DYNAMICS OF A FELINE RETROVIRUS (FELV) IN HOST POPULATIONS WITH VARIABLE SPATIAL STRUCTURE

The predictions of epidemic models are remarkably affected by the unde rlying assumptions concerning host population dynamics and the relatio n between host density and disease transmission. Furthermore, hypothes es underlying distinct models are rarely tested. Domestic cats (Felis catus) can be used to compare models and test their predictions, becau se cat populations show variable spatial structure that probably resul ts in variability in the relation between density and disease transmis sion. Cat populations also exhibit various dynamics. We compare four e pidemiological models of Feline Leukaemia Virus (FeLV). We use two dif ferent incidence terms, i.e. proportionate mixing and pseudo-mass acti on. Population dynamics are modelled as logistic or exponential growth . Compared with proportionate mixing, mass action incidence with logis tic growth results in a threshold population size under which the viru s cannot persist in the population. Exponential growth of host populat ions results in systems where FeLV persistence at a steady prevalence and depression of host population growth are biologically unlikely to occur. Predictions of our models account for presently available data on FeLV dynamics in various populations of cats. Thus, host population dynamics and spatial structure can be determinant parameters in paras ite transmission, host population depression, and disease control.