Photobleaching of fluorescent dissolved organic matter in Beaufort Sea and North Atlantic Subtropical Gyre

Abstract

Fluorescent dissolved organic matter (FDOM) samples collected from water masses of Beaufort Sea (n = 12) and the North Atlantic Subtropical Gyre (NASG; n = 6) were assessed based on susceptibilities of its parallel factor analysis (PARAFAC) modeled fluorescent components to photobleaching over 72 hours of simulated solar irradiation. 315 excitation–emission matrix spectra (EEMs) were PARAFAC modeled yielding 4 humic-like (C1–3 and C5) and 1 protein-like (C4) fluorescent components. Protein-like C4 and humic-like C5 did not adhere to first order kinetics and did not yield decay rate constant (k) or half-life (t1/2) values as part of this study. Humic-like C1 was found to be most resilient to photodegradation with lowest k values. Principal component analysis (PCA) illustrated a shift towards C1% with photo-exposure in addition to distinguishing compositions of C1–5% in Beaufort Sea halocline and Atlantic layer as well as NASG deep water masses. When Beaufort Sea water masses were considered, k of C2 was found to be lower in upper (UH) and mid (MH) halocline layers when compared to lower halocline (LH) and underlying Atlantic layer (AL; p < 0.10). This suggested that C2 in Pacific-derived seawater was less photoreactive than C2 in Atlantic-derived seawater in Beaufort Sea, likely as a result of distinct formation pathways of these water masses. Similarly, C2 in upper Labrador Seawater (ULSW) and ‘classical’ Labrador seawater (LSW) was less photolabile than in deeper NASG layers (Iceland-Scotland and Denmark Strait overflow waters; ISOW and DSOW). Interestingly, despite similar formation pathways of AL in Beaufort Sea and deeper waters sampled from NASG (ISOW + DSOW), lower k for C1 was found in NASG.