A series of long-term field and glasshouse experiments were conducted in the semi-arid Sandveld Bioregion of southern Africa, across a wide range of post-mining site and soil conditions, on the two major plant groups, leaf-succulent shrubs (LSS, which establish in open micro-habitats) and non-succulent shrubs (NSS, which establish in sheltered micro-habitats). Seeding experiments applied a sequential ecological framework intended to reduce the impact of impeding natural ecological processes whilst incorporating processes that facilitate plant establishment. Experimental plots were scaled to naturally operating ecological dynamics at the patch scale, and various potentially facilitative types of shelter and soil amelioration, tested separately. Results are reported over three or six year periods and from four or seven field sites. Shelter generally had a greater effect on plant establishment than soil ameliorants, but this could be either positive or negative, and varied for the two plant groups: LSS establishment was inhibited or neutral under all forms of shelter, NSS were facilitated by types of artificial shelter that could be removed or biodegraded without impacting young plants. Both groups were strongly inhibited under transplanted nurse plants. Water absorbing gels improved LSS establishment but other soil ameliorants had no effect, while milled kelp or nutrients improved NSS establishment (measured as a biomass index or plant abundance, respectively). Finally, site had a greater effect than either shelter or soil ameliorants, and could not be related to any soil or other factors measured, supporting recent studies that explicitly quantify the substantial unknown effects of site and year. Specific recommendations are made for comprehensive, seed-based, nucleated restoration designs for the Sandveld and similar bioregions, while the over-arching results are of universal import to dryland restoration and ecology.

在南部非洲半干旱的桑德维尔德生物区，针对两种主要植物群：叶肉质灌木（LSS、LSS、建立在开放的微生境中）和非肉质灌木（NSS，建立在有遮蔽的微生境中）。播种实验采用了连续的生态框架，旨在减少阻碍自然生态过程的影响，同时纳入促进植物生长的过程。实验地块按斑块规模自然运行的生态动态进行调整，并对各种潜在的有利类型的庇护所和土壤改良进行了单独测试。结果是在三到六年的时间内、从四到七个现场进行报告的。庇护所对植物建立的影响通常比土壤改良剂更大，但这可能是积极的，也可能是消极的，并且对于两个植物组来说是不同的：在所有形式的庇护所下，LSS 建立受到抑制或中立，NSS 受到各种类型的人工庇护所的促进可以在不影响幼苗的情况下去除或生物降解。两组在移植的苗木下均受到强烈抑制。吸水凝胶改善了 LSS 的建立，但其他土壤改良剂没有效果，而磨碎的海带或营养物质则改善了 NSS 的建立（分别以生物量指数或植物丰度衡量）。最后，场地比庇护所或土壤改良剂具有更大的影响，并且与任何土壤或其他测量因素无关，支持最近的研究，这些研究明确量化了地点和年份的重大未知影响。针对桑德维尔德和类似生物区的全面的、基于种子的、有核的恢复设计提出了具体建议，而总体结果对旱地恢复和生态学具有普遍重要性。