Rapid agricultural intensification has led to the decline of grassland species richness and the disruption of biogeochemical cycles, ultimately contributing to the loss of grassland ecosystem services and multifunctionality. Restoring grassland biodiversity through the reintroduction of herb and legume species into grass-dominated pastures can offer a practical, low-cost grassland degradation mitigation strategy by improving the soil physiochemical characteristics that underpin ecosystem service delivery. In the UK, herbal leys (multispecies swards) are rapidly gaining in popularity due to their potential to deliver greater agronomic and environmental benefits than conventional grass-clover pastures. However, despite their promotion in agri-environment schemes, little is known about the effect of high-diversity (e.g., 9-18 species) commercial herbal leys on lowland sheep production, soil N cycling, and soil quality. The overall aims of this thesis were therefore to: i) determine if herbal leys can improve sward nutritional quality and productivity while enhancing livestock productivity and health (Chapter 3); ii) investigate if a commercial herbal ley can reduce urine N excretion and soil N losses (e.g., NO3-, NH3, N2O) associated with lowland lamb grazing (Chapter 4); and iii) evaluate if a commercial herbal ley can improve short-term soil physiochemical characteristics responsible for below-ground ecosystem service delivery in a sheep grazed grassland (Chapter 5). To explore this, a 2-ha split-field experiment using either a herbal or grass-clover ley (0.33 ha paddock-1, n = 3 per sward) was established at the Henfaes Research Centre (North Wales, UK) in July 2020 and rotationally grazed by Welsh mountain lambs (ca. n = 40 per sward) over two experimental seasons: autumn 2020 (males) and spring 2021 (females). Chapter 3 revealed that sward nutritional quality (e.g., crude protein) did not differ between the herbal and grass-clover ley in either experimental season, however, the herbal ley contained higher concentrations of macro- and micronutrients in both seasons. Subsequently, spring liveweight gain was greater in lambs grazing the herbal (172 ± 7 g d-1) vs. grass-clover ley (144 ± 7 g d-1), while autumn liveweight gain showed no difference, driven by the high gastrointestinal parasite burden in both treatments. Spring lambs grazing the herbal ley compared to the grass-clover had elevated plasma cobalt (2.0 ± 0.1 nmol l-1 vs. 1.6 ± 0.1 nmol l-1) and selenium (0.7 ± 0.04 µmol l-1 vs. 0.5 ± 0.01 µmol l-1), with lower blood urea (7.7 ± 0.3 mmol l-1 vs. 10.4 ± 0.4 mmol l-1). Chapter 4 found that while there was no difference in lamb dung and urine N concentration between swards, seasonal variations in urination volume increased the urine N loading rate in the herbal ley in autumn (1020 vs. 555 kg N ha-1), but not spring. A higher sward sodium content drove greater urinary sodium excretion in herbal ley grazed lambs in autumn (439 vs. 71 mg Na l-1) and spring (389 vs. 47 mg Na l-1). Surprisingly, sward type did not affect soil N cycling or gaseous N losses (e.g., N2O) associated with lamb grazing in either season. Chapter 5 examined soil characteristics 2-years after ley establishment and discovered no overall difference in soil physical (e.g., aggregate stability), chemical (e.g., soil organic carbon) and biological (e.g., earthworm abundance) properties between sward types. Despite clear differences in root architecture, X-ray µCT analysis revealed greater pore connectivity in grass-clover ley intact soil cores (0-10 cm depth, 7.5 cm diameter) than herbal ley cores dominated by Plantago lanceolata. The findings of this thesis have implications for the design and promotion of agri-environment schemes aimed at promoting soil quality and sustainability, with further research needed to optimise commercial herbal ley mixtures to deliver greater agronomic and environmental benefits and help the UK agri-food system achieve sustainable agricultural intensification.