Host and environmental risk factors associated with Cryptosporidium scophthalmi (Apicomplexa) infection in cultured turbot, Psetta maxima (L.) (Pisces, Teleostei)

Abstract

An epidemiological cohort study of Cryptosporidium scophthalmi in cultured turbot Psetta maxima L. of Northwestern Spain was conducted along a four-year period. Four different ongrowing cohorts were monitored monthly from introduction into the ongrowing tanks (10–50 g) until reaching market size (400–1400 g). The association of host and environmental factors with five categories of parasite abundance was assessed using a multivariable regression framework. Epidemiological factors assessed here were water temperature, weight, length, month of collection, season, age, origin, condition factor, water filtration, and status to the myxozoan Enteromyxum scophthalmi infection. E. scophthalmi was included into the analysis because it targets the same organ than C. scophthalmi and it was prevalent in the studied population. The multivariable analysis demonstrated the statistically significant association between several factors and parasite abundance. C. scophthalmi abundance was associated ( P < 0.05) with age, condition factor, season, and status to E. scophthalmi infection. Young animals, with poor condition factor, during spring or summer, and not infected with the myxozoan were most likely to be highly infected by C. scophthalmi. Inclusion of these four variables significantly ( P < 0.05) improved the model, compared to the model that did not include any of these epidemiological factors. Increasing levels of C. scophthalmi abundance were associated ( P < 0.01) with higher severity of C. scophthalmi-compatible lesions. The frequency of distribution of C. scophthalmi abundance was clearly right-skewed and fitted a negative binomial distribution, whereas the intensity of infection fitted a Poisson distribution. The quantification of the variance-to-mean ratio stratified by age demonstrated overdispersion for 8–16 months old fish, although this bivariate association is likely affected by several other factors, as suggested by the results of the multivariable analysis. The negative relation between C. scophthalmi abundance and status to E. scophthalmi infection suggests differences in the transmission, onset, and course of both infections. The coarse filtration used in some cohorts did not significantly affect the levels of infection. C. scophthalmi was probably introduced into the ongrowing tanks mainly with carrier fish, though the involvement of infective oocysts from the water supply cannot be disregarded. Infection prevalence and mean intensity decreased with fish age and a seasonal distribution was found. Results presented here will help to understand the epidemiology of C. scophthalmi in turbot, to estimate the expected levels of infection associated with presence or absence of epidemiological factors, and to quantify the impact that the disease may have on susceptible turbot populations. The multivariable model used here is more powerful than the visual inspection of graphics for exploring associations in cooperative processes and can be easily extended to the assessment of epidemiological associations in other population and parasitic diseases.