Assessment of the Wildlife
Reproduction and Deformities
Beneficial Use Impairment in the
Thunder Bay Area of Concern

 Environment and Climate Change Canada – Ecotoxicology & Wildlife
Health Division
K.D. Hughes, D. Crump, K. Williams and P.A. Martin
September 2016

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 2

 Citation: Hughes, K.D., D. Crump, K. Williams, and P.A. Martin. Assessment of the
wildlife reproduction and deformities Beneficial Use Impairment in the Thunder Bay
Area of Concern. 2016. Environment and Climate Change Canada – Ecotoxicology
and Wildlife Health Division report. September 2016. 40 pp.

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 1

 ABSTRACT
Reproduction and development were examined in herring gulls (Larus argentatus)
breeding within the Thunder Bay Area of Concern in 2012, 2014 and 2015. Freshlylaid eggs were collected from two colonies within the Area of Concern (AOC) and a
reference colony outside of the AOC, artificially incubated in the laboratory and
assessed for embryonic viability, incidence of embryonic deformities, contaminant
burdens and other biochemical endpoints (i.e., stable isotopes). Productivity was
determined at the colonies when chicks were > 21 days old and chicks were
examined for morphological deformities as well as other biological endpoints. A
second aquatic-feeding colonial waterbird species, the double-crested cormorant
(Phalacrocorax auritus), nesting at a colony within the AOC was also selected for
assessment of contaminant burdens. Overall, embryonic viability of herring gulls
was high at AOC colonies and herring gull productivity at these colonies was within
the range required to maintain a stable population. No embryonic deformities were
evident in gull eggs incubated in the laboratory and no morphological deformities
were found in > 21-day-old herring gull chicks from AOC colonies (based on sample
sizes of 10-33 chicks). Importantly, contaminant burdens (e.g., PCBs, 2,3,7,8-TCDD,
and mercury) in gull embryos and double-crested cormorant eggs from the Thunder
Bay AOC were not notably elevated and were comparable to or lower than burdens
in embryos or eggs from respective reference colonies in these years. No adverse
health effects were evident in gull chicks based on measurements of biochemical
endpoints relating to growth and development and immune function. Population
trends based on nest count surveys indicate greater numbers of nests for both
species at AOC study colonies in 2007 and 2012 compared to earlier decades. Lakewide population trends based on decadal surveys conducted on the Canadian and
American sides of Lake Superior indicate that nest numbers of herring gulls have
fluctuated while cormorant nest numbers have increased steadily since the late
1970s. Large declines in concentrations of sum PCBs and other organochlorines,
dioxins and TEQs in herring gull tissues since the early 1990s indicate that exposure
to these compounds has decreased in herring gulls foraging in the AOC. While no
change in mercury levels was found in herring gulls in the AOC, this pattern was
similar to that found at other Lake Superior herring gull colonies during this period.
In summary, concentrations of PCBs, other organochlorine compounds, PBDEs,
dioxins and furans, and mercury were not sufficiently elevated to adversely impact
the reproductive success and development of herring gulls and cormorants nesting
in the Thunder Bay AOC.

INTRODUCTION
The Thunder Bay Area of Concern (AOC) is one of 43 Great Lakes AOCs that were
initially identified by Canada, the United States and the International Joint
Commission (IJC) as specific locations where local environmental degradation had
severely impacted the area’s ability to support aquatic life. The border of the AOC
extends approximately 28 kilometres along the shoreline of Lake Superior from
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 1

 north of Bare Point to Flatland Island and up to 9 kilometres offshore, including the
Welcome Islands (Figure 1). Historical discharges of pollutants from local pulp and
paper industries and wastewater treatment plants, as well as atmospheric
deposition and urban runoff impaired water quality and contaminated sediment in
the Bay (Thunder Bay RAP Team 1991). Contaminants of concern included dioxins
and furans, mercury, and polychlorinated biphenyls (PCBs) which contributed to
exceedances of water

Figure 1. The Thunder Bay Area of Concern.

quality objectives, sediment quality guidelines and/or fish consumption guidelines
(Thunder Bay RAP Team 1991, 2004).
Fourteen beneficial use impairments (BUIs) were used by the Canadian and U.S.
federal governments to identify and assess the extent of environmental degradation
and direct restoration and remedial activities. One of these BUIs, “bird or animal
deformities or reproduction problems”, addresses contaminant exposure or other
anthropogenic environmental stressors on reproductive success or deformity rates
in wildlife. Initially, this BUI was designated as “requires further assessment” since
no formal studies had been conducted (Thunder Bay RAP Team 1991). The status of
the BUI was subsequently changed to “not impaired” following no reports of bird
and animal deformities in the AOC and observations that cormorant populations
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 2

 were reproducing at normal levels and numbers of colonies on Lake Superior had
increased since the early 1970s (Weseloh et al. 1995; Thunder Bay RAP Team 2004).
However pending the results of additional studies conducted by the Canadian
Wildlife Service (CWS) initiated in 2000, the status of this BUI was changed back to
“requires further assessment” (Environment Canada and the Ontario Ministry of the
Environment 2011). In 2012, studies of the potential effects of contaminants on
reproduction and development in aquatic-feeding wildlife were initiated by
Environment and Climate Change Canada (ECCC) to more fully evaluate and assess
the status of this BUI in the Thunder Bay AOC. The approach used here has been
implemented in the St. Marys River (Ontario) and Hamilton Harbour AOCs for similar
assessments of this BUI.
Fish-eating wildlife, such as colonial waterbirds, are important indicators of exposure
to persistent contaminants in the aquatic environment (Fox and Weseloh 1987). As
top predators, they occupy a high trophic level in the aquatic food web and
therefore can accumulate high levels of contaminants, which may in turn adversely
affect their reproductive health and development. Two colonial waterbird species,
that breed and forage within the Thunder Bay AOC, were selected for assessment
purposes. The herring gull (Larus argentatus) is a long-lived, primarily fish-eating
colonial waterbird that from the time it reaches breeding age is a year-round
resident in the Great Lakes basin. A second aquatic-feeding colonial waterbird
species, the double-crested cormorant (Phalacrocorax auritus), was selected for
assessment of contaminant burdens. This species feeds almost exclusively on fish
compared to herring gulls, which are opportunistic feeders that will consume
terrestrial prey if fish are not readily available. This close connection to the aquatic
environment is vital for assessment of local conditions in the AOC.
In 2012, 2014 and 2015, breeding colonies of herring gulls were studied by ECCC in
the Thunder Bay AOC. Freshly-laid eggs were collected for artificial incubation in the
laboratory to assess embryonic viability, incidence of embryonic deformities,
contaminant burdens and biochemical endpoints (i.e., stable isotopes). Under
controlled laboratory conditions, this method assesses the effects of embryonic
exposure to potentially high levels of contaminants during critical periods of
development. Reproduction and development were also assessed in wild
populations with visits to colonies to monitor productivity and examine chicks for
morphological deformities and to measure additional biochemical endpoints (i.e.,
stress hormone, thyroxine levels) relating to growth and development as well as
immune function, which could be affected by increased contaminant exposure. The
results of this study will be used to assess the status of the wildlife reproduction and
deformities BUI in this AOC. The results of earlier studies conducted by CWS and
initiated in 2000 in the AOC are referenced in the discussion with further details
provided in the Supplementary Information section.

METHODS

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 3

 Two herring gull colonies were selected within the Thunder Bay AOC at Mutton
Island (48o22’15”N, 89o11’43”W) and an unnamed island (48o21’0”N, 89o08’34”W)
located further out from the City of Thunder Bay shoreline and south of the two
main Welcome Islands (Figure 2). Specifically, there are two small, unnamed islands
situated approximately 0.4 kilometers and 0.6 kilometers south of the Welcome
Islands. The AOC colony for this study is the latter island that is also smaller in size
(0.34 hectares). For our purposes, this study colony will be hereafter referred to as
“Welcome Islands-a”. Historically, colonial waterbird studies have also been
conducted on the larger of these two unnamed islands (0.98 hectares) and this site
will be subsequently referred to as “Welcome Islands-b” (48 o21’07”N, 89o08’35”W).
Initially Granite Island (48o43’11”N, 88o27’33”W), located approximately 65 km
northeast of the AOC in Black Bay of Lake Superior, was selected as the AOC
reference colony. However, during the first visit in late April of 2012, this portion of
the lake was not yet ice-free and gulls had not begun laying eggs. Since it was not
possible to collect eggs for artificial incubation or to build enclosures for the field
study at Granite Island at that time, Double Island (46 o10’24”N, 82o51’50”W) in the
North Channel of Lake Huron was selected as the alternate reference site for
assessment of these components. This coincided with work that was underway for a
similar study in the St. Marys River AOC (Ontario) where Double Island served as the
non-AOC reference site.
Figure 2. Colony locations for herring gulls and double-crested cormorants in the
Thunder Bay AOC in 2012, 2014, and 2015.

Visits to each colony were made at two times during the breeding season for gulls:
1) egg laying (late April) and 2) when chicks were > 21 days old (mid-May) in 2012
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 4

 and 2014 to assess reproduction and various parameters of health. In 2014, due to
late spring and high water levels, productivity could not be sufficiently determined
at Welcome Islands-a because 10 of the 12 nest enclosures were washed out. As a
result, an additional year of field study (2015) was required to estimate productivity
at AOC colonies. Health parameters were also examined in 2015 to supplement the
two earlier years of data. During the first visit, 15 freshly-laid eggs (i.e., not
incubated) were collected from one-egg nests at each colony for artificial incubation
in the laboratory at the National Wildlife Research Centre (NWRC) in Ottawa.
Embryonic viability, incidence of embryonic deformities, contaminant burdens, and
stable isotope signatures were determined. In addition, a thorough nest count was
conducted and contents were recorded at the colony. Individual nest enclosures
(~1m in diameter and 16” high) were constructed around twelve 3-egg nests at
each colony. As a measure of colony health, egg measurements for up to forty-two
3-egg clutches were recorded (in millimetres) and egg volume calculated as:
Egg volume (cm3) = 0.489 x (length x breadth2)/1000
Total clutch volume was determined as the sum volume of the three eggs in the
clutch. Intraclutch variation in egg size was calculated as the difference in volume
between the largest and smallest egg in the clutch divided by the largest egg size
(i.e., volume) and multiplied by 100.
During the second visit, when chicks from enclosed nests were > 21 days old,
productivity was calculated as:
Productivity = no. of > 21-day-old chicks/no. of enclosed nests
Enclosed nests that had been abandoned, damaged by high water levels, or where
there was evidence that chicks had escaped were not included in estimates of
productivity. Body measurements of chicks, including mass, tarsus, wing cord, and
culmen length were recorded and chicks were banded with a stainless steel USFW
band. Chicks were examined for morphological deformities when they were > 21
days old. Health effects were also assessed in chicks at this stage. A blood sample
was collected from the brachial vein of chicks using a 25 gauge x 5/8” needle and
heparinized syringe to examine thyroxine concentrations in plasma (see below for
further details). In addition, two secondary covert feathers were collected to
quantify corticosterone concentrations as a measure of stress over time in herring
gull chicks. Immune status of chicks was evaluated using a phytohemmagglutinin
(PHA) skin test.
Double-crested cormorant eggs were also collected for assessments of contaminant
burdens. Cormorants nesting on Welcome Islands-a served as the AOC colony and
Black Rock (46o06’59”N, 82o50’00”W) in the North Channel served as the
corresponding reference colony. In May of 2012 and 2014, 10 or 13 eggs were
collected from nests containing > 3 eggs at each colony. After collection, the eggs
were sent to NWRC where contents were placed in chemically-cleaned glass jars,
homogenized, and frozen until chemical analysis. Eggs from each colony were
pooled together as a single sample for chemical analysis.
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 5

 Artificial Incubation of Eggs:
In 2012 and 2014, unincubated herring gull eggs were collected in the field from
nests containing a single egg, transported to NWRC in insulated coolers with foam
inserts and set in a Petersime incubator (model# MX-1) at 37 oC, 58% humidity and
turned every two hours. It was not possible to follow this protocol at Double Island in
2014 due to late ice-out conditions and the inability to access the colony during the
early-laying period. Consequently, subsequent collection of eggs and transport to
the laboratory at this later egg-laying period likely compromised embryonic viability
of eggs under artificial incubation conditions. Following this, the results of the
artificial incubation study conducted at Double Island in 2013 were used for
comparisons with the AOC colonies in 2014. The incubated eggs that were collected
from Double Island in 2014 however, were analyzed for contaminant burdens (see
details below).
Just prior to the pipping stage of development (i.e., embryonic day 26-27), embryos
were removed from their shells and euthanized by decapitation. Each embryo was
examined for physical deformities. Embryonic viability was determined as the
number of viable embryos that survived to the designated embryonic day (i.e., just
prior to pipping) divided by the total number of fertile eggs. Eggs that were
nonviable were staged if possible (e.g., infertile; early, mid or late embryo death).
Egg contents, including yolk sac, whole carcass, and shell membranes, were
collected in chemically-cleaned glass jars, homogenized, and frozen until chemical
analysis for contaminants. Ten to fifteen embryos were randomly selected from each
colony for chemical analysis in the two study years.

Contaminant Analyses:
Chemical analyses of herring gull embryos and cormorant eggs for organochlorine
compounds and polybrominated diphenyl ethers (PBDEs) were conducted at the
Great Lakes Institute for Environmental Research at the University of Windsor (2012AOC colonies) and at NWRC (2012-reference colonies, 2014-AOC and reference
colonies). Organochlorine compounds measured included p,p’-DDE
(dichlorodiphenyldichloroethylene), oxychlordane, cis-chlordane, trans-chlordane,
cis-nonachlor, trans-nonachlor, hexachlorobenzene (HCB), dieldrin, heptachlor
epoxide (HE), mirex, octachlorostyrene (OCS), and polychlorinated biphenyls (PCBs).
Sum chlordane is based on the sum concentrations of oxychlordane, cis-chlordane,
trans-chlordane, cis-nonachlor, and trans-nonachlor. Prior to chemical analysis,
thawed embryos and eggs were homogenized and then underwent neutral
extraction and removal of lipids and biogenic compounds by gel permeation
chromatography and further clean up by Florisil column chromatography.
Quantitative analysis of organochlorine compounds was performed using capillary
gas chromatography coupled with a mass selective detector (GC-MSD) operated in
selected ion monitoring mode. PBDEs were quantified by gas chromatography high
resolution mass spectrometry methods using a time-of-flight mass spectrometer
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 6

 (GC-MS-TOF) at GLIER and by GC-MSD operated in the NICI mode at NWRC. Sum
PCBs were based on the sum concentrations of 35-62 individual or co-eluting PCB
congeners found above the limit of detection depending on the year and laboratory
performing the analysis. Similarly, sum PBDEs were based on the sum
concentrations of 15 or 25 individual or co-eluting PBDE congeners found above the
limit of detection. Certified internal standards were used for quantification and
certified reference materials, blanks and duplicate samples were analyzed for
quality assurance purposes. Concentrations of organochlorines and PBDEs are
reported in µg/g on a wet weight basis. Detection limits for organochlorine
compounds and individual PBDE congeners ranged from 0.00002-0.0003 µg/g.
Embryos and eggs were analyzed for non-ortho substituted PCBs, polychlorinated
dibenzo-p-dioxins, including 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), and
polychlorinated dibenzofurans using gas chromatography-high resolution mass
spectrometry (GC/HRMS) at RPC Laboratory in Fredericton, New Brunswick. Methods
were based on US EPA Method 1613B and 8290A for dioxins and furans and on US
EPA Method 1668C for non-ortho PCBs. Reference materials, blanks, and duplicates
were analyzed for quality assurance purposes. Samples were analyzed as a single
pool consisting of 10-15 gull embryos at each colony. Embryos from Double Island in
2012 were not analyzed for these compounds since the results of analysis from the
year prior (2011) were available when eggs were also collected as part of the St.
Marys River AOC study and artificially incubated in the lab under identical
conditions (single pool of five embryos). Double-crested cormorant eggs were
analyzed for non-ortho PCBs, dioxins and furans in 2014 only and as a single pool
comprising 10 eggs from each colony.
Total mercury was quantified on a dry weight basis using an Advanced Mercury
Analyzer (AMA-254) as described in CWS Method No. MET-CHEM-AA-03I at NWRC.
Certified reference materials and duplicate samples were also analyzed to ensure
correct calibration, accuracy, and reproducibility of test methods. Mercury
concentrations in 2012 and 2014 embryos and eggs are reported in µg/g on a wet
weight basis using percent moisture content.
As part of earlier research studies examining health effects and exposure in herring
gulls in the Thunder Bay AOC, eggs were collected from Mutton Island for several
years since 1992 and analyzed for contaminants (ECCC unpublished). Compared to
current data, these early data allow for an assessment of temporal trends in
contaminant exposure for birds foraging within the AOC. Double-crested cormorant
eggs were also collected for contaminant analysis from Cone Island (48o13’17”N,
89o01’39”W) in 1989, a site that is just beyond the AOC boundary and east of Pie
Island. All historic data are presented where available and compared to current
burdens in herring gull embryos and cormorant eggs.

Stable Isotopes:

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 7

 Stable isotope analyses of samples were conducted at the University of Ottawa’s
G.G. Hatch Stable Isotope Laboratory in Ottawa, Ontario. Following lipid-extraction,
samples were weighed into tin capsules and loaded into an elemental analyser. The
sample was flash combusted at ~1800oC (Dumas combustion) and the resultant gas
products were carried by helium through columns of oxidizing/reducing chemicals
optimised for CO2 and N2. The gases were separated by a purge and trap adsorption
column and then sent to the Delta Advantage isotope ratio mass spectrometer
coupled with Conflo III. Samples were normalized to internal standards and
calibrated to international standards. Stable isotope ratios are expressed in δ
notation as the deviation from standards in parts per thousand (‰) according to the
following relationship:
δX = (Rsample-Rstandard)/Rstandard x 1000
where X is 15N or 13C and R is the corresponding ratio 15N/14N or 13C/12C. In this study,
δ15N signatures were compared to infer relative (and not absolute) trophic position
at colonies.

Feather Corticosterone:
Feathers from chicks were prepared by first removing the calamus, i.e., the proximal
end of the quill to where the feathers start (~ 5 to 10 mm, depending on length of
feather). The remaining portion was then minced with scissors until homogenous
and 1.5 ml methanol was added. The sample was vortexed, sonicated for 30
minutes at room temperature, and incubated overnight at 50 oC while shaking. After
centrifuging at 13,000 rpm for 20 minutes, the supernatant was transferred to a
fresh tube and the original feather sample was re-extracted with 1 ml methanol.
This re-extraction was again vortexed for 5 minutes and the supernatant was
removed and added to the original fraction. The entire methanol fraction was
evaporated to dryness overnight in a fume hood and then the sample was
reconstituted with 200 µl steroid diluent.
Corticosterone concentrations in the extracts of the feather sample were
determined using a corticosterone EIA kit (Assay Designs – corticosterone enzyme
immunoassay kit; product no. 900-097; 96 well kit). This is a competitive
immunoassay for the quantitative determination of corticosterone in biological fluids
and it uses a polyclonal antibody to corticosterone to bind in a competitive manner.
In the presence of the corticosterone antibody, corticosterone in the sample
competes with a known amount of corticosterone that has an alkaline phosphatase
molecule covalently attached to it. The excess reagents are washed away and the
substrate is added. Following a short incubation time, the enzyme reaction is
stopped and the samples are read on a SpectraMax 190 UV-VIS microplate reader
(Molecular Devices, Sunnyvale, California, USA) at 405 nm. The intensity of colour
produced in the reaction is inversely proportional to the concentration of
corticosterone in the sample. The measured optical density was used to calculate

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 8

 the concentration of corticosterone. The average net optical density (OD) bound for
each standard and sample is calculated by subtracting the average NSB (nonspecific binding) OD from the average OD bound:
Average Net OD = Average Bound OD - Average NSB OD
Then the binding of each pair of standard wells as a percentage of the maximum
binding wells (Bo) was calculated using the following formula:
Percent Bound = Net OD x 100 Net Bo OD
The plot of percent bound versus concentration of corticosterone for the standards
was graphed using Prism software (GraphPad, La Jolla, California, USA) and a line
fitted through the points. The concentration of corticosterone in the sample was
then determined. An in-house quality control sample was used in each plate. The
sensitivity of the assay is 27 ρg/ml.

Thyroid Status:
Blood samples were centrifuged for 5 minutes at 14,000 rpm to separate plasma
from red blood cells. Plasma was stored at -80°C and red blood cells were stored at
4°C. Concentrations of free thyroxine in the plasma of chicks sampled in the three
study years were determined using a commercially available kit (AccuBind Elisa
microwells product no. 1225-300 - Monobind Inc., Lake Forest, CA, 92630, USA). The
method is based on a competitive enzyme immunoassay format in which a
competition is set up between an immobilized antibody, the enzyme-antigen
conjugate and the free thyroxine in the sample. When equilibration is reached, the
unbound antigen fraction is removed and the enzyme activity in the bound fraction
is measured which is inversely proportional to the concentration of free thyroxine in
the sample. As per the method, standards, controls and sample plasma, as well as
the enzyme reagent, were added to the microplate wells, which contained the
immobilized antibody. After an incubation period, the unbound fraction was washed
from the wells and the substrate added. The reaction was stopped and the plate
was read at 450 nm on a SpectraMax 190 UV-VIS microplate reader (Molecular
Devices, Sunnyvale, California, USA). Concentrations were determined from the
standard curve, which was fitted using a variable slope (four parameter) method
using Prism software (GraphPad, La Jolla, California, USA).
Controls used were Randox product no. HS2611 assayed human sera levels 2 & 3
(Randox Laboratories Ltd., Antrim, UK). The method gave the intra-assay precision
as 10.98%, 4.26%, and 3.25 % coefficient of variation for low, medium and high
controls, respectively. The inter-assay precision was 10.81%, 6.01%, and 7.90 %
coefficient of variation for low, medium, and high controls, respectively. The
detection limit of this assay was 0.32 ng/dl.

Immune Response - Phytohemagglutinin (PHA) Skin Test:
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 9

 Feathers were plucked from the patagium (“wing web”) of each wing in order to
measure the thickness to the nearest 0.05 mm using pressure-sensitive calipers
(Dyer 304-196). Two measurements were taken of each wing web and the mean of
the two measurements was determined for each wing. To assist in measuring the
same location on day 2, a small dot was made with a permanent marker at the site
of measurement. Following the thickness determination, 0.1 ml of a 1 mg/ml
solution of PHA (Sigma) dissolved in sterile phosphate buffered saline (PBS; 0.01M
phosphate buffer, 0.137M NaCl, pH 7.4) was injected intra- or sub-dermally into one
wing web and 0.1 ml of PBS solution was injected into the other wing web as a
control using a 1 ml syringe fitted with a 26 gauge intradermal needle. Injections of
PHA and PBS were randomly alternated between right and left wings of birds tested.
Twenty-four hours (+6 hours) after injection, the thickness of each wing web was remeasured in duplicate at the site of injection (indicated by marker dot) and the
mean of the two measurements again determined. For each individual chick, the
PHA stimulation index was calculated as the increase in wing web thickness caused
by PHA in one wing web minus the increase caused by PBS in the other wing web.
This test was conducted using chicks at study sites in 2015 only.

Statistical Analysis:
Contaminants and other biological endpoints were statistically analyzed using either
the Student’s t-test for between colony comparisons or a one-way ANOVA for
among-colony comparisons, which when significant, was followed by Tukey’s HSD
test. Data were log-transformed (log10) to meet conditions of equal variance and
normality for parametric analysis. If data failed these assumptions, comparisons
were made using either a Mann-Whitney U non-parametric test or Kruskal-Wallis
one-way analysis of variance by ranks; post-hoc tests were conducted using nonparametric multiple comparison tests for unequal sample sizes. Concentrations of
chemical and biochemical endpoints that were below the limit of detection were
given a value of one-half of the detection limit. Since numbers of PCB and PBDE
congeners differed depending on the year and laboratory performing the chemical
analysis, statistical analysis of contaminant burdens among colonies were
conducted using congeners that were common to all chemical analyses. For sum
PCBs and sum PBDEs, this represented 35 and 14 common congeners, respectively.
Mercury concentrations in samples were statistically analyzed on a dry weight basis;
however, concentrations are reported on a wet weight basis for comparisons to
published values and thresholds. A Spearman rank correlation analysis was
performed to examine the relationship between the two stable isotopes in gull
embryos. Data for productivity and the three health parameters relating to stress,
thyroid status and immune response were combined for chicks from the two AOC
colonies to increase sample size for comparisons to the reference colony. All results
were considered significant at p<0.05.
Concentrations of 2,3,7,8-TCDD toxic equivalents (TEQs) were calculated for dioxinlike PCBs, furans, and dioxins and are based upon toxic equivalency factors
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC
10

 developed by van den Berg et al. (1998) for birds. Dioxin-like PCBs include four nonortho PCB congeners (77, 81, 126, and 169) and eight mono-ortho PCB congeners
(105, 114, 118, 123, 156, 157, 167, and 189). Of the eight mono-ortho PCB
congeners, three were quantified in all analyses (105, 118, and 156), three were
quantified in some analyses (114, 157 and 189) and two were not quantified in any
analyses (123 and 167). Mean mono-ortho PCB concentrations, quantified in the
chemical analysis for organochlorines, were calculated for individuals used to create
the pool (where data were available) for reporting of TEQs associated with monoortho PCBs. Total TEQ concentration is based on the sum concentration of TEQs
calculated for the 4 non-ortho PCBs, 3 or 6 mono-ortho PCBs (where data were
available), and 17 dioxin and furan congeners.

RESULTS
A) Artificial Egg Incubation Study
Embryonic Viability and Deformities:
Embryonic viability was consistently high at 92% overall in herring gulls from the
Thunder Bay AOC and ranged between 83% to 100% at the two AOC colonies,
Mutton Island and Welcome Islands-a, in 2012 and 2014 (Table 1). Embryonic
viability was also high overall at 95% in gulls from the Double Island reference
colony and equal to 96% and 93% in 2012 and 2013, respectively. Of 51 fertile eggs
examined in 2012, a total of three embryos died: one from Mutton Island (early
development; could not stage because the contents were rotten), one from
Welcome Islands-a (stage 44) and one from Double Island (stage 29). Of 39 fertile
eggs examined in 2013/4, two embryos died from Mutton Island (one very early, i.e.,
stage 6-8, and stage 33) and one embryo died from the reference colony (stage 3637). No embryonic deformities (0%) were evident in incubated eggs from either of
the AOC colonies or the reference colony in all study years. Overall, artificially
incubated herring gull eggs collected from AOC colonies had high rates of
embryonic viability and no embryonic deformities, which was similar to the results
observed at the reference colony in the two study years.

Contaminants:
Concentrations of organochlorine compounds were generally low in herring gull
embryos from the two AOC colonies and the reference colony in 2012 and 2014 with
sum PCBs found at the highest concentrations (Table 2a). Mean sum PCB
concentrations ranged from 0.75 µg/g in embryos from Welcome Islands-a in 2012
to 1.54 µg/g in embryos from Double Island in 2014. Mean concentrations of the
remaining organochlorines in embryos were below 0.4 µg/g at the study colonies.
Concentrations of p,p’-DDE in embryos were similar to or lower than sum PBDEs
and, based on comparisons of means, were at least eight times higher than
concentrations of the remaining organochlorines. The highest concentrations of sum
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC
11

 PCBs (2.83 µg/g) and p,p’-DDE (0.82 µg/g) in this study were found in an embryo
from Double Island in 2014. The maximum sum PBDE concentration (4.84 µg/g) was
also found in an embryo from Double Island with the next highest concentration
(1.48 µg/g) in an embryo from Mutton Island in 2014. Significant differences in
concentrations of two organochlorines were found in herring gull embryos among
the three colonies in 2012 and for several more compounds, including PCBs, in
2014. Noteworthy is that for compounds where this was evident, mean
concentrations in embryos from the Double Island reference colony were
significantly higher than concentrations in either one or both of the AOC colonies.
Mean percent lipid content was not significantly different among colonies within a
study year and ranged from 7.2% in embryos from Mutton Island in 2014 to 8.7% in
embryos from Double Island in 2014. Organochlorine and sum PBDE burdens in
herring gull embryos from AOC colonies were overall comparable to or lower than
burdens in embryos from the Double Island reference colony.
Concentrations of organochlorines and sum PBDEs in double-crested cormorant
eggs were also consistently lower in pooled samples from the AOC colony on
Welcome Islands-a compared to the Black Rock reference colony in the two study
years (Table 2b). Sum PCBs in cormorant eggs were found at the highest
concentrations (1-3 µg/g) followed by p,p’-DDE (1 µg/g), sum PBDEs (0.05 µg/g) and
then the remaining organochlorines. Comparing burdens between the two colonial
waterbird species at Welcome

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC
12

 Table 1. Embryonic viability and incidence of embryonic deformities in artificially incubated herring gull eggs collected
from Thunder Bay AOC colonies (Mutton Island, Welcome Islands-a) in 2012 and 2014 and the reference colony (Double
Island) in 2012 and 2013*.

Mutton I.

AOC

2012
2014

15
15

2
3

13
12

12
10

1
2

Embryoni
c
Viability
(%)
92%
83%

Welcome I.a

AOC

2012
2014

Overall
Double I.

AOC
Ref

Overall

Ref

15
15
60
26
15
41

2
3
10
1
0
1

13
12
50
25
15
40

12
12
46
24
14
38

1
0
4
1
1
2

92%
100%
92%
96%
93%
95%

Colony

AOC/
Ref

Year

Total
No.
Eggs

No.
Infertile
Eggs

No.
Fertile
Eggs

No.
Viable
Eggs

No.
Dead
Eggs

2012
2013

No.
Deformiti
es

Deformiti
es (%)

0
0

0%
0%

0
0
0
0
0
0

0%
0%
0%
0%
0%
0%

*Due to late ice-out conditions and the inability to access the Double Island reference colony during the early-laying period in
2014, the results of the artificial incubation study conducted in 2013 were used.

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC

13

 Table 2. Concentrations of organochlorines and sum PBDEs in embryos of herring gulls (a) and eggs of double-crested
cormorants (b) from Thunder Bay AOC colonies (Mutton Island, Welcome Islands-a) and corresponding reference colonies
(Double Island or Black Rock) in 2012 and 2014 (µg/g, wet weight). Mean concentrations (SD) in herring gulls are based on
analysis of ten individual embryos following incubation in the lab while concentrations in double-crested cormorant eggs are
based on analysis of a single pool consisting of 10 or 13 eggs. Different uppercase letters indicate significant differences
among gull colonies within a study year. Statistical comparisons of sum PCBs and sum PBDEs in 2012 are based on
congeners that were common among colonies, i.e., the sum of 35 PCB congeners and 14 PBDE congeners, respectively.
a) Herring gull embryos:
Colony

AO
C/
Ref

Year

p,p’DDE

Sum
Chlord
ane

Mutton I.

AOC

2012

0.261
(0.076)

0.023
(0.011)

Welcome I.a

AOC

2012

0.227
(0.135)

0.021
(0.012)

Double I.

Ref

2012

0.218
(0.048)

0.025
(0.005)

Mutton I.

AOC

2014

0.258
(0.195)

Welcome I.a

AOC

2014

0.246
(0.129)

Double I.

Ref

2014

0.374
(0.199)

0.027
(0.015)
B
0.030
(0.015)
AB
0.046
(0.016)
A

HCB
0.008
(0.006
)
0.005
(0.002
)
0.008
(0.005
)
0.009
(0.005
)B
0.007
(0.002
)B
0.019
(0.015
)A

Dieldri
n
0.007
(0.005)
0.008
(0.012)
0.007
(0.005)
0.017
(0.019)
AB
0.007
(0.003)
B
0.022
(0.014)
A

HE

Mirex

OCS

0.006
(0.002
)
0.011
(0.012
)
0.006
(0.001
)
0.008
(0.004
)B
0.008
(0.003
)B
0.013
(0.004
)A

0.005
(0.006)
AB
0.003
(0.002)
B
0.010
(0.013)
A

0.0007
(0.0010)
AB
0.0003
(0.0004)
B

HE

0.002
(0.002)A

0.017
(0.026)

0.001
(0.001)B

0.006
(0.004)

0.0009
(0.0002)
B

0.033
(0.048)

0.004
(0.003)A

Mirex

OCS

PCB
1:1a

Sum
PCBsb

NA

0.746
(0.241)

NA

0.745
(0.419)

2.187
(1.177
)
2.730
(1.161
)
2.417
(0.990
)
3.828
(1.800
)

1.131
(0.704)
1.068
(0.457)
AB
0.862
(0.317)
B
1.543
(0.768)
A

Sum
PBDE
sc
0.258
(0.08
4)
0.286
(0.15
2)
0.332
(0.18
5)
0.689
(0.39
4)
0.379
(0.19
4)
0.787
(1.42
9)

b) Double-crested cormorant eggs:
Colony

AO
C/
Ref

Year

p,p’DDE

Sum
Chlord
ane

HCB

Dieldri
n

PCB
1:1a

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC

Sum
PCBsb

Sum
PBDE
sc
14

 Welcome
I.-a
Black Rock
Welcome
I.-a
Black Rock

AOC

2012

1.337

0.010

0.006

0.024

0.007

0.003

0.001

NA

0.921

0.036

Ref

2012

1.540

0.028

0.043

0.042

0.018

0.009

0.004

5.480

3.370

0.114

AOC

2014

1.070

0.014

0.008

0.028

0.008

0.008

0.001

2.320

1.180

0.044

Ref

2014

1.470

0.016

0.009

0.032

0.008

0.006

0.003

3.570

1.470

0.054

Based on 1:1 ratio of Aroclor 1254:1260; NA denotes that data are not available
Based on sum concentrations of 35-62 PCB congeners
c
Based on sum concentration of 15 or 25 PBDE congeners
a

b

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC

15

 Islands-a, concentrations of all organochlorines and sum PBDEs in cormorant eggs
were within the ranges of concentrations in gull embryos in 2012 and 2014 with the
exception of p,p’-DDE where concentrations in pooled cormorant eggs were at least
two times higher than the maximum concentration in gull embryos.
Concentrations of four non-ortho PCBs, 2,3,7,8-TCDD, and total TEQs were lower in
embryos from the Thunder Bay AOC colonies compared to the Double Island
reference colony in 2011/2 and 2014 (Table 3a). Of the four non-ortho PCBs
measured in herring gull embryos from study colonies in both years, concentrations
of PCB-126 > 169 > 81 > 77. Overall, 2,3,7,8-TCDD concentrations were relatively
lower in embryos from the AOC colonies in the two years (range=0.95-2.27 pg/g)
compared to concentrations at the Double Island reference colony which were at
least two times higher. Total TEQ concentrations ranged from 49.44 pg TEQ/g to
53.87 TEQ/g in embryos from AOC colonies and were below concentrations at the
reference colony in the two years. Toxicity associated with non-ortho PCBs, dioxins
and furans, and mono-ortho PCBs contributed (as means) 82%, 10%, and 8%,
respectively, to the mean total TEQ concentration in embryos from AOC colonies in
2012 and 2014. Similar contributions of toxicity associated with non-ortho PCBs,
dioxins and furans, and mono-ortho PCBs to mean total TEQ concentrations were
found for embryos from the reference colony (as means, 79%, 14%, 8%,
respectively). With respect to spatial patterns for TEQ burdens in herring gulls from
colonies throughout the Great Lakes, embryos from the two Thunder Bay AOC
colonies had the lowest mean TEQ concentration (51.89 pg TEQ/g; N=4 samples) in
2012 and 2014 compared to eggs collected annually from Great Lakes colonies from
2012-2014 (Figure 3). Results for artificially incubated herring gull embryos from the
St. Marys River (Ontario) and eggs collected from Spanish Harbour in 2011 and
2012 are also provided for comparison purposes.
The total TEQ concentration in cormorant eggs from the AOC colony in 2014 was
78.64 pg TEQ/g which was nearly one half of that found in pooled eggs from the
Black Rock reference colony (140.37 pg TEQ/g; Table 3b). Toxicity associated with
non-ortho PCBs, dioxins and furans, and mono-ortho PCBs contributed 78%, 16%,
and 6%, respectively, to the mean total TEQ concentration in eggs from the AOC
colony. These relative contributions to total TEQ concentrations were also similar to
those found at the reference colony. Comparing burdens between the two colonial
waterbird species at Welcome Islands-a, concentrations of non-ortho PCBs, 2,3,7,8TCDD and total TEQs were higher in cormorant eggs compared to gulls embryos in
2014.

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC
16

 Table 3. Concentrations of non-ortho PCBs, 2,3,7,8-TCDD, and 2,3,7,8-TCDD toxic equivalents (TEQ) in embryos of
herring gulls (a) and eggs of double-crested cormorants (b) from Thunder Bay AOC colonies (Mutton Island, Welcome
Islands-a) and corresponding reference colonies (Double Island or Black Rock) in 2011/2 and 2014 (pg/g, wet weight).
Each sample is a single pooled sample consisting of 5-15 embryos or 10 eggs. TEQs associated with 4 non-ortho PCBs,
17 dioxins and furans (PCDD/Fs), and 3 mono-ortho PCBs (#105, #118, and #156) and which together comprise total
TEQs are also provided.
a) Herring gulls embryos:

Colony

AOC/
Ref

Mutton I.

AOC

Welcome I.-a

AOC

Double I.

Ref

Mutton I.

AOC

Welcome I.-a

AOC

Double I.

Ref

Yea
r
201
2
201
2
201
1
201
4
201
4
201
4

TEQ –
nonortho
PCBs

TEQ –
PCDD/
Fs

TEQ –
monoortho
PCBs

Total
TEQs

PCB77

PCB81

PCB126

PCB169

2,3,7,
8–
TCDD

37.50

52.50

368.00

89.10

1.44

44.01

5.33

4.53

53.87

17.90

60.40

343.00

67.40

0.95

41.30

3.96

4.18

49.44

36.70

98.90

680.00

121.00

5.62

79.85

13.89

9.73*

103.4
6

12.00

49.60

373.00

71.40

2.27

42.93

6.06

4.50

53.49

9.06

33.20

381.00

105.00

0.97

41.98

4.98

3.79

50.75

64.40

73.10

570.00

120.00

4.93

67.65

11.93

5.41

84.99

* Includes 3 additional mono-ortho PCB congeners, #114, #157, and #189, since these data are available; combined these congeners
contributed less than 10% to the total TEQ – mono-ortho PCB concentration.

b) Double-crested cormorant eggs:

Colony

AOC/
Ref

Welcome I.-a

AOC

Black Rock

Ref

Yea
r
201
4
201

PCB77

PCB81

PCB126

PCB169

2,3,7,
8–
TCDD

TEQ –
nonortho
PCBs

TEQ –
PCDD/
Fs

TEQ –
monoortho
PCBs

Total
TEQs

85.60

134.00

438.00

102.00

3.07

61.58

12.24

4.81

78.64

291.00

146.00

797.00

138.00

6.23

108.99

25.31

6.08

140.3

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC

17

  

 

 

Assessment of Wildlife Reproduction Deformities BUI in the Thunder Bay ADC

18

Figure 3. Mean total TEQ concentrations (SD) in embryos (N=4 samples total) from
Thunder Bay AOC colonies in 2012 and 2014 and in herring gull eggs from Great
Lakes colonies from 2012-2014 (pg/g, wet weight). Data for St. Marys River
(Ontario) are based on artificially incubated herring gull embryos (N=6 samples)
and for Spanish Harbour are based on eggs collected in 2011 and 2012 (N=2 pools).
The contributions of TEQ concentrations associated with mono-ortho PCBs, dioxins
and furans, and non-ortho PCBs to the total TEQ concentration are shown. Means
are arranged in decreasing order from highest to lowest concentrations. Colony
locations are associated with the following lakes/rivers: LH=Lake Huron, LO=Lake
Ontario, LE=Lake Erie, SLR=St. Lawrence River, NR=Niagara River, LS=Lake
Superior, and LM=Lake Michigan.

Mean mercury concentrations (SD) in herring gull embryos in 2012 were statistically
similar among the three study colonies ranging from 0.09 (0.04) µg/g at Mutton
Island to 0.12 (0.07) µg/g wet weight at Double Island (Figure 4a). In 2014, some
spatial differences among colonies were evident with mean mercury concentrations
(SD) ranging from 0.05 (0.03) µg/g at Welcome Islands-a to 0.14 (0.04) µg/g at
Double Island (F2,27=16.58, p=0.00002). Mercury concentrations in embryos from
Double Island were significantly higher than concentrations in embryos from Mutton
Island and Welcome Islands-a, which were not significantly different from each
other. The maximum mercury concentration (0.28 µg/g) was found in an embryo
from Double Island in 2012. Relative to mercury concentrations in herring gull eggs
from other Great Lakes colonies from 2012 to 2014, mean mercury concentrations
in gull embryos from the two AOC colonies were the lowest of 17 herring gull
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 19

 colonies monitored (Figure 5; ranked as means from highest to lowest; ECCC
unpublished). Similar to the TEQ data presented as a broad comparison of Great
Lakes trends, results for herring gull embryos incubated in the lab from the St.
Marys River (Ontario) and for eggs collected from Spanish Harbour in 2011 and
2012 are also included here.
Figure 4. Concentrations of total mercury (µg/g, wet weight) in embryos of herring
gulls (a) and eggs of double-crested cormorants (b) from Thunder Bay AOC colonies
(Mutton Island, Welcome Islands-a) and corresponding reference colonies (Double
Island or Black Rock) in 2012 and 2014. Mean concentrations (SD) in herring gulls
are based on 10 or 15 embryos per colony. Different uppercase letters show
significant differences in mean concentrations among gull colonies within a study
year and are based on statistical analysis of dry weight mercury concentrations.
Concentrations in double-crested cormorant eggs are based on analysis of a single
pool consisting of 10 or 13 eggs.
a) Herring gulls embryos:

b) Double-crested cormorant eggs:

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 20

 Figure 5. Mean mercury concentrations (SD) in embryos from Thunder Bay AOC
colonies in 2012 and 2014 (based on N=4 mean concentrations) and in herring gull
eggs from Great Lakes colonies from 2012-2014 (µg/g, wet weight). Data for St.
Marys River (Ontario) are based on artificially incubated herring gull embryos (N=4
mean concentrations) and for Spanish Harbour are based on eggs collected in 2011
and 2012 (N=2 pools). Codes for lakes and rivers associated with colony locations are
provided in caption for Figure 3.

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 21

 Mercury concentrations in double-crested cormorant eggs (as pooled samples) from
Welcome Islands-a were 0.13 µg/g in 2012 and 0.27 µg/g in 2014 (Figure 4b).
Concentrations of mercury in eggs from Black Rock were more similar between
years and equal to 0.20 µg/g and 0.19 µg/g in 2012 and 2014, respectively. No
consistent trends were apparent between AOC and reference colonies in the two
study years. Comparing burdens between the two colonial waterbird species at
Welcome Islands-a, the concentration of mercury in cormorant eggs was within the
range of concentrations in gull embryos in 2012 but not in 2014 when the
concentration in eggs was over two times higher than the maximum concentration
in gulls embryos.

Temporal Trends of Contaminants in Herring Gulls and
Cormorants:
Concentrations of sum PCBs in herring gulls from Mutton Island declined
dramatically in the five study years when eggs were collected from the Thunder Bay
AOC. Overall, sum PCBs decreased by 85.4% based on an initial concentration of
7.33 µg/g in eggs in 1992 (analyzed as a single pool) and a mean concentration of
1.07 µg/g in embryos in 2014 (Figure 6a). With the exception of OCS, large
decreases of at least 56% were also found for concentrations of other
organochlorines and 2,3,7,8-TCDD between these two years (Supplementary
Information, Table S1). Concentrations of total TEQs also declined
Figure 6. Temporal trends for concentrations of sum PCBs, total TEQs and mercury in
herring gull eggs and embryos from Mutton Island in the Thunder Bay AOC in 1992,
1993, 1996, 2000, 2007, 2012 and 2014 (where data are available). Depending on
the compound analyzed, concentrations are based on analysis of a single pooled
sample of eggs or embryos or as means (SD) of two pools of eggs (1993 and 1996) or
10 individual embryos (2012 and 2014).
a) Sum PCBs

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 22

 b) Total TEQs:

c) Mercury

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 23

 steadily from 196.22 pg TEQ/g in eggs in 2000 to 53.49 pg TEQ/g in embryos in
2014 representing a decrease of 72.7% between these two years (Figure 6b). A
temporal decline in mercury concentrations was not evident in gulls when eggs
were collected from Mutton Island over seven years from 1992 to 2014 and
concentrations varied between years. Mercury concentrations ranged from 0.05
µg/g (wet weight) in embryos in 2014 to 0.15 µg/g in eggs collected in 2000 and
analyzed as a pooled sample (Figure 6c; Supplementary Information, Table S2).
Large decreases in contaminant burdens were also evident in cormorant eggs
collected from Cone Island (just beyond the boundary of the AOC) in 1989 relative
to current concentrations reported in eggs collected from Welcome Islands-a. Sum
PCB burdens in cormorants decreased by 70.8% from an initial concentration of
4.04 µg/g in eggs in 1989 (analyzed as a single pool) to 1.18 µg/g in eggs from
Welcome Islands-a in 2014 (Supplementary Information, Table S1). Similar large
decreases were found for other organochlorines in cormorant eggs between these
two years (range=39%-91%). Concentrations of 2,3,7,8-TCDD also decreased by
74.4% from an initial concentration of 12 pg/g in cormorant eggs in 1989 to 3.07
pg/g in eggs in 2014.

Stable Isotopes:
Significant spatial differences for mean δ 15N values in herring gull embryos were
found among study colonies in 2014 only (H=12.32, p=0.002; Table 4a). As an
indicator of trophic position, mean δ15N values were significantly higher in gulls from
the Double Island reference colony compared to both AOC colonies. A similar spatial
trend was evident in gull embryos in 2012, however, this was not statistically
significant. Mean δ13C values in gull embryos from Welcome Islands-a were
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 24

 significantly more depleted (i.e., more negative) than mean values in embryos from
the other AOC colony in 2012, while values at the Double Island reference colony
were not significantly different from either of the two AOC colonies (H=12.17,
p=0.002). No significant differences for mean δ13C values were found in 2014
among colonies. No significant correlations were found between δ 15N and δ13C
values when herring gull embryos from all colonies and years were grouped
together or when examined separately by colony and year.
Table 4. Values for δ15N and δ13C in embryos of herring gulls (a) and eggs of doublecrested cormorants (b) collected from Thunder Bay AOC colonies (Mutton Island,
Welcome Islands-a) and corresponding reference colonies (Double Island or Black
Rock) in 2012 and 2014. Mean (SD) values in herring gulls are based on 10-15
embryos per colony. Different uppercase letters show significant differences in mean
values among gull colonies within a study year. Values in double-crested cormorant
eggs are based on analysis of a single pool consisting of 10 or 13 eggs.
a) Herring gull embryos:
Colony

AOC/R
ef

Mutton I.

AOC

Welcome I.-a

AOC

Double I.

Ref

δ15N
2012
9.16
(0.35)
9.28
(0.48)
9.47
(1.03)

δ13C

2014
8.76 (0.38)
B
8.74 (0.33)
B
9.95 (1.05)
A

2012
-21.14 (0.34)
A
-21.98 (0.34)
B
-21.43 (0.79)
AB

2014
-21.14
(0.55)
-21.32
(0.54)
-21.60
(0.69)

b) Double-crested cormorant eggs:
Colony
Welcome I.-a
Black Rock

AOC/R
ef
AOC
Ref

δ15N
2012
11.56
13.20

δ13C
2014
13.27
13.26

2012
-24.59
-21.20

2014
-22.31
-22.34

Stable isotope values in cormorant eggs varied more between years at the AOC
colony compared to the Black Rock reference colony in the two study years (Table
4b). Comparing isotopic signatures between the two colonial waterbird species at
Welcome Islands-a, cormorants occupied a higher trophic position with δ 15N values
that were higher than all gull embryos. They also had more depleted carbon values
in eggs in 2012 and 2014 compared to gulls in the two study years.

B) Field Study
Egg Size Parameters:
To assess potential food stress in birds, total clutch volume and intraclutch variation
in egg size were examined in 3-egg clutches in 2012, 2014, and 2015. Mean total
clutch volume (SD) in herring gull eggs from Thunder Bay AOC colonies ranged from
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 25

 231.0 (12.7) cm3 at Welcome Islands-a in 2014 to 251.0 (18.3) cm3 at Mutton Island
in 2012 (Figure 7a). Total clutch volume in gull eggs at the Double Island colony
ranged from 241.0 (18.7) cm3 in 2014 to 251.7 (18.1) cm3 in 2012. Mean total clutch
volumes were not significantly different among clutches from AOC colonies and the
Double Island reference colony in the three study years.
Mean intraclutch variation (SD) in egg size at AOC colonies ranged from 7.9 (4.1)%
at Welcome Islands-a in 2015 to 14.7 (4.3)% at Welcome Islands-a in 2012. In 2012,
mean intraclutch variation was

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 26

 Figure 7. Total clutch volume (a) and intraclutch variation (b) in 3-egg clutches of
herring gulls from Thunder Bay AOC colonies (Mutton Island, Welcome Islands-a)
and the reference colony (Double Island) in 2012, 2014, and 2015. Numbers of 3egg clutches ranged from 8-42 at each colony. Different uppercase letters show
significant differences in mean estimates within a study year.
a) Mean total clutch volume (SD):

b) Mean intraclutch variation (SD) in egg size:

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 27

 significantly different among colonies (F2,51=13.66, p=0.00002; Figure 7b) and was
significantly higher between clutches from Welcome Islands-a compared to Double
Island with clutches from Mutton Island not significantly different from either of the
two colonies. No significant differences were found for this parameter among study
sites in either 2014 or 2015.

Productivity & Prevalence of Deformities in Wild Populations:
In 2012, herring gull productivity, defined as the number of > 21-day-old chicks
produced per nest, was equal to 1.9 chicks per nest at the two Thunder Bay AOC
colonies (N=17 enclosed nests, combined) and 2.0 chicks per nest at Double Island
(N=7 nests; Figure 8). Productivity was consistently high at AOC colonies in 2014
and 2015 and equal to 2.0 chicks per nest (N=5 enclosed nests) and 1.8 chicks per
nest (N=8 enclosed nests), respectively. Productivity at the reference colony was
slightly lower at 1.3 chicks per nest in 2014 and 1.7 chicks per nest in 2015 (N=8
and 10 enclosed nests, respectively). Overall, productivity estimates at AOC
colonies in the three study years were well within the range of productivity levels
required to maintain a stable population (0.8-1.4 chicks per nest; Kadlec and Drury
1968). In addition, many herring gull chicks were observed at the colony across the
channel at Welcome Islands-b in 2015. A formal survey was not conducted on this
island due to concerns that disturbance might impact a colony of American white
pelicans (Pelecanus erythrorhynchos) that were also observed nesting there.
Figure 8. Herring gull productivity, as the number of > 21-day-old chicks produced
per nest, at Thunder Bay AOC colonies (Mutton Island, Welcome Islands-a,
combined) and the reference colony (Double Island) in 2012, 2014 and 2015. The
solid line indicates the minimum productivity level of 0.8 chicks per nest associated
with maintaining a stable herring gull population (range in levels=0.8-1.4 chicks per
nest; Kadlec and Drury 1968).

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 28

 In 2012, 2014 and 2015, no morphological deformities were found in herring gull
chicks from enclosed nests at the two AOC colonies (combined) or the reference
colony (N=10-33 chicks; Table 5).
Table 5. Prevalence of morphological deformities (%) in herring gull chicks examined
in enclosed nests from Thunder Bay AOC colonies (Mutton Island & Welcome
Islands-a, combined) and the reference colony (Double Island) in 2012, 2014 and
2015.
Colony

Thunder
Bay AOC

Double I.

Yea
r
201
2
201
4
201
5
201
2
201
4
201
5

No. Chicks
Examined

%
Deformities

33

0%

10

0%

14

0%

14

0%

10

0%

17

0%

Corticosterone in Feathers of Herring Gull Chicks:

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 29

 Mean corticosterone concentrations (SD) in feathers of herring gull chicks from
Thunder Bay AOC colonies (combined) ranged from 0.86 (0.65) ρg/mm to 1.69
(0.11) ρg/mm in 2012, 2014 and 2015 (Figure 9). Mean corticosterone
concentrations (SD) in feathers from the Double Island reference colony ranged
from 2.01 (2.44) ρg/mm to 2.77 (4.11) ρg/mm in the three study years. Significant
spatial differences were found between the Thunder Bay AOC colonies and the
reference colony in two of three study years in which mean corticosterone
concentrations were significantly higher at the reference colony compared to the
AOC colonies (t>2.10, p<0.05). Relative to AOC chicks, these results suggest that
chicks from the reference colony experienced higher levels of stress during the
period of feather growth.

Thyroxine in Plasma of Herring Gull Chicks:
Mean thyroxine concentrations (SD) in plasma of herring gull chicks from Thunder
Bay AOC colonies (combined) ranged from 0.24 (0.07) ng/dl to 0.69 (0.29) ng/dl in
2012, 2014 and 2015 (Figure 10). Mean thyroxine concentrations (SD) in plasma of
chicks from Double Island were comparatively lower and ranged from 0.19 (0.07)
ng/dl to 0.59 (0.39) ng/dl in the three study years. Mean thyroxine concentrations
were significantly higher in plasma of chicks from AOC colonies compared to the
reference colony in 2012 (Mann Whitney test U=98.50, p=0.014). No significant
differences in thyroxine concentrations were found between sites in 2014 and 2015.

Figure 9. Mean corticosterone concentrations (SD), expressed as picograms of
corticosterone per millimetre of feather, in herring gull chicks from Thunder Bay
AOC colonies (Mutton Island, Welcome Islands-a, combined) and the reference
colony (Double Island) in 2012, 2014 and 2015. Numbers of chicks sampled ranged
from 4-29 per site. Different uppercase letters show significant differences in mean
concentrations within a study year.

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 30

 Figure 10. Mean thyroxine concentrations (SD) in plasma of herring gull chicks from
Thunder Bay AOC colonies (Mutton Island, Welcome Islands-a, combined) and the
reference colony (Double Island) in 2012, 2014 and 2015 (ng/dl). Numbers of chicks
sampled ranged from 5-29 per site. Different uppercase letters show significant
differences in mean concentrations within a study year.

Immune Response in Herring Gull Chicks:
Estimates of the PHA stimulation index as means (SD) were 0.28 (0.15) mm and
0.36 (0.12) mm for chicks from the Thunder Bay AOC colonies (combined) and
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 31

 Double Island reference colony, respectively, in 2015 (Figure 11). There was no
significant difference in PHA response in herring gull chicks between the two sites.
Figure 11. Mean PHA stimulation index (SD) in herring gull chicks from Thunder Bay
AOC colonies (Mutton Island, Welcome Islands-a, combined) and the reference
colony (Double Island) in 2015. Fourteen chicks were tested at each site.

DISCUSSION
Using a multi-tiered lab and field study approach, assessments of reproduction and
deformities in herring gulls nesting in the Thunder Bay AOC were conducted in three
study years. Artificial incubation of freshly-laid eggs under controlled conditions in
the laboratory were valuable for assessing the importance of intrinsic factors such
as contaminants that may induce early embryonic mortality or result in
developmental abnormalities at this critical life stage. In this study, embryonic
viability in artificially incubated eggs from AOC colonies was consistently high in
herring gulls and on average equal to 92% in 2012 and 2014. No embryonic
deformities (0%) were evident in incubated eggs from AOC colonies in these two
years. Egg viability in herring gulls from Mutton Island in 2007, based on field
assessments, was also high and equal to 94.1% on average in three egg clutches
(Supplementary Information, Table S3; ECCC unpublished).
Reproduction based on productivity values for herring gulls at two nesting colonies
in the Thunder Bay AOC was considered good with productivity ranging from 1.8 to
2.0 chicks per nest in the three study years. These values were well above
productivity levels required to maintain a stable herring gull population (range=0.81.4 chicks per nest; Kadlec and Drury 1968). These results are also consistent with
productivity estimates for herring gulls nesting at Welcome Islands-a and b
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 32

 determined in earlier research studies conducted in 2000 and 2008 (ECCC
unpublished). Herring gull productivity at Welcome Islands-b in 2000 was estimated
to be 1.1 chicks per nest based on total counts of 237 chicks at 216 nests. This
value likely represents an underestimation since chicks can easily hide on the island
or swim off the island, thus decreasing the census numbers. While no herring gulls
nested on Welcome Islands-a in 2000, they nested on this island in 2008 and
productivity was estimated to be 1.8 chicks per nest (21 chicks from 12 nests in a
single large enclosure). Double-crested cormorant productivity was also estimated
on Welcome Islands-a and b in 2000 (ECCC unpublished). Productivity was equal to
1.4 chicks per nest at the two colonies with 187 chicks counted at 131 nests at
Welcome Islands-a and 584 chicks counted at 421 nests at Welcome Islands-b.
In this study, no morphological deformities were found in juvenile herring gull chicks
from both AOC and reference colonies in 2012, 2014 and 2015. Similarly, no
morphological deformities were found in gull and cormorant chicks examined as
part of earlier research studies at Welcome Island-a and b in 2000 and 2008 when
productivity was also assessed (as described above; ECCC unpublished). Two largescale surveys of bill deformities in cormorant chicks were conducted on the Great
Lakes by the Canadian Wildlife Service in 1979-1987 and 1988-1996 (Fox et al.
1991; Ryckman et al. 1998). Cormorant colonies assessed in both studies were
largely located in the region encompassing the three large bays, including Thunder
Bay, on the northwest shore of Lake Superior. Two chicks with bill deformities were
found in total in these two surveys. Based on these results, the prevalence of bill
defects in cormorant chicks from Lake Superior was 5.5 per 10,000 chicks at five
colonies in 1979-1987 and 1.2 per 10,000 chicks at 18 colonies in 1988-1996. While
these rates were not significantly different between the two time periods, the
prevalence of chicks with bill deformities at Lake Superior colonies was relatively
lower in the latter time period and was also the lowest of all of the Great Lakes
where surveys were conducted (Ryckman et al. 1998).
Based on the low overall contaminant burdens reported in herring gull embryos and
double-crested cormorant eggs in 2012 and 2014, concentrations of PCBs, other
organochlorines and PBDEs were not sufficiently elevated to adversely impact the
reproductive success of these species foraging in the Thunder Bay AOC. In a broad
literature review of PCB effects in birds, Hoffman et al. (1996) concluded that sum
PCB concentrations in the range of 8 to 25 μg/g in eggs were associated with
decreased hatching success for terns and cormorants. Sum PCB concentrations in
embryos and eggs of both species were well below the 8 μg/g threshold. Similarly,
concentrations of p,p’-DDE in embryos and eggs were well below threshold levels
associated with significant effects on reproductive success as reported in blackcrowned night-herons (Nycticorax nycticorax; 8 μg/g; Henny et al. 1984) and
cormorants (10 μg/g; Pearce et al. 1979). Sum PBDE concentrations were generally
low in gulls from AOC colonies and below the lowest-observed effect level on
pipping and hatching success in American kestrels (Falco sparverius) equal to 1.8
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 33

 μg/g in eggs (McKernan et al. 2009). This threshold was exceeded in one gull
embryo from the Double Island reference colony in 2014. Overall, concentrations of
compounds in both species were largely comparable between AOC colonies and the
reference colonies and for some compounds, e.g., PCBs and sum chlordane, were on
occasion significantly lower in gulls at an AOC
colony than the respective reference colony in a study year.
Concentrations of 2,3,7,8-TCDD, dioxin-like PCBs and total TEQs in herring gull
embryos and cormorant eggs from Thunder Bay AOC colonies were not notably
elevated relative to birds from the respective reference colonies in 2011/2 and
2014. Total TEQs in herring gulls from the Thunder Bay AOC were the lowest of
mean TEQ concentrations in herring gull eggs from other Great Lakes colonies,
including those elsewhere on Lake Superior, from 2012-2014 (Figure 3). Mean
concentrations of total TEQs and sum PCBs in gull embryos from Thunder Bay AOC
colonies were also lower than median concentrations in herring gull eggs from three
Lake Superior colonies in Michigan, one of which included a colony near Isle Royale,
from 2008-2012 (Fuentes et al. 2014). Large declines in concentrations of sum PCBs
and other organochlorines, 2,3,7,8-TCDD and total TEQs in herring gull tissues since
the early 1990s indicate that exposure to these compounds has decreased in
herring gulls foraging in the AOC (Figure 6; Supplementary Information, Table S1).
These trends are consistent with larger scale declines reported for organochlorines
and 2,3,7,8-TCDD in herring gull eggs collected from two Lake Superior colonies,
Granite Island and Agawa Rocks, that have been monitored annually since the early
1970s (de Solla et al. 2016).
With respect to concentrations associated with toxicity, 2,3,7,8-TCDD (<4 pg/g) and
total TEQs (<80 pg TEQ/g) in herring gull embryos and cormorant eggs from AOC
colonies were below concentrations associated with effects on reproduction in avian
species. Median concentrations of 2,3,7,8-TCDD (37 pg/g) and total TEQs (2175 pg
TEQ/g) in eggs of Forster’s tern (Sterna forsteri) from Lake Michigan were associated
with a significant reduction in hatching success (TEQs based on concentrations of
2,3,7,8-TCDD and dioxin-like PCBs only; Kubiak et al. 1989). However, no effect on
hatching success was observed at relatively lower median concentrations of 2,3,7,8TCDD (8 pg/g) and total TEQs (201 pg TEQ/g) in eggs from the reference colony in
1983 (Kubiak et al. 1989). Total TEQ concentrations in gull embryos and cormorant
eggs were also well below 1200-1300 pg TEQ/g at which high rates (6-10%) of
embryonic deformities were found in cormorants at two Lake Michigan colonies in
1988 (Yamashita et al. 1993). Concentrations of 2,3,7,8-TCDD were three orders of
magnitude below the lethal dose associated with 50% embryonic mortality in
cormorants (4000 pg/g based on egg injection; Powell et al. 1998) and were well
below those associated with decreased embryonic growth and edema in herons
(150-250 pg/g; Hoffman et al. 1996). As effectively demonstrated in the artificial
incubation component of this study, these compounds were not sufficiently elevated
to impact embryonic viability in gull eggs from study colonies in the two years.
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 34

 Exposure to high concentrations of mercury can have significant impacts on
reproductive success in birds and result in teratogenic effects in avian embryos, as
demonstrated in egg injection studies with methylmercury in the laboratory
(Fimreite 1974; Hill et al. 2008; Heinz et al. 2011). Overall, mercury concentrations
in all embryos were below the predicted threshold of 0.6 μg/g (wet weight) in eggs
set to be protective against adverse reproductive effects for 95% of non-marine
avian species (Shore et al. 2011). The highest mercury concentrations were
approximately one-half of the threshold concentration and were found in a herring
gull embryo from Double Island in 2012 and cormorant eggs (pooled sample) from
Welcome Islands-a in 2014. Mercury burdens in gull embryos were not elevated at
AOC compared to the references colonies and based on mean concentrations were
the lowest of all Great Lakes colonies where gull eggs were collected from 20122014 (Figure 5). Mean mercury concentrations in embryos from AOC colonies were
also lower, i.e., one quarter, of median mercury concentrations in herring gull eggs
from three Lake Superior colonies in Michigan from 2008-2012 (Fuentes et al. 2014).
Low concentrations of mercury in gull embryos from AOC colonies are consistent
with the normal embryonic development of artificially incubated eggs in the lab.
While large and significant declines in mercury were found in herring gull eggs from
the two Lake Superior colonies, Agawa Rocks and Granite Island, from 1974-2009,
no significant changes in mercury concentrations were found at these colonies from
1994-2009 (Weseloh et al. 2011). This is similar to the temporal pattern found for
mercury concentrations in eggs from the AOC during a comparable time span
(Figure 6; Supplementary Information, Table S2). Overall, current mercury burdens
were low and it is unlikely that mercury concentrations would impact reproduction
of breeding colonial waterbirds in the Thunder Bay AOC.
Long-term population trends of nesting herring gulls and cormorants at AOC study
colonies provide additional evidence that populations are not adversely impacted by
conditions that are specific to the AOC. Using nest count surveys as an indicator of
population trends of breeding herring gulls, total numbers of gull nests at Mutton
Island and Welcome Islands-a and b increased from 450 nests in 1978 to 519 nests
in 1999 to 868 nests in 2007 and then decreased to 757 nests in 2012
(Supplementary Information, Table S4). While gull nest numbers have not changed
remarkably at each of the AOC colonies over 11 survey years (where survey data
are available), total numbers of nests at the three AOC colonies were overall higher
in 2007 and 2012 relative to nest numbers in 1978, 1989 and 1999. In contrast,
numbers of cormorant nests at Welcome Islands-a and b increased dramatically
from a single nest in 1978 to 1,574 nests in 2007. Lake-wide population trends
based on decadal surveys conducted on the Canadian and American sides of Lake
Superior indicate that nest numbers of herring gulls have fluctuated while
cormorant nest numbers have increased steadily since the late 1970s
(Supplementary Information, Table S4). Numbers of herring gull nests on Lake
Superior were 13,269 in the first decadal survey and peaked to nearly 25,000 nests
in the second survey. Nest numbers subsequently decreased to 18,987 in the third
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 35

 survey (1996-2001), and decreased again by nearly 20% to 15,255 nests in the
fourth survey (2007-2010). Cormorant nests have increased steadily over the four
surveys from 35 nests in 1978 to nearly 4,800 nests in 2007-2010 on Lake Superior.
Over the four binational Great Lakes colonial waterbird decadal surveys, eight
colonial waterbird species, including great blue heron (Ardea herodias) and common
tern (Sterna hirundo), have been reported nesting on Lake Superior. Only two
species, herring gulls and cormorants, have been found nesting within the Thunder
Bay AOC since 1978 and both have been consistently recorded in years when
surveys were conducted. A new colonizer to this area, the American white pelican,
was observed breeding on Welcome Islands-b in 2015.
Stable isotopes of nitrogen and carbon are used to provide information on trophic
position and carbon source in the food web, respectively (Hobson 1999).
Significantly higher δ15N values were found in gull embryos at the Double Island
reference colony compared to AOC colonies in 2014 and a similar, albeit nonsignificant, trend was found in 2012. This suggests that gulls occupied a relatively
higher trophic level at the reference colony compared to gulls from AOC study
colonies. Specifically, gulls at this colony may have fed more on fish (or larger fish)
compared to gulls from AOC colonies, which fed at a relatively lower trophic level
and on a diet that may have included terrestrial food sources such as small
mammals, refuse and plant material (Fox et al. 1990). Differences in trophic levels
between colonies may have contributed to higher concentrations of several
organochlorines reported in gull embryos from the reference colony compared to
AOC colonies. While a significant difference was found for δ13C values in gull
embryos between the two AOC colonies in 2012, no significant differences were
found between either of these and the reference colony in the two study years. As a
strict piscivore, the double-crested cormorant occupied a relatively higher trophic
level, which may have contributed to differences in burdens of TEQs between
species while concentrations of most other contaminants were largely similar
between species.
Determinations of total clutch volume and intraclutch variation in 3-egg clutches
can be used as an indicator of potential food stress for laying females during the
egg production period. Total clutch volume did not differ significantly among the
three herring gull colonies in all three study years indicating that food availability
was likely not limited during this critical time. While mean intraclutch variation in
egg size was statistically comparable among colonies in two of three years, mean
intraclutch variation in egg size was significantly higher at Welcome Islands-a
compared to Double Island in 2012. This variation in egg size at Welcome Islands-a
however did not appear to impact chick survival, i.e., productivity values were
normal. In 2001 and 2002, similar measurements of 3-egg clutches at Mutton Island
were taken to determine total clutch volume and intraclutch variation in egg size
(Supplementary Information, Table S5 and Table S6; ECCC unpublished). Egg size
parameters at Mutton Island were comparable between 2001 and 2002 and a
decade later in 2012, 2014 and 2015. Over a relatively longer time period from
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 36

 1981 to 2014, significant declines in egg volume were found in herring gulls from
colonies on the upper Great Lakes including those on Lake Superior (Hebert et al. in
prep.). This finding has been associated with temporal changes in gull diet that may
be related to reductions in prey fish availability.
Three additional endpoints relating to growth, development, and immune function
of chicks were also measured in this study. Corticosterone deposited in growing
feathers provides important insight into the physiology of stress during the period of
feather growth (Bortolotti et al. 2009). Unexpectedly, feathers from gull chicks at
the reference colony had significantly higher mean concentrations of corticosterone
compared to chicks from AOC colonies in two of three study years. Possible
stressors at the reference colony could include low food availability, inclement
weather, and/or threats of predation. Depressed thyroxine concentrations in
reference colony chicks (albeit significant in 2012 only) could have implications in
terms of body condition and growth and warrants further investigation. Immune
function assessed using the PHA skin response test was suppressed in herring gull
chicks that had high levels of PCBs, 2,3,7,8-TCDD, and total TEQs (Grasman et al.
2013). In this study, there was no evidence of suppressed immune function in chicks
from AOC colonies in 2015. Results for all three of these endpoints suggest that no
adverse health effects were found in AOC chicks compared to reference colony
chicks. Early studies examining corticosterone in plasma of gull chicks from
Welcome Islands-b in 2000 and 2008 indicated no significant differences in stress
response in gulls from the AOC colony compared to the reference colony (ECCC
unpublished); further details of these studies are provided in the Supplementary
Information, Figure S1.
In conclusion, the results of this three-year study suggest that there is no evidence
of impaired reproduction or deformities in herring gulls attributable to local
contamination effects within the Thunder Bay AOC. Embryonic viability of herring
gulls was high at AOC colonies in 2012 and 2014 and productivity was sufficiently
high to maintain a stable population in 2012, 2014 and 2015. No embryonic
deformities were evident in gull eggs incubated in the laboratory and no
morphological deformities were found in > 21-day-old herring gull chicks from AOC
colonies. Importantly, contaminant burdens (e.g., PCBs, 2,3,7,8-TCDD, and mercury)
in gull embryos and double-crested cormorant eggs from the Thunder Bay AOC were
not notably elevated and were comparable to or lower than burdens in embryos or
eggs from respective reference colonies in these years. Population trends based on
nest count surveys indicate greater numbers of nests for both species at AOC study
colonies in 2007 and 2012 compared to earlier decades. Large declines in
concentrations of sum PCBs and other organochlorines, 2,3,7,8-TCDD, and total
TEQs in herring gull tissues since the early 1990s indicate that exposure to these
compounds has decreased in herring gulls foraging in the AOC. While no change in
mercury levels was found in herring gulls in the AOC, this temporal pattern was
similar to that found at two other Lake Superior herring gull colonies during this
period. The results of early studies conducted by CWS in the AOC in the 2000s
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 37

 support many of the findings presented in this current study. In summary,
concentrations of PCBs, other organochlorine compounds, PBDEs, dioxins and
furans, and mercury were not sufficiently elevated to adversely impact the
reproductive success and development of herring gulls and cormorants nesting in
the Thunder Bay AOC.

ACKNOWLEDGEMENTS
Our thanks to Chris Allen and Ryan Mutz of Zimmer Air Services for helicopter
transport and fieldwork assistance. We also thank Sarah DaSilva (Environment and
Climate Change Canada) and Aaron Nicholson (former Thunder Bay RAP
coordinator) for their assistance with egg collections and field surveys. We also
thank the Thunder Bay AOC public advisory committee and specifically, Jim Bailey
(current RAP coordinator) and Brian Ratcliffe (independent contractor), for providing
local weather/ice conditions in Thunder Bay. We would like to thank Lab Services at
NWRC and the Great Lakes Institute for Environmental Research for tissue
preparations and chemical analyses. Finally, we acknowledge Dr. Laird Shutt and Dr.
Craig Hebert (ECCC) for their contributions to colonial waterbird health assessments
in the Thunder Bay AOC in 2000 and 2007/8.

REFERENCES
Blokpoel, H. and G.D. Tessier. 1993. Atlas of colonial waterbirds nesting on the
Canadian Great Lakes, 1989-1991. Part 1. Cormorants, gulls and island-nesting
terns on Lake Superior in 1989. Canadian Wildlife Service Technical Report Series
No. 181. Canadian Wildlife Service, Environmental Conservation Branch, Ontario
Region. 104 pp.
Blokpoel, H. and G.D. Tessier. 1998. Atlas of colonial waterbirds nesting on the
Canadian Great Lakes, 1989-1991. Part 5. Herons and egrets in 1991. Canadian
Wildlife Service Technical Report Series No. 272. Canadian Wildlife Service,
Environmental Conservation Branch, Ontario Region. 36 pp.
Blokpoel, H., J.P. Ryder, I. Seddon, and W.R. Carswell. 1980. Colonial waterbirds
nesting in Canadian Lake Superior in 1978. Canadian Wildlife Service Progress Notes
118. 13 pp.
Bortolotti, G.R., T. Marchant, J. Blas, and S. Cabezas. 2009. Tracking stress:
localisation, deposition and stability of corticosterone in feathers. J. Exper. Biol. 212:
1477-1482.
Cuthbert, F. J., J. McKearnan, L.R. Wires and A.R. Joshi. 2001. Distribution and
abundance of colonial waterbirds in the U.S. Great Lakes, 1997-1999. Final report to
U.S. Fish and Wildlife Service, Ft. Snelling, MN.
De Solla, S.R., D.V. Weseloh, K.D. Hughes, and D.J. Moore. 2016. 40 year decline of
organic contaminants in eggs of herring gulls (Larus argentatus) from the Great
Lakes, 1974 to 2013. Waterbirds 39 (Special Publication 1): 166-179.

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 38

 Environment Canada and the Ontario Ministry of the Environment. 2011. Thunder
Bay Area of Concern Status of Beneficial Use Impairments. September 2010. 8 pp.
Fimreite, N. 1974. Mercury contamination in aquatic birds in northwestern Ontario. J.
Wildl. Manage. 38: 120-131.
Fox, G.A. and D.V. Weseloh. 1987. Colonial waterbirds as bio-indicators of
environmental contamination in the Great Lakes. In: The value of birds, eds. A.W.
Diamond and F.L. Fillion. International Council for Bird Conservation Technical
Publication No. 6, Cambridge, U.K.
Fox, G.A. L.J. Allan, D.V. Weseloh, and P. Mineau. 1990. The diet of herring gulls
during the nesting period in Canadian waters of the Great Lakes. Can. J. Zoology 68:
1075-1085.
Fox, G.A., B. Collins, E. Hayakawa, D.V. Weseloh, J.P. Ludwig, T.J. Kubiak, and T.C.
Erdman. 1991. Reproductive outcomes in colonial fish-eating birds: a biomarker for
developmental toxicants in Great Lakes food chains. II. Spatial variation in the
occurrence and prevalence of bill defects in young double-crested cormorants in the
Great Lakes, 1979-1987. J. Great Lakes Res. 17: 158-167.
Fuentes, L., W. Bowerman, L. Moore, K. Leith, and K. Drouillard. 2014.
Concentrations of environmental contaminants in herring gull eggs from Great
Lakes colonies in Michigan 2002-2006 and 2008-2012. Report to Michigan
Department of Environmental Quality, Lansing, Michigan. MI/DEQ/WRD-14-028. 49
pp.
Grasman, K.A., K.R. Echols, T.M. May, P.H. Peterman, R.W. Gale, and C.E. Orazio.
2013. Immunological and reproductive health assessment in herring gulls and blackcrowned night herons in the Hudson-Raritan estuary. Environ. Toxicol. Chem. 32:
548-561.
Heinz, G.H., D.J. Hoffman, J.D. Klimstra, K.R. Stebbins, S.L. Kondrad, and C.A. Erwin.
2011. Teratogenic effects of injected methylmercury on avian embryos. Environ.
Toxicol. Chem. 30(7): 1593-1598.
Henny, C.J., L.J. Blus, A.J. Krynitsky, and C.M. Bunck. 1984. Current impact of DDE on
black-crowned night-herons in the intermountain west. J. Wildl. Manage. 48(1): 1-13.
Hill, E.F., C.J. Henny, and R.A. Grove. 2008. Mercury and drought along the lower
Carson River, Nevada: II. Snowy egret and black-crowned night-heron reproduction
on Lahontan Reservoir, 1997-2006. Ecotoxicology 17: 117-131.
Hobson, K.A. 1999. Tracing origins and migration of wildlife using stable isotopes: a
review. Oecologia 120: 314-326.
Hoffman, D. J., C.P. Rice, and T.J. Kubiak. 1996. PCBs and dioxins in birds. In:
Environmental contaminants in wildlife: interpreting tissue concentrations, eds. W.N.
Beyer, G.H. Heinz and A.W Redmon-Norwood, pp. 165-207. SETAC Special
Publications, Lewis Publishers, Boca Raton, FL.
Kadlec, J.A. and W.H. Drury. 1968. Structure of the New England herring gull
population. Ecology 49: 644-676.
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 39

 Kubiak, T.J., H.J. Harris, L.M. Smith, T.R. Schwartz, D.L. Stalling, J.A. Trick, L. Sileo,
D.E. Docherty, and T.C. Erdman. 1989. Microcontaminants and reproductive
impairment of Forster’s tern on Green Bay, Lake Michigan – 1983. Arch. Environ.
Contam. Toxicol. 18: 706-727.
Mayne, G. J., P.A. Martin, C.A. Bishop, and H.J. Boermans. 2004. Stress and immune
responses of nestling tree swallows (Tachycineta bicolor) and eastern bluebirds
(Sialia sialis) exposed to nonpersistent pesticides and p,p′dichlorodiphenyldichloroethylene in apple orchards of southern Ontario, Canada.
Environ. Toxicol. Chem. 23: 2930-2940.
McKernan, M.A., B.A. Rattner, R.C. Hale, and M.A. Ottinger. 2009. Toxicity of
polybrominated diphenyl ethers (DE-71) in chicken (Gallus gallus), mallard (Anas
platyrhynchos), and American kestrel (Falco sparverius) embryos and hatchlings.
Environ. Toxicol. Chem. 28: 1007-1017.
Pearce, P.A., D.B. Peakall, and L.M. Reynolds. 1979. Shell thinning and residues of
organochlorines and mercury in seabird eggs, eastern Canada, 1970-1976. Pest.
Monit. J. 13(2): 61-67.
Powell, D.C., R.J. Aulerich, J.C. Meadows, D.E. Tillitt, M.E. Kelly, K.L. Stromberg, M.J.
Melancon, S.D. Fitzgerald, and S.J. Bursian. 1998. Effects of 3,3’,4,4’,5pentachlorobiphenyl and 2,3,7,8-tetrachlorodibenzo-p-dioxin injected into the yolks
of double-crested cormorants (Phalacrocorax auritus) eggs prior to incubation.
Environ. Toxicol. Chem. 17: 2035-2040.
Ryckman, D.P., D.V. Weseloh, P. Hamr, G.A. Fox, B. Collins, P.J. Ewins, and R.J.
Norstrom. 1998. Spatial and temporal trends in organochlorine contamination and
bill deformities in double-crested cormorants (Phalacrocorax auritus) from the
Canadian Great Lakes. Environ. Monitor. Assess. 53: 169-195.
Scharf, W.C. 1978. Colonial birds nesting on manmade and natural sites in the U.S.
Great Lakes. Tech Report D-78-10, U.S. Army Engineer Waterways Experiment
Station, Vicksburg, MS. 136p + Append.A-D.
Scharf, W.C. 1998. Distribution and abundance of tree-nesting heron and marshnesting tern colonies of the U.S. Great Lakes, 1991. W.W. Bowerman and A.S. Roe
(eds.). Publication No. 2, Gale Gleason Environmental Institute, Lake Superior State
University Press, Sault Ste. Marie, MI.
Scharf, W.C. and G.W. Shugart 1998. Distribution and abundance of gull, tern and
cormorant nesting colonies of the U.S. Great Lakes, 1989 and 1990. Gale Gleason
Environmental Institute, Publication No. 1, Lake Superior State University Press,
Sault Ste. Marie, MI.
Shore, R.F., M.G. Pereira, L.A. Walker, and D.R. Thompson. 2011. Mercury in
nonmarine birds and mammals. In: Environmental contaminants in biota (2nd
edition), eds. W.N. Beyer and J.P. Meador, pp. 609-624. CRC Press, Boca Raton, FL.
Thunder Bay RAP Team. 1991. Thunder Bay Area of Concern Remedial Action Plan
Stage 1: environmental conditions and problem definition. 109 pp.

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 40

 Thunder Bay RAP Team. 2004. Thunder Bay Remedial Action Plan Stage 2: Remedial
strategies for ecosystem restoration. 74 pp. + app.
Van den Berg, M., L. Birnbaum, A.T.C. Bosveld, B. Brunström, P. Cook, M. Feeley, J.P.
Giesy, A. Hanberg, R. Hasegawa, S.W. Kennedy, T. Kubiak, J.C. Larsen, F.X.R. van
Leeuwen, A.K. Dijen Liem, C. Nolt, R.E. Peterson, L. Poellinger, S. Safe, D. Schrenk,
D. Tillitt, M. Tysklind, M. Younes, F. Waern, and T. Zacharewski. 1998. Toxic
equivalency factors (TEFs) for PCBs, PCDDs, PCDFs for humans and wildlife. Environ.
Health Perspect. 106: 775-792.
Weseloh, D.V., P.J. Ewins, J. Struger, P. Mineau, C.A. Bishop, S. Postupalsky, and J.P.
Ludwig. 1995. Double-crested cormorants of the Great Lakes: changes in population
size, breeding distribution and reproductive output between 1913 and 1991.
Colonial Waterbirds 18: 48-59.
Weseloh, D.V.C., D.J. Moore, C.E. Hebert, S.R. de Solla, B.M. Braune and D.
McGoldrick. 2011. Current concentrations and spatial and temporal trends in
mercury in Great Lakes herring gull eggs, 1974-2009. Ecotoxicology 20: 1644-1658.
Yamashita, N., Tanabe, S., J.P. Ludwig, H. Kurita, M.E. Ludwig, and R. Tatsukawa.
1993. Embryonic abnormalities and organochlorine contamination in double-crested
cormorants (Phalacrococorax auritus) and Caspian terns (Hydroprogne caspia) from
the upper Great Lakes in 1988. Environ. Pollut. 79: 163-173.

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 41

 Supplementary Information

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC 42

 i) Temporal Trends of Contaminants in Eggs of Herring Gulls and Cormorants in/near the Thunder
Bay AOC:
Table S1. Temporal trends for concentrations of organochlorine pesticides, PCB 1:1, sum PCBs, sum PBDEs (µg/g)
and 2,3,7,8-TCDD and 2,3,7,8-TCDF (ρg/g, wet weights) in herring gull eggs (a) from Mutton Island in the Thunder
Bay AOC in 1992, 2000, and 2007 (where data are available). Concentrations of organochlorines, 2,3,7,8-TCDD and
2,3,7,8-TCDF in eggs of double-crested cormorants (b) from Cone Island, just outside of the AOC boundary, in 1989
are also shown. Samples are based on analysis of a single pooled sample consisting of 9 or 10 eggs. “-“ indicates
that data are not available. Concentrations in gull embryos from Mutton Island (as a mean) and cormorant eggs
from Welcome I.-a (as a single pool) in 2012 and 2014 are also shown for reference.
a) Herring gull eggs/embryos:

Year
199
2
200
0
200
7
201
2
201
4

Colony

p,p’DDE

Sum
Chlorda
ne

Mutton I.

3.815

0.324

0.040

0.157

0.075

0.039

Mutton I.

1.417

0.106

0.011

0.031

0.032

0.036

Mutton
I.***

0.659

0.052

0.009

0.020

0.012

0.080

Mutton I.

0.261

0.023

0.008

0.007

0.006

0.005

Mutton I.

0.258

0.027

0.009

0.017

0.008

0.017

HE

Mirex

HCB

Dieldr
in

HE

Mirex

2,3,7
,8
–
TCD
D

2,3,7
,8
–
TCD
F

PCB
1:1

Sum
PCBs*

Sum
PBDE
s**

14.380

7.334

NA

8.20

2.5

6.688

3.165

0.455

5.40

0.75

4.137

2.204

0.567

2.33

I

NA

0.746

0.258

1.44

I

2.730

1.068

0.689

2.27

I

OCS

PCB
1:1

Sum
PCBs*

2,3,7,
8–
TCDD

2,3,7
,8–
TCD
F

OCS
0.001
0
0.001
7
0.001
1
0.000
7
0.001
3

b) Double-crested cormorant eggs:
Year
198
9
201

Colony

p,p’DDE

Sum
Chlorda
ne

HCB

Dieldr
in

Cone I.

3.531

0.158

0.013

0.143

0.043

0.026

0.004

8.953

4.043

12

2

Welcome

1.337

0.010

0.006

0.024

0.007

0.003

0.001

NA

0.921

-

-

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC

43

 2
201
4

I.-a
Welcome
I.-a

1.070

0.014

0.008

0.028

0.008

0.008

0.001

2.320

1.180

3.07

I

* Based on the sum concentration of 35-62 PCB congeners
** Based on the sum concentration of 14 or 15 PBDE congeners; NA denotes that data are not available
*** Based on collections of dead eggs from egg viability study
I = possible interference

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC

44

 Table S2. Mercury concentrations (µg/g, wet weight) in herring gull eggs from
Mutton Island in the Thunder Bay AOC in 1992, 1993, 1996, 2000, and 2007. Results
are based on analyses of a single pooled sample of 9-13 eggs with the exception of
eggs from Mutton Island in 1993 and 1996 where two pools of five eggs were
analyzed and mean concentrations (SD) are shown. Note that as a result of
differences in chemical methodology, concentrations in eggs collected from 19921996 have been adjusted as described in Weseloh et al. (2011) to allow for
comparisons to current data. Mean mercury concentrations (SD) in gull embryos
from Mutton Island in 2012 and 2014 are also shown for reference.
Yea
r
199
2
199
3
199
6
200
0
200
7
201
2
201
4

Colony

Mercury

Mutton I.

0.140

Mutton I.
Mutton I.

0.081
(0.016)
0.138
(0.043)

Mutton I.

0.150

Mutton I.

0.111

Mutton I.
Mutton I.

0.086
(0.042)
0.054
(0.028)

ii) Egg Viability (2007)
Herring gull egg viability was determined at Mutton Island in 2007 as part of a broad
study of biological effects associated with contaminant exposure in herring gulls
from the Great Lakes from 2000-2007. Egg viability was determined using an
embryonic viability detector (EVD) that detects vibrations associated with
embryonic movement and heartbeat. These vibrations are turned into sound waves
allowing for a determination of whether an embryo is dead or alive. Eggs were
examined at 15-20 days of incubation and egg viability was calculated as the
proportion of live eggs in a 3-egg clutch. Egg viability was also determined at
Chantry Island on Lake Huron, which serves as the non-AOC Great Lakes reference
site for comparison purposes. Herring gull egg viability at Mutton Island in the
Thunder Bay AOC in 2007 was high and equal to 94.1% on average (Table S3). Egg
viability was also high and equal to 96.8% at Chantry Island in 2000-2004. No
significant difference in egg viability was found between the two colonies.
Table S3. Mean percentage (SD) of viable herring gull eggs in 3-egg clutches from
Mutton Island in the Thunder Bay AOC and Chantry Island in Lake Huron as the
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC
45

 reference colony in years shown and where data are available. N indicates the
number of 3-egg clutches.
Colony

Year

% Viable

N

Mutton I.
Chantry
I.

2007

94.1 (12.8)
96.8 (10.3)

45
229

2000-2004

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC
46

 iii) Populations of Nesting Colonial Waterbirds at Thunder Bay AOC
Study Colonies and
Lake Superior:
Table S4. Census data of herring gull and double-crested cormorant nests (=pairs)
at study colonies in the Thunder Bay AOC (A) and for all colonial waterbirds on the
Canadian and American sides of Lake Superior (B) during the 1 st (1977/1978), 2nd
(1989-1991), 3rd (1996-2001) and 4th (2007-2010) decadal surveys (Scharf 1978,
1998; Blokpoel et al. 1980; Blokpoel and Tessier 1993, 1998; Scharf and Shugart
1998; Cuthbert et al. 2001; CWS unpublished; Cuthbert and Wires unpublished).
Decadal survey years in the Thunder Bay AOC are shown by an asterisk and nest
counts for other years were collected during assorted field studies by CWS in the
AOC. Percent change in nest numbers for colonial waterbirds in Lake Superior are
between the 3rd to 4th decadal surveys.
A) Herring gulls – AOC study colonies:
200
1
182

200
2
196

2007

114

200
0
NC

70

164

0

NC

229

241

216

432

519

-

1978

1989

1999

*

*

*

133

Mutton Island
50
Welcome Islands0
a
Welcome Islands400
b
Total
450

180

200
8
177

201
2
171

201
4
142

201
5
129

NC

255

NC

239

148

118

NC

NC

433

NC

347

NC

NC

-

-

868

-

757

-

-

*

NC=not counted

Double-crested cormorants – AOC study colonies:
1978

1989

1996

*

*

200
0

2007

*

62

69

131

455

73

261

421

1,119

135

330

552

1,57
4

Welcome Islands0
a
Welcome Islands1
b
Total

1

*

B) Colonial waterbirds – Lake Superior:
Percent
Change

Census Year
Species
1977/78
Herring Gull
13,269
Ring-billed Gull
7,471
Double-crested Cormorant 35
Great Blue Heron
527
Common Tern
328
American White Pelican
0
Caspian Tern
0
Black-crowned Night0

19891991
24,896
15,742
1,975
685
282
0
0
0

19962001
18,987
18,761
4,270
466
316
0
0
0

20072010
15,255
15,656
4,789
343
315
20
4
3

-19.7%
-16.6%
+12.2%
-26.4%
-0.3%
-

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC
47

 Percent
Change

Census Year
Species
Heron
Totals

1977/78

19891991

19962001

20072010

21,630

43,580

42,800

36,385

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC
48

 iv) Egg Size Parameters (2001 & 2002):
Significant spatial variation was found for total clutch volume in 3-egg clutches from
Mutton Island compared to other colonies in both 2001 and 2002 (Table S5). In both
years, mean clutch volume was significantly smaller in clutches from Mutton Island
compared to the Chantry Island reference colony. In one of two years, total clutch
volume was significantly smaller in clutches from Mutton Island compared to
Granite Island, another gull colony on Lake Superior. Mean intraclutch variation
(ICV) in egg size was not significantly different among the three colonies in both
years. There was no evidence of temporal changes for either of these endpoints in
2001 and 2002 compared to current estimates in clutches of eggs in 2012, 2014,
and 2015 (Table S6).
Table S5. Mean total clutch volume (SD) and intraclutch variation (SD) in 3-egg
clutches of herring gulls from Mutton Island, Granite Island and Chantry Island in
2001 and 2002.
Colony

Year

Mutton I.

2001

Granite I.

2001

Chantry
I.

2001

Mutton I.

2002

Granite I.
Chantry
I.

2002
2002

Clutch Volume
(cm3)
241.23 (19.43) B
251.08 (18.05)
AB
256.48 (16.87) A
240.51 (22.76) B
257.17 (16.12) A
255.0 (14.23) A

ICV (%)

N

9.48 (4.75)
10.51 (6.35)

152
30

8.23 (4.32)

63

12.35
(13.91)
8.84 (5.02)
9.63 (3.64)

31
30
48

Table S6. Temporal trends in mean total clutch volume (SD) and intraclutch variation
(SD) in 3-egg clutches of herring gulls from Mutton Island in 2001, 2002, 2012,
2014, and 2015. N indicates the number of 3-egg clutches.

2001

Clutch Volume
(cm3)
241.23 (19.43)

2002

240.51 (22.76)

2012
2014
2015

250.96 (18.32)
236.94 (9.83)
243.68 (18.18)

Year

ICV (%)

N

9.48 (4.75)
12.35
(13.91)
10.66 (3.88)
10.02 (7.66)
8.64 (3.20)

152
31
12
9
12

v) Stress Response in Juvenile Herring Gulls in the Thunder Bay AOC
(2000 & 2008):
In order to assess adrenal responsiveness, two different methods were employed to
simulate stress in juvenile chicks from Welcome Islands-b in 2000 and 2008. In
Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC
49

 2000, chicks underwent a standardized capture, handling and restraint procedure
known to elicit an increase in circulating corticosterone levels. In 2008, chicks were
injected with ACTH as a standardized stress challenge. Details for both methods are
provided in Mayne et al. (2004). Following the stress event, blood samples were
collected at regular intervals to measure the bird’s ability to respond through the
production of corticosterone. Juvenile chicks, i.e., between 21 and 28 days of age,
were captured and housed separately prior to the stress test. Response was
measured in eight chicks from Welcome Islands-b in 2000 and 12 chicks in 2008. In
2000, where handling chicks represented the stress challenge, serial blood samples
were taken at 0, 20, and 40 minutes. For the ACTH challenge, birds were injected
with ACTH (Cortosyn, Amphastar Pharmaceuticals, 5 µg/kg body weight) at
staggered times. Serial 0.5 ml blood samples were then taken from each individual
at 10, 25, and 45 minutes post-injection. Hereafter, methods for quantifying
corticosterone in plasma were similar in the two study years. Blood was transferred
to a microcentrifuge tube containing 5 µl heparin and mixed by inversion and kept
on ice until centrifugation. The separated plasma and red blood cells were then
frozen in liquid nitrogen. The plasma samples were analyzed for corticosterone by
radioimmunoassay (IMMUNOCHEM double antibody corticosterone 125I RIA kit [for rats
and mice]; ICN #07-120103, Santa Ana, CA, USA) as per the kit instructions. In this
test, a limited amount of specific antibody is reacted with 125I-labeled corticosterone.
When an additional amount of corticosterone is available there is less 125I-labelled
corticosterone bound to the antibody. After separation of the bound from the free
125
I-labelled corticosterone, the amount of radioactivity is measured and the amount
of hormone present in the sample can be determined from a standard curve.
Statistical comparisons were made to juvenile chicks from Chantry Island on Lake
Huron in 2000 (N=7 chicks) and 2003 (N=16 chicks) where these tests were also
done.
In 2000, corticosterone response was not significantly different between juvenile
chicks from the Thunder Bay AOC colony and the Chantry Island reference colony
following the handling challenge (repeated measures ANOVA; Figure S1a). As
expected, there was a significant difference in the secretion of corticosterone over
the three time periods following the handling challenge (time effect; p=0.0003). No
significant difference was found in the pattern of corticosterone secretion between
the two colonies (site x time effect).
In 2008, a similar pattern was found following the ACTH challenge (Figure S1b).
Corticosterone response was not significantly different between juvenile chicks from
the Thunder Bay AOC colony and the Chantry Island reference colony or the pattern
of secretion between the two colonies (repeated measures ANOVA). There was a
significant difference in the secretion of corticosterone over the three time periods
following the ACTH challenge (p=0.000001).

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC
50

 Figure S1. Mean (SD) corticosterone levels in plasma of juvenile herring gulls from the
Thunder Bay AOC (Welcome Islands-b) with no ACTH challenge (a) and following an
ACTH challenge (b) in 2000 and 2008 (N=8 and 12 chicks, respectively). Chantry Island
on Lake Huron served as the reference colony in both studies where chicks were
sampled in 2000 and 2003 (N=7 and 16 chicks, respectively).
a)

b)

Assessment of Wildlife Reproduction & Deformities BUI in the Thunder Bay AOC
51

  

Assessment of Wildlife Reproduction Deformities BUI in the Thunder Bay ADC
52