Bifurcation of the separatrix skeleton in some 1-parameter families of planar vector fields (Q2343526)

This work concerns planar differential systems of the form \[ \dot{x} \, = \, y^3-x^{2k+1}, \quad \dot{y} \, = \, -x+m y^{4k+1}, \] which depend on a real parameter \(m\), for arbitrarily but a fixed integer value of \(k \geq 1\). The vector field associated to this system is denoted by \(X_m^k\). This article is inspired by the results presented in [\textit{J. D. García-Saldaña} et al., J. Math. Anal. Appl. 413, No. 1, 321--342 (2014; Zbl 1308.34048)] where systems of the aforementioned form with \(k=1\) are considered. In that work [loc. cit.], the authors conjecture the existence of exactly one value of the parameter \(m\) for which the system presents a \(2\)-saddle cycle. If this is true, then the bifurcation diagram of the global phase portraits is given including the configuration of limit cycles. In this case, at most one limit cycle can exist. The authors also bound the values of \(m\) for which this limit cycle exists. In the present work under review, this conjecture is proved. Besides, the author provides the bifurcation diagram of global phase portraits up to configuration of limit cycles for \(k \geq 2\). One of the main tools used in the proof is a change of variables which transforms \(X^k_m\) into a semi-complete family of indefinitely rotated vector fields. Moreover, the proofs provided in the paper illustrate a whole arsenal of methods classically used in the field and, thus, it is very interesting for a wide range of readers.