Heavy metals concentrations often vary at small spatial scales not captured by air monitoring networks, with implications for environmental justice in industrial-adjacent communities. Pollutants measured in moss tissues are commonly used as a screening tool to guide use of more expensive resources, like air monitors. We piloted a community science approach, engaging over 55 people from nine institutions, to map heavy metals using moss in two industrial-adjacent neighborhoods. Local youth led sampling of the moss Orthotrichum lyellii from trees across a 250×250-m sampling grid (n = 79). We compared their data with expert-collected samples (n = 19) using Principal Components Analysis and Procrustes Analysis. We mapped 21 chemical elements measured in moss, focusing on 6 toxic ‘priority’ metals: arsenic, cadmium, chromium, cobalt, lead, and nickel. We compared local data, using t-tests and boxplots, with two ‘reference datasets’ of O. lyellii collected in Portland, Oregon, and in Seattle City Parks. We also use Principal Components Analysis to describe major gradients in metals in the study area. Our data submission includes two R scripts and four datasets of heavy metals in moss, including the two reference datasets, which will enable replication of our analyses as well as novel analyses.

SAMPLING (For main dataset. Reference datasets are similar. Citations provided in metadatafile)

Moss sampling protocols were adapted from prior studies using the widespread epiphytic (i.e. “tree-dwelling”) moss species Orthotrichum lyellii (Donovan et al. 2016, Gatziolis et al. 2016). The Duwamish Valley Youth Corps (DVYC) and adult study partners collected the moss from a 250×250-m grid across two industrial-adjacent neighborhoods in Seattle's Duwamish Valley (Georgetown and South Park) with longstanding air pollution concerns. Moss was collected at the suitable tree nearest the centroid of each grid cell. The DVYC led moss sampling excursions on four warm, dry days in 2019 (May 25^th, June 1^st, 4^th, and 8^th). Participants met the following week in a local high school science laboratory to harvest the upper 2/3^rds of living moss stems for heavy metals analysis (Gatziolis et al. 2016). To assess sampling precision, the youth-led teams immediately collected a replicate moss sample at 18 sites where ample moss was available. Their final analytical dataset had 79 samples from 61 grid cells. To check sampling accuracy, experts re-sampled 19 grid cells although this occurred about 2 weeks later on June 13, 2019 due to scheduling difficulties.

LAB METHOD (For main dataset. Reference datasets are similar. Citations provided in metadatafile)

All moss samples were sealed and mailed to the US Forest Service Grand Rapids, MN laboratory where they underwent the same treatment. Samples were prepared for heavy metals analysis by oven drying at 40 °C for 24 h and homogenizing by grinding to a fine powder (IKA tube-mill, 1 min grinding time for each sample at 15,000 rpm). A 0.500-g subsample of each moss sample was processed using a modified microwave-assisted digestion with 10 mL concentrated HNO₃ + 2 mL 30% H₂O₂ + 2 mL concentrated HCl (CEM, 2019). An overnight pre-digestion of the samples with added reagents was done at room temperature. Following the microwave-assisted digestion cycle, digests were transferred by rinsing with deionized water to 50 mL volumetric flasks, diluted to volume with deionized water, and filtered through 0.45-um membrane filters into plastic storage bottles prior to analysis. Concentrations of 25 elements in total were measured by inductively coupled plasma optical emission spectrophotometry (Thermo 7000 series dual-view (axial and radial) ICP-OES).

Quality control steps included use of method blanks, instrument calibration standards, instrument performance check standards, and reference lichen samples. The measurement quality objectives (MQOs) are the same as the confidence and tolerance levels that accompany each standard reference material or check standard certification sheet. Quality control/quality assurance (QC/QA) for air-quality sampling in our study followed the EPA Quality Assurance Guidance document (USEPA 2016). Details of the laboratory QC/QA steps are provided in Appendix A of the Ecosphere article.

All data processing/analysis is documented in the two R-scripts provided and summarized briefly in the metadata.

We provide 2 main datasets, one used for most statistical analyses ('Unaveraged_data_all_metals_June8.csv") and one used to create maps ("Final_forMAPSV3"). "Unaveraged_data_all_metals_June8.csv" contains raw values of heavy metals measured in all moss samples along with basic information for each sample, including collector (expert vs youth), collection date, lat and long of sampled tree. ""Final_forMAPSV3" was derived from the prior dataset; it includes only youth-collected data and replicate samples were averaged, leaving one value for mapping. We also include two reference datasets we compared values from the main dataset with ("Portland_fulldataset" and "seattleparks").

Note that Mn and Se values are not reliable and are included for informational purposes only. Mn in moss is influenced by canopy leachates and ICP-OES detection limits for Se are high. Units for all elements are mg/kg unless specified otherwise in the metadata. Additionally, a flash thunderstorm occurred before experts collected moss samples on June 13, which we believe led to systematically lower heavy metals values in expert-collected samples (although results, for the most part, were not statistically significantly different from youth-collected samples; See Derrien et al. 2020).