Plant soil feedbacks (PSF) are increasingly believed to play a role in determining plant growth and community development, yet most PSF research is conducted using short-term greenhouse experiments. Further, the importance of PSF is rarely tested. Here we use a seven-year, field-based, common garden experiment to measure PSF of seven native and six non-native species common to the western USA. To explore mechanisms of PSF we measured microbial community composition using pyrosequencing to describe plant-microbial feedbacks; we measured soil penetration resistance to describe plant-physical soil feedbacks; we measured nitrogen cycling to describe plant-chemical feedbacks. Finally, we use these PSF values to predict plant cover and biomass in both experimental and naturally-occurring plant communities.
PSF values improved predictions of plant cover and biomass in four native and four non-native plant communities. Native PSF values were also correlated with plant abundance on the landscape. Native PSF values were more negative than non-native PSF values. Mechanisms driving these PSFs were not determined, but differences in bacterial and fungal community composition among soils cultivated by different plant species were consistent with the presence of plant-microbial PSF. Nitrogen cycling dynamics also differed among plant types, but were not correlated with PSF. Plants also created soils with different levels of soil penetration resistance, but these values were not consistent with PSF. While the number of species was low (seven) there was a correlation between native plant lifespan and PSF – longer lived plants had more positive PSF. Results from this relatively long-term, field-based experiment provided clear support for the role of PSF in determining native and non-native plant growth on the landscape.