Laminar free hyperbolic 3-manifolds (Q878319)

A lamination on a 3-manifold is a foliation of a closed subset. A lamination is essential if it has no sphere leaves, no tori leaves bounding solid tori, the complement of the lamination is irreducible and the leaves in the boundary of the complement are incompressible and end incompressible in their respective complementary components (see \textit{D. Gabai} and \textit{U. Oertel} [Ann. Math. (2) 130, No.~1, 41--73 (1989; Zbl 0685.57007)]). The author answers in the negative the question of D. Gabai and U. Oertel: does every closed hyperbolic 3-manifold admit an essential lamination? Let \(M\) be a torus bundle over the circle with the monodromy induced by the matrix \[ A = \begin{pmatrix} m&-1\cr 1&0 \end{pmatrix}, \] let \(O\) be the image in the torus \(T^2\) of the origin of \(\mathbb R^2\) in the universal covering map, and \(\delta\) be the orbit of the suspension flow on \(M\) through \(O\). Then \(\delta\) is a closed orbit intersecting \(T^2\) once. Theorem. Let \(M_{p/q}\) be the manifold obtained from \(M\) as a result of \((q, p)\) Dehn surgery on \(\delta\). If \(m \leq 4\) and \(| p - 2q| = 1\), then the manifold \(M_{p/q}\) does not admit essential laminations. In particular, \(M_{p/q}\) admits no Reebless foliation. The main tool of the proof is studying group actions on trees. The author shows that the group \(\pi_1(M_{p/q})\) admits no non-trivial action on any topological tree. This contradicts the existence of an essential lamination on \(M_{p/q}\).