To meet these objectives, we used a variety of experimental approaches.
To address Objective 1, we adopted and altered a long-term drought field experiment. This allowed us to test how 20 years of summer drought affects plant and soil microbial communities, and how these changes determine future plant growth and response to new drought. We found that chronic drought exposure strongly shifted the composition of both soil bacterial and fungal communities, and drought legacy reduced the amount of carbon stored in soils underneath older heather plants, through it effect on bacterial communities. These findings indicate that the active management of heathlands is essential not only for keeping aboveground vegetation dynamics, but also for maintaining below-ground soil nutrient and carbon pools. These findings have been published in Journal of Applied Ecology (Gliesch et al. 2024, output x).
To assess whether these changes in soil microbial communities as a result of chronic drought persistent after drought , and whether these changes affect the response of the ecosystem to a new drought, we subjected both control plots and drought legacy plots of this long-term experiment to a new drought treatment. We found that the persistent changes in soil bacterial and fungal communities did not recover, but that the effect of new drought on these communities was minimal. Drought legacy also shifted plant community composition, with the strongest effect of new drought on building stage Calluna. These findings highlight a mismatch between plant and soil microbial community response to drought. Importantly, we found that ecosystem CO2 fluxes were determined by shifts in plant community composition, and that drought legacy only had a minor effect on how ecosystem CO2 fluxes respond to a new drought.
To assess the effect of these persistent shifts in soil fungal and bacterial communities for future plant growth and community composition, we grew the grass Molinia caerulea and heather (Calluna vulgaris) in soils collected from underneath these species in drought and control plots, alone and in combination. We found that drought increased the invasion growth rate of Molinia, which is in line with our field observations that grasses and fast-growing forbs are increasing in abundance after drought.
To address Objective 2, we set up field drought experiments in six sites, representing a chronosequence of time since land-abandonment. In the third year of these field experiments, we performed a manipulation experiment within control plots, drought plots, and ‘after-drought’ plots to quantify the effects of plant-plant and plant-microbial interactions on plant growth in response to drought. These results show that the presence of plant-microbe interactions mitigates the effect of drought on plant growth, and promotes more facilitative interactions in early and late successional stages.
To address Objective 3, we set up a series of mechanistic experiments, ranging from large, outdoor pots with different plant communities, to individual species growing in the greenhouse. In our long-term, field-based mesocosm experiment we found that drought only shifts plant-soil feedback when plants are grown in communities in the feedback phase, and that in particular the fast-growing grass experienced more positive feedback under drought, which may explain the shift towards dominance of this plant species in response to drought. We also found, similar to the field experiments for Objective 1, that aboveground and belowground responses to drought show a mismatch.
In an experiment growing individual plants of 12 different species, grasses and forbs had negative plant-soil feedback, while legumes had neutral to positive feedback, but that overall, feedback was not affected by drought and only marginally predicted by root traits. In a separate experiment, we found that the feedback caused by root exudates was distinct from total feedback effects of individual plants, and we identified the bacterial and fungal taxa that are linked to these responses. These findings highlight that drought effects on plant-soil feedback may be community and trait-dependent.
