Modelling of stream metabolism under human-induced disturbances

With this study, components of stream metabolism; gross primary production, net ecosystem production, and community respiration were estimated by using the diel oxygen technique for the first time in Turkey. "Two-station" method developed by Odum (1956), was used as the diel oxygen technique. This study aimed at investigating effects of discharges of the urban wastewater treatment plant on stream metabolism of Büyüksu Stream (Bolu, Turkey), examinating and statistically modelling the relationships between components of stream metabolism and environmental variables (water temperature, turbidity, specific conductivity, pH, biochemical oxygen demand, total nitrogen, total phosphorus, atmospheric pressure, air temperature, relative humidity, etc.). Within this scope, required variables (dissolved oxygen and water temperature) were measured in order to estimate metabolism components at the points of before and after discharges of Bolu Urban Wastewater Treatment Plants, and springs (headwaters) of Abant Creek and Mudurnu Creek feeding Büyüksu Stream for 17 months between August 2015 and December 2016. According to naturality gradient (Reach disturbance: "Before discharge", After discharge", "Spring of Abant Creek", Spring of Mudurnu Creek"), the group means of stream metabolism components were compared by performing one-way analysis of variance. Results of variance analysis showed that the discharges of the urban wastewater treatment plant influence the stream metabolism; increase the average community respiration (before discharge: -30.6 gr O2 m-2 day-1, after discharge: -130.9 gr O2 m-2 day-1), statistically make no significant difference on the average gross primary production (before discharge: 15.6 gr O2 m-2 day-1, after discharge: 9.1 gr O2 m-2 day-1). Validations of the multiple non-linear regression models between environmental variables and net ecosystem metabolism, and community respiration gave coefficients of determination of 74.9 % ve 66.6 %, respectively. Moreover, the process model estimating dissolved oxygen and net ecosystem metabolism was improved by using STELLA software for the spring of Abant Creek, and approximate validation results to the field measurement values (MAE for dissolved oxygen: 0.096 – 0.809 mg L-1, MAE for net metabolism: 0.0009 – 0.0054 gr O2 m-2 min-1) were obtained by the model.