Two decades of dune slack restoration in North Wales: Diversity, Community and Habitat Specialists - Dataset and analysis code

For the restoration scrub and trees were removed along with 10cm of nutrient rich topsoil to explose bare sand. Ten 2m x 2m quadrates were established and surveyed in subsequent years (surveyed between May and July in: 2006, 2007, 2008, 2009, 2010, 2011, 2013, 2014, 2018, 2020, 2021 and 2022) to monitor botanical succession. These quadrats are referred to as "scraped" in the dataset which a comparable non-restored area was survery in 2022 as a point of comparision and referred to in the dataset as "unscraped".
Species data was collated from all previous survey years as follows: individual quadrat and abundance data (the DOMIN scale, a 1 to 10 scale for abundance (Rodwell, 2006) was used in previous surveys) for 2006, 2007, 2008, 2018, 2020 and 2021 (termed ‘full survey years’); total species list only for 2009, 2010, 2011, 2013 and 2014. Due to the nature of the vegetation – short with numerous very small plants – distinguishing between some closely related species can be difficult. This is likely to have resulted in differences in professional opinion between different surveyors over the years. Data was cleaned to account for these potential differences and enable anaylsis (see paper appendix for the full specifics of what was done). To be able to analyse comparable data, the DOMIN scores from previous full survey year were converted to percentage values using the mid-point percentage value of each DOMIN grouping. Soil data was not collected as part of the previous surveys, therefore EIVs were used as a proxy measure to assess any changes in abiotic factors over time. Mean EIVs for light (L), moisture (F), reaction (R), nitrogen (N), and salt (S) (Hill et al., 1999) were calculated for all species on the total species list for all survey years to assess any changes in functional type over time. The EIV’s were not weighted as the species lists contained no abundance data.
Processed data and R code for Indicator Species Analysis, NMDS and EIVs are also included.

为开展修复清理工作，研究人员移除了区域内的灌丛与树木，并刮除了10厘米厚的富营养表层土壤以暴露裸沙。共设立10个2米×2米的样方（quadrate），并在后续各年份开展调查，调查时间为各年份的5月至7月，具体年份包括2006、2007、2008、2009、2010、2011、2013、2014、2018、2020、2021及2022年，以监测植物群落演替。本数据集将此类样方命名为“刮除组”（scraped）；2022年同时对一处未进行修复的对照区域开展了调查作为参照，该区域在数据集中被命名为“未刮除组”（unscraped）。

物种数据整合自既往各调查年份，具体规则如下：2006、2007、2008、2018、2020及2021年（称为“完整调查年份”）记录了单个样方的物种及丰度数据，丰度评级采用DOMIN等级（DOMIN scale，1至10级丰度评级，Rodwell, 2006）；2009、2010、2011、2013及2014年仅记录了物种总名录。

由于该区域植被低矮且伴生大量小型植株，部分近缘物种难以区分，这可能导致多年来不同调查人员间的专业判断存在偏差。研究人员对数据进行了清洗以修正这类潜在差异，以支持后续分析（具体处理细节参见论文附录）。

为保证分析数据的可比性，既往完整调查年份的DOMIN评级被转换为百分比数值，转换方式为取每个DOMIN等级对应的中点百分比值。既往调查未采集土壤数据，因此采用埃伦贝格指示值（Ellenberg Indicator Values，EIVs）作为替代指标，以评估非生物因子随时间的变化。针对所有调查年份的总物种名录中的物种，计算其光照（L）、湿度（F）、酸碱度（R）、氮素（N）及盐分（S）的平均EIV值（Hill et al., 1999），以分析功能群随时间的变化。由于物种名录未包含丰度数据，因此未对EIV值进行加权。

本数据集同时包含指示物种分析、非度量多维标度分析（Non-metric Multidimensional Scaling，NMDS）及EIV值计算所用的处理后数据与R代码。