TEXT

Top Abstract Text References

In a previous study, we demonstrated that consumption of raw turtle eggs is a risk factor for cholera-like diarrhea caused by Vibrio mimicus, due to the fact that some strains produce the choleric toxin (CT) (2, 4). Most of the turtle eggs consumed in Costa Rica come from the National Wildlife Refuge at Ostional, located at 10°00'00"N and 86°45'50"W. Ostional is one of the most important beaches in the world for the massive arrival and nesting of more than 100,000 olive ridley sea turtles (Lepidochelys olivacea) (3, 7). During the arrivals, which usually occur 7 to 11 times a year and last five consecutive nights, the turtles come to the beach to dig holes about 30 cm deep (6). In this sand nest they deposit 98 to 123 eggs, and immediately afterward they cover them with sand and return to the ocean (3, 7, 12). The eggs deposited on the first and second nights of arrival have a very low probability of hatching, because most of them are destroyed by turtles that arrive on the following nights (7). This massive destruction also causes an increase in the protein content of the sand, which stimulates proliferation of fungi and bacteria (7, 13). The level of microbial contamination of the sand on this beach varies constantly, not only because of those effects of the arrival but also because of the tides, climatic factors, and predatory excavations, among other factors (7). It has been suggested that the rise in microbial contamination of the sand contributes to the decrease in viability of the eggs in the nests that have survived destruction by other turtles or predators (3, 7). Because of this natural loss of eggs, in 1985 the people from Ostional organized themselves for the commercial exploitation of turtle eggs under the supervision of the Ministry of Environment and Energy (MINAE), the Costa Rican Institute of Fisheries (INCOPESCA), and the Ocean Turtle Project of the University of Costa Rica (9). Currently about 100 low-income families that live in this community and many others involved in egg commercialization depend on this economic activity (1, 7). The collection of up to 1,000,000 eggs deposited during the first and second arrival nights is an activity carried out during the day in a purely manual way: while some people dig the nests, extracting the eggs with their hands, others collect the eggs in sacks and later pack them in plastic bags that hold 200 eggs each, with no processing and with the sand still adhering to them. During the two collection days these bags are kept at room temperature (23.1 to 34.5°C during the dry season and 22.8 to 27.9°C in the rainy season), and then they are shipped in nonrefrigerated trucks to different parts of the country.

To our knowledge, the present study is the first one to identify the sources and factors that contribute to the contamination of turtle eggs with V. mimicus. This information could lead to the development of low-technology-level measures for the improvement of the sanitary quality of the turtle eggs and in this way reduce the risk of diseases caused by the consumption of the eggs.

Since in Costa Rica the eggs that are legally sold are those collected during the first two days of arrival, three of the four samplings for this study were carried out at this time (Table 1). Cloaca swabs were obtained during the night, when the turtles were starting to spawn. For that, a sterile swab was introduced approximately 3 cm into the cloaca and rubbed against all the internal walls and then placed in Cary Blair transportation medium. Eggs were also collected directly from the cloaca during spawning; sterile gloves were used to catch them before they touched the sand. Immediately afterward, they were placed in sterile plastic bags and the nest was marked for easy identification the next morning. The morning after spawning, between 6:00 and 7:00 a.m., each nest was located and excavated with gloves in one of its sides to collect an average of 5 to 10 eggs (nest eggs), and then the nest hole was carefully covered with sand. This procedure was repeated at noon, with excavation being conducted on the opposite side. Sand samples were obtained with a sterile ladle and container. The first sand sample was collected from the bottom of the nest at the time of spawning. The second and third samples were collected at a depth of 35 cm, 6 and 12 h after spawning, at the same time that nest eggs were obtained. For estuarine water samples, 100-ml sterile containers were introduced against the current. All the samples were placed immediately in coolers and sent to the laboratory for processing within 24 h. In the laboratory, the samples were inoculated in alkaline peptone water enrichment, pH 8.5 (APA), and subcultured 6 h later in TCBS agar. Eggs were individually processed, each one being submerged in 150 ml of APA. From each sample, five to eight TCBS colonies, including green and yellow ones, were selected. Isolation and identification of the Vibrio species were conducted by conventional methods and reconfirmed with the API 20E system (Biomérieux SA) with modifications recommended previously (8, 10, 11). All of the enrichment broths were also tested for Vibrio cholerae O1 with SMART by the procedure described by the manufacturer (New Horizons Diagnostic Corporation). All V. mimicus and V. cholerae non-O1 strains isolated were tested to detect the ctxA gene of V. cholerae O1 by the PCR technique utilizing the primers Col-1 and Col-2, with modifications described previously (2, 14).

In this study, most V. mimicus-positive samples yielded both ctxA⁺ and ctxA mutant strains, with a predominance of mutant strains, as was expected from environmental samples (4). This situation must be considered in designing procedures for the detection of ctxA⁺ strains from these sources. For statistical analysis, those samples with at least one ctxA⁺ strain were designated V. mimicus ctxA(+). During the arrival periods V. mimicus was recovered from eggs of all the nests (13/13) studied along the beach, and toxigenic strains were recovered from 31% of them (4/13). All nests with ctxA(+) strains were located near the estuary, in the area of higher anidation density, where eggs are collected for commercial sale. Even in nests that yielded V. mimicus-positive eggs, not all of the eggs were contaminated and only 47% (89/190) of the eggs from the nests had V. mimicus, while ctxA(+) strains were found only in 5% (9/190) of the eggs. A significant difference was found between the contamination of eggs collected during arrival and those collected before arrival time (2% [1/61]; P < 0.0005; ²). Moreover, as was observed in eggs, it was demonstrated that V. mimicus contamination of the sand was significantly higher during the arrival time (42% [13/31]) than before arrival (0% [0/16]; P = 0.003; ²). Aside from detection in nest eggs, toxigenic strains were detected in only 6.4% (2/31) of the arrival period sand samples. Also, the three V. mimicus ctxA mutant strains isolated from estuarine water were recovered during the arrival period. It is not clear why the bacterial density increases during the arrival. Some factors which could explain this phenomenon are (i) the increase in the numbers of turtles and predators, such as vultures, dogs, pigs, and others, that come to the beach, (ii) the dramatic increase in the protein content of the sand due to the destroyed eggs during the first arrival evenings, and (iii) the variations in temperature, humidity, incidence of UV light, and other climatic variables (7). Further studies are necessary to determine the cause. This is important because the collection of turtle eggs for commercial purposes is carried out during the first two days of arrival. Therefore, it can be expected that the later the collection, the greater the probability that eggs will be contaminated.

V. mimicus contamination levels varied according to the season and the time at which eggs were collected. In the dry season, the level of contamination of the eggs was significantly higher at 6:00 a.m. than at noon, 50% (15/30) and 20% (6/30), respectively (P = 0.03; ²). In the rainy season, contamination tended to increase the longer the eggs remained on the sand, from 46% (30/65) at 6:00 a.m. to 57% (37/65) at noon; however, this difference was not significant.

Since V. mimicus was also isolated in 15% (4/26) of the cloaca swabs and 11% (11/100) of the eggs extracted directly from the cloaca, we evaluated whether the V. mimicus contamination of the nest eggs was associated with contamination of the cloaca or contamination of the sand. For these statistical analyses, we included only data for those turtles and nests from which cloaca swabs, cloaca eggs, nest eggs, and sand samples for the same turtle were available. We found no statistical association when evaluating the influence that the turtle cloaca has on the contamination of the cloaca eggs, nest eggs, and sand (P = 0.09, 0.5, and 0.3, respectively). Thus, the isolation of V. mimicus from those samples could be due to the difficulty in getting an adequate sample (free of sand, seawater, or bacterial flora of the chelonian "skin") since these were taken at the precise time of spawning, under natural conditions (in open air with wind at night with bad lighting and with a large quantity of mucus present when the eggs were coming out). In contrast, in the study of the influence of the sand on the contamination of nest eggs, cloaca eggs, and cloaca, a significant relationship was observed only with nest eggs exposed to the sand (P = 0.01; Fisher's exact test).

V. cholerae O1 was not detected in any of the samples ana-lyzed by either the traditional method or the SMART test. Other members of the family Vibrionaceae recovered from the different samples were V. mimicus and an Aeromonas sp. from the cloaca, cloaca eggs, nest eggs, sand, and estuarine water; ctxA mutant V. cholerae (non-O1), V. parahaemolyticus, and V. vulnificus from nest eggs, sand, and estuarine water; and V. alginolyticus from the cloaca swabs only. These bacteria are considered to be among the emerging microorganisms which cause not only gastroenteritis but also fatal septicemia in humans (11). The finding of V. mimicus in estuarine water is also important because it could cause otitis in people who swim in it. This should be an important consideration for children, immunodeficient individuals, or tourists who are exposed to this agent for the first time.

In conclusion, contamination of turtle eggs with V. mimicus and other vibrios of medical importance was demonstrated. The proportion of V. mimicus-contaminated eggs found during the time that commercial collection is allowed was high (47%), but the proportion of ctxA(+) isolates was much lower (5%). Although it has been shown that CT is not the only toxin involved in V. mimicus pathogenesis, it is probably the most important one for the watery diarrhea presentation (4, 5). Thus, the relatively low proportion of ctxA(+) strains found in nest eggs indicates that not all of the people who eat this product raw will have the same risk of developing cholera-like diarrhea. Since samples for this study were collected even out of the main commercial collecting area (away from the estuary), the proportion of toxigenic strains in commercial eggs might be even higher than 5%. An understanding of why toxigenic strains were found only in nests near the estuary was not in the scope of the present work but should be the focus of future investigations.

As far as the source of V. mimicus egg contamination is concerned, the results of this study suggest that it is more likely that the eggs acquired the contamination when they became exposed to the contaminated sand in the nest rather than from the mother.

Further research should also be done to study possible corrective measures to improve the sanitary quality of the eggs and/or to change the consumption habits of local people to prevent illnesses transmitted by this mechanism.