Soil microbial community influences plant diversity–productivity relationship in an environmental context dependent way
Soil pathogens are thought to drive small–scale positive plant diversity–productivity relationships by suppressing the productivity of low–diversity assemblages. However, arbuscular mycorrhizal fungi (AMF) can boost plant productivity and counteract negative effects of pathogens. Moreover, nutrient additions can reduce the positive effects of AMF, and increase the negative effects of pathogens. Along a soil nutrient gradient, there might be a switchover point where the relative strength of positive mycorrhizal and negative pathogenic effects changes. Therefore, we hypothesize that nutrient availability can change soil community composition that in turn leads to an altered plant diversity–productivity relationship. Specifically, we hypothesize that in nutrient poor soils, positive mycorrhizal effects dominate and therefore the presence of soil microbes enhances productivity overall. Conversely, in nutrient rich soils, we hypothesize negative pathogenic effects dominate, and thus productivity will be depressed by soil micorbes in monocultures. To test this hypothesis, we assembled plant communities at four diversity levels (1, 2, 4 and 8 species) with seedlings of eight woody species. Each composition was grown in four different ‘environments’: low and high levels of nutrient availability were crossed with sterilized and unsterilized soil treatments. After 15 months of growth, plants were harvested to measure productivity.
We found that soil nutrient and microbial communities concurrently influenced the slopes of plant diversity–productivity relationship (nutrient treatment × soil treatment × plant diversity interaction; F1,40 ＝ 5.75, P＝ 0.021). Under the low nutrient conditions, productivity marginally increased with diversity in both sterilized (F1, 39 ＝ 3.15, P ＝ 0.084) and unsterilized (F1, 39 ＝ 3.30, P ＝ 0.077) soils. Under the high nutrient conditions, significant positive diversity–productivity relationship existed in sterilized soils (F1, 39 ＝ 5.22, P ＝ 0.028), whereas this relationship disappeared in unsterilized soils (F1,37 ＝ 0.003, P ＝ 0.958). Though we did not detect positive soil microbial effects in nutrient poor soils, our study did show that soil communities suppressed productivity only in nutrient rich soils and surprisingly in high–diversity assemblages, leading to non–significant plant diversity–productivity relationship. These findings provide new insights that soil microbial community does not act as driver of plant diversity–productivity relationship in all situations, instead the specific environmental context should be taken into account.