Abstract:
Levels of fecal-associated bacteria that may be harmful to human health can be very high
near haul-out sites of marine mammals. Water from sites near Seal Beach at Hopkins Marine
Station and from varying distances from the haul-out site was tested for bacterial abundance with
both epifluorescence microscopy and flow cytometry, and was not found to have increased total
numbers of bacteria near the haul-out site, even though it has been shown that the microbial
community at these sites is different than others around Monterey Bay. There was, however,
significant variation in bacterial abundance based on the day the sample was taken. Various
environmental factors including number of seals on the beach, water temperature, air
temperature, and wind speed were compared to the daily mean bacterial abundance counts to test
for a significant effect. Sea surface temperature was found to have the strongest correlation on
bacterial abundance of the factors tested.
tracchi 20
Introduction:
Water quality in coastal areas is frequently monitored to ensure human health and safety
in recreational or seafood-producing areas (Boehme 2005). Human activities can cause high
levels of pathogenic bacteria in coastal seawater, but many other factors contribute to bacterial
abundance as well (Stewart 2008). Marine mammal haul-out sites have been cited as a cause of
high levels of fecal-associated bacteria, due to the high density of animal feces being introduced
to the water (Calambokidis 1987). In winter 2011, the microbiology class at Hopkins Marine
station used 4-5-4 sequencing to examine bacterial community composition in various sites
around Monterey Bay. Seal Beach at Hopkins, so named because of the hundreds of harbor seals
that use it as a haul-out site, had a significantly different bacterial composition than any of the
other sites tested (Hopkins Marine Station, unpublished data). This study examines bacterial
abundance at the Seal Beach site and many sites nearby, to test if proximity to the seal haul-out
location would increase bacterial abundance, as well as change community composition. The
data were tested against a statistical null hypothesis that bacterial abundance was unaffected by
proximity to seal beach. In addition, other environmental factors that may have an effect on
bacterial abundance, such as time and weather conditions were examined over the course of the
study and compared to variability in bacteria counts. The number of seals was expected to
positively correlate with bacterial counts, as more fecal production would likely increase
bacterial concentration. Wind speed was expected to effect swell height and water mixing (Pinet
2005), which could possibly suggest that high winds could cause bacteria created at the beach to
be flushed out to sea, decreasing bacterial abundance (Boehme 2005). The temperature factors
were measured and compared post hoc, and had no initial hypothesis attached to them.
Beltracchi 2011 3
Methods:
I chose seven sites in the Hopkins Marine Reserve for sampling, and numbered them O-
6, with site 0 on Seal Beach, and site 6 the farthest from Seal Beach (Figure 1).
I sampled Site 0 only once during the experiment to minimize disturbance of the seals,
while I sampled site 1 a total of eight times to get a stronger temporal resolution of bacterial
variation close to the loading site as well as the weather station (ancillary data source). Sites 4
5, and 6 were sampled four times over a three week period, while sites 2 and 3 were sampled
three times each, with one sample taken from a site halfway between sites 2 and 3, called site
2.5, sampled on April 27, due to a high wave that caused me to lose my sampling bottle at site 2.
I collected water at high tide in IL plastic bottles. In the lab, I fixed 57 mL of each sample with
3mL 16% paraformaldehyde solution and refrigerated it overnight. I then filtered this through a
0.6um filter. 30mL of this filtrate was then filtered again through a 0.22um black polycarbonate
filter and stained with 2% DAPI saline solution (according to the Kirchman 2009 protocol). The
filter was then mounted on a slide, so that there were two filter slides for every sample. 1
counted bacteria on the slides in ten fields of view using a Nikon FXA microscope fitted with
DAPI filter 31000 (Chroma). The mean of 20 total fields of view was taken and compared for
each sample. One 60 mL sample of deionized water was also filtered, stained, and counted as a
control.
For flow-cytometric analyses, I fixed 1.5 mL of each unfiltered sample with 100uL of
16% PFA solution in 2mL vials and froze them until a trip to the flowcytometer could be
arranged. Once the samples thawed, I stained them with the cell-permanent, green-fluorescent
SYTO BC nucleic acid stain for 2 to 5 minutes in the dark. I then added 1OuL of a solution
containing 6um beads for comparison before the sample was run through the flowcytometer.
Beltracchi 2011 4
Abundaces were estimated with a BD Influx: Four-Laser Flowcytometer (MEGAMER, UCSC)
equipped with an air-cooled argon laser (488 nm, 25 mW). I analyzed scatter plots of particle
fluorescence using FlowJo software, and found 3 distinct populations of bacteria in most
samples, although all of the samples from May 16 and 17, as well as some of the samples from
May 2, only showed 2 distinct bacterial populations in addition to the beads.
In addition to bacterial counts, I obtained data on conditions at the time of sampling
including wind speed, air temperature, and sea surface temperature from the Hopkins Marine
Life Observatory Weather Station and the Scripps Institute of Oceanography Wave and
Temperature buoy. I counted the seals on Seal Beach at the time of sampling for five out of the
eight sampling days. Daily means and site means from both the flowcytometer counts and the
microscope counts from each sampling day were compared with a single-factor ANOVA.
Significant ANÖVA results were tested again with a Student-Newman-Keuls test. Linear
regressions were performed to test correlation between the bacteria count data sets and the
environmental factors (wind speed, air temperature, water temperature, and seal count).
Results:
The bacterial counts from the microscope and from the flow cytometer showed very
similar patterns for each sample (Figure 2), and served as a control to ensure both sampling
methods detected the same variation between sites and sampling days.
The variability between sites was not significant by a single-factor ANÖVA in either the
microscope counts or the flowcytometer data (Figure 3). Therefore, the null hypothesis that
distance from Seal Beach does not effect bacterial abundance was supported. There was,
however, a great deal of temporal variation, and a single-factor ANÖVA analyzing the day of
sampling was significant for microscope bacterial counts at the .05 level. Flowcytometer daily
Beltracchi 2011 5
counts were only significant at the .25 level. The microscope data was then tested with a post
hoc Student-Newman-Keuls test, which showed that May 2 is significantly different than the
sites with the lowest means (May 9 and 10), but no other differences were detected. The control
sample of deionized water from the lab had a very high number of cells, and although the
bacterial counts were lower than any of the ocean samples, the Student-Newman-Keuls test did
not show that it was significantly different than any sample other than the very highest ones.
This daily variation was compared to daily variation of environmental factors that may
have influenced bacterial abundance, including the number of seals on the beach, sea surface
temperature, wind speed, and air temperature. The environmental factor with the strongest
correlation to bacterial abundance was sea surface temperature, with an r°2 value of 29989
when compared to the microscope daily means, and r°2- 647 when compared to the daily means
of flowcytometer counts (Figure 3). The samples taken on May 2 in particular showed an
unusually high abundance of bacteria in both the microscope and the flowcytometer counts,
when the water was 13C at the time of sampling. This temperature was warmer than any of the
other sampling times (average between 11 and 12 degrees). In addition, the May 2 samples, as
well as those taken on May 16 and 17 showed only 2 distinct populations of bacteria in the
flowcytometer, while the samples from April 27 and May 6,9, 10, and 11 showed 3 populations
of bacteria (Figure 5). May 16 and 17 did not have outstandingly high sea surface temperatures,
but May 16 had the highest air temperature of 13.58C at the time of sampling.
Air has a lower heat capacity than water, so these two variables did not correlate
particularly well (r42=.057), and bacterial abundance did not correlate with air temperature as
well as it did with water temperature. The could have been some heat transfer, suggesting the
high bacterial count on May 16 and 17 could have been effected by warmer weather, although
Beltracchi 2011 6
more data is needed to support this, particularly considering the springtime upwelling in
Monterey Bay which would have effected water temperature more strongly.
Wind speed, which effects swell intensity, showed no correlation whatsoever with
bacterial abundance (r*2=.017). These were not analyzed any further, although it would be
interesting in a future study to investigate bacterial community composition based on wind¬
driven mixing.
When both sets of bacterial abundance data were compared to the number of seals on the
beach at the time of sampling (Figure 6), fewer seals seemed to correlate with higher amounts of
bacteria (with an r22 value of .94 in the comparison between seal counts and bacterial daily
means from microscope counts), when it was expected that more seals would mean more fecal
production, and therefore more bacteria in the water.
Discussion:
The microscope and flowcytometer counts showed many similar patterns in daily
variation of bacterial abundance, but the comparison of the two counts showed some distinct
differences as well (Figure 1). Differences between the results obtained from both may be due to
the fact that the water analyzed using the flow cytometer was not pre-filtered before the bacteria
counts were taken, so larger cells and pieces of phytoplankton may have been counted in the
sample. Any particles larger than 6um were filtered out of the sample before microscope counts
were taken, minimizing the potential for larger cells or particles to be included in the counts. In
addition, many of the flowcytometer samples showed a population of unstained cells (Figure 4),
likely Synechococcus cyanobacteria, which would not have been detectable in the microscope
counts.
Beltracchi 2011 7
The variation of bacterial abundance showed no significant difference based on distance
from the loading site, even though it has been shown in the past that bacterial community
composition is greatly different near seal beach. Abundance of bacteria, regardless of species
composition, showed little spatial variation at any given time (Figure 8), but other factors do
contribute to increased amounts of bacteria, particularly water temperature. A higher sea surface
temperature usually meant a higher number of bacteria in the water, perhaps suggesting warmer
temperatures is more conducive to bacterial growth. However, there could very well be other
confounding factors, such as upwelling, which effects both water temperature and nutrient
abundance (Pinet 2000), that were not tested in this experiment. There seemed to be
unexpectedly more seals hauled out on days when there were lower counts of bacteria in the
water, but these were also days when the water was cooler. A possible logical explanation for
this correlation is the seals were more inclined to be out swimming when the water was warmer,
and hauled out to sleep in the sun when the water was cold, as it has been shown that at least
during the molting season, seals will haul out for the purposes of increasing skin temperature
(Sydeman 1999). This experiment, too, showed a general trend of more seals on the beach when
the air was warmer, or when the water was cooler (Figure 7). The roughness of calmness of the
sea (as measured by my personal observations) did not show any correlation with bacterial
abundance, which may imply that fecal bacteria are not flushed from seal beach is by high
swells. However all samples in this study were taken at high tide, so perhaps tidal currents have
a stronger flushing effect than wind-driven waves (Boehme 2005). Tidal flushing of bacteria,
seal haul-out habits, and bacterial abundance changes based on water temperature could all be
investigated in future studies. To better understand the impact on the health of human
Beltracchi 201
beachgoers and seafood consumers caused by the seal haul-outs, water should be tested for fecal
bacteria, and the bacterial community in this water can be better analyzed.
Acknowledgements:
I would first like to thank Dr. Vanessa Michelou and Dr. Sarah Lee for their valuable
feedback and support of this project throughout. I would also like to thank Professor Steve
Palumbi for his supervision of the project, and the rest of the Hopkins Marine Station faculty for
their feedback, especially Professor James Watanabe for teaching me how to do the statistics for
this project. In addition, I would like to thank Brandon Carter for allowing me to use the
MEGAMER flowcytometer, and Chris Patton for helping me use the microscope. Finally, I
would like to thank my classmates Acata Felton, lan Markham, and Walker Clayton for their
contributions to my learning experience.
Literature Cited:
Boal, J., 1980. Pacific Harbor Seal (Phoca vitulina richardii). Haul Out Impact on the Rocky
Midtidal Zone. Marine Ecology Progress Series, 2, pp.265-269. Available at:
http: /www.int-res.com/articles/ meps/2/m002p265.pdf.
Boehm, A.B. & Weisberg, S.B., 2005. Tidal forcing of enterococci at marine recreational
beaches at fortnightly and semidiurnal frequencies. Environmental science & technology,
39(15), pp.5575-83. Available at: http://www.ncbi.nlm.nih.gov/pubmed/16124289.
Calambokidis, J. & McLaughlin, B., 1987. Harbor Seal Populations and their contributions to
fecal coliform pollution in Quilcene Bay, Washington.
Hopkins Marine Station. 2011 unpub. data.
Izbicki, J.A. et al., 2009. Sources of fecal indicator bacteria in urban streams and ocean beaches,
Santa Barbara, California. Annals of Environmental Science, 3,p.139-178.
Kirchman Lab Methods, 2009. Direct Counts of Microbes by Epifluorescence Microscopy.
Pinet, P.R., 2000. Invitation to Oceanography. 2nd ed. Jones and Bartlett Publishers, Sudbury,
Massachusetts.
Beltracchi 2011 9
Schneider, D.C. & Payne, P.M., 1983. Factors Affecting Haul-Out of Harbor Seals at a Site in
Southeastern Massachusetts. Journal of Mammalogy, 64(3), p.518-520. Available at:
http://www.jstor.org/stable/1380370 [Accessed April 15,20111.
Sydeman, W. J. and S. G. Allen (1999). "Pinniped population dynamics in central California:
correlations with sea surface temperature and upwelling indices." Marine Mammal Science
15(2): 446-461.
Simpkins, M. a et al., 2003. Stability in the Proportion of Harbor Seals Hauled Out Under
Locally Ideal Conditions. Marine Mammal Science, 19(4), pp.791-805. Available at:
http://doi.wiley.com/10.1111/j.1748-7692.2003.tb01130.x.
Stewart, J.R. et al., 2008. The coastal environment and human health: microbial indicators,
pathogens, sentinels and reservoirs. Environmental health : a global access science
source, 7 Suppl 2, p.S3. Available at:
http: /www.pubmedcentral. nih.gov/articlerender. fegi? artid=25867 16& tool-pmcentrez&re
ndertype-abstract.
Yochem, P.K. et al., 1987. Diel Haul-Out Patterns and Site Fidelity of Harbor Seals (Phoca
Vitulina Richardsi) on San Miguel Island, California, in Autumn. Marine Mammal
Science, 3(4), pp.323-332. Available at: http://doi.wiley.com/10.1111/j.1748-
7692.1987.tb003 19.X.
tracchi 201
Figures:
Figure 1:


Map of seven sites from which water samples were taken.
Beltracchi 2011 11
Figure 2:
Scope Mean vs. log Flowcytometer count, every sample
8
+
+
++
8 6
+
—
-+-
55.25
80.375
30.125
105.5
mean cells/view
Log Flowcytometer daily means
08
2 06
8 04
8 02
00
Apr 27, May 2, May 6, May 9, May 10, May 11, May 16, May 17, DI
2011 2011 2011 2011 2011 2011 2011 2011 control
Scope mean by day
100.00
80.00
60.00
40.00
§ 2000
E
.00
May 9, May 10, May 11, May 16, May 17, DI control
Apr 27,
May 2,
May 6,
2011 2011 2011 2011 2011 2011 2011
2011
Microscope counts show a similar pattern to flowcytometer counts, as seen directly compared
for each sample (top) and in the daily means of both types of counts from each sampling day
(bottom).
Beltracchi 2011 12
Figure 3:
SST Vs. Scope Mean
81.80
63.85

45.90
â”
27.95
10,00 —
SST vs. Log Flowcytometer mean
10.000 —
8.787
7.573 —
6.360
++
5.140
15.00
10.00
11.25
13.75
12.50
Both sets of bacterial abundance data show a moderately strong positive correlation with water
temperature.
Beltracchi 20
Figure 4
Scope count vs. site
105.5000
90.4286
75.357
+
60.2857 +

452143
30.1429

*

15.0714
6
—
2
3
Site number
Log FlowCytometer Count vs. site number
8.626
7.394
6.162 +
+
4

4929
3.697
2.465
1.232
Site
Mean number of cells in each field of view in the microscope for every sample, with site 7 being
the deionized water control sample. There is no discernible pattern or significant relationship
between bacterial abundance and distance from Seal Beach in the microscope count (top) or in
the flowcytometer count (bottom).
Beltracchi 2011 14
Figure 5
10*
102
101

10
10
10° 101
102 103 10
10° 10
102 103
572-27
572-27
A typical sample from May 2, 16, and 17 show 2 distinct populations of bacteria, plus the beads
added to the sample for comparison (right), while a typical sample from all the other days
showed 3 populations of bacteria in addition to the beads (left). The extra population is most
likely cyanobacteria, which did not react to the stain.
racchi 20
Figure 6
Seal count vs. log cells/mL (Flowcytometer)
7.876
7.194
6.511
5.829
5.146
170.00
91.00
110.75
130 50
150.25
Seal count vs. cells/viewfield (Microscope)
90.000
70.575

51.150
31.725
12.300
150
130
170
110
Surprisingly, there seemed to be fewer bacteria on average on sampling days with a higher
number of seals, and this was reflected in both data sets.
Beltracchi 2011 1
Figure 7:
SST vs. Seal Count
170.0
127.5

85.0
42.5
—
11.25
3.75
7.50
15.00
Air Temperature Vs. Seal Count
170.00
129.75

89.50
49.25
9.00
T140
+ 135
105
15.0
The temperature measurements compared to the seal counts show a general trend of more seals
out of the water when the air is warmer, or when the water is colder.
Beltracchi 201
Figure 8:
6 control Daily
Scope
mean
counts
(cells/vie
wfield)
47.5
59.8
40.52
46.7
Apr 27
23.6
23.6
77.8
61.2
88.15
105.5
78.05
81.8
80.1
May 2
33.3
30.15
44.467
74.65
47.5
May 6
33.5
May 9
34.7
34.7
May 10
884
51.925
45.45
May 11
56.05
56.05
May 16
24.95
52.116
38.85
74.9
56.1
57.9
May 17
12.3
12.3
DI
control
44.94375 51.8
48.2375
60.625
61.875
54.125
12.3
58.4
site
mean
control
day mean
flowcyt
(cell/mL)
1.19E406
9.84E405 3.65E406 3.65E406 1.12E406 1.52E406 1.15E406 X
Apr 27
7.52E407
2.78E+06 4.23E+08 7.91E406 1.02E+07 5.81E+06 1.60E+06 X
May 2
1.30E406 1.27E406 1.18E406 6.55E+05 1.99E406 1.39E406 X
1.30E+06
May 6
1.59E406 X
1.59E406
May 9
5.50E+05
5.50E+05 X
May 10
1.88E+06 6.37E405 X
1.26E+06
May 11
2.87E+06
May 16
2.87E+06 X
4.47E406
4.40E406 4.74E406 2.50E406 8.73E406 3.13E406 3.37E406
May 17 X
1.40E405 1.40E405
DI
Contro
1880000 1.89E406 1.08E408 3.81E406 5.18E+06 3.11E406 1.88E406 1.40E+05
site
mean
Tables of microscope and flowcytometer counts, as well as the means by day and by site.
Beltracchi 2011 1