Data from: Using filter-based community assembly models to improve restoration outcomes

1. Ecological filter models derived from community assembly theory can inform restoration planning by highlighting management actions most likely to affect community composition. Despite growing interest in these models, many restoration studies solely manipulate a single filter—the biotic filter by altering interspecific competition in studies—while ignoring abiotic and dispersal filters that may also influence restoration success. 2. To examine how manipulating all three filters (biotic, abiotic, dispersal) affected restoration in an annual-type grassland, we seeded native forbs from the same functional group as a target invader to increase biotic resistance to invasion (biotic filter), cut standing biomass and either removed it or returned it to plots as litter to alter light conditions (abiotic filter), and added native forbs at different seeding rates to alter density of establishing native populations (dispersal filter). We measured restoration success by recording native species and invader cover in plots. 3. The addition of native species with phenological and morphological traits similar to the target invader reduced invasion and increased native populations, but only in litter-free plots when high-densities of native seed were added. 4. Seeding two species with functional traits similar to the invader was more effective at reaching restoration goals than seeding just one functionally similar species. As such, trait differences among restoration species, even species belonging to the same functional group, may increase biotic resistance to invasion in restored communities. 5. Litter removal altered native-invader interactions. When litter was kept, added natives did not reduce invader cover. However, when litter was removed, added natives led to declines in invader cover. 6. Increasing native seeding rates led to larger native populations and increased invasion resistance. 7. Synthesis and applications. In this study, simultaneously manipulating biotic, abiotic and dispersal filters was necessary to optimize restoration outcomes. In particular, the biotic filter only contributed to successful restoration outcomes under abiotic and dispersal conditions that were created through management actions specifically targeting these two additional filters. Restoration planning based on filter models should incorporate actions that target all three filters, rather than solely focusing on the biotic filter.1. 基于群落组装理论（community assembly theory）的生态过滤模型（ecological filter models），可通过识别最有可能影响群落组成的管理措施，为恢复规划提供科学依据。尽管学界对这类模型的关注度与日俱增，但诸多恢复研究仅针对单一过滤因子开展调控——即通过改变种间竞争来调控生物过滤（biotic filter），却忽略了同样可能影响恢复成效的非生物过滤（abiotic filter）与扩散过滤（dispersal filter）。 2. 为探究同时调控三类过滤因子（生物、非生物与扩散）对一年生草原恢复的影响，本研究开展如下处理：通过播种与目标入侵种同功能群的本土非禾本科草本植物，提升群落对入侵的生物抗性（调控生物过滤）；刈割现存生物量，并将其移除或作为枯落物返还样地，以改变光照条件（调控非生物过滤）；以不同播种率添加本土非禾本科草本植物，改变定植本土种群的密度（调控扩散过滤）。本研究通过记录样地内本土物种与入侵种的盖度，评估恢复成效。 3. 播种具备与目标入侵种相似物候与形态性状的本土物种，可抑制入侵并提升本土种群规模，但该效果仅在无枯落物的样地中，且高播种密度的条件下才会显现。 4. 播种两种具备与入侵种相似功能性状的本土物种，相比仅播种一种同功能群物种，更易达成恢复目标。由此可见，即便同属一个功能群，恢复用物种间的性状差异仍可提升恢复群落对入侵的生物抗性。 5. 枯落物移除会改变本土物种与入侵种的相互作用。当保留枯落物时，添加的本土物种并未降低入侵种盖度；而当移除枯落物后，添加的本土物种则会导致入侵种盖度下降。 6. 提高本土物种的播种率，可提升本土种群规模并增强入侵抗性。 7. 总结与应用：本研究表明，若要优化恢复成效，需同时调控生物、非生物与扩散三类过滤因子。具体而言，生物过滤仅在通过管理措施同时调控非生物与扩散过滤的条件下，才能助力恢复取得成功。基于生态过滤模型的恢复规划，应纳入针对三类过滤因子的管理措施，而非仅聚焦于生物过滤。