in East Lansing. The concentration of S0 in soil was 0.19 mg·cm−3 (Tabatabai, 1974), and most probable number estimates were 3 × 103 to 7 × 105 SRBs/g soil (Alexander, 1982; Berndt and Vargas, 2006). The per minute rate of 35SO4 2– reduction was 4.9 × 10−3 nmol·cm−3 for intact cores (Berndt and Vargas, 2006) and 36.6 × 10−3 nmols·cm−3 for disturbed soil (Berndt, 1990).
Recovering 35S2– required an anaerobic still and S2– trapping train, similar to the one used by Berndt and Vargas (2006) (Fig. 1). The reaction vessel was a 125-mL Wheaton serum bottle (Wheaton Science Products, Millville, NJ) linked to an upstream source of O2-free N2 and a downstream series of 35S2– traps. Oxygen was removed from the N2 stream by sparging it over hot Cu2+ filings in a Hungate apparatus (Kaspar and Tiedje, 1982). Sulfide traps were 20-mL glass scintillation vials containing 3 mL 2% cadmium chloride (CdCl2).
Citation: HortScience horts 43, 5; 10.21273/HORTSCI.43.5.1615
Elemental 35S0 with a SA of 3.7 × 104 Bq·mg−1 (Perkin-Elmer, Waltham, MA) was dissolved in benzene, producing a solution with a SA of 4.8 × 105 Bq·mL−1. The SA of this solution was verified by liquid scintillation counting (LSC) using a Beckman LS 8100 LSC instrument (Beckman Coulter, Fullerton, CA).
One milliliter of the 35S-benzene solution was injected into the empty reaction vessel. Benzene was evaporated by purging the bottle with O2-free N2 at a flow rate of 100 cm3·min−1 for several minutes, leaving 35S0 adsorbed to the inside of the glass reaction vessel.
As purging was continued, 18 cm3 of BL soil was put into the vessel along with 50 mL of boiled, distilled water cooled to room temperature under N2. The water was amended with cysteine to establish low redox (i.e., Eh ≈ –340 mV) in the reaction vessel (Kaspar and Tiedje, 1982). Resazurin indicator was used to verify existence of low redox (Kaspar and Tiedje, 1982). The vessel opening was then plugged with butyl rubber, crimped with an aluminum seal, and the vessel was attached to both the upstream O2-free N2 gas stream and the downstream trapping train through the butyl rubber using syringe needles and Teflon tubing. Traps containing 3 mL 2% CdCl2 were installed into the trapping train, and the flow rate of the O2-free N2 was adjusted to 100 cm3·min−1.
If the 35S0 was reduced by the soil in the reaction vessel, then AS35S precipitates (i.e., Fe35S) and free H2 35S would be produced. Any free H2 35S gas that evolved from the soil was swept from the reaction vessel into the trapping train with the gas stream, where it reacted with the Cd solution contained in the traps forming a Cd35S precipitate (Lindsay, 1979):
Radioactivity contained in each of the traps as Cd35S was then counted with a Beckman LS 8100 liquid scintillation counter (Beckman Coulter). Counting required addition of counting cocktail for aqueous solutions (Research Products International, Mt. Prospect, IL). All counted samples were corrected for quenching by the H# method (Beckman Coulter). Traps were changed and Cd35S was counted at 60, 120, 180, 360, 1080, and 1440 min.
After the trap change at 1440 min, 3 mL 10% w/v NaN3 and 2 mL anoxic 37% HCl were injected into the reaction vessel. Azide sterilized soil, which stopped reduction of 35S0, and the HCl lowered vessel solution pH to ≈1, which released AS35S from soil as free H2 35S gas. The flow rate of O2-free N2 was once again adjusted to 100 cm3·min−1, and the distillation was continued. Traps were changed as necessary until levels of radioactivity being swept into the trapping train with the gas stream were near background.
The per-minute rate of S0 reduction was calculated using the following formula (Sorokin, 1962):
The (S) was the concentration of S0 in the reaction vessel in nmol, (a) was Bq of 35S2− recovered, 1.06 was an isotope correction factor, (A) was Bq added, (V) was soil volume (cm3), and (t) was the incubation time in minutes. The mean residence time (t mrt) of the S0 pool was calculated by dividing (S) by the calculated rate of reduction, and the residence half-life of the 35S-label (t rhl) was (A) ÷ (a)(2) (Berndt and Vargas, 2006).
The concentration of 35S2– recovered (nmol·cm−3) was plotted versus time to depict the release pattern of H2 35S over 1440 min and accrual of AS35S in the reaction vessel during the 1440 min. From these measurements, a first-order rate constant (k) was calculated (Segel, 1976):
In this equation, [S]0 was the original concentration of 35S0 added as a percent (i.e., 100%), [S] was the percentage of 35S0 remaining at 1440 min (t), and k was the rate constant min−1.
After k was estimated, [S] was calculated for times of 60, 120, 180, 360, and 1080 min using:
Subtracting [S] from [S]0 for each time gave the amount of 35S2– released (nmol·cm−3), which was plotted versus time. The [S] at 1440 min was determined experimentally.
The velocity (v) of the reduction of 35S0 per minute (nmol·cm−3) was estimated by v = k[S] and plotted versus time. A plot of v versus the concentration of 35S0 and corresponding double reciprocal plot (Segel, 1976) were developed to show the relationship between v and 35S0.
Regression analysis was used to describe accrual of 35S2– versus time, velocity of reduction versus time, and other velocity curves. Graphics and regressions were developed using SigmaPlot 10 (Steel and Torrie, 1980; Systat Software, 2006).
Radioactive 35S2– was trapped with no apparent lag time (Fig. 2A). Approximately 31.5% of 35S0 was recovered as 35S2– with the ratio of total 35S2– trapped to H2 35S trapped being 4.95. The H2 35S trapped over 1440 min was 6.4% of the added label, or 20.2% of trapped 35S2–. The AS35S liberated by HCl was 25.2% of the added label, or 79.8% of trapped 35S2–. The per-minute rate of S0 reduction was 5.3 nmols·cm−3. The t mrt of the S0 pool was 77,900 min and the t rhl of the label was 2304 min.
Citation: HortScience horts 43, 5; 10.21273/HORTSCI.43.5.1615
The k at 1440 min was 8.1 × 10−4 per min. Velocity (v) of the reduction of 35S0 decreased with time in curvilinear style (R2 = 0.9998; Ŷ = 18.4 − 0.137X + 3.0 × 10 − 6X2) (Fig. 2B). The per-minute v at 1440 min of 5.77 nmol·cm−3 was very close to the rate of 5.33 nmol·cm−3 calculated using Sorokin's (1962) formula. When v was plotted versus concentration of 35S0 (Fig. 3A), the curve was linear (R2 = 1.0; Ŷ = –0.0023 + 0.0008X), which meant v was proportional to concentration of 35S0. When plotted as a Lineweaver-Burk plot (Fig. 3B), the curve intercepted both axes too closely to the origin to determine either maximum velocity (i.e., Vmax ) or the Michaelis constant (i.e., Km ). This meant that enzyme saturation did not occur, Vmax and Km appeared infinite, and the reaction was first order.
Citation: HortScience horts 43, 5; 10.21273/HORTSCI.43.5.1615
Understanding the reduction of S0 is important, because S0 is recommended for turf in a variety of publications (Beard, 1973; Landschoot, 2007; Vargas, 2005; Voight et al., 2007). Sulfur can be used to help mitigate alkaline soil conditions and offset S deficiency (Beard, 1973; Vargas, 2005). It has also been used to try to prevent encroachment of annual bluegrass (Poa annua L.) into greens (Joseph Vargas, personal communication). Conversely, the negative consequence of applying S0, especially in relation to its reduction, has not been adequately addressed in the turfgrass management literature.
Applying S0 can help to generate and perpetuate favorable environmental conditions for release of S2– to soil, hence BL formation (Berndt and Vargas, 1992). These include low redox and abundant S. Consider that applying S0 to soil increases acidity through microbial oxidation of S0 to H2SO4:
This reaction is widely known, but what has not been emphasized is this process consumes O2 and increases H+ thereby reducing redox potential as pe + pH (Berndt and Vargas, 1992).
Oxidation of S0 is obviously not the only source of low redox in soils. It is well known that water-logging and the presence of restrictive layers in soil can exclude diffusion of O2 and in turn generate low redox. Excessive microbial respiration can also cause low redox. The point is, however, that once low redox exists, and provided S0 is abundant, as when it is applied to greens, the formation of S2– through S0 reduction may occur rapidly, and the rate of BL formation and development may be accelerated. In the current research introducing S0 and cysteine into the reaction vessel and continually purging it with O2-free N2 generated favorable environmental conditions for formation of S2–, a redox potential of ≈–340 mV (Kaspar and Tiedje, 1982), and plenty of S0 to be reduced.
Avoiding an accelerated rate of BL formation in putting greens requires limiting input of S0 and sustaining high soil redox potential through chemical or physical means such as fertilizing with nitrate (NO3 –) and aerifying.
Contributor Notes
To whom reprint requests should be addressed; e-mail lberndt@edison.edu
