Field transect surveys

Field surveys of eelgrass meadow sites were conducted at mid-summer low tides at field sites along the west coast of North America in the U.S. and Canada. Samples and data were collected within the intertidal area of 32 eelgrass meadows distributed in six regions (Alaska, British Columbia, Washington, Oregon, California -Bodega Bay, and California -San Diego). Surveys were conducted between late June and early August in 2019, 2020, and 2021 by teams from six institutions.

For each site, three 20 meter transects were laid parallel to the shore at the shoreward (upper edge) of continuous eelgrass, and three lower (intertidal) 20 meter transects were laid at least 4 meters closer to the water. Along each transect, individual third-rank eelgrass leaves were collected at meters 1, 2, 3, 5, 6, 7, 9, 10, 11, 13, 14, 15, 17, 18, and 19, and full eelgrass shoots were collected at meters 4, 8, 12, 16, and 20. Leaf and shoot samples were transported in individual containers on ice to the laboratory for immediate processing.

Transect locations were recorded using a hand-held GPS (exact model varied between field locations). Salinity was measured at the time of sampling using a refractometer. Temperature loggers (HOBO MX 2201 and UA-001-64, Onset, Bourne, MA) were deployed at each eelgrass meadow site to provide a continuous record of in situ temperature. For HOBO data, see https://www.bco-dmo.org/dataset/877355 and Related Datasets section below.

Laboratory (Imaging and Disease Metrics)

In the lab, eelgrass blades were cleaned and prepared for imaging to capture disease metrics. Cleaned blades were imaged at high resolution (600 dpi) using an Epson Perfection V550 scanner. Shoot morphology measurements (sheath length, number of leaves, canopy height) were taken by hand in the laboratory (see Shoot Metrics, https://www.bco-dmo.org/dataset/878857). The third-rank leaf from each shoot was analyzed for epiphyte load and grazing scars.

Third-rank leaves were further analyzed for disease metrics through imaging. Cleaned leaves were placed between sheets of acetate and imaged at high resolution (600 dpi) using an Epson Perfection V550 scanner. The high-resolution images were saved in TIFF format and then processed using a program developed by the authors. The Eelgrass Lesion Image Segmentation Application (EeLISA) uses machine learning to identify healthy and diseased eelgrass tissue and outputs the following metrics:

disease prevalence (presence or absence of disease on a given leaf)
disease lesion area (absolute size of wasting disease lesions), and
disease severity (proportion of leaf area damaged by disease).

~ For details on the development, testing, and training of EeLISA, see Rappazzo et al. (2021).
~ For methodology details, see Aoki et al. (2022)
~ Additional details for the field surveys are available in the Eelgrass Disease Project Handbook.
~ For 16S rRNA amplicon sequencing of eelgrass associated bacteria, refer to NCBI BioProject PRJNA802566 in the Related Datasets section below.

disease_metrics.csv