Few studies have explored the effect of precipitation on the fraction of total primary production allocated belowground (fBNPP). Evidence is contradictory showing that, as precipitation increases, fBNPP can decrease, increase, or remain unaffected. We designed a greenhouse experiment where we manipulated both water availability and soil fauna composition to investigate whether the effects of water availability result from changes in plant allocation patterns or in belowground herbivory. The belowground-herbivory hypothesis postulates that fBNPP increases with increasing water availability and is based on our previous field study showing that drought suppresses nematode predators and promotes root herbivores in sub-humid grasslands. Decreases in fBNPP with increasing water availability would support the plant-response hypothesis which is based on flexible allocation patterns of grasses that reduce leaf area relative to root area during a drought. To test these hypotheses, we exposed blue grama (Bouteloua gracilis) to a gradient of root-herbivore percent abundance relative to the nematode community which also includes bacterivores, fungivores, predators and omnivores. This root-herbivore gradient had four treatments (0%, 35%, 50%, and 100%) that were nested within three levels of water availability from extreme low to intermediate and extreme high.
Results/Conclusions
The effect of water availability on fBNPP depended on the proportion of root herbivores in the nematode community (significant interaction; F6,59 = 2.59, P = 0.0271, R2 = 0.36). Increasing water availability decreased fBNPP at both the absence and natural relative abundance of root herbivores (i.e., 0% and 35% of roots feeders in the nematode community). On the other hand, fBNPP became insensitive to changes in water availability with the addition of extra root herbivores that increased its percent abundance to 50%, and the slope of this relationship ultimately became slightly positive when only root feeders were added. These results support both the plant-response and the root-herbivory hypotheses. The root-herbivory phenomenon becomes stronger and the plant-response weaker as the abundance of root herbivores increase. At root-herbivore levels similar to what was observed under extreme drought in our previous field experiments (50%), both mechanisms postulated for the plant and herbivory hypotheses compensate each other yielding a similar fBNPP across the gradient of water availability. Therefore, drought-induced changes in the abundance of root herbivores may impede the physiological plant mechanism that increases biomass allocation to roots relative to leaves in response to dry conditions.