Evaluating the success of upland hay meadow restoration in the North Pennines, UK, using green hay transfer

1. Traditionally managed mesotrophic species-rich upland hay meadows conforming to the National Vegetation Classification (NVC) MG3b, are one of the rarest grassland types in the UK, with substantial declines in botanical diversity over the last 50 years. Intensive spring grazing, earlier cut dates and increases in soil fertility causes a decline in characteristic positive indicator species in MG3b meadows, shifting communities from species-rich MG3b, to NVC MG6 meadows, and finally to species-poor NVC MG7 meadows. 2. The North Pennines Area of Outstanding Natural Beauty (AONB) Partnership's Hay Time project aimed to improve the knowledge of upland hay meadows, and to investigate the success of seed addition of key positive indicator species. A landscape-scale restoration programme was undertaken between 2006 and 2012, harvesting seed from 82 species-rich donor meadows and spreading seed onto 89 receptor meadows (2282ha). Seed was harvested as green hay, using two types of donors: an MG6 donor (classed as restoration, with species such as Rhinanthus minor), or using an MG3b donor (classed as enhancement, with species such as Geranium sylvaticum). All 89 meadows were monitored, with a baseline botanical survey, and a repeat survey three to five years after seed addition. In addition, 41 meadows that did not have seed addition were monitored (controls). 3. Species-richness, diversity and floristic composition improved in 77 meadows three to five years after seed addition. Eighteen plant species had an increase in frequency in the receptor meadows but did not increase in frequency in the control meadows. The most successful were eight positive indicators which were annuals or fast-growing perennial plants (Anthoxanthum odoratum, Euphrasia spp., Myosotis discolor, Plantago lanceolata, Ranunculus acris, R. minor, Trifolium dubium and Trifolium pratense). However, rarer characteristic MG3b plants such as Alchemilla spp., G. sylvaticum, and Cirsium heterophyllum showed little signs of establishing. 4. Botanical evidence is demonstrating that seed addition using green hay is a successful way of restoring meadows to an MG6 community. What is now needed is an effective method to establish characteristic MG3b plants. Hand-collecting seeds and establishing plug plants, alongside seed addition and maintaining traditional management practises is one possible way forward.Calculating meadow variables For each meadow, the botanical survey data was used to calculate the species richness (total number of species found) and Shannon Diversity Index (H') (Shannon, 1948). At the beginning of the Hay Time project, grassland plant species recorded in the North Pennines were assigned an indicator score ranging from -2 through to +4. Positive indicator species are plants that are typical of upland hay meadows. Positive indicator species in MG3b meadows were given +3 scores, whereas MG6 plants were given +1 or +2 scores. Neutral indicator species are plants that are common to all grasslands, and these were given zero scores. Negative indicator species are plants that are competitive, injurious weeds or invasive, and these were given negative scores (Table 1). Positive Indicator Species Scores and Negative Indicator Species Scores were calculated by multiplying the frequency on a 1 to 5 scale (DAFOR) with the species score and summing all species per meadow (O'Reilly/North Pennines AONB Partnership, 2006; Starr-Keddle, 2018). Ellenberg Fertility Index (N) was similarly calculated by multiplying the species frequency with the species fertility score and summing all species per meadow, then averaged (weighted means) (Hill et al., 1999). TABLEFIT was used to calculate the goodness of fit to MG3b for each meadow survey (Hill et al., 1996). This was calculated to see if the receptor meadows had improved in botanical composition similar to an MG3b meadow.Botanical surveys and monitoring Botanical surveys consisted of a W-shaped walk-through the meadow, taking approximately one hour to complete. All vascular plants were recorded (nomenclature according to Stace, 2010) and allocated a DAFOR frequency score, between 1 to 5 (1=rare, 2=occasional, 3=frequent, 4=abundant, 5=dominant) (O'Reilly/North Pennines AONB Partnership, 2006; Starr-Keddle, 2018). All 89 receptor meadows were surveyed just before seed addition, with a repeat monitoring survey undertaken three to five years later. Forty-one meadows that did not have seed addition were monitored (controls), with a baseline survey, and repeat monitoring survey undertaken three to five years later. The control meadows were often on the same farm as the receptor meadows. Monitoring surveys were undertaken from mid-June to mid-July, between the years 2009 – 2017. Monitoring surveys were undertaken close to the same day as the previous baseline survey