Abstract Soil bioengineering for riverbank stabilization involves the use of living plant materials to treat unstable or eroding riverbanks. These near-natural structures may harbor a higher plant richness and vegetation cover compared to classical civil engineering structures such as ripraps, but little information exists on vegetation dynamics during secondary succession on stabilized riverbanks. We hypothesized that soil bioengineering, by means of active introduction of early successional Salix shrubs, can foster successional trajectories of riparian plant communities, unlike civil engineering. We sampled three types of riverbank stabilization structures: pure bioengineering structures, mixed structures (combining riprap and bioengineering techniques) and ripraps, across a 14-year sequence on 42 sites located along 23 different streams running through the foothills of the Alps and the Jura Mountains (France and Switzerland). We quantified species richness and density and compared the temporal patterns of four groups of species that normally appear sequentially in natural succession on riverbanks (ruderal, hygrophilous, shade-tolerant, competitive species), as well as non-native species. Plant community composition differed greatly between ripraps and the two types of bioengineered sites, and ligneous species typical of advanced successional stages (Cornus sanguinea, Corylus avellana) spontaneously established in the oldest bioengineered sites. In general, richness of total species was higher in stabilization structures using soil bioengineering (including mixed structures) than in riprapped sites. In particular, the number of shade-tolerant and competitive species in bioengineered sites was double that found at ripraps after 14 years. Yet, richness of shade-tolerant species increased over time only on purely bioengineered sites, and their density there was almost twice that in mixed structures. Neither the type of stabilization structure nor time explained the variability in richness and density of non-native species across sites. Our study showed that along streams running through foothills, where erosion processes are usually intense, vegetation of bioengineered riverbanks exhibits successional dynamics similar to those theoretically found in natural conditions. Bioengineering can therefore foster ecological processes while stabilizing eroding riverbanks along foothill streams, thus satisfying human needs for infrastructure protection with less impact on the riparian ecosystem than riprap structures.

中文翻译：

---

沿河岸生物工程控制侵蚀的河岸植物群落的多样性和演替：阿尔卑斯山和侏罗山脉山麓的案例研究

摘要 河岸稳定的土壤生物工程涉及使用活的植物材料来处理不稳定或侵蚀的河岸。与抛石等经典土木工程结构相比，这些近自然结构可能具有更高的植物丰富度和植被覆盖度，但关于稳定河岸次生演替期间植被动态的信息很少。我们假设土壤生物工程通过积极引入早期演替柳灌木，可以促进河岸植物群落的演替轨迹，这与土木工程不同。我们采样了三种类型的河岸稳定结构：纯生物工程结构、混合结构（结合抛石和生物工程技术）和抛石，跨越阿尔卑斯山脚下和侏罗山脉（法国和瑞士）的 23 条不同溪流沿线的 42 个地点的 14 年序列。我们量化了物种丰富度和密度，并比较了通常在河岸自然演替中顺序出现的四组物种（粗野、喜湿、耐阴、竞争性物种）以及非本地物种的时间模式。抛石和两种类型的生物工程站点之间的植物群落组成差异很大，并且在最古老的生物工程站点中自发建立的高级演替阶段（Cornus sanguinea，Corylus avellana）典型的木质物种。一般来说，使用土壤生物工程（包括混合结构）的稳定结构中的总物种丰富度高于抛石场地。特别是，在 14 年后，生物工程场地中耐阴和竞争物种的数量是在抛石场中发现的两倍。然而，仅在纯生物工程场地上，耐阴物种的丰富度会随着时间的推移而增加，它们的密度几乎是混合结构中的两倍。稳定结构的类型和时间都不能解释不同地点非本地物种丰富度和密度的变化。我们的研究表明，沿着流经山麓的溪流，侵蚀过程通常很激烈，生物工程河岸的植被表现出类似于自然条件下理论上发现的演替动态。因此，生物工程可以促进生态过程，同时稳定沿山麓溪流侵蚀的河岸，