45608-AC2
Organic Matter Preservation in Marine Environments: Insights from a Study of Marine Dissolved Organic Matter - Composition, Age, and Source
As noted in the previous report the solid phase extraction (SPE) method described in the original proposal was applied (using a divinyl-benzene resin and multi-step solvent elution scheme) together with an acid hydrolysis/solvent partitioning scheme to isolate several polar and hydrophobic fractions of dissolved organic matter (DOM) from the surface ocean and Delaware river and estuary. For comparison, the acid hydrolysis/solvent partitioning scheme was also applied to DOM samples isolated from the same environment by size separation (ultrafiltration using a 1000 Da (1 nm) cutoff). Our goal was to address the question is there a relationship between chemical composition and radiocarbon age of dissolved organic fractions? In summary, the results show that the two distinct DOC isolation methods -- SPE and ultrafiltration -- are capable of removing compounds with parallel chemical characteristics from various aquatic environments. Each method isolates both contemporary and 14C-depleted organic matter; and, in each method, compounds with comparable chemical properties (for example, sugars or lipids) share similar 14C signatures. SPE is often used to isolate fractions referred to as humic substances, which have been typically considered to be refractory due to their complex structure and depleted 14C signatures. However, results from the method employing acid hydrolysis coupled to solvent extraction show that these fractions include compounds of contemporary origin that are strongly associated with 14C-depleted compounds. These tight associations are necessary to explain both the isolation of carbohydrates and other polar compounds onto hydrophobic resins and the release of compounds soluble in ethyl acetate only following acid hydrolysis. Results identify both structural features that facilitate DOC preservation in marine environments on long timescales and molecular interactions that could allow transient accumulation of bioavailable compounds in the upper ocean. Ultimately these results challenge models that utilize a homogenous reservoir to represent the behavior of DOC in the carbon cycle.
The second question proposed still remains elusive: is the average age of DOC a result of a broad DOC age continuum or a few discrete ages? Though hydrophobic fractions of each sample were always depleted in 14C they still exhibited a range of Δ14C values (unlike polar fractions). In the eastern North Pacific a seasonal shift in Δ14C values was observed for these fractions suggesting some recent production of lipids. However, the overall 14C signature of each hydrophobic fraction was depleted indicating a major contribution from refractory material. These hydrophobic compounds are currently being characterized using gas chromatography (GC) time-of-flight (TOF) mass spectrometry (MS) and GC electron ionization (EI) MS. GC EI-MS has identified several common lipids such as fatty acids and alkanes (which are probably seasonally produced) but there are several complex structures that still need resolving. To resolve some of these more complex regions of the gas chromatograph a 2D-GC TOF MS method is also being developed. Ultimately, the age distribution in the hydrophobic fraction and identity of refractory lipids will be determined by isolating particular lipids for compound specific radiocarbon measurements (via preparatory GC).
We are continuing to investigate the third question posed in our proposal - is the old component of DOC derived from terrestrial/riverine environments and if so, is it compositionally distinct? As stated in the last report, the Delaware study showed that the modern, terrestrial, hydrophobic fraction isolated from river water was replaced downstream in the estuary by 14C-depleted, δ13C-enriched marine compounds. It was also shown that glucose (produced after acid hydrolysis of the SPE fraction), unlike other sugar fractions, retained a terrestrial δ13C signature and showed no shift in Δ14C with increased salinity. These data indicated that glucose (a relatively labile carbohydrate) is exported to marine systems from the terrestrial environment.
Two more rivers have now been sampled and together, they encompass a range of Δ14C-DOC values. The rivers, in order of increasing DOC age, are Santa Clara (-73‰; despite the fact that POC had values between -230‰ and -544‰), Delaware (-114‰) and Sacramento (average value of -237‰). These samples allow an investigation into whether a relationship exists between total Δ14C-DOC and the 14C signature of hydrophobic DOC fractions (i.e., is the hydrophobic fraction primarily responsible for the 14C-depleted signature of DOC). As with the Delaware River, DOC in CA rivers shows a terrestrial δ13C signature, so both polar and hydrophobic fractions should be derived from terrestrial environments. The CA river samples are currently being processed for radiocarbon measurements. The new graduate student on this project, G. Ian Ball, is also developing a method to measure the 14C signature of lignin oxidation products in CA river DOM. Lignin is an unambiguous biomarker for terrestrial material and its 14C signature can be compared to that of hydrophobic fractions to determine the diversity of 14C-ages in terrestrial OM transported by these rivers.
