Uniform persistence in nonautonomous delay differential Kolmogorov-type population models (Q1335213)

The paper presents a result of uniform persistence for a system of two time-dependent delay differential equations with a bounded delay: \[ dx/dt= x(t) f(t,x_ t, y_ t); \qquad dy/dt= g(t, x_ t, y_ t). \] Both competition and predator-prey conditions are considered. Uniform persistence is meant in the sense that there exists a compact region \(K\) in the interior of \(\mathbb{R}^ 2_ +\), such that every trajectory eventually enters \(K\). The method used here can be referred to as a direct method. It is based on the use of differential inequalities or, even, ordinary differential equations whose solutions dominate in some sense those of the delay system. Clearly, the framework adopted here precludes the use of dynamical system analysis [as it is treated in \textit{J. K. Hale} and \textit{P. Waltman}, SIAM J. Math. Anal. 20, No. 2, 388-395 (1989; Zbl 0692.34053)] as well as the analysis by means of a Lyapunov functional [as in \textit{T. Burton} and \textit{V. Hutson}, J. Math. Anal. Appl. 137, No. 1, 240-263 (1989; Zbl 0677.92016)]. Examples provided show the applicability of the results. The examples indicate dependence upon the delays in contrast with related results [notably by W. Wang and Zh. Ma (1991), or Y. Cao et al. (1993), quoted in the references] where delays are shown to be harmless. This seems to be the price to pay when comparing the delay system to an ordinary one, which is confirmed by the fact that the conditions given here are the best ones in the restricted case where the system is autonomous and without delays.