Date of Award

Winter 2020

Project Type


Program or Major

Natural Resources

Degree Name

Master of Science

First Advisor

Wilfred M Wollheim

Second Advisor

William H McDowell

Third Advisor

Gopal Mulukutla


Metabolism in aquatic ecosystems influences food webs and water quality but is potentially altered by changes in land use and flow regime. The interacting effects of storms and land use on stream metabolism are largely understudied. The goal of this study was to understand how flow variability and land use interact to affect biogeochemical cycling in headwater streams. In situ measurements of stage, water temperature, and dissolved oxygen (DO) were made in three headwater streams draining different land uses (urban, forest, and mixed) and one larger river site from July 2017 to January 2018. All streams were located in the Oyster River Watershed in coastal New Hampshire, USA. Metabolism was quantified using the single station dissolved oxygen method and the streamMetabolizer package in the R statistical program. StreamMetabolizer is a state space inverse Bayesian model with partial pooling to constrain process error. It was hypothesized that Gross Primary Productivity (GPP) and ecosystem respiration (ER) in the urbanized stream (URB) would be higher than the mixed (MIX) and forested (FOR) streams because of more open canopy cover and higher nutrients, but that GPP and ER would decline more following storm events because of the flashier (larger and quicker flows) hydrology. It was also hypothesized that GPP and ER in the larger mainstem (MAIN) would remain constant following storms due to more attenuated storm peaks. All streams were net heterotrophic with GPP estimates from 0.0 to 0.96 g O2 m-2 day-1 and ER from 0.0 to 14.2 g O2 m-2 day-1. GPP was higher at the urban stream (0.32 g O2 m-2 day-1) than at all other streams. Pre and post storm disturbance metabolism estimates were compared across all streams for 11 to 28 storms. In general, GPP declined more due to storms than ER at the urbanized and forested streams respectively. Mainstem GPP and ER were most affected by storms; however; GPP and ER were very low. The urbanized stream demonstrated that, despite being subject to constant flashy flow events and being in a constant state of recovery, urbanized streams could rebound quickly and still exhibit high GPP and ER. Future changes in global climate and land use could lead to more frequent and more harmful episodic disturbances in headwater streams.