Detail of the variables included in the Lipoma_etal_database_2019.txt file Field separator character = "," Character used in the file for decimal points = "." More detailes about mothods in: Lipoma, M.L.; Cuchietti, A.; Gorné, L.D and Díaz, S.M, 2019. Not gone with the wind: Vegetation complexity increases seed retention during windy periods in the Argentine Semiarid Chaco. Accepted for publication in Journal of Vegetation Science Varibles: year ET Transect Quadr SeedType Tot.Ret.pro NseedType LseedType ST.Ret.per ST.Ret.pro PC1 PC2 Year: year of the experiment, categorical variable ET: Ecosystem type, categorical variable. Levels: PF (Primary forest), SF (Secondary forest), CS (Closed species-rich shrubland), OS (Open shrubland). Transect: Transect in each site where the experiment was developed, categorical variabel. Levels: 1-16 Quadr: quadrats from each transect where the experiment was developed, categorical varaibel. Levels: 1-160. SeedType: each seed type used in the experiment, categorical variable. Levels: LF (large and flattened) and SS (small and spherical). Tot.Ret.pro: Proportion of total seed retention, response variable. Range: 0-1. NseedType: Number of seeds retain by each seed type, responde variable. Range: 0-20. LseedType: Numer of seeds lost by each seed type, 20-NseedType, response variable. Range: 0-20. ST.Ret.per: Percentage of seed retention for each seed type, response variable. Range: 0-100. ST.Ret.pro: Proportion of seed retention for each seed type, response variable. Range: 0-1. PC1: Position of each quadrat along the PC1 axe. PC1 captured vegetation structure, with higher scores indicating higher % of ground cover and more complex vertical structure PC2: Position of each quadrat along the PC2 axe. PC2 captured ground cover types, with graminoids and Sellaginella selowii dominating at the positive end, and litter and forbs at the negative end.