Freshwater Aquaculture Nurseries and Infection of Fish with Zoonotic Trematodes, Vietnam

TOC summary: Aquaculture nurseries should be included in control programs to produce parasite-free fish for human consumption.

The epidemiology of FZTs is complex because humans and reservoir hosts, such as dogs, cats, pigs, and fi sh-eating birds, harbor egg-shedding adult stages (12,13). These hosts are infected by consumption of raw, inadequately cooked, or pickled fi sh. For many inhabitants in the Red River Delta provinces of northern Vietnam, the consumption of such fi sh dishes is a traditional behavior that is diffi cult to alter (14)(15)(16). In the Nam Dinh and Ninh Binh provinces, the widespread habit of eating raw fi sh is associated with a high FZT prevalence of 30%-40% in humans (3,4). Aquaculture fi sh species commonly used to prepare raw fi sh dishes, such as carp, frequently also have high a prevalence of FZT metacercariae (12,(17)(18)(19).
The infl uence of FZTs on the food safety of aquaculture products can have a noticeable adverse economic and public health effect because fi sh farming in Asia is expanding rapidly. Farm-raised fi sh are a main protein source consumed domestically and an essential product for exporting to other countries (2,10,20). Therefore, the production of FZT-free fi sh for human consumption should be a key objective for the aquaculture industry. Achieving this goal is seriously hampered, however, because the present state of knowledge on FZT infection in the fi sh production chain is inadequate to devise practical and sustainable prevention strategies, especially for small-scale and integrated freshwater aquaculture. The available knowledge of FZT infection is mainly obtained from studies of fi sh in growout ponds, where fi sh are harvested for human consumption (12,17,19).
Freshwater fi sh hatcheries in Nam Dinh, Ninh Binh, and Bac Ninh provinces include facilities such as a water reservoir, water storage facilities, breeding tanks, and incubators for hatching eggs. Depending on the hatchery, the water reservoirs are cement tanks or consist of earthen ponds from which the water is either pumped into cement breeding tanks or supplied from a tower to the breeding Freshwater Aquaculture Nurseries and Infection of Fish with Zoonotic Trematodes, Vietnam tanks and egg incubators. Water used for the breeding tanks and incubators is fi ltered through a net to remove different microbiota, e.g., zooplankton. Brood stock are moved from earthen ponds into cement breeding tanks for induced spawning. Fertilized eggs are then incubated in round cement incubators with running water for ≈5 days, depending on fi sh species and temperature. Newly hatched fi sh are termed fry. Fry are kept in tanks for 3-5 days, after which they are sold and subsequently stocked in earthen ponds in so-called nurseries.
The fi sh raised in nursery ponds are called juveniles. The nursery ponds in Nam Dinh and Ninh Binh provinces are mainly backyard earthen ponds located close to households and premises housing livestock and poultry. Juveniles are nursed up to 4 weeks and then sold for further nursing to bigger size or to be stocked to reach market size in growout ponds. Juveniles may be kept in ponds during the winter months (overwintered juveniles) for sale in early spring.
Management of nurseries in northern Vietnam often involves the application of livestock manure as fertilizer before stocking fi sh to increase the density of plankton that serves as a food source for juveniles. Additionally, farmers may also apply night soil (human manure) as fertilizers to the nursing ponds.
We report an investigation that aimed to determine the FZT infection status in integrated small-scale hatcheries and nurseries in Nam Dinh, Ninh Binh, and Bac Ninh provinces, which are major areas endemic for FZTs in Vietnam. By assessing the FZT metacercariae prevalence in fi sh from the initial stages of production, namely the hatcheries and nurseries, the study provides knowledge needed for a comprehensive assessment of FZT infection during the entire fi sh production cycle.
Juveniles were collected from 14 nurseries in Nam Dinh and 13 nurseries in Ninh Binh provinces by cast net (4-week old and overwintered juveniles) or scoop (1-weekold juveniles) nets. Fish were sampled twice from the nursing ponds, initially 1 week after the fry had been stocked and a second time at the end of the 1-month nursing period. In addition, overwintered juveniles were sampled from the same nurseries to compare their infection status with ju-veniles sampled before the winter period. On the basis of sampling size calculations, 15 juveniles were collected for each species at each time of sampling in each nursery.
The sample fi sh were transported to the laboratory for metacercariae examination and kept alive in plastic bags with added oxygen. The length and weight of each fi sh was recorded before samples were processed.
FZT infections in fry were examined by placing 5 fry on a glass slide, compressing them with another slide, and then examining for trematodes under a stereo microscope (×4) and a compound microscope (×100). Juveniles were digested in 1% pepsin to release metacercariae, following the procedure described by the World Health Organization (1) and modifi ed by Chi et al. (12). Identifi cation of the metacercariae was made according to morphologic features detailed by Pearson and Ow-Yang (21), Scholtz et al. (22), and Kaewkes (23).

Data Analysis
SAS version 9.1 (SAS Institute, Inc., Cary, NC, USA) was used for statistical analysis with fi sh as the study unit. Descriptive analyses of FZT infection in fry and juveniles were conducted by frequency distribution. Density of metacercariae in 1-week-old juveniles was calculated as number of metacercariae per whole fi sh, and density of metacercariae in 4-week-old juveniles and overwintered juveniles was calculated as number of metacercariae per gram of fi sh. Because of a skewed distribution of the density of metacercariae in fi sh, descriptive analysis was performed by means of median, Q 1 (25% percentile), Q 3 (75% percentile), minimum values, and maximum values.
Logistic regression analysis was used to evaluate prevalence differences of FZT between provinces, nurseries systems and fi sh species. A p value <0.05 was considered signifi cant.

FZT Species
No FZT metacercariae were found in fi sh fry from hatcheries in Nam Dinh, Ninh Binh, and Bac Ninh provinces. Table 1 shows the FZT species recovered from the various fi sh species and nursery systems examined in Nam Dinh and Ninh Binh provinces. The liver fl uke Clonorchis sinensis was found in only 1.5% of fi sh. Intestinal fl ukes, including Haplorchis pumilio and Centrocestus formosanus, were found in 55.6% and 41.0%, respectively, of the total number of FZT-infected juveniles. Several FZT species were found in individual FZT-infected juveniles. Table 2 shows the prevalence of FZTs in fi sh juveniles from Nam Dinh and Ninh Binh provinces and in different age groups of juveniles. Juveniles were already infected (14.1%) during the fi rst week of exposure in the nursery ponds with a further signifi cant increase (48.6%) after 4 weeks nursing when the juveniles were either sold for further stocking or kept in ponds during the winter months. Some additional but nonsignifi cant infection took place during the 5-6-month overwintering period.

FZT Density in Juvenile Fish
Among infected 1-week-old juveniles, 50% contained only 1 metacercaria/fi sh, 25% contained 2-4 metacercariae/fi sh, and 25% contained >5-18 metacercariae/fi sh. Among the 4-week-old juvenile fi sh, 50% had metacercariae densities <1.7/g fi sh tissue, 25% had densities of 1.7-<6.7/g fi sh tissue, and 25% had <6.7-173.3/g of fi sh tissue. In overwintered juveniles, 50% had a density <0.8/g, 25% had a density of 0.8-<2.3/g, and 25% had a density of 2.3-170.2/g. An explanation for this apparent decrease in density over the winter is that when overwintered juveniles have the same density as the non-overwintered juveniles at the start of the overwintering period, then as the fi sh grows during the overwinter period, the density will decrease if FZT transmission is reduced during the colder winter. The signifi cant increase in FZT prevalence seen among overwintered juveniles compared with 4-week-old juveniles indicates that some transmission of FZTs occurs during the winter. Table 3 shows the result of logistic analysis of FZT prevalence in 5 fi sh species and 3 nursery systems in Nam Dinh and Ninh Binh provinces. Fish juveniles from Ninh Binh Province had signifi cantly higher risk for being infected with FZT than those from Nam Dinh Province (p = 0.012). The odds of FZT infection for grass carp was 6× higher than for Rohu (p<0.0001), and silver carp had a 1.3× higher risk for FZT infection than Rohu. Combined for all fi sh species, overwintered juveniles and juveniles cultured for 4 weeks had odds of FZT infection that were 12.9× and 3.3× higher, respectively, than 1-week-old nursed fry. The nursing system had a signifi cant effect (p<0.0001) on FZT prevalence in 1-week-old juveniles.

Discussion
The fi nding in this study that fi sh become infected with FZT as early as in the fi rst week of growth in the nurseries and have a 48.6% infection prevalence when 4 weeks of age is of great consequence for public health and the aquaculture industry. It indicates that infection in nursery juveniles may account for most of the FZT prevalence reported for fi sh from grow-out ponds (12,17). Furthermore, because juveniles produced at nurseries are sold and distributed to numerous grow-out farms, the risk for FZT infection becomes widely distributed throughout the fi sh farming areas, complicating efforts to control this zoonosis. For example, the acquisition of FZT infection by overwintered juveniles in Nam Dinh and Ninh Binh provinces has implications for grow-out farms in nearby provinces such as Thanh Hoa, Thai Binh, and Ha Nam, as well as in provinces further north, which are major customers for these fi sh.
The increase of FZT prevalence during the fi rst month in the nursing ponds is much higher than that during the ≈12-month grow-out period in Nam Dinh Province (19).
Among possible explanations for this fi nding are more risky management practices by nursery operators or biological factors such as age-related fi sh susceptibility, resistance factors, and immunity. Further studies are urgently *n, no. fish infected with specific fishborne zoonotic trematode species; N, no. fishborne zoonotic trematode-infected fish; -, no parasites found. needed to understand the causes for this higher prevalence during the nursing phase and to develop effi cient prevention management. Grass carp, one of the favored sources of raw fi sh for human consumption, had a signifi cantly higher FZT prevalence than other fi sh species, which is consistent with results of a previous study (24). This may be due to the carp's food preference for grass and aquatic plants. In these nursing ponds, the grass and aquatic plants used for feeding were often harvested outside the farm, which may have introduced FZT-infected snails into the ponds. According to Phan et al. (19), feeding with green vegetation collected outside the farm is a risk factor for FZT infection in cultured fi sh. Another factor may be that grass carp spend more time in the littoral zone where vegetation and therefore snails may be concentrated. Furthermore, grass carp might be more susceptible to FZT infection than other fi sh species.
The higher prevalence of FZT metacercariae in Ninh Binh Province may be due to the common practice of using night soil to fertilize ponds to increase growth of plankton, a major source of carp feed. Nam Dinh Province also has a high human and domestic animal FZT prevalence (3,4,13), making fecal waste from these hosts a highly risky fertilizer. To develop effective control interventions, more research on methods to inactivate FZT eggs in manure must be conducted.
The lack of infection in fi sh fry is consistent with the report of Thien et al. (24), who investigated fry from hatcheries in the Mekong Delta. Fry were raised in well-controlled hatcheries by using tanks or ponds containing fi ltered water free of zooplankton, snails, and other microbiota.
One limitation of our study is that bias could occur when sampling fi sh by using 2 types of net, i.e., cast net and scoop net. However, cast nets cannot be used in fry incubators in hatcheries or to catch 1-week-old juveniles.
Different methods for examining and isolating FZT metacercariae in fi sh fry and juveniles are needed because compression of whole fi sh bodies is effi cient for processing of small fry and detecting any metacercariae present. For larger fi sh such as juveniles, however, the thickness of the body and muscle tissue precludes using compression, and instead, pepsin digestion (a more laborious procedure) is the method of choice. We also have compared pepsin digestion for several cups of fry with compressing methods and found no signifi cant difference in detection effi ciency.
Liver fl uke (C. sinensis) metacercariae had a low prevalence (1.5%); most FZT species found were intestinal fl ukes, particularly Haplorchis spp. and Procerovum varium. Although generally not considered to have a clinical role compared with liver fl ukes, several intestinal FZTs can cause serious pathologic effects, even death, when infecting the heart, brain, and spinal cord of humans (1,9,11). The dominance of H. pumilio trematodes in juveniles is similar to the fi ndings of other studies of nurseries (24) and growout systems (12,17,18). Stellantchasmus falcatus, however, is a major FZT species found in fi sh from southern and central Vietnam. Although Trung Dung et al. (4) recently recovered S. falcatus trematodes from humans in Nam Dinh, this species was not found in fi sh from this province, possibly because humans acquired their infections from fi sh originating from provinces other than Nam Dinh.
The life cycle and epidemiology of FZTs in aquaculture systems is complicated, and fi sh are vulnerable to multiple risks for infection. However, by developing a hazard analysis of critical control points approach to identifying points for intervention and enlisting the collaboration between different sectors such as human health, animal health, snail vector control, and aquaculture management and extension, we believe a major reduction in FZTs in farmed fi sh can be achieved, with the benefi t of being able to provide aquaculture products free of FZTs for human consumption. In conclusion, the results from this study strongly suggest that a program for the prevention for FZTs in farmed fi sh must include nurseries and grow-out farms. An integrated program will, however, require more research on the infection events occurring in the nursery pond, particularly the regulation of snail populations, fi sh risk behavior, and on methods to inactivate FZT eggs in manure intended for fertilizer use.