1. Understanding species coexistence and regulation of biodiversity are major research challenges, yet there is no consensus on the effects of diversity on diversity, including their mediation through plant–plant interactions. 2. We examined how the diversity of recipient communities impacted on the establishment of colonising species. We ran a greenhouse-based community experiment, creating artificial arable crop communities with varying levels of barley genotype and weed species diversity, analysed with structural equation modelling of responses across trophic levels. 3. Colonising arable weed species establishment was significantly reduced when the recipient communities’ weed species richness, but not barley genotype richness, increased. Recipient plant communities with higher species richness occupied more total niche space and showed stronger competition with colonising species. In contrast, recipient plant communities with higher genotype richness, even though they had slightly greater niche space occupancy, had no significant effects on the colonisation of additional species. 4. Responses were species-specific: colonising common weed species establishment was less affected by the recipient plant communities than that of colonising rare weed species. Strong responses of colonising rare species were related to the abiotic rather than the biotic environment created by the recipient communities. The stronger response of colonising rare species suggests greater niche overlap with and competitive inferiority to the recipient plant community, whereas the relative independence of colonising common species indicates less niche overlap and/or competitive equality or superiority. 5. We observed a negative species diversity effect on colonising species, with the effect size increasing with increasing rarity of the colonising species. This differential responsiveness of colonising rare and common arable weed species to the presence of and diversity in the recipient arable plant communities may be crucial in understanding rarity and may result from differential evolutionary pathways in response to interactions.

datasheet for plot-level measurementsData file containing contrast coding of the experimental treatments, trait metrics of the recipient communities, resource measurements and principal component of the microbial communities of all plots and abundance of the colonising weed species. The file also contains a meta data sheet for further information.database_20160516.xlsxR code for all the calculations, analyses and figuresR code to run all the calculations, analyses and figures using the datasheet providedRcode.R