To test the applicability of alternate land use systems for improving soil functionality in restored ecologies, soils were sampled from 0-15, 15-30 and 30-45 cm deep layers of Leucaena leucocephala (LL), Hardwickia binata (HB), Emblica officinalis (EO), Azadiracta indica (AI)-based silviculture systems; Acacia nilotica-based silvipasture systems (AN); and natural grassland (NG). These were compared with samples from fallow land (F). They were evaluated for their carbon management index (CMI), nutrient supply capacity (NSC), soil functionality (SF), ecorestoration efficiency (ERE) and 21-day cumulative microbial respiration (CO 2 -21) to assess their applicability in semiarid India. Soil functionality and functional diversity as impacted by restoration have remained largely overlooked. The LL had $12, 7 and 11% higher CMI than fallow in sampled soil layers. ERE of LL was $ 55, 65 and 79% higher than fallow land in sampled soil layers, respectively. However, ERE in surface layer was poorer than subsequent soil layers for all systems. The LL, HB and AN improved NSC and SF by: a) $ 2.5-, 2.2-and 1.6-times; and b) 9.3-, 5.3-and 5.1-times over fallow land in the surface soil layer. A similar trend was observed for SF in lower layers. However, the topsoil layer had >16% mean SF values than subsequent layers. LL, HB and AN systems had $4.2-, 2.3-and 1.9-times higher CO 2 -21 than fallow land in the top layer. CO 2 -21 was positively correlated with NSC and SF but could not predict ERE well. Hence, legume tree-based restoration tactics might be useful for improving land restoration and soil functionality in semiarid regions.