The effects of marine particle composition on phytoplankton coagulation efficiency and the use of Th-234 as a proxy for particulate organic carbon flux

Loading...
Thumbnail Image
Issue Date
1-Aug-12
Authors
Chow, Jennifer Szlosek
Publisher
The Graduate School, Stony Brook University: Stony Brook, NY.
Keywords
Abstract
This dissertation investigates two parameters important for estimating and predicting organic matter export from the surface ocean: controls on the POC/<super>234<super>Th of marine particles and the coagulation efficiency of algal cells. Specifically, the sources of variability in POC/<super>234<super>Th and coagulation efficiency of <italic>Emiliania huxleyi<italic> are studied. The research explores two methods commonly employed &ndash the <super>234<super>Th -proxy method for estimating POC flux, and Couette flow device experiments for experimentally deriving coagulation efficiency values. Applications of <super>234<super>Th /<super>238<super>U disequilibrium as a flux proxy for particulate organic carbon (POC) in the oceans commonly rely on characterizing the POC/<super>234<super>Th in filterable particles and using this as representative of the sinking flux. To better understand the relationship between <super>234<super>Th and POC, marine particles were collected in the northwestern Mediterranean Sea (spring, 2003 and 2005). First, we evaluated the role of particle settling velocity and chemical composition on POC/<super>234<super>Th of material collected by settling velocity sediment traps. This study provides evidence that marine particle source (e.g., phytoplankton aggregates, zooplankton fecal pellets, degraded biogenic material) has a stronger influence on the POC/<super>234<super>Th than the particle size (vis-a-vis particle settling velocity). We found no consistent trend in POC/<super>234<super>Th with settling velocity, contrary to the canonical view of differential scaling of POC and <super>234<super>Th with volume and surface area. The source material of the slow and fast settling particles result in similarly low POC/<super>234<super>Th due to carbon degradation of material in slow settling velocity classes and carbon assimilation of material in fast settling velocity classes. Results revealed a factor of 3 drop in POC/<super>234<super>Th from 313 to 1918 m due to a combination of POC losses (dissolution and degradation) and <super>234<super>Th gains (continued scavenging of <super>234<super>Th and disaggregation of fast settling particles with depth). Motivated by the variability in POC export estimates of the pumps versus the traps, we investigated relationships and linkages between filterable particles and those collected in sediment traps by comparing their POC/<super>234<super>Th ratios and compositions (organic and radiochemical). Principal components analysis showed little overlap in composition between filterable particles and sediment trap-collected material, despite similar (within a factor of ~2) POC/<super>234<super>Th and POC flux estimates of the 1-70 åµm pump fraction and sediment traps. The relative separation of particles according to freshness were: small pump fraction (1-70 åµm) > large pump fraction (>70 åµm) > time series and settling velocity sediment trap fractions. The small pump fraction was enriched in indicators of fresh phytoplankton and calcifying algae whereas traps contained biomarkers indicative of fecal pellets and bacterially degraded organic matter. Indicators enriched in the large pump fraction appeared to be consistent with algal material from diatoms and coccolithophores as well as fecal pellets. The pigment composition data of the filterable particles indicated that lateral advection and vertical mixing occurred as well as large short-term changes (5 d) in POC/<super>234<super>Th (1.7-3 &times) at the base of the mixed later (~300 m). Coupled with significant short-term variation in water column <super>234<super>Th distributions, these results imply that a thorough understanding of a given oceanic region is required to validate the use of <super>234<super>Th as a proxy for POC flux. To experimentally assess the controls on and rates of phytoplankton aggregation, we used Couette flow device experiments to derive coagulation efficiencies of <italic>Emiliania huxleyi<italic>; to our knowledge, this has not been done previously. In chemostat experiments conducted in 2006 and 2007, <italic>E. huxleyi<italic> was grown at different growth rates and we conducted replicate Couette flow coagulation experiments at multiple growth rates and assessed the relationship of cell coagulation efficiency with cell growth rate and indicators of cell stickiness. Analogs for cell stickiness that were characterized include the sugar composition of high molecular weight dissolved material and the abundance of transparent exopolymer particles (TEP). Coagulation efficiencies increased from 0.2 to 1 as cell growth rates declined from ~0.6 to ~0.1 d-1. We observed a significant positive correlation between growth rates of cells and total alkalinity and a negative correlation between growth rates of cells, TEP/chlorophyll a, concentrations of detached coccoliths, and sugars specific to coccolith formation (p<0.01). Experiment replication showed coagulation efficiencies of cells at similar growth rates were enhanced by the presence of TEP on cell surfaces and calcite coverage of the coccosphere. Based on our findings we propose that the method by which TEP is included within an aggregate (e.g., TEP matrix embedding cells, free TEP gluing cells together, or exopolymer coating on cells leading to cell-cell contacts) could influence aggregate strength and its likelihood to sink intact beyond the mixed layer. This dissertation highlights the utility of organic matter composition and settling velocity data of trap and filterable particles for evaluating a representative POC/<super>234<super>Th and assessing the accuracy of the <super>234<super>Th proxy method for estimating POC flux. Results also show that Couette flow coagulation experiments and algal exudate characterization help increase our mechanistic understanding of the tie between cell environment and physiology changes and phytoplankton aggregation and export dynamics.
This dissertation investigates two parameters important for estimating and predicting organic matter export from the surface ocean: controls on the POC/<super>234<super>Th of marine particles and the coagulation efficiency of algal cells. Specifically, the sources of variability in POC/<super>234<super>Th and coagulation efficiency of <italic>Emiliania huxleyi<italic> are studied. The research explores two methods commonly employed &ndash the <super>234<super>Th -proxy method for estimating POC flux, and Couette flow device experiments for experimentally deriving coagulation efficiency values. Applications of <super>234<super>Th /<super>238<super>U disequilibrium as a flux proxy for particulate organic carbon (POC) in the oceans commonly rely on characterizing the POC/<super>234<super>Th in filterable particles and using this as representative of the sinking flux. To better understand the relationship between <super>234<super>Th and POC, marine particles were collected in the northwestern Mediterranean Sea (spring, 2003 and 2005). First, we evaluated the role of particle settling velocity and chemical composition on POC/<super>234<super>Th of material collected by settling velocity sediment traps. This study provides evidence that marine particle source (e.g., phytoplankton aggregates, zooplankton fecal pellets, degraded biogenic material) has a stronger influence on the POC/<super>234<super>Th than the particle size (vis-a-vis particle settling velocity). We found no consistent trend in POC/<super>234<super>Th with settling velocity, contrary to the canonical view of differential scaling of POC and <super>234<super>Th with volume and surface area. The source material of the slow and fast settling particles result in similarly low POC/<super>234<super>Th due to carbon degradation of material in slow settling velocity classes and carbon assimilation of material in fast settling velocity classes. Results revealed a factor of 3 drop in POC/<super>234<super>Th from 313 to 1918 m due to a combination of POC losses (dissolution and degradation) and <super>234<super>Th gains (continued scavenging of <super>234<super>Th and disaggregation of fast settling particles with depth). Motivated by the variability in POC export estimates of the pumps versus the traps, we investigated relationships and linkages between filterable particles and those collected in sediment traps by comparing their POC/<super>234<super>Th ratios and compositions (organic and radiochemical). Principal components analysis showed little overlap in composition between filterable particles and sediment trap-collected material, despite similar (within a factor of ~2) POC/<super>234<super>Th and POC flux estimates of the 1-70 åµm pump fraction and sediment traps. The relative separation of particles according to freshness were: small pump fraction (1-70 åµm) > large pump fraction (>70 åµm) > time series and settling velocity sediment trap fractions. The small pump fraction was enriched in indicators of fresh phytoplankton and calcifying algae whereas traps contained biomarkers indicative of fecal pellets and bacterially degraded organic matter. Indicators enriched in the large pump fraction appeared to be consistent with algal material from diatoms and coccolithophores as well as fecal pellets. The pigment composition data of the filterable particles indicated that lateral advection and vertical mixing occurred as well as large short-term changes (5 d) in POC/<super>234<super>Th (1.7-3 &times) at the base of the mixed later (~300 m). Coupled with significant short-term variation in water column <super>234<super>Th distributions, these results imply that a thorough understanding of a given oceanic region is required to validate the use of <super>234<super>Th as a proxy for POC flux. To experimentally assess the controls on and rates of phytoplankton aggregation, we used Couette flow device experiments to derive coagulation efficiencies of <italic>Emiliania huxleyi<italic>; to our knowledge, this has not been done previously. In chemostat experiments conducted in 2006 and 2007, <italic>E. huxleyi<italic> was grown at different growth rates and we conducted replicate Couette flow coagulation experiments at multiple growth rates and assessed the relationship of cell coagulation efficiency with cell growth rate and indicators of cell stickiness. Analogs for cell stickiness that were characterized include the sugar composition of high molecular weight dissolved material and the abundance of transparent exopolymer particles (TEP). Coagulation efficiencies increased from 0.2 to 1 as cell growth rates declined from ~0.6 to ~0.1 d-1. We observed a significant positive correlation between growth rates of cells and total alkalinity and a negative correlation between growth rates of cells, TEP/chlorophyll a, concentrations of detached coccoliths, and sugars specific to coccolith formation (p<0.01). Experiment replication showed coagulation efficiencies of cells at similar growth rates were enhanced by the presence of TEP on cell surfaces and calcite coverage of the coccosphere. Based on our findings we propose that the method by which TEP is included within an aggregate (e.g., TEP matrix embedding cells, free TEP gluing cells together, or exopolymer coating on cells leading to cell-cell contacts) could influence aggregate strength and its likelihood to sink intact beyond the mixed layer. This dissertation highlights the utility of organic matter composition and settling velocity data of trap and filterable particles for evaluating a representative POC/<super>234<super>Th and assessing the accuracy of the <super>234<super>Th proxy method for estimating POC flux. Results also show that Couette flow coagulation experiments and algal exudate characterization help increase our mechanistic understanding of the tie between cell environment and physiology changes and phytoplankton aggregation and export dynamics.
Description
247 pg.
DOI