We collected samples using SCUBA at three sites in Sitka Sound, all within 6 km of each other: Harris Island (N 57.03165, W 135.27754), Breast Island (N 57.03896, W 135.33309), and Samsing Pinnacle (N 56.98750, W 135.35718). Samples were collected during one week in two seasons: winter (January 2019) and summer (July 2019). Generally, we collected ~4 replicate samples for each species at each of the three sites in each season for fatty acids analysis. For seaweeds, this resulted in 4 replicates from one site in each season, and for invertebrates, this resulted in ~12 replicates (across 3 sites) for each season, with the exception of the isopod Pentidotea resecata in the summer, for which only one sample was collected due to a paucity of isopods residing in the seaweeds across our 3 sampling sites.

We sampled six seaweeds, including three Laminarian kelps (Ochrophyta, urn:lsid:marinespecies.org:taxname:345465) and three red algae from the Gigartinales (urn:lsid:marinespecies.org:taxname:871)  and Ceramiales (urn:lsid:marinespecies.org:taxname:860), that are present in both seasons in these sites: Neoagarum fimbriatum, Macrocystis pyrifera, Hedophyllum nigripes, Cryptopleura ruprechtiana, Opuntiella californica, and Osmundea spectabilis (See supplemental species_list.csv for identifiers).
Note: "Neoagarum fimbriatum" appears as the unaccepted synonym "Agarum fimbriatum" used in this dataset. 

We sampled six macro-invertebrates (two each of gastropods, crustaceans, echinoderms) that are present and dominant herbivores in these sites: Haliotis kamtschatkana, Tegula pulligo, Pentidotea resecata, Pugettia producta, Mesocentrotus franciscanus, and Strongylocentrotus droebachiensis (See supplemental species_list.csv for identifiers). 

M. pyrifera blades were collected from the distal ends of the thallus near the scimitar area from a boat at the surface, but other collections were made by SCUBA divers between 5-8 m depth. In general, all seaweeds were sampled in a consistent way within each species, targeting the same general tissues, avoiding proximal and distal areas of blades, avoiding sori, holdfasts, and stipes. All seaweeds were wiped clean with a paper towel and inspected to ensure they were not contaminated by epiphytes or endophytes. Invertebrates were sacrificed by freezing (-20°) for ~30 minutes before subsequent dissection. Invertebrate tissues were sampled in a consistent way for each taxon: Haliotis were sampled by snipping the foot muscle; Tegula were crushed with a hammer and the foot muscle was snipped; whole P. resecata individuals were sampled owing to their small size; for Pugettia, muscle tissue was removed from the legs; for Mesocentrotus and Strongylocentrotus, the test was cut open and gonadal tissue sampled. Sampling equipment was wiped clean with kimwipes and ethanol between samples. All samples were placed into 1.5 ml Eppendorf tubes, which were immediately frozen in a -20°C freezer. No solvents were introduced to samples in plastic storage vials.Within 3 days, all samples were moved to a -80°C freezer where they were stored until shipment to the -80°C freezer at the Oregon Institute of Marine Biology, where all fatty acid extractions were performed. 

Tissue samples were kept at -80°C until extraction; all samples for both collection efforts were extracted within 9 months of collection. Samples were lyophilized to dryness for 48 hours. Once tissues were dried they were ground to fine, homogeneous powder with a stainless-steel mortar and pestle. Homogenized tissues were then immediately digested in chloroform for ~12 hrs sealed under nitrogen at -20°C. Total lipid extraction and subsequent derivatization of fatty acid methyl esters (FAME) was performed using a modification of the method described by (Taipale et al. 2016) and briefly summarized here. All sample extractions were performed in pre-combusted glassware.

Following initial tissue digestion in chloroform, 70 ul of unmethylated nonadecanoic acid (C19) was added as an internal standard. Total lipids were extracted using a 2:1:0.75 solution of chloroform: methanol: 0.9% NaCl solution, which was subsequently sonicated, vortexed, and centrifuged so that the more dense organic layer could be removed and evaporated to dryness under a steady stream of nitrogen gas. The organic phase was immediately resuspended in toluene and 1% sulfuric acid-methanol solution for 90 minutes at 90°C to transesterify FAME. Following transesterification, solutions were cooled to room temperature and KHCO3, to neutralize the acidic solution, and hexane was added. The combined solution was then vortexed and centrifuged to separate FAME in the hexane solution, which was concentrated to 1.5 ml in glass vials for gas chromatography. 

FAME were analyzed with a gas chromatograph mass spectrometer (GC-MS, Shimadzu, Model QP2020), fitted with a DB-23 column (30 x 0.25mm x 0.15 µm, Agilent, Santa Clara, CA, USA), using helium as the carrier gas. To ensure sufficient separation between fatty acid peaks we utilized a heating program modified from Taipale et al. (2016) and described by Thomas et al. (2020). Individual fatty acids were identified using relative retention times of a FAME standard (GLC 566C, Nu-Chek Prep, Elysian, MN, USA) and specific ions. Fatty acids were quantified using integration of ­the major ion peaks (Taipale et al. 2016) with Shimadzu Lab Solutions software. Following identification and quantification, individual peak areas for all fatty acids identified were converted to proportions, representing the % contribution of all fatty acids identified in each sample.