Fig. 1: Experimental design and hypothesized relationships.

Left: control mesocosm communities were composed of plants and soil derived from a reference field community with a decade of shared history. Treatment communities were lacking either community-specific plant history, soil history or both. Right: Soil fauna multitrophic functioning should be positively related to plant species diversity, but this relationship builds up over time. It would therefore be stronger in the control communities and would diminish when soil history or plant history are lacking.