Published online by Cambridge University Press: 13 June 2016
Theoretical models predict that plant interactions with free-living soil microbes, pathogens and fungal symbionts are regulated by nutrient availability. Working along a steep natural gradient of soil fertility in a Costa Rican tropical dry forest, we examined how soil nutrients affect plant–microbe interactions using two complementary approaches. First, we measured mycorrhizal colonization of roots and soil P availability in 18 permanent plots spanning the soil fertility gradient. We measured root production, root colonization by mycorrhizal fungi, phosphatase activity and Bray P in each of 144 soil cores. Next, in a full-factorial manipulation of soil type and microbial community origin, tree seedlings of Albizia guachapele and Swietenia macrophylla were grown in sterilized high-, intermediate- and low-fertility soils paired with microbial inoculum from each soil type. Seedling growth, biomass allocation and root colonization by mycorrhizas were quantified after 2 mo. In the field, root colonization by mycorrhizal fungi was unrelated to soil phosphorus across a five-fold gradient of P availability. In the shadehouse, inoculation with soil microbes had either neutral or positive effects on plant growth, suggesting that positive effects of mycorrhizal symbionts outweighed negative effects of soil pathogens. The presence of soil microbes had a greater effect on plant biomass than variation in soil nutrient concentrations (although both effects were modest), and plant responses to mycorrhizal inoculation were not dependent on soil nutrients. Taken together, our results emphasize that soil microbial communities can influence plant growth and morphology independently of soil fertility.