Seasonal variation on the ectoparasitic communities of Nile tilapia cultured in three regions in southern Brazil

A total of 240 Nile tilapia were examined between April 2007 and March 2008, gathered from three different fish farms, 20 fish in each fish farm, in the four seasons of the year. Fish ponds were located in Joinville, Blumenau and Ituporanga, Santa Catarina state, Brazil and each pond had a different culture system. Prevalence, mean intensity, mean abundance and mean relative dominance were compared among fish ponds and seasons. During this period, the water quality was kept in normal values. Piscinoodinium pillulare (Dinoflagellida) was the most dominant parasite followed by Trichodina magna e T. compacta (Ciliophora), Cichlydogyrus sclerosus, C halli, C. thurstonae, Scutogyrus longicornis (Monogenoidea), copepodids Lernaeidae gen. sp. The highest prevalence, mean intensity and mean abundance of ectoparasites were found on the body surface in fish from Joinville followed by Blumenau and Ituporanga. In the gills, the highest mean intensity and mean abundance were found in fish from Blumenau and Ituporanga in the winter. Piscinoodinium pillulare showed prevalence 100% during autumn in Blumenau and Ituporanga. In winter P. pillulare occurred in all study facilities. Fish from Joinville showed 100% prevalence of Monogenoidea during all seasons, as well as the highest mean intensity and abundance. The results showed that the majority of examined fish had higher infestations by protozoan during autumn and winter and higher infestations by metazoan have occurred in spring and summer.

The hypothesis of this study was to verify whether parasitological indexes in Nile tilapia may vary according to the season of the year in different production systems in the state of Santa Catarina, Southern Brazil. Every time the fishes were collected, water parameters such as pH, temperature, oxygen and transparency were availed and a 500 mL water sample was frozen for further ammonia analysis, according to Grasshoff (1976). After biometry, the fish were sacrificed (Ethics Commission n° 23080055748/2006-04/CEUA/UFSC) for parasitological analysis according to Ghiraldelli et al. (2006) methodology. The collection and parasites fixation followed Kritsky et al. (1995) and Eiras et al. (2006) methods.

Material and Methods
The quantification of protozoan was done by sampling five aliquot of 0.3 mL from a tissue (skin or gills) homogenate, counting the number of protozoan on the aliquot using a McMaster chamber and estimating the number of possible protozoan by the total homogenate volume. Monogenoidea and crustacean were counted under a stereomicroscope using marked Petri dish (Ghiraldelli et al., 2006). In tables, the species identified are referred as Monogenoidea. Trichodinids were identified according to Lom (1958), Van As and Basson (1989) and Ghiraldelli et al. (2006), Monogenoidea were identified according to Paperna and Thurston (1969), Ergens (1981), Douëllou (1993 and Pariselle and Euzet (1995).
Prevalence, mean intensity and abundance data were obtained according to Bush et al. (1997) and the mean relative dominance according to Rohde et al. (1995). The results were submitted to ANOVA and, upon significance, to Tukey test for comparison among arithmetic means. Significance level adopted was of 5% (Zar, 1999). The data were compared between seasons for each fish farm and between farms for each season. Table 1 shows that Joinville was characterised by its traditional method of fish culture with stocking density of 0.75 fish/m 2 , using 10% daily water renewal, sometimes using aeration and fish fed once a day. On the other hand,

Introduction
Ectoparasites are among the major etiological agents in Brazil and their presence is directly related to water quality and pond management (Moraes and Martins, 2004).
Parasitism occurs as a result of an interaction between host, parasite and environment (Buchmann and Lindestrøm, 2002). Some factors or substances are responsible for lowering the host immune response resulting in unbalanced host/parasite/environment interaction, factors such as water temperature, stress level (Xu et al., 2007), nutritional quality (Cavichiolo et al., 2002), age and natural immunity (Buchmann and Lindestrøm, 2002).
Studying the seasonal dynamics of carp infestation by T. nobilis Chen, 1963 in Yugoslavia, Nikolic and Simonovic (1998) observed higher infestation percentages in autumn and spring. On the other hand, Özer (2003) observed higher mean intensities of round goby (Neogobius melanostomus) infestations by T. domerguei Wallengren 1897 during spring and summer in Sirakirkagadar River, Sinop coast. In Brazil, Ranzani-Paiva et al. (2005) related high infestations of tilapia (Oreochromis niloticus) to the lowest temperature and water quality in Guarapiranga reservoir, São Paulo. Consequently, Schalch and Moraes (2005), observed a constant presence of this parasite during summer, autumn and winter, not exceeding 50% prevalence in fee fishing ponds.
High infestations by Monogenoidea were related to high ammonia levels in the water (Skinner, 1982) and high prevalence rates were related to water temperature in fish from India (Singhal et al., 1986) and from Finland lakes (Halmetoja et al., 1992). Koskivaara et al. (1991) observed higher mean intensity of Gyrodactylus von Nordmann, 1832 infestation on fish from eutrophic and polluted lakes. In intensive cultures, high fish density, low water flow and high organic matter concentration favors the growth and reproduction of parasites (Moraes and Martins, 2004).
In Mexico, Flores-Crespo et al. (1992), studying the seasonal variation of tilapia parasitised by Dactylogyrus Diesing, 1850, observed lower infestation intensity during autumn and winter and related the parasite presence to water temperature increase. In turn, Rawson and Rogers (1973) observed high infestation levels by Gyrodactylus macrochiri Hoffman and Putz, 1964 on Lepomis macrochirus and Micropterus salmoides during winter in Walter F. George Reservoir, Georgia.
In southern Brazil, the first fish parasitological study was performed by Azevedo et al. (2006) in tilapia from Nova Trento, Santa Catarina state, whom related higher highest water temperatures were observed in Blumenau (Table 2). Table 3 shows the biometrical data from the availed fish for each season.
There was a higher prevalence rate, mean intensity and mean abundance (p < 0.05) of ectoparasites in the Blumenau was a fee fishing facility in which the introduction of fish from other fish farms once a week was a common practice, fish stocking density 2.0 fish/m 2 , and fish fed once a day. Due to the fact that the fish stocking density in Blumenau was frequently uncontrolled, they used aeration three times a day. Consequently, Ituporanga practiced the consorted system between fish and pig manure, fish fed only at the finish of cycle and kept 3.5 to 4.0 fish/m 2 without water renewal. Only Joinville and Ituporanga assessed the water quality regularly. Water pH was stable in all fish ponds during the whole study period. In Joinville, dissolved oxygen was higher in autumn, while the lowest dissolved oxygen levels were found during summer and winter, respectively in Blumenau and Ituporanga. The

Discussion
The water quality was maintained within the acceptable values for tilapia, fish that supports broad variation in water quality (Boyd, 1979;Zaniboni Filho, 2004). The higher dissolved oxygen and water transparency values observed in Joinville were related to a high pond water renewal, a fact that distinguished it from the other fish ponds. In Ituporanga, oxygen levels were low during winter and summer, due to water stratification, and despite the pig manure deposition, ammonia levels were tolerable, as also observed by Azevedo et al. (2006).
Water transparency in Blumenau and Ituporanga was lower due to aspects of pond management, especially lack of fish entrance control and pig manure deposition, respectively. Temperature levels were below those recommended for tilapia culture, whose thermal comfort is between 27 and 32 °C (Kubitza, 2000). As related by Ghiraldelli et al. (2006), water quality assessment must be emphasised throughout the fish culture. The variation in water quality parameters could present severe consequences on fish health in high temperature climate (Tavares-Dias et al., 2008).
With the increase of intensive tilapia culture in Brazil, trichodinids started to play an important role in the list of potential fish pathogens. (Moraes and Martins, 2004;Martins and Ghiraldelli, 2008). Studying eels (Anguilla anguilla) in a recirculation system, Madsen et al. (2000) classified infestation by T. jadranica in four categories: skin mucus of fish from Joinville's fish pond. Trichodinids were the most representative parasites in the skin mucus (Table 3).
The highest mean intensities and abundance (p < 0.05) for Trichodina parasiting gills were found in fish collected in Blumenau and Ituporanga during winter (Table 3).
A prevalence rate of 100% was observed for P. pillulare (Table 4) in Blumenau and Ituporanga during autumn. Infestation levels for this parasite were high during winter in all farms.
Regarding infestation by Monogenoidea on the gills, fish from Joinville presented prevalence of 100% during all seasons, as well as higher mean intensity and abundance values when compared to Blumenau and Ituporanga (p < 0.05). In Joinville, there was a significant increase on Monogenoidea parasitism during winter and spring. Contrarily, in Blumenau, there was no significant difference (p < 0.05) between seasons, whereas Ituporanga presented higher values during winter (p < 0.05) ( Table 5).
The ectoparasites mean relative dominance values found in this study (   Ghiraldelli et al. (2006), fish presented lower mean intensities of parasitism. In this study, neither mortality nor clinical signs of disease were observed, probably due to lower temperatures when compared to Southeast Brazil (Tavares-Dias et al., 2001a. Massive infestations by the dinoflagellate P. pillulare most of the time culminates in high mortality rates in cultured fish (Martins et al., 2001) and its dissemination is related to water quality (Shaharom-Harrison et al., 1990;Moraes and Martins, 2004). Prevalence of P. pillulare in tilapia was higher in the present study than those observed by Tavares-Dias et al. (2001a in pacu (Piaractus mesopotamicus), piauçu (Leporinus macrocephalus), matrinxã (Brycon amazonicus) and in the hybrid tambacu (P. mesopotamicus x Colossoma macropomum). Contrary to that related in this study, Tavares-Dias et al. (2001a) did not verify seasonality on the occurrence of P. pillulare, when the parasite number was lower in winter.
Great infestation values by P. pillulare on gills of tambacu can cause excessive skin mucus production, loss of epithelium, paleness and lamellar hyperplasia on gills, petechiae and congestion, resulting on the death of 4,000 fish in the first 24 hours and 3,000 more in the next 15 days (Martins et al., 2001). With its fixation structures, named rhizocysts, P. pillulare causes severe damage to gill tissue in high infestations (Lom and Schubert, 1983). Hundred percent mortalities were also observed in cultured 0 (no parasites), 1 (from 1 to 10 parasites), 2 (from 11 to 100), 3 (from 100 to 1000).
In the present study the highest mean intensities for trichodinids were found on the skin mucus of fish from Joinville, during autumn and spring, corresponding to category 3 of Madsen et al. (2000). Trichodinids were present in all seasons, but higher infestation levels were seen in spring.
On the gills, trichodinids parasitism was also classified as category 3 during autumn in Joinville, winter in Blumenau, spring and summer in Ituporanga. This study confirms that the high eutrophication caused by the deposition of pig manure in Ituporanga was responsible for keeping high levels of trichodinids, which agrees with the findings of Afifi et al. (2000).
Regarding seasonality, this study corroborates Nikolic andSimonovic (1998) andÖzer (2003), who found higher infestation intensities by T. nobilis and T. domerguei during spring and autumn, and during spring and summer, respectively. Özer (2000) also found high mean intensity of T. mutabilis Kazubski and Migala, 1968 during spring. Comparatively, fish in this particular study presented higher infestation intensities than those studied by Özer (2000). Özer and Erdem (1999) verified that trichodinids occurred during all seasons, but with higher infestation levels during spring, which was also verified in the present study.
Prevalence of trichodinids found in this study were higher than those found in tilapia from Guarapiranga reservoir studied by Ranzani-Paiva et al. (2005) and than those found in the same farms by Ghiraldelli et al. (2006)  Puntius gonionotus from Malaysia and were related to high ammonia levels in water by Shaharom-Harrison et al. (1990).
In this study, the highest values of P. pillulare infestation on tilapia gills during winter in Joinville, autumn and winter in Blumenau and autumn in Ituporanga were probably related to low water temperature, which is in agreement with the data presented by Martins et al. (2001). These authors registered mortalities in Brazilian fish cultured during winter in southeast Brazil. Dinoflagellate was the most dominant parasite in all three fish farms, with higher mean abundance during autumn and winter, probably due to lower fish resistance in low temperatures. Also, it is worth mentioning that management characteristics do not seem to exert an influence on the infestation levels.
Monogenoidea, parasites with high host specificity, play an important role in intensive fish pond culture of both ornamental and food fish in Brazil (Garcia et al., 2003;Moraes and Martins, 2004;Lizama et al., 2007). Cichlidogyrus sclerosus was considered a chemical pollution biomarker for tilapia (Sanchez-Ramirez et al., 2007).
The highest prevalence rates, mean intensities and mean abundance for Monogenoidea, for all study seasons, were observed on fish gills from Joinville. Water quality can influence on Monogenoidea parasitism, as reported by Flores-Crespo et al. (1992), who verified higher intensity of dactylogyrids in tilapia at higher temperatures. Similar results were verified by Cecchini et al. (1998) studying the life cycle of Diplectanum aequans Wagener, 1857. Contrarily, Koskivaara et al. (1991) reported that gyrodactylids on Rutilus rutilus were more abundant during autumn in Finland. Mortality on carp fry parasitised by Dactylogyrus vastator Nybelin, 1924 exposed to low water oxygen concentrations was observed by Molnár (1994). Garcia et al. (2003) reported that parasitism by Urocleidoides Mizelle and Price, 1969 was negatively influenced by pH, temperature and electric conductivity of water.
In the present study, though, it was not possible to associate the presence of Monogenoidea, neither on the gills nor on the skin mucus, with water quality. Something worth reporting is the significantly higher mean parasitism intensities on fish gills from Joinville. These values were higher than those reported by Tavares-Dias et al. (2001b) in tilapia from fee-fishing ponds; by Ghiraldelli et al. (2006) in tilapia from the same farms assessed in this study; and by Ranzani-Paiva et al. (2005) in tilapia from Guarapiranga reservoir, southeast Brazil. Lizama et al. (2007), registered occurrence of Monogenoidea during all studied period in a fish pond from Assis city, such as observed in this study. The farm from Joinville monitored water quality regularly in its ponds and kept a high water renewal rate, which contributed to the higher water transparency observed. The only water parameter that distinguished Blumenau from Ituporanga was also water transparency.
These results show that most of the analysed fish had greater infestations by protozoan during autumn and winter and by metazoan during spring and summer seasons. The use of trichodinids infestation as a good water quality biomarker for fish pond cultures can be