Open Collections

Abstract

The methylated sulfur compounds dimethyl sulfide (DMS), dimethyl sulfoniopropionate (DMSP), and dimethyl sulfoxide (DMSO; colloquially referred to together as DMS/O/P) are essential components to the marine sulfur cycle. These compounds are ubiquitous in oceanic waters and have putative roles ranging from regulating cellular physiology to facilitating multi-trophic interactions and regulating regional climate. Despite over three decades of intense study, there remains significant uncertainty in the mechanisms and processes influencing DMS/O/P cycling and distributions. This PhD thesis focuses on applying novel machine learning techniques and isotopic shipboard experiments to better characterize DMS distributions and DMS/O/P cycling within the NE subarctic Pacific (NESAP) and Southern Oceans, two “hotspot” regions for DMS/O/P cycling globally. Chapters 2 and 3 applied ensembled random forest regression and artificial neural networks to build surface ocean DMS climatologies in the NESAP and Southern Oceans, representing smaller-scale (20-27 km) variability. Results from these chapters demonstrated significantly improved predictive accuracy over traditional empirical models, although later in situ work in Chapter 5 revealed limitations in these models. Nonetheless, this modelling work revealed several oceanographic factors, including irradiance, nutrient availability, and mixing, as key drivers of DMS variability in these regions, with temperature gradients and sea surface height anomalies controlling finer-scale DMS features. Chapters 4 and 5 further explored the mechanisms and processes underpinning the relationships between DMS/O/P cycling, mixing, irradiance and temperature. Ship-board irradiance manipulation and photosynthetic inhibition experiments in the NESAP revealed that DMSO reduction is upregulated in low-light acclimated phytoplankton assemblages exposed to surface level irradiance, and this pathway was linked to photosynthetically derived electrons. From these results, a novel electron scavenging hypothesis was proposed as a physiological function for DMSO in phytoplankton. In Chapter 5, DMS/O/P cycling was examined in response to a marine heatwave event in the NESAP, providing evidence for an indirect influence of temperature through floristic shifts towards high-DMS producing species, and notable coastal-oceanic transitions in sulfur metabolism. Collectively, the chapters in this thesis provide new insights into the application of novel statistical methods to characterize DMS distributions, as well as the mechanisms and processes underpinning DMS/O/P cycling within two globally significant ocean regions.