EVOLUTION OF SILICA ACCUMULATION STRATEGIES IN LAND PLANTS AND THE ADAPTIVE VALUE OF PHYTOLITHS

The ability to take up silica in the form of monosilicic acid from soil water and precipitate it as hydrated silica inside and around cells is widespread among land plants, occurring in virtually all major living clades. As a result, extant vegetation, with ecosystems dominated by phytolith-rich grasses as a prime example, contribute vitally to the terrestrial aspects of the silica cycle, including weathering of silicates and retention of easily mobilized (bio)silica on land. In modern plants, phytoliths are thought to function as mechanical support—an economic alternative to lignin—and defending against herbivores, by limiting nutrient access and abrading herbivore mouthparts. It has commonly been assumed that high (active) phytolith accumulation evolved early on, as plants transitioned to life on land, and that subsequent variation in phytolith production reflects adaptive evolution to serve one or more of the functions observed today. For example, grasses are thought to have evolved their high silica levels through antagonistic coevolution with large, mammalian grazers during the Cenozoic. However, these hypotheses have not been tested in detail.

We evaluate the pattern of evolution of active phytolith accumulation within the land plant clade, focusing on vascular plants and with grasses as a special case. We do so by conducting ancestral character state reconstructions of silica content on time-calibrated land plant and grass phylogenies. We then compare our results to paleontological evidence for the timing of the ‘demand’ for the hypothesized function (support, herbivore defense). Our phylogenetic mapping analyses indicates that high silica deposition evolved multiple times independently within vascular plants, rather than being ancestral in land plants. However, a clear temporal link between these events and putative functional demands is still missing. Moreover, comparative phylogenetic analysis does not support the hypothesis that Cenozoic coevolution between grasses and herbivorous ungulates led to increased accumulation of phytoliths in grasses. Taken together, our results indicate that high phytolith accumulation was not widespread among early land plants, suggesting that their important role in the terrestrial silica pump may have emerged more gradually than previously thought.