Questions:We investigated the transfer of seeds from species-rich grassland using green hay and seed material from on-site threshing to answer the following questions: do these two methods differ in terms of restoration success; and how do the two methods behave in their long-term effect concerning species composition and species presence? Location: Styria, Austria. Methods:We harvested seed material of a species-rich Arrhenatherion meadow by green hay and on-site threshing and transferred it to a receptor site. Transferable species were identified by vegetation survey and seed sample analyses. We analysed transfer rates and vegetation cover for the donor site species pool and Arrhenatherion target species, covering a 10-year observation period. Species composition and presence were derived from plant surveys. Results:The restoration success of both methods was similar within the observation period. Although donor site species declined, the establishment of target species was satisfactory and species-rich grassland was successfully established, despite different site conditions between donor and receptor sites and strong dynamics in species composition over time. Conclusions:Both harvesting methods are well suited for restoring species-rich grassland, and the actual choice ultimately depends on the costs and the given circumstances. Even if donor and receptor site differ in their site conditions, species-rich grassland can be established successfully. The prerequisite for this is that: (a) a high proportion of transferred species is able to establish; and (b) specific restoration goals do not exclude vegetation development according to specific receptor site conditions.

We studied the performance of different seed mixtures and the influence of liming on the establishment of a montane pasture under acidic site conditions during a wood‐pasture separation process from 1998 to 2001 in Styria, Austria. Two treatments (with and without liming) were established, each with four seeding treatments (without seeding, commercial, site‐adapted and site‐specific mixtures), and all plots were regularly fertilized from 1998 to 2002. Vegetation surveys and forage analyses regarding development of vegetation cover, seeded species persistence, quality and yield parameters were conducted again in 2014 and 2016; soil samples were taken in 2015. After cessation of regular fertilization, the analysed soil parameters declined to levels that were sufficient or low for grasslands. pH decreased again over time, but still was significantly higher (p = .0034) in the liming treatment. The single liming during experimental set‐up also caused better long‐term performance for nearly all vegetation, quality and yield parameters. The site‐specific seed mixture performed best regarding most analysed parameters, under both liming treatments. Concerning seeded species, the habitat‐typical species Agrostis capillaris, Festuca rubra agg. and Trifolium repens persisted best with and without liming. All analysed vegetation and yield parameters stayed far behind their results of 2001 and their potential. The use of site‐specific seed mixtures supports the establishment of productive grassland on specific locations. The long‐term results also illustrated that even low‐input management like liming in multi‐annual intervals can help to improve acidic site and growing conditions significantly and thereby increases the sustainability of cost‐intensive wood‐pasture separation processes under similar conditions.