Dynamics of acyclic interval maps (Q2431375)

The paper under review studies acyclic interval maps. A map \(f:I\to I\) is acyclic if the only periodic points are fixed points. It follows from the famous Sharkovskii theorem that \(f\) is acyclic if and only if it has no 2-cycles. \textit{W. A. Coppel} [``The solution of equations by iteration'', Proc. Camb. Philos. Soc. 51, 41--43 (1955; Zbl 0064.12303)] and \textit{A. N. Sharkovskij} [``On cycles and the structure of a continuous mapping'' (Russian), Ukr. Mat. Zh. 17, No. 3, 104--111 (1965; Zbl 0168.20806)] proved that for an acyclic map \(f\) we have the following: (1) \(Orb(x)=\{f^{n}(x):n=0,1,2,\dots\}\) converges to a point for each \(x\in I\); and (2)\(f(x)\leq x\) iff \(f^n(x)\leq x\) for every \(n>0\). \textit{H. R. Thieme} [Mathematics in population biology. Princeton, NJ: Princeton University Press (2003; Zbl 1054.92042)] proved that an acyclic map \(f:(0,\infty)\to(0,\infty)\) with a unique fixed point \(p\), and \(x_1<p<x_2\) with the property that \(f(x_1)>x_1\) and \(f(x_2)<x_2\) is globally stable (i.e., \(\lim_{n\to\infty}Orb(x)=p\) for every \(x\)). \textit{P. Cull} studied acyclicity in the context of population models and showed that on \((0,\infty)\) it is equivalent to global stability [``Local and global stability for population models'', Biol. Cybern. 54, 141--149 (1986; Zbl 0607.92018)]. He also proved some sufficient conditions on the derivatives \(f',f''\) and \(f'''\) in order for \(f\) to be globally stable. The author uses the convergence result due to Coppel and Sharkovskii to extend the results of Thieme and Coppel to acyclic maps that have several fixed points. He also proves various other sufficient conditions for acyclicity of interval maps and suggests nine related problems for future research. Reviewer's remark: One must be careful applying the results of the paper. By the time the reviewer received the present assignment there had been known concerns about validity of some of the results. For example, according to H. P. Bruin [Mathematical Reviews MR2558488] Theorem 6.2 is incorrect. Some related results, that are not cited in the paper can be found in [\textit{G. Di Lena, M. Martelli} and \textit{B. Messano}, ``Global stability and plus-global stability. An application to forward neural networks'', Rend. Ist. Mat. Univ. Trieste 42, 1--18 (2010; Zbl 1243.37021)].