Data from Castro et al[Taxonomic invisibility and knowledge shortfalls in terrestrial molluscs of the Caatinga Dominion, a seasonally dry tropical region]

This dataset supports the study “Taxonomic invisibility and knowledge shortfalls in terrestrial molluscs of the Caatinga Dominion, a seasonally dry tropical region”. We analyse whether the currently available occurrence data for Caatinga terrestrial molluscs are insufficient to represent the known and expected spatial distribution of the group, and whether many species become analytically “invisible” when minimum occurrence-record thresholds are applied. The dataset contains curated occurrence records and R scripts used to assess observed richness, expected richness, sampling completeness, spatial sampling gaps and species visibility under increasing occurrence thresholds. Occurrence records were compiled from scientific collections, GBIF, and the published literature, then taxonomically standardised and spatially filtered to retain valid records of terrestrial molluscs in the Caatinga Dominion, north-eastern Brazil. The data show that 155 terrestrial mollusc species, belonging to 62 genera and 25 families, are currently documented from 944 occurrence records inside the Caatinga Dominion. However, these records are highly sparse and spatially uneven. Only a small fraction of 10 km grid cells contain occurrence records, whereas most cells are included in the expected-occupancy area but lack any formal record. The outputs also show very low sampling completeness, low overlap between observed and expected richness, and a rapid decline in the number of species eligible for spatial analyses as the minimum occurrence threshold increases. Expected richness should be interpreted as a standardised, data-driven hypothesis of potential occupancy. It was derived from simple geometric occupancy envelopes, using buffers for poorly recorded species and concave hulls for species with enough records to construct polygons. These outputs are intended to support reproducibility, transparency and future studies on biodiversity knowledge shortfalls, spatial sampling gaps and conservation planning for terrestrial molluscs in tropical dry regions.本数据集支持题为《卡廷加生物区（Caatinga Dominion）——季节性干燥热带区域的陆生软体动物分类学不可见性与知识缺口》的研究。我们将分析当前可获取的卡廷加陆生软体动物出现数据是否不足以表征该类群已知及预期的空间分布，以及当设置最低出现记录阈值时，是否有大量物种在分析中变得“不可见”。 本数据集包含经整理的出现记录与R脚本，用于评估不同出现阈值梯度下的观测物种丰富度、预期物种丰富度、采样完整性、空间采样缺口及物种可见性。出现记录汇编自科学馆藏、全球生物多样性信息设施（GBIF, Global Biodiversity Information Facility）及已发表文献，随后经过分类学标准化与空间筛选，以保留巴西东北部卡廷加生物区内地生软体动物的有效记录。 数据显示，目前在卡廷加生物区内已有944条出现记录，记录了隶属于25科、62属的155种陆生软体动物。然而这些记录极为稀疏且空间分布极不均衡。仅极少数10公里网格单元存有出现记录，而多数网格单元虽处于预期分布范围内，却无任何正式记录。分析结果还显示采样完整性极低，观测与预期物种丰富度的重叠度不足，且随着最低出现阈值升高，可用于空间分析的物种种数会快速减少。 预期物种丰富度可被理解为一种标准化的、由数据驱动的潜在分布假说。其通过简单的几何占用包络线推导得出：对记录稀少的物种使用缓冲区，对拥有足够记录以构建多边形的物种则使用凹包（concave hulls）。本数据集的产出旨在助力热带干旱区域陆生软体动物生物多样性知识缺口、空间采样缺口及保护规划相关研究的可重复性、透明度提升与后续研究开展。