Introduction
The chub mackerel, Scomber japonicus Houttuyn (Scombridae), a commercially important species in the Madeira Economic Exclusive Zone, inhabits the epi-pelagic regions of the Atlantic, Indian and Pacific Oceans and adjacent seas (Collette, Reference Collette, Whitehead, Bauchot, Hureau, Nielsen and Tortonese1986). Research on the morphology, occurrence and infection dynamics of the helminth parasites of this fish species in Madeiran waters (Atlantic Ocean, Portugal), has been conducted since 1999. The results can be used to study the biogeographical patterns of the observed helminths, which in turn can be important to examine the existence of latitudinal gradients in species diversity (Rohde, Reference Rohde1992; Rohde & Heap, Reference Rohde and Heap1998). Other studies on the parasite composition of scombrids have characterized helminths both from the visceral cavity and gills (Rego & Santos, Reference Rego and Santos1983; Pozdnyakov & Vasilenko, Reference Pozdnyakov and Vasilenko1994; Cremonte & Sardella, Reference Cremonte and Sardella1997). In particular, the characterization of ectoparasite assemblages for species of the genus Scomber was carried out in temperate regions of the North Atlantic (Somdal & Schram, Reference Somdal and Schram1992) and tropical waters of Australia (Hayward et al., Reference Hayward, Perera and Rohde1998). To date, six species of monogeneans (Monogenea: Mazocraeidae) have been described from the gills of Scomber spp., including the chub mackerel, namely Pseudokuhnia minor Rohde & Watson, Kuhnia scombri (Kuhn) Sproston, Kuhnia scombercolias Nasir & Fuentes Zambrano, K. sprostonae Price, Grubea cochlear Diesing, and Grubea australis Rohde. These monogeneans were found in the gills and pseudobranchs of mackerels. The present work analyses the results on the occurrence and infection dynamics of five species of monogenean in Scomber japonicus from Madeiran waters, followed by a comparison of their occurrence in other oceanic regions. The biogeographic patterns of these parasites are also discussed.
Materials and methods
One hundred and fifty one chub mackerel, S. japonicus, (sample 1) from Madeiran waters, Atlantic Ocean, Portugal, were examined for the occurrence of gill monogeneans. Fish samples ranging in total length from 23 to 40 cm, were obtained from the local fish market of Funchal, from November 2004 to February 2005. Additionally a sample of 30 chub mackerel, ranging in length from 24 up to 35 cm, was purchased in August 2005 (sample 2). Chub mackerel available in Funchal were caught locally, by a purse seining technique called ‘ruama’. Fish were dissected one day after capture, measured in centimetres, weighed in grams, and the gills dissected out and examined for the presence of monogeneans under a Zeiss Stemi 2000 C stereomicroscope, equipped with a Sony digital camera. Monogeneans were counted, fixed in 70% ethanol, cleared in lactophenol, and mounted in glycerol jelly. Following microscopic examination, measurements of the monogeneans, taken with the aid of an ocular micrometer, are given in millimetres (mm) or micrometres (μm) and presented as means ± standard deviations. Prevalence, intensity and abundance indices were calculated after Bush et al. (Reference Bush, Lafferty, Lotz and Shostak1997). Confidence intervals for the prevalence and intensity of infection, aggregation indices (D), (k) and variance to mean ratios (s 2/x), were calculated using the program Quantitative Parasitology (QP 2.0) of Reiczigel & Rózsa (Rózsa et al., Reference Rózsa, Reiczigel and Majoros2000), according to Poulin (Reference Poulin1993) and Bush et al. (Reference Bush, Fernández, Esch and Seed2001). The significance of the relationship between prevalence and abundance of infection with host length, sex, weight and date of sampling were tested by univariate analysis of variance using SPSS 12.0. Results were considered significant where P ≤ 0.05.
Results
The sample of 151 chub mackerel (sample 1) consisted of 89 females and 62 males. The length of female fish ranged from 23 to 38 cm (28.54 ± 2.74, n = 89) whereas the length of males was 23.5–40.0 cm (28.99 ± 3.11, n = 62). A total of 2332 monogeneans belonging to five different species were found attached to the gill filaments of the chub mackerel, namely Pseudokuhnia minor (length 1.15–2.26 mm), Kuhnia scombri (length 1.52–4.12 mm), Kuhnia scombercolias (length 1.62–2.64 mm), Kuhnia sprostonae (length 4.46 mm) and Grubea cochlear (length 10.3 mm) (table 1). The most abundant monogenean species was Pseudokuhnia minor (n = 2072), followed by Kuhnia scombri (n = 148), K. scombercolias (n = 106), Grubea cochlear (n = 5) and K. sprostonae (n = 1) (table 2).
Table 1 Measurements of total body length, clamp and hamuli size for four species of Kuhnia recovered from the gills of the chub mackerel Scomber japonicus from Madeiran waters.
Table 2 Prevalence, intensity and abundance of monogeneans on the gills of 151 chub mackerel Scomber japonicus with body lengths ranging from 23 to 40 cm, examined from November 2004 to February 2005.
The highest prevalence, 98.68%, (95% confidence limits 95.29 to 99.84%) was found for P. minor, followed by a prevalence of 43.71% for K. scombri (95% confidence limits 35.66 to 52.01%) and K. scombercolias with 39.07% of fish infected (95% confidence limits 31.24 to 47.34%) (table 2). Grubea cochlear occurred in only five fish and K. sprostonae in one fish. Pseudokuhnia minor was also the species with the highest mean abundance (13.72, 95% confidence limits 12.40 to 15.06; bootstrap indices: 46, 1945, sample size = 151) and mean intensity (13.91, 95% confidence limits 12.62 to 15.30; bootstrap indices: 54, 1954, sample size = 149). The mean abundance with K. scombri was 0.98 (95% confidence limits 0.76 to 1.23; bootstrap indices: 55, 1955, sample size = 151) and the mean intensity 2.24 (95% confidence limits 1.89 to 2.64; bootstrap indices: 52, 1952, sample size = 66), while K. scombercolias had a mean abundance of 0.71 (95% confidence limits 0.54 to 0.89, bootstrap indices: 56, 1956, sample size = 151) and a mean intensity of 1.81 (95% confidence limits 1.54 to 2.15; bootstrap indices: 52, 1952; sample size = 59) (table 2). Sample 2 consisted of 30 fish, 13 females and 17 males. The length of female fish ranged from 25.5 to 34.5 cm (30.86 ± 2.74) and the length of males ranged from 24.2 to 34.5 cm (31.34 ± 2.94). Again the highest prevalence was found for P. minor (86.7%, 95% confidence limits 69.27 to 96.25%). The prevalence value for K. scombri was about the same as in sample 1 (43.3%, 95% confidence limits 25.46 to 62.58%), whereas the prevalence with K. scombercolias decreased (13.3%, 95% confidence limits 37.5 to 30.73%). With reference to mean intensity and abundance values, an increase was observed in both values for P. minor, the mean intensity being 28.23 (95% confidence limits 21.58 to 35.00, bootstrap indices: 53; 1953) and the mean abundance 24.47 (95% confidence limits 18.67 to 31.50, bootstrap indices: 71; 1966). The mean intensity and abundance of both K. scombri and K. scombercolias were lower and did not differ much from values obtained in sample 1 (mean intensity = 2.69; mean abundance = 1.17 for K. scombri; mean intensity = 1.50, mean abundance = 0.20 for K. scombercolias). The total number of P. minor was 734, with 35 K. scombri and six specimens of K. scombercolias respectively (table 3).
Table 3 Prevalence, intensity and abundance of monogeneans on the gills of 30 chub mackerel Scomber japonicus with lengths ranging from 24 to 34.5 cm, examined in August 2005.
When combining both samples, the prevalence of K. scombri and K. scombercolias was found to be independent of host sex, length, weight and date of sampling (univariate ANOVAs). On the other hand for P. minor, the date of sampling and to lesser degree fish weight, had a slight significant effect on prevalence which explained 8% of the variation (F = 13.58, P = 0.000; F = 3.945, P = 0.049, univariate ANOVA). The number of parasites found in fish increased markedly from November 2004 up to August 2005 (fig. 1) indicating a seasonal variation for P. minor, where a significant correlation was found between sampling date and the abundance of infection (r2 = 0.191, F = 42.16 P = 0.001) (fig. 1 and table 4). The abundance of infection with P. minor varied significantly with sampling date and fish weight (F = 22.32, P = 0.000; F = 4.319, P = 0.039 respectively). For K. scombri males had slightly higher abundances than females (F = 6.715, P = 0.010). However, the effects of the date of sampling, fish length, weight and sex on the abundance of the parasites could explain only 20% of the variation. Pairwise comparisons of monogenean intensities resulted in a positive association between P. minor and K. scombri (rs = 0.263, P = 0.033).
Fig. 1 Variation in the abundance of monogenean infections on the gills of 181 chub mackerel Scomber japonicus from Madeira relative to sampling dates, from November 2004 to August 2005. Pseudokuhnia minor (□), Kuhnia scombri ( × ), Kuhnia scombercolias (●), Grubea cochlear (▲). Trend lines: (——, P. minor; — —, K. scombri; — ● —, K. scombercolias; — ●● —, G. cochlear).
Table 4 Correlation coefficients of four species of monogeneans on the gills of 181 chub mackerel Scomber japonicus from November 2004 to August 2005.
Estimates of the aggregation indices, variance to mean ratios (s2/x), the index of discrepancy (D) and the parameter (k) for the three most prevalent species in sample 1 (table 5), suggested that their distribution follows the negative binomial model, with good agreement between observed and expected frequencies of the parasites (chi-square = 2.09, df = 4, k = 0.72 for K. scombercolias; chi-square = 5.62, df = 5, k = 0.611 for K. scombri; chi-square = 25.68, df = 19; k = 2.79 for P. minor). These results demonstrate that parasites have a tendency to be clumped in some of their hosts. Kuhnia scombri and K. scombercolias show few hosts infected and many hosts uninfected or infected with just one parasite. Pseudokuhnia minor occurs with a high intensity of infection and is less aggregated.
Table 5 Aggregation indices, variance to mean ratio (s 2/x), index of discrepancy (D) and parameter (k) for the monogeneans Pseudokuhnia minor, Kuhnia scombri and K. scombercolias on the gills of 181 chub mackerel Scomber japonicus.
W/S: Winter/summer values.
(-) data not calculated by QP2.0 as frequency distribution of the parasites was not a negative binomial.
In contrast, in sample 2 the observed and expected frequencies of P. minor in each abundance class differed significantly meaning that this monogenean was more regularly dispersed (chi-square = 29.505, df = 18). The number of K. scombri and K. scombercolias was low in this sample, therefore the frequency distribution of the parasites was not analysed. The aggregation indices, variance to mean ratios and index of discrepancies are presented in table 5.
Discussion
Size and number of monogenean species
Six monogenean species, Pseudokuhnia minor, Kuhnia scombri, K. sprostonae, K. scombercolias, Grubea cochlear and G. australis have been reported to date from the gills and pseudobranchs of scombrid fishes (see Rohde & Watson, Reference Rohde and Watson1985a, Reference Rohde and Watsonb; Rohde Reference Rohde1986, Reference Rohde1989a, Reference Rohdeb). Considerable variability has been found in both body and hamuli length in specimens of the same species collected from different geographical areas and from different hosts (Rohde & Watson, Reference Rohde and Watson1985a, Reference Rohde and Watsonb). Values obtained for total body length and size of the large hamuli for Pseudokuhnia minor in Madeiran samples were close to those obtained for the same fish species from South African Atlantic waters (mean length = 1.72 and 1.56 mm respectively; hamuli 42 μm mean length and 38–42 μm respectively). On the other hand specimens from the Mediterranean and Spanish Atlantic coastal waters, were smaller (mean length 1.12 mm) (Rohde & Watson, Reference Rohde and Watson1985a). The total body length of K. scombri from Madeira was closer to measurements obtained for South African samples of this species (3.18 and 2.87 mm respectively), when compared with values for Mediterranean (1.1–2.4 mm) and south Atlantic waters of Brazil (1.49–3.44 mm) (Rohde & Watson, Reference Rohde and Watson1985b). Kuhnia scombercolias collected from Madeira also showed a larger body size (1.6–2.64 mm) and slightly larger hamuli (37.5–67.5 μm), while in specimens from Ecuador and Brazil the size of hamuli varied between 36 and 55 μm. The three species P. minor, K. scombri and K. scombercolias recovered from Madeiran waters, in summary, are more closely related to South African specimens than to those from Mediterranean and south-western Atlantic waters, in terms of body length and hamuli size, thus confirming the existence of geographical variation in these parameters.
Prevalence, abundance and intensity of infection
The most common monogenean species found on the gills of S. japonicus was P. minor reaching a maximum prevalence of 98.7% (n = 151). Kuhnia scombri and K. scombercolias also showed relatively high prevalences (43.7 and 39.1% respectively in sample 1), whereas Grubea cochlear and K. sprostonae were rarely observed (see table 2). The prevalence of K. scombri in other regions of the Atlantic differed considerably, with high values being obtained for the Northeast Atlantic (60–75% in S. scombrus; Somdal & Schram, Reference Somdal and Schram1992) and off continental Portugal (73.9% in S. japonicus; Rego, Reference Rego1985). Contrasting with those high prevalences, only 1.1% of chub mackerel from Argentinean waters were infected with K. scombri (Cremonte & Sardella, Reference Cremonte and Sardella1997), 11% in Brazilian waters (Alves et al., Reference Alves, Luque and Abdallah2003) and 6.6% in North Pacific waters (Pozdnyakov & Vasilenko, Reference Pozdnyakov and Vasilenko1994). These data suggest that this parasite is better adapted to colder temperate waters of the North Atlantic, with Madeira serving as an intermediate region between the colder temperate waters of the North Atlantic and the south Atlantic waters. Of course there are other factors that can mask the differences in prevalence, as for example fish size, although in the present study no correlation between prevalence and fish size was observed. Furthermore, the lengths of chub mackerel examined by Alves et al. (Reference Alves, Luque and Abdallah2003) were similar to those examined in Madeira (20–32 cm). Likewise, the mean length of fish collected from the Skagerrak (North Atlantic) was 30 cm, and prevalence with K. scombri ranged from 76 to 100% (Somdal & Schram, Reference Somdal and Schram1992). Kuhnia sprostonae which, according to Rohde & Watson (Reference Rohde and Watson1985b), is characterized by having a wide geographical distribution, encompassing Pacific waters of Australia, south Atlantic waters of Brazil, coastal waters of Ecuador and the North Sea, was not recovered by Alves et al. (Reference Alves, Luque and Abdallah2003) in coastal waters of Brazil, nor from Argentinean waters by Cremonte & Sardella (Reference Cremonte and Sardella1997). Although present in Madeira, this species was extremely rare (present work). On the other hand K. scombercolias, which is morphologically similar to K. sprostonae, but with a smaller total body and hamuli length (see Rohde & Watson, Reference Rohde and Watson1985b; Rohde, Reference Rohde1989b), was a relatively frequent species in Madeira (39.1%) and in the South Atlantic waters of Brazil (46%, Alves et al., Reference Alves, Luque and Abdallah2003). This monogenean was not recovered from North Atlantic and Argentinean waters. It may be better adapted to warmer waters, being replaced in colder waters by K. sprostonae (found in Helgoland, Plymouth, Madeira, southern Australia; while K. scombercolias found in Ecuador, northern New South Wales, Australia, Perth, Western Australia, São Paulo, Brazil and Madeira). Pseudokuhnia minor, the monogenean with the highest prevalence, intensity and abundance in Madeiran waters (present work) was surprisingly not found by previous authors in Atlantic waters off Portugal (Rego, Reference Rego1985; Rego et al., Reference Rego, Carvalho-Varela, Mendonça and Afonso-Roque1985) and the southwestern Atlantic (Cremonte & Sardella, Reference Cremonte and Sardella1997; Alves et al., Reference Alves, Luque and Abdallah2003). Nevertheless Rohde & Watson (Reference Rohde and Watson1985a) studied this monogenean from South African Atlantic waters, Mediterranean and Atlantic waters of Spain, although they did not consider prevalence and intensity. Therefore it is not possible to compare the frequency of occurrence of this monogenean in different geographical regions where it is present, in order to reach more accurate conclusions about the links of those regions and the distribution of this parasite. Apparently P. minor has a more restricted distribution, with preference for warmer waters (Madeira, present work; New South Wales in Hayward et al., Reference Hayward, Perera and Rohde1998). Finally, Grubea cochlear was rare in both S. japonicus and S. scombrus, with low prevalences ranging from 3.3% in Madeira, 6% and 18% in Brazilian waters (Rego & Santos, Reference Rego and Santos1983; Alves et al., Reference Alves, Luque and Abdallah2003), while it occurred in S. scombrus with a prevalence of 2.5% in Portuguese waters (Rego et al., Reference Rego, Carvalho-Varela, Mendonça and Afonso-Roque1985) and 3.4% in the Mediterranean (Parukhin et al., Reference Parukhin, Naidenova, Nikolaeva and Vodjanichky1971). Grubea cochlear differs significantly from the other species by its larger size (up to 10 mm in length). The low number of individuals of G. cochlear both in the present study (five individuals) and elsewhere (see Rohde, Reference Rohde1986) do not allow conclusive statements about the geographical variation of the species, although the species has been recovered from S. scombrus and S. japonicus from the Atlantic Ocean, Mediterranean and Indo-Pacific waters (Rego & Santos, Reference Rego and Santos1983; Rohde, Reference Rohde1986; Somdal & Schram, Reference Somdal and Schram1992; Alves et al., Reference Alves, Luque and Abdallah2003).
The mean abundance of K. scombri was higher in Madeiran waters (0.98–1.17) and at the austral region (0.14–2.25) (Rohde & Watson, Reference Rohde and Watson1985b) whereas in Brazil and Argentinean waters lower values were obtained (0.2 and 0.01 respectively) (Cremonte & Sardella, Reference Cremonte and Sardella1997; Alves et al., Reference Alves, Luque and Abdallah2003). The mean abundance of K. scombercolias was higher in Brazil (3.7) followed by Madeiran waters (0.20–0.71) and the austral region (0.39–1.20) (Alves et al., Reference Alves, Luque and Abdallah2003; present work). The mean abundance of P. minor reached 13.72–24.47 in Madeiran waters, and from 1.29 to 6.13 at the austral region. Taking into account that P. minor is much smaller than K. scombri and K. scombercolias, its high intensities could also mean that this species adopted a strategy of closeness to other members of the species, for defence purposes or cooperative feeding (see Bush et al., Reference Bush, Fernández, Esch and Seed2001). An increase in parasite numbers for P. minor from November to August observed in Madeiran waters, may be due to parasite reproduction occurring on the fish host when the water temperature rises. However, this assumption would need an experimental study in order to fully comprehend the life cycle of this monogenean.
In a study of the monogenean Gyrodactylus turnbulli of guppies, Poecilia reticulata, Scott (Reference Scott1987) concluded that parasites became more aggregated as their numbers increased, due to direct reproduction of parasites on individual fish. Furthermore, according to this author, the best indicator of the degree of aggregation is the s 2/x ratio. In our case this ratio was higher for P. minor and increased in the summer (14.16, see table 5). For K. scombri, relatively high values of the s 2/x ratio were obtained (see table 5) indicating that at least for these two species, the parasites were aggregated.
The prevalence and abundance of the four monogeneans, Pseudokuhnia minor, K. scombri, K. scombercolias and K. sprostonae from the scombrid Scomber australasicus in New South Wales showed similar trends to values obtained at Madeira (Hayward et al., Reference Hayward, Perera and Rohde1998). Consequently, for fish of the same length range, the highest prevalence and intensity was found for P. minor (58.7–78.2%; 1.56–6.13), followed by K. scombercolias (38.1–59%; 0.87–1.15), K. scombri (34.4–74.5%; 0.49–2.25) and K. sprostonae with the lowest values (1.6–5.5%; 0.03–0.05). A positive association found in the intensities of P. minor and K. scombri in the present work may have several explanations (Hayward et al., Reference Hayward, Perera and Rohde1998). One possible explanation could be an increase in susceptibility to infection, once one of the parasites settles on the host gills or one species could establish the microhabitat for the second species.
The occurrence of monogeneans in Scomber japonicus at Madeira, with the disparity of degrees of prevalence and intensity, may lead to the conclusion that Madeiran waters represent an interesting mixing point for the monogeneans, with species more frequent in colder waters, such as K. sprostonae, species characteristic of warmer waters, such as K. scombercolias, and species with wide geographical distributions such as K. scombri. Pseudokuhnia minor, which seems to be more host-specific, having been recorded from S. japonicus and S. australasicus only, apparently has a preference for warmer waters, with higher intensities occurring when the water temperature rises (present study). Further studies on the influence of temperature in the abundance of P. minor are in progress in order to test this assumption. Its presence in South African and Madeiran samples in S. japonicus may indicate a relationship between host populations in these areas, in contrast to its absence from Brazilian populations.
Acknowledgements
The authors would like to express their gratitude to the Portuguese Foundation of Science and Technology for financial support, through the POCTI/Plurianual Program and to two anonymous reviewers.
