In the 2003 and 2004 blooms, there were distinct physical and chemical differences between the plume of the Maumee River flowing into western Lake Erie and non-plume waters. Highest Microcystis density was closely aligned with the plume.  We have also observed seasonal and inter-annual differences in plume conditions between non-bloom years (2002) and bloom years (2003, 2004).  Knowledge gained in previous studies of cyanobacteria ecology suggest that in the clearer waters of the post-Dreissena era, conditions will generally favor green algae and diatoms, but that we may expect the highly buoyant Microcystis to be favored within turbid river plumes.  In pre-Dreissena years, more evenly distributed turbidity would likely have led to more spatially diffuse growth of Microcystis, not creating the impression of “blooms.”

We will approach our objective using two strategies:  1) We will enhance our already-extensive collection of field data with the addition of targeted light, nutrient, and phytoplankton data from the lake and Maumee River to determine whether contrasting physical and chemical conditions inside and outside of the plume can predict bloom events.  2) We will test our hypotheses using molecular and physiological markers of stress to determine the response of phytoplankton communities to differing light conditions both in the field and in laboratory incubations.  These measurements will indicate the relative physiological status and potential growth of Microcystis and competing algal taxa within natural algal assemblages, inside and outside of the river plume.

     In August 2003 and 2004, massive blooms of the cyanobacteria Microcystis aeruginosa formed in western Lake Erie and persisted for nearly a month (Figure 1).  Surface scums of Microcystis containing high concentrations of the toxin microcystin washed ashore in Michigan and Ohio, resulting in foul-smelling, rotting, algal mats.  Beaches and recreational boating areas were rendered unusable and sport fishing was adversely affected.  The Microcystis bloom of 2003 was perhaps the most severe in Lake Erie’s recent history, but of greater concern is a trend towards increasing frequency of Microcystis blooms in the last decade.  Based on our field observations in 2003 and 2004, we hypothesize that low-light and high-nutrient conditions in the Maumee River plume may potentially explain why outer Maumee Bay, the epicenter of both 2003 and 2004 blooms, is especially prone to bloom formation.  We will test our hypothesis using a combination of traditional field measurements and recently-developed molecular and physiological techniques.  Recommendations resulting from this research may include controls on sediment loading and shifts in the timing of dredging activity.

If, as we hypothesize, high-turbidity/low-light conditions in river plumes play an important role in the promotion of Microcystis blooms, the results of our research will help to focus future efforts in preventing blooms.  The results of our study could lead to recommendations that would seek to suppress Microcystis blooms by reducing sediment loading, and therefore turbidity into rivers.  Additional recommendations might include shifting shipping-channel dredging activities (which produce additional turbidity) from July and August when blooms begin to other months with lower risk of blooms.  Results obtained from the Maumee River would be applicable to other turbid rivers that produce plumes as they enter the great lakes such as the Saginaw River on Lake Huron and the Sandusky River on Lake Erie.  This project will also provide advanced training for a graduate student at the University of Toledo, and ecological field and laboratory experience for summer students through the NSF-supported Research Experience for Undergraduates Program based at the University of Toledo.  The results of our research will be disseminated by publication in peer-reviewed journals and by presentation at scientific conferences.  Field data from this project will also be incorporated into the data set produced by our monitoring efforts which have been ongoing since 2002.  By the completion of the project we will have a six-year time series (2002-2007) on Maumee Bay and western Lake Erie that will likely prove extremely valuable for future research.

Light:  We propose to test our hypotheses concerning the existence of three distinct light climates in and outside of the Maumee River plume by extensive field measurements of underwater irradiance in the Maumee Bay area of western Lake Erie.  Measurements will be taken at 10-day intervals from mid-April through mid-October using Lake Erie Center research boats – a 25-ft. Sportcraft for offshore work and 18-ft. zodiac for shallow water.  Sampling intensity will increase to weekly cruises during the major bloom season (late-July through September).  Measurements will be taken every 4 km along an approximately 80 km route that extends from the mouth of the Maumee River to open waters 25 km offshore and transects the river plume several times.  This route has been chosen, in part, to take advantage of locations for which historic data is available (Fraleigh 1979).  At each stop, the following data will be collected at 2 meter intervals from surface to bottom using a Hydrolab datasonde (model 4a): temperature, turbidity, in situ fluorescence, conductivity, and dissolved oxygen.  Temperature and dissolved oxygen (mg L^-1) will be used to indicate the degree and duration of vertical mixing.  Wind speed, recorded by monitoring stations in the bay and western Lake Erie (NOAA/NWS) will also be used to determine the average degree of mixing for summer months.  Turbidity (NTUs) is a direct measure of scattering due to suspended particles.  In situ fluorescence will be used along with extracted chlorophyll samples to determine chlorophyll a concentrations and conductivity (mS cm^-1) will be used to track the river plume. 

Additional measurements and collections will be made at 6 locations chosen to represent inner bay, plume, and offshore non-plume conditions.  At these locations, in addition to Secchi depth, underwater profiles of photosynthetically active radiation (PAR) will be recorded using a photometer equipped with a spherical sensor.  Light profiles will be used to calculate the light extinction coefficient (K_PAR).  Measurements of K_PAR and Secchi depth will be used to develop a regression that will predict K_PAR at all locations where Secchi depth was measured.  Preliminary data from 2003 indicates a strong correlation between K_PAR and Secchi depth (R² = 0.94).  In turn, K_PAR will be used to calculate the depth of the euphotic zone.  With this information, we will be able to map areas of Maumee Bay, the river plume, and offshore non-plume waters to determine in which regions light reaches the lake floor.

Nutrients and Chlorophyll a

Water for nutrient and chlorophyll a analysis will be collected during each cruise at six locations chosen to represent inner bay, plume, and offshore non-plume conditions.  Water samples will be collected at 1m, mid-water, and near-bottom depths using a Van Dorn sampler.  Water samples will be processed within six hours of collection and preserved for later analysis of nutrients (total phosphorus, soluble reactive phosphorus, total nitrogen, nitrate, nitrite, ammonia, silica) and chlorophyll a.  Nutrient analysis will be conducted by the Water Quality Laboratory at Heidelberg College.  Chlorophyll a analysis will be performed at the L. Erie Center.  On four occasions (2 pre-bloom, 2-during bloom) vertical distribution, abundance, and growth of algae and Microcystis will be determined on samples using microscopy and molecular and physiological status techniques described in the Molecular and Physiological Techniques section.

Lake water and algae from the river plume will be collected in 20 L bottles and kept in the dark for transportation to the University of Toledo Plant Science Center.  Water will be transferred to 36 L experimental chambers measuring 60 cm long x 10 cm wide x 60 cm tall.  If the samples do not contain Microcystis (a non-bloom year), Microcystis will be added from cultures obtained from Bowling Green State University.  Tanks will be darkened on all sides, with the open top exposed to natural sunlight.  Light extinction will be measured using a 4-pi submersible photometer.  In turbidity treatments, powdered clay will be added until light penetration is reduced to 1% at one half the depth of the chamber.  Water in the chambers will be stirred by blowing air across the long axis of the chamber surface.  If additional mixing is required, pumps will be installed.  Since phosphate may adsorb to clay particles, additional phosphate may need to be added to maintain initial phosphate concentrations.  Incubations will be performed in triplicate replicates and run for 1-5 days (period to be determined in early trials).  At the beginning and end of each incubation, water samples will be collected from the chambers for analysis.  Microcystis will be separated from other algae by allowing samples to sit for 10 minutes during which Microcystis will float to the surface where it can be siphoned off.  Purity of Microcystis sub-samples will be determined microscopically.  The following molecular and physiological analyses will then be performed.

Table 1 lists the parameters and expected responses that will be used to measure the physiological status and growth of Microcystis relative to competing algae.  The responses will be used to test the hypothesis that Microcystis has an advantage in highly turbid river plumes.

RNA-based assessment Microcystis environmental fitness.

Bacteria (including cyanobacteria, e.g. Microcystis) will rarely expend energy unnecessarily and are generally found in environmental samples in a state of energy conservation.  Stressed bacteria will limit energy expenditure by entering a state of limited metabolic activity.  Conversely, during conditions of nutrient and/or light optima, bacteria will exhibit rapid growth and high levels of metabolic activity (e.g. cyanobacterial blooms).  Analysis of RNA provides a reliable measure of organism activity, gene expression, and general fitness in the environment as evidenced by the correlation of cellular activity with cellular ribosome content and changes in mRNA levels (Sheridan et al., 1998).  Furthermore, due to its labile nature, RNA detection can be correlated with recent gene expression and therefore organism activity.  To capitalize on the sensitivity of RNA-based methodology, we will determine the fitness of Microcystis in conditions of differing turbidity by applying two approaches that utilize detection of cellular RNA.

Fluorescence in situ hybridization (FISH) analysis of Microcystis populations.  

Enumeration of active Microcystis spp. and cyanobacteria spp. will be performed by FISH according to the method of Zarda et al. (1997).  This method will quantify the abundance of active Microcystis within and outside of the turbidity plume, which will be used to determine the ratio of the abundance of metabolically active Microcystis to that of the total cyanobacteria community.  It is assumed that a high ratio will suggest that site conditions are favorable for Microcystis proliferation, while a low ratio will suggest the opposite. We expect that because the buoyancy regulation capacity of Microcystis enables it to seek out optimal light levels, it is likely that Microcystis spp. will maintain a greater abundance over other cyanobacteria within the plume as compared to outside the plume.

Population structure of active Microcystis and total cyanobacteria. 

Immediately following harvest, water samples (volume dependent on cyanobacterial density) will be filtered, extracted in nucleic acid buffer, and then flash-frozen in liquid nitrogen.  Nucleic acids will be extracted (Sigler and Zeyer 2002), total RNA will be isolated, and reverse transcription (RT) of Microcystis - and total cyanobacterial !6S rRNA will be performed.  The resulting cDNAs will be PCR-amplified according to the Microcystis spp.-specific protocol of Knut et al. (1998) and the total cyanobacteria-specific protocol of Nübel et al. (1997).  The PCR products will be used for two purposes.  First, we will quantify relative differences in Microcystis activity within- and outside of the plume and with depth, and the ratio of Microcystis to total cyanobacteria. These techniques will allow us to compare Microcystis RNA activity with the activity of other cyanobacteria as well as environmental parameters such as turbidity and depth.  Second, we will monitor changes in the active cyanobacteria- and Microcystis population structure using the DGGE genetic fingerprinting method of Sigler et al. (2003).  DGGE analysis will provide quantitative data to determine relative differences in bacterial population dynamics within- and outside of the plume.  Therefore, by understanding how cyanobacteria populations in general, and Microcystis spp. in particular, each respond to changes in their environment, we can begin to describe population-based mechanisms of bloom formation.

In general, total cellular protein (and often N) content increases with cell health and with growth light-level (if no photoinhibition) (Larcher 1996).  Total protein (per g and chlorophyll a) will be determined after extraction with a buffer containing sodium dodecyl sulfate (SDS) detergent and protease inhibitors (e.g., Heckathorn et al. 2002).  Total N (and C, simultaneously) will be determined using the combustion technique (C:N analyzer, Perkin-Elmer model 2400 series II).  Light limitation decreases the content of total non-structural (labile) carbohydrate (TNC) in algae relative to total C (e.g., starch & soluble sugars, relative to cellulose).  TNC will be determined spectrophotometrically, as in Robyt and White (1987).

Light limitation induces the increased production of photosynthetic accessory pigments (e.g., certain carotenoids in all taxa, and phycobilin in cyanobacteria and red algae), and an increase in chlorophyll b relative to chlorophyll a (in green algae); excess light reduces both chlorophyll and accessory pigment content (Hall & Rao 1994; Larcher 1996).  Since we will have a mix of taxa at our site (cyanobacteria, green algae, and possibly others), we will quantify low-light effects on photosynthetic pigments by measuring absorption spectra on samples, and from these, determine the absorption ratios of chl b:a and carotenoids:chl a, and phycobilins:chl a (Hall & Rao 1994); total chl a content will be determined as before (Heckathorn et al. 2002).  These data will also help, along with the DNA profiling data, to determine the relative abundance of algae from different taxa.

            Small (low-molecular-weight) heat-shock proteins (Hsps) are general stress proteins found in all organisms, and these increase in response to virtually all abiotic stresses (Parsell and Lindquist 1994).  Small Hsps typically are expressed only with stress, and play major roles in protection from photoinhibition and increase with high-light stress (e.g., Schroda et al. 1999; Downs et al. 1999).  During oxidative stress, which occurs during nearly all abiotic stresses, especially photoinhibitory high light, oxidative damage to proteins and lipids occur, and a major form of this damage is production of protein carbonyl (Robinson et al. 1999) and lipid peroxides.  The content of small Hsp and protein carbonyl will be determined using immunological methods; lipid peroxidation will be determined spectrophotometrically (using commercial kits).  Total cell protein will be extracted as above, then proteins will be fractionated by SDS-PAGE, transferred to membranes by electroblotting, and detected and quantified using protein specific antibodies and colorimetric reactions (Downs et al. 1999; Preczewski et al. 2000; Heckathorn et al. 2002).  Commercially available antibodies will be used for protein carbonyl, and our own antibodies to several small Hsps (Heckathorn et al. 2002).

To access the effects of light on algal photosynthesis, we will measure the quantum yield of electron transport (Φ_et) of Photosystem II (PSII), PSII efficiency/damage, electron transport rate (ETR), and non-photochemical quenching (NPQ) (Schreiber et al. 1994; Heckathorn et al. 1997; 2002).  Φ_et is typically proportional to net photosynthesis at a given light level, and is highest at very low light, decreases with increasing light, and further decreases during photoinhibition.  PSII efficiency decreases with increasing light (beyond very low levels), and PSII damage occurs at photoinhibitory high-light.  We will measure PSII efficiency by determining the ratio of variable-to-maximum chl fluorescence in both light- and dark-adapted samples (light-adapted decreases indicate damage and/or photo-protection, while dark-adapted decreases indicate damage).  ETR increases with light, unless photoinhibition occurs.  NPQ is a measure of the amount of light energy protectively dissipated as heat, and increases with increasing light level (beyond very low levels).  Lastly, we will determine the relative content of rubisco (as in Heckathorn et al. 1997), the enzyme that catalyzes the photosynthetic fixation of CO₂ during the first step of the Calvin cycle (dark reactions).  Rubisco is often the most abundant protein in photosynthetic cells, and its content decreases at low light.

Accomplishments:

Field work for this project commensed in May 2007 and sampling for the first summer season was completed in October.  The table below lists sampling dates, samples collected and progress of sample analysis for 2007.  In 2008, field work resumed in May and was completed in September.  Laboratory experiments were run during the month of September.  We expect that analyses will be completed by early 2009 and that manuscripts and presentations will be ready by spring 2009.