Summer hypoxia in the northern Gulf of Mexico and its prediction from 1978 to 1995.

An 18-year monitoring record (1978-1995) of dissolved oxygen within a region having hypoxia (dissolved oxygen less than 2 mgl(-1)) in the bottom layer was examined to describe seasonal and annual trends. The monitoring location was near or within a well-described summer hypoxic zone whose size has been up to 20,000 km(2). The monitoring data were used to hindcast the size of the hypoxic zone for before consistent shelfwide surveys started, and to predict it for 1989, when a complete shelfwide survey was not made. The concentration of total Kjeldahl nitrogen (TKN) in surface waters and concentration of bottom water oxygen were directly related, as anticipated if organic loading from surface to bottom was from in situ processes. The TKN data were used to develop a predictive relationship that suggested there was no substantial hypoxia before the 1970s, which was before nitrate flux from the Mississippi River to the Gulf of Mexico began to rise. The peak frequency in monthly hypoxic events is two to three months after both the spring maximum in discharge and nitrate loading of the Mississippi River. These results support the conclusion that persistent, large-sized summer hypoxia is a recently-developed phenomenon that began in the 1970s or early 1980s.