The decline of global biodiversity and the increasing spread of invasive alien plants raise critical questions about how native species diversity and biological invasions interact to influence community resistance to disturbance, particularly global climate change. 

To explore the effects of native species diversity and biological invasions on community resistance to drought, we established experimental plant communities with varying species richness levels (1, 2, 4, 8, and 16 species), introduced the invasive species Symphyotrichum subulatum into half of these communities, and then subjected all communities to no, moderate, or intense drought. 

In both native and invaded communities, community resistance to drought was negatively correlated with biomass of the communities without drought, regardless of species richness. Additionally, lower biomass of the communities without drought was associated with smaller drought-induced increases in root:shoot ratio, which in turn conferred higher community drought resistance. However, the mechanisms underlying drought resistance differed between native and invaded communities: changes in biomass triggered by complementarity effects positively determined drought resistance in native communities, but those triggered by selection effects governed drought resistance in invaded communities. 

Synthesis. These findings reveal distinct mechanisms governing drought resistance in native communities and those invaded by S. subulatum, offering valuable insights for managing the invasion of this species in the context of global climate change.

We constructed experimental plant communities with five levels of species richness (1, 2, 4, 8, and 16 species) on the campus of Taizhou University, Zhejiang Province, China (28°39’ N, 121°23’ E). The fieldwork did not need permission. We used a pool of 16 native plant species commonly found in areas of mixed brush and grassland in nearby valleys and hillsides in Taizhou. We established a total of 46 communities with different species compositions: one monoculture for each of the 16 species, ten two-species mixtures, ten four-species mixtures, nine eight-species mixtures, and one 16-species mixture. Each mixture contained a unique combination of species randomly selected from the species pool. Monocultures and two-, four- and eight-species mixtures were replicated six times. The 16-species mixture was replicated 30 times, resulting in 300 experimental communities. These replicates were randomly assigned to six treatment combinations consisting of three drought levels (no, moderate, and intense drought) and two invasion treatments (with or without the invasive species), as described below. Each of the six treatment combinations contained five replicates of the 16-species mixture and one replicate of the other 45 different communities.

Experimental plant communities were established in plastic containers (72 × 64 × 42 cm) with five drainage holes at the bottom. Each container was filled with a 27 cm layer of soil from hillsides in Taizhou and then topped with a 10 cm layer of a 1:1 mixture of soil and nutrient-rich potting compost (Table S3). A target density of 32 seedlings per container was achieved through sowing (800 seeds per container) and subsequent thinning, retaining only vigorous seedlings of each species. Each species within a mixture was represented by an equal number of seedlings (e.g., eight seedlings per species in four-species mixtures). The 32 seedlings were evenly distributed within the container, avoiding adjacency of conspecific seedlings where possible. Additionally, seedlings that emerged after thinning were removed.

Treatments

All containers were randomly placed under a rain shelter at Taizhou University. Three drought treatments (no, moderate and intense drought) were implemented using automatic drip irrigation systems, controlling irrigation duration to achieve target soil moisture levels. Gravimetric soil water content was maintained at 15.5%-19.8% for the treatment with no drought (similar to ambient conditions in the surrounding hills), 12.4%-15.4% for moderate drought, and 10.0%-12.6% for intense drought. Drought treatments began on March 12, 2015, 16 months after sowing the seeds of the native plant species.

In December 2015, half of the experimental communities (150) were each sown with 50 seeds of the invasive annual herb Symphyotrichum subulatum, evenly distributed within the container. Four monoculture replicates of S. subulatum (50 seeds per replicate) were also established under each drought treatment. Thus, our study simulated the invasion of the alien annual, which started from seeds and covered the entire growth cycle (seeds, seedlings, and adult plants). The density of the seeds of S. subulatum was within the range of the densities observed in the field. In each container with monocultures of the invasive species, the number of seedlings of S. subulatum emerged was 10–12 under no drought, 7–8 under moderate drought, and 3–4 under intense drought.

Measurements

We harvested all communities in October 2016, by which time plants of S. subulatum had reached the adult stage. Aboveground parts were sorted by species, dried to a constant mass at 80°C, and weighed. All six monocultures of the native species Medicago sativa failed to establish (one replicate in each of the six treatment combinations), and these containers were excluded from the analysis. The final sample size was, therefore, 49 communities (container) per treatment. Five soil cores (3 cm in diameter) were randomly collected from each container. Roots were extracted by rinsing the soil cores with water, dried to a constant mass at 80°C, and weighed. Total root biomass per container was estimated based on the area of the soil cores.

The colonization rates of AMF in roots were quantified using the line-intersect method after clearing roots in 10% KOH and staining them in 0.05% trypan blue (Mcgonigle et al., 1990). For each native species in uninvaded and invaded communities with 16 native species, at least 200 intersects of root segments of ~1 cm long were scored under a light microscope at ×200 magnification. The AMF colonization rate of a species was calculated by dividing the number of the infected intersects by the total number of intersects observed.