Impact of milk fish farming in the tropics on potentially pathogenic vibrios

Marine Pollution Bulletin 77 (2013) 325–332

Contents lists available at ScienceDirect

Marine Pollution Bulletin journal homepage: www.elsevier.com/locate/marpolbul

Impact of milk fish farming in the tropics on potentially pathogenic vibrios W.T. Reichardt ⇑, J.M. Reyes, M.J. Pueblos, A.O. Lluisma Marine Science Institute, University of the Philippines, Diliman, 1101 Quezon City, Philippines

a r t i c l e

i n f o

Keywords: Sucrose-negative vibrios Enterovibrionaceae Health risk Milkfish mariculture Philippines

a b s t r a c t Ratios of sucrose-negative to sucrose-positive vibrios on TCBS agar (suc/suc+) indicate the abundance of potential human pathogenic non-cholera vibrios in coastal mariculture environments of the Lingayen Gulf (Philippines. In guts of adult maricultured milkfish (Chanos chanos) of suc vibrios reached extreme peak values ranging between 2 and 545 million per g wet weight. Suc vibrios outnumbered suc+ vibrios in anoxic sediments, too, and were rarely predominant in coastal waters or in oxidized sediments. Suc/suc+ ratios in sediments increased toward the mariculture areas with distance from the open sea at decreasing redox potentials. There is circumstantial evidence that suc vibrios can be dispersed from mariculture areas to adjacent environments including coral reefs. An immediate human health risk by pathogenic Vibrio species is discounted, since milkfish guts contained mainly members of the Enterovibrio group. A representative isolate of these contained proteolytic and other virulence factors, but no genes encoding toxins characteristic of clinical Vibrio species. Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction Marine coastal pollution caused by intensified milk fish farming in the Lingayen Gulf (South China Sea, Philippines) shows chemical and biochemical effects in water and sediment (Holmer et al. 2003; Reichardt et al. 2011; Nacorda et al. 2012). Repeated fish kills, diseases of maricultured invertebrates, a cholera epidemic in 2005, and enhanced mortality of reef-building corals raise concerns about sustainability of intensive milk fish mariculture in relation to environmental health (Villanueva et al. 2006; Reichardt et al. 2007). In terms of bacterial health risks, vibrios with at least a dozen pathogenic Vibrio species play a dominant role in coastal marine environments used for mariculture (Tantillo et al. 2004; Das et al. 2009, Senderovitch et al., 2010). As copiotrophic heterotrophs, marine vibrios are frequently associated with maricultured organisms including fish intestines (De Paola et al., 1994, Beneduce et al. 2010; Ganesh et al. 2010; Matsunaga et al. 2011). Diversified use of coastal waters at Cape Bolinao (Lingayen Gulf, Philippines) for mariculture, restoration of marine wildlife, and tourism offers a promising basis for assessing the impact of tropical mariculture on potential health risks. Whereas numerous marine vibrios have been classified as pathogens affecting marine animals as well as humans (Farmer and Hickman-Brenner 2006; Austin 2010), only a small fraction of these pathogenic species may be virulent in their marine environment (Oberbeckmann et al. 2011; Bier et al., ⇑ Corresponding author. Tel.: +63 2 922 3959; fax: +63 2 924 7678. E-mail address: [email protected] (W.T. Reichardt). 0025-326X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.marpolbul.2013.09.018

2013). Yet, this assumption can be biased, as currently available information stems from non-tropical marine environments. Virulent pathogenic Vibrio species may be expected more frequently in tropical marine environments, since the expression of virulence genes seems to increase at elevated environmental temperatures (Mahony et al., 2010). Hence tropical coastal waters with yearround temperatures near 30 °C would ensure most favorable growth conditions for these pathogens. Initial steps to examine this bacteriological health risk relied on selective viable counts of (presumptive) vibrios on TCBS agar (Bolinches et al., 1988). This highly selective medium allows distinction between sucrose positive and sucrose negative phenotypic subgroups of vibrios that comprize valid Vibrio species as well as reclassified former Vibrio species (Farmer and Hickman-Brenner 2006). Since sucrose negative phenotypes harbor the bulk of potential human pathogens – ratios of sucrose negative to sucrose positive vibrios (suc/suc+) are considered as useful health risk indicator in sea food microbiology (Lopez-Joven et al. 2011). Human pathogenic vibrios such as Vibrio vulnificus and Vibrio parahaemolyticus can occur in the intestine of certain finfish (de Paola et al. 1994; Das et al. 2009). But our information about selective forces governing both the enrichment in fish guts and dispersal of milkfish-borne vibrios is scarce (Matsunaga et al. 2011). Routine bacteriological surveys of fish farming sites leave only limited scope for diagnostic analyses at the species level. Therefore this investigation examines suc/suc+ ratios as possible candidate for a practicable indicator of the likely presence of non-cholera human pathogenic vibrios that would be suited for low cost and

326

W.T. Reichardt et al. / Marine Pollution Bulletin 77 (2013) 325–332

routine environmental monitoring. At the same time it addresses as yet unresolved environmental health risks associated with certain bacterial loads that accompany the dispersal of fish farming waste in adjacent marine environments. Bacteriological monitoring data in 2007 had suggested a temporary spillover of sucrose-negative water-borne vibrios from a fish cage site into the waters of a nearby coral reef. Subsequent analyses of intestinal contents of milk fish, water and sediment samples in 2011/12 were to provide suc/suc+ ratios for vibrios on a larger scale during NE monsoonal dry season. This choice minimized the possibility of interfering selective salinity effects on Vibrio populations during the SW monsoonal wet season and covered both peak harvest and near fallow periods for milk fish. Intestinal contents were obtained exclusively from adult milk fish, because only these had indicated a predominance of sucrose-negative vibrios in previous analyses (Reichardt et al. (2007).

2. Materials and methods Mariculture zones at Cape Bolinao are located in close proximity to coral reefs and areas devoted to tourism. Milkfish farming using 18  18 m cages in the Anda–Bolinao sections of the Lingayen Gulf has been restricted to a southern (Anda) and a northern fish farming zone (Bolinao) along the Caquiputan Strait that diverges at Siapar Island (Fig. 1). Seasonal distribution patterns of water borne viable bacterial counts were obtained at monthly intervals in 2006/2007 from a fish cage site at Siapar (N16°20.90 , E119°57.50 (11 m water depth) and from a neighboring shallow coral reef site at Cangaluyan Island (N21°58.90 , E119° 58.30 (1–2 m w.d.). A survey of surface water and sediment extended over a distance of more than 10 km from the northern fish farming zone of the Caquiputan Strait toward the mouth of the main tidal channel. This survey was conducted on eight sampling dates at intervals of roughly three weeks during NE monsoonal seasons in 2011/2012. It included the peak harvest season in December followed by a subsequent nearly fallow period in January. Water samples were taken (A) at the main tidal channel exit near Lucero, N16°24.6060 , E 119°53.9160 at 18 m water depth (w.d.), (B) at a coral reef site at Malilnep (Fig. 1,# 1), N16°26.3200 , E119°56.4580 (5 m w.d.), (C) near inhabited shoreline at Bolinao (Fig. 1,# 9), N16°22.8400 , E119° 54.6260 (21 m w.d.), (D) in the fish cage area of Bolinao (Fig. 1, #10), N16°23.1600 , E119°55.4800 (15 m w.d.), (E) at Poro Panaien islet (Fig. 1, # 11), N16°22.2470 , E119°, 55.7750 (6 m w.d.), and (F) near Siapar island (Fig. 1, # 12), N16°21.0710 , E119°57.6310 (7 m w.d.).  On each sampling date single specimens of adult milk fish (Chanos chanos) were obtained from the fish cage site of Bolinao (Fig. 1, # 10). Complimentary sediment samples were obtained from the same locations as the water samples, except for replacing the coral reef site at Malilnep with fine sandy sediment from Gawa (Fig. 1, #4), (N 16°23.5400 , E 119°54.5390 ,13 m water depth). In March 2012 additional sediment samples were obtained from a total of 17 locations extending from the southern fish cage areas of Anda to Lucero near the open sea. These sediment sampling sites are indicated by numbers in Fig. 1. Triplicate water samples from 1 m depth were collected using a Niskin sampler and were immediately filled into sterile 50 ml Corning tubes. Triplicate sediment subsamples were obtained using plexiglass tubes (inner diameter: 5 cm) either by scuba divers or with a gravity corer. The top 2 cm layers of sediment were subsampled using sawed-off 5 cm3 sterile syringes from which 0.5 ml aliquots were distributed into serial dilution tubes containing sterile sea water. A special sediment survey in 2012 followed a redox potential gradient stretching from anoxic and sulfidic

sediments in the fish farming zones of Anda and Bolinao toward oxidized sediments northwest of Santiago Island. Redox potential readings were taken with a Pt redox electrode (WTW) in combination with an Ag/AgCl electrode serving as reference at 194 mV. Freshly caught milkfish were aseptically dissected within 1–2 h. Weighed contents of the distal end of the intestine were distributed into 50 ml Corning tubes containing 40 ml of sterile seawater for serial dilutions and dispersed three times for 5 s using a high speed ultraturrax blender, before serial dilutions were completed. For comparison with a filter feeding organism as alternative, the same procedure was applied to the aseptically dissected soft tissue of freshly caught bivalve ‘‘tahong’’ (Perna viridis). Samples of this bivalve were collected at the same time from immersed structures of the same fish cages that provided the milkfish specimens. From serial dilutions of water, sediment, and animal samples targeting roughly estimated titer ranges, 0.1 ml aliquots were distributed and spread plated onto petridishes. Means of viable counts of presumptive vibrios (mesophilic Vibrionaceae and other closely related vibrios) were based on triplicate plate counts after 1–2 d of incubation at 37–38 °C on selective thiosulfate-citrate-bile salts sucrose (TCBS) agar (Difco; Bolinches et al. 1988). Sucrose fermenting and sucrose non fermenting colonies (yellow and green in the presence of bromothymolblue) were counted separately. To mark the predominance of sucrose negative (sucrose non fermenting) colony forming units (CFUs), ratios of sucrose-negative to sucrose positive CFUs (suc/suc+) were recorded according to LopezJoven et al. (2011). For comparison, plate counts of total copiotrophic bacteria were obtained on ZoBell’s marine agar 2216 (HiMedia) after incubation for two weeks at 30 °C using 3–4 replicates. Sucrose-negative isolates from TCBS agar were grown on marine agar plates and maintained in alkaline peptone ‘‘CDC 1494’’ liquid medium (Farmer and Hickman-Brenner 2006). Phenotypic characterization focused mainly on tests recommended by Noguerola and Blanch (2008) using an API 20E kit (Biomerieux, France). A representative isolate from milkfish intestine (V6) was chosen for whole genome sequencing using the Roche 454 GS Junior System. The protocols and kits recommended by the manufacturer were used for the preparation of the sequencing libraries. The sequences were then assembled using the software Newbler (ver. 2.7). The resulting contigs were searched for sequences encoding putative virulence factors, including the known toxin genes of clinically pathogenic vibrios (Chen et al., 2005). In addition, phylogenetic analysis of selected markers (16S rRNA, rpoA, pyrH; Thompson et al. 2005a) obtained from the genomic data was also carried out using MEGA5 (Tamura et al. 2011) to provide the basis for the taxonomic classification of the isolate. Statistics were based on GraphPad InStat 3, version 3.06 for Windows (Graph Pad Software, Inc.).

3. Results 3.1. Bacterial load of vibrios in maricultured organisms Intestinal contents of freshly collected adult milkfish from the Bolinao fish farming zone showed strong anoxia with redox values ranging between -150 and 290 mV. Viable counts of vibrios yielded almost exclusively sucrose negative CFUs. Ratios of sucrose non-fermenting to sucrose fermenting CFUs (suc/suc+) exceeded by far the equilibrium ratio of one. Milk fish specimens analyzed on seven sampling dates between January 2011 and January 2012 contained from 2.0  106 to 545  106 CFUs per g of wet weight (median: 52.3  106) on TCBS agar (Table 1). Suc/suc+ ratios ranging between 2.1 and 59, with a mean of 22.4, indicated an extreme

327

W.T. Reichardt et al. / Marine Pollution Bulletin 77 (2013) 325–332

Fig. 1. Map of sampling area along main tidal channel of Caquiputan Strait and coral reefs at Malilnep and Cangaluyan, northern (Bolinao) fish farming zone (# 6 and 11), and southern (Anda) fish farming zone (# 13 and 17). Numbers 1–17 indicate sediment sampling sites at different redox potential values in 2012. Sediment sites # 3, 1, 9, 10, 12, and 14, are identical with sites A, B, C, D, E, and F, respectively. – Insert: location of Cape Bolinao on the Philippine island of Luzon.

prevalence of sucrose-negative vibrios. In contrast, samples of the maricultured bivalve Perna viridis collected on the same sampling dates contained only between 0.1 and 31.3  106 CfUs per g wet weight. Both mean and median (suc/suc+) ratios of 0.8 and 0.4 obtained for the bivalve indicated a prevalence of sucrose positive vibrios. This was significantly different from the mean (22.4) and median (11.2) (suc/suc+) ratios obtained for the milkfish intestine (Table 1, Mann-Whitney test, two-tailed P value = 0.0012).

When extrapolated to a ten years (2002-2011) average of 493 fish cages in the Bolinao fish farming zone, the approximate total cage area of 159732 m2 with an average stocking density of 50 000 specimens per cage contained 7.99 109 fish. With a median of 52.3  106 CFUs per fish (Table 1) and based on minimum intestinal contents of 1 g (wet weight) per fish, the standing stock of intestinal milkfish-borne vibrios in the fish farming zone can be estimated at 418  1015 CFUs.

Table 1 Total viable counts on TCBS agar and prevalence of sucrose-non-fermenting vibrio CFUs expressed as ratio suc/suc+ in milkfish intestine as compared with mussel (Perna viridis) from Bolinao mariculture zone for n sampling dates between January 2011 and January 2012. Sample

(106) Total CFU/g

Ratio (suc/suc+)

Range

Median

Mean ± SE

Range

Median

Mean ± SE

Milkfish intestine (n = 7) Mussel (n = 7)

2.0–545 0.1–31.3

52.3 0.9

113 ± 79.8 10.6 ± 5.4

2.1–59.0 0.1–3.7

11.2 0.4

22.4 ± 9.7 0.8 ± 0.5

328

W.T. Reichardt et al. / Marine Pollution Bulletin 77 (2013) 325–332 Table 2 Comparison of selected phenotypic characteristics of majority of 8 isolates (V6 phenotype) out of a total of 16 oxidase positive, sucrose-negative isolates on TCBS agar from adult milkfish intestines collected in Bolinao mariculture zone from January to March 2011. Characteristic

Majority Group (V6 phenotype)

Vibrio (tasmaniensisa

Enterovibrionaceaeb

Growth at 38 °C Arginine dihydrolase Lysine decarboxylase Ornithine decarboxylase Indol production Acetoin production Gelatinase activity Mannitol utilization b-Galactosidase activity

+    +    +

    + +  + 

 v   v   v +

Reference taxa: a Thompson et al. (2003a, 2003b, 2004). b Thompson et al. (2004,2005b), Pascual et al. (2009).

Adult milkfish collected in January and March 2011 yielded 3.4  106 and 52.3  106 CFUs on TCBS agar at suc/suc+ ratios of 14.4 and 2.1. A total of sixteen sucrose-negative colonies were isolated from TCBS agar and purified for taxonomic analyses using a phenotypic diagnostic key (Noguerola and Blanch 2008). This rapid identification scheme classified eight of sixteen isolates as Vibrio tasmaniensis; but their ability to grow in 8% NaCl and at 40 °C, and their inability to utilize glucose, mannitol, and amygdalin, and to produce acetoin (Voges-Proskauer test) and bgalactosidase (ONPG test) deviated from the phenotypic description of that species (Thompson et al., 2003a; Thompson et al., 2003b). Using the same diagnostic key, the other less frequent phenotypes resembled Vibrio splendidus, Vibrio gigantis, Vibrio mimicus and Vibrio mediterranei. Isolate V6 considered as median of the dominating

presumptive V. tasmaniensis phenotype was subjected to whole genome sequencing. Preliminary data based on whole-genome BLAST and multi-locus phylogenetic analysis using 16S, rpoA, and pyrH as markers) indicate, however, that isolate V6 is most closely related to Enterovibrio calvensis, E. norvegicus, and Grimontia (previously: Vibrio) hollisae (Enterovibrionaceae, Table 2). The molecular markers applied were, however, not informative enough to permit classification of the isolate to the species level.

3.2. Water analyses Near the margin of the southern fish farming zone (Siapar Island) and at a shallow coral reef site (Cangaluyan) situated approximately 2 km apart from each other, monthly water samples (1 m water depth) were analyzed from November 2006 to December 2007. Viable counts indicated similar trends for vibrios and total heterotrophs (Fig. 2). Sucrose negative vibrios at the coral reef remained near their detection limit (as indicated by suc/suc+ ratios far below one), with the remarkable exception of a single peak in September that occurred at both sampling sites. At the fish farming site of Siapar predominance of sucrose-negative vibrios (suc/suc+ ratios > 1) began shortly after a fish kill event (documented on June 15, 2007) and lasted until October. That fish kill event coincided further with a peak value of 8% for the calculated relative abundance of total vibrios as percent of total copiotrophs in June 2007. Water analyses in the northern fish farming zone (Bolinao) and its neighboring environments were conducted during the NE monsoonal season from January 2011 to March 2012. Predominance of

Table 3a Viable counts of vibrios on TCBS agar [CFU/ml] with pertaining suc-/suc+ ratios in water samples from 1 m depth on TCBS in 2011/2012. Sampling site

Lucero

Malilnep

Bolinao

Fish cage

Poro panaien

Siapar

No. of sampling dates

8

8

8

8

8

8

150 (45) 154 32– 395

912 (797) 192 67–6127

1513 (1405) 180 123– 10 710

332 (154) 167 43– 1237

579 (363) 210 17–2870

198 (78) 114 17– 540

0.63 (0.22) 0.66 0.02– 1.67

0.21 (0.07) 0.29 0.01– 0.57

0.72 (0.20) 0.69 0.15– 1.70

1.72 (1.21) 0.43 0.15– 9.57

1.04 (0.56) 0.88 0.13– 4.33

0.86 (0.17) 0.81 0.39– 1.54

TCBS CFU Mean (SE) Median Range

Fig. 2. Annual cycle (2006/2007) of mean viable counts (n = 3) of vibrios on TCBS agar and of total copiotrophic heterotrophs on ZoBell’s marine agar in surface waters of mariculture zone (Siapar) and at nearby coral reef (Cangaluyan). Suc/ suc+ ratios of vibrio counts are shown as open triangles. Asterisk indicates widely spread fish kill event.

Suc/suc+ Mean (SE) Median Range

W.T. Reichardt et al. / Marine Pollution Bulletin 77 (2013) 325–332 Table 3b Viable counts of vibrios on TCBS agar [CFU cm-3]  103 with pertaining suc-/suc+ ratios in surface sediment samples in 2011/2012. Significant differences (P = 0.0004, Dunns multiple comparisons test) between certain sites are indicated in brackets. Sampling site

A. Lucero

B. Gawa

C. Bolinao

D. Fish Cage

E. Poro Panaien

D. Siapar

No. of sampling dates TCBS CFU Mean (SE) Median Range

7

6

8

8

4

4

42.1 9.4 35.6 16.2– 261.7

94.3 83.2 9.4 2.3– 471.2

16.8 3.5 13.8 13.2– 433.4

23.2 10.4 13.4 0.5– 353.4

11.5 6.3 10.8 1.2–23.3

14.7 7.6 12.1 2.5– 32.2

Suc/suc+ Mean (SE) Median Range

(C)(D)(F) 0.81 (0.03) 0.04 0.03– 0.20

0.31 (0.10) 0.45 0.03– 0.50

(A) 1.27 (0.41) 1.2 0.07– 8.20

(A) 1.46 (0.33) 1.3 0.5– 3.40

1.38 (0.82) 0.85 0.40–3.50

(A) 1.84 (0.48) 2.10 0.70– 2.50

sucrose negative vibrios (in 11 of a total of 48 samplings) was mainly confined to fish cage sites and their vicinity (Bolinao, Poro Panaien, Siapar), but not completely absent from Lucero waters at the exit of the main tidal channel. Due to high variances between sampling dates no significant differences between water sampling sites outside (Lucero, Malilnep) and inside fish farming areas (remaining sites) could be established (Table 3a, non-parametric ANOVA, Kruskal-Wallis test). 3.3. Sediment analyses During NE monsoonal season sediment and water analyses were conducted on the same sampling dates. In sediment samples

Fig. 3. Suc/suc+ ratios of vibrios based on means of triplicate subsamples in sediment on 14 March 2012 at (A) incremental distances from reference site near open sea on; and (B) measured redox potential.

329

sucrose-negative vibrios prevailed in the fish farming area and its immediate vicinity at Bolinao, Poro Panaien and near Siapar (Table 3b). At greater distance from the fish farming zone, however, oxidized sediments showed an overwhelming predominance of sucrose positive vibrios nearest to the open sea (Lucero sediment). This predominance (suc/suc+ < 1) was weaker in sediment at Gawa, a sampling site closer toward the fish farming area that still showed redox potential readings in the positive range. Significant differences (P = 0.0004) were noted for suc/suc+ ratios between oxidized sandy sediments at Lucero and strongly reduced sediments from the fish farming area (fish cage and locations at Bolinao and Siapar – (Table 3b, Dunn’s multiple comparisons test). When sediment analyses were further extended into the southern fish farming area of Anda in 2012, the resulting topographic gradient revealed an even stronger predominance of sucrose negative vibrios with increasing distance from a reference site close to the open sea; for suc/suc+ ratios were significantly correlated with the distance of sampling sites from the open sea (r = 0.89, two tailed P < 0.0001, Spearman Rank correlation, Fig. 3a). Furthermore, suc/suc+ ratios were also negatively correlated with the redox potential values of the sediment sampled (r = 0.56, twotailed P = 0.0184, Spearman Rank correlation). This correlation was significant, but only moderately as compared with the strong distance-dependent decline of suc/suc+ ratios toward the open sea. In the northern fish farming area at Bolinao, frequently sulfidic sediments contained, despite their extremely low redox potential values, less sucrose negative vibrios as compared with sediments from locations near Anda at a greater distance from the open sea (Fig. 3b).

4. Discussion Coastal marine environments that serve as resource base for mariculture can cause human health hazards by favoring toxigenic Vibrio species (Rivera et al. 1989; Barbieri et al. 1999; Beneduce et al. 2010). As these human pathogenic vibrios are mesophiles with growth temperature optima around 37 C, maximum human health risks ought to be expected in permanently warm tropical environments. Yet, currently published health hazards arising from non-cholera mesophilic vibrios have raised particular awareness of global warming in non-tropical climates (Vezzulli et al. 2012). This provokes the question whether human health hazards caused by vibrios go usually unreported in the tropics, or whether they are considered here as less alarming. This could be either due to mitigating environmental factors or because of better immunization of local human populations in combination with lowered awareness. Growth and virulence of mesophilic vibrios can rise with increasing temperature even beyond 30 °C (Farmer and Hickman-Brenner 2006; Mahony et al., 2010). On the other hand, a temperature of 30 °C has already been considered as an upper threshold for detecting maximum abundances of classic clinical pathogens (Tantillo et al. 2004). Intestines of marine finfish can harbor 105 to 106 CFU/g of Vibrio vulnificus (De Paola et al. 1994; Fukushima and Seki 2004). Also toxigenic Vibrio parahaemolyticus and even Vibrio cholerae are known to reside in finfish guts (Das et al. 2009, Senderovitch et al., 2010). Vibrio vulnificus (biotype 2) serovars can infect both fish and humans (Fouz et al. 2010). With maximum viable counts approaching half a billion CFU/g (Table 1) intestines of adult milkfish at Bolinao contain a remarkably high load of presumptive vibrios representing almost exclusively sucrose negative vibrios. The result that filter feeding organisms (bivalves) from the same sampling location as the milkfish show an opposite prevalence of sucrose- positive vibrios, suggests the existence of different modes of selective enrichment in milk fish intestine and Perna viridis

330

W.T. Reichardt et al. / Marine Pollution Bulletin 77 (2013) 325–332

mussels. Strongly anoxic enrichment conditions that prevail in milk fish intestines seem to favor different physiological groups of vibrios than the entrails of filter feeding mussels with their permanent exposure to oxygenated water from the same fish cage. The subgroup of sucrose negative vibrios comprizes toxigenic pathogens such as Vibrio vulnificus, Vibrio parahaemolyticus, Vibrio (now: Grimontia) hollisae, Vibrio damselae (now:Photobacterium damselae) and Vibrio mimicus (Farmer and Hickman-Brenner 2006). This subgroup has become a practical target in depuration techniques for shellfish in sea food microbiology where elevated ratios of sucrose negative to sucrose positive CFUs on TCBS agar serve as health risk indicators of toxigenic vibrios (Lopez-Joven et al. 2011). Its members are also associated with vibrioses in maricultured animals including finfish (Paillard 2004; Das et al. 2009; Fouz et al., 2010; Austin 2010). Phenotype based numerical taxonomy has identified a mesophilic cluster of sucrose negative coastal Vibrio species with elevated growth temperature optima (>38 °C) and moderate NaCl requirement (Simidu and Tsukamoto 1985). While most human pathogenic non-cholera Vibrio spp. are sucrose negative (Farmer and Hickman-Brenner 2006), confirmation of pathogenicity of environmental isolates may not always be achieved by conventional phenotypic diagnostics and can even go beyond proper taxonomic classification. Classification as pathogenic Vibrio species using both classic phenotyping as well as genotyping needs to be supplemented by the identification of toxin genes and additional virulence factors as these are characteristic of clinical pathogenic strains. Current taxonomic revisions of vibrios do not only suffer from mismatches between classic phenotypes and 16S r DNA-based genotypes, but also from genotypic differences that require multilocus genetic analyses (Thompson et al. 2005a). Apparent mismatches between rapid phenotypic identification schemes based on dichotomous keys (Noguerola and Blanch, 2008) and genotyping seem also to apply to our isolates of sucrose negative vibrios from milkfish intestine. Phenotype-based keys have led to members of the Vibrio splendidus group dominated by Vibrio tasmaniensis. But this classification does not match with whole genome analysis of isolate V6 (representing half of the total) that leads to Enterovibrionaceae. Since there is considerable sequence divergence between Enterovibrionaceae and the Vibrio splendidus group including Vibrio tasmaniensis (Thompson et al. 2004), isolate V6 has been classified as a member of the Enterovibrio group (Enterovibrionaceae). Furthermore the key-based phenotypic identification proved to be in disagreement with the complete phenotypic description of Vibrio tasmaniensis (2004). In contrast to Vibrio tasmaniensis the phenotypic majority of milk fish isolates led by isolate V6 does neither produce acetoin nor use mannitol, but possesses b-galactosidase activity and can grow above 35 °C (Table 2). The latter preference for elevated growth temperatures is absent from both Vibrio tasmaniensis and Enterovibrionaceae. With several phenotypic key characteristics reported as variable for the novel family of Enterovibrionaceae, this is the final taxonomic classification we are currently able to reach for a representative of the majority group of isolates from milkfish intestine. Universal virulence markers such as haemolysin genes are often absent from environmental isolates of pathogenic Vibrio species as confirmed for Vibrio parahaemolyticus in temperate or cold marine waters (Oberbeckmann et al. 2011). Expression of such virulence genes seems to increase with elevated temperatures and salinities (Mahoney et al. 2010). Hence such virulence factors of pathogenic Vibrio spp. could be assumed to occur at higher frequencies in coastal marine environments of the tropics. Yet no known toxin genes related to clinically pathogenic vibrios were detected in the analyzed isolate. However, other classes of virulence factors were found in the genome of isolate V6. These included metallo-

proteinases, iron sequestration and transport components, and possibly antibiotic resistance proteins (JR, unpubl.). These have been previously documented to contribute to the pathogenicity of non-toxigenic Vibrio vulnificus (Jones and Oliver 2009), Vibrio parahaemolyticus (Goshima et al. 1978) and Vibrio mimicus (Hasan et al. 2010). Among Enterovibrionaceae Grimontia (previously: Vibrio) hollisae has been reported to be causative of gastroenteritis and sepsis in humans as well (Edouard et al. 2009; Hinestrosa et al. 2007). It is also noteworthy that certain bacterial functions encoded in ‘‘genomic islands’’ can be related, though not restricted, to pathogenicity. As these genomic islands may be transferred en bloc into recipient genomes, horizontal gene transfer in the rather selective microenvironments of milk fish guts can be most effective (Thompson et al. 2004). Equally low redox potential values in both the intestine of adult milkfish and in the sediments underneath fish cages (<-200 mV) would eventually provide potent sites for subsequent enrichment and dispersal of members of the Enterovibrio group. More in-depth analyses will be required to fully assess potential, as yet unknown health risks associated with this group. Predominance of sucrose-negative vibrios in surface layers of sediment in the vicinity of fish cages (Table 3, Fig. 3) matched with the extreme prevalence of this phenotypic subgroup in the intestine of adult milkfish (Table 1). Selective enrichment of sucrose-negative vibrios in anaerobic milkfish guts (as opposed to enrichment of sucrose-positive vibrios in ‘‘better aerated’’ bivalves) would plausibly explain that prevalence in sediments of the fish farming area. Fecal pellets leaving the fish cages via sedimentation are deposited in an anoxic environment with redox characteristics quite similar to those prevailing in milk fish guts. Hence presumably selective anoxic conditions that favor enrichment could be carried over from intestinal to sea floor environments. This most likely explanation has yet to be confirmed by specific taxonomic analyses in sediments. Prevalence of sucrose-negative vibrios in fish farming areas proved significant for sediments only (Table 3b). As a result of extreme variances conferred upon water analyses under a tidal regime and depending on changing weather conditions, benthic records prove more stable and better buffered against short-term fluctuations and interferences by drifting ‘‘vectors’’ of vibrios such as sea wrack in the water column. Abundance of sucrose-negative vibrios in sediment correlates with redox potential values (Fig. 3b) and even more strongly so with distance from the open sea (Fig. 3a). Extremely high densities in the southern fish farming zone of Anda suggest an involvement of further factors other than low redox potentials that would favor selective enrichment. Previous water analyses in the southern part of Caquiputan Strait indicate a drastic decrease of salinities during SW monsoonal (wet) season in 2004 when high densities of sucrose-negative vibrios coincided with minor fish kills (Reichardt et al. 2007). It is possible that temporarily enhanced freshwater inflow creating brackish water conditions in this area constitutes another selective advantage for sucrose-negative coastal vibrios such as ‘‘cluster S 2’’ type mesophiles (Simidu and Tsukamoto 1985) that may be considered as main target of viable counts on TCBS agar at 37–38 °C. Usually observed absence of sucrose-negative vibrios from water samples outside fish farming areas could be either due to severely limited resuspension from sediment surfaces, or to adverse conditions for growth in the presence of oxygen. The latter appears more likely; for sucrose-negative vibrios are almost absent from permanently oxic sediments at Lucero where sucrose-positive vibrios abound (Table 3b). Despite adverse conditions for growth in the water column, sucrose-negative vibrios are likely to be transported by tidal currents. This would at least explain the incidence of a single peak of water-borne sucrose-negative vibrios near a fish farming area at Siapar that had been mirrored at a nearby coral reef (Fig. 2). This apparent culmination of suc/suc+

W.T. Reichardt et al. / Marine Pollution Bulletin 77 (2013) 325–332

ratios had been preceded by a fish kill event at Siapar that occurred at peaking relative abundance of vibrios (among heterotrophic bacteria counted on marine agar). Determining the predominance of mesophilic sucrose-negative vibrios in mariculture environments is a practical, but rather rough assessment of a phenotypic group of vibrios that is to contain the main human pathogenic Vibrio species except Vibrio cholerae. This study has indicated that none of the established pathogenic Vibrio species, but members of the newly described family of Enterovibrionaceae make up for the bulk bacterial constituents associated with milk fish in tropical mariculture. This result diminishes the specificity of suc/suc+ ratios as a hypothesized indicator of human pathogens. Enterovibrionaceae comprize both Grimontia hollisae (Vibrio hollisae), an ‘‘old’’ re-classified human pathogen, and apparently non-pathogenic inhabitants of the intestine of sea fish larvae (Thompson et al. 2002). The absence of classic virulence genes from a representative isolate (V6) substantially reduces the likelihood of human health risks in tropical milk fish mariculture. Future risk assessment might have to pay particular attention to the frequency at which pathogenic traits can be acquired via horizontal gene transfer in fish intestine. Extremely high loads of sucrose negative vibrios in the anoxic intestine of adult milk fish and predominance of this group in the strongly anoxic sediments of fish farming areas would suggest that the bulk of sucrose negative vibrios encountered has originated from fecal pellets of milk fish and are most likely Enterovibrionaceae. This assumption has yet to be confirmed, as their dispersal with resuspended sediment into neighboring marine environments including coral reefs can only be inferred from circumstantial evidence. Particularly high prevalence of sucrose negative vibrios in mariculture areas located at greatest distance from the open sea (Anda) suggests the presence of further selective enrichment factors beyond low redox potentials. Previous investigations indicate that milk fish mariculture at Anda has frequently been affected simultaneously by lowered salinities, extremely low frequencies of fingerling survival and high abundances of water-borne sucrose-negative vibrios during SW monsoonal wet seasons. At the same time sucrose-negative vibrios prevailed only in the intestine of adult milk fish, not, however, in fingerlings (Reichardt et al. 2007, WR unpubl.). While milkfish fingerlings or other cultivated animals including holothurians and giant clams continue to be affected by diseases in the Lingayen Gulf, the pathogenic potential of sucrose –negative vibrios toward marine animals has yet to be elucidated. Certain as yet unclassified sucrose-negative vibrios that cause ulcerative white spot disease in Porites spp. corals at Bolinao may exemplify a latent pathogenic potential of sucrose negative vibrios of the study area toward marine animals (Arboleda and Reichardt, 2010), that deserves further taxonomic examination.

5. Conclusion In coastal marine environments affected by milk fish mariculture viable counts of sucrose-negative vibrios as indicators of non-cholera pathogens predominate nearly exclusively under anoxic conditions. These favorable conditions are provided in the intestines of adult milkfish and in the strongly reduced sediments of fish farming areas. It can be assumed that the enormous load of sucrose-negative vibrios in milk fish intestines is largely dispersed via sedimentation of fecal pellets onto sediments that turn rapidly anoxic under permanent sedimentation of fish farming waste. Human health risks arising from this accumulation of sucrose-negative vibrios prove lower than to be expected, since the predominant bacterial constituents are not pathogenic Vibrio

331

species, but apparently non-toxigenic Enterovibrionaceae. On the other hand, microbial habitats in fish guts remain a high risk environment for genetic exchanges of pathogenic traits. Further investigations will also have to focus on the most probable involvement of sucrose-negative vibrios from fish farming areas on diseases of milk fish fingerlings and maricultured invertebrates as well as on coral diseases on reefs in the vicinity of fish farming zones. Acknowledgments Access to municipal fish farming records by the Mayor of Bolinao is gratefully acknowledged. This investigation was supported by funds from WB-COE and from DAAD (Germany) to WTR, by a Philippine government grant (CHED) to the Marine Science Institute, and by a UP-OVPAA Emerging S&T grant to AOL. References Arboleda, M.D.M., Reichardt, W.T., 2010. Vibrio sp. causing Porites ulcerative white spot disease. Dis. Aquat. Org. 90, 93–104. Austin, B., 2010. Vibrios as causal agents of zoonoses. Veterinary Microbiology 140, 310–317. Barbieri, E., Falzano, L., Fiorentini, C., Pianetti, A., Baffone, W., Fabbri, A., Matarrese, P., Casiere, A., Katouli, M., Kuehn, I., Moellby, R., Bruscolini, F., Donelli, G., 1999. Ocurrence, diversity and pathogenicity of halophilic Vibrio spp. and non-01 Vibrio cholerae from estuarine waters along the Italian Adriatic coast. Appl. Environ. Microbiol. 65, 2748–2753. Beneduce, L., Vernile, A., Spano, G., Massa, S., Lamacchia, F., Oliver, J.D., 2010. Occurence of Vibrio vulnificus in mussel farms from the Varano lagoon environment. Lett. Appl. Microbiol. 51, 443–449. Bier, N., Bechlars, S., Diescher, S., Klein, F., Hauck, G., Duty, O., Strauch, E., Dieckmann, R., 2013. Appl. Environ. Microbiol. 79, 3570–3581. Bolinches, J., Romalde, J.L., Toranzo, A.E., 1988. Evaluation of selective media for isolation and enumeration of vibrios from estuarine waters. J. Microbiol. Meths. 8, 151–160. Chen, L.H., Yang, J., Yu, J., Yao, Z.J., Sun, L.L., Shen, Y., Jin, Q., 2005. VFDB: a reference database for bacterial virulence factors. Nuc. Acids Res. (Database issue) 33, D325–D328. Das, B., Manna, S.K., Sarkar, P., Batabyal, K., 2009. Occurrence of Vibrio parahaemolyticus in different finfish and shellfish species. J. Food Safety 29, 118–125. De Paola, A., Capers, G., Alexander, D., 1994. Densities of Vibrio vulnificus in the intestines of fish from the U.S. Gulf coast. Appl. Environ. Microbiol. 60, 984–988. Edouard, S., Daumas, A., Branger, S., Durand, J.-M., Raoult, D., Fournier, P.-E., 2009. Grimontia hollisae, a potential agent of gasteroenteritis and bacteraemia in the Mediterranean area. Eur. J. Clin. Microbiol. & Infectious Diseases 28, 705–707. Farmer III, J.J., Hickman-Brenner, F.W., 2006. The genera Vibrio and Photobacterium. The Prokaryotes 6, 508–563. Fouz, B., Llorens, A., Valiente, E., Amaro, C., 2010. A comparative epizootiologic study of the two fish-pathogenic serovars of Vibrio vulnificus biotype 2. Journal of Fish Disease 33, 383–390. Fukushima, H., Seki, R., 2004. Ecology of Vibrio vulnificus and Vibrio parahaemolyticus in brackish environments of the Sada river in Shimane Prefecture. Japan. FEMS Microbiol. Ecol. 48, 221–229. Ganesh, E.A., Das, S., Chandrasekar, K., Arun, G., Balamurugan, S., 2010. Monitoring of total heterotrophic bacteria and Vibrio spp. in an aquaculture pond. Current Research Journal of Biological Sciences 2, 48–52. Goshima, K., Owaribe, K., Yamanaka, H., Yoshino, S., 1978. Requirement of calcium ions for cell degeneration with a toxin (vibriolysin) from Vibrio parahaemolyticus. Infect. Immun. 22, 821–832. Hasan, N.A., Grim, C.J., Haley, B.J., Chun, J., Alam, M., Taviani, E., Hoq, M., Munk, C., Saunders, E., Brettin, T.S., Bruce, D.C., Challacombe, J.F., Detter, J.C., Han, C.S., Xie, G., Nair, G.B., Huq, A., Colwell, R.R., 2010. Comparative genomics of clinical and environmental Vibrio mimicus. PNAS 107, 21134–21139. Hinestrosa, F., Madeira, R.G., Borbaeu, P.P., 2007. Severe gastroenteritis and hypovolemic shock caused by Grimontia hollisae (Vibrio hollisae) infection. J. Clin. Microbiol. 45, 3462–3463. Holmer, M., Duarte, C.M., Heilskov, A., Olesen, B., Terrados, J., 2003. Biogeochemical conditions in sediments enriched by organic matter from net-pen fish farms in the Bolinao area, Philippines. Marine Pollution Bulletin 46, 147–01479. Jones, M.K., Oliver, J.D., 2009. Vibrio vulnificus: disease and pathogenesis. Infect. Immun. 77, 1723–1733. Lopez-Joven, C., Ruiz-Zarzuela, I., de Blas, I., Furones, M.D., Roque, A., 2011. Persistence of sucrose fermenting and non-fermenting vibrios in tissues of Manila clam species, Ruditapes philippinarum, depurated in seawater at two different temperatures. Food Microbiology 28, 951–956. Mahony, J.C., Gerding, M.J., Jones, S.H., Whistler, C.A., 2010. Comparison of the pathogenic potentials of environmental and clinical Vibrio parahaemolyticus

332

W.T. Reichardt et al. / Marine Pollution Bulletin 77 (2013) 325–332

strains indicates a role for temperature regulation of virulence. Appl. Environ. Microbiol. 76, 7459–7465. Matsunaga, N., Chisada, S., Fujioka, H., Takashima, K., Okino, N., Ito, M., 2011. Glycosphingolipid receptors for pathogenic vibrios in intestines of mariculture fish. Fisheries Science 77, 583–590. Nacorda, H.M.E., Obliosca, J.M., Tentia, M.C.L., Jacinto, G.S., San Diego-McGlone, M.L.C., 2012. Deterioration of soft bottom macrofaunal communities in a milkfish mariculture zone off Bolinao-Anda, Pangasinan (NW Philippines). In: Kawaguchi, M., Misaki, K., Sato, H., Yokokawa, T., Itai, T., Nguyen, T.M., Ono, I., Tanabe, S. (Eds.), Interdisciplinary studies on environmental chemistryenvironmental pollution and ecotoxicology. Terrapub, Tokyo, pp. 387–395. Noguerola, I., Blanch, A.R., 2008. Identification of Vibrio spp. with a set of dichotomous keys. J. Appl. Microbiol. 105, 175–185. Oberbeckmann, S., Wichels, A., Wiltshire, K.H., Gerdts, G., 2011. Occurrence of Vibrio parahaemolyticus and Vibrio alginolyticus in the German Bight over a seasonal cycle. Antonie van Leeuwenhoek 100, 291–307. Paillard, C., 2004. A short review of brown ring disease, a vibriosis affecting clams Ruditapes philippinarum and Ruditapes decussatus. Aquatic living resources 17, 467–475. Pascual, J., Macian, M.C., Arahal, D.R., Garay, E., Pujalte, M.J., 2009. Description of Enterovibrio nigricans sp. nov., reclassification of Vibrio calviensis as Enterovibrio calviensis comb. nov. and emended description of the genus Enterovibrio Thompson et al. 2002. International Journal of Systematic and Evolutionary Microbiology 59, 698–704. Reichardt, W., San Diego-McGlone, M.L., Jacinto, G.S., 2007. Organic pollution and its impact on the microbiology of coastal marine environments: A Philippine perspective. Asian Journal of Water Environment & Pollution 4, 1–9. Reichardt, W., Bolalin, G., Cutchon, X., Que, F., Balgos, M., 2011. Enzymatic degradation of proteinaceous compounds in sediments exposed to intensive mariculture in the tropics. Aquatic Ecosystem Health & Management 14, 318– 324. Rivera, S., Lugo, T., Hazen, T.C., 1989. Autecology of Vibrio vulnificus and Vibrio parahaemolyticus in tropical waters. Wat. Res. 23, 923–931.

Senderovitch, Y., Izhaki, I., Halpern, M., 2010. Fish as reservoirs and vectors of Vibrio cholerae. PLOS ONE 5, e8607. Simidu, U., Tsukamoto, K., 1985. Habitat segregation and biochemical activities of marine members of the family Vibrionaceae. Appl. Environ. Microbiol. 50, 781– 790. Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., Kumar, S., 2011. MEGA5: Molecular Evolutionary Genetics Analysis Using Maximum Likelihood, Evolutionary Distance, and Maximum Parsimony Method. Mol. Biol. Evol. 28, 2731–2739. Tantillo, G.M., Fontanarosa, M., Di Pinto, A., Musti, M., 2004. Updated perspectives on emerging vibrios associated with human infections. Letters Appl. Microbiol. 39, 117–126. Thompson, F.L., Hoste, B., Vandemeulebroecke, K., Swings, J., 2003a. Reclassification of Vibrio hollisae as Grimontia hollisae gen nov., comb. nov. International Journal of Systematic and Evolutionary Microbiology 53, 1615–1617. Thompson, F.L., Thompson, C.C., Swings, J., 2003b. Vibrio tasmaniensis sp. nov., isolated from Atlantic Salmon (Salmo salar L.). System. Appl. Microbiol. 26, 65– 69. Thompson, F.L., Iida, T., Swings, J., 2004. Biodiversity of Vibrios. Microbiology and Molecular Biology Reviews 68, 403–431. Thompson, F.L., Gevers, D., Thompson, C.C., Dawyndt, P., Naser, S., Hoste, B., Munn, C.B., Swings, J., 2005a. Phylogeny and molecular identification of vibrios on the basis of multilocus sequence analysis. Appl. Environ. Microbiol. 71, 5107–5115. Thompson, F.L., Thompson, C.C., Naser, S., Hoste, B., Vandenmeulebroecke, K., Munn, C., Bourne, D., Swings, J., 2005b. Photobacterium rosenbergii sp. nov. and Enterovibrio coralii sp. nov., vibrios associated with coral bleaching. International Journal of Systematic and Evolutionary Microbiology 55, 913–917. Vezzulli, L., Brettar, I., Pezzati, E., Reid, P.C., Colwell, R.R., Hofle, M.G., Pruzzo, C., 2012. Long-term effects of ocean warming on the prokaryotic community: evidence from the vibrios. ISME J 6, 21–30. Villanueva, R.D., Yap, H.T., Montano, M.N.E., 2006. Intensive fish farming in the Philippines is detrimental to the reef-building coral Pocillopora damicornis. Mar. Ecol. Prog. Ser. 316, 165–174.