Radiorespirometric assays for the detection of biogenic sulfides from sulfate-reducing bacteria



Antonio Carlos Augusto da Costa, Instituto de Química, Programa de Pós-Graduação em Química, Universidade do Estado do Rio de Janeiro, Pavilhão Haroldo Lisboa da Cunha, R. São Francisco Xavier 524, Maracanã, Rio de Janeiro, Brasil. E-mail:



The detection of trace concentrations of biogenic sulfides can be carried out through radiorespirometric assays. The objective of this work was to improve the methodology for detection of H2S in trace concentrations, to correlate with sulfate-reducing bacterial activity.

Methods and Results

Serial dilutions of synthetic sea water with a pure culture of Desulfovibrio alaskensis, a mixed anaerobic microbial culture and a natural saline sample from a petroleum offshore platform indicated that dilutions were followed, accordingly, by sulfate reduction.


Tests performed indicated that increasing the time of incubation of a mixed anaerobic microbial culture contributed to an increase in the sulfate reduction rates, as well as the amount of carbon source and inoculum.

Significance and Impact of the Study

The technique here developed proved to be a rapid test for the detection of biogenic sulfides, particularly those associated with corrosion products, being an useful tool for monitoring and controlling oil/water storage tanks, petroleum continental platforms and several types of reservoirs.


One of the first attempts to use radiorespirometric tests for the detection of microbial activity was performed by Levin and Straat (1976), studying the labelled release of carbon for extraterrestrial life detection onboard the Viking spacecraft on the surface of Mars. At that time, the authors described the intention of detecting heterotrophic life by supplying a dilute solution of radioactive organic substrates to a sample of Martian soil and monitoring the evolution of gas.

Later, Hardy and Syrett (1983) used the same technique to evaluate inhibitors of sulfate-reducing bacteria (SRB), concluding for its effectiveness and short-time duration of the technique.

After that, a few papers were published on the subject, all of them emphasizing the classical monitoring methods may give misleading results and that radiorespirometric techniques may give a better understanding of the processes taking place within the systems (Maxwell and Hamilton 1986; Hamilton et al., 1989; NACE 2004)

Sulfate-reducing bacteria constitute a common, widespread and harmful microbial group of great environmental and economic impact for the petroleum industry, commonly found in injection waters, corrosion products and usually constituting a biofilm matrix scraped from the surface of corrosion coupons (Sanders 1988). They have the ability of using sulfate as final electron acceptor in the respiration, with H2S as the final metabolic product (Barton 1995; Roychoudhury et al. 2003). Problems associated with the anaerobic corrosion due to the activity of SRB cells become even worse due to the use of sea water for secondary oil recovery in offshore platforms (Gaylarde 1990; Edyvean 1991). The resulting environment favours bacterial growth, offering anaerobic conditions with sulfate and nutrients. In this particular case, during the production and storage of petroleum, there are many points at which biocorrosion is a problem (Araújo-Jorge et al. 1992). In the case of growth of a consortium of microbes, SRB adhere to inert surfaces with subsequent development of biofilms, which mediate the interaction between metal surfaces and the environment (Videla 1989). Routine methods to prevent and treat microbial contamination and biodeterioration involve the use of biocides, which are toxic and with some degree of environmental impact. Then, an accurate diagnosis of biocorrosion is always required prior to a treatment decision (Gaylarde and Johnston 1984; Mckenzie and Hamilton 1992). According to Maxwell et al. (2002), it is extremely difficult to estimate the costs related to corrosive processes attributed to the activity of micro-organisms (SRB and other bacteria) in the oil industry. However, in the last few years, the costs for the control of their activity are significant, with annual income values ordering about US$150 000 per continental platform, only considering the use of biocides.

Recently, considerable efforts have been carried out for the development of new methods for SRB enumeration. In general, the most usual methods to enumerate SRB can be divided in two categories: direct detection methods and culture methods (Flemming and Ingvorsen 1998). The last category includes methods based on the most probable number (MPN) technique, a worldwide used method. However, this technique does not provide an accurate determination of the number of SRB cells within a natural ecosystem or sample. According to Oliveira (2005), the uncertainty in this method can reach 68%, and according to American FDA, several improvements and statistical changes were introduced in the MPN method to decrease bias and uncertainties. McCrady (1915) published the first precise estimation of the bacterial numbers by the MPN, followed by Halvorson and Ziegler (1933), Eisenhar and Wilson (1943) and Cochran (1950), who published articles on the statistical foundations of the MPN technique. Woodward (1957) recommended that MPN tables should omit those combinations of positive tubes (high for low concentrations and low for high concentrations) that are so improbable that they raise concerns about laboratory errors or possible contaminations. de Man (1983) published a confidence interval method that was modified to make the tables for this procedure. In relation to direct detection methods, like radiorespirometric assays, they constitute a promising technique for SRB activity determination, but still require some improvement for practical applications.

The radiotracer technique used for radiorespirometric assays for sulfide determination was introduced by Ivanov (1956). This method consists in adding trace amounts of math formula to a natural sample of sea water, injection water or any other process water. In the end of an incubation period, the biologically reduced sulfate can be quantified. Rosser and Hamilton (1983) developed a radiorespirometric assay for the study of sea sediments and, later, adopted it to monitor the microbial sulfate reduction activity in biofilms formed on mild steel corrosion coupons. According to them, the efficiency of 35S-sulfide recovery under their experimental conditions (16–20 h equilibrium in a shaking water bath at 100 rev min−1 at 35°C) was about 96%.

Fossing and Jorgensen (1989) developed a more advanced technique, which separates the reduced 35S-sulfur by reflux distillation. The disadvantage of this method is the time-consuming procedures for the test and the more expensive equipment required. As stated by Ulrich et al. (1997), the method developed for quantifying reduced inorganic sulfur from sediments and water may also be used for estimation of sulfate reduction rates. In this case, the methodology based on the use of radiotracers was adapted to a simple and less time-consuming procedure. The great advantage of these assays is the use of math formula by SRB and, consequently, the detection of the H235S produced, through liquid phase scintillation, in shorter periods, and with greater specificity. Hamilton et al. (1989) concluded that sulfide levels may offer the most reliable parameter for the detection of long-term ongoing corrosive processes in which SRB are implicated. Based on these considerations, the main objective of this work was to improve the methodology for the detection of H2S in low concentrations, to correlate sulfate reduction rates with SRB activity and enumeration by the MPN method.

To reach this goal, a series of experiments were conducted aiming at determining ideal conditions for the production of biogenic sulfides under conservative conditions of nutritional substances in the medium. This seemed to be a suitable simulation of environments with low availability of nutrients, as expected in the natural environments where SRB cells are present in sea water. At last, this work attempted to correlate the number of SRB cells present in distinct laboratory-made and natural samples and the corresponding production of biogenic sulfides. All the experiments were performed in a reaction flask specially designed for this purpose and patented by the authors and others at Brazilian INPI under number PI 0904216-4 A2 (da Costa et al. 2012).

Materials and methods

Strains used in the study

For the development of the present tests, distinct strains/samples were tested, as follows: (i) A pure culture of the sulfate-reducing bacteria Desulfovibrio alaskensis (NCIMB 13491T or DSM 16109T) was used in radiorespirometric test 1. This bacterium was recovered from a soured oil well in Alaska. These are Gram-negative, vibrio-shaped and motile by means of a single polar flagellum. The carbon and energy sources used by the isolate, and the salinity, temperature and pH ranges facilitating its growth, proved to be typical of a partial lactate-oxidizing, moderately halophilic, mesophilic and sulfate-reducing bacterium. It clustered closely with Desulfovibrio vietnamensis DSM 10520T, from a similar habitat. However, there is sufficient dissimilarity at the DNA sequence level between D. vietnamensis DSM 10520T and D. alaskensis strain (10·2% similarity) to propose that it belongs to a separate species within the genus Desulfovibrio. Based on the results obtained, the name D. alaskensis sp. nov. is therefore proposed, with Al1T (or NCIMB 13491T or DSM 16109T) as the type strain; (ii) A mixed culture of anaerobic bacteria, obtained from the bottom of an oil/water storage tank, without biological characterization, however, with a high sulfide production activity, was used in radiorespirometric tests 2, 5 and 6; (iii) A natural saline sample from an offshore platform was used in radiorespirometric tests 3 and 4.

Culture conditions

The pure culture tested in the radiorespirometric assays was the species D. alaskensis grown in Postgate C medium in synthetic Tropic Marin™sea water. The inoculation of the cells (10% v/v), in 50-ml flasks, was carried out to obtain a test culture aged 18 h, in the exponential growth phase of the culture. Some variations in this procedure were included in the present work: substitution of the pure bacterial culture by a mixed culture of anaerobic bacteria and substitution of the pure bacterial culture by a natural saline sample from an offshore platform, with a high microbiological activity. From test to test, parameters such as volume of saline solution, amount of Postgate C medium, number of dilutions and time of incubation were changed. Specific objectives for every radiorespirometric test performed are presented in Table 1.

Table 1. Summary of the radiorespirometric tests performed
TestTropic marin sea water or 3·5% saline solution (ml)Postgate medium (ml)Type of microbial culture or sampleNumber of dilutionsTime of incubation (h)Objective of the test
  1. a

    In this test, two sets of experiments were conducted: one using TM synthetic sea water and a second one using 3·5% saline solution.

14·3 ml0·2Desulfovibrio alaskensis (0·5 ml)34To check the effect of dilutions in a pure culture in the biogenic production of sulfides
24·3 ml0·2Anaerobic mixed culture (0·5 ml)14To check the effect of dilutions in an anaerobic mixed culture in the biogenic production of sulfides
33·8 ml0·2Natural saline sample (1·0 ml)24To check the effect of dilutions in a natural saline sample from an offshore petroleum platform
43·8 ml0·2Natural saline sample (1·0 ml)14 and 6To test the effect on the incubation time
53·8 mla0·2/0·5Anaerobic mixed culture (1·0 and 0·7 ml)06To check more conservative conditions for the production of biogenic sulfides. Also to investigate the importance of Postgate C medium. Simulation of deep-sea conditions
63·8 ml0·2Anaerobic mixed culture (1·0 ml)126To check the reproducibility of the method for decreasing most probable number populations

The steps of each radiorespirometric assay performed are described as follows. To each test, initially, it is necessary to add the culture medium (Postgate C medium) to a glass tube, with the help of a syringe and needle; the side and upper septa of the flask are closed; and the system is sterilized by autoclaving at 120°C for 20 min. The second step of the test is the addition of a 1 μCi math formula (IPEN/USP) solution to the sterilized flask, and the further addition of a filter paper support to the flask, containing 0·75 ml of a 2 mol l−1 zinc acetate solution, through the upper septum of the flask. The filter paper support must not touch the liquid in the flask. It is essential to observe whether both septa of the flask are locked, to start the introduction of nitrogen gas, to create an anaerobic atmosphere inside the flask, suitable for SRB cells. Immediately after, a known volume of microbial culture or natural sample with biological activity is transferred to the flask, through the side septum. The septa of the flasks are closed with parafilm, to prevent the release of H235S, and the glass tubes are incubated at 30°C at predetermined periods. At the final incubation period, 0·5 ml of a 6 mol l−1 hydrochloric acid solution is introduced in the solution through the side septum of the flask, stimulating the release of biogenic sulfides due to the acid attack. To stimulate the release, this procedure is conducted in a rotary shaker, at 35°C and 100 rpm, during 16 h. At last, the filter paper supports are removed from the tubes to have the biosulfides quantified through liquid phase scintillation. The filter papers are introduced in a glass flask containing 15 ml of scintillation solution. The time needed for the quantification of biogenic sulfides was predetermined as 100 min for each sample (counts of ß-emissions from 35S) and also for the control flasks of standard solutions.

The results obtained can be expressed as% math formula reduced that represents a direct function of the biological activity. Alternatively, as a mean of comparison, biological reduction of sulfates can be correlated with MPN cells quantification (McCrady 1915), to check the possible correlation between the two techniques.

Each radiorespirometric test was followed by a quantification of MPN cells in the same sample or inoculum. The tests also included radioactive and microbiological controls. In the case of the radioactive control test, all the reagents are added to the tubes, except the inoculum. In the case of the microbiological control, all the reagents are introduced, except the radiotracer.

Inoculum preparation

If necessary, when dilution of the test cultures and samples was performed during the test, 1·0 ml of the culture was transferred to a penicillin flask containing 9·0 ml of artificial sea water and homogenized. From that initial dilution, 1·0 ml was transferred to another flask in the same conditions, until the final dilution needed. After each dilution performed, the MPN of cells was performed.

Procedures for the preparation of the prototype flasks used in the tests were performed inside a laminar flux chamber and equipped with nitrogen gas fluxes. The locker and rubber septum located at the top of the prototype flask were removed and the radiotracer was introduced (math formula in a known mass of a solution with a fixed activity) directly in the solution containing the sea water and Postgate C medium immediately, except in the microbiological control tests. Following this procedure, flasks were washed with 1·0 ml of sterile sea water. After that, the filter paper support was introduced; the prototype was completely closed and purged with nitrogen gas to initiate the tests. We can confirm that the total amount of H2S produced during sulfate reduction, released in the reaction flask, was recovered by the filter paper. We tested more than one type of solvent solution to wet the filter paper to recover H2S, and the shaking conditions during H2S evolution were enough to ensure total H2S recovery. To confirm this, we also made scintillometric determinations in the remaining liquid solution to validate that all H2S produced was in the filter paper. This information was incorporated in the present version of the article.

Radiorespirometric assays performed

All the tests described below were conducted five times to ensure reproducibility of the results. Results reported in this work correspond to the average values obtained from five equivalent experiments with standard deviations. Sulfide quantification by scintillometry presented a detection limit of 4 × 10−3 cpm. These counts were converted to nmol math formula reduced per day. Final results presented were expressed as per cent math formula reduction. These radiorespirometric tests for the detection of biogenic sulfides were carried out with the addition of a radiotracer (math formula) to the sample assayed at the trapping system as represented below (Fig. 1). This prototype was protected for industrial safety at Brazilian INPI, as previously informed.

Figure 1.

Schematic representation of the trapping system used in the development of the radiorespirometric assays of math formula reduction by sulfate-reducing bacterial cells present in natural samples.

Radiorespirometric test 1

The medium used was the synthetic sea water Tropic Marin™ (Forsters & Smith, Rhinelander, WI, USA) (4·3 ml) added to 0·2 ml of Postgate C medium and 0·5 ml of a pure culture of D. alaskensis. Three consecutive tenfold dilutions of this solution were prepared. The incubation period was equal to 4 h. This test was performed according to the previous description.

Radiorespirometric test 2

The medium used was the synthetic sea water Tropic Marin (4·3 ml) added to 0·2 ml of Postgate C medium and 0·5 ml of a mixed anaerobic microbial culture. One tenfold dilution of this solution was prepared. The incubation period was equal to 4 h. This test was performed according to the previous description.

Radiorespirometric test 3

The medium used was the synthetic sea water Tropic Marin™(3·8 ml) added to 0·2 ml of Postgate C medium and 1·0 ml of a natural saline sample from an offshore platform, with a high microbiological activity. Two tenfold dilutions of this solution were prepared. The incubation period tested was 4 h. This test was performed according to the previous description.

Radiorespirometric test 4

The medium used was the synthetic sea water Tropic Marin (3·8 ml) added to 0·2 ml of Postgate C medium and 1·0 ml of a natural saline sample from an offshore platform, with a high microbiological activity. One tenfold dilution of this solution was prepared. The incubation periods were 4 and 6 h.

Radiorespirometric test 5

The medium used was the synthetic sea water Tropic Marin™(3·8 ml) added to 0·2 or 0·5 ml of Postgate C medium and 1·0 ml of a mixed anaerobic microbial culture. In substitution to the Tropic Marin™, a saline solution (3·5%) was tested in the same previous conditions. The incubation period was 6 h, for all the experiments performed in this set.

Radiorespirometric test 6

An anaerobic culture containing SRB cells at 109 MPN ml−1 was used in this test, after serial dilutions that produced distinct anaerobic cultures containing SRB cells from 108 to 102 MPN ml−1. The medium used was the synthetic sea water Tropic Marin™(3·8 ml) added to 0·2 ml of Postgate C medium and 1·0 ml of each diluted mixed anaerobic microbial culture containing SRB cells. The incubation period was 6 h, for all the experiments performed in this set.

A summary of all the experiments can be found in Table 1.


For the first set of radiorespirometric assays, we expected to find a direct correlation between math formula reduction and MPN quantification. However, from the results obtained, we could observe that it was not easily observed due to the uncertainty associated with cell quantification by MPN technique.

Figure 2 presents the results obtained from test 1 in terms of math formula reduction, against the corresponding MPN results used in the same test.

Figure 2.

Radiorespirometric test 1. Conditions: synthetic sea water Tropic Marin™ solution, Postgate C medium, Desulfovibrio alaskensis pure culture, incubation period of 4 h, three serial tenfold dilutions.

The results indicate a decrease in the per cent values of sulfate reduced ranging from 4·5 to 3·5 × 10−4%, for the three dilutions made; however, it can be seen that these results were statistically significant only from the nondiluted to the first dilution. For the remaining dilutions, the results were statistically equivalent.

The results observed for the MPN of sulfate-reducing bacteria showed a concentration of 7·5 × 104 MPN ml−1, in the first sample. For the remaining dilutions, cell concentrations ranged from 55 to 5·5 MPN ml−1, a fact that was not expected, as serial dilutions were performed.

In the second radiorespirometric test, we expected to observe math formula reduction by a mixed microbial anaerobic culture in comparison with the previous experiment performed with the use of a pure culture of Desulfovibrio alaskensis. According to our expectation, math formula reduction was much more pronounced in comparison with the previous experiment.

In test 2, however, just one dilution of the inoculum was performed. The results from radiorespirometric test 2 are presented in Fig. 3. The results indicated a markedly higher sulfate reduction by the mixed anaerobic bacterial population: 18·5 × 10−4% for the original bacterial culture and 9·0 × 10−4% for the diluted bacterial culture.

Figure 3.

Radiorespirometric test 2. Conditions: synthetic sea water Tropic Marin™solution, Postgate C medium, mixed anaerobic culture, incubation period of 4 h, one tenfold dilution.

In the third radiorespirometric test, authors investigated math formula reduction obtained by a natural saline sample from an offshore platform, probably associated with a more specific and active microbial consortium, able to reduce sulfate at rates considerably higher than previously observed.

Analogously, results from test 3 are presented in Fig. 4. The results obtained in this test indicated a much higher microbial sulfate reduction, markedly greater than the results obtained with pure and mixed microbial cultures. The sample that was used as inoculum presented a high population of SRB cells and a high sulfate reduction activity, equal to 366 × 10−4% without dilution and 183 and 22·4 × 10−4% after the subsequent tenfold dilutions and 4 h of incubation.

Figure 4.

Radiorespirometric test 3. Conditions: synthetic sea water Tropic Marin™solution, Postgate C medium, natural saline sample from an offshore platform with high microbiological activity, incubation period of 4 h with two tenfold consecutive dilutions.

After concluding that the anaerobic consortium was much more active for math formula reduction, in comparison with the previous investigated conditions, we decided to study the effect of more conservative conditions on math formula reduction. This would bring valuable information about a more realistic condition, probably found in deep-ocean environments.

Again, during test 4, the effect of the dilution of the sample was tested, to evaluate the performance of the method under more conservative conditions (lack of nutrients for microbial growth, simulating seawater conditions). Results are presented in Fig. 5.

Figure 5.

Radiorespirometric test 4. Conditions: synthetic sea water Tropic Marin™solution, Postgate C medium, natural saline sample from an offshore platform with high microbiological activity, incubation periods of 6 h (1st and 2nd Bars) and 4 h (3rd and 4th bars) with one tenfold dilution for each incubation period.

In order to try to optimize math formula reduction, based on the previous results obtained, we decided to perform the fifth radiorespirometric assay, combining optimized factors, such as inoculum, time, carbon source and salts concentration.

Results from test 5 are presented in Fig. 6. The results presented in Fig. 6 indicated, for all the tests performed, a high microbial reduction of sulfates, in comparison with all the previous tests. This is probably due to a combination of factors that were changed in test 5, based on previous tests: (i) The amount of inoculum used in the tests, equal to 0·7 or 1·0 ml, was higher than the previous ones; (ii) Time of incubation was 6 h; (iii) The increase in the amount of carbon source (Postgate C medium) was associated with a decrease in the total volume of reaction medium; and, (iv) Change in the medium was due to the substitution of Tropic Marin™sea water by saline solution.

Figure 6.

Radiorespirometric test 5. Conditions: 1st bar: saline solution (3·5%), 0·2 ml of Postgate C medium and 0·7 ml of mixed anaerobic culture; 2nd bar: saline solution (3·5%), 0·5 ml of Postgate C medium and 0·7 ml of mixed anaerobic culture; 3rd bar: saline solution (3·5%), 0·2 ml of Postgate C medium and 1·0 ml of mixed anaerobic culture; 4th bar: saline solution (3·5%), 0·5 ml of Postgate C medium and 1·0 ml of mixed anaerobic culture.

As it was previously observed, those sulfate reduction rates are dependent on a series of parameters and environmental conditions; thus, a new test was planned under optimized conditions (radiorespirometric test 6).

In this test, we worked with a mixed anaerobic culture with a highly active sulfate-reducing activity. This culture presented a SRB population equal to 108 MPN ml−1; after serial dilutions, subcultures were produced with concentrations up to 102 MPN ml−1. After that, with each dilution obtained, a new radiorespirometric test was performed, including the optimized conditions obtained in previous tests. The results obtained are presented in Figs 7 and 8.

Figure 7.

Radiorespirometric test 6. Correlation between most probable number technique and per cent sulfate reduced, for microbial concentrations found up to 102–105 MPN ml−1 (x axys). Dark circles (MPN ml−1) and open circles (% sulfate reduction). MPN, most probable number.

Figure 8.

Radiorespirometric test 6. Correlation between most probable number technique and per cent sulfate reduced, for microbial concentrations found higher than 105 MPN ml−1 (x axys). Dark circles (MPN ml−1) and open circles (% sulfate reduction). MPN, most probable number.

From Fig. 7, it is clearly observed that for cell concentrations ranging from 102 up to 105 MPN ml−1, accordingly sulfate reduction took place, indicating that for this population range, sulfate reduction can be correlated with SRB cell numbers. In addition, from sulfate reduction measurements, an approximate estimation of SRB cell numbers can be achieved. Of course, the idea is to have a fast procedure for the quantification of sulfide production, irrespective of the cell number present in the sample. However, the opposite procedure can be performed: from sulfate reduction rates, an estimation of the number of active microbial cells can be achieved, to evaluate possible biocide treatments.

From Fig. 8, the same observations from Fig. 7 apply, however, with a different correlation than that previously observed. Here, it can be seen that for SRB concentrations higher than 105 MPN ml−1, the reduction in sulfate does not fit the exact pattern, as previously observed. Anyway, it can be seen that if we consider the errors associated with the quantification of SRB cells, sulfate reduction rates can be correlated with cell numbers, as well.


Before discussing the main results obtained, a few statistical concepts involved in the application of MPN technique will be presented, according to procedures adopted by the United States Food and Drug Administration (2010).

The 95% confidence intervals in the MPN tables indicate that before tubes are inoculated, there is a chance that at least 95% of the confidence interval associated with any result obtained will enclose the actual concentration of cells. Consequently, there are many possibilities of intervals that meet this criterion. de Man (1983) suggested to calculate confidence limits iteratively from the smaller to the higher concentrations. However, there is an increasing tendency to estimate cell concentrations (when MPN technique is selected) based on slight shifts in intervals by iterating from the greater to the smaller concentrations.

It is widely known that a MPN can be obtained for any number of tubes and dilutions, and it is also known that MPN based on three dilutions corresponds to very close approximations to the procedure based on four or more dilutions. Several possibilities arise from this assumption, such as one or more dilutions can show all tubes positive, or no dilutions show all tubes positive. Conventionally, the available methods require that no excluded lower dilutions may have any negative tubes. However, based on the procedure suggested by FDA, when the highest dilution that makes all tubes positive follows a lower dilution that has one or more negative tubes, no tubes should be excluded. This is performed to reduce underestimations in the detection of target groups of microbes.

This procedure, however, must be used considering low and high confidence values. So, in the present work, although average values are reported, the confidence limits were considered; in these cases, distinct results reported are statistically different results.

The most common problem with SRB in offshore systems is sulfide corrosion. Monitoring of SRB has been almost solely by counting of bacteria and chemical analysis of the bulk phase. However, bacterial counts give little information on the in situ bacterial activity, that is, the sulfate reduction/sulfide production rates. This technique that uses labelled sulfate has been implemented, to allow for the determination of SRB activity, trying to correlate activity with microbial numbers.

The data from Fig. 2 confirm the difficulty to correlate the two distinct techniques used; the liquid phase scintillation is a highly precise technique, while the most probable number technique is associated with a high level of uncertainty, particularly for small cell concentrations, as those observed in this test.

These results indicated that it was not recommended to work with such number of dilutions, under the conditions here stated. Those conclusions were partially solved in the radiorespirometric test 2, performed with the use of a mixed culture of anaerobic bacteria, under the same conditions used in the previous test 1.

One more time, from the results presented in Fig. 3, it was not observed a direct and perfect quantitative correlation between MPN of cells and sulfate reduction. However, a good correlation could be reached. This conclusion could be seen because the bacterial population decreased five times after dilution and the same level of sulfate reduction was not observed. Again, it is important to emphasize that these results were obtained in the previously described conditions of the test, and it must be considered that not all the cells in the culture are present in the same level of metabolic activity.

From test to test, the optimization of distinct parameters was performed to reach a closer correlation between sulfate reduction and microbial quantification.

If results from test 1 are compared with the results obtained in test 2, it can be observed a higher microbial sulfate reduction in test 2, indicating that the mixed culture is presently metabolically more active than the pure culture of Desulfovibrio alaskensis, even though it is known that the mixed culture is not exclusively constituted of sulfate-reducing species. These results, however, indicate an expected microbial activity, although not directly and quantitatively proved: serial dilutions are followed by a decreasing sulfate-reducing activity. This is the expected result that confirms that the use of the MPN technique, without a combination with another technique, is not recommended. As known, MPN technique is a time-consuming technique, beyond being highly imprecise due to the high errors associated. Thus, the use of an accessory chemical technique that could be used to indicate the presence of metabolic activity associated with SRB cells can lead to a fast and time-saving procedure.

Considering the results obtained from test 1 (pure culture of SRB) and test 2 (mixed culture of anaerobic bacteria), test 3 was planned. The objective of this test was to observe whether it would be possible to quantify the biogenic sulfides produced by a natural saline sample, collected from an offshore petroleum platform, probably with a high population of microbial cells, particularly SRB cells. This would serve as a standard for natural samples from several similar environments, to detect the potential of those samples to produce biogenic sulfides, irrespective of the number of cells present in the samples.

The decreasing microbial sulfate reduction was not quantitatively followed by the SRB number detected, although the qualitative behaviour followed the expected profile. Based on these results, test 4 was performed with a natural saline sample from an offshore platform, again during 4 and 6 h of incubation. This was performed to check the need for higher incubation periods and the corresponding sulfide production. Under the present conditions of test, it could be confirmed that the best incubation period for radiorespirometric assays of microbiological sulfate reduction is 6 h, both for pure and mixed cultures, as well as for natural samples with a high biological activity.

The natural sample used in test 4 presented a higher microbiological sulfate-reducing activity after 6 h of incubation in comparison with 4 h, both for the diluted and nondiluted samples tested. Results obtained after 6 h of incubation presented markedly higher microbiological activity, emphasizing the importance of the incubation period of the samples. A comparison between results from test 4 and test 3 confirmed that an increase in the microbial population present in the sample not necessarily produces a higher amount of biogenic sulfides.

For instance, microbial populations in test 3 ranged from 2·25 to 12·5 × 103 MPN ml−1, and the corresponding biological reduction in sulfates ranged from 22·4 to 366 × 10−4%35SO4. On the other hand, microbial populations in test 4 ranged from 1·25 to 125 × 105 MPN ml−1, and the corresponding biological reduction in sulfate ranged from 1·9 to 28·1 × 10−4%35SO4. This clearly indicated that the higher population from test 4 produced less sulfide than the smaller population from test 3. These results corroborate the importance of the quantification of the metabolic product and not necessarily the microbial population; this can be high, but with a decreased biological activity.

After concluding that 6 h was a better incubation time, the next test (test 5) aimed at introducing changes in the composition of the medium, for a more conservative environment. To reach this goal, Tropic Marin™sea water was substituted by a saline solution (3·5% w/v) and the amount of Postgate C medium used was also changed, consequently changing the amount of carbon source for microbial reduction of sulfates. However, it is important to emphasize that without a minimum amount of carbon source, no microbiological sulfate reduction could be possible. To reach this goal, the objective of this test was to verify whether the water sample itself, when transferred to the reaction flask, could supply the reaction with the adequate amount of carbon for bacterial growth. The importance of this procedure is to simulate seawater conditions, where the availability of carbon source for SRB cells is limited. Beyond these conditions, this procedure would facilitate the preferential use of math formula by SRB cells, once the availability of math formula would also be reduced.

The change in the composition of the medium from Tropic Marin™sea water to saline solution in the presence of 0·2 ml of Postgate C medium (Fig. 6, 1st and 3rd bars) did not contribute to a higher microbial reduction of sulfates, although a higher amount of mixed anaerobic culture has been used. It can be concluded that sulfate reduction was observed in various media, including a saline solution with a low availability of carbon. This medium contributed to the preferential use of the radiotracer by the bacterial cells.

On the other hand, the microbial reduction of sulfates was markedly increased when the saline solution was tested, in higher concentrations of Postgate C medium. A possible explanation for this fact is that the mixed culture used in these experiments was obtained from a natural environment, where the availability of organic substances is quite limited. It can be thus concluded that under these conditions, we tried to simulate a conservative natural environment that contributes to the maintenance and activity of this type of microbial population, where the cells were already adapted. This is a possible explanation for higher sulfide production under more conservative conditions if compared with previous experiments performed. However, it was still necessary to find what type of correlation could be envisaged between sulfate reduction and the amount of microbial cells responsible for this reduction.

From 102 to 105 MPN ml−1 of an anaerobic mixed culture, sulfate reduction took place, following the same behaviour as observed for cells serial dilutions. This is an indication of the wide range of sample applications of the technique. For higher SRB cell concentrations, this equivalent production of sulfides was not directly observed, only after considering the errors associated with cell quantification. Thus, qualitatively it was possible to correlate microbial reduction of sulfates (activity) and MPN results for samples under 105 MPN ml−1.

Ulrich et al. (1997) compared a passive extraction method and a distillation method described by Fossing and Jorgensen (1989) and found virtually identical results in highly and moderately active sediments, with some advantages of passive extraction, as the capacity to process a large number of samples in a short time.

In a comparison among radiotracer techniques, Meier et al. (2000) discussed three different procedures to quantify sulfate reduction rates – two passive extractions, as described by Rosser and Hamilton (1983) and Ulrich et al. (1997), and reflux distillation, as suggested by Fossing and Jorgensen (1989). They concluded that sulfate reduction rates are reproducible both for passive and active extraction methods for recovery of reduced 35S-sulfur. They consider that the diffusion procedure of Ulrich et al. (1997) is a quick and simple method for the extraction of total reduced inorganic sulfurs, with a good efficiency, comparable with that of the distillation procedure.

Maxwell and Hamilton (1986) used the method proposed by Rosser and Hamilton (1983) with the addition of a 3·0 × 1·0 × 0·5 cm 50D mild steel corrosive coupon. They described that the applicability of this assay to high sulfide production systems, as water injection systems and oil storage cells, must be carefully assessed. It is because the filter paper strip can be saturated with sulfide. Furthermore, it is possible to face a decrease in sulfate reduction rates, due to a nutrient limitation in the closed assay system. This is a possible explanation for the results observed in Fig. 8.

Hardy and Syrett (1983) tested a rapid and sensitive method for measuring respiration of SRB with a paper wick containing zinc acetate to trap the labelled sulfur produced by biological activity. Authors observed that with a contact time of 2 h, it was possible to observe distinct inhibiting effects of quaternary ammonium compound–based products. The technique could be used to assess inhibitor efficiencies in <6 h.

Rosser and Hamilton (1983) detected labelled reduced sulfur species in sediments contaminated with domestic and industrial effluents, with an efficiency of 99·3%.

Maxwell (1986) used the radiorespirometric technique to assess SRB-mediated corrosion concluding that it is not possible to correlate microbial production of sulfide and corrosion rates, due to the chemical and physical nature of the materials tested, that play a more important role. The author indicates that understanding of the processes taking place within a particular system, associated with the detection of microbial activity measured by radiorespirometric techniques, can help in the remedial measures to be carried out, with a greater assurance than the conventional methods.

Maxwell and Hamilton (1986) proposed a modified radiorespirometric assay for determining sulfate-reducing activity on biofilms formed on metal surfaces. They concluded that the great advantage of the proposed technique is such that the biofilm can be studied without removing it from the surface. The authors obtained the recovery of practically all sulfide produced by a mixed culture containing SRB cells. However, no attempts to correlate their results with cell numbers were reported.

Hamilton et al. (1988) studied the mechanism of anaerobic microbial corrosion in the marine environment using radiorespirometric techniques, concluding that only a qualitative agreement is possible between sulfate reduction and corrosion rates, probably due to the cathodic protection. The authors indicate that knowledge of the microbial numbers is needed to confirm this, but emphasizes that measuring microbial activity is more important than microbial numbers.

The work of Mckenzie and Hamilton (1992) described the use of a mixed culture of SRB cells, containing 2·5 × 108 cells ml−1, obtained from sediments from the vicinity of an oil production platform. Authors observed a 65% sulfide recovery. If we compare their results with the ones obtained in the present work, we can see that the closest correlation can be found in data from Fig. 7. However, attempts to correlate their results with the present ones must consider not only the size of the inoculum but also the initial concentration of labelled sulfate in the medium and the chemical and physical nature of the samples used.

Edenborn and Brickett (2001) investigated the activity of immobilized SRB cells with the sulfide production from microbial cultures in wetland sediments. The sediment cores were obtained from an underground mine complex with a SRB concentration of 4·5 × 106 cells cm−1 of gel probe. In that work, authors observed a 98·5% sulfate reduction, equivalent to the results obtained in the present work. Although the results obtained by those authors are quite similar to the ones obtained in the present work (Fig. 7), if we consider the concentration of sulfate used by those authors (6 μCi of labelled suldate), we can conclude that our results were much better, considering per cent sulfate reduction, from a more diluted solution, indication a much lower detection of biogenic activity.

However, we did not find similar articles in the scientific literature to compare the present tests performed. Papers found based on the use of this methodology included experiments under distinct conditions, with a different prototype reaction flask, thus making any comparison unsuccessful. No correlation between MPN technique and radiorespirometric techniques were found. That is why, in the present article, authors highlight the results for field applications due to its precision, safety, time-saving operation and novelty for the petroleum industry.

The radiorespirometric assay proved to be a useful tool for the quantification of the microbial sulfate-reducing activity, both for synthetic and natural samples. The great advantage of the present methodology is the possibility of quantification of microbial activity in a maximum of 48 h, in comparison with the conventional 672 h (28 days) of the MPN technique, conventionally used for the determination of the sulfate-reducing activity.

The obtained results showed a good detection limit, making it possible the evaluation of the presence of SRB in concentrations not detected by conventional methods.

The radiorespirometry proved to be an excellent and rapid alternative to quantify microbial sulfate-reducing activity, irrespective of the knowledge available about the microbial population.


The authors would like to thank Comissão Nacional de Energia Nuclear for a scholarship, Instituto de Pesquisas Energéticas e Nucleares/Universidade de São Paulo for providing the radiotracer, Petrobras and Financiadora de Estudos e Projetos for previous financial support to work in this field.