Let E be an elliptic curve over Q and let p ≥ 5 be a prime of good supersingular reduction for E. Let K be an imaginary quadratic field satisfying a modified "Heegner hypothesis" in which p splits, write K∞ for the anticyclotomic Zp-extension of K and let Λ denote the Iwasawa algebra of K∞/K. By extending to the supersingular case the Λ-adic Kolyvagin method originally developed by Bertolini in the ordinary setting, we prove that Kobayashi's plus/minus p-primary Selmer groups of E over K∞ have corank 1 over Λ. As an application, when all the primes dividing the conductor of E split in K, we combine our main theorem with results of Ç iperiani and of Iovita-Pollack and obtain a "big O" formula for the Zp-corank of the p-primary Selmer groups of E over the finite layers of K∞/K that represents the supersingular counterpart of a well-known result for ordinary primes.2010 Mathematics Subject Classification. 11G05, 11R23.Heegner points that satisfy a kind of trace-compatibility relation of the sort needed to study restricted Selmer groups as in [2].As an application, combining our main theorem with results of Ç iperiani and of Iovita-Pollack, we obtain the following "big O" formula (Theorem 6.1) for the Z p -corank of the p-primary Selmer groups of E over the finite layers of K ∞ /K.This is the supersingular analogue of a well-known result for ordinary primes; in fact, Theorem 1.5 proves [3, Conjecture 2.1] when p is supersingular and K ∞ is the anticyclotomic Z p -extension of K. Here we would like to emphasize that, due to the failure of Mazur's control theorem in its "classical" formulation, knowledge of the Λ-corank of Sel p ∞ (E/K ∞ ) as provided by [7, Theorem 3.1] is not sufficient to yield the growth result described in Theorem 1.5 (see Remark 6.3 for more details). Moreover, assuming the finiteness of the p-primary Shafarevich-Tate group of E over K m for m ≫ 0, standard relations between Mordell-Weil, Selmer and Shafarevich-Tate groups of elliptic curves over number fields lead (at least when D = 1) to a formula (Corollary 6.5) for the growth of the rank of E(K m ).As already mentioned, the techniques employed in this paper are close to those of Bertolini [2]. Similar results could presumably be obtained via different approaches, for example by adapting the arguments of Ç iperiani in [7] (which rely on the techniques developed in [8]) or, following ), by using Λ-adic Kolyvagin systems as is done by Howard in [15]. In particular, we hope that extending the point of view of [15] to the supersingular setting would lead to an understanding of the torsion submodule of X ± ∞ : we plan to come back to these issues in a future project.Acknowledgements. It is a pleasure to thank Mirela Ç iperiani for helpful discussions and comments on some of the topics of this paper. We would also like to thank Christophe Cornut for useful correspondence on his joint work with Vinayak Vatsal.
Anticyclotomic Iwasawa algebrasWe briefly review the definition of the anticyclotomic Z p -extension K ∞ of K and then introduce the Iwasawa a...