- Email: [email protected]
Identification of conditions underlying production of geosmin and 2-methylisoborneol in a recirculating system
Aquaculture 279 (2008) 85–91
Contents lists available at ScienceDirect
Aquaculture j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / a q u a - o n l i n e
Identification of conditions underlying production of geosmin and 2-methylisoborneol in a recirculating system Lior Guttman, Jaap van Rijn ⁎ Department of Animal Sciences, Faculty of Agricultural, Food and Environmental Quality Sciences, The Hebrew University of Jerusalem, P.O. Box 12, Rehovot, 76100, Israel
A R T I C L E
I N F O
Article history: Received 1 November 2007 Received in revised form 26 March 2008 Accepted 26 March 2008 Keywords: Geosmin 2-Methylisoborneol Off-flavor Recirculating aquaculture system (RAS) Streptomycetes
A B S T R A C T Geosmin and 2-methylisoborneol (MIB) are semi-volatile terpenoid compounds produced as secondary metabolites by benthic and planktonic cyanobacteria, several genera of fungi, and various actinomycetes. These off-flavor compounds pose a heavy economic burden in the aquaculture industry rendering fish unmarketable unless purified by purging with large quantities of clean water. In the present study, the presence of off-flavor compounds was examined in a recirculating aquaculture system (RAS) for tilapia culture. In this zero-discharge system, where water from the fish basins is recirculated through parallel aerobic (drum filter and a trickling filter) and anaerobic treatment loops (sedimentation/digestion basin), concentrations of geosmin and, in particular, MIB were highest in the aerobic treatment loop. Lowest concentrations were detected in the anaerobic treatment loop. This latter finding pointed toward a possible reduction of these compounds in this basin. Two bacterial strains of the streptomycetes family were isolated from the aerobic, organic-rich, drum filter and the nitrifying trickling filter. In vitro tests with these isolates, closely related to Streptomyces roseoflavus and Streptomyces thermocarboxydus, revealed that MIB production exceeded geosmin production under all conditions tested and was significantly higher under aerobic than under anoxic conditions. Under the latter conditions, with nitrate as an electron donor, the S. roseoflavus-like isolate was capable of denitrification. Based on the results obtained in this study, it was concluded that aerobic, organic-rich conditions stimulate the growth of actinomycetes and subsequent production of geosmin and MIB in the system. The observed reduction of these compounds in the anaerobic water treatment component may serve in designing treatment steps aimed at alleviating the problem of geosmin and MIB accumulation in recirculating systems. © 2008 Elsevier B.V. All rights reserved.
1. Introduction Geosmin and 2-methylisoborneol (MIB) are two of the most common compounds that impart an earthy–musty taste and odor to water. In addition to drinking water supplies (Persson, 1983; Silvey and Roach, 1953; Terashima, 1988; Watson, 2004; Yagi, 1983), these compounds are detrimental in many aquaculture facilities (Tucker, 2000). When released into the culture water of such facilities, geosmin and MIB are absorbed through the gills, skin or gastrointestinal tract by lipid-rich fish tissues and often render the fish unmarketable (Howgate, 2004). Geosmin and MIB are semi-volatile terpenoid compounds produced as secondary metabolites by benthic and planktonic cyanobacteria, several genera of fungi, and various actinomycetes (Wood et al., 2001). While cyanobacteria are often associated with geosmin and MIB production in conventional, outdoor ponds (Tucker, 2000), streptomyces are thought to be the organisms responsible for ⁎ Corresponding author. Tel.: +972 8 9489302; fax: +972 8 9465763. E-mail address: [email protected] (J. van Rijn). 0044-8486/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.aquaculture.2008.03.047
production of these compounds in indoor recirculating systems. However, this latter assumption has not been verified by experimental evidence. In both conventional and recirculating aquaculture systems, abatement strategies have mainly been focused on purging offflavored fish with clean water before marketing (Tucker and van der Ploeg, 1999). In conventional ponds, identification of geosmin and MIB-producing cyanobacteria and an understanding of the conditions promoting their growth, have triggered studies aimed at preventing the proliferation of these organisms (Tucker and van der Ploeg, 1999; Zimba et al., 2001; Schrader et al., 2003). Preventive measures such as these have not been applied in recirculating systems where an understanding of the organisms responsible for geosmin and MIB production and their environmental requirements is lacking. In the present study, production of geosmin and MIB was examined in a zero-discharge recirculating systems with aerobic and anaerobic treatment components. Production sites of geosmin and MIB within the system were identified and production rates of these compounds were determined in vitro in crude samples from different system components. In addition, two geosmin and MIB-producing
86
L. Guttman, J. van Rijn / Aquaculture 279 (2008) 85–91
bacteria of the Streptomyces genus were isolated from the system and examined for the production of these compounds under different environmental conditions.
acid method (Strickland and Parsons, 1968). Organic matter was measured as chemical oxygen demand by oxidation of the organic matter with dichromate in the presence of concentrated sulphuric acid according to APHA (1995).
2. Materials and methods 2.3. Analysis of geosmin and MIB 2.1. Experimental system The indoor, zero-discharge recirculating system examined in this study has been previously described by Shnel et al. (2002). Briefly, it comprises 24 culture basins (volume 5 m3, each) of which 12 were operated in a recirculation mode (Fig. 1). Water from the fish basins is led through a mechanical drum filter for removal of organic matter. From here, water is led through two separate treatment stages. One stage, the aerobic treatment stage, consists of a nitrifying trickling filter for ammonia removal followed by an oxygen enrichment step. The second, anoxic treatment stage, consists of a sedimentation/ digestion basin in which organic matter derived from the drum filter is digested and where nitrate is denitrified. Geosmin and MIB concentrations in the system were examined during part (January–July, 2003) of a tilapia (Oreochromis niloticus × Oreochromis aureus) growth cycle lasting from October 2002 until September 2003. 2.2. Chemical analyses Inorganic nutrients were determined monthly during the complete tilapia growth cycle lasting from October, 2002 until September, 2003. Determinations were performed on duplicate samples derived from the effluent of the fish culture basins. Samples for total ammonia nitrogen (NH3-N and NH+4-N, from hereon referred to as TAN), nitrite, nitrate and phosphorus analyses were filtered through 25 mm glass micro fiber filters immediately after sampling and kept cool at 4 °C until analyzed by colorimetric methods using a Spectrosonic 1001 spectrophotometer (Bausch & Lomb, Rochester, NY). TAN was determined after oxidation to indophenol as described by Scheiner (1976). Nitrite was analyzed using the sulfanilamide reagent according to Strickland and Parsons (1968). Nitrate was determined by ultraviolet spectrophotometry as previously described (APHA 1995). Dissolved reactive phosphorus levels were measured by the ascorbic
During the period from January until July, 2003, geosmin and MIB were determined monthly on triplicate samples derived from: (a) the outlet of the fish basins, (b) water within the drum filter, (c) water within the digestion basin, (d) trickling filter inlet, and (e) trickling filter outlet. Analysis of these compounds was performed by using the solid phase micro extraction method based on extraction of these compounds onto a StableFlex fiber (Supelco, Bellefonte, PA) from the headspace of 40 mL glass vials, containing 25 mL water sample. After an initial incubation of 10 min in a water bath at 65 °C, fibers were injected through the Teflon faced silicone septa (Supelco, Bellefonte, PA) of the airtight vials. After an extraction period of 20 min, the fibers were introduced for 1.5 min at 250 °C into the splitless operated injector of a HP5890 (Palo Alto, CA) gas chromatograph with a flame ionized detector (GC-FID). The GC was operated with a MDN-5 fused silica capillary column (30 m × 0.25 mm) of 0.25 μm film thickness (Supelco, Bellefonte, PA). Helium was used as the carrier gas at constant flow rate of 1 mL min− 1. Oven temperature was held at 60 °C for 0.5 min from injection, increased to 100 °C at 30 °C min− 1, followed by an increase to 185 °C at 20 °C min− 1 and to 250 °C at 40 °C min− 1 and held at this maximum temperature for 2.3 min. FID temperature was 280 °C. Identification of geosmin and MIB peaks detected by GCFID was verified by parallel analysis of selected samples with a gas chromatograph coupled to a mass spectrometer detector, (GC–MS model Saturn 2000, Varian Inc. Palo Alto, CA.). 2.4. Laboratory incubation of geosmin and MIB-rich organic matter derived from the system Crude samples of organic matter were collected from the drum filter and trickling filter of the system. Tests were performed on triplicate samples. Ten grams of organic matter were placed in 500 mL Erlenmeyers containing 250 mL distilled water. The Erlenmeyers were
Fig. 1. A schematic presentation of the investigated RAS (not to scale). Solid arrows indicate the direction of the water flow and broken arrows the direction of water + sludge flow. The main water purification processes in the various compartments are indicated.
L. Guttman, J. van Rijn / Aquaculture 279 (2008) 85–91
incubated at 37 °C with vigorous shaking. Geosmin and MIB concentrations in water were measured during the next 27 h. Samples containing 10 g of presterilized organic matter were used as controls. The dry weight of the organic matter was determined in 5 replicates after drying at 105 °C and subsequent burning of the samples at 550 °C according to APHA (1995). 2.5. Isolation of streptomyces Streptomyces spp. isolation was achieved by using casein and starch agar used for enrichment of actinobacteria (Kuster and Williams, 1964). Distinctive streptomyces-like colonies were repeatedly transferred to selective agar plates containing nystatin, polymixin B sulfate, penicillin and cyclohexamide (Sigma-Aldrich, Israel). MIB and geosmin producing bacteria were selected from among the different bacterial colonies growing on the plates based on the typical odors associated with the off-flavor compounds they produced. 2.6. Molecular analyses of isolates Identification of the isolates was based on polymerase chain reaction amplifications of 16 S rRNA with the general bacterial primer 1392-R (Lane, 1991) and the GC-rich actinobacteria-specific primer A235-F (Stach et al., 2003). Primers S661-F and S1218-R, designed by Inbar et al. (2005), were used for further classification of the isolates to the genus level. Each 50 μL of PCR reaction mixture contained the following components: 1.5 units Taq DNA Polymerase (Red Taq; Sigma Chemical Co., St. Louis, MO), 5 μL Sigma Taq Buffer with a final magnesium concentration of 2 mM and 4 mM for actinobacteria and streptomyces primers, respectively, 0.2 mM PCR nucleotide mixture (Promega, Madison, WI), 25 pmol of each primer and 1 μL of DNA template. DNA was extracted by heating live colonies in 30 μL presterilized double distilled water at 95 °C for 10 min. All PCR reactions were conducted in a gradient thermal cycler (model PTC200; MJ Research, Watertown, MA) using the following protocol: initial denaturation at 95 °C for 3 min, followed by 35 (for 235-F/1392-R) or 30 cycles (for S661-F/S1218-R) of denaturation at 94 °C for 30 s, annealing at 62 °C for 30 s and elongation at 72 °C for 1 min. The final elongation step was 2 min at 72 °C. The presence and size of PCR fragments were determined by agarose gel electrophoresis (1.8%) with ethidium bromide staining. 2.7. Sequencing and phylogenetic analysis of isolates DNA was purified from the PCR products by using the UltraClean PCR Clean-up kit (MO BIO, Solana Beach, CA) according to the protocol described by the manufacturer. For bacterial identification, purified products of the PCR reactions were sequenced using the Applied Biosystems PRISM Dye Terminator Cycle Sequencing Ready Reaction kit with Ampli Taq DNA polymerase and the A235-F primer, described above. The sequencing products were analyzed with an Applied Biosystems 377 DNA sequencer. Nucleotide sequences (550 bp) were added to a pre-aligned database of a 16 S rRNA sequences using the aligning tool supplied by ARB Phylogenetic program package (Strunk et al., 2004). Phylogenetic trees were generated with the neighborjoining and maximum likelihood methods with the ARB program package using the Olsen correction method applying a 50% cut-off filter (Strunk and Ludwig, 1998). The topologies of the resulting trees were compared. Branching order was supported by both methods.
87
temperatures for a maximum of 168 h. Aerobic conditions were maintained by vigorous shaking of the Erlenmeyers. Subsamples (27 mL) were taken, every 3 h for the first 12 h, every 4 h for the next 12 h and then every 8 h until no further growth was observed. Samples were examined for MIB and geosmin and for bacterial growth. Bacterial counts were performed by the most probable number method. For this determination, aliquots of the cultures (200 μL) were transferred into the wells of sterile ELISA plates containing 1800 μL casein liquid media and further decimal dilutions were performed for calculation of the bacterial most probable number in the original sample. Mean values of bacterial counts performed during the exponential phase of growth of the isolates were used for calculation of the hourly growth rate constants of the isolates as described by Koch (1981). The aerobic and anoxic growth performance of the isolates was examined in the above described growth medium. Erlenmeyers were kept aerobic by vigorous shaking and anoxic conditions were secured by bubbling argon gas into the airtight Erlenmeyers for 20 min prior to incubation. Absorbance of the culture medium at 600 nm was used as an indicator for growth of the isolates. Geosmin and MIB were measured by the above described solid phase micro extraction method. 3. Results 3.1. System performance Data concerned with the growth performance of tilapia were only available for the period from February until June, 2003. During this period, the average tilapia size increased from 47 g to 326 g. Total production during this seven months period was 31.5 kg m− 3 and maximum fish density was 37 kg m− 3. The average daily specific growth rate was 1.4% day− 1. Fresh water addition was limited to compensate for evaporation losses only and averaged 3.6% of the total system volume per day. The water requirement for production of 1 kg of fish was 213 L. During this period, two organoleptic tests were performed by a certified analytical laboratory (Bactochem, Nes Ziona, Israel) on randomly selected fish collected from the system at 19.5.03 and 30.6.03. In both tests, fish were rated as highly off-flavored. 3.2. Water quality conditions during the culture period Monthly analyses of the main water quality parameters over the entire growth cycle (October, 2002 until September, 2003) in the fish culture basins (Fig. 2) revealed that TAN levels were generally lower than 1 mg TAN-N L− 1, nitrite levels were lower than 0.5 mg NO2-N L− 1,
2.8. In vitro assays with isolates Growth rate measurements at different temperatures were conducted by introduction of one colony of each of the Streptomyces spp. isolates into 1 L Erlenmeyers containing 250 mL casein liquid media (Kuster and Williams, 1964) followed by incubation at different
Fig. 2. Monthly examination of inorganic nitrogen, phosphorus and COD concentrations in the culture water of the RAS. Samples were derived from the combined effluent off all 12 fish tanks at a sampling point just before the discharge of this effluent water into the drum filter (see Fig. 1). Concentrations represent average values of duplicate analyses of two separate water samples.
88
L. Guttman, J. van Rijn / Aquaculture 279 (2008) 85–91
Fig. 3. Geosmin and MIB concentrations in: (A) the culture water of the RAS and (B) the different components of the RAS. Concentrations in the culture water (A) represent the mean value (and standard deviation) of triplicate water samples. Concentrations in the different components (B) represent the average monthly values of duplicate determinations conducted during each month of the examined period (n = 10).
phosphorus levels were around 10 mg PO4-P L− 1, nitrate levels fluctuated between 30 and 70 mg NO3-N L− 1, and organic matter (expressed as COD) fluctuated between 70 and 160 mg O2 L− 1 (Fig. 2). 3.3. Geosmin and MIB production in recirculating system. Geosmin and MIB could be detected at all sampling dates in the fish basin water with higher concentrations of MIB than geosmin. Maximum MIB concentrations were around 170 ng L− 1 compared to 75 ng L− 1 for geosmin (Fig. 3A). Geosmin and, in particular, MIB concentrations differed among the various compartments (Fig. 3B). Generally, concentrations of MIB were higher than those of geosmin in all compartments. Highest mean concentrations of 260 ng L− 1 for MIB and 47 ng L− 1 for geosmin were detected in the drum filter and at the outlet of the trickling filter, respectively. Geosmin and especially MIB were relatively low in the digestion basin and, based on the concentrations of geosmin and MIB in the inlet water to this basin (water from the drum filter), it was here where a reduction of these compounds took place. 3.4. In vitro production of geosmin and MIB by organic matter Differences between inlet and outlet geosmin and MIB concentrations clearly pointed to a production of both compounds in the aerobic trickling filter (Fig. 4A) and drum filter components (Fig. 4B).
Maximum production rates during laboratory incubation (expressed per gram of dry organic matter) of geosmin were 1.725 ng g− 1 h− 1 and 2.446 ng g− 1 h− 1 and maximum MIB production rates were 1.781 ng g− 1 h− 1 and 3.786 ng g− 1 h− 1 in sludge derived from the drum and trickling filters, respectively. Rates of MIB production were higher than those for geosmin in both samples. Production of geosmin and MIB was sustained for longer periods in sludge derived from the tricking filter. 3.5. Isolation and identification of geosmin and MIB-producing streptomyces Enrichment of sludge derived from the drum and trickling filters resulted in the isolation of two bacterial strains capable of geosmin and MIB production. Initial selection of these isolates from a large number of isolates was based on the distinctive earthy–musty smell of the agar plates harboring these isolates. Sequence analyses of 16 S rRNA products obtained with the general and actinobacterial primer pair 1392R/A235F followed by phylogenetic BLAST analysis resulted in the identification of two members of streptomyces. One isolate (Iso. 2A) revealed a 99% sequence identity to Streptomyces roseoflavus and the other (Iso. 4C) a 99% sequence identity to Streptomyces thermocarboxydus. Additional analysis, using the phylogenetic tree analysis ARB program, confirmed the identification of both isolates as members of the streptomyces genus (Fig. 5).
Fig. 4. Production of geosmin and MIB during laboratory incubation of organic matter derived from (A) the trickling filter and (B) the drum filter in distilled water. Concentrations are mean values of triplicate analyses. Bars represent the standard deviation.
L. Guttman, J. van Rijn / Aquaculture 279 (2008) 85–91
89
Fig. 5. Phylogenetic tree based on 16S rRNA gene sequences of the streptomycetes-like isolates Iso 2A and Iso 4C. The topology of the tree is based on the consensus of trees of N 1000 informative positions generated using the Neighbor Joining and Maximum Likelihood methods (ARB phylogenetic package). Scale bar = 1% estimated difference in nucleotide sequence position.
3.6. In vitro geosmin and MIB production by Streptomyces spp.
4. Discussion
Growth rates constants of the two species were conducted at different temperatures (Fig. 6). The optimal temperature for S. roseoflavus-like isolate was around 37 °C while it was higher for the S. thermocarboxydus-like isolate. However, due to a too narrow range of temperatures tested, we were unable to determine the optimal temperature for this latter organism. Incubation of the isolates resulted in considerable growth and geosmin and MIB production in water derived from the fish culture basin and trickling filter outlet (Fig. 7). Pre-sterilized control samples showed no offflavor compounds production and growth of the isolates. A comparison of Streptomyces spp. under aerobic and anoxic conditions revealed that both isolates produced geosmin and MIB during exponential phase of growth with MIB production exceeding geosmin production. Anoxic incubation resulted in similar growth rates for the S. roseoflavus-like isolate while lower growth rates were measured for S. thermocarboxydus-like isolate. MIB production rates under anoxic conditions were considerably lower than under aerobic conditions (Fig. 8). Growth of the S. roseoflavus-like isolate under anoxic conditions coincided with a nitrate decrease in the medium with only a slight increase in ammonia and nitrite (Fig. 9). The observed disappearance of inorganic nitrogen from the medium points to a denitrification capability of this organism.
Although off-flavor of fish in recirculating systems is often encountered, information on geosmin and MIB concentrations in these systems is scarce and to the best of our knowledge, this is the first study in which such information is provided. High concentrations of both compounds were found within all water treatment components during routine operation of the system. Unlike most of the previously investigated aquaculture farms (Schrader and Blevins, 1993; Robertson et al., 2005; Robin et al., 2006), MIB was shown to be the primary off-flavor compound in all parts of the recirculating system examined in this study. Off-flavor in fish culture systems is experienced at MIB concentrations as low as 18 ng L− 1 (Persson and York, 1978) for MIB and at geosmin concentrations as low as 15 ng L− 1 (Persson, 1980). Concentrations of both compounds in the system examined in this study were significantly higher than above threshold concentrations and were also higher than the few systems in which ambient MIB and geosmin concentrations were measured (Klausen et al., 2005; Robertson et al., 2006; Vallod et al., 2007). Streptomyces, present in organic-rich, aerobic parts of recirculating system, were found to be responsible for MIB and geosmin production, as shown by in vitro experiments. In accordance to the field observations, MIB production was higher than geosmin production upon incubation of both isolates in either culture medium or water derived from the
Fig. 6. The effect of temperature on the hourly growth rate constant (μ) of streptomycetes-like isolates Iso. 2A and Iso. 4C. Isolates were incubated in casein liquid media under aerobic conditions and growth rate constants were determined during exponential phase of growth.
Fig. 7. MIB and geosmin production by the two streptomycetes-like isolates in water derived from different sources within the RAS system. Incubation in water derived from the fish culture basin of Iso. 2A (A) and Iso. 4C (B) and incubation in water derived from the trickling filter of Iso. 2A (C) and Iso. 4C (D). Concentrations are mean values of triplicate analyses. Bars represent the standard deviation.
90
L. Guttman, J. van Rijn / Aquaculture 279 (2008) 85–91
Fig. 8. Cell density (determined as O.D.600) and production of MIB and geosmin by the two streptomycetes-like isolates under either aerobic or anaerobic conditions. Incubation under anoxic conditions of Iso. 2A (A) and Iso. 4C (B) and under aerobic conditions of Iso. 2A (C) and Iso. 4C (D). Concentrations are mean values of triplicate analyses. Bars represent the standard deviation.
recirculating system. MIB and geosmin production by streptomyces present in the organic-rich trickling filter and drum filter suggest a benthic nature of these organisms. The relatively high geosmin and MIB concentrations in the aerobic components of the recirculating aquaculture system and the higher in vitro production under aerobic conditions by both Streptomyces isolates present a strong evidence for the important role of oxygen in geosmin and MIB production by streptomyces as previously suggested by Dionigy and Ingram (1994). MIB and geosmin concentrations in the organic-rich digestion basin were relatively low. The lower production rates of these compounds under anaerobic conditions by the isolates might explain why the lowest MIB and geosmin concentrations were found in this compartment. Furthermore, removal of these compounds in this basin seems to take place as concentrations of MIB and geosmin within the basin were lower than in the inlet water (water derived from the drum filter). Preliminary results indicate that both absorption to the lipid
phases of organic compounds and biological degradation underlie the observed decrease of MIB and geosmin in this basin (Guttman and van Rijn, unpubl.). Finally, nitrate reduction without a concomitant increase in either ammonia or nitrite by the S. roseoflavus-like strain, provide circumstantial evidence for the denitrification ability of this isolate. Evidence for the denitrification capability of a number of actinomycetes was provided by Shoun et al. (1998) but to the best of our knowledge has not been determined in off-flavor producing streptomyces. The extent of geosmin and MIB production by the S. roseoflavus-like strain under denitrification conditions is currently under investigation. 5. Conclusions Based on in-situ analyses as well as studies with the bacterial isolates, it was concluded that aerobic, organic-rich conditions stimulate the growth of actinomycetes and subsequent production of geosmin and MIB. Since such conditions are intrinsic, not only in this particular RAS system but in most recirculating systems operated with nitrifying filters, it seems unlikely that alleviation of the problem by modification of the nitrification stage is feasible. Conversely, the observed reduction of these compounds in the anaerobic water treatment component might be considered an important finding toward the alleviation of the problem. Presently, studies are ongoing in which the processes responsible for geosmin and MIB reduction under anaerobic, organic-rich conditions are examined. Acknowledgments
Fig. 9. Inorganic nitrogen fluctuations during anoxic incubation of streptomyces-like Iso. 2A in liquid mineral salt medium in the presence of nitrate. Concentrations are mean values of triplicate analyses. Bars represent the standard deviation.
This research was supported by grant #820-0185-03 from the Chief Scientist Office of the Ministry of Agriculture & Rural Development, Israel. The authors would like to thank Mr. Tamir Ezer and the other staff members of the Ginosar Intensive Aquaculture Research Station for daily operation of the recirculating system.
L. Guttman, J. van Rijn / Aquaculture 279 (2008) 85–91
References APHA., 1995. Standard methods for the examination of water and waste water (19th ed.). American Public Health Association, Washington, DC. Dionigy, C.P., Ingram, D.A., 1994. Effects of temperature and oxygen concentration on geosmin production by Streptomyces tendae and Penicillium expansum. J. Agric. Food Chem. 42, 143–145. Howgate, P., 2004. Tainting of farmed fish by geosmin and 2-methyl-iso-borneol: a review of sensory aspects and of uptake/depuration. Aquaculture 234, 155–181. Inbar, E., Green, S.J., Hadar, Y., Minz, D., 2005. Competing factors of compost concentration and proximity to root affect the distribution of streptomycetes. Microbial Ecol. 50, 73–81. Klausen, C., Nicolaisen, M.H., Strobel, B.W., Warnecke, F., Nielsen, J.L., Jorgensen, N.O.G., 2005. Abundance of actinobacteria and production of geosmin and 2-methylisoborneol in Danish streams and fish ponds. FEMS Microbiol. Ecol. 52, 265–278. Koch, A.L., 1981. Growth measurement. In: Gerhardt, P., Costilow, R.N. (Eds.), Manual of Methods for General Bacteriology. ASM, Washington, DC, pp. 179–208. Kuster, E., Williams, S.T., 1964. Selection of media for isolation of streptomycetes. Nature 202, 928–929. Lane, D.J., 1991. 16S/23S rRNA sequencing. In: Stackebrandt, E., Goodfellow, M. (Eds.), Nucleic acid techniques in bacterial systematics. John Wiley & Sons, Chichester, UK., pp. 115–175. Persson, P.E., 1980. Sensory properties and analysis of two muddy odour compounds, geosmin and 2-methylisoborneol, in water and fish. Water Res. 14, 1113–1118. Persson, P.E., 1983. Off-flavors in aquatic ecosystem—an introduction. Water Sci. Technol. 15, 1–11. Persson, P.E., York, R.K., 1978. Notes on muddy odor: I. Sensory properties and analysis of 2-methylisoborneol in water and fish. Aqua Fennica 8, 83–88. Robertson, R.F., Jauncey, K., Beveridge, M.C.M., Lawton, L.A., 2005. Depuration rates and the sensory threshold concentration of geosmin responsible for earthy–musty taint in rainbow trout, Onchorhynchus mykiss. Aquaculture 245, 89–99. Robertson, R.F., Hammond, A., Jauncey, K., Beveridge, M.C.M., Lawton, L.A., 2006. An investigation into the occurrence of geosmin responsible for earthy–musty taints in UK farmed rainbow trout, Onchorhynchus mykiss. Aquaculture 259, 153–163. Robin, J., Cravedi, J.P., Hillenweck, A., Deshayes, C., Vallod, D., 2006. Off flavor characterization and origin in French trout farming. Aquaculture 260, 128–138. Scheiner, D., 1976. Determination of ammonia and Kjeldahl nitrogen by the indophenol method. Wat. Res. 10, 31–36. Schrader, K.K., Blevins, W.T., 1993. Geosmin-producing species of Streptomyces and Lyngbya from aquaculture ponds. Can. J. Microbiol. 39, 834–840. Schrader, K.K., Nanayakkara, N.P.D., Tucker, C.S., Rimando, A.M., Ganzera, M., Schaneberg, B.T., 2003. Novel derivatives of 9,10-anthraquinone are selective
91
algicides against the musty-odor cyanobacterium Oscillatoria perornata. Appl. Environ. Microbiol. 69, 5319–5327. Shnel, N., Barak, Y., Ezer, T., Dafni, Z., van Rijn, J., 2002. Design and performance of a zerodischarge tilapia recirculating system. Aquacult. Eng. 26, 191–203. Shoun, H., Kano, M., Baba, I., Takaya, N., Matsuo, M., 1998. Denitrification by actinomycetes and purification of dissimilatory nitrite reductase and azurin from Streptomyces thioluteus. J. Bacteriol. 180, 4413–4415. Silvey, J.K.G., Roach, A.W., 1953. Actinomycetes in the Oklahoma City water supply. J. Amer. Water Works Assoc. 64, 35–39. Stach, J.E., Maldonado, L.A., Ward, A.C., Goodfellow, M., Bull, A.T., 2003. New primers for the class Actinobacteria: application to marine and terrestrial environments. Environ. Microbiol. 5, 828–841. Strickland, J.D., Parsons, T.R.A., 1968. A practical handbook of seawater analysis. Fish. Res. Board Canada 167, 77–80. Strunk, O., Ludwig, W., 1998. ARB: A software environment for sequence data. Tech. Univ., Munich. Germany. Strunk, O., Ludwig, W., Westram, R., Richter, L., Meier, H., Yadhukumar, Buchner, A., Lai, T., Steppi, S., Jobb, G., Forster, W., Brettske, I., Gerber, S., Ginhart, A.W., Gross, O., Grumann, S., Hermann, S., Jost, R., Konig, A., Liss, T., Lussmann, R., May, M., Nonhoff, B., Reichel, B., Strehlow, R., Stamatakis, A., Stuckmann, N., Vilbig, A., Lenke, M., Ludwig, T., Bode, A., Schleifer, K.H., 2004. ARB: A software environment for sequence data. Nucleic Acids Res. 32, 1363–1371. Terashima, K., 1988. Reduction of musty odor substances in drinking water — a pilot plant study. Wat. Sci. Technol. 20, 275–281. Tucker, C.S., 2000. Off-flavor problems in aquaculture. Rev. Fish. Sci. 8, 45–88. Tucker, C.S., van der Ploeg, M., 1999. Managing off-flavor problems in pond-raised catfish. Southern Regional Aquaculture Center publication, 192. Vallod, D., Cravedy, J.P., Hillenweck, A., Robin, J., 2007. Analysis of the off-flavor risk in carp production in ponds in Dombes and Forez (France). Aquacult. Int. 15, 287–298. Watson, S.B., 2004. Aquatic taste and odor: a primary signal of drinking-water integrity. J. Toxicol. Environ. 67, 1779–1795. Wood, S., Williams, S.T., White, W.R., 2001. Microbes as a source of earthy flavours in potable water — a review. Int. Biodeterior. Biodegrad. 48, 26–40. Yagi, S., 1983. Taste and flavor of water. J. Japan Water Works Assoc. 590, 65–79. Zimba, P.V., Grimm, C.C., Dionigi, C.P., Weirich, C.R., 2001. Phytoplankton community structure, biomass, and off-flavor: pond size relationships in Louisiana catfish ponds. J. World Aquacult. Soc. 32, 96–104.
Contents lists available at ScienceDirect
Aquaculture j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / a q u a - o n l i n e
Identification of conditions underlying production of geosmin and 2-methylisoborneol in a recirculating system Lior Guttman, Jaap van Rijn ⁎ Department of Animal Sciences, Faculty of Agricultural, Food and Environmental Quality Sciences, The Hebrew University of Jerusalem, P.O. Box 12, Rehovot, 76100, Israel
A R T I C L E
I N F O
Article history: Received 1 November 2007 Received in revised form 26 March 2008 Accepted 26 March 2008 Keywords: Geosmin 2-Methylisoborneol Off-flavor Recirculating aquaculture system (RAS) Streptomycetes
A B S T R A C T Geosmin and 2-methylisoborneol (MIB) are semi-volatile terpenoid compounds produced as secondary metabolites by benthic and planktonic cyanobacteria, several genera of fungi, and various actinomycetes. These off-flavor compounds pose a heavy economic burden in the aquaculture industry rendering fish unmarketable unless purified by purging with large quantities of clean water. In the present study, the presence of off-flavor compounds was examined in a recirculating aquaculture system (RAS) for tilapia culture. In this zero-discharge system, where water from the fish basins is recirculated through parallel aerobic (drum filter and a trickling filter) and anaerobic treatment loops (sedimentation/digestion basin), concentrations of geosmin and, in particular, MIB were highest in the aerobic treatment loop. Lowest concentrations were detected in the anaerobic treatment loop. This latter finding pointed toward a possible reduction of these compounds in this basin. Two bacterial strains of the streptomycetes family were isolated from the aerobic, organic-rich, drum filter and the nitrifying trickling filter. In vitro tests with these isolates, closely related to Streptomyces roseoflavus and Streptomyces thermocarboxydus, revealed that MIB production exceeded geosmin production under all conditions tested and was significantly higher under aerobic than under anoxic conditions. Under the latter conditions, with nitrate as an electron donor, the S. roseoflavus-like isolate was capable of denitrification. Based on the results obtained in this study, it was concluded that aerobic, organic-rich conditions stimulate the growth of actinomycetes and subsequent production of geosmin and MIB in the system. The observed reduction of these compounds in the anaerobic water treatment component may serve in designing treatment steps aimed at alleviating the problem of geosmin and MIB accumulation in recirculating systems. © 2008 Elsevier B.V. All rights reserved.
1. Introduction Geosmin and 2-methylisoborneol (MIB) are two of the most common compounds that impart an earthy–musty taste and odor to water. In addition to drinking water supplies (Persson, 1983; Silvey and Roach, 1953; Terashima, 1988; Watson, 2004; Yagi, 1983), these compounds are detrimental in many aquaculture facilities (Tucker, 2000). When released into the culture water of such facilities, geosmin and MIB are absorbed through the gills, skin or gastrointestinal tract by lipid-rich fish tissues and often render the fish unmarketable (Howgate, 2004). Geosmin and MIB are semi-volatile terpenoid compounds produced as secondary metabolites by benthic and planktonic cyanobacteria, several genera of fungi, and various actinomycetes (Wood et al., 2001). While cyanobacteria are often associated with geosmin and MIB production in conventional, outdoor ponds (Tucker, 2000), streptomyces are thought to be the organisms responsible for ⁎ Corresponding author. Tel.: +972 8 9489302; fax: +972 8 9465763. E-mail address: [email protected] (J. van Rijn). 0044-8486/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.aquaculture.2008.03.047
production of these compounds in indoor recirculating systems. However, this latter assumption has not been verified by experimental evidence. In both conventional and recirculating aquaculture systems, abatement strategies have mainly been focused on purging offflavored fish with clean water before marketing (Tucker and van der Ploeg, 1999). In conventional ponds, identification of geosmin and MIB-producing cyanobacteria and an understanding of the conditions promoting their growth, have triggered studies aimed at preventing the proliferation of these organisms (Tucker and van der Ploeg, 1999; Zimba et al., 2001; Schrader et al., 2003). Preventive measures such as these have not been applied in recirculating systems where an understanding of the organisms responsible for geosmin and MIB production and their environmental requirements is lacking. In the present study, production of geosmin and MIB was examined in a zero-discharge recirculating systems with aerobic and anaerobic treatment components. Production sites of geosmin and MIB within the system were identified and production rates of these compounds were determined in vitro in crude samples from different system components. In addition, two geosmin and MIB-producing
86
L. Guttman, J. van Rijn / Aquaculture 279 (2008) 85–91
bacteria of the Streptomyces genus were isolated from the system and examined for the production of these compounds under different environmental conditions.
acid method (Strickland and Parsons, 1968). Organic matter was measured as chemical oxygen demand by oxidation of the organic matter with dichromate in the presence of concentrated sulphuric acid according to APHA (1995).
2. Materials and methods 2.3. Analysis of geosmin and MIB 2.1. Experimental system The indoor, zero-discharge recirculating system examined in this study has been previously described by Shnel et al. (2002). Briefly, it comprises 24 culture basins (volume 5 m3, each) of which 12 were operated in a recirculation mode (Fig. 1). Water from the fish basins is led through a mechanical drum filter for removal of organic matter. From here, water is led through two separate treatment stages. One stage, the aerobic treatment stage, consists of a nitrifying trickling filter for ammonia removal followed by an oxygen enrichment step. The second, anoxic treatment stage, consists of a sedimentation/ digestion basin in which organic matter derived from the drum filter is digested and where nitrate is denitrified. Geosmin and MIB concentrations in the system were examined during part (January–July, 2003) of a tilapia (Oreochromis niloticus × Oreochromis aureus) growth cycle lasting from October 2002 until September 2003. 2.2. Chemical analyses Inorganic nutrients were determined monthly during the complete tilapia growth cycle lasting from October, 2002 until September, 2003. Determinations were performed on duplicate samples derived from the effluent of the fish culture basins. Samples for total ammonia nitrogen (NH3-N and NH+4-N, from hereon referred to as TAN), nitrite, nitrate and phosphorus analyses were filtered through 25 mm glass micro fiber filters immediately after sampling and kept cool at 4 °C until analyzed by colorimetric methods using a Spectrosonic 1001 spectrophotometer (Bausch & Lomb, Rochester, NY). TAN was determined after oxidation to indophenol as described by Scheiner (1976). Nitrite was analyzed using the sulfanilamide reagent according to Strickland and Parsons (1968). Nitrate was determined by ultraviolet spectrophotometry as previously described (APHA 1995). Dissolved reactive phosphorus levels were measured by the ascorbic
During the period from January until July, 2003, geosmin and MIB were determined monthly on triplicate samples derived from: (a) the outlet of the fish basins, (b) water within the drum filter, (c) water within the digestion basin, (d) trickling filter inlet, and (e) trickling filter outlet. Analysis of these compounds was performed by using the solid phase micro extraction method based on extraction of these compounds onto a StableFlex fiber (Supelco, Bellefonte, PA) from the headspace of 40 mL glass vials, containing 25 mL water sample. After an initial incubation of 10 min in a water bath at 65 °C, fibers were injected through the Teflon faced silicone septa (Supelco, Bellefonte, PA) of the airtight vials. After an extraction period of 20 min, the fibers were introduced for 1.5 min at 250 °C into the splitless operated injector of a HP5890 (Palo Alto, CA) gas chromatograph with a flame ionized detector (GC-FID). The GC was operated with a MDN-5 fused silica capillary column (30 m × 0.25 mm) of 0.25 μm film thickness (Supelco, Bellefonte, PA). Helium was used as the carrier gas at constant flow rate of 1 mL min− 1. Oven temperature was held at 60 °C for 0.5 min from injection, increased to 100 °C at 30 °C min− 1, followed by an increase to 185 °C at 20 °C min− 1 and to 250 °C at 40 °C min− 1 and held at this maximum temperature for 2.3 min. FID temperature was 280 °C. Identification of geosmin and MIB peaks detected by GCFID was verified by parallel analysis of selected samples with a gas chromatograph coupled to a mass spectrometer detector, (GC–MS model Saturn 2000, Varian Inc. Palo Alto, CA.). 2.4. Laboratory incubation of geosmin and MIB-rich organic matter derived from the system Crude samples of organic matter were collected from the drum filter and trickling filter of the system. Tests were performed on triplicate samples. Ten grams of organic matter were placed in 500 mL Erlenmeyers containing 250 mL distilled water. The Erlenmeyers were
Fig. 1. A schematic presentation of the investigated RAS (not to scale). Solid arrows indicate the direction of the water flow and broken arrows the direction of water + sludge flow. The main water purification processes in the various compartments are indicated.
L. Guttman, J. van Rijn / Aquaculture 279 (2008) 85–91
incubated at 37 °C with vigorous shaking. Geosmin and MIB concentrations in water were measured during the next 27 h. Samples containing 10 g of presterilized organic matter were used as controls. The dry weight of the organic matter was determined in 5 replicates after drying at 105 °C and subsequent burning of the samples at 550 °C according to APHA (1995). 2.5. Isolation of streptomyces Streptomyces spp. isolation was achieved by using casein and starch agar used for enrichment of actinobacteria (Kuster and Williams, 1964). Distinctive streptomyces-like colonies were repeatedly transferred to selective agar plates containing nystatin, polymixin B sulfate, penicillin and cyclohexamide (Sigma-Aldrich, Israel). MIB and geosmin producing bacteria were selected from among the different bacterial colonies growing on the plates based on the typical odors associated with the off-flavor compounds they produced. 2.6. Molecular analyses of isolates Identification of the isolates was based on polymerase chain reaction amplifications of 16 S rRNA with the general bacterial primer 1392-R (Lane, 1991) and the GC-rich actinobacteria-specific primer A235-F (Stach et al., 2003). Primers S661-F and S1218-R, designed by Inbar et al. (2005), were used for further classification of the isolates to the genus level. Each 50 μL of PCR reaction mixture contained the following components: 1.5 units Taq DNA Polymerase (Red Taq; Sigma Chemical Co., St. Louis, MO), 5 μL Sigma Taq Buffer with a final magnesium concentration of 2 mM and 4 mM for actinobacteria and streptomyces primers, respectively, 0.2 mM PCR nucleotide mixture (Promega, Madison, WI), 25 pmol of each primer and 1 μL of DNA template. DNA was extracted by heating live colonies in 30 μL presterilized double distilled water at 95 °C for 10 min. All PCR reactions were conducted in a gradient thermal cycler (model PTC200; MJ Research, Watertown, MA) using the following protocol: initial denaturation at 95 °C for 3 min, followed by 35 (for 235-F/1392-R) or 30 cycles (for S661-F/S1218-R) of denaturation at 94 °C for 30 s, annealing at 62 °C for 30 s and elongation at 72 °C for 1 min. The final elongation step was 2 min at 72 °C. The presence and size of PCR fragments were determined by agarose gel electrophoresis (1.8%) with ethidium bromide staining. 2.7. Sequencing and phylogenetic analysis of isolates DNA was purified from the PCR products by using the UltraClean PCR Clean-up kit (MO BIO, Solana Beach, CA) according to the protocol described by the manufacturer. For bacterial identification, purified products of the PCR reactions were sequenced using the Applied Biosystems PRISM Dye Terminator Cycle Sequencing Ready Reaction kit with Ampli Taq DNA polymerase and the A235-F primer, described above. The sequencing products were analyzed with an Applied Biosystems 377 DNA sequencer. Nucleotide sequences (550 bp) were added to a pre-aligned database of a 16 S rRNA sequences using the aligning tool supplied by ARB Phylogenetic program package (Strunk et al., 2004). Phylogenetic trees were generated with the neighborjoining and maximum likelihood methods with the ARB program package using the Olsen correction method applying a 50% cut-off filter (Strunk and Ludwig, 1998). The topologies of the resulting trees were compared. Branching order was supported by both methods.
87
temperatures for a maximum of 168 h. Aerobic conditions were maintained by vigorous shaking of the Erlenmeyers. Subsamples (27 mL) were taken, every 3 h for the first 12 h, every 4 h for the next 12 h and then every 8 h until no further growth was observed. Samples were examined for MIB and geosmin and for bacterial growth. Bacterial counts were performed by the most probable number method. For this determination, aliquots of the cultures (200 μL) were transferred into the wells of sterile ELISA plates containing 1800 μL casein liquid media and further decimal dilutions were performed for calculation of the bacterial most probable number in the original sample. Mean values of bacterial counts performed during the exponential phase of growth of the isolates were used for calculation of the hourly growth rate constants of the isolates as described by Koch (1981). The aerobic and anoxic growth performance of the isolates was examined in the above described growth medium. Erlenmeyers were kept aerobic by vigorous shaking and anoxic conditions were secured by bubbling argon gas into the airtight Erlenmeyers for 20 min prior to incubation. Absorbance of the culture medium at 600 nm was used as an indicator for growth of the isolates. Geosmin and MIB were measured by the above described solid phase micro extraction method. 3. Results 3.1. System performance Data concerned with the growth performance of tilapia were only available for the period from February until June, 2003. During this period, the average tilapia size increased from 47 g to 326 g. Total production during this seven months period was 31.5 kg m− 3 and maximum fish density was 37 kg m− 3. The average daily specific growth rate was 1.4% day− 1. Fresh water addition was limited to compensate for evaporation losses only and averaged 3.6% of the total system volume per day. The water requirement for production of 1 kg of fish was 213 L. During this period, two organoleptic tests were performed by a certified analytical laboratory (Bactochem, Nes Ziona, Israel) on randomly selected fish collected from the system at 19.5.03 and 30.6.03. In both tests, fish were rated as highly off-flavored. 3.2. Water quality conditions during the culture period Monthly analyses of the main water quality parameters over the entire growth cycle (October, 2002 until September, 2003) in the fish culture basins (Fig. 2) revealed that TAN levels were generally lower than 1 mg TAN-N L− 1, nitrite levels were lower than 0.5 mg NO2-N L− 1,
2.8. In vitro assays with isolates Growth rate measurements at different temperatures were conducted by introduction of one colony of each of the Streptomyces spp. isolates into 1 L Erlenmeyers containing 250 mL casein liquid media (Kuster and Williams, 1964) followed by incubation at different
Fig. 2. Monthly examination of inorganic nitrogen, phosphorus and COD concentrations in the culture water of the RAS. Samples were derived from the combined effluent off all 12 fish tanks at a sampling point just before the discharge of this effluent water into the drum filter (see Fig. 1). Concentrations represent average values of duplicate analyses of two separate water samples.
88
L. Guttman, J. van Rijn / Aquaculture 279 (2008) 85–91
Fig. 3. Geosmin and MIB concentrations in: (A) the culture water of the RAS and (B) the different components of the RAS. Concentrations in the culture water (A) represent the mean value (and standard deviation) of triplicate water samples. Concentrations in the different components (B) represent the average monthly values of duplicate determinations conducted during each month of the examined period (n = 10).
phosphorus levels were around 10 mg PO4-P L− 1, nitrate levels fluctuated between 30 and 70 mg NO3-N L− 1, and organic matter (expressed as COD) fluctuated between 70 and 160 mg O2 L− 1 (Fig. 2). 3.3. Geosmin and MIB production in recirculating system. Geosmin and MIB could be detected at all sampling dates in the fish basin water with higher concentrations of MIB than geosmin. Maximum MIB concentrations were around 170 ng L− 1 compared to 75 ng L− 1 for geosmin (Fig. 3A). Geosmin and, in particular, MIB concentrations differed among the various compartments (Fig. 3B). Generally, concentrations of MIB were higher than those of geosmin in all compartments. Highest mean concentrations of 260 ng L− 1 for MIB and 47 ng L− 1 for geosmin were detected in the drum filter and at the outlet of the trickling filter, respectively. Geosmin and especially MIB were relatively low in the digestion basin and, based on the concentrations of geosmin and MIB in the inlet water to this basin (water from the drum filter), it was here where a reduction of these compounds took place. 3.4. In vitro production of geosmin and MIB by organic matter Differences between inlet and outlet geosmin and MIB concentrations clearly pointed to a production of both compounds in the aerobic trickling filter (Fig. 4A) and drum filter components (Fig. 4B).
Maximum production rates during laboratory incubation (expressed per gram of dry organic matter) of geosmin were 1.725 ng g− 1 h− 1 and 2.446 ng g− 1 h− 1 and maximum MIB production rates were 1.781 ng g− 1 h− 1 and 3.786 ng g− 1 h− 1 in sludge derived from the drum and trickling filters, respectively. Rates of MIB production were higher than those for geosmin in both samples. Production of geosmin and MIB was sustained for longer periods in sludge derived from the tricking filter. 3.5. Isolation and identification of geosmin and MIB-producing streptomyces Enrichment of sludge derived from the drum and trickling filters resulted in the isolation of two bacterial strains capable of geosmin and MIB production. Initial selection of these isolates from a large number of isolates was based on the distinctive earthy–musty smell of the agar plates harboring these isolates. Sequence analyses of 16 S rRNA products obtained with the general and actinobacterial primer pair 1392R/A235F followed by phylogenetic BLAST analysis resulted in the identification of two members of streptomyces. One isolate (Iso. 2A) revealed a 99% sequence identity to Streptomyces roseoflavus and the other (Iso. 4C) a 99% sequence identity to Streptomyces thermocarboxydus. Additional analysis, using the phylogenetic tree analysis ARB program, confirmed the identification of both isolates as members of the streptomyces genus (Fig. 5).
Fig. 4. Production of geosmin and MIB during laboratory incubation of organic matter derived from (A) the trickling filter and (B) the drum filter in distilled water. Concentrations are mean values of triplicate analyses. Bars represent the standard deviation.
L. Guttman, J. van Rijn / Aquaculture 279 (2008) 85–91
89
Fig. 5. Phylogenetic tree based on 16S rRNA gene sequences of the streptomycetes-like isolates Iso 2A and Iso 4C. The topology of the tree is based on the consensus of trees of N 1000 informative positions generated using the Neighbor Joining and Maximum Likelihood methods (ARB phylogenetic package). Scale bar = 1% estimated difference in nucleotide sequence position.
3.6. In vitro geosmin and MIB production by Streptomyces spp.
4. Discussion
Growth rates constants of the two species were conducted at different temperatures (Fig. 6). The optimal temperature for S. roseoflavus-like isolate was around 37 °C while it was higher for the S. thermocarboxydus-like isolate. However, due to a too narrow range of temperatures tested, we were unable to determine the optimal temperature for this latter organism. Incubation of the isolates resulted in considerable growth and geosmin and MIB production in water derived from the fish culture basin and trickling filter outlet (Fig. 7). Pre-sterilized control samples showed no offflavor compounds production and growth of the isolates. A comparison of Streptomyces spp. under aerobic and anoxic conditions revealed that both isolates produced geosmin and MIB during exponential phase of growth with MIB production exceeding geosmin production. Anoxic incubation resulted in similar growth rates for the S. roseoflavus-like isolate while lower growth rates were measured for S. thermocarboxydus-like isolate. MIB production rates under anoxic conditions were considerably lower than under aerobic conditions (Fig. 8). Growth of the S. roseoflavus-like isolate under anoxic conditions coincided with a nitrate decrease in the medium with only a slight increase in ammonia and nitrite (Fig. 9). The observed disappearance of inorganic nitrogen from the medium points to a denitrification capability of this organism.
Although off-flavor of fish in recirculating systems is often encountered, information on geosmin and MIB concentrations in these systems is scarce and to the best of our knowledge, this is the first study in which such information is provided. High concentrations of both compounds were found within all water treatment components during routine operation of the system. Unlike most of the previously investigated aquaculture farms (Schrader and Blevins, 1993; Robertson et al., 2005; Robin et al., 2006), MIB was shown to be the primary off-flavor compound in all parts of the recirculating system examined in this study. Off-flavor in fish culture systems is experienced at MIB concentrations as low as 18 ng L− 1 (Persson and York, 1978) for MIB and at geosmin concentrations as low as 15 ng L− 1 (Persson, 1980). Concentrations of both compounds in the system examined in this study were significantly higher than above threshold concentrations and were also higher than the few systems in which ambient MIB and geosmin concentrations were measured (Klausen et al., 2005; Robertson et al., 2006; Vallod et al., 2007). Streptomyces, present in organic-rich, aerobic parts of recirculating system, were found to be responsible for MIB and geosmin production, as shown by in vitro experiments. In accordance to the field observations, MIB production was higher than geosmin production upon incubation of both isolates in either culture medium or water derived from the
Fig. 6. The effect of temperature on the hourly growth rate constant (μ) of streptomycetes-like isolates Iso. 2A and Iso. 4C. Isolates were incubated in casein liquid media under aerobic conditions and growth rate constants were determined during exponential phase of growth.
Fig. 7. MIB and geosmin production by the two streptomycetes-like isolates in water derived from different sources within the RAS system. Incubation in water derived from the fish culture basin of Iso. 2A (A) and Iso. 4C (B) and incubation in water derived from the trickling filter of Iso. 2A (C) and Iso. 4C (D). Concentrations are mean values of triplicate analyses. Bars represent the standard deviation.
90
L. Guttman, J. van Rijn / Aquaculture 279 (2008) 85–91
Fig. 8. Cell density (determined as O.D.600) and production of MIB and geosmin by the two streptomycetes-like isolates under either aerobic or anaerobic conditions. Incubation under anoxic conditions of Iso. 2A (A) and Iso. 4C (B) and under aerobic conditions of Iso. 2A (C) and Iso. 4C (D). Concentrations are mean values of triplicate analyses. Bars represent the standard deviation.
recirculating system. MIB and geosmin production by streptomyces present in the organic-rich trickling filter and drum filter suggest a benthic nature of these organisms. The relatively high geosmin and MIB concentrations in the aerobic components of the recirculating aquaculture system and the higher in vitro production under aerobic conditions by both Streptomyces isolates present a strong evidence for the important role of oxygen in geosmin and MIB production by streptomyces as previously suggested by Dionigy and Ingram (1994). MIB and geosmin concentrations in the organic-rich digestion basin were relatively low. The lower production rates of these compounds under anaerobic conditions by the isolates might explain why the lowest MIB and geosmin concentrations were found in this compartment. Furthermore, removal of these compounds in this basin seems to take place as concentrations of MIB and geosmin within the basin were lower than in the inlet water (water derived from the drum filter). Preliminary results indicate that both absorption to the lipid
phases of organic compounds and biological degradation underlie the observed decrease of MIB and geosmin in this basin (Guttman and van Rijn, unpubl.). Finally, nitrate reduction without a concomitant increase in either ammonia or nitrite by the S. roseoflavus-like strain, provide circumstantial evidence for the denitrification ability of this isolate. Evidence for the denitrification capability of a number of actinomycetes was provided by Shoun et al. (1998) but to the best of our knowledge has not been determined in off-flavor producing streptomyces. The extent of geosmin and MIB production by the S. roseoflavus-like strain under denitrification conditions is currently under investigation. 5. Conclusions Based on in-situ analyses as well as studies with the bacterial isolates, it was concluded that aerobic, organic-rich conditions stimulate the growth of actinomycetes and subsequent production of geosmin and MIB. Since such conditions are intrinsic, not only in this particular RAS system but in most recirculating systems operated with nitrifying filters, it seems unlikely that alleviation of the problem by modification of the nitrification stage is feasible. Conversely, the observed reduction of these compounds in the anaerobic water treatment component might be considered an important finding toward the alleviation of the problem. Presently, studies are ongoing in which the processes responsible for geosmin and MIB reduction under anaerobic, organic-rich conditions are examined. Acknowledgments
Fig. 9. Inorganic nitrogen fluctuations during anoxic incubation of streptomyces-like Iso. 2A in liquid mineral salt medium in the presence of nitrate. Concentrations are mean values of triplicate analyses. Bars represent the standard deviation.
This research was supported by grant #820-0185-03 from the Chief Scientist Office of the Ministry of Agriculture & Rural Development, Israel. The authors would like to thank Mr. Tamir Ezer and the other staff members of the Ginosar Intensive Aquaculture Research Station for daily operation of the recirculating system.
L. Guttman, J. van Rijn / Aquaculture 279 (2008) 85–91
References APHA., 1995. Standard methods for the examination of water and waste water (19th ed.). American Public Health Association, Washington, DC. Dionigy, C.P., Ingram, D.A., 1994. Effects of temperature and oxygen concentration on geosmin production by Streptomyces tendae and Penicillium expansum. J. Agric. Food Chem. 42, 143–145. Howgate, P., 2004. Tainting of farmed fish by geosmin and 2-methyl-iso-borneol: a review of sensory aspects and of uptake/depuration. Aquaculture 234, 155–181. Inbar, E., Green, S.J., Hadar, Y., Minz, D., 2005. Competing factors of compost concentration and proximity to root affect the distribution of streptomycetes. Microbial Ecol. 50, 73–81. Klausen, C., Nicolaisen, M.H., Strobel, B.W., Warnecke, F., Nielsen, J.L., Jorgensen, N.O.G., 2005. Abundance of actinobacteria and production of geosmin and 2-methylisoborneol in Danish streams and fish ponds. FEMS Microbiol. Ecol. 52, 265–278. Koch, A.L., 1981. Growth measurement. In: Gerhardt, P., Costilow, R.N. (Eds.), Manual of Methods for General Bacteriology. ASM, Washington, DC, pp. 179–208. Kuster, E., Williams, S.T., 1964. Selection of media for isolation of streptomycetes. Nature 202, 928–929. Lane, D.J., 1991. 16S/23S rRNA sequencing. In: Stackebrandt, E., Goodfellow, M. (Eds.), Nucleic acid techniques in bacterial systematics. John Wiley & Sons, Chichester, UK., pp. 115–175. Persson, P.E., 1980. Sensory properties and analysis of two muddy odour compounds, geosmin and 2-methylisoborneol, in water and fish. Water Res. 14, 1113–1118. Persson, P.E., 1983. Off-flavors in aquatic ecosystem—an introduction. Water Sci. Technol. 15, 1–11. Persson, P.E., York, R.K., 1978. Notes on muddy odor: I. Sensory properties and analysis of 2-methylisoborneol in water and fish. Aqua Fennica 8, 83–88. Robertson, R.F., Jauncey, K., Beveridge, M.C.M., Lawton, L.A., 2005. Depuration rates and the sensory threshold concentration of geosmin responsible for earthy–musty taint in rainbow trout, Onchorhynchus mykiss. Aquaculture 245, 89–99. Robertson, R.F., Hammond, A., Jauncey, K., Beveridge, M.C.M., Lawton, L.A., 2006. An investigation into the occurrence of geosmin responsible for earthy–musty taints in UK farmed rainbow trout, Onchorhynchus mykiss. Aquaculture 259, 153–163. Robin, J., Cravedi, J.P., Hillenweck, A., Deshayes, C., Vallod, D., 2006. Off flavor characterization and origin in French trout farming. Aquaculture 260, 128–138. Scheiner, D., 1976. Determination of ammonia and Kjeldahl nitrogen by the indophenol method. Wat. Res. 10, 31–36. Schrader, K.K., Blevins, W.T., 1993. Geosmin-producing species of Streptomyces and Lyngbya from aquaculture ponds. Can. J. Microbiol. 39, 834–840. Schrader, K.K., Nanayakkara, N.P.D., Tucker, C.S., Rimando, A.M., Ganzera, M., Schaneberg, B.T., 2003. Novel derivatives of 9,10-anthraquinone are selective
91
algicides against the musty-odor cyanobacterium Oscillatoria perornata. Appl. Environ. Microbiol. 69, 5319–5327. Shnel, N., Barak, Y., Ezer, T., Dafni, Z., van Rijn, J., 2002. Design and performance of a zerodischarge tilapia recirculating system. Aquacult. Eng. 26, 191–203. Shoun, H., Kano, M., Baba, I., Takaya, N., Matsuo, M., 1998. Denitrification by actinomycetes and purification of dissimilatory nitrite reductase and azurin from Streptomyces thioluteus. J. Bacteriol. 180, 4413–4415. Silvey, J.K.G., Roach, A.W., 1953. Actinomycetes in the Oklahoma City water supply. J. Amer. Water Works Assoc. 64, 35–39. Stach, J.E., Maldonado, L.A., Ward, A.C., Goodfellow, M., Bull, A.T., 2003. New primers for the class Actinobacteria: application to marine and terrestrial environments. Environ. Microbiol. 5, 828–841. Strickland, J.D., Parsons, T.R.A., 1968. A practical handbook of seawater analysis. Fish. Res. Board Canada 167, 77–80. Strunk, O., Ludwig, W., 1998. ARB: A software environment for sequence data. Tech. Univ., Munich. Germany. Strunk, O., Ludwig, W., Westram, R., Richter, L., Meier, H., Yadhukumar, Buchner, A., Lai, T., Steppi, S., Jobb, G., Forster, W., Brettske, I., Gerber, S., Ginhart, A.W., Gross, O., Grumann, S., Hermann, S., Jost, R., Konig, A., Liss, T., Lussmann, R., May, M., Nonhoff, B., Reichel, B., Strehlow, R., Stamatakis, A., Stuckmann, N., Vilbig, A., Lenke, M., Ludwig, T., Bode, A., Schleifer, K.H., 2004. ARB: A software environment for sequence data. Nucleic Acids Res. 32, 1363–1371. Terashima, K., 1988. Reduction of musty odor substances in drinking water — a pilot plant study. Wat. Sci. Technol. 20, 275–281. Tucker, C.S., 2000. Off-flavor problems in aquaculture. Rev. Fish. Sci. 8, 45–88. Tucker, C.S., van der Ploeg, M., 1999. Managing off-flavor problems in pond-raised catfish. Southern Regional Aquaculture Center publication, 192. Vallod, D., Cravedy, J.P., Hillenweck, A., Robin, J., 2007. Analysis of the off-flavor risk in carp production in ponds in Dombes and Forez (France). Aquacult. Int. 15, 287–298. Watson, S.B., 2004. Aquatic taste and odor: a primary signal of drinking-water integrity. J. Toxicol. Environ. 67, 1779–1795. Wood, S., Williams, S.T., White, W.R., 2001. Microbes as a source of earthy flavours in potable water — a review. Int. Biodeterior. Biodegrad. 48, 26–40. Yagi, S., 1983. Taste and flavor of water. J. Japan Water Works Assoc. 590, 65–79. Zimba, P.V., Grimm, C.C., Dionigi, C.P., Weirich, C.R., 2001. Phytoplankton community structure, biomass, and off-flavor: pond size relationships in Louisiana catfish ponds. J. World Aquacult. Soc. 32, 96–104.