Summary1. Plant-soil feedback (PSF), plant trait and functional group concepts advanced our understanding of plant community dynamics, but how they are interlinked is poorly known. 2. To test how plant functional groups (FGs: graminoids, small herbs, tall herbs, legumes) and plant traits relate to PSF, we grew 48 grassland species in sterilized soil, sterilized soil with own species soil inoculum and sterilized soil with soil inoculum from all species, and quantified relative growth rate (RGR), specific leaf area (SLA), specific root length (SRL) and per cent arbuscular mycorrhizal fungi colonization (%AMF). 3. Plant growth response to the plant species' own soil biota relative to sterilized soil (PSFsterilized) reflects net effects of all (generalist + specialized) soil biota. Growth response to the plant species' own soil biota relative to soil biota of all plant species (PSFaway) reveals effects of more specialized soil organisms. 4. PSFsterilized showed that graminoids and small herbs have a negative and tall herbs a positive response to their own soil biota, whereas legumes responded neutrally. However, PSFaway showed that on average, all plant FGs benefitted from growing with other species' soil biota, suggesting that pathogens are more specialized than plant growth-promoting soil biota. Feedback to plant growth from all soil biota (PSFsterilized) was stronger than from more specialized soil biota (PSFaway) and could be predicted by SRL and especially by %AMF colonization. Species with high SRL and low %AMF colonization when grown in away soil experienced most negative soil feedback. 5. Synthesis. Plant species from all plant FGs grow better in soil from other species because of less net negative effects of soil biota (in graminoids), or because of more net positive soil biota effects (in tall herbs). Explorative plant species (high SRL, low %AMF colonization) suffer most from negative feedback of all soil biota, whereas more resource conservative species (low SRL, high %AMF colonization) benefit from soil feedback of all soil biota. These findings help to understand replacement of explorative species during succession. Moreover, we suggest a potentially larger role for species with positive feedback than for species with negative feedback to contribute to maintain plant community productivity of diverse communities over time.

Earth is home to over 350,000 vascular plant species that differ in their traits in innumerable ways. A key challenge is to predict how natural or anthropogenically driven changes in the identity, abundance and diversity of co-occurring plant species drive important ecosystem-level properties such as biomass production or carbon storage. Here, we analyse the extent to which 42 different ecosystem properties can be predicted by 41 plant traits in 78 experimentally manipulated grassland plots over 10 years. Despite the unprecedented number of traits analysed, the average percentage of variation in ecosystem properties jointly explained was only moderate (32.6%) within individual years, and even much lower (12.7%) across years. Most other studies linking ecosystem properties to plant traits analysed no more than six traits and, when including only six traits in our analysis, the average percentage of variation explained in across-year levels of ecosystem properties dropped to 4.8%. Furthermore, we found on average only 12.2% overlap in significant predictors among ecosystem properties, indicating that a small set of key traits able to explain multiple ecosystem properties does not exist. Our results therefore suggest that there are specific limits to the extent to which traits per se can predict the long-term functional consequences of biodiversity change, so that data on additional drivers, such as interacting abiotic factors, may be required to improve predictions of ecosystem property levels.

4 6 * These authors contributed equally 4 7 4 8 ABSTRACT 4 9Earth is home to over 350,000 vascular plant species 1 that differ in their traits in 5 0 innumerable ways. Yet, a handful of functional traits can help explaining major differences 5 1 among species in photosynthetic rate, growth rate, reproductive output and other aspects of plant 5 2 performance 2-6 . A key challenge, coined "the Holy Grail" in ecology, is to upscale this 5 3 understanding in order to predict how natural or anthropogenically driven changes in the identity 5 4 and diversity of co-occurring plant species drive the functioning of ecosystems 7,8 . Here, we 5 5 analyze the extent to which 42 different ecosystem functions can be predicted by 41 plant traits 5 6 in 78 experimentally manipulated grassland plots over 10 years. Despite the unprecedented 5 7 number of traits analyzed, the average percentage of variation in ecosystem functioning that they 5 8 jointly explained was only moderate (32.6%) within individual years, and even much lower 5 9 (12.7%) across years. Most other studies linking ecosystem functioning to plant traits analyzed 6 0 no more than six traits, and when including either only six random or the six most frequently 6 1 studied traits in our analysis, the average percentage of explained variation in across-year 6 2 ecosystem functioning dropped to 4.8%. Furthermore, different ecosystem functions were driven 6 3 by different traits, with on average only 12.2% overlap in significant predictors. Thus, we did not 6 4 find evidence for the existence of a small set of key traits able to explain variation in multiple 6 5 ecosystem functions across years. Our results therefore suggest that there are strong limits in the 6 6 extent to which we can predict the long-term functional consequences of the ongoing, rapid 6 7 changes in the composition and diversity of plant communities that humanity is currently facing. 6 8 6 9 BODY 7 0 Worldwide, ecological communities are rapidly changing due to various anthropogenic 7 1 activities 9-12 . This biodiversity change is non-random, and the functional traits of organisms 7 2 driving their growth, survival and reproduction are key in determining which species thrive and 7 3 which perish under global change 13-15 . This may have important implications, as traits not only 7 4 affect individual plant performance, but they may also drive various ecosystem functions such as 7 5biomass production, and the services these functions provide to human well-being 7,8,15 . 6Predicting rates of ecosystem functioning, such as biomass production or carbon 7 7 sequestration, from the composition or diversity of traits in plant communities has been coined 7 8 the "Holy Grail" in ecology 7,8 . Various studies have shown links between plant traits and 7 9 species-level variation in photosynthetic rate, growth, and reproductive output present in the 8 0 plant kingdom 3-5 . However, in natural communities, plants occur in various abiotic 8 1 environments, and they interact with individuals from other species, so ...

No abstract