Families of Lagrangian fibrations on hyperkähler manifolds (Q2452077)

Let \(X\) be a compact hyperkähler manifold of dimension \(n\). A holomorphic line bundle \(L\) on \(X\) is called Lagrangian if it is semiample and \((c_1 (L))^n =0 \in H^{n,n} (X, \mathbb{C})\). The pair \((X, L)\) is called a holomorphic Lagrangian fibration. The article under review proves that given a compact simple hyperkähler manifold with second Betti number \(b_2 \geq 7\), there exists only finitely many deformation types of holomorphic Lagrangian fibrations. It is well known (due to Huybrechts) that for a fixed compact manifold there are at most finitely many deformation types of hyperkähler structures on it. The main result of the article implies that there exists only finitely many deformation types of the Lagrangian fibration pair \((X,L)\) under the assumptions that \(X\) is simple hyperkähler and \(b_2 (X) \geq 7\). The proof of the main result of the article relies on a density result about lattices, and also a study of the orbits of \(\Gamma_I\) on \(S\), where \(S \subset H^2 (M, \mathbb{Z})\) is the set of classes which are parabolic and primitive, and \(\Gamma_I\) is the monodromy group on \(H^2 (M, \mathbb{Z})\) fixing the complex structure \(I\) on \(M\). An interesting consequence of the main result is that an irreducible holomorphic symplectic manifold is not Kobayashi hyperbolic (i.e., it contains a nontrivial entire curve) if it is a deformation of a Hilbert scheme of points on \(K3\), a deformation of generalized Kummer variety, or a deformation of one of two examples by O'Grady.