The survival probability for a class of multitype subcritical branching processes in random environment (Q2191950)

Consider a multitype branching process where the reproduction law of the particles (at any time step) depends on the status of an external environment. The random environment evolves as a sequence of i.i.d. variables with values in the set of admissible reproduction laws (branching measures). Putting a set of (merely analytical) conditions on the mean value matrices of the reproduction laws for the particles which smooth the processes' behaviour, the authors investigate the finite (large times) behaviour of the multitype branching process. It is proved for subcritical processes starting with exactly one particle of each type at time zero that the total population size at time \(n\) is of order \(\lambda^n\cdot n^{-1/2}\), where \(\lambda\) is defined as a limit in terms of the products of the mean value matrices of the reproduction laws.