This chapter considers a transaction management problem in collaborative data management. We propose a new approach corresponding to a distributed version of conservative two-phase lock (C2PL). This approach is efficient when the contention rate is high, because it effectively prevents distributed deadlock. First, we introduce two notions to express update propagation scope: family record set and update participant scope. The family record set expresses an atomic unit for distributed locking during update propagation. The update participant scope effectively reduces the search space of participants for acquiring locks. Second, we design put rules for views defined with monotonic SPJU queries in order to ensure that BX always propagates updates between the records in the same family record set. Finally, we propose an adaptive mechanism that appropriately chooses either the simple approach (2PL mode) or the family record set based approach (C2PL mode) based on the contention rate of transactions. Intensive experiments verify the effectiveness of our proposals.