Prevalence and Distribution of Vibrio spp. in Wild Aquatic Birds of the Southern Caribbean Sea, Venezuela, 2011–12

The genus Vibrio consists of more than 100 species associated with aquatic organisms, birds, mammals, and humans, in commensal or pathogenic relationships (Thompson et al. 2004). Although data on the association between Vibrio and birds is limited, the main human pathogenic species (Vibrio cholerae, Vibrio parahaemolyticus, and Vibrio vulnificus) have been reported in aquatic birds, mainly from temperate latitudes (Miyasaka et al. 2006; Vezzulli et al. 2010); limited studies of Vibrio occurrence in tropical aquatic birds have been conducted. We evaluated the prevalence and distribution of Vibrio in aquatic bird feces from the southern Caribbean Sea.

Fieldwork was conducted at two protected coastal areas in Venezuela from October 2011 through March 2012: Cuare Wildlife Refuge (10°57′N; 68°21′W) in the northwest, and Margarita Island (10°52′N; 64°03′W) in the northeast (Fig. 1). Both areas have substantive bird richness and abundance. We took fresh fecal samples (n=160) from resident, migratory (regional and long-distance), and nonidentified birds from mangroves and soils from shallow waters (Table 1). Nonidentified bird feces were collected on soils where no specific groups were identified. Soils close to feces were collected as controls. Long-distance migratory individuals: Semipalmated Sandpiper (Calidris pusilla; n=2), Least Sandpiper (Calidris minutilla; n=1), and Semipalmated Plover (Charadrius semipalmatus; n=1) were captured using mist nets. Live birds were placed in boxes to collect feces and then released. Samples were collected with Venezuelan Ministry of Environment permits; specimens were transported frozen to the laboratory, and kept at −80 C.

We purified DNA with Power soil DNA kit (MO BIO Inc., Carlsbad, California, USA). The Vibrio 16S rRNA gene was amplified using genus-specific primers (Liu et al. 2006). We tested V. cholerae and cholera toxin using the 16S-23S rRNA intergenic spacer region (ISR; Chun et al. 1999) and the ctxA (Lipp et al. 2003), respectively. Vibrio parahaemolyticus thermolabile haemolysin (tl; Rizvi and Bej 2010) and V. vulnificus haemolysin/cytolysin (vvhA; Warner and Oliver 2008), were evaluated. We performed PCR using Ready-To-Go PureTaq PCR kit (Amersham Biosciences, Piscataway, New Jersey, USA), containing 2–8 μL of DNA, 3 μL of primers (5 μmol/L), and sterile water to 25 μL in a Gen AMP 9700 thermal cycler (Applied Biosystems, Foster City, California, USA).

Genus-specific 16S rRNA fragments (∼500 base pairs [bp]) were purified with QIAquick PCR Purification Kit (Qiagen, Valencia, California, USA) and sequenced at Macrogen Inc. (Seoul, Korea). All fecal sequences were 99–100% homologous to Vibrio 16S rRNA sequence and were deposited in GenBank (KP762347–KP762357). The sequences were aligned with the closest matches using SINA Software (Pruesse et al. 2012). We constructed a phylogenetic tree using the neighbor-joining algorithm and the Jukes–Cantor model in Molecular Evolutionary Genetics Analysis 5.0 (Tamura et al. 2013). The stability of grouping was estimated by bootstrap analysis (10,000 replicates).

We used the nonparametric Mann-Whitney test for two independent samples to determine differences (P<0.05) among sites and bird migratory status in the Vibrio and V. cholerae PCR results using the PAST program (Hammer et al. 2001).

Vibrio was amplified by PCR in 40% of samples collected from both sites. In Cuare Wildlife Refuge, the prevalence of Vibrio was 47% of resident, 36% of migratory, and 32% of nonidentified birds. In Margarita Island, the prevalence was 33% of resident, 44% of migratory, and 47% of nonidentified birds. A higher prevalence of Vibrio was detected for migratory birds at Margarita Island than Cuare Wildlife Refuge; however, the difference between sites was not statistically significant (P=0.485; Table 1). Additionally, there were no statistically significant differences in the prevalence of Vibrio between resident and migratory birds (P>0.05). Soil controls did not have detectable Vibrio DNA. These results are consistent with previous findings related to the association of Vibrio spp. with aquatic birds worldwide (Miyasaka et al. 2006; Vezzulli et al. 2010).

We did not detect V. parahaemolyticus and V. vulnificus genes in any samples. The ISR of V. cholerae was detected only in Cuare Wildlife Refuge with a prevalence of 18% in residents, 7% in migratories, and 12% for nonidentified birds; no significant differences were found between resident and migratory birds (P=0.205; Table 1). Although we did not detect the ctxA gene for cholera toxin, the emergence of new toxigenic clones has been reported from marine environments (Thompson et al. 2004). Few investigators have found V. cholerae in aquatic bird feces (pelicans [Pelecanus spp.], herons [Ardeidae], gulls [Larus spp.], cormorants [Phalacrocorax spp.], ducks [Anatidae], and geese [Anatidae]; Lee et al. 1982; Ogg et al. 1989). To our knowledge, this is the first report of V. cholerae in Wattled Jacana (Jacana jacana), terns (Phaetusa spp.), sandpipers (Calidris spp.), and flamingos (Phoenicopterus spp.). The increasing number of Vibrio spp. in these sites suggests their spread in the digestive tract of aquatic birds in the Neotropics.

As inferred from the 16S rRNA partial sequencing and phylogenetic analysis, most fecal sequences from Cuare Wildlife Refuge formed clusters with V. cholerae and only one with V. vulnificus, whereas sequences from Margarita Island were grouped with Vibrio alginolyticus and Vibrio harveyi (Fig. 2). Accurate identification of vibrios at the family and genus levels is obtained by 16S rRNA sequencing. However, exclusive comparisons of 16S rRNA at the species levels might provide limited resolution for identifying Vibrio spp. (Thompson et al. 2004). In Cuare Wildlife Refuge, four Vibrio sequences formed a clade with V. cholerae in agreement with most ISR results by PCR, which confirms its presence in Large-billed Tern (Phaetusa simplex), American Flamingo (Phoenicopterus ruber), and Semipalmated Sandpiper. In particular, the detection of V. cholerae in a sequence from a Semipalmated Sandpiper (KP762351) demonstrates the presence of this bacterium in the gastrointestinal tract.

Another sequence from an American Flamingo (KP762350) clustered with V. vulnificus, which has only been reported in gulls and ducks from Japan (Miyasaka et al. 2006). On the other hand, the sequences from Margarita Island grouped mainly with species related to the Harveyi clade (V. harveyi, V. alginolyticus, Vibrio campbellii, and V. parahaemolyticus). However, these bacteria have not been reported in aquatic wild birds before. Although a previous study identified V. cholerae by biochemical and serologic tests in Wilson's Plover (Charadrius wilsonia) and Greater Yellowlegs (Tringa melanoleuca) samples from a coastal lagoon close to Margarita Island, this report should be confirmed by molecular tests (Rodriguez et al. 2010). There is a surprising clear distribution of Vibrio species between sites (Fig. 2), considering that aquatic birds are very mobile animals. Long-distance migrants such as sandpipers or flamingos are able to move hundreds of kilometers in one day (Halpern et al. 2008).

Vibrio presence in bird feces could be explained by their feeding habits as a source of bacteria in the bird's gut. This is supported by a previous report of V. cholerae in seawater and plankton from Cuare Wildlife Refuge (Fernández-Delgado et al. 2009), whereas species related to the V. alginolyticus, V. harveyi, and V. campbellii group have been found in the Cariaco water column close to Margarita Island (Bozo-Hurtado et al. 2013). Fish have been demonstrated to be reservoirs of Vibrio spp. (Chatterjee and Haldar 2012), including V. cholerae (Halpern et al. 2008; Senderovich et al. 2010). Herons, Large-billed Terns, Black Skimmers (Rynchops niger), Magnificent Frigatebirds (Fregata magnificens), and Neotropic Cormorants (Phalacrocorax brasilianus) feed on fish (Hilty 2003). Although our PCR results do not allow us to establish if the Vibrio we identified came from the bird's diet or their intestinal microbiota, most reports indicate that invertebrates—mainly copepods (Crustacea), chironomids (Diptera), crabs (Crustacea), protozoa, and bivalves (Bivalvia)—and fishes—tilapia (Oreochromis niloticus), salmon (Salmo salar), rainbow trout (Oncorhynchus mykiss), and eels (Anguilliformes)—are consumed by aquatic birds (Halpern et al. 2008; Vezzulli et al. 2010; Chatterjee and Haldar 2012). More studies are necessary to understand the Vibrio dynamics and interactions with birds and aquatic reservoirs. The ubiquity of certain resident birds (herons, terns, skimmers, jacanas, and ibises) along coastlines, and their association with human activities, such as fishing, shellfish beds, and tourism, might pose significant risk for the spread of vibrios to areas of ecologic and human health importance (Miyasaka et al. 2006; Vezzulli et al. 2010). Our data indicate that several species of Vibrio are common in aquatic birds along the southern Caribbean Sea. Culture and isolation of Vibrio is needed to gain more information regarding the species, the strains, and their pathogenic potential. Our results contribute to understanding the role of birds as reservoirs of pathogenic bacteria.

We gratefully acknowledge F. Perozo, H. Chirinos, and C. Azpurua for sampling facilities, G. Velasquez from Unidad de Sistemas de Información Geográfica—Instituto Venezolano de Investigaciones Científicas (IVIC) for support with Figure 1 and M. López for English manuscript improvements. This work was funded by the IVIC and the Universidad Simón Bolívar.