I’ve packaged our 12-month Puget Sound manta trawl (333-micron mesh) + FTIR workflow into a field-ready SOP with contamination controls and an R script for QA/QC and power analysis, geared toward comparable baselines that inform source-control decisions. If you’re running nearshore surveys around stormwater outfalls, I can share the folder and would appreciate any validation data or notes on detection limits from your sites.
On our near‑outfall manta tows, we run field blanks at the ramp and swap in a stainless 63 µm sieve stack for the cod‑end rinse, plus bake filters at 450°C; tiny caveat, 333 µm misses a lot of fibers so we flag those separately in power calcs (DHARMa helps: CRAN: Package DHARMa) — our hood gasket used to shed like a golden retriever. Mind sharing the folder, @OP, and does your R script handle zero‑inflation?
We’ve had good luck clipping a mid‑tow air blank at net‑mouth height (47 mm GF/F, 5 min) and subtracting it — near outfalls it knocked counts down about 5–10% beyond ramp blanks. Re: “zero‑inflation”, @OP, glmmTMB with a ziformula and an offset for filtered volume has been stable for us, and pairing tows by tide stage tightened power. Can you share the folder so I can test your FTIR match thresholds against our weathered PE library?