BI 341 Final Project

E.coli and Seagulls

Lindsay Wielichowski and Kathleen Manke

Abstract

Recently, South Shore Beach and other area Lake Michigan beaches have closed their waters to swimmers due to high Escherichia coli 0157:H7 (E.coli) levels (Schultze 2001). Many theories have been created to explain the poor water quality; the latest is the relationship between the high presence of seagulls and the high levels of E.coli (Schultze 2001). To determine if the air temperature had anything to do with the poor water quality, we hypothesized that E.coli levels will be lower, regardless how many seagulls are present, as the temperature decrease. Findings did not support our hypothesis; E. coli levels were high in an area where no seagulls were present. Keywords: Seagulls, E. coli, Temperature, E. coli Coliform Density

Introduction

When beaches close due to poor water quality, many people would blame the deep tunnel project and Jones Island (Milwaukee’s sewage treatment facility) for releasing raw sewage (Schultze 2001). However, this summer a group of student researchers working with the University of Wisconsin Milwaukee (UWM) Great Lakes Water Research Institution have found evidence that links the high E. coli levels at South Shore Beach to seagulls (Larus argentatus) which are a main carrier for E. coli. (Schultze 2001). The high presence of seagulls and their droppings may be the reason for the dozen of beach closings this summer, and past summers. On July 22 1999, South Shore Beach had an E. coli density of 15,400 counts per 100ml. The federal standards for E. coli are no greater then 235 counts per 100 ml (Schultze & Rohde 2000).

High temperatures, of the air and water often contribute to rapid growth of bacteria (Schultze & Rohde 2000).. Seagulls tend to be more abundant in warmer weather. Is it possible that the high number of seagulls and the warm temperatures play a role in the poor water quality of Lake Michigan’s beaches? To find out, we have hypothesized that the numbers of seagulls and the high temperature of the air contribute to high levels of E. coli in the water.

Method and Materials

Eight water samples were taken from South Shore beach -- four from the Marina and four from the rock formation just east from the playground -- on October 28, November 1, November 5, November 9, 2001.

Each sample was taken no sooner than twenty-four hours after a rainstorm, to prevent contamination from run-off. After each sample was taken it was cultured within forty-eight hours using the following procedure obtained from the Seventh Edition of Microbiological Applications by Harold Benson. This method, utilizing the membrane filter, has been recognized by the United States Public Health Service as a “reliable method for the detection of coliforms in water” (Benson 1994). Bacteria larger than 0.47 micrometers cannot pass through the 150-micrometer thick filter disk (Benson 1994). When the sample is run through the filtration system, all of the bacteria present remain on the filter disk (Benson 1994). An absorbent pad is soaked with the nutrient medium, m Endo MF broth; the filter disk is placed on the pad and is incubated for twenty-two to twenty-four hours at thirty-five degrees Celsius (Benson 1994). Coliforms, displaying a “golden metallic sheen” are then counted (Benson 1994).

“The advantages of this method over the multiple tube test are (1) higher degree of reproducibility of results; (2) greater sensitivity since larger volumes of water can be used; and (3) shorter time (one-fourth) for getting results” (Benson 1994).

The following is a modified procedure, “Bacteriological Examination of Water: The Membrane Filter Method” from Benson, page 178:

Materials:

air compressor

membrane filter assemblies (sterile)-Bucker Funnel

side-arm flask, 1000 ml size

sterile graduates (250 ml size)

sterile, plastic Petri dishes

sterile membrane filter disks

sterile filter paper

distilled water

2 ml pipettes

m Endo MF broth (50 ml) (see below for preparation)

water samples

Making m Endo MF Broth

“This medium is extremely hygroscopic in the dehydrated form and oxidizes quickly to cause deterioration of the medium after the bottle has been opened. Once a bottle has been opened, it should be dated and discarded after one year. If the medium becomes hardened within that time, it should be discarded. Storage of the bottle inside a larger bottle that contains silica gel will extend shelf life.” (Benson 1994)

It is best to make up the medium the day it is to be used. It should not be stored over 96 hours prior to use. However, we discarded all medium the day of use. The Millipore Corporation recommends the following method for preparing this medium. (These steps are not exactly as stated in the Millipore Application Manual AM302.)

1. Into a 250 ml screw cap Erlenmeyer flask, we placed the following:

Distilled water ..................……………………………... 50 ml

95% ethyl alcohol ................………………………….. 2 ml

Dehydrated medium (m Endo MF broth)....................... 4.8 grams

We shook the above mixture by swirling the flask until the medium was dissolved and then we added another 50 ml of distilled water.

2. The flask was loosely capped and immersed it into a pan of boiling water. Our medium never boiled, although the lab directions called for simmering of the medium so we allowed the medium to sit in the water bath for 1 hour, on the approval of Cathy Simmerling.

3. We cooled the medium to 45 degrees Celsius.

Filtration of Sample

1. We prepared a small plastic Petri dish (100 by 15 mm) as follows:

a. With a flamed forceps, we transferred a sterile absorbent pad to a sterile plastic Petri dish.

b. Using a 2 ml pipette, we transferred 2.0 ml of m Endo MF broth to the absorbent pad.

2. A membrane-filtering unit was assembled as follows:

a. We inserted a sterile Buckner filter holder base into the neck of a 1-liter side-arm flask, and inserted the sterile Buckner funnel into the holder base creating a tight seal. We had two separate membrane filtering assemblies, one for water samples taken from the marina and one for the water collected from the rocks.

b. With a flamed forceps, we placed sterile filter paper inside the funnel and placed the membrane filter disk, grid side up, on the filter paper.

3. We attached the rubber hose to the air compressor and poured 50 ml of the sample water into the funnel, using a sterile graduated cylinder. We turned on the pump and filtered the water until all the water had filtered through. We then turned off the pump.

4. Carefully, we transferred the filter disk with sterile forceps to the Petri dish of m Endo MF broth, keeping the grid side up.

5. We repeated these steps until all of our desired number of samples were filtered. We rinsed the Buckner funnel with distilled water after each sample was filtered.

6. We incubated the plates at 35 degrees C for 22 to 24 hours. The plates were not inverted.

7. We counted the colonies on the disks, ignoring all colonies that lacked the golden metallic sheen.

Results

Marina Results

Date	Plate 1 # of colonies	Plate 2 # of colonies	Plate 3 # of colonies	Temperature degrees Celsius	Number of seagulls
10/28/01	0	0	0	11	50
11/01/01	12	6	4	13	70
11/05/01	10	17	10	1	100
11/09/01	0	8	13	0	40

It appears from this data that the opposite of our hypothesis is occurring. On the 5^th of November the colonies were the highest of the four dates and the temperature is several degrees colder than the previous dates. However, the number of seagulls present was higher compared to the other dates.

Rock Results

Date	Plate 1 # of colonies	Plate 2 # of colonies	Plate 3 # of colonies	Temperature degrees Celsius	Number of seagulls
10/28/01	0	0	0	11	0
11/01/01	69	33	0	13	0
11/05/01	3	67	45	1	0
11/09/01	12	56	37	0	0

E. coli coliform colonies were higher on the rock plates than the marina plates and seagulls were absent from the area on all sampling dates. Again, November 5^th had the highest number of colonies when the temperature was several degrees lower than the previous dates.

Next, we calculated the averages density of the E. coli colonies for each date using the following equation obtained from the "Water Analysis Handbook":

"Average coliform density for more than one filter/sample:

Coliform colonies per 100 mL= Sum of colonies in all sample *100 "

Sum of volumes (in mL) of all samples

Marina Results

Date	Average coliform density	Temperature (Celsius)	Number of Seagulls
10/28/01	0	11	50
11/01/01	22	13	70
11/05/01	37	1	100
11/09/01	21	0	40

The E. coli coliform density was calculated for each site for comparison to standard measurements used by health department officials when they calculate water quality.

Rock Results

Date	Average coliform density	Temperature (Celsius)	Number of Seagulls
10/28/01	0	11	0
11/01/01	102	13	0
11/05/01	115	1	0
11/09/01	108	0	0

On average, the rock formation site had higher E. coli coliform densities than the marina site.

Finally, we graphed our data to check for any correlation between the average E. coli density and the number of seagulls present/temperature.

The r-squared value for all three graphs are below one, in fact they are below .55, which means that a correlation between the E. coli density and the temperature is weak and the correlation between E. coli and seagulls present is moderate. The closer the r-squared value is to one, the more of a correlation.

The difference in E. coli coliform density between the two sites (the marina and the rocks) is significant. The p-value was .00036, with the omission of the data from 10/28/01; we believe that we ran the filtration wrong; therefore, the colonies did not grow.

Discussion

We found that our results for this experiment did not support our hypothesis. This could have been because we could not follow our set criteria to have all our water samples taken when the temperature outside was decreasing. We thought that doing this experiment in October would allow us to gather information as the weather decreased in temperature. However, Wisconsin weather is very unpredictable and October ended up have higher temperatures then normal.

To discuss our results, we met with the head researcher, Sandy McLellan, of the UWM Great Lakes Water Research Institution on Monday November 12, 2001. One of our first concerns was that the E. coli levels were higher at the rocks, where no seagulls were seen when sampling the water, and where we found dozens of seagulls by the marina E. coli levels were lower. Sandy suggested that this could have been the result of several factors. First, the birds maybe gathering near the rocks at night and then in the morning they move to the marina side, which Sandy and her team call the "bird cove". If this theory is true, then collecting the water samples in the morning (which was practice in the experiment) would not provide enough time for E. coli levels to increase at the "bird cove." Sandy also mentioned her team took samples periodically throughout the day and their E. coli levels changed constantly. Factors that could influence these changes are direction of the wind and the location of the seagulls.

The team from UWM found that if the seagulls were located out on the breakwater wall, the E. coli levels at the beach were lower and the water near the break wall contained almost no E. coli. Therefore, they believe that the depth of the water is a factor in the E. coli growth. The low E. coli levels near the shore and the absence of seagulls strengthened the belief that the gulls are the main cause of the E. coli levels.

Sandy mentioned that research has shown that seagulls produce 340 million E. coli per gram of defecation. This amount is two times greater then the levels found in geese droppings. Thus, they doubt that any other waterfowl are to blame for the contamination.

If we were to perform this experiment again, we would change our sampling method. First, we would sample in the morning and at night to test the hypothesis at the seagulls rest near the rocks at night. Second, we would take the water temperature of the water each time we sampled. Third, we would perform this experiment later in the season when temperatures are constantly low.

Literature Cited

Benson, H. (1994). Bacteriological Examination of Water: The Membrane Filter Method. Seventh Edition. Microbiological Applications. (p. 178-179 & 430). Boston, MA: McGraw-Hill.

Hach Company. (1997). Water Analysis Handbook, 3^rd Edition. Loveland, Colorado: Hach Company.

Schultze, S. (2001). Research Ties Gulls to Beach Pollution. Retrieved October 2, 2001 from the World Wide Web: http://www.jsonline.com

Schultze, S. & Rohde, M. (2000). Beach Bacteria Once Hit Sixty-five Times Dangerous Level in '99. Retrieved October 27, 2001 from the World Wide Web: http://www.jsonline.com