De-identified tree-level dataset on mortality, recruitment, vigor dynamics and local proximity variables of the native wild population of Metasequoia glyptostroboides
Description
This dataset provides a de-identified tree-level database for the native wild population of Metasequoia glyptostroboides Hu & W.C. Cheng in its natural range in south-central China. The dataset was compiled from historical records, long-term field patrol archives and repeated censuses conducted in 2007, 2017 and 2023. It supports analyses of demographic vulnerability, mortality processes, natural recruitment and local environmental exposure in the remnant native population of this endangered relict tree species. The dataset includes three main components. First, it provides records for 153 dead native mother trees, including the census period of death and field-classified proximate causes of mortality. These mortality categories represent field-based proximate causes rather than complete mechanistic diagnoses. Second, it includes natural recruitment records from the 2023 census, including seedlings and saplings recorded beneath native mother trees. Third, it provides a de-identified tree-level table for surviving and registered mother trees, including vigor status across census years and binary local proximity variables indicating whether roads, farmland, water bodies or residences occurred within 40 m of each tree. For conservation sensitivity and data-protection reasons, original mother-tree tag numbers, geographic coordinates and dendrometric measurements, including DBH, tree height and crown traits, are not provided. The tree IDs in this dataset are anonymized and randomly reordered and do not correspond to the original field tag numbers. This de-identified version is intended to allow reproducibility of the main demographic and spatial-association analyses while protecting sensitive location and individual-tree information for this endangered species.
Files
Steps to reproduce
This dataset was compiled from historical census records, long-term patrol archives and repeated field censuses of the native wild population of Metasequoia glyptostroboides Hu & W.C. Cheng in south-central China. Historical baseline records from 1972–1974 and 1983 were used to define the registered native mother-tree population. Follow-up censuses were conducted in 2007, 2017 and 2023 in collaboration with local conservation authorities. During each census, registered native mother trees were relocated and checked in the field. Tree vigor status was assessed according to the forestry industry standard Authentication of Ancient and Famous Trees (LY/T 2737–2016), which classifies growth condition into four categories: vigorous, declining, critical and dead. For each tree, local proximity variables were recorded as binary indicators showing whether roads, farmland, water bodies or residences occurred within 40 m. These variables were obtained from field inspection and distance measurement. Mortality records were compiled from long-term patrol archives and field verification. Each dead mother tree was assigned to a field-classified proximate mortality category based on visible evidence, patrol records and site conditions. These categories represent proximate causes of death rather than complete mechanistic diagnoses. Natural recruitment was surveyed in 2023. For each native mother tree, the area beneath and around the tree was searched for identifiable natural regeneration. Seedlings were defined as individuals <1.5 m in height, and saplings as individuals 1.5–5 m in height. The dataset reports recruitment occurrence and juvenile counts associated with registered mother trees. To protect sensitive information for this endangered species, original tree tag numbers, geographic coordinates and dendrometric measurements, including DBH, tree height and crown traits, were removed. The tree IDs in the public dataset are anonymized and randomly reordered. The de-identified tables can be used to reproduce the main descriptive analyses, including mortality counts, proximate mortality composition, recruitment incidence, juvenile counts and changes in vigor status across census years. The 40-m proximity variables can also be used for tabular comparisons and regression analyses relating vigor or mortality status to local proximity variables. Analyses requiring exact geographic coordinates, such as spatial point-pattern or nearest-neighbour clustering tests, cannot be fully reproduced from the public de-identified dataset alone.
Institutions
- Hubei University for NationalitiesHubei, Enshi
- Chinese Academy of ForestryBeijing, Beijing