Data from: Dispersal and establishment limitation slows plant community recovery in post-agricultural longleaf pine savannas

Abandoned agricultural lands often have distinct plant communities from areas with no history of agriculture because plant species fail to recolonize. This may be due to dispersal limitation from a lack of seeds, or establishment limitation because of unsuitable environmental conditions. However, few experiments have directly tested how restoration activities may overcome these limitations.
We studied longleaf pine savannas in South Carolina abandoned from agriculture >60 years ago that were immediately adjacent to remnant habitats (areas with no history of agriculture). Using 27 sites, we conducted a factorial experiment that sowed seeds of 12 species indicative of remnant communities and conducted restoration thinning of overstorey trees in half of 126, 1-ha patches to mimic canopy density of natural savannas. We also established vegetation transects to examine if restoration promotes spread of remnant species into post-agricultural areas.
We found strong evidence for dispersal limitation in post-agricultural areas as over 99% of the occurrences of our focal species were in seed addition plots. Seed additions increased total species richness by 27%.
Restoration thinning increased establishment in seed addition plots (measured as richness of sown species) by 126% and increased total richness by 88%. Restoration thinning also increased seed production in remnant habitats by an average of 6506% across our focal species. However, after 4 years, restoration thinning did not facilitate the natural spread of remnant species into adjacent post-agricultural sites.
Synthesis and applications. We show that both dispersal and establishment limitation are key factors causing some plant species to be absent from post-agricultural sites. Dense canopy conditions limit seed production in remnant habitats and reduce establishment in post agricultural areas. Restoration thinning helps overcome these limitations and should facilitate the natural spread of species from remnant habitats but natural recovery may still be slow. Our results suggest that accelerating the recovery of post-agricultural habitats will require active restoration that reduces dispersal limitation (seed additions) and reinstates appropriate ecological conditions.

弃耕农田往往与无农业开垦历史的区域拥有截然不同的植物群落，原因在于植物物种未能成功重新定植。这一现象可能源于种子匮乏导致的扩散限制（dispersal limitation），或是不适宜的环境条件引发的建群限制（establishment limitation）。然而，鲜有研究直接检验修复活动能否克服这些限制因素。
我们以美国南卡罗来纳州弃耕超过60年、且紧邻残留生境（remnant habitats，即无农业开垦历史的区域）的长叶松稀树草原（longleaf pine savannas）为研究对象。依托27个样地，我们开展了一项析因实验（factorial experiment）：播种12种代表残留群落的物种，并对126个1公顷样地中的一半进行上层乔木（overstorey trees）修复疏伐（restoration thinning），以模拟天然稀树草原的冠层密度（canopy density）。此外，我们还设置了植被样带（vegetation transects），以探究修复活动是否能促进残留物种向弃耕区域扩散。
我们发现弃耕区域存在显著的扩散限制：目标物种（focal species）超过99%的出现记录均位于种子添加样地（seed addition plots）中。种子添加使总物种丰富度（species richness）提升了27%。修复疏伐使种子添加样地内的建群成功率（以播种物种的丰富度为衡量指标）提升了126%，并使总物种丰富度提升了88%。同时，修复疏伐使残留生境中目标物种的种子产量平均提升了6506%。但在4年后，修复疏伐并未促进残留物种自然扩散至邻近的弃耕区域。
综合与应用：我们的研究表明，扩散限制与建群限制均是导致部分植物物种无法出现在弃耕区域的关键因素。稠密的冠层条件会抑制残留生境的种子生产，并降低弃耕区域的物种建群成功率。修复疏伐可有效克服这些限制，理论上应能促进残留生境的物种向周边自然扩散，但自然恢复过程仍可能较为缓慢。本研究结果显示，若要加速弃耕生境的恢复进程，需开展主动修复工作，同时缓解扩散限制（如进行种子添加）并恢复适宜的生态条件。