NOAA’s Great Lakes harmful algal bloom program Part 1: Prediction, monitoring and experiments

Timothy Davis HABs occur throughout the Great Lakes basin Potentially -toxic cyanobacteria of Lake Erie

Anabaena / Dolichospermum Planktothrix

Microcystin Anatoxin-a Overarching research statement: Understanding the drivers of bloom ecology will aid in enhancing predictive models that forecast bloom size, location AND toxicity An integrated approach to studying HABs NOAA’s cross-line office CHAB team

Monitoring & Experiments Timothy Davis Tom Johengen* Duane Gossiaux Ashley Burtner* Danna Palladino* Heidi Purcell* Hank Vanderploeg Alicia Ritzenthaler* Remote Sensing Steve Ruberg Richard Stumpf Andrea Vander Woude* Steve Constant George Leshkevich Ron Muzzi Russ Miller Modeling Richard Stumpf Eric Anderson Tim Wynne Danielle Dupuy Mark Rowe* Greg Lang Communications BOLD = HAB team leads Sonia Joseph Joshi * * = CILER employees Joe Smith* NOAA and academic partners predict bloom severity

2015 was the largest bloom this century

The forecast uses a 12-year Lake Erie nutrient flow data set, collected by Heidelberg University’s National Center for Water Quality Research, and analysis of satellite data from the European Space Agency’s ENVISAT and NASA’s Terra and Aqua satellites. Western Lake Erie weekly sampling NOAA GLERL HAB webpage

NOAA GLERL HAB webpage Critical water quality data provided to stakeholders

• Weekly data shared with over 30 regional partners

Data for 8/10/2015 Particulate Microcystins (g/L • Allows stakeholders to view microcystinLR equivalents): 9.19 trends in bloom biomass and toxicity

• Surface and bottom measurements

8/15 NOAA GLERL HAB webpage MODIS HAB Detection Model Initialization

From HAB Bulletin

Vertical Mixing Analysis

HAB Tracker

5day Forecast

NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION I GREAT LAKES ENVIRONMENTAL RESEARCH LAB I ANN ARBOR MI 14 NOAA GLERL HAB webpage Autonomous near real-time toxin detection for Lake Erie Funded by EPA-GLRI

Will be extremely valuable in the development of more accurate bloom forecasting models Deployment of ESP niagara

Bolted lifting assembly

Pressure housing for ESP − Rated to 48 m depth 3-way sample valve Sampling manifold • 5’6” tall • 5ft x 5ft footprint • < PSI pressure on ESP battery assemblies footprint (and opposite) • Deck weight ~ 1800 lbs Each hold 200 D cell batteries • In-water weight ~ 750lbs Tagline attachment points

June ‘16 July Sept/Oct

Test of communication Science testing at Ohio State Full Mission Deployment system and general University Stone Laboratory on • Microcystin sampling every other day operations South Bass/Gibraltar Islands • Archived samples for future DNA sequencing Toledo Water Intake (Station 12) ) -1

Weekly sampling reveals important (mg L trends that highlight potential ) environmental drivers and inform -1 (μg L (μg future experiments

• Toxicity changes throughout the

bloom (nM) 2 O 2 H • Relationship between nitrogen,

toxicity and toxic Microcystis ) -1 (μg L (μg • Microcystis blooms occur even when phosphorus concentrations are low ) -1 (μg L (μg

• Other factors beyond nutrients Toxicity may be important in driving bloom structure and function % Toxic Toxic % Microcystis Analysis of Microcystis subpopulations:

Other cyanobacteria

Microcystis (16S rDNA)

NIES1050 NIES44 Potentially-toxic Non-toxic Microcystis (toxic) (non-toxic) Microcystis (16S without mcy ) (mcy -containing)

% toxic Microcystis = proportion of Microcystis cells containing the genetics to produce microcystins (mcyD or E / 16S x 100)

Determined by (1) qPCR and (2) metagenomics What is omics?

Also metabolomics, lipidomics, etc….

Dick and Lam, Elements Magazine , in review Eco-transcriptomic surveys of LE CHABs

Harke, Davis et al., 2016; ES&T Microcystis is an excellent P scavenger Soluble reactive reactive SolubleP [µM]

Harke, Davis et al., 2016 ES&T Western Lake Erie relationships Toxic strains decline with lower N

20 1.2 Par culate MCs SRP Nitrate

% Toxic Microcystis % Toxic Microcystis 1.0 )

-1 >100µm = 63 ± 31% >100µm = 4 ± 3% 15 53100µm = 48 ± 31% 53100µm = 29 ± 26%

± ± Nitrate conc. (mg L 353µm = 5.6 5.2% 353µm = 5.7 10.1% 0.8 ) -1 Microcystins = 4.0 ± 2.6 µg L 1 Microcystins = 0.7 ± 0.6 µg L 1

± ± 10 (Average SD of 5 sampling dates) (Average SD of 4 sampling dates) 0.6 -1 SRP conc. (μg L (μg conc. SRP ) 0.4

5 Particulate MCs (μg MC-LR equiv. L equiv. MC-LR (μg MCs Particulate Nitrate = 0.41 ± 0.47 mg L 1 Nitrate = 0.14 ± 0.03 mg L 1 0.2 Ammonium = 2.8 ± 2.6 µg L 1 Ammonium = 2.3 ± 2.7 µg L 1 SRP = 1.9 ± 1.4 µg L 1 SRP = 10.5 ± 5.2 µg L 1

0 0.0 4 4 4 4 4 4 4 4 4 4 4 /1 /1 /14 /1 /1 /1 /14 /1 /1 /1 /1 10 17 /1 /8 29 12/1 26 /2 23 30/14 21 28 5/27/14 6/3/14 6/ 6/ 6/24/14 7 7 7/15/14 7/22/14 7/ 8/5 8/ 8/19/14 8/ 9 9/9/14 9/16/14 9/ 9/ 10/7/1 0/14/14 0/ 0/ 1 1 1 Nitrogen constrains growth and toxicity in Sandusky Bay +NO +PO +NH +Urea only Urea Urea + PO NO NH Control 3 4 4 3 4 + PO + PO only only only 4 4 4

After 48 hours of incubation

Davis et al., 2015 ES&T From:Gobler, Burkholder,

Soluble Reactive P concentrations Davis , et al., 2016, Harmful Harmful Algae , 2016, al., et Toledo Water Intake (Station 12) ) -1

Weekly sampling reveals important (mg L trends that highlight potential ) environmental drivers and inform -1 (μg L (μg future experiments

• Toxicity changes throughout the

bloom (nM) 2 O 2 H • Relationship between nitrogen,

toxicity and toxic Microcystis ) -1 (μg L (μg • Microcystis blooms occur even when phosphorus concentrations are low ) -1 (μg L (μg

• Other factors beyond nutrients Toxicity may be important in driving bloom structure and function % Toxic Toxic % Microcystis Distrabution of antioxidant enzymes among select cyanobacterial genera

Catalase (2(H H2O2)2O2) KatG (H2O2(H 2O2 oror ROOH)ROOH) External

Mn Catalase (2(H H2O2)2O2) cytochrome c peroxidase (ROOH(H 2O2 or or ROOH) H2O2) dyP-type peroxidase (ROOH(H O oror ROOH)H2O2) Haloperoxidase (ROOH(H O oror ROOH)H2O2) 2 2 2 2 Internal Peroxiredoxin (H2O2(H 2O2 oror ROOH) ROOH) glutathione peroxidase (ROOH) - superoxide dismutase (O2-)(O 2 )

100 copy

90 gene

1 80

70 least

at

60 50 40 containing

30 20 genomes

10 of

% 0 Microcys s Anabaena Planktothrix Pseudanabaena Prochlorococcus Synechococcus n = 19 n = 6 n = 16 n = 13 n = 45 n = 36 n = number of genomes queried

MS = Methanol soluble MCs (Free) MI = Methanol insoluble MCs (protein-bound) ROS experimental design Sample collection: • Chlorophyll a • Phycocyanin • Particulate microcystins •Control • Dissolved microcystins 48 hours @ • Lugols •+N +P (20 & 2 µM) i • DNA •+ROS (H O ; 1µM) ambient light 2 2 and • RNA 2L •+Ni+P + H 2O2 temperature • Picoplankton bottles • Dissolved and total nutrients N = NH NO x3 x3 i 4 3 • CDOM/FDOM P = KH 2PO 4 • DNA Sterivex • RNA sterivex

• H2O2 • PhytoPAM

• Four experiment were conducted

• At t = 2, 4, 8, 12, 24 subsamples were collected for RNA and H 2O2 • At t = 48 all parameters listed above were sampled (one per bottle) except for total nutrients and RNA sterivex which were omitted

• Based on calculated H 2O2 degradation rate in LE, 500nM H 2O2 was added every 2 hours to maintain a 1µM concentration • Nutrients were replenished every 12 hours to maintain concentrations of 20 & 2 µM for N and P, respectively Conjugated MCs change over time but not influenced by ROS

August 20, 2014 September 3, 2014

20 Par culate Conj. Par culate Free 8 ) Par culate Conj. Par culate Free -1

16 6

12

4 8

2 4

Microcystins (μgequivalents MC-LR L 0 0 Ini al Control ROS N+P ROS+N+P Ini al Control ROS N+P ROS+N+P

100% 100% 12 ± 2 8 ± 1 13 ± 1 15 ± 2 18 ± 1

49 ± 8 47 ± 4 46 ± 3 23 ± 2 31 ± 4 80% 80%

60% 60%

40% 40% % totalmicrocystins %

20% 20%

0% 0% Ini al Control ROS N+P ROS+N+P Ini al Control ROS N+P ROS+N+P Microcystis protected from external ROS by synergistic relationships?

Distribu on of anitoxidant enzymes among select cyanobacterial genera

Catalase KatG Mn Catalase

cytochrome c peroxidase dyP-type peroxidase Haloperoxidase

Peroxiredoxin glutathione peroxidase superoxide dismutase

100 copy

90

gene 80

1

70 least

60 at

50

40 containing

30

20 genomes

10 of

% 0 Microcys s Anabaena Planktothrix Pseudanabaena Prochlorococcus Synechococcus Genomic database will fuel research in the Great Lakes and beyond

• Searchable genomic database for CHAB genera including Great Lakes strains

• Publicly available resource

• Serves as a link between environmental and genomic data

• Critical to further understand the response of CHAB genera to environmental drivers Future directions and challenges

• Develop an ESP network for Western Lake Erie

• Further develop the ESP capabilities

• Investigate the ecological adaptations of Great Lakes CHAB species Microcystins >100 congeners • Further understanding the interactive roles of light, nutrients, and temperature on toxin production and community composition

• Investigate changes in microcystin congeners over time Questions?

Place Photo Here