Data from: Divergent investment in direct versus indirect defense in Populus tremuloides across a latitudinal gradient
Abstract
Theory predicts that plant defense against herbivores should be greatest towards the equator due to the strong selective pressure of more populous and diverse herbivore communities, yet recent work suggests variation in plant defense is more complex, predicting divergent patterns of investment across defense types and underlying environmental drivers. To address gaps in previous research, we measure multiple defense types in the same system and consider both biotic and abiotic variables hypothesized to underlie latitudinal gradients. Using the tree species trembling aspen (Populus tremuloides), we combine field surveys of biotic interactions and defense traits with climate data and metabolomic analysis of plant chemistry across a continental-scale latitudinal gradient. While some metrics of defense investment were greater at lower latitudes, we found that the environmental variables underlying variation in defense traits differed across defense types, with direct defense decreasing along a precipitation gradient and indirect defense increasing with temperature and growing season length. Incorporating proxies for predation and herbivory pressure further suggested that variation in defensive traits is driven by an interplay between biotic and abiotic variables. Together, our findings reinforce that, even within a single species, different defense strategies can correlate with distinct environmental variables across latitude, consistent with complexity in the selective mechanisms shaping defense variation across space. These results illustrate the importance of considering the full scope of defense strategies and environmental pressures within a system when testing classic hypotheses of plant defense.
https://doi.org/10.5061/dryad.t1g1jwtbc
Description of the data and file structure
These data include a combination of in situ field observations of biotic interactions, including herbivory and predation; quantification of extrafloral nectaries; metabolomics analysis of leaf chemistry; and long-term climate data. All data were collected to address inconsistencies in empirical tests of the Latitudinal Biotic Interactions Hypothesis.
Files and variables
File: data.zip
Description: This file contains all the data utilized in metabolomics processing and downstream statistical analyses and is organized as described below.
- chemistry
- Extraction_table.csv - data file containing extraction numbers of each tree sample, the date they were extracted, and the weight (mg) of the dried leaf material used.
- LCMS_master_list.csv - data file containing extraction number, LC-MS file name, the date samples were run, and file quality (0, 1) for inclusion in further analysis. Raw LC-MS files are included in the attached .ZIP file 'mzML'.
- XCMS processing.csv - data file to keep track of which raw LC-MS samples (.mzML files) have been processed using XCMS (1_xcms_peak_picking.R).
- xcms_feature_table.csv - output file of XCMS processing (1_xcms_peak_picking.R); raw LC-MS chromatographic features and its corresponding retention time, m/z and abundance (TIC) within each sample.
- compound_object_prefeat.pkl - pickle object containing Python dictionary of XCMS processed chromatographic features from LC-MS samples (2_group_features_compounds.py).
- compound_object_precomps.pkl - pickle object containing Python dictionary with chromatographic features grouped across all LC-MS samples (2_group_features_compounds.py).
- features_matched_across_samples.csv - data file containing chromatographic feature IDs identified across samples and their cluster m/z and retention time values (in seconds) (2_group_features_compounds.py).
- compound_feature_table.csv - data file containing the m/z, retention time, TIC, and relative abundance of each chromatographic feature within each compound (2_group_features_compounds.py).
- compound_object_grouped_comps.pkl - pickle object containing Python dictionary with chromatographic features grouped into compounds via m/z and retention time values across all LC-MS samples (2_group_features_compounds.py).
- comp_feat_pcid_mzrt_rtwindow.csv - data file containing specific m/z, retention time, and TIC values for chromatographic features and compounds detected within each sample (2_group_features_compounds.py).
- known_phenolic_glycoside_mass_adducts.csv - data file product from* msac* to calculate possible LC-MS mass adducts of known salicinoids.
- defense_SIRIUS_CANOPUS.csv - output file of CANOPUS Natural Product classifications in SIRIUS for putative defensive compounds (4_SIRIUS_defense_filtering.R). The MGF files uploaded are included in the attached .ZIP file 'MGF'.
- non-defense_SIRIUS_CANOPUS.csv - output file of CANOPUS Natural Product classifications in SIRIUS for putative non-defensive compounds (4_SIRIUS_defense_filtering.R). The MGF files uploaded are included in the attached .ZIP file 'MGF'.
- defensive_comps_in_samps.csv - list of unique defensive compounds found within each sample (5a_comp_rich_abund_diversity_calc.R).
- sampsbycomps_def_raw.csv - samples x compounds abundance table for putative defenses (5a_comp_rich_abund_diversity_calc.R).
- defense_chem_div_abun_rich_rawTIC.csv** **- data file containing the compound richness, chemical abundance, evenness (Pielou's J), and Shannon-Wiener diversity (H) for the defensive chemical profile of each aspen tree sample (5a_comp_rich_abund_diversity_calc.R).
- NONdefensive_comps_in_samps.csv - list of unique non-defensive compounds found within each sample (5a_comp_rich_abund_diversity_calc.R).
- sampsbycomps_NONdef_raw.csv - samples x compounds abundance table for non-defenses (5a_comp_rich_abund_diversity_calc.R).
- NONdefense_chem_div_abun_rich_rawTIC.csv - data file containing the compound richness, chemical abundance, evenness (Pielou's J), and Shannon-Wiener diversity (H) for the non-defensive chemical profile of each aspen tree sample (5a_comp_rich_abund_diversity_calc.R).
- all_comps_in_samps.csv - list of all unique (defenses + non) compounds found within each sample (5a_comp_rich_abund_diversity_calc.R).
- sampsbycomps_all_raw.csv - samples x compounds abundance table for all compounds (5a_comp_rich_abund_diversity_calc.R).
- allcomps_chem_div_abun_rich_rawTIC.csv** **- data file containing the compound richness, chemical abundance, evenness (Pielou's J), and Shannon-Wiener diversity (H) for the entire chemical profile of each aspen tree sample (5a_comp_rich_abund_diversity_calc.R).
- PG_abun_in_samps_spectral_matches.csv - data file with abundances for spectral matches to known salicinoids (5b_known_PGs_presabs_abun.R)
- NONdefense_pairwise_comps.csv - data file containing the pairwise comparisons of non-defensive compounds based on MSMS spectral similarity, created from GNPS data products (6_pairwise_comps_chemsim.R). The MGF files uploaded are included in the attached .ZIP file 'MGF'. GNPS data products are included in the attached .ZIP file 'GNPS'.
- defense_pairwise_comps.csv - data file containing the pairwise comparisons of defensive compounds based on MSMS spectral similarity, created from GNPS data products (6_pairwise_comps_chemsim.R). The MGF files uploaded are included in the attached .ZIP file 'MGF'. GNPS data products are included in the attached .ZIP file 'GNPS'.
- allcompounds_pairwise_comps.csv - data file containing the pairwise comparisons of all compounds (defense + non) based on MSMS spectral similarity, created from GNPS data products (6_pairwise_comps_chemsim.R). The MGF files uploaded are included in the attached .ZIP file 'MGF'. GNPS data products are included in the attached .ZIP file 'GNPS'.
- ramet_similarity_defenses.csv - data file of compositional-structural cosine similarity (CSCS) comparisons between samples based on the abundance of shared defensive compounds and the spectral similarity of non-shared defensive compounds (6_pairwise_comps_chemsim.R).
- ramet_similarity_NONdefenses.csv - data file of compositional-structural cosine similarity (CSCS) comparisons between samples based on the abundance of shared non-defensive compounds and the spectral similarity of non-shared non-defensive compounds (6_pairwise_comps_chemsim.R).
- ramet_similarity_all_compounds.csv - data file of compositional-structural cosine similarity (CSCS) comparisons between samples based on the abundance of all shared compounds and the spectral similarity of all non-shared compounds (6_pairwise_comps_chemsim.R).
- MGF.zip - contains five subfolders with the MGF files of large molecules (>850 Da) and small molecules for SIRIUS processing, as well as known phenolic glycosides (PGs), SIRIUS defenses, and SIRIUS non-defenses for post SIRIUS data processing. Files are named with compound number, feature number, and source file separated by hyphens (e.g., 1-6842_0-E103.mgf). The same MGF file may appear in up to two folders (e.g., in SIRIUS defenses and large molecules).
- GNPS.zip - contains three subfolders for defenses, non-defenses, and all compounds based on SIRIUS annotation. All files result directly from the output of GNPS completed job download 'View All Raw Spectra'.
- mzML.zip - contains all original mzML files converted from raw data format. Files are named with prefixes according to type where BLANK = solvent blank for contaminant removal, QC = quality control sample admixture for retention time and mass corrections, and E = raw samples. File suffixes are ordered according to when they were processed via LC-MS starting at 001.
- climate
- DayMet_BM_highres_1980-2022.csv - daily weather data for the site outside of Butte, Montana in Beaverhead-Deerlodge National Forest, downloaded from DayMet (from homepage > Get Data > Singe Pixel Extraction Tool > Coordinates = Enter Latitude & Longitude [averages for BM site] > Variables = ALL > Years = 1980 - 2022 > Download).
- DayMet_GM_highres_1980-2022.csv - daily weather data for the site outside of West Glacier, Montana in Flathead National Forest, downloaded from DayMet (from homepage > Get Data > Singe Pixel Extraction Tool > Coordinates = Enter Latitude & Longitude [averages for GM site] > Variables = ALL > Years = 1980 - 2022 > Download).
- DayMet_FA_highres_1980-2022.csv - daily weather data for the site outside of Flagstaff, Arizona in Coconino National Forest, downloaded from DayMet (from homepage > Get Data > Singe Pixel Extraction Tool > Coordinates = Enter Latitude & Longitude [averages for FA site] > Variables = ALL > Years = 1980 - 2022 > Download).
- DayMet_PI_highres_1980-2022.csv - daily weather data for the site outside of Pocatello, Idaho in Caribou-Targhee National Forest, downloaded from DayMet (from homepage > Get Data > Singe Pixel Extraction Tool > Coordinates = Enter Latitude & Longitude [averages for PI site] > Variables = ALL > Years = 1980 - 2022 > Download).
- DayMet_TA_highres_1980-2022.csv - daily weather data for the site outside of Tucson, Arizona in Coronado National Forest, downloaded from DayMet (from homepage > Get Data > Singe Pixel Extraction Tool > Coordinates = Enter Latitude & Longitude [averages for TA site] > Variables = ALL > Years = 1980 - 2022 > Download).
- DayMet_WU_highres_1980-2022.csv - daily weather data for the site outside of Widtsoe, Utah in Dixie National Forest, downloaded from DayMet (from homepage > Get Data > Singe Pixel Extraction Tool > Coordinates = Enter Latitude & Longitude [averages for WU site] > Variables = ALL > Years = 1980 - 2022 > Download).
- DayMet_SLU_highres_1980-2022.csv - daily weather data for the site outside of Salt Lake City, Utah in Uintah-Wasatch-Cache National Forest, downloaded from DayMet (from homepage > Get Data > Singe Pixel Extraction Tool > Coordinates = Enter Latitude & Longitude [averages for SL site] > Variables = ALL > Years = 1980 - 2022 > Download).
- site_annual_climate_data_1980-2022.csv - daily weather data for each site averaged per year from 1980-2022 (climate_data_per_site.R).
- field observations
- tree_location.csv - raw observations of tree location, elevation, and size data. Units provided in column headers.
- location_lat_elevation_corrections.csv - tree location, elevation, and size data with additional columns for location and elevation corrections (1_latitude_elevation_correction.R).
- EFN.tsv - raw observations of extrafloral nectary abundance per leaf for 9 leaves per aspen tree/ramet.
- herbivory.tsv - raw observations of percent herbivory and herbivory type per leaf for 15-20 leaves per aspen tree/ramet.
- predation.tsv - raw observations of predation presence / absence and type per clay model insect attached to 5 leaves per aspen tree/ramet.
Column descriptions - Raw data files
Description: Data dictionary for columns in raw field, climate, and chemistry data. Downstream data products in chemistry pipeline and averaged climate data are not included because they are described in code annotations.
- chemistry
- Extraction_table.csv
- date: Date of extraction.
- sample_ID: Unique identifier for each aspen ramet (clone_ramet, i.e., BM04_R2).
- extraction_number: Order of sample extraction for chemical analysis.
- empty_vial_wt_mg: Mass of empty microcentrifuge tube before adding dried leaf powder.
- vial_ground_sample_mg: Combined mass of microcentrifuge tube and dried leaf powder sample.
- sample_wt: Mass of sample (vial_ground_sample_mg - empty_vial_wt_mg).
- vial_leftover_sample: Combined mass of microcentrifuge tube and sample post-extraction and drying.
- dry_marc: Remaining sample mass post-extraction (vial_leftover_sample - empty_vial_wt_mg)
- percent_extracted: Percentage of sample mass lost to extraction solvent ([sample_wt - dry_marc] / sample_wt).
- QC_mixture: List of samples (extraction_number) used in QC for associated LC-MS runs.
- grinding_method: Method used to grind dried plant material to powder. BM = ball mill, CM = cyclone mill, MP = mortar + pestle.
- LC_data_check: Binary indicating whether the sample was run on LC (0 = no, 1 = yes).
- notes: Notes about sample processing including weighing, extraction, and LC-MS.
- LCMS_master_list.csv
- sampleID: Unique identifier for samples (extraction_number) + QCs + Blanks in order of sample run.
- sample_type: Type of extraction run including Sample, QC, and Blank.
- asoc_blank: Blank associated with each sample; used for baseline removal downstream.
- asoc_qc: QC associated with each sample; used for retention time and signal correction downstream.
- sample_position: Position of LC vial in tray during runs; 'No Injection' = methanol solvent directly from LC line.
- date: Date of LC-MS data collection.
- quality: Binary indicating whether the sample is included in downstream analysis (0 = no, 1 = yes).
- Extraction_table.csv
- climate
- DayMet_BM_highres_1980-2022.csv
- year: Year of data collection.
- yday: Julian date of data collection for given year (above).
- dayl_sec: Amount of daylight at site on date of collection in seconds.
- prcp_mm: Amount of precipitation at site on date of collection in millimeters.
- srad_W_m2: Amount of solar radiation at site on date of collection in Watts/sq meter.
- swe_kg_m2: Mass of water contained in snow pack (snow water equivalent) at site on date of collection in kilograms/sq meter.
- tmax_C: Maximum temperature at site on date of collection in degrees Celsius.
- tmin_C: Minimum temperature at site on date of collection in degrees Celsius.
- vp_Pa: Amount of vapor pressure at site on date of collection in Pascals.
- DayMet_GM_highres_1980-2022.csv
- Same as 'DayMet_BM_highres_1980-2022.csv'
- DayMet_FA_highres_1980-2022.csv
- Same as 'DayMet_BM_highres_1980-2022.csv'
- DayMet_PI_highres_1980-2022.csv
- Same as 'DayMet_BM_highres_1980-2022.csv'
- DayMet_TA_highres_1980-2022.csv
- Same as 'DayMet_BM_highres_1980-2022.csv'
- DayMet_WU_highres_1980-2022.csv
- Same as 'DayMet_BM_highres_1980-2022.csv'
- DayMet_SLU_highres_1980-2022.csv
- Same as 'DayMet_BM_highres_1980-2022.csv'
- DayMet_BM_highres_1980-2022.csv
- field observations
- tree_location.csv
- clone: Unique identifier for each clone (site code + clone number at each site).
- date: Date of field observations and leaf collections for chemistry.
- ramet_num: Ramet number per clone (R1-R4).
- lat: Latitude in decimal degrees.
- long: Longitude in decimal degrees.
- elevation_ft: Elevation in feet.
- DBH_cm: Diameter at breast height for each aspen ramet in centimeters.
- notes: Notes about sample collection.
- location_lat_elevation_corrections.csv
- clone_ram: Unique identifier for each aspen tree (ramet); clone + ramet_num from 'tree_location.csv'.
- site: Site code for each collection location; BM = Butte, Montana, GM = Glacier, Montana, TA = Tucson, Arizona, FA = Flagstaff, Arizona, PI = Pocatello, Idaho, SL = Salt Lake City, Utah, WU = Widtsoe, Utah.
- lat: Latitude in decimal degrees.
- long: Longitude in decimal degrees.
- elevation_m: Elevation in meters.
- lat_elev_corrected: Latitude in decimal degrees with 1 degree added for every 100 m of elevation (similar to Ashton et al. 2016).
- elev_site_scale: Elevation of each site scaled using the 'scale' function in R.
- DBH_cm: Diameter at breast height for each aspen ramet in centimeters.
- EFN.tsv
- clone: Unique identifier for each clone (site code + clone number at each site).
- date: Date of EFN data collection.
- ramet_num: Ramet number per clone (R1-R4).
- excision_num: Number of twig excised from canopy for EFN count (3 per ramet).
- leaf_position: Position of leaf on twig used for EFN count (1, 3, 5).
- EFN: Presence (1) vs. absence (0) of EFNs on each leaf.
- count: Number of EFNs on each leaf.
- notes: Notes about leaves used for EFN data collection.
- herbivory.tsv
- clone: Unique identifier for each clone (site code + clone number at each site).
- plant_num: Ramet number per clone (H1-H4).
- date: Date of herbivory data collection.
- size: Categorical description of size class for tree; sap = sapling, juv = juvenile, YA = young adult, adult = adult.
- location: Description of tree location within aspen clone (i.e., proximity to edge).
- leaf_num: Number of each leaf recorded for herbivory (1-15 or 1-20; subsetted to 15 in downstream analysis).
- percent_herb: Percentage of leaf damaged due to herbivory (0 - 100).
- type: Categorical description of damage type (chew = chewing, LC = light chewing, roll = rolling, fold = folding, fung = fungi, mine = leaf miner).
- notes: Notes about leaves used for herbivory data collection.
- predation.tsv
- clone: Unique identifier for each clone (site code + clone number at each site).
- date: Date of predation data collection.
- plant_num: Ramet number per clone (P1 - P4).
- larva_num: Clay caterpillar number per ramet (1-5).
- size: Categorical description of size class for tree; sap = sapling, juv = juvenile, YA = young adult, adult = adult.
- location: Description of tree location within aspen clone (i.e., proximity to edge).
- predation: Categorical description of predation presence and type; arth = arthropod, bird = bird, none = predation not observed, N/A = clay larva missing.
- notes: Notes about predation data collection.
- tree_location.csv
Code/software
File: code.zip
Software required: R, Python
Description: This file contains all the code utilized in metabolomics processing and downstream statistical analyses and is organized as described below.
- chemistry
- 1_xcms_peak_picking.R - raw LC-MS chromatographic data processing; peaks picked for samples within clones for optimized iteration.
- Input: .mzML files for aspen samples (E001, E002, etc.), blanks (BLANK), and quality control (QC).
- Output: xcms_feature_table.csv, XCMS_processing.csv
- Requires: build_compound_table_with_classes.py
- 2_group_features_compounds.py - grouping of peaks into features and compounds across aspen samples using a cosine similarity of m/z, retention time, abundance.
- Input: xcms_feature_table.csv
- Output: compound_object_prefeat.pkl, compound_object_precomps.pkl, features_matched_across_samples.csv, compound_feature_table.csv, compound_object_grouped_comps.pkl, comp_feat_pcid_mzrt_window.csv
- 3a_MSMS_dataexport_samples.R - identify and mine the MSMS spectral data for each compound's highest abundance feature.
- Input: .mzML files for aspen samples (E001, E002, etc.), compound_feature_table.csv, comp_feat_pcid_mzrt_rtwindow.csv
- Output: .MGF files for each compound, MSMS_match_processing_check.csv, precursor_mass_all_compounds.csv
- 3b_MSMS_dataexport_knownPGs.R - mine MSMS spectral data for known salicinoids based on possible mass adducts.
- Input: .mzML files for known salicinoids, known_phenolic_glycoside_mass_adducts.csv
- Output: .MGF files for known salicinoid spectral data matched to a potential adduct.
- 4_SIRIUS_defense_filtering.R - separation of compounds into putative defenses and non-defenses based on SIRIUS and CANOPUS data.
- Input: CANOPUS formula summary and formula id output files from uploading .MGFs to SIRIUS GUI.
- Output: defense_SIRUS_CANOPUS.csv, non-defense_SIRIUS_CANOPUS.csv
- 5a_comp_rich_abund_diversity_calc.R - calculations of chemical diversity and abundance metrics per aspen tree sample, separated by defenses vs. non.
- Input: comp_feat_pcid_mzrt_rtwindow.csv, defense_SIRIUS_CANOPUS.csv, non-defense_SIRIUS_CANOPUS.csv, Extraction_table.csv
- Output: defense_chem_div_abun_rich_rawTIC.csv, sampsbycomps_def_raw.csv, defensive_comps_in_samps.csv, NONdefense_chem_div_abun_rich_rawTIC.csv, sampsbycomps_NONdef_raw.csv, NONdefensive_comps_in_samps.csv, all_comps_in_samps.csv, allcomps_chem_div_abun_rich_rawTIC.csv, sampsbycomps_all_raw.csv
- 5b_known_PGs_presabs_abun.R - matching of compounds in samples to known salicinoids and calculations of presence/absence and total abundance of known salicinoids in aspen tree samples.
- Input: comp_feat_pcid_mzrt_rtwindow.csv, defense_SIRIUS_CANOPUS.csv, PGs_detected_peaks_SIRIUS.csv, known_phenolic_glycoside_mass_adducts.csv
- Output: PG_abun_in_samps_spectral_matches.csv, PG_abun_in_samps_theoretical_matches.csv
- 6_pairwise_comps_chemsim.R - calculation of chemical similarity between samples based on abundance of shared compounds and the spectral similarity of non-shared compounds.
- Input: GNPS output files of cosine similarity based on compound MSMS spectra, sampsbycomps_def_raw.csv, sampsbycomps_NONdef_raw.csv, sampsbycomps_all_raw.csv, defense_chem_div_abun_rich_rawTIC.csv, NONdefense_chem_div_abun_rich_rawTIC.csv, allcomps_chem_div_abun_rich_rawTIC.csv, Extraction_table.csv
- Output: ramet_similarity_defenses.csv, ramet_similarity_NONdefenses.csv, ramet_similarity_all_compounds.csv, allcompounds_pairwise_comps.csv, NONdefense_pairwise_comps.csv, defense_pairwise_comps.csv
- Requires: chem_similarity_function.R, create_pairwiseComps.R
- 1_xcms_peak_picking.R - raw LC-MS chromatographic data processing; peaks picked for samples within clones for optimized iteration.
- climate
- climate_data_per_site.R - calculation of annual averages for weather data from DayMet for each sampling site.
- Input: DCS_climate.csv, DayMet_WU_highres_1980-2022.csv, DayMet_SLU_highres_1980-2022.csv, DayMet_BM_highres_1980-2022.csv, DayMet_GM_highres_1980-2022.csv, DayMet_TA_highres_1980-2022.csv, DayMet_FA_highres_1980-2022.csv, DayMet_PI_highres_1980-2022.csv
- Output: site_annual_climate_data_1980-2022.csv
- climate_data_per_site.R - calculation of annual averages for weather data from DayMet for each sampling site.
- latitude_elevation_correction
- 1_latitude_elevation_correction.R - calculate latitude corrected for elevation and scaled per location.
- Input: tree_location.tsv
- Output: location_lat_elev_corrections.csv
- 1_latitude_elevation_correction.R - calculate latitude corrected for elevation and scaled per location.
- statistical_analysis_make_figures
- 1a_defense_latitude.R - GLMM models of defense measurements (chem + EFNs) as a function of latitude.
- Input: location_lat_elev_corrections.csv, PG_abun_in_samps_spectral_matches.csv, defense_chem_div_abun_rich_rawTIC.csv, Extraction_table.csv, EFN.tsv
- Output: Figures 2a-c, Table S1
- 1b_environment_latitude.R - GLMM models of environmental variables (biotic + climate) as a function of latitude.
- Input: location_lat_elev_corrections.csv, herbivory.tsv, predation.csv, site_annual_climate_data_1980-2022.csv
- Output: Table S1
- 2a_defense_biotic.R - GLMM models of defense measurements (chem + EFNs) as a function of biotic interactions.
- Input: location_lat_elev_corrections.csv, defense_chem_div_abun_rich_rawTIC.csv, PG_abun_in_samps_spectral_matches.csv, Extraction_table.csv, EFN.tsv, herbivory.tsv, predation.csv
- Output: Figure 3d, Table S3
- 2b_defense_abiotic.R - GLMM models of defense measurements (chem + EFNs) as a function of climate.
- Input: location_lat_elev_corrections.csv, defense_chem_div_abun_rich_rawTIC.csv, PG_abun_in_samps_spectral_matches.csv, Extraction_table.csv, EFN.tsv, site_annual_climate_data_1980-2022.csv
- Output: Figures 3a-c,e-f, Table S3
- 3_chemical_similarity.R -
- Input: location_lat_elev_corrections.csv, defense_chem_div_abun_rich_rawTIC.csv, Extraction_table.csv, ramet_similarity_defenses.csv, ramet_similarity_NONdefenses.csv, site_annual_climate_data_1980-2022.csv, herbivory.tsv, predation.csv
- Output: Figure 2d, Table S2
- 4_defense_tradeoffs.R - GLMM models of chemical defense measurements as a function of EFN abundance.
- Input: location_lat_elev_corrections.csv, defense_chem_div_abun_rich_rawTIC.csv, PG_abun_in_samps_spectral_matches.csv, Extraction_table.csv, EFN.tsv
- Output: Table S4
- 5_non-defense_chem_environment.R - all above described measurements and models used for chemical defenses (except chemical similarity) as a function of latitude, climate, and biotic interactions data, repeated for non-defensive chemistry.
- Input: location_lat_elev_corrections.csv, NONdefense_chem_div_abun_rich_rawTIC.csv, Extraction_table.csv, site_annual_climate_data_1980-2022.csv, herbivory.tsv, predation.csv
- X_site_map.R - creation of site map + elevation across latitude inlay.
- Input: GADM / RDS data from rworldmap package, tree_location.tsv
- Output: Figure 1 (map)
- 1a_defense_latitude.R - GLMM models of defense measurements (chem + EFNs) as a function of latitude.
Supplemental Data/Figures
Files: figures.zip, Supplementary Tables.docx
Description: These files contains all figures and tables generated from the data and analysis described above.
Access information
Data was derived from the following sources:
- Original data, no license required
- SIRIUS (CANOPUS), Academic License
- GNPS, License
- Daymet, Creative Commons CC0 (Public Domain)
