SFS Annual Meeting

5/23/2019  |   11:00 - 11:15   |  251 AB

DEVELOPMENT OF EXTENSIBLE SOFTWARE TO INFER ECOSYSTEM METABOLIC RATES FROM MULTIVARIATE METABOLITE SIGNALS IN STREAMS Advances in sensor technology are rapidly increasing the number of different metabolite signals that can be measured in situ in streams over long periods of time. We define “metabolite signal” as relatively high-frequency measurements of chemical concentrations that are influenced by ecosystem metabolism (e.g., dissolved oxygen, dissolved inorganic carbon, nitrate, etc.). To keep pace with continued advances in sensor development, we suggest that the inferential modeling software needed to estimate ecosystem metabolic rates from metabolite signals should be developed in an extensible, object-oriented design. Here, we use the Unified Modeling Language (UML) to share our current design of an R package (in active development) intended to meet these needs. Abstract implementations in the code base are designed to minimize the amount of new code (i.e. extensions of R6 classes) necessary to make metabolic inferences from new types of metabolite signals and using new types of models. We provide examples of a Bayesian optimization scheme and multivariate metabolic models to illustrate use of the design, and we demonstrate how confidence in inferences from new signals and new models are assessed using Monte Carlo analyses.

Robert Payn (Primary Presenter/Author), Montana State University, Montana Institute on Ecosystems, rpayn@montana.edu;


Elizabeth Mohr (Co-Presenter/Co-Author), Montana State University, elizabethjmohr@gmail.com;


Elfrida Isaksen-Swensen (Co-Presenter/Co-Author), Montana State University, frida.e.swensen@gmail.com;


Todd Schlotfeldt (Co-Presenter/Co-Author), Montana State University, toddschlotfeldt@gmail.com;


Geoffrey Poole (Co-Presenter/Co-Author), Montana State University, Montana Institute on Ecosystems, gpoole@montana.edu ;


Ann Marie Reinhold (Co-Presenter/Co-Author), Montana State University, Montana Institute on Ecosystems, reinhold@montana.edu;


Michael DeGrandpre (Co-Presenter/Co-Author), University of Montana, michael.degrandpre@umontana.edu;


Joanna Blaszczak (Co-Presenter/Co-Author), University of Nevada, Reno, jblaszczak@unr.edu;


Robert O. Hall (Co-Presenter/Co-Author), Flathead Lake Biological Station, University of Montana, bob.hall@flbs.umt.edu;


Clemente Izurieta (Co-Presenter/Co-Author), Montana State University, Montana Institute on Ecosystems, clemente.izurieta@cs.montana.edu;


H. Maurice Valett (Co-Presenter/Co-Author), University of Montana, Division of Biological Sciences, maury.valett@umontana.edu;


Presentation:
This presentation has not yet been uploaded.

5/23/2019  |   11:15 - 11:30   |  251 AB

THE PREVALENCE OF HYPOXIA IN RIVER NETWORKS AND THE ROLE OF NOT SO DEAD ZONES IN ELEMENT CYCLING Though well understood to be a widespread challenge for water quality managers, hypoxia (defined as [DO] < 5mgL-1) in fluvial systems is understudied by stream ecologists who tend to work in well-aerated stream segments. A high resolution synoptic sampling of 20km of the New Hope Creek river network in the Piedmont of North Carolina found that more than 40% of the network was hypoxic. Time series data from nine locations along this network show that some hypoxic segments oscillate on a diel cycle while others remain hypoxic for weeks between storm flows. Hypoxia was more prevalent in incised channel reaches where slopes have declined as a result of flow impoundments. In addition to being stressful for many aquatic organisms, hypoxia has implications for microbial metabolism. In many locations along this continuum, we measured high co-occurring concentrations of methane, nitrous oxide, hydrogen sulfide and reduced iron. Predictions of metabolic pathway rates based on these concentrations and on thermodynamic constraints (using the GANGSTA model) indicate that anaerobic metabolism is significant for energy cycling. Incorporating hypoxia into the conceptual model of river networks provides a more complete understanding of metabolism in rivers.

Alice M. Carter (Primary Presenter/Author), Flathead Lake Biological Station, University of Montana, alicecarter05@gmail.com;


Joanna Blaszczak (Co-Presenter/Co-Author), University of Nevada, Reno, jblaszczak@unr.edu;


Martin Doyle (Co-Presenter/Co-Author), Duke University, martin.doyle@duke.edu;


Ashley Helton (Co-Presenter/Co-Author), University of Connecticut, ashley.helton@uconn.edu;


Jim Heffernan (Co-Presenter/Co-Author), Duke University, james.heffernan@duke.edu;


Emily Bernhardt (Co-Presenter/Co-Author), Duke University, emily.bernhardt@duke.edu;


Presentation:
This presentation has not yet been uploaded.

5/23/2019  |   11:30 - 11:45   |  251 AB

THE ROLE OF DENITRIFICATION IN CARBON REMOVAL FROM STREAM ECOSYSTEMS For every 4 mol of nitrate used in denitrification, 5 mol of organic carbon (C) is removed from an ecosystem. This potentially large removal of carbon often goes overlooked when considering C fluxes in streams, which are often derived from O² based modeling and do not account for anaerobic forms of respiration including denitrification. Our objective is to evaluate the role of denitrification in C removal in 36 streams across the United States. Denitrification measured using acetylene block assays ranged from 0-22900 µg N/m²/hr, with the highest rates measured in McDiffitt Creek, KS during a drought and under hypoxic conditions. We estimate that denitrification removed 24536 µg C/m²/hr in this stream, when aerobic respiration rates measured using light-dark chambers were close to 0. In contrast, in Hop Brook, MA, aerobic respiration rates were high (29000 µg C/m²/hr), but denitrification removed an additional 2000 µg C/m²/hr; a roughly 6% increase over aerobic respiration measurements alone. The role denitrification plays in carbon removal can range widely depending on stream and substrate conditions but must be considered to accurately develop carbon budgets.

Kevin Nevorski (Primary Presenter/Author), Michigan Technological University, kcnevors@mtu.edu;


Amy Marcarelli (Co-Presenter/Co-Author), Michigan Technological University, ammarcar@mtu.edu;


Presentation:
This presentation has not yet been uploaded.

5/23/2019  |   11:45 - 12:00   |  251 AB

FLOOD-PULSE CONTROLS ON HETEROTOPIC PROCESSES AND FOOD-WEB PRODUCTIVITY IN THE MEKONG RIVER BASIN The magnitude and pathways by which carbon and energy enter food webs is of fundamental ecological importance and relevant to ecological theory describing community interaction, spatial subsidies, and ecosystem function. The Lower Mekong River basin, including Tonle Sap Lake, is a classic flood-pulse ecosystems and one of the largest inland fisheries in the world that provides a critical source of animal protein for much of Southeast Asia. Recent research has shown that fishery productivity is positively related to flood-pulse magnitude and the duration of the low water period between floods. This talk will summarize a series of investigations into mechanisms driving this pattern, including flood-pulse controls on ecosystem metabolism, oxidative and non-oxidative heterotropic processes, and tropic transfers. Floodplain GPP and ER are strongly related to time of inundation as is both methanogensis and methanotrophy. A combination of stable isotope and fatty acid tracers identify distinct carbon transfer pathways originating from strong autotrophic and heterotrophic production regimes. Ultimately, flood-pulse controls on basal resources and food web dynamics leads to large differences among species in key macromolecules critical for human nutrition, with implications for human food security.

Benjamin Miller (Co-Presenter/Co-Author), University of Washington, blm8@uw.edu;


John Sabo (Co-Presenter/Co-Author), Arizona State University, John.L.Sabo@asu.edu;


Gordon Holtgrieve (Primary Presenter/Author), University of Washington, gholt@uw.edu;


Presentation:
This presentation has not yet been uploaded.