Abstract: It is crucial to study the nature and origin of the mantle heterogeneity for understanding the solid Earth evolution. In this study we evaluated the mantle heterogeneity through oceanic basalts, e.g., ocean island basalts (OIBs) and mid-ocean ridge basalts (MORBs), and discussed the reason for insufficient understandings of the mantle heterogeneity. Studies in the past three decades show that the enriched mantle end-members include HIMU (('μ'=²³⁸U/²⁰⁴Pb) (low ⁸⁷Sr/⁸⁶Sr and high ²⁰⁶Pb/²⁰⁴Pb)), EMI (medium ⁸⁷Sr/⁸⁶Sr, low ²⁰⁶Pb/²⁰⁴Pb), EMII (high ⁸⁷Sr/⁸⁶Sr, medium ²⁰⁶Pb/²⁰⁴Pb), FOZO, and depleted mantle end-member of DMM (including Indian-type and Pacific-type). Origins of enriched mantle end-members are supposed to be related to plate tectonics, such as oceanic plate subduction, continent rifting and detachment. However, the origin of compositional differences between the end-members remains highly controversial. We consider that studies on depleted mantle end-members are also important for understanding the diversity of the compositions of mantle end-members. Debates on the origin of mantle end-members are mainly caused by lack of knowledge on the role of plate subduction in driving geochemical cycling and the mantle depletion in the early stage of melting of the Earth. We, therefore, suggest the following study aspects: Firstly, phase transition and the fate of subducted oceanic plate in the lower mantle; Secondly, test if detached continental crust can be mixed into the upper mantle due to continent rifting; Thirdly, the role of carbonate in the genesis of alkali basalts; Fourthly, the geochemical fractionation of subducting oceanic plate in the mantle.