Table of Contents

• Coverage
• Dataset Description
  • Methods & Sampling
  • Data Processing Description
  • BCO-DMO Processing Description
• Data Files
• Related Publications
• Parameters
• Instruments
• Deployments
• Project Information
• Funding

Water chemistry data, including dissolved and particulate phosphorus, chlorophyll a, carbon dioxide, total dissolved inorganic carbon, and dissolved organic carbon were collected from a Lake Michigan transect between Milwaukee, WI and Muskegon MI from 2017 to 2020.

Field sampling: Research cruises were conducted on& R/V Neeskay and R/V Osprey. Water samples were collected using 5-liter Niskin sampling bottles suspended on a 1/4" cable from a hydrographic winch. Immediately after collection, samples were transferred to 4-liter HDPE sample bottles. Sample bottles were rinsed with sample water 3 times before filling. Prior to use, sample bottles were acid washed (48 hours in 5% HCl), followed by multiple rinses with distilled, deionized water. Samples in bottles were stored in a cooler on ice until return to the laboratory. Samples were filtered immediately upon return to the laboratory (usually less than 8 hours after collection). Samples were filtered through pre-combusted Whatman GF/F filters. Filters were retained for particulate Phosphorus (P), stable isotope (particulate Carbon (C) and Nitrogen (N)), and chlorophyll a analyses. At least twice during the field season, field blanks are collected, which consist of clean bottles brought into the field where they are filled with distilled water, followed by analysis for dissolved and particulate phosphorus.

Nutrients: Samples were collected and analyzed as described in Mosley and Bootsma (2015). Soluble Reactive Phosphorus (SRP) was analyzed using the standard molybdate method and a 10 cm path length in the spectrophotometer. Total Dissolved Phosphorus (TDP) and Particulate Phosphorus (PP) were digested to convert to phosphate, followed by analysis with the standard molybdate method. SRP and TDP were measured within 12 hours of sample filtration.

Chlorophyll a: Samples were collected and analyzed as described in Mosley and Bootsma (2015). Chl a was extracted with a 68:27:5 methanol-acetone-deionized water extraction solvent for 24 hours at -28 degrees Celsius and measured on a Turner Model 10 Series fluorometer, which was calibrated using a chlorophyll extract, the concentration of which was determined spectrophotometrically (Stainton et al. 1977).

CO₂ / DIC: Samples for CO2 and Dissolved Inorganic Carbon (DIC) analyses were collected in stoppered 120 mL glass serum bottles. Prior to sampling, bottles were flushed with nitrogen gas and then evacuated, to ensure they contained no CO2. At the time of sampling, a double-ended needle was inserted into the discharge tube of the Niskin bottle while water was flowing out, and the other end of the needle was inserted through the rubber cap of the serum sample bottle, allowing the vacuum in the bottle to draw in the sample water. The bottle was filled approximately 3/4. CO2 and DIC analyses were carried out following the method described by Davies et al. (2003). Briefly, 50 microliter (uL) subsamples are taken from the bottle headspace using a pressure-lok syringe and injected into a gas chromatograph, calibrated with CO2 standard gases. Samples are run in triplicate. Dissolved CO2 is then determined based on the temperature-dependent solubility of CO2, corrected for CO2 lost to the headspace and for the change in inorganic carbon species distribution accompanying the CO2 loss to headspace. Following CO2 analysis, samples are acidified by adding 150 uL of concentrated phosphoric acid, converting all inorganic carbon to CO2, after which the above analysis was repeated to determine total dissolved inorganic carbon concentration. In-lake CO2 concentrations are determined by correcting for any difference between in situ temperature and temperature at time of analysis, which affects the inorganic carbon partitioning coefficients. CO2 samples were measured within 24 hours of collection, and DIC samples were measured within 3 days of collection.

Continuous CO2: The components of the continuous CO2 monitoring system include a peristaltic pump that forces water through an air-water equilibrator (Membrana mini-module membrane contactor). Reverse-flow air from the equilibrator is pumped through desiccant, after which it flows through an infrared gas analyzer (Li-Cor Li-820) which measures the partial pressure of CO2 normalized to 1 atmosphere. The system also includes a temperature sensor and a WETLabs flow-through fluorometer. The system is controlled by a Campbell CR1000 Controller / Datalogger. Input from a GPS on the ship's upper deck allows all data to be geo-referenced. The system is mounted in the engine room of the Lake Express high-speed ferry, where it draws water from a sea chest that has a residence time of several seconds. Additional details are provided in Zagorski and Bootsma (2006).

Stable isotopes: Samples for stable isotope (13C:12C and 15N:14N ratios) analyses were collected by filtering lake water samples through GF/F glass fiber filters (nominal pore size = 0.7 to 0.8 micrometers (um)). Following filtration, filters were doused with 5% HCl for ~ 3 minutes to remove any inorganic carbon, followed by rinsing with distilled, deionized water. Filters were then freeze dried and packed in tin foil disks. Samples were then analyzed on an isotope ratio mass spectrometer, following the methods as described in Turschak et al. (2014). After every 12th sample, an acetanilide control was run to ensure instrument calibration.

Dissolved organic carbon: 25 ml of filtered water was transferred to an amber glass ampule and acidified to a pH of less than 2 by adding 2-3 drops of 1 N hydrochloric acid (HCl), converting all inorganic carbon to CO2, which was then purged from the sample bubbling with carbon-free gas prior to OC analysis. DOC was then measured using the combustion catalytic oxidation method on a total organic carbon analyzer (Shimadzu TOC-L analyzer equipped with an ASI-5000 auto sampler). The analyzer was calibrated with a dilution series of reagent grade potassium hydrogen phthalate in 0.3 molar hydrochloric acid.

Data Processing:
All nutrient data are stored in a common database. Following analyses, nutrient standard curves are examined to ensure that calibration coefficients are within the range of variability of a long-term (5-year) dataset (±3%). Fluorometer measurements are entered into a spreadsheet containing the fluorometer calibration coefficients, which are used to calculate chlorophyll a and phaeophytin concentrations. The fluorometer is calibrated annually against extracted chlorophyll a standards. CO2 and DIC gas chromatograph measurements are entered into a spreadsheet program that calculates all inorganic carbon species concentrations, as well as pH and carbonate alkalinity. Concentrations are then corrected for any temperature difference between in situ and time of analysis. Stable isotope measurements are stored in a stable isotope database, while DOC measurement data are stored along with nutrient, chlorophyll and inorganic carbon measurements in a chemistry database.

Versions 1 and 2:
- modified parameter names to conform with BCO-DMO naming conventions;
- re-formatted date to ISO format;
- replaced missing data with nd ("no data");
- updated to version 2 on 2019-05-20.

Version 3:
- converted dates to YYYY-MM-DD format;
- converted times to hh:mm format;
- modified parameter names to conform with BCO-DMO naming conventions;
- removed 'nd' as the missing data identifier (missing data are empty/blank in the final CSV file);
- updated to version 3 on 2023-12-20 as file name "737176_v3_lake_michigan_chemistry.csv".

Overview:
While benthic filter feeders are known to influence plankton and nutrient dynamics in shallow marine and freshwater systems, their role is generally considered to be minor in large, deep systems. However, recent evidence indicates that profundal quagga mussels (Dreissena rostriformis bugensis) have dramatically altered energy flow and nutrient cycling in the Laurentian Great Lakes and other larges aquatic systems, so that conventional nutrient-plankton paradigms no longer apply. Observed rates of phosphorus grazing by profundal quagga mussels in Lake Michigan exceed the passive settling rates by nearly an order of magnitude, even under stably stratified conditions. We hypothesize that the apparently enhanced particle deliver rate to the lake bottom results from high filtration capacity combined with vertical mixing processes that advect phytoplankton from the euphotic zone to the near-bottom layer. However, the role of hydrodynamics is unclear, because these processes are poorly characterized both within the hypolimnion as a whole and within the near-bottom layer. In addition, the implications for phytoplankton and nutrient dynamics are unclear, as mussels are also important nutrient recyclers. In the proposed interdisciplinary research project, state-of-the-art instruments and analytical tools will be deployed in Lake Michigan to quantify these critical dynamic processes, including boundary layer turbulence, mussel grazing, excretion and egestion, and benthic fluxes of carbon and phosphorus. Empirical data will be used to calibrate a 3D hydrodynamic-biogeochemical model to test our hypotheses.

Intellectual Merit:
This collaborative biophysical project is structured around two primary questions: 1) What role do profundal dreissenid mussels play in large lake carbon and nutrient cycles? 2) How are mussel grazing and the fate of nutrients recycled by mussels modulated by hydrodynamics at scales ranging from mm (benthic boundary layer) to meters (entire water column)? The project will improve the ability to model nutrient and carbon dynamics in coastal and lacustrine waters where benthic filter-feeders are a significant portion of the biota. By so doing, it will address the overarching question of how plankton and nutrient dynamics in large, deep lakes with abundant profundal filter feeders differ from the conventional paradigm described by previous models. Additionally, the project will quantify and characterize boundary layer turbulence for benthic boundary layers in large, deep lakes, including near-bed turbulence produced by benthic filter feeders.

Broader Impacts:
The project will provide new insight into the impacts of invasive dreissenid mussels, which are now threatening many large lakes and reservoirs across the United States. Dreissenid mussels appear to be responsible for a number of major changes that have occurred in the Great Lakes, including declines of pelagic plankton populations, declines in fish populations, and, ironically, nuisance algal blooms in the nearshore zone. As a result, conventional management models no longer apply, and managers are uncertain about appropriate nutrient loading targets and fish stocking levels. The data and models resulting from this project will help to guide those decisions. Additionally, the project will provide insight to bottom boundary layer physics, with applicability to other large lakes, atidal coastal seas, and the deep ocean. The project will leverage the collaboration and promote interdisciplinary education for undergraduate and graduate students from two universities (UW-Milwaukee and Purdue). The project will support 3 Ph.D. students and provide structured research experiences to undergraduates through a summer research program. The project will also promote education of future aquatic scientists by hosting a Biophysical Coupling Workshop for graduate students who participate in the annual IAGLR conferences, and the workshop lectures will be published for general access through ASLO e-Lectures and on an open-access project website.

Background publications are available at:
http://onlinelibrary.wiley.com/doi/10.1002/2014JC010506/full
http://link.springer.com/article/10.1007/s00348-012-1265-9
http://aslo.net/lomethods/free/2009/0169.pdf
http://www.sciencedirect.com/science/article/pii/S0380133015001458

Note: This is an NSF Collaborative Research Project.