Overview of a Novel Green Technology:

Biological Control of Zebra and Quagga Mussels

with Pseudomonas fluorescens
(Version 6: Updated August 24, 2007)

Daniel P. Molloy and Denise A. Mayer

New York State Museum
Improvements Over Existing Chemical Control Methods
Power generation facilities require annual maintenance and preventive programs to keep infestations
of fouling zebra and quagga mussels (Dreissena spp.) in their cooling water intake systems under control.
Currently it is necessary at many of these and other infested raw-water dependent infrastructures to
administer controlled dosages of chlorine or other types of biocides for this purpose. Although such
applications meet all existing water pollutant
discharge regulatory limits, evidence exists to
suggest that natural resource interest groups
and regulatory agencies are reexamining the
negative long-term use of biocides for this
purpose. Both groups have made it clear that
safe, non-chemical alternatives for controlling
mussel fouling would be environmentally
beneficial. Chlorination, for example, is a
common control method, and when chlorine
combines with organic compounds in water,
potentially carcinogenic substances such as
trihalomethanes and dioxins are formed
(United States Environmental Protection
Agency 1999; Thornton 2000). Should future
regulatory actions result in the loss of
chemical biocides, without an alternative
control
option,
electric
generation
Coal-fired power plant on Lake Ontario.
organizations
and
many
other
industries
that
(Photo credit: Rochester Gas & Electric)
rely on the withdrawal of surface waters for
operational reasons are certain to experience economic penalties. These losses would be the result of
decreased production brought on by increased facility maintenance and downtime. Thus, the availability
of an equally effective, yet far more environmentally benign, mussel control method to replace chlorine
and other biocides is critically needed by power plants and other infested facilities.

Research Paradigm
Why would one look to use a naturally-occurring, non-parasitic, non-infectious microbe, such as the
ubiquitous soil-water bacterial species, Pseudomonas fluorescens, to serve as an innovative, novel
strategy for mussel management in power generation facilities? Sounds illogical? Well, it is widely
accepted that the screening of diverse biochemicals found in tropical plant species is a worthwhile activity
due to the discovery of drugs that can prevent or cure animal diseases, particularly cancers. Production
of these biochemicals, however, did not evolve in these plants for this purpose, and the effect of these
plant substances on animal diseases, although fortuitous, is purely coincidental. Using the same logic,
we can also look to microorganisms for unique biochemicals or toxins which have potential as highly
selective biopesticides. In fact, the use of microbial toxins already has a clear record of commercial
success and environmental safety in the control of invertebrate pests in North America, as well as globally

Bacterial Project Overview - Page 2

(Rodgers 1993), and our New York State Museum (NYSM) laboratory has been involved in such research
for over two decades as discussed in the following section.

Prior Participation in Commercial Success

In the interest of eliminating polluting pesticides
and thereby protecting biodiversity in New York State,
our NYSM Field Research Laboratory assisted in the
commercial development of Bti the selectively toxic
bacterium, Bacillus thuringiensis subsp. israelensis
as the first biological control agent for black flies
(Simuliidae).
This bacterium, because of its
extraordinary nontarget safety (Molloy 1982, 1990,
1992; Molloy and Jamnback 1981; Molloy and Struble
1989), has now completely replaced broad-spectrum,
chemical pesticides throughout New York State and
elsewhere in North America for the control of these
biting flies. The commercial use of this microbial
agent is not small scale; large waterbodies, such as
the Susquehanna River in Pennsylvania and the New
River in West Virginia, are routinely treated with this
bacterial species to control larval black fly
populations.

Treatment of the Susquehanna River with Bti bacteria

for black fly larval control.
(Photo credit: Pennsylvania Department of
Environmental Resources)

Research Progress To Date

1. Inception of Project: Research Funded by Private Electric Power Industry
1

The Empire State Electric Energy Research Corporation (ESEERCO ) faced with the threat of
zebra mussels fouling electric power facilities within New York State contracted with our NYSM Field
Research Laboratory in 1991 for the screening of bacteria as potential biological control agents. Based
on the successful development of the environmentally safe, biological control agent for aquatic black fly
larvae (see above), it was hypothesized (Molloy 1991) that
bacteria also existed in nature whose toxins could be used as
lethal agents for these new aquatic pests, zebra and quagga
mussels. The research efforts funded by ESEERCO proved
this hypothesis to be true (Molloy 1998). Extensive laboratory
screening trials of more than 700 bacterial strains identified a
North American isolate, strain CL145A of Pseudomonas
fluorescens, to be lethal to these mussels. Of all 10 strains of
P. fluorescens that have been laboratory tested to date, only PfCL145A has been found to be highly lethal, i.e., at dosages that
produce >90% zebra mussel kill with Pf-CL145A, the other 9
Individual cells of P. fluorescens.
strains of P. fluorescens caused only 0-11% mortality.
Pseudomonas fluorescens is worldwide in distribution and is present in all North American
waterbodies. In nature it is a harmless bacterial species that is found protecting the roots of plants from
rot and mildew. It is so ubiquitous that it is a common food spoilage organism in the average household
refrigerator. Our research, however, has shown that the Pf-CL145A strain of this species may be
fortuitously used for another purpose the control of Dreissena spp. (Molloy 2002). A patent for this
purpose has been issued in both the United States (Molloy 2001) and Canada (Molloy 2004).

A research consortium of New York States electric power generation companies.

Bacterial Project Overview - Page 3

2. Mussels Die from a Natural Toxin: Dead Bacteria Kill Equally As Well
Although phytoplankton is their preferred food, Dreissena mussels can filter out and consume
bacteria as a food source (Mikheev and Sorokin 1966; Frischer et al. 2000). When a zebra or quagga
mussel ingests artificially high densities of strain CL145A, however, a toxin within these bacterial cells
destroys the mussel's digestive system. Dead cells are equally as lethal as live cells, providing clear
evidence that the mussels die from a toxin, not from infection. Techniques have already been developed
at our laboratory that kill the bacteria without any reduction in their lethality to the mussels. Future
commercial products based on this microbe will contain dead cells, thus further reducing environmental
concerns.

In healthy mussels, epithelial cells (arrows) appear as a

thick layer lining the tubules of the digestive gland.

Following bacterial treatment, epithelial cells are

destroyed. Blood cells are abundant as the digestive
gland hemorrhages.

3. Mussel Feeding: Bacteria Are Readily Ingested

Although ingestion of CL145A cells is clearly a suicidal behavior for Dreissena mussels, they appear
to have no adverse reaction to feeding on the cells and filter normally throughout a typical 6-hr, oncethrough pipe treatment. In contrast, biocides, like chlorine, that are currently being used for mussel
control cause them to quickly shut their valves since the mussels apparently sense an adverse effect.
This necessitates more prolonged chlorination periods, such as continuous treatments of three weeks or
more. The apparent acceptance of CL145A cells as "normal" bacterial food by these mussels facilitates
the use of this microbe as a biocontrol agent.

4. Mussel Length: All Mussel Sizes Can Be Killed

All Dreissena mussel sizes tested to date (length, ca. 1-25 mm) appear to be equally susceptible to
kill by CL145A. Thus, the bacteria are capable of killing, irrespective of mussel size. Susceptibility of the
planktonic mussel stage has not yet been tested.

All mussel sizes show the same susceptibility.

Bacterial Project Overview - Page 4

5. Mussel Species: Both Species Can Be Killed

The bacterium can kill zebra mussels (Dreissena
polymorpha) and quagga mussels (Dreissena rostriformis
bugensis) the 2 species that invaded North America in the
1980s. Tests to date consistently achieve higher kill against
zebra mussels.

6. Water Hardness: Mussel Kill is Highest in Hard Water the Preferred Dreissena Mussel Habitat
Tests to date suggest that bacterial treatments may have reduced efficacy in soft waters with pH
values less than ca. 7.4. Dreissena mussels, however, rarely reach high population densities in such
(near neutral or acidic) waters, and thus, infested pipes in power plants typically will have more alkaline
waters where bacterial efficacy will not be impaired.

8. Water Temperature: Higher Kill at Warmer Temperatures

Susceptibility increases with water temperature, with >90%
mortality consistently achieved in routine lab testing at 23C
against zebra mussels. High mortality is still achievable even in
very cold waters, e.g., near 80% kill at 7C, indicating that the
bacteria are actually more effective at lower temperatures than
currently commercialized chemical molluscicides used for
Dreissena control. The latter commercial biocides, e.g., chlorine,
can not achieve such high mussel kill below about 10C, thus
limiting their application to warm water periods. Mussels die more
quickly at higher water temperatures, with all kill typically
achieved within two weeks at 23C and two months at 7C.

Mean mussel mortality (%)

7. Dissolved Oxygen: Keep Oxygen Levels High to Ensure Highest Kill

Laboratory tests indicate that very low oxygen levels (<2 ppm) can sometimes result in a 20% decline
(e.g., 75% vs. 95%) in mussel kill. This is possibly due to lower feeding by the mussels on suspended
bacteria under such low oxygen conditions. Thus, wherever possible, bacterial treatments should occur
in waters of high dissolved oxygen the preferred environment of Dreissena mussels.

100
90
80
70
60
50
40
30
20
10
0

High mussel mortality

achieved from 7 to 23C

7C
12C
17C
23C
Treatment and holding temperature

Mortality increases with temperature.

9. Suspended Particles: Avoid Treating in Periods of Very High Particle Loads To Ensure Highest Kill
Tests to date indicate that unusually high levels of naturally-occurring particles in water (e.g.,
suspended mud at greater than 100 ppm) can result in a 20% decline (e.g., 75% vs. 95%) in mussel kill.
This is possibly due to competitive displacement, i.e., lower feeding by the mussels on the suspended
bacteria vs. mud particles. Thus, wherever possible, bacterial treatments should not occur in waters of
very high particle loads.

10. Mussel Siphoning Behavior: Do Not Disturb Normal Mussel Feeding

In nature, a dreissenid mussel typically has its two shells
spread apart and extends an inhalant siphon tube from between
its shells to take food particles into its mantle cavity. After
passing through the digestive system, fecal material is egested
through the exhalant siphon. Testing has generally indicated
that the more active this siphoning behavior is, the higher the
mortality that will be achieved by a bacterial treatment. Thus,
any stress factors (e.g., vibrations, shadows) that cause the
mussels to close their shells during treatment will likely reduce
mortality.

Bacterial Project Overview - Page 5

11. Treatment Concentration and Duration: Treat for about 6 hr for Maximum Kill
Laboratory and facility trials indicate that ca. 6-hr treatments of ca. 50-100 ppm (dry bacterial mass
per unit volume) consistently obtain the highest mussel mortality. Exposing mussels to longer treatment
durations or higher bacterial concentrations is (particularly >12 hr) achieves very limited further benefit.

12. Nontarget Trials: Outstanding Specificity

Laboratory trials to date have been very encouraging regarding nontarget safety. At dosages which
produced high zebra mussel mortality (76100%), no bacteria-induced mortality was recorded among any
of the nontargets, including fish, ciliates, daphnids, and bivalves:

Fish: No biotoxin-induced mortality has been observed in the three fish species thus far tested:
fathead minnows (Pimephales promelas), young-of-the-year brown trout (Salmo trutta), and juvenile
bluegill sunfish (Lepomis macrochirus). Laboratory and facility trials have indicated that fish can not
tolerate exposure to high levels of live bacteria. Fish trials conducted with dead bacteria, however,
have indicated that applications of killed cells were harmless to the fish, but yet were still highly lethal
to the Dreissena mussels. To protect fish, as well as to reduce overall environmental concerns,
future commercial products based on this microbe will contain almost exclusively dead cells.

Ciliates: Trials with the common freshwater ciliate Colpidium colpoda indicated that the bacteria were
not only nonlethal, but served as a food source permitting higher rates of ciliate reproduction than
ciliates held in untreated streamwater.

Daphnids: The microcrustacean Daphnia magna is an aquatic filter feeder that ingests small
suspended particles including bacteria, making it an appropriate organism for non-target tests.
Laboratory assays indicate that the bacteria are not lethal to this species.

Bivalves: Bacterial exposures caused no mortality to blue mussels (Mytilus edulis) or any of 6 native
North American unionid clam species (Pyganodon grandis, Lasmigona compressa, Strophirus
undalatus, Lampsilis radiata, Pyganodon cataracta, and Elliptio complanata).
Brown trout

Daphnids

Fathead minnows

Ciliates

Blue mussels

Sunfish

Unionids

There has been no non-target mortality from the bacterial toxin yet observed.

Bacterial Project Overview - Page 6

13. Trials at Power Plants: High Kill Can Be Achieved in Service Water
Trials routinely achieving high zebra mussel mortality (ca. 7097%) in pipes have been conducted at
the New York Power Authority (NYPA) electric power station on the Mohawk River (Crescent, New York).
Trials at the Rochester Gas & Electrics Russell Power Station on Lake Ontario (Rochester, New York)
have been primarily against quagga mussels (the less susceptible of the 2 North American Dreissena
spp.) and routinely achieved 50-70% mortality.

Very high zebra mussel kill (>95%) was consistently

achieved in 6-hr treatments at 100 ppm inside a
NYPA hydropower plant under flow-through
conditions (3 replicate pipes 17 m in length were
used in this trial). Experiments to date indicate that
there should be no limit on the length of pipe that
can be successfully treated.

Pouring suspension of bacterial cells in preparation for

pipe treatments within power plants. Advances in
fermentation have allowed increasingly larger volumes of
bacteria to be produced, thus allowing larger volumes of
water to be treated in pipes.

14. Identity of the Natural Product that is Lethal to Dreissena Mussels

Research was undertaken to characterize, isolate, and identify the specific Dreissena-killing natural
product that is associated with P. fluorescens strain CL145A cells. Treatment of toxic cells with lysozyme
or deoxycholate appeared to separate the toxin molecules from the bacterial cells, suggesting that the
toxin was associated with the outer membrane of the cells. Protease treatments also decreased toxicity,
suggesting that the membrane-associated toxin was likely a protein. Cells that were mildly heated lost
their ability to kill mussels, providing evidence that the toxin was heat-labile and protein in nature.
A biochemical approach was undertaken in an attempt to get the toxin to adhere to particles which
could hopefully be used to identify the toxin. Even though the toxin was able to be separated from the
cells by chemical treatment (i.e., make the cells nontoxic), efforts to develop an effective method to
deliver the solubilized toxin molecules to the mussels on particles that they would ingest were
unsuccessful. As a result, we altered their biochemical experimental approach and decided to search the
literature for documented products from P. fluorescens that matched characteristics of the mussel-killing
toxin. A candidate molecule investigated was glycine dehydrogenase, an enzyme that catalyzes the
conversion of the amino acid glycine to hydrogen cyanide (HCN). Analysis of strain CL145A confirmed
that it did produce trace amounts of HCN. Testing then focused on determining whether HCN was the
toxin that was responsible for causing mussel death. These experiments demonstrated that treating
CL145A cells with an irreversible flavoenzyme inhibitor, diphenyleneiodonium chloride (DPI), successfully
blocked the ability of strain CL145A to produce HCN. Even though DPI-treated cells no longer produced
trace amounts of HCN, the cells still remained equally lethal to the mussels, demonstrating that HCN was

Bacterial Project Overview - Page 7

not the toxin that caused mussel death. Efforts are now underway at the NYSM to use genetic
approaches to determine the identity of the toxin (see next section).

15. Culturing Optimization and Fermentation Scale-up:

Toxic P. fluoresens-CL145A cells are harvested during late linear to stationary growth phase under
defined culture conditions. Toxic cells are routinely harvested from shake-flasks and 0.5-L fermentors at
the bench-scale and fermentation has been scaled to 100-L. Media components and fermentations
parameters have been optimized to result in a proprietary medium and protocol that consistently produce
toxic cells. The optimized medium has been further modified to define a commercial medium formulation
at a fraction of the cost.

Culturing has been optimized and scaledup successfully from shake-flasks to 100-L
fermentors.

16. Current and Recent Project Activities

Research on this NYSM project is currently supported by the U.S. Department of Energy National
Energy Technology Laboratory (DOE-NETL 2007). With this funding, the complete genome of strain
CL145A has been sequenced, and research is currently underway to identify the gene(s) that produce(s)
the mussel killing biotoxin. Successful identification of the gene(s) could lead to identification of the
biotoxin molecule and ultimately to the production of cells of higher toxicity.
Recently a grant from the National Science Foundation (NSF 2007) supported experimentation which
developed a protocol to produce a dead-cell, powdered commercial formulation that would have good
shelf life due to biotoxin stability.
The lack of non-target impact when treating with dead cells may allow this novel green technology to
also be used for Dreissena mussel control in open waters, such as lakes and rivers. Thus far, however,
the research focus of the project has been controlling these mussels within power plant pipes. Recent
preliminary trials against zebra and quagga veligers have suggested that this planktonic larval stage may
actually be even more susceptible to the bacterial treatment that attached mussel stages. This suggests
that bacterial treatments in open waters could significantly reduce veliger densities, thereby preventing
high mussel density buildup and slowing the spread of mussel populations.

17. Commercialization Partnership

Following a nation-wide search, the NYSM has chosen to partner with Marrone Organic Innovations
(http://www.marroneorganicinnovations.com/), a biopesticide company whose staff have unparalleled
experience in the discovery, development, and marketing of natural products for pest management.
These joint research efforts will be directed toward increasing bacterial cell toxicity via additional
fermentation work and further understanding the chemistry of the toxic moiety so that cells can achieve
even higher mussel kill and thus be more competitive with current polluting chemical control methods.
Other critical steps toward commercialization include analytical method development, identification of the
mussel-killing toxin, formulation improvement, and additional nontarget toxicology studies that will be
mandated by the USEPA for product registration.

Bacterial Project Overview - Page 8

References
DOE-NETL (Department of Energy National Energy Technology Laboratory). 2007. Power Systems
Advanced Research Environmentally-Safe Control of Zebra Mussel Fouling Fact Sheet.
Accessed March 14, 2007 @ http://www.netl.doe.gov/publications/factsheets/project/Proj291.pdf
Frischer, M., Nierzwicki-Bauer, S., Parsons, R., Vathanodorn, K., and Waitkus, K. 2000. Interactions
between zebra mussels (Dreissena polymorpha) and microbial communities. Canadian Journal
of Fisheries and Aquatic Sciences 57:591-599.
Mikheev, V. P. and Sorokin, Y. L. 1966. Quantitative studies of Dreissena polymorpha habits using the
radiocarbon method. Zhurnal Obshchei Biologii 27:463-472. (In Russian.)
Molloy, D. (ed.) 1982. Biological Control of Black Flies (Diptera: Simuliidae) with Bacillus thuringiensis var.
israelensis (Serotype 14): A Review with Recommendations for Laboratory and Field Protocol.
Miscellaneous Publications of the Entomological Society of America 12(4):30 pp.
Molloy, D. P. 1990. Progress in the biological control of black flies with Bacillus thuringiensis var.
israelensis, with emphasis on temperate climates. Pages 161-186 in Bacterial Control of
Mosquitoes and Black Flies: Biochemistry, Genetics & Applications of Bacillus thuringiensis
israelensis and Bacillus sphaericus (H. de Barjac and D. Sutherland, eds.) Rutgers University
Press, New Brunswick, New Jersey.
Molloy, D. P. 1991. Biological control of zebra mussels: Use of parasites and toxic microorganisms.
Journal of Shellfish Research 10:260.
Molloy, D. P. 1992. Impact of the black fly (Diptera: Simuliidae) control agent Bacillus thuringiensis var.
israelensis on chironomids (Diptera: Chironomidae) and other nontarget insects: Results of ten
field trials. Journal of the American Mosquito Control Association 8:24-31.
Molloy, D. P. 1998. The potential for using biological control technologies in the management of
Dreissena spp. Journal of Shellfish Research 17:177-183.
Molloy, D. P. 2001. A Method for Controlling Dreissena Species. United States Patent and Trademark
Office, U. S. Department of Commerce. Patent No. 6,194,194. (Filed December 17, 1997 &
issued February 27, 2001.) 4 pp.
Molloy, D. P. 2002. Biological control of zebra mussels. Pages 86-94 in Proceedings of the Third
California Conference on Biological Control. University of California, Davis.
Molloy, D. P. 2004. A Method for Controlling Dreissena Species. Canadian Intellectual Property Office,
Industry Canada. Patent No. 2,225,436. (Filed December 27, 1997 & issued December 21,
2004.) 11 pp.
Molloy, D. and Jamnback, H. 1981. Field evaluation of Bacillus thuringiensis var. israelensis as a black
fly (Diptera: Simuliidae) biocontrol agent and its effect on nontarget stream insects. Journal of
Economic Entomology 74:314-318.
Molloy, D. P. and Struble, R. H. 1989. Investigation of the microbial control of black flies (Diptera:
Simuliidae) with Bacillus thuringiensis var. israelensis in the Adirondack Mountains of New York.
Bulletin of the Society of Vector Ecology 14:266-276.
NSF (National Science Foundation) 2007. STTR Phase I: Stabilization of a Protein Toxin - An Essential
Step in the Commercialization of an Innovative Method for Controlling Zebra Mussels. Accessed
March 14, 2007 @ http://www.nsf.gov/awardsearch/showAward.do?AwardNumber=0610252
Rodgers, P. B. 1993. Potential of biopesticides in agriculture. Pesticide Science 39:117-129.
Thornton, J. 2000. Pandoras Poison: Chlorine, Health, and a New Environmental Strategy. MIT Press,
Cambridge, Massachusetts. 599 pp.
United States Environmental Protection Agency. 1999. Wastewater technology fact sheet: Chlorine
disinfection. U. S. Environmental Protection Agency, Washington, DC. EPA/832-F99-062. 7 pp.

Bacterial Project Overview - Page 9

Acknowledgments
Funding for this zebra mussel biocontrol project from the following agencies is gratefully
acknowledged:
Private
ESEERCO New York State Power Generation Utilities
State
New York Sea Grant
New York State Department of Environmental Conservation
New York State Energy Research and Development Authority
Federal
National Science Foundation
U.S. Army Corps of Engineers
U.S. Department of Energy National Energy Technology Laboratory
U.S. Fish and Wildlife Service