The marine geochemistry of methane

Abstract

In the highly productive coastal surface waters near Walvis Bay, methane is present in concentrations considerably above those which would be predicted from solubility equilibrium with the atmosphere. A one dimensional diffusive model and a one dimensional horizontal advection diffusion model were used to describe the methane distribution. Evaluation of the model fits to the data suggests that both advective supply of methane-rich coastal waters and in situ biological methane production are important sources for the mixed layer methane excess. The complexity of the hydrographic regime near Walvis Bay makes it impossible to make a quantitative estimate of the rate of methane production. In the less productive Murray-Wilkinson Basin in the Gulf of Maine, a mixed layer methane excess is also observed. Methane concentrations are closely correlated with hydrographic parameters and the source of methane at a middepth maximum appears to be the highly anoxic sediments in the adjoining Franklin Basin. Diffusion of methane from the middepth maximum is probably adequate to maintain the surface methane excess against loss across the air-sea interface. Coastal waters are frequently enriched in methane, and it has been shown that advective supply of these methane-rich waters may be a significant source of methane for the mixed layer near the coast. Thus the widespread occurrence of a methane maximum at the base of the mixed layer in the open ocean, coupled with surface waters typically 30-70% supersaturated with respect to solubility equilbrium, suggests that advective supply of methane might be an important methane source for the open ocean as well. However, a study of the western subtropical Atlantic shows that advective transport can probably supply only a fraction of the methane present in the maximum. Also the loss of methane across the air-sea interface was observed to be twenty times greater than the flux from the maximum. Thus in situ methane production must be very important to the open ocean methane distribution. A series of phytoplankton culture experiments demonstrated that cultures of both Coccolithus huxleyi and Thalassiosira pseudonana produce trace amounts of methane during logarithmic growth. (Because the cultures are highly oxygenated, anaerobic methane bacteria can be neglected as methane sources. However heterotrophic bacteria cannot be excluded as possible sources of methane to the cultures.) After three algal generations, the rate of methane increase closely parallels the growth curve suggesting that the methane is in fact coming from the algae. A methane production rate of 2 x 10-10 nmole methane/viable cell/hr was calculated from the data. This rate is three to four orders of magnitude slower than the rates of oxygen consumption and glutamate and glucose uptake measured by other workers. for algae and bacteria. The methane production rate calculated from the culture experiments is the correct order of magnitude to account for the methane production occurring in the open ocean. Methane is present in quite low concentrations in the deep ocean. By calculating water mass ages from GEOSECS and other data, it is possible to estimate methane consumption rates in the deep sea. Methane consumption is rapid at first (probably greater than 0.06 nmole/l/yr). At depth consumption appears extremely slow. This may be due to the fact that the methane concentrations in the deep sea are so low that methane oxidizing bacteria cannot use methane as a substrate, or due to reduced metabolic activity in the bacteria at the high pressures and low temperatures of the sea floor. Methane is present in very high concentrations in anoxic basins, indicating that methanogenic bacteria are active. However, near the anoxic-oxic interface in both the Black Sea and the Cariaco Trench a one dimensional advection diffusion model predicts that methane consumption is occurring in the anoxic zone. In the Black Sea the methane depletion may be indicative of the presence of rapid methane oxidation near the Bosporus overflow. However in the Cariaco Trench the validity of such an explanation is difficult to evaluate since the overflow process is so poorly understood. A box model for the Trench has been developed which incorporates time dependence and supply of chemical species to the water from the sediments at all depths in the Trench. This model can explain the silica and sulfide data quite well, but methane depletion near the interface, relative to the model predictions, still occurs. Thus either anaerobic methane oxidation or decreased methane production in the sediments must be hypothesized.