Heegard splittings of the form \(H+nK\) (Q856790)

In 1978 \textit{F. Waldhausen} conjectured that given a 3-manifold \(M\) and \(g> 0\), \(M\) cannot admit an infinite collection of non-isotopic Heegaard splitting all of genus at most \(g\). This is known to be false for toroidal manifolds, where (in some cases) infinitely many non-isotopic splittings can be obtained by twisting a given Heegaard splitting about an essential torus. On the other hand, Johanson proved Waldhausen's conjecture for \(a\)-toroidal manifolds that admit an essential surface of genus at least 2. Casson and Gordon produced the first examples of \(a\)-toroidal manifolds that admit infinitely many non-stabilized Heegaard splittings. In each example, there are only finitely many splittings of any given genus. New examples are given in the paper under review. The authors explore a conjecture of Eric Sedgwick that asserts that if a manifold admits infinitely many non-stabilized Heegaard splittings then the manifold admits an essential surface. They prove this conjecture in a special case which we now describe. Assume that \(M\) is a 3-manifold admitting infinitely many non-stabilized Heegaard splittings. If any of these splittings weakly reduces, by Casson and Gordon \(M\) admits an essential surface. We may therefore assume all the splittings are strongly irreducible. In that case, each splitting is isotopic to a normal or almost normal surface, and as such given as a Haken sum of fundamental solutions. The authors assume infinitely many surfaces can be given as \(H+ nK\), for fixed normal surfaces \(H\) and \(K\), and prove that \(K\) is essential. (We note that \(H\) and \(K\) need not be fundamental.) Sedgwick's conjecture and Waldhausen's conjecture were proved by \textit{Tao Li} in [Invent. Math. 167, No. 1, 135--177 (2007; Zbl 1109.57012) and J. Am. Math. Soc. 19, No. 3, 625--657 (2006; Zbl 1108.57015)].