conspecific negative density dependence across tree life stage in global forest communities

Paper searchWe limited our synthesis to studies that focused on natural forest ecosystem. A search was conducted to collect all the studies cite Janzen (1970) and has contains “forest*” in title, abstract or keywords in the Web of Science (https://clarivate.com/webofsciencegroup/), and CNKI (a Chinese database; https://www.cnki.net/) through to 10th February 2025. We harvested totally 1479 papers. From these papers, we first screened the titles and abstracts to select papers on the correct topic. We then examined each paper to determine whether it met the following criteria: (1) it focused on individual survival or growth; (2) it included both conspecific neighbour density and heterospecific (or phylogenetic weighted) neighbour density as explanatory variables in one model; (3) the study was conducted at community level, but not for specific species. We finally yield 50 papers. In each selected paper, we focused on three different type of density dependent studies: (1) Seedling-seedling neighborhood interactions within seedling plots, The target individuals are seedlings located within designated seedling plots. For each seedling, the explanatory variables are the density of conspecific and heterospecific seedling neighbors within the same seedling plot. (2) Tree-seedling neighborhood interactions within certain radius, the target individuals are seedlings located within designated seedling plots. For each seedling, the explanatory variables are the density of conspecific and heterospecific tree neighbors (including saplings and adult trees) within a circular neighborhood of a fixed radius centered on the seedling plot. (3) Tree-tree neighborhood interactions within certain radius, the target individuals are trees (including saplings and adults) within a forest plot. For each tree, the explanatory variables are the density of conspecific and heterospecific tree neighbors within a circular neighborhood of a certain radius centered on the focal tree. For papers that contained multiple pair of conspecifics and heterospecific neighbour densities, for example, a model that included both seedling-seedling neighbour densities and tree-seedling neighbour densities, each pair of densities was considered separately, and for papers that tested multiple life stages, each life stage was considered independently.  Data extractionFor each study, we extracted the mean, standard error (or confidence interval) of each coefficient of density variables. For the studies that only reported the mean value of confidence, we only recorded so. For the studies that reported odd ratio as coefficients, we have log-transformed odd ratio into log odd ratio. Information from studies where results were only presented as graphs were extracted using Plot Digitizer (http://plotdigitizer.sourceforge.net/). Other key variables extracted were life-history stage (i.e., seedling or tree), as well as the stages of seedlings (early-stage and late-stage) and the stages of tree (sapling, juveniles and adults). We also recorded metadata and covariates including author, year, study site location, and, the size of seedling plots. We recorded the latitude, mean annual temperature (MAT), and mean annual precipitation (MAP) for each study. To assess the influence of aridity on the geographic variation in PSF, we extracted aridity values from the Global Aridity Index (version 3; Zomer et al., 2022).