In experiments that test plant diversity-productivity relationships, the common practice of weeding unsown species and disallowing species colonization may have the unintended consequence of favoring priority effects that maintain niche complementarity in determining productivity. However, in naturally assembled communities where colonization occurs, resource competition may favor dominant traits, which eventually have the greatest influence on productivity. Here, in naturally developed long-term subalpine meadows (from 4-year to at least 40 years meadows) in the Qinghai-Tibetan Plateau, we investigated the relationships between species richness and productivity to testify whether positive diversity-productivity relationships can still exist in naturally developed long-term communities. We also measured five functional traits (specific leaf area, photosynthesis rate, leaf proline content, seed mass and seed germination rate) to calculate two functional diversity indices: community-weighted mean trait values (CWM) and Rao's quadratic entropy (RaoQ) which are highly correlated to functional traits of dominating species and variety of functional trait among all species. Finally, we quantified the relative contribution of species diversity, functional traits of dominating species and functional diversity among all species to productivity along the succession. We demonstrated strong positively diversity-productivity relationships in the natural sub-alpine meadow communities across time. The five traits of dominating species explained a large proportion (54-80%) of the variation in productivity during succession, whereas species diversity and functional diversity (FD) for each of the five traits explained much less (24-48% for species richness and 0-40% for FD for each of the five traits respectively). We found unequivocal evidence that significantly positive diversity-productivity relationships in the natural sub-alpine meadow communities across time are up to superior performers (dominant traits) in naturally developed communities where colonization occurs. As a result, understanding diversity-productivity relationships under the full range of community assembly processes therefore merits further investigation. Global biodiversity is declining sharply 1 with the potential to impair ecosystem functioning in the near future, but the mechanisms that connect biodiversity to ecosystem function are not well understood 2. In diversity-function relationships, the connection between plant diversity and productivity is considered particularly important 3. Although the classic diversity-productivity relationship is thought to be hump-shaped, with species richness highest at intermediate levels of productivity 4 , experimental studies have, however, mostly yielded linear positive diversity-productivity relationship 5-9. In long-term experiments, the positive diversity-productivity relationships

More than half of the world's tropical lowland rainforests have been lost due to conversion to agricultural land (such as rubber plantations). Thus, ecological restoration in degraded tropical lowland rainforests is crucial. The first step to restoration is restoring soil functioning (i.e., soil fertility, carbon, and nitrogen cycling) to levels similar to those in the primary tropical lowland rainforest. This requires understanding soil nematode community assembly in primary tropical lowland rainforest, which has never been explored in this habitat. In this study, we measured species compositions of plant and soil nematode communities and soil characteristics (pH, total and available nitrogen, phosphorus, and soil water content) in a primary tropical lowland rainforest, which is located on Hainan Island, China. We performed two tests (the null-model test and distance-based Moran's eigenvector maps (MEM) and redundancy analysis-based variance partitioning) to quantify the relative contribution of the deterministic (abiotic filtering and biotic interactions) and stochastic processes (random processes and dispersal limitation) to the soil nematode community. We found that a deterministic process (habitat filtering) determined nematode community assembly in our tropical lowland rainforest. Moreover, soil properties, but not plant diversity, were the key determinants of nematode community assembly. We have, for the first time, managed to identify factors that contribute to the nematode community assembly in the tropical lowland rainforest. This quantified community assembly mechanism can guide future soil functioning recovery of the tropical lowland rainforest.

1. Nitrogen (N) fertilization and warming are two crucial global change factors affecting the soil nematode communities. The effects of N fertilization and warming, however, on nematode communities in soils are inconsistent across ecosystems and maybe be even opposite.