Temporal Distribution Of Primary And Secondary Production Estimated From Water Quality Data In The Seto Inland Sea, Japan

Estuaries and coastal areas support highly productive fisheries and aquaculture operations. The high productivity is partly supported by inputs of nutrients from land; consequently, these areas are sensitive and vulnerable to human activities. Excess anthropogenic nutrient input can lead to accelerated phytoplankton growth and associated environmental problems. Good management practices are needed to achieve high productivity and healthy environmental conditions simultaneously. This requires close monitoring of environmental conditions and productivity, especially at lower trophic levels. Although environmental conditions, especially water quality, are widely monitored, primary production (PP) and secondary production (SP) are rarely monitored even though these basic parameters govern the structure and functioning of marine ecosystems. In this study, we developed methods to estimate PP and the SP from commonly collected water quality indicators and without the use of any information on phytoplankton or zooplankton except chlorophyll a (Chl.a) concentration. For this purpose, we developed the method to determine the half-saturation constants as a function of nutrient concentration, and the method to determine the SP from Chl.a concentration. Then, we used our newly developed methods to estimate the temporal distribution of PP and SP in the Seto Inland Sea from the 1980s to the present. To do so, we monitored water quality at three sites with different levels of eutrophication. We also measured PP by using C-13-labeled sodium bicarbonate ((NaHCO3)-C-13) incubation method, and calculated copepod SP based on the species and body size of copepods collected in a survey. To estimate PP from water quality data, we modeled PP using a half-saturation constant for nutrient uptake that was allowed to vary as a function of nutrient concentration. Seasonal differences in PP could be successfully estimated by using this method. The estimation error for PP was within -16% to 23%. We found that SP calculated from copepod species and body size data was significantly correlated with chlorophyll a concentration. Therefore, we proposed a model to estimate SP from chlorophyll a concentration. Then, we used our newly developed models to estimate the temporal distribution of both PP and SP in the Seto Inland Sea from seasonal water quality data. According to our estimates, annual mean PP and SP peaked in the late 1990s and have trended downward since around 2000 or 2005. No significant inter-decadal changes in autumn or winter PP were observed during our study period; by contrast, spring and summer PP decreased significantly from the 1990s to the present. Estimated annual mean PP and SP over the most recent 5-year period were 73% and 88%, respectively, of equivalent values estimated for the 1990s.