Data
We used data from the National Ecological Observatory Network (NEON, National Ecological Observatory Network. 2020), the Botanical Information and Ecology Network (BIEN, Maitner et al., 2020, Enquist et al., 2009) and TRY (Kattge et al., 2020) to link information on leaf traits, species identity, and approximate locations for individual trees. We used Foliar Physical and Chemical Properties (DP1.10026.001) and Vegetation Structure data (DP1.10098.001) from NEON to build joint trait distribution models with environmental drivers (climate and topography) alone, phylogenetic drivers (species identity and phylogeny) alone, and both (combined model). Linking the two different NEON datasets produced individual tree data with stem geolocation and measures of eight leaf traits (LMA, chlorophyll A and B, carotenoids, lignin, cellulose, C, N) for 542 trees in 21 sites across the US (Figure S.1). Since foliar trait concentrations vary significantly with phenology and canopy position (Niinemets et al., 2015), foliar samples were collected at the “peak of greenness” and from the sunlit portion of the canopy. We tested the generalizability of our approach outside of NEON by evaluating predictions from independent (out of sample) data available from the BIEN and TRY datasets (Appendix S1). These two datasets provide measures for C, N and LMA for a total of 223 individual trees. We used data from the Open Tree of Life (Redeling,s 2017) to measure phylogenetic distance between species.
Data for environmental drivers included average monthly climate data from 1995 to 2015 (Appendix S1) extracted from Daymet (Thornton et al., 2018) and topographic variables (elevation, slope and aspect) reported in the NEON and FIA datasets. For three common eastern US tree species (Acer rubrum , Fagus grandifolia , and Abies balsamea ), we used all publicly available leaf N% data from the TRY database to quantify intraspecific variation in leaf N% across each species’ geographic range in the US. We selected these three species because: (1) Abies balsamea is the needleleaf species with the most leaf N% data in TRY for the US; (2) Fagus grandifolia is the broadleaf species with the most leaf N% data in TRY for the USA; and (3) Acer rubrum occurs throughout much of the eastern US in a wide variety of habitats (e.g., from xeric to mesic; Burns and Honkala 1990) and has abundant leaf N% data in TRY. We combined our trait modeling approach with forest survey data from the Forest Inventory and Analysis (FIA) database (https://www.fia.fs.fed.us/) to estimate traits for all individual trees surveyed in the FIA across the eastern USA from 2016 to 2019. We used Lv.3 ecoregions and Lv. 2 ecoprovinces as defined by the Environmental Protection Agency (McMahon 2001, Omernik et al., 2014) to analyze trait distributions at different scales.