Impacts of increasing dissolved inorganic nitrogen discharged from Changjiang on primary production and seafloor oxygen demand in the East China Sea from 1970 to 2002

Abstract

In recent years, benthic hypoxia has been observed in the outflow region of the Changjiang River in the East China Sea. Because the nitrogen input to the Changjiang watershed, mainly from human activities, has increased by 3 fold in the last four decades and the nitrogen load had grown exponentially, it is speculated that anthropogenic nutrients may be responsible for the hypoxia in the East China Sea shelf. We employ a coupled 3-D physical-biogeochemical model of the East China Sea to investigate how the changing Changjiang nutrient loads from 1970 to the end of 2002 may have impacted on primary production in the water column and the seafloor oxygen demand (SOD) on the seafloor. The model predicts an average value of 437 mgC m- 2 d- 1 for primary production and 10.0 mmol O2 m- 2 d- 1 for SOD for the ECS shelf over the entire modeling period. The model results compare reasonably with observations during the period from December 1997 to October 1998. Responding to the increase of the Changjiang DIN loading by a factor of ~ 2.4, the modeled primary production in the East China Sea shelf has increased by 17%, and the modeled SOD by 22%. In the inner shelf, where the impact is the strongest, the SOD increases by 30%. We are able to identify areas of potential hypoxia using two criteria: SOD > 30 mmol O2 m- 2 d- 1 and water depth > 25 m. The maximum area of potential hypoxic region in any month of a year has increased dramatically after 1991; the change appears related to the Changjiang DIN loads from May to July that showed a sudden increase after 1990. The responses in potential hypoxic area are more pronounced than the increases in DIN loads, suggesting strong nonlinear effect in the development of hypoxia, which warrants further investigation. It is cautioned that the SOD calculation was based on the Redfield C/N ratio, but the actual C/N ratio may deviate from it. Direct observations of the sediment oxygen consumption are needed to valid our modeling approach. We also assessed the potential impacts of particulate organic matter from Changjiang by introducing a load of reactive particulate nitrogen (PN), which was assumed proportional to DIN based on estimated yields in the watershed. The modeled impacts on primary productivity and SOD are significant, but more accurate quantification of the monthly PN load and better characterization of its reactivity are required for better assessment.