Homotopy minimal periods of holomorphic maps on surfaces (Q1032750)

Given a self-map \(f: M\to M\), the set \(\text{Per}(f)\) of periods of \(f\) is a subset of the natural numbers. A natural number \(n\) is in \(\text{Per}(f)\) if and only if \(f\) has periodic points of minimal period \(n\). The set \(\text{Hper}(f)\) of homotopy minimal periods of \(f\) is defined to be the intersection \(\cap_{g\sim f} Per(g)\), where \(g\) ranges over the maps homotopic to the given map \(f\). Such a set is not only a homotopy invariant from the algebraic topology point of view but also includes some characteristics of a map in the sense of dynamical systems. When \(f\) is a holomorphic map on a Riemann surface, the authors extend this concept to the following ones: \(\text{Hper}^{\mathcal{H}}(f)\) and \(\text{Hper}_{\mathcal{H}}(f)\), which are both the intersection \(\cap_{g} Per(g)\), where \(g\) ranges over the holomorphic maps homotopic to \(f\) in the former case and ranges over the holomorphic maps being connected to \(f\) by a homotopy \(\{h_t\}\) such that every \(h_t\) is holomorphic in the latter case. By definition, \(\text{Hper}(f)\subseteq\text{Hper}^{\mathcal{H}}(f)\subseteq \text{Hper}_{\mathcal{H}}(f)\). The authors try to determine these sets for holomorphic maps on Riemann surfaces. If \(f\) is a holomorphic map on \(S^2\), \(\text{Hper}^{\mathcal{H}}(f)\) and \(\text{Hper}_{\mathcal{H}}(f)\) coincide with each other, but are very different from \(\text{Hper}(f)=\{ 1\}\). This is not surprising because holomorphic maps are more restricted and should have more periodic points. On the other hand, these three sets turn out to be the same for holomorphic maps on the torus, which was already proved by \textit{Ll. Alsedà, S. Baldwin, J. Llibre, R. Swanson} and \textit{W. Szlenk} [in: McCord, Christopher K. (ed.), Nielsen theory and dynamical systems. AMS- IMS-SIAM summer research conference, June 20--26, 1992, South Hadley, MA (USA), supported by the National Science Foundation. Providence, RI: American Mathematical Society. Contemp. Math. 152, 1--7 (1993; Zbl 0804.54032)]. Actually, the idea of homotopy minimal periods originates from that paper. In the case of surfaces of genus greater than one, it is shown that \(\text{Per}(f) = \text{Hper}^{\mathcal{H}}(f) = \text{Hper}_{\mathcal{H}}(f)\) for each holomorphic map. Some more descriptions for these sets are given. In fact, any holomorphic map on a Riemann surface of genus greater than one must be a periodic one. The set \(\text{Hper}(f)\) of homotopy minimal periods for a periodic map \(f\) on a closed surface was addressed by the second author and the reviewer in [Topology Appl. 156, No.~15, 2527-2536 (2009; Zbl 1176.55002)].