Coexistence in the unstirred chemostat (Q1126629): Difference between revisions

This paper is concerned with the coexistence of positive steady-state solutions for the system of parabolic equations \[ {\partial S\over \partial t} =d_0\Delta S-\nabla \varphi \cdot \nabla S- \sum^n_{i=1} m_if_i(S,u) u_i, \] \[ {\partial u_i \over \partial t} =d_i \Delta u_i-\nabla \varphi \cdot \nabla u_i+ m_if_i(S,u)u_i, \quad i=1, \dots,n,\;x\in \Omega, \] \[ {\partial S \over \partial \nu} +\gamma_0 S=S^0, \quad {\partial u_i \over \partial \nu}+ \gamma_i u_i=0, \quad i=1, \dots, n,\;x\in \partial \Omega \] Our primary interest is to investigate models for exploitative competition where \(S\) represents a resource for which the \(n\) species, \(u_i\), compete, the parameter \(m_i\) represents the ideal growth rate for the species \(u_i\), \(d_i\) determines a diffusion rate for species \(u_i\), and \(\varphi\) is a potential for fluid flow within the domain. Our results will establish coexistence without the assumption of equal diffusion rates, and will establish some local estimates on the size of the coexistence region near a bifurcation point in the parameter space. Our proofs rely on standard maximum principle and bifurcation techniques. Questions regarding stability of solutions and conservation of biomass remain open. We also include a section with several interesting numerical computations. In particular, we confirm the conclusion in \textit{J. W.-H. So} and \textit{P. Waltman} [Appl. Math. Comput. 32, No. 2/3, 169-183 (1989; Zbl 0691.92018)]that the coexistence region is thin. However, we see that if we adopt a more practical definition of coexistence, such as requiring no species to lose over the half of its population over the next several hundred time units, then the coexistence region is relatively large.