Do silica-based defences drive plant-herbivore dynamics

A large enclosure experiment involving was established in early 2009 with replicated treatments of 6 different vole densities. The impact on Silicon (Si) levels in the voles’ major over-winter food species, the grass Deschampsia caespitosae was measured until spring the following year and demonstrated that a) induction of Si was proportional to vole densities and positively related to damage levels; b) there is a threshold to induction in that no significant induction occurs unless vole densities exceed ~200voles/ha; c) there is a seasonal component to both Si content and plant responses, with maximum induction occurring in response to damage in spring when plants are actively growing; d) there is a lag of ~9months to maximum induction and the increases in Si are similar to those causing significant impacts on vole performance in lab experiments.
We also manipulated population densities of field voles on approximately 1 ha grassland plots in Northern England over one year. Subsequently, we tested for the impact of past density on vole life history traits in spring, and whether these effects are driven by the induction Si. We predicted that female voles on sites with previously high density, and consequently high grass Si concentrations, would be slower to gain weight and so would enter reproduction later in spring than conspecifics on sites with previously low density. Several months after commencing the population density manipulations, leaf silicon concentrations diverged and were on average 22% lower on plots where vole densities had been reduced. However, these effects did not persist beyond the period of the density manipulations. Contrary to our predictions, for every additional 100 voles/ha in the previous spring and early summer, vole mass in spring increased by 2.4g and the mean date for the onset of spring reproduction was advanced by ~20 days. These findings show that grazing by field voles can induce increased silicon defences in grasses on a landscape-scale, but that the effect is relatively transient and thus unlikely to drive the multi-annual population cycles of field voles in Northern England. Unusually this study found that reducing vole densities had a negative rather than the expected positive impact on vole performance the following spring, suggesting that negative density dependent processes may only operate once vole populations have exceeded a certain density threshold.
In addition to this field work, we have carried out some initial modelling work, based on our measured magnitude and time-scale of silica-based defence induction, which did provide a quantitatively robust explanation for population cycles in voles (Reynolds et al 2012).
We have also collaborated with researchers in Norway to investigate Si -herbivore interactions in the sub-arctic (Soininen et al 2012), developed a new analytical method for Si measurement (Reidinger et al 2012) and examined the impact of Si defences on plant competition (Garbuzov et al 2011).