Case study: planting methods and beneficial substrate microbes effect on the growth of vegetated roof plants in Finland

Vegetated roofs, often called “green roofs”, are popular and necessary in urban greening in densely populated areas. Well-functioning vegetated roofs can provide various ecosystem services to urban residents (e.g., stormwater management, air pollution mitigation, and aesthetic value). Plants essentially determine the actualization of the ecosystem services, thus finding effective ways to establish and maintain the roof plants is important. While greenhouse experiments can be better controlled than field experiments, it is critical to test whether results gained in the greenhouse hold in actual roof conditions. Therefore, we investigated the effects of microbial inoculant, plant species, planting method, and their interactions on plant growth and the beneficial microbes in the roof substrate at the initial establishment of vegetated roofs. The selected plants (i.e., Antennaria dioica, Campanula rotundifolia, Fragaria vesca, Geranium sanguineum, Lotus corniculatus, Thymus serpyllum, Trifolium repens, and Viola tricolor) were established using pre-grown vegetation mats, plug plants, and seeds, each with and without the co-inoculation with Rhizophagus irregularis and Bacillus amyloliquefaciens, two plant growth-promoting microbial species. Eventually, only F. vesca, T. serpyllum, T. repens, and V. tricolor were found successfully settled in either of the three planting methods. Dry aboveground plant biomass was measured to assess the effects of co-inoculation on plant growth. R. irregularis colonization level and B. amyloliquefaciens bacterial density were detected from root and substrate samples, respectively. The results indicated that co-inoculation with R. irregularis and B. amyloliquefaciens successfully colonized target plant species and significantly increased the initial growth of the vegetated roof plants by 18–292%. Additionally, the abundance of R. irregularis was affected by plant species (F. vesca > T. serpyllum > T. repens) and planting methods (seed > plug > mat), while the bacterial density of B. amyloliquefaciens was higher in T. repens roots than the other plant species, and was not affected by planning methods.