Characterisation of brush-harvested hayseed quality (purity, seed content and germination) for the calculation of optimal sowing densities

Several projects demonstrate the utility of hayseed for ecological restoration, but how hayseed should be employed remains ambiguous as no standard procedure currently exists for hayseed characterisation. In Europe an absence of guidelines partly reflects the lack of a requirement by EC Directive 2010/60/EU for specific data quality parameters for directly harvested seed mixtures. However, the Directive states that directly harvested seed mixtures should be appropriate for restoration of each given habitat, and a density of 8000 established seedlings m-2 has been shown as a requirement for successful revegetation in alpine grasslands (Florineth 2007). We developed a method for the standardised characterisation of hayseed lots to allow labelling of recommended sowing densities for hayseed from different grassland types, inspired by the quality tests usually applied to commercial seed (such as those of the International Seed Testing Association; ISTA 2017). Hayseed was collected from 49 grassland donor sites in Lombardy, northern Italy, using three different types of brush harvester, and characterised in terms of purity (seed:inert plant material), seed content (number of seeds per unit hayseed weight) and germination capacity (density of seedlings emerging over a given area). Hayseed was also used for restoration and restored areas monitored in terms of seedling density and cover. The mean yield across all donor sites was 60.6 kg ha-1 (min./max. 19.6 to 131.8 kg ha-1). Seed content exhibited a mean of 536.5 seeds g-l, ranging from 82.4 to 2635.0 seeds g-l, and differed between brush harvester types. The mean purity across sites was 28% seed, ranging from 5 to 68%, and was also affected by brush harvester type (p<0.05). Mean germinability was 7000 seedlings m-2 (at 30d from sowing). Recent tests conducted on sub-lots of hayseed preserved for several years (3, 4 or 5), demonstrate a significant increase in germination during short term storage (germination differences in the first and fifth years for three hayseed lots): similar effects have been noted for single species of Poaceae (e.g. Shaidaee et al. 1969). During revegetation trials, complete soil cover was generally achieved after three months, associated with declining seedling density due to shifts in competitive dominance. Monitoring of sites over several years shows the progressive and spontaneous ingress of other species from surrounding vegetation, leading to floristic enrichment and development of the grassland. Optimal sowing density for each hayseed lot ranged between 5 and 94 g of hayseed per m2: values for most lots were below the sowing densities for commercial seed mixtures advised by ERSAF (2001) for establishment of vegetation for erosion control (30 to 50 g m2). Standardised characterisation of hayseed lot quality is possible in a way that allows sowing density recommendations and labelling to be carried out for each type of donor grassland.