Data Documentation for Predicting Recovery paper General Information Name of data set: Predicting Recovery Name of data files in data set predicting_recovery.csv Date the data set was last modified: June 2018 Funder: National Science Foundation How to cite data: K. D. Holl, J. L. Reid, F. Oviedo-Brenes, A. J. Kulikowski, and R. A. Zahawi. 2018. Rules of thumb for predicting tropical forest recovery **** Methodology for data collection: For full details of experimental design and data collection see Holl et al. 2018. Tree recruits and vegetation cover The study was conducted at 13, 50×50 m plots spread across an ~100 km2 agriculture-forest landscape between the Las Cruces Biological Station (LCBS; 8° 47' 7" N; 82° 57' 32" W) and Agua Buena (8° 44' 42" N; 82° 56' 53" W) in southern Costa Rica. We recorded vegetation measurements (i.e., grass cover and woody recruitment) annually beginning 0.5 years after plot abandonment (starting in 2007 to 2009 depending on the site). We sampled tree recruits in all plots using a stratified sampling procedure with area sampled adjusted for recruit size. In four permanent, nested quadrats at each site, we recorded tree seedlings (>=0.2 and <1 m tall) in eight adjacent 1×1 m quadrats; saplings (>=1 m tall and <5 cm diameter at breast height (DBH) in four adjacent 2×4 quadrats; and small trees (>=5 cm and <10 cm DBH) in 8×8 m quadrats. Trees >=10 cm DBH were recorded throughout the entire 50×50 m plot. We estimated total percent grass cover, as well as cover of individual grass species with >5% cover, in alternating 1×1 m seedling quadrats (n = 16 per plot) using a modified Braun-Blanquet cover-abundance scale: 0, 1-5, 5-10, 10-25, 25-50, 50-75, 75-95, and 95-100%. We recorded canopy closure over the same quadrats by taking spherical densiometer measurements in four directions and averaging values. We categorized individual grass species into forage (mostly aggressive non-native species) and non-forage species (mostly native species). Surface soil nutrients In August 2007, we collected 25, 2.5 × 15-cm soil cores across each plot. Cores were mixed, air dried, and then passed through a 2-mm sieve. Samples were analyzed for pH, organic matter and Mehlich P all macronutrients following standard procedures at Brookside Laboratories, New Bremen, OH (www.blinc.com/resources/testing-methods). Landscape forest cover Forest cover within 100- and 500-m radius from the center of each plot was hand-digitized from orthorectified 2005 aerial photographs (Cole et al. 2010). Tree cover surrounding plots at the start of the study ranged from (0-61%) and (11-89%) at 100- and 500-m buffers, respectively. Primary data collector(s): Karen Holl, Federico Oviedo-Brenes, Rakan Zahawi, J. Miguel Chaves-Fallas, Juan Abel Rosales Date of data collection: June-July 2007-2017 Software (including version #) used to prepare data set: Microsoft Excel 2017. Data processing that was performed: Sites were abandoned in three subsequent years (2007-2009) so we standardized data to 0.5, 1.5, 2.5 and 8.5 years following abandonment. Recruit richness and density were calculated at the site level. Change in values over time were calculated by subtracting the value in the year 0.5, 1.5, or 2.5 from values in year 8.5. Person/people who entered data: Karen Holl either entered or supervised a host of different people entering the data. Person/people who proofed data: Karen Holl either proofed or supervised a host of different people who proofed the data over several years. Person to contact with questions: Karen Holl (kholl@ucsc.edu) Variables in predicting_recovery.csv site: two-letter code denoting study site. recdy0.5: recruit density per meter squared 0.5 years after abandonment recdy1.5: recruit density per meter squared 1.5 years after abandonment recdy2.5: recruit density per meter squared 2.5 years after abandonment recdy8.5: recruit density per meter squared 8.5 years after abandonment sry0.5: recruit species richness per site 0.5 years after abandonment sry1.5: recruit species richness per site 1.5 years after abandonment sry2.5: recruit species richness per site 2.5 years after abandonment sry8.5: recruit species richness per site 8.5 years after abandonment grcy0.5: percent grass cover 0.5 years after abandonment grcy1.5: percent grass cover 1.5 years after abandonment grcy2.5: percent grass cover 2.5 years after abandonment grcy8.5: percent grass cover 8.5 years after abandonment forgrcy1.5: percent forage grass cover 1.5 years after abandonment ccy0.5: percent canopy closure 0.5 years after abandonment ccy1.5: percent canopy closure 1.5 years after abandonment ccy2.5: percent canopy closure 2.5 years after abandonment ccy8.5: percent canopy closure 8.5 years after abandonment CEC2007: cation exchange capacity (me 100 per g)in 2007 OMper2007: percent soil organic matter in 2007 PMehl2007: Mehlich P (mg per kg) in 2007 perfor100m: percent forest cover at 100-m radius perfor500m: percent forest cover at 500-m radius rankpasture: ranked years of pasture use, higher numbers were used longer rankclear: ranked years since forest cleared, higher numbers were cleared earlier elevm: site elevation in meters