Watershed Protection


Using Mushroom Mycelium to Filter Bacteria from EBMUD Watershed – Bay Area Applied Mycology

East Bay Municipal Utility District (EBMUD) in collaboration with Bay Area Applied Mycology.



Scott Hill, Manager of Watershed and Recreation

Jonathan Price, Fisheries and Wildlife Aide

Virginia Northrop, Senior Ranger



Bay Area Applied Mycology (BAAM) , a group that utilizes fungus and other natural organisms to remediate the environment, has formed a partnership with East Bay Municipal Utility District (EBMUD) in Orinda, California to develop a filtration system made from mushroom spawn (mycelium) to filter harmful bacteria from EBMUD watersheds. EBMUD provides drinking water for the greater East Bay of California, and EBMUD intends to prevent harmful bacteria from contaminating their water source.

EBMUD allows cattle to graze their lands for a short period of time. During rainstorms, bacteria and protozoa present in cattle dung leach into creek beds that eventually lead to water reservoirs. These reservoirs supply drinking water to the greater East Bay of California. EBMUD currently utilizes Chloramine to eliminate harmful bacteria and protozoa. EBMUD intends to prevent water contamination by employing filtration systems.

This report details BAAM’s method for developing a system using mushroom mycelium to satisfy EBMUD’s request to prevent their water from becoming contaminated with harmful bacteria and protozoa.



The purpose of this project was to develop a low-cost method to filter harmful bacteria from water by utilizing fungus. We developed a filtration system utilizing large amounts of mycelium that was intended to capture and eliminate bacteria that would otherwise be harmful to humans if ingested.

Due to the success of Paul Stamets, who has shown success in filtering and eliminating Escherichia coli using mushroom mycelium as a filter, BAAM and EBMUD intend to find similar success. “The filter” is an aggregate of many burlap sacks filled with a mixture of straw and wood chips colonized by mushroom mycelium.

This project has established strategies for developing these filters and applying them to creek beds which may experience turbulent flows of water.

EBMUD currently removes cattle from their lands before rain falls, however cow dung remains standing when rains occur. By developing a system to filter bacteria. EBMUD intends to prevent any harmful bacteria, such as Escherichia coli, and protozoa such as Giardia lamblia and species of Cryptosporidium, from reaching the reservoir.

The reservoir is currently treated with Chloramine. Chloramine is considered more stable than chlorine, does not dissipate as rapidly as chlorine (refer to “Dealing with chlorine and chloramine”), and imparts less flavor. EBMUD extracts and tests water samples from several towers located in the reservoir.

BAAM fostered the development of the mycelium used in the filter. EBMUD has supplied counseling for the instruments and efforts needed to test for bacteria levels within the water.

The mycelium of several species of mushrooms have shown to eliminate E. coli in water by capturing and eliminating bacterias (Stamets, 62). Mycelium, the larger body of the mushroom organism, exudes metabolites into its surrounding environment which capture and consume bacterias, as well as breaks down and degrades organic material into simpler structures.

Mycelium is a ropy web of filaments that lives inside it’s food. Because of its microscopic nature, it battles with other microscopic organisms for territory. Mycelium expands to colonize and feed from available organic material (known as substrate). By processing straw and introducing mycelium to that material, we will “ramp up” the amount of mycelium that can be bagged and used as part of this filtration system.



Several considerations have risen while developing this project. These topics will be further elaborated in this report:

  • which species of mycelium can best filter bacteria from water when applied strategically
  • how to develop a repeatable method to prove the effectiveness of this system
  • how to construct a filter that will withstand the forces of a heavy water current
  • the optimal location of our filter
  • the optimal way to structure the filter so that all water flowing into its path will collect and pass through the mycelium
  • how to generate massive amounts of spawn effectively
  • the cost of this system
  • cost reduction for future applications
  • what are other methods of filtration and how does ours compare with labor, cost, and effectiveness


Choosing a Fitting Mushroom Species for Filtration

Each mushroom species excels at deconstructing and consuming specific material. For this project, based on observation and information previously published (refer to “Mycoremediation Species and Spawn” and Stamets pgs 51, 60, and 192), we decided to use Stropharia rugoso-annulata. S. rugoso-annulata can be purchased from commercial spawn growers, and has already shown to grow well on straw in home tests. Because it’s grown commercially, we were able to obtain blocks of it easily to implement into our project (at a cost).

Although BAAM prefers to use a local mushroom species found on EBMUD property, we selected this species of mushroom because of its reputed ability to thrive in bacteria-rich environments and its ability to colonize large volumes of material quickly (refer to “Mycoremediation in Action!”).

However, while surveying possible sites to apply the filter, we observed at least three species of mushrooms growing directly from cow dung. Because these mushrooms have already shown preference for bacteria-rich environments, and because they grow locally, we believe future tests utilizing this fungus may yield exceptional results. The speed at which the mycelium of these species grows and their ability to colonize large volumes of material quickly must be observed before proper implementation.

Strains of Pleurotus pulmonarius and Pleurotus ostreatus were also considered for their ability to quickly colonize available organic material and in future testing we hope to incorporate them.


Developing a Filter that Will Withstand the Forces of Heavy Currents

One of the observations made concerning this project was that during rare events of extremely heavy rainfall the water in the creek beds flows at turbulent speeds. BAAM installed a system that is intended to withstand the forces of this heavy water current.

We considered using a combination of rebar and chicken wire to prevent the burlap sacks from moving, or to use rebar and stakes to hold the burlap sacks in place. We settled on the latter.

Placing the filter upstream of any cow pies could likely result in a failure to eliminate harmful bacterias from reaching the EBMUD reservoir. We placed our filter in an area downstream of cattle grazing.


Generating Mass Amounts of Spawn Effectively

Intending to eliminate the use of nonrenewable resources, BAAM sought to expand the amount of available mycelium by preparing substrate using the cold fermentation method. As opposed to pasteurizing straw, which uses large quantities of fuel to heat water, cold fermentation is a method of preparing straw for inoculation without any heat source.

To process straw using cold fermentation, straw was submerged in water and left to soak for eleven days. Although temperature may play a role in the fermentation process, a week is typically sufficient. This method destroys aerobic bacteria and organisms living dormant on the straw. The presence of a sour smell was a good indicator that the process was succeeding. After eleven days, the straw was removed and allowed to drain.


Jonathan and Virginia of EBMUD fill a trough with straw. Mino and Mark look on.

The trough was filled with water and the straw allowed to soak for a week. November 8, 2012. Photo by Joe Soeller.

The straw was organized into small piles and mushroom spawn was scattered equally into the fermented straw at a rate of 1:10 to 1:20, spawn to straw, by volume. The straw was then consolidated into one pile.

Small piles of straw were first made and inoculated with equal amounts of spawn, then all

piles were consolidated and covered with a tarp. November 19, 2012. Photo by Joe Soeller.

The water that was used to soak the straw was emptied onto the blended substrate as a final dampening. The straw pile was then thinned to no more than a foot high to allow the mycelium opportunity to breathe. The edges were packed together to lend more structure to the colonizing mycelium. A tarp was used to cover the material and checked weekly to monitor the spawn’s effectiveness at colonizing the substrate.

The straw was checked about 3 weeks after being inoculated. Ropy strands

of mycelium dominate the fermented straw. December 10, 2012. Photo by Joe Soeller.

The spawn effectively colonized the substrate when the straw held tenaciously together by ropy white strands of mycelium. The external layer of straw did not exhibit white mycelium because that layer dried out and was exposed to weathering.
After the straw was colonized by Stropharia rugoso-annulata mycelium, it was blended at a 1:1 ratio with fresh woodchips. We then filled burlap sacks with this straw-wood chip blend. The burlap sacks were sewn shut to prevent the substrate blend from spilling out of the bags.

The advantage of this method of first applying sterile spawn to fermented straw then to wood chips and “ramping up” the colony allowed us to cheaply and effectively produce twenty to forty times the amount of spawn originally introduced to the substrate, all without using fuel.

The disadvantage is that this system for “ramping up spawn” is slower than pasteurizing straw and applying spawn immediately. However, with foresight and planning, this process was handled with little effort.

This method saves the cost of purchasing fuel such as propane, and doesn’t deal with potential hazards of heat and flame. Since the spawn was expanded prior to filling burlap sacks, this method also saved us the cost of purchasing enough spawn to inoculate both the straw and the wood chips that filled the burlap sacks.

Volunteers mix wood chips with myceliated straw into burlap sacks. Photo by Sean Parnell

Method Costs

The cost of this method depended on how spawn was produced and with what materials the spawn was grown on. With our current system, we decided to purchase eleven 5.5 lb bags of commercial spawn of Stropharia rugoso-annulata from Field and Forest Products located in Wisconsin at $15.75 a piece (price-break at different quantities). If we developed the spawn in a private cultivation laboratory, the cost could be significantly reduced, considering the expense of laboratory equipment. However, labor times would increase.

A bale of straw (60-90 lbs) typically costs between five and ten dollars. We used five bales of straw.

Although farmed wheat straw was used as a substrate, it isn’t necessary. Straw is inexpensive and easy to acquire, however. If using local plant life is an important consideration, local grasses and plant material can be mowed or hacked down and left to dry before processing them in the cold fermentation method. Again, doing so would raise labor times.

Box elder and willow wood chips were provided for the project by EBMUD.

Burlap sacks, originally used to transport green coffee beans, were provided free-of-charge by Peet’s Coffee.

As part of the study plan (see below), the cost of developing an isolated system to test levels of bacteria and protozoa using gravity fed water in a tube and constructing a filter will be considered when definite plans are developed.


For spawn generation:

Spawn, straw, water, troughs/containers, tarps, wood chips (box elder, willow), pitchforks

Constructing the grids of burlap sacks:

Burlap sacks, wood chips, myceliated substrate (straw), rebar, stakes, wheelbarrows, mallets

Developing a Repeatable Method to Prove the Effectiveness of the Mycelium Filter, a Study Plan

In an effort to illustrate the effectiveness of this system, to eliminate as many variables as possible, and as a way for others to repeat the experiment to prove its worth, EBMUD suggested BAAM develop a system that was isolated from the natural environment.

The purpose of this system is to show that bacteria-contaminated water, when gravity fed through a mycelium filter, will exit the system into a catch showing either no or less bacteria than was previously present.

Cost is an important factor in developing this method.

Several tests of the creek water can be held:

the raw creek water itself, to demonstrate the presence of bacteria and protozoa
the same test of the the water post-filtration
Water that is poured onto the colonized substrate may take a long while before the water reaches a catch. A trickle-fed system may be the best method of introducing water to the filter, and most closely emulates a natural environment. A trickle-fed system will allow water to saturate the mycelium and continue flowing through the filter without overflowing.

Given the freedom of time and testing, several variables can be eliminated from this test:

How thick must the colonized substrate be to filter bacteria from water?
What strain of mycelium most efficiently filters bacteria from water?
What substrate works best to support the filter? For instance, straw, wood chips, or a blend?
And to combine the two variables: what strain of mycelium growing on which substrate works best to filter water? (This consideration could be scenario-based. There might be no silver bullet.)
What is more optimal: A) a loosely colonized filter of larger volume, or B) a tightly colonized filter of smaller volume. (The rate of flow of water is important to consider when handling high-flow creeks. Think of a tightly knit barricade versus a loosely woven course).
At what rate does water flow through the system (once the colonized substrate has been saturated by water)? This information is important to consider when building large-scale filter.
Unfortunately, we’ve run into some roadblocks. Although we currently don’t have the funds to run sufficient testing, we are raising funds for such tests. If you are interested in donating or supporting this effort, please contact project director Mino de Angelis, deanglismino@gmail.com, to support the matter.


“Dealing with chlorine and chloramine.” The Skeptical Aquarist. Published March 21, 2011. Accessed

February 20, 2013. http://www.skepticalaquarist.com/chlorine-chloramine

“Mycoremediation in Action!” Perma Dub Dream. Published November 12, 2010. Accessed November 18,


“Mycoremediation Species and Spawn.” Mushroom Mountain. Accessed February 20, 2013.


Srivastava, P. “Alleviating Water Quality Impacts of Animal Waste Through Mycoremediation and

Mycofiltration.” Auburn University. Accessed November 18, 2012


Stamets, Paul. Mycelium Running. Berkeley: Ten Speed Press, 2005.


Tuesday, January 15, 2013

With a large group of volunteers, BAAM was able to install the bio-filtration patch in a swale on the EBMUD property. Myceliated straw was mixed with uninoculated wood chips and stuffed into burlap sacks. The sacks were then sewn shut with yarn to prevent the spawn and wood chips from spilling out.

Next, the burlap sacks were wheelbarrowed to the swale and formed into two grids, the upstream grid being on the cattle grazing side, the downstream grid being on the other side of the fence where cattle have not grazed.

72 total bags were installed between two grids. With 16 volunteers, the installation took about 3 to 4 hours to complete.

We used pitchforks for mixing, shovels for filling the burlap bags, and large needle pullers to tie the burlap bags. We used sledge hammers for driving the stakes.

Once at the site, we slightly overlapped the burlap sacks over each other and stakes were driven through them at various locations. We used 24″ wooden stakes and 3/8″ rebar U’s to pin them to the creek bed. We used the U’s at both section’s front and back. They straddled two bags at once.

Period photographs will be taken to observe the growth of the mycelium, the decomposition of materials, and how flowing water affects the installation.

Photo May 17 A happy Sean Parnell with a single cluster of Stropharia

We are very pleased by the viability of the spawn, we think that this method placed in the appropriate location can prove very effective in mitigating bacterial effluent. Although for our test we chose a seasonal stream channel the watershed area is just too great to mitigate it all. A future test of interest is to site it around cattle holding corrals where slow moving but constant discharge can slowly pass through the filtration bags and become decontaminated.

Want to learn more? Contact Sean or Mino at email@email.email.


Interested in supporting projects like this?