The fate of clades in a world of recurrent climatic change: Milankovitch oscillations and evolution

Variations in Earth's orbit with periods of 10-100 thousand years (kyr) (Milankovitch oscillations) have led to recurrent and rapid climatic shifts throughout Earth's history. These cause changes in the geographical distributions of clades, which we term orbitally forced range dynamics (ORD). The magnitude of ORD varies geographically, e.g., with latitude. Climatic shifts cause extinction, splitting, and merging of gene pools and clades. They select among individuals and clades for traits enhancing the ability to survive in situ and to establish new populations. There is also nonadaptive sorting caused by the large geographical variation in ORD, as only gene pools that are in the right place when climate shifts survive. ORD lead to sorting at many levels of genealogic inclusiveness. Clades that have survived climatic shifts during at least one entire period of the longest significant Milankovitch oscillations (100 kyr), we name beta-clades. The products of more recent cladogenesis are alpha-clades, which are always nested within a beta-clade. We conclude that ORD may promote alpha-clade formation but curb rates of beta-clade formation. In areas with little ORD, where gene pools persist without going extinct or merging, clade splits and divergence may accumulate leading to high rates of beta-clade formation and beta-anagenesis (evolutionary change persisting > 100 kyr). High ORD should lead to low numbers of beta-clades, beta-clades with low levels of spatial genetic divergence, little geographical subdivision and large ranges, organisms with high vagility and low specialization, high proportions of beta-clades formed by polyploidization, and little beta-anagenesis. We predict global and interregional geographic patterns in these variables caused by differential ORD. Thus, ORD potentially explains a wide array of patterns, suggesting ORD as a fundamental factor in evolution. The vulnerability of biotas to many human activities should vary with the magnitude of ORD.