Horizontal versus vertical transmission of parasites in a stochastic spatial model (Q5926101): Difference between revisions

The paper deals with a stochastic spatial model for the study of horizontal versus vertical transmission of parasites in a population. The starting point is a deterministic nonspatial model in which healthy individuals always give birth to healthy children and infected give birth to infected individuals. Healthy individuals get infected by infected ones. On the other hand, the spatial stochastic model proposed by the author has the following features: the space is assumed to be of the lattice type \(Z^{d},\) \(Z\) being the set of integers and each site \(x\) on the lattice may be in one of three states: 0, 1 and 2. The state 0 is an empty site while state 1 is a healthy one and 2 an infected site. Assuming that \(n_{1}(x,j)\) and \(n_{2}(x,j)\) are the number of nearest neighbours of \(x\) (among the \(2d\) nearest neighbours of \(x\)) that are in states 1 and 2, respectively, and that the state at a given \(x\) evolves with transition rates for the transitions \(0\rightarrow 1,\;0\rightarrow 2,\;1\rightarrow 2,\;1\rightarrow 0\) and \(2\rightarrow 0\), respectively \(\lambda_{1}n_{1}(x,j)+\gamma n_{2}(x,j),\;\lambda_{2}n_{2}(x,j),\;\beta n_{2}(x,j),\;1,\;1,\) the main results are best summarized by the following two theorems. \textbf{Theorem 1.} (a) Consider the spatial stochastic model on \(Z\) with \(\lambda_{2} < \lambda_{c}\) and \(\gamma = 0.\) Then for any \(\beta > 0\) the infected individuals die out. That is, starting with a 2 at the origin and an arbitrary configuration of 0s and 1s elsewhere, the 2s will die out after a finite time with probability one. (b) Consider the spatial stochastic model on \(Z^{d},\;d\geq 1,\) with \(\lambda_{2}=0\) and \(\beta > \lambda_{c}.\) If \(\gamma\) is large enough, then there is a positive probability that the 2s will survive at all times. \textbf{Theorem 2.} Consider the spatial stochastic model on \(Z^{d},\;d\geq 1,\) starting with an infected individual at the origin and an arbitrary configuration of 0s and 1s elsewhere. (a) If \(\min(\lambda_{2},\beta) > \lambda_{c},\) for any \(\gamma \geq 0\) and in any \(d\geq 1,\) there may be an epidemic. That is, there is a positive probability that there will be 2s at all times. (b) If \(\min(\lambda_{2},\beta) > \lambda_{c},\;\gamma =0\) and \(d=1,\) then the 1s may die out. That is, there is a positive probabiliy that for any site \(x\) there is a finite time after which there will never be a 1 at \(x\). (c) If \(\lambda_{1} > \lambda_{2},\) for any \(\gamma \geq 0\) and in any \(d\geq 1,\) there is a \(\beta_{c}> 0\) (depending on \(\lambda_{1}\) and \(\lambda_{2}\)) such that if \(\beta < \beta_{c},\) then the infected individuals die out. These theorems are elegantly presented with a discussion in the introductory section and the proofs are relegated to the concluding part of the paper.