Abstract
Megalocytivirus pagrus1, formerly known as Infectious spleen and kidney necrosis virus (ISKNV) poses a significant threat to Nile tilapia (Oreochromis niloticus) production worldwide. Since mathematical models of disease transmission can be helpful in evaluating and comparing the efficacy of control measures, we built a deterministic, compartmental model of ISKNV transmission within tilapia pens and then modified it to simulate several practices recommended by the government of Brazil. Our model considers net pens with densities of 10-10,000 small, susceptible fish (<25-30 grams) per m(3) pen and two routes of infection-exposure to infectious live fish or to moribund animals and carcasses-which are both likely important in these production environments. We varied the carcass decay or removal rate, simulated the rate of removal of clinically infected fish, and simulated different levels of hypothetical host resistance. With baseline parameters and a starting population of 1000, we project that almost all susceptible fish would be infected within six days and 80.8% would die by day 30. At a higher density of 10,000 fish per pen susceptible fish were infected faster, within three days, but overall mortality was slightly lower mortality (75%). Reducing carcass persistence times (e.g., simulating daily removal) had very little influence on cumulative mortality at densities >= 100 fish per pen. The effects of removing clinically ill fish switched from reducing cumulative mortality at low fish densities (<= 100 fish per pen), to consistently increasing mortality at higher densities. Lastly, when we modelled a less susceptible fish population the course of the epidemics slowed (i.e., less transmission, a longer time to deplete the susceptible hosts), resulting in lower cumulative mortality, but many infectious fish remained at the end of the simulated outbreak. Thus, we are concerned that sub clinically infected tilapia might serve as hidden carriers of ISKNV and could spark new outbreaks. Collectively, our results indicate that control will be difficult to achieve, in large part because transmission seems to be very rapid, especially in high-density net pens. We hope this modelling perspective will improve our understanding aquatic diseases and their control.