The reciprocal interaction of angiosperm evolution and tetrapod herbivory

We have interpreted the history of angiosperm-herbivorous tetrapod interactions based on the fossil record of the two groups. This history can be divided conveniently into four stages. Stage 1, lasting ∼ 40 Ma (Barremian-Campanian), was characterized by diverse large herbivores, few species of small herbivores, and r-selected angiosperms. The dominant interaction between herbivores and angiosperms during this stage was generalized herbivory. During Stage 2 (∼ 10 Ma, Campanian - Maestrichtian), small herbivores increased in diversity; larger angiosperms and larger angiosperm diaspores became more common. Generalized herbivory was still the dominant interaction in this stage, but frugivory/dispersal of angiosperm diaspores by small herbivores became more important. In Stage 3 (∼ 25 Ma, Paleocene - mid-Eocene) large angiosperms and large angiosperm diaspores were diverse; large herbivores were initially absent, later low in diversity. Frugivory/dispersal was common during this stage, generalized herbivory much less so. During Stage 4 (∼ 30 Ma, Oligocene - Recent), the relative importance of large vs. small herbivores and large vs. small angiosperms has varied by community, as has the relative importance of generalized herbivory vs. frugivory/dispersal. We infer the following evolutionary effects of angiosperms and tetrapods on each other. During Stage 1 generalized herbivory/disturbance by dinosaurs favored angiosperms that remained relatively small and r-selected. Increasing abundance and geographic spread of these r-selected angiosperms fueled the Late Cretaceous diversification of low-browsing ornithopod dinosaurs. The rarity of angiosperms with large diaspores provided little resource for a radiation of small herbivores. The low diversity of small herbivores created few opportunities for the evolution of small herbivore dispersal syndromes among angiosperms. The stability of angiosperm-vertebrate herbivore interactions during Stage 1 suggests this system of ecological relationships had internally self-reinforcing properties. The modest radiation of small herbivores during Stage 2 may indicate increased frugivory/dispersal. More common large angiosperm axes and diaspores show some angiosperm lines were becoming more K-selected. These imply a modification of the stable system formed during Stage 1. Extinction of all large herbivores at the K/T boundary destroyed Stage 1 ecological interactions and changed selective pressures on angiosperms. In the early Paleocene, generalized herbivory/herbivore disturbance was absent or uncommon. Denser vegetation, increased competition between plants, and increased seed dispersal/predation by small animals resulted in selection for larger seeds, diaspores and sporophytes. The distribution of resources in closed angiosperm vegetation permitted the radiation of small, arboreal, frugivorous birds and mammals, which in turn were important to the success of angiosperms with large, animal-dispersed seeds. Spread of arborescent angiosperm vegetation reduced resources and evolutionary opportunities available to larger mammalian herbivores, retarding their diversification, and in turn perpetuating low levels of generalized herbivory/disturbance. Stage 3 was another period during which angiosperm-vertebrate herbivore interactions were stabilized by self-reinforcing behavior of the system. During Stage 4 increased seasonality began to favor shorter life cycles among angiosperms; herbaceous plants diversified and spread. Open, high-productivity vegetation was better exploited by large herbivores, which diversified, increasing generalized herbivory/disturbance that in turn created more habitat for r-selected angiosperms.