lable at ScienceDirect
Toxicon 129 (2017) 36e43

Contents lists avai

Toxicon
journal homepage: www.elsevier.com/locate/toxicon

The role of a PSP-producing Alexandrium bloom in an unprecedented
diamondback terrapin (Malaclemys terrapin) mortality event in
Flanders Bay, New York, USA
Theresa K. Hattenrath-Lehmann a, Robert J. Ossiboff b, 1, Craig A. Burnell c,
Carlton D. Rauschenberg c, Kevin Hynes d, Russell L. Burke e, Elizabeth M. Bunting b,
Kim Durham f, Christopher J. Gobler a, *
a Stony Brook University, School of Marine and Atmospheric Sciences, Southampton, NY, 11968, USA
b Cornell University, College of Veterinary Medicine, Department of Population Medicine and Diagnostic Sciences, Ithaca, NY, 14853, USA
c Bigelow Analytical Services, Bigelow Laboratory for Ocean Sciences, East Boothbay, ME, 04544, USA
d New York State Department of Environmental Conservation, Wildlife Health Unit, Delmar, NY, 12054, USA
e Hofstra University, Department of Biology, Hempstead, NY, 11549, USA
f Riverhead Foundation for Marine Research and Preservation, Riverhead, NY, 11901, USA




a r t i c l e i n f o
Article history:
Received 11 November 2016
Received in revised form
3 February 2017
Accepted 11 February 2017
Available online 14 February 2017
Keywords:
Alexandrium
Diamondback terrapins
PST
Saxitoxin
Turtle
Chelonian

* Corresponding author.
E-mail address: christopher.gobler@stonybrook.ed
1 Present address: Zoological Pathology Program, Co
University of Illinois, Brookfield, IL, USA.
http://dx.doi.org/10.1016/j.toxicon.2017.02.006
0041-0101/© 2017 Elsevier Ltd. All rights reserved.

a b s t r a c t
Diamondback terrapins (Malaclemys terrapin) are a threatened or endangered species in much of their
range along the U.S. Atlantic and Gulf coasts. Over an approximately three-week period from late April to
mid-May 2015, hundreds of adult diamondback terrapins were found dead on the shores of Flanders Bay,
Long Island, New York, USA. Concurrent with the mortality event, elevated densities of the paralytic
shellfish toxin (PST)-producing dinoflagellate, Alexandrium fundyense (>104 cells L
1) and high levels of
PST in bivalves (maximal levels ¼ 540 mg STX eq. 100 g
1 shellfish tissue) were observed in the Flanders
Bay region, resulting in shellfish bed closures in regional tributaries. Gross and histologic postmortem
examinations of terrapins revealed no physical trauma to individuals or a common, underlying disease
process to explain the deaths. PST compounds (0.2e12.5 mg STX eq. 100 g
1) were present in various
M. terrapin tissues collected over the duration of the mortality event. High-throughput sequencing
revealed that the ribbed mussel (Geukensia demissa, a PST vector) was present in the gastrointestinal
tracks of all terrapin samples tested. While the potential of PST to cause mortality in chelonians has not
been well-characterized, in the absence of other significant findings from necropsies and pathological
analyses, we provide evidence that PST in shellfish was likely high enough to cause or contribute to the
mortality in these small (<2.0 kg) animals.
© 2017 Elsevier Ltd. All rights reserved.








1. Introduction
Several genera of phytoplankton can produce a variety of bio-
toxins (i.e. saxitoxin, domoic acid, brevetoxins, microcystins) that
can have serious health, ecosystem, and economic impacts
(Anderson et al., 2008, 2012, 2002; Hoagland et al., 2002; Sunda
et al., 2006; Van Dolah, 2000). Beyond the well-established ef-
fects these harmful algal blooms (HABs) and their toxins can have

u (C.J. Gobler).
llege of Veterinary Medicine,

on humans via shellfish poisoning events (Anderson et al., 2012;
Van Dolah, 2000), many of these toxins are transferred through
the food chain and thus can affect a variety of wildlife including fish
(Landsberg et al., 2014), marine mammals (Geraci et al., 1989;
Lefebvre et al., 2016; Miller et al., 2010; Van Dolah et al., 2003),
seabirds (Shearn-Bochsler et al., 2014; Shumway et al., 2003), and
sea turtles (Amaya et al., 2014; Capper et al., 2013; Fauquier et al.,
2013; Licea et al., 2008).
Paralytic shellfish toxins (PST) are composed of more than 20
different saxitoxin (STX) analogs (Rourke et al., 2008), a particularly
potent class of algal toxins. Saxitoxin and associated analogs have a
long history of causing seabird deaths (Shearn-Bochsler et al., 2014;
Shumway et al., 2003) as well as marine mammal strandings and

Fig. 1. Locations of Alexandrium sampling sites and terrapin collection sites. Black
star ¼ Meetinghouse Creek, White star ¼ James Creek. Black diamond ¼ Terrapin
sample CU1. Gray diamond ¼ Terrapin sample SB1. White diamond ¼ terrapin sample
CU2. Black circle ¼ all other terrapin samples. Dead terrapins were found along the
north and south shore of western Flanders Bay during late April and early May 2015.
T.K. Hattenrath-Lehmann et al. / Toxicon 129 (2017) 36e43
37

mortalities (Geraci et al., 1989; Lefebvre et al., 2016; Van Dolah
et al., 2003). Globally, dinoflagellates in the genus Alexandrium
are the primary causative species of paralytic shellfish poisoning
(PSP; Landsberg et al., 2014) although since the 1970s there have
been a series of sea turtle deaths attributed to STX poisoning from
dinoflagellate blooms caused by Pyrodinium spp. (Amaya et al.,
2014; Licea et al., 2008; Maclean, 1975). To our knowledge, there
have been no reports of STX poisoning in turtle species inhabiting
freshwater or brackish environments.
Diamondback terrapins (Malaclemys terrapin) are medium-sized
emydid turtles with strong sexual dimorphism (carapace length of
adult females to 23 cm, adult males to 14 cm) that inhabit brackish
water habitats, marshes, and mangroves along the U.S. Atlantic and
Gulf coasts (Ernst and Lovich, 2009). M. terrapin was heavily har-
vested for food during the 19th and early 20th centuries (Schaffer
et al., 2008) and more recently has suffered population declines
from habitat loss, road mortality, and as by-catch in crab fishing
gear (Bishop, 1983; Dorcas et al., 2007; Grosse et al., 2011; Ner and
Burke, 2008; Roosenburg et al., 1997; Szerlag and McRobert, 2006).
Unlike sea turtles, terrapins are more sedentary, showing strong
fidelity to foraging, nesting, and overwintering habitats generally
spanning no more than one km (Baldwin et al., 2005; Tucker et al.,
2001). Terrapins eat a variety of marine invertebrates including
hard shelled prey such as salt marsh periwinkles (Littorina irrorata;
Tucker et al., 1995), Atlantic ribbed mussels (Geukensia demissa),
and blue crabs (Callininectes sapidus) (Erazmus, 2012; Petrochic,
2009). Given that terrapins consume shellfish, exposure to STX is
plausible via these known PST vectors. To date, no study has
demonstrated the effects of- or the lethal dose (LD50) of- PSTs in
terrapins or any other reptile.
Here we report on aM. terrapinmortality event that occurred in
Flanders Bay, NY, USA, that coincided with Alexandrium blooms in
this system. This study presents Alexandrium bloom dynamics,
toxicity in shellfish, and a description of terrapin deaths. Subsets of
dead terrapins from different stages of the mortality event were
necropsied, histologically examined, and various tissues were
analyzed for the presence of PST. High-throughput sequencing of
DNA isolated from intestinal contents was used to assess prey
consumed by the terrapins. Additionally, we present a series of
dose-response scenarios based on the range of known lethal doses
for vertebrates.


2. Materials and methods
2.1. Alexandrium monitoring
Field samples were collected on a weekly basis from April
through June during 2015. Samples were collected from Meeting-
house Creek (4056012.600N 7237006.300W) and James Creek
(4059010.700N 7232005.600W), both tidal tributaries located in the
Peconic Estuary, NY, USA (Fig. 1). Concentrated water samples were
made by sieving 2 L of subsurface water through a 200 mmmesh (to
eliminate large zooplankton) and then onto a 20 mm sieve that was
backwashed into a 15 mL centrifuge tube. Alexandrium fundyense
densities were enumerated using a highly sensitive molecular
probe procedure described by Anderson et al. (2005). Briefly, ali-
quots of phytoplankton concentrates (formalin and then methanol
preserved) were hybridized with an oligonucleotide probe specific
for the NA1 North American (Group I, recently revised as
A. fundyense (John et al., 2014)) ribotype Alexandrium fundyense/
catenella/tamarense with Cy3 dye conjugated to the 50 terminus
(50-/5Cy3/AGT GCA ACA CTC CCA CCA-30). Cells were enumerated
using a Nikon epifluorescence microscope with a Cy3™ filter set
(Anderson et al., 2005) as described in Hattenrath et al. (2010).


Shellfish toxicity data was supplied by the New York State
Department of Environmental Conservation (NYSDEC). Briefly,
hundreds of bluemussels (Mytilus edulis) were collected from Stony
Brook Harbor, NY, USA, where Alexandrium and PSP have never
been detected (Hattenrath-Lehmann, Gobler, NYSDEC, unpub.).
Mussels were placed in mesh bags in aliquots of 15 mussels per bag
and were deployed within Meetinghouse Creek and James Creek in
early April of 2015. Mussels were collected weekly from each site,
shucked of their meat, and prepared for PST analysis using standard
techniques (AOAC, 1990). Toxin levels in shellfish were quantified
using standard mouse bioassays (AOAC, 1990) performed by FDA
approved laboratories (NYSDEC lab and Resource Access Interna-
tional, LLC, ME, USA).


2.2. Diamondback terrapin postmortem examinations and
ranavirus testing
Complete postmortem examinations were performed on 12
carcasses submitted to the New York State Veterinary Diagnostic
Laboratory at Cornell University (Ithaca, NY, USA) and on three
carcasses submitted to the New York State Department of Envi-
ronmental Conservation Wildlife Health Unit at the Wildlife Re-
sources Center (Delmar, NY, USA). Sample size during this study
was based on the availability of carcasses and their condition upon
collection. Samples of fresh liver, kidney and heart were collected
and archived frozen (
20 C); samples of gastrointestinal contents
were collected and archived frozen (
20 C). The choana and cloaca
were swabbed with a single rayon-tipped, plastic shaft, fine tip
swab (Medical Wire and Equipment, Wiltshire, England) and swabs
were archived frozen (
20 C). A complete set of tissues was
collected, fixed and stored in 10% neutral buffered formalin at room
temperature, and processed routinely for histologic examination.
Paraffin sections were cut at 5 mm and stained with hematoxylin
and eosin (HE). Choanal/cloacal swabs and gastrointestinal con-
tents (when present) were collected from an additional 18 car-
casses in poor postmortem condition and archived frozen (
20 C).
An additional three gross postmortem examinations were con-
ducted by the Riverhead Foundation for Marine Research and

T.K. Hattenrath-Lehmann et al. / Toxicon 129 (2017) 36e43
38

Preservation (Riverhead, NY, USA); samples of kidney, heart, liver
and gastrointestinal contents from one, freshly killed individual
were archived frozen (
20 C).
Swab nucleic acid extractionwas performedwith Prepman Ultra
(Applied Biosystems, Foster City, California, USA) per the manu-
facturer's recommendations. Extracts were diluted 1:10 in molec-
ular grade water. Quantification of extracted nucleic acids was
performed on a subset of samples using Qubit fluorometric quan-
tification (Invitrogen, Carlsbad, California, USA). Nucleic acid ex-
tracts were screened for ranavirus using a previously published
hydrolysis probe quantitative polymerase chain reaction (qPCR)
targeting a portion of the viral major capsid protein (Pallister et al.,
2007). Briefly, qPCR reactions were composed of sense (50-CTC ATC
GTT CTG GCC ATC AA-30) and antisense primers (50-TCC CAT CGA
GCCGTT CA-30), a Taqman hydrolysis probe (50-/6FAM/CAC AAC ATT
ATC CGC ATC/MGBNFQ/e30), mixed with TaqMan Fast-Virus 1-step
Master Mix (Life Technologies, Carlsbad, CA, USA), TaqMan Exoge-
nous Internal Positive Control (Life Technologies, Carlsbad, CA,
USA), and 2 ml choanal-cloacal swab extracts. Plates were run on a
StepOne Real-Time PCR System (Life Technologies, Carlsbad, CA,
USA) under the following conditions: 1 cycle at 50 C for 5 min; 1
cycle at 95 C for 20 s; 40 cycles of 95 C for 3 s followed by 60 C for
30 s and a data collection point. A commercially synthesized ssDNA
oligomer including ranavirus primer and probe binding sites (In-
tegrated DNA Technologies, Coralville, IA, USA) and a positive
clinical sample extract from an eastern box turtle (Terrapene car-
olina carolina) were used as positive controls.


2.3. Paralytic shellfish toxin measurements in diamondback
terrapin tissues
Paralytic shellfish toxin (PST) concentrations in terrapin tissues
were investigated at Bigelow Analytical Services, Bigelow Labora-
tory for Ocean Sciences (East Boothbay, Maine, USA). Lyophilized
tissue samples were reconstituted in deionizedwater and extracted
with 0.1 M acetic acid (Bricelj et al., 1991) using 1.65 mL g
1 wet
tissue and a Brinkmann Instruments Polytron homogenizer. Sam-
ples were centrifuged at 4000g for 20 min at 5 C. Supernatants
were loaded onto preconditioned Waters Sep-Pak C-18 Light single
use cartridges and eluted by syringe pressure. Samples were
analyzed via high performance liquid chromatography with post-
column oxidation and fluorescence detection (HPLC-PCOX-FLD;
Van de Riet et al., 2011).
The total toxicity for each sample was calculated from the sum
of the individual saxitoxin-normalized toxicities for all PSTs present
(Van de Riet et al., 2009). Each sample was analyzed for: 1)
carbamate (GTX4, GTX1, GTX3, GTX2, NEO, and STX), decarbamoyl
(dcGTX3, dcGTX2, and dcSTX) and N-sulfocarbamoyl (GTX5) PST
congeners; and 2) two N-sulfocarbamoyl congeners (C1 and C2).
Certified calibration standards (Institute for Marine Biosciences,
National Research Council of Canada) were used to identify and
quantify the individual compounds. Checks were conducted to
ensure identified peaks were not due to autofluorescence by non-
PST compounds. The limit of quantitation (LOQ) for this approach
is approximately 0.1 mg STX eq 100 g
1, based on values from the
analysis of terrapin tissues that were not exposed to saxitoxin.


2.4. Sequencing COI gene of diamondback terrapin intestinal
contents
Intestinal contents from three terrapins were extracted using
the Qiagen Dneasy Blood and Tissue kit (Leray et al., 2013). Aliquots
of 25 mg were lysed at 56 C for 1.5 h and extractions were con-
ducted according to the manufacturers' tissue protocol. Following


extraction, double-stranded DNA was quantified on a Qubit® fluo-
rometer using a dsDNA BR (Broad-Range) Assay kit (Qubit®). All
samples were normalized and sent to Molecular Research Labs
(Shallowater, TX, USA) for amplicon sequencing. The mitochondrial
cytochrome c oxidase subunit I gene (COI) was amplified using
miCOIintF and jgHCO2198 (50-TAI ACY TCI GGR TGI CCR AAR AAY
CA-30) degenerate primers from Leray et al. (2013). The forward
primer, however, was modified to allow for amplification of local
shellfish species (miCOIintF modified: 50-GGW RSD GGR TGR ACW
MTT TAY CCK CC-30), resulting in an ~300 bp product. In addition,
annealing blocking primers were created to prevent the amplifi-
cation of predator-specific DNA (M. terrapin) as to not mask the
prey DNA signal (Leray et al., 2013). The blocking primer (50- TAY
CCYCCATTAGCC GGA AAC CTA/3SpC3/-30) was created to anneal to
the forward universal primer, was modified at the 30 end with a
Spacer C3 CPG to prevent elongation, and was added at 10 the
concentration of the universal primers (Leray et al., 2013). For each
sample, an identifying barcode was placed on the forward primer
and a 30 cycle PCR using the HotStarTaq Plus Master Mix Kit
(Qiagen, USA) was performed. The following PCR conditions were
used: 94 C for 3 min, followed by 28 cycles of 94 C for 30 s, 53 C
for 40 s and 72 C for 1 min, and a final elongation step at 72 C for
5 min. Successful amplification was determined by examining PCR
products using a 2% agarose gel. Each bar-coded sample was pooled
in equal proportions based on its molecular weight and DNA con-
centration. Pooled samples were then purified using calibrated
Ampure XP beads and subsequently used to prepare a DNA library
by following Illumina TruSeq DNA library. Paired-end (2  300)
sequencing was performed on an Illumina MiSeq following the
manufacturer's guidelines. Sequence data was processed using the
Quantitative Insights Into Microbial Ecology v1.9.1 pipeline (QIIME,
http://qiime.org; Caporaso et al., 2010b). Raw sequences were
depleted of barcodes, paired-end reads joined, depleted of primers,
demultiplexed, and quality-filtered using the default parameters in
QIIME v1.9.1. The resulting quality filtered sequences were then
clustered into operational taxonomic units (OTUs) at 99% similarity
with UCLUST (Edgar, 2010) using the open reference clustering
protocol and a custom COI database as the reference set. A custom
COI database was created from sequences of local fish and shellfish
species deposited in GenBank (http://www.ncbi.nlm.nih.gov). The
representative sequence set was aligned using PyNAST (Caporaso
et al., 2010a), taxonomically classified using BLAST (Altschul et al.,
1990), and an OTU table was generated.


3. Results
3.1. Alexandrium blooms and diamondback terrapin deaths
Alexandrium cells were observed in Meetinghouse Creek
beginning on 9 April 2015 with densities peaking at ~47,000 cells
L
1 on 29 April and the bloom ending two weeks later (12 May;
37 cells L
1; Fig. 2). At the peak of the bloom, mussels deployed in
Meetinghouse Creek had accumulated low levels of PST (44 mg STX
eq 100g
1) but became highly toxic, almost seven times greater
than the federal closure limit (80 mg STX eq 100g
1), one week later
(5 May; 510 mg STX eq 100g
1; Fig. 2) with toxin levels dissipating
thereafter. Meetinghouse Creek, and the adjoining Terry Creek,
were closed to all shellfish harvest on 6May and reopened on 1 July
to harvest. In James Creek, Alexandrium densities increased from
91 cells L
1 (20 April) to ~5100 cells L
1 (4 May) and peaked at
~14,000 cells L
1 on 15 May (Fig. 2). Concurrently, PST levels in
mussels were 350 mg STX eq 100g
1 on 16 May prompting the
closure of shellfish beds in this region (Figs. 1 and 2). During spring
2015, hundreds of dead terrapins washed up in the western

Fig. 2. (A) Meetinghouse Creek and (B) James Creek Alexandrium densities (cells per L) and corresponding shellfish toxicity of blue mussels measured via mouse bioassay during the
spring of 2015. Gray box indicates the mortality period of terrapins collected during this study. Black dotted line indicates the first report of dead terrapins.
T.K. Hattenrath-Lehmann et al. / Toxicon 129 (2017) 36e43
39

Flanders Bay region (Figs. 1 and 2). The first report of dead dia-
mondback terrapins was on 21 April 2015 and continued through
mid-May (Fig. 2, gray box), during which several sets of terrapins
were collected for examination (Fig. 1; Table 1). During the mor-
tality event, 27 terrapin carcasses were collected and submitted by
a local turtle rescue center and 6 additional terrapin carcasses were
collected by NYSDEC biologists. A singular survey during this three-
week mortality event on 15 May 2015 quantified more than 100
dead terrapins along Iron Point, a peninsula representing a very
small fraction of the Flanders Bay coastline (Fig. 1).
Alexandrium densities have been historically monitored in the
Flanders Bay region (including Reeves Bay, Terry Creek and East
Creek, all of which are < 1 km from or are connected to Meeting-
house Creek) in the late 1980s (Nuzzi and Waters, 1993) and have
been monitored specifically in Meetinghouse Creek every spring
since 2008 (Table 2), with peak bloom densities occurring between
late-March to mid-May. To date, the 2015 bloom (~47,000 cells L
1)
achieved the highest Alexandrium densities ever observed in

Table 1
PST concentrations (mg STX eq 100 g
1) in the organs of- and condition of- diamondba
University, Cornell and Delmar. GI ¼ Gastrointestinal. n/a ¼ not analyzed. * ¼ sample G
(freshly collected), F ¼ fair postmortem condition (some autolysis), P ¼ poor postmortem
Project ID #
Mortality period
PST in tissue (mg STX eq 10
Kidney
Heart
SB1
Late (5/15)
0
1.2
CU1
Early (4/24)
1
0
CU2
Early (4/24)
0
12.5
CU3
Middle (4/28e5/3)
0
n/a
CU4
Middle (4/28e5/3)
0
n/a
CU5
Middle (4/28e5/3)
n/a
n/a
CU6
Middle (4/28e5/3)
n/a
n/a
Del1
Middle (4/28e5/3)
n/a
0
Del2
Middle (4/28e5/3)
n/a
0
Del3
Middle (4/28e5/3)
n/a
n/a

Meetinghouse Creek or the Peconic Bay region (Table 2). Similarly,
PST levels measured in 2015 (540 mg STX eq 100g
1) were the
highest ever measured in the Meetinghouse Creek or the Peconic
Estuary (Table 2).
3.2. Diamondback terrapin postmortem examinations and
ranavirus testing
During the mortality event, 33 fresh or previously frozen car-
casses were submitted for postmortem examination. Fifteen car-
casses (5 males, 10 females) were selected for complete
postmortem examination based on the degree of postmortem
autolysis and the date at which the carcass was found during the
mortality event. All animals were mature, with males having a
medianweight of 345 g (range 270e370 g) and a median carapacial
length of 13.5 cm (range 11.5e15 cm) and females having a median
weight of 1596 g (range 1323e1905 g) and a median carapacial
length of 21.5 cm (range 20.5e23 cm). All females had active

ck terrapins collected from the Flanders Bay region and examined at Stony Brook
I contents used for high-throughput sequencing. G ¼ good postmortem condition
condition (decay/autolysis).
0 g
1)
Condition of terrapins
Liver
GI contents
0
0
G
1.2
n/a
G
0
n/a
G
0
n/a
P
0
n/a
F
n/a
0.2
P
n/a
0.3
P
0
n/a
F
n/a
n/a
F
n/a
0
P
Table 2
Maximum Alexandrium densities (cells L
1; peak date in parentheses) and
maximum shellfish toxicity (mg STX eq 100 g
1 shellfish tissue) measured via mouse
bioassay in deployed blue mussels (Mytilus edulis) from Meetinghouse Creek, NY,
USA. n.m. ¼ not monitored, and * denotes no closure implemented, although over
the federal closure limit. a, b, c¼ data collected from Nuzzi andWaters 1993 and are
from Reeves Bay, Terry's Creek and East Creek, respectively. Italics indicates year of
terrapin mortalities.
Year
Maximum Alexandrium
densities (cells L
1)
Maximum shellfish toxicity
(mg STX eq 100 g
1 shellfish tissue)
1986
14,000 (20-May)a
190a*
1987
500 (20-April)a
50a
1988
1600 (9-April)a
60a
1989
5700 (30-March)a
60a
1989
480 (4-May)b
<40b
1989
1000 (4-May)c
58c
2008
4733 (29-Apr)
n.m.
2009
19,868 (23-Apr)
<40
2010
1982 (15-Apr)
57
2011
1166 (5-May)
48
2012
17,206 (11-Apr)
380*
2013
1058 (10-Apr)
40
2014
7480 (8-May)
53
2015
46,690 (29-April)
540
2016
550 (16-May)
<40
T.K. Hattenrath-Lehmann et al. / Toxicon 129 (2017) 36e43
40

ovarian folliculogenesis at the time of death. Individuals not in an
advanced state of decay had full GI tracts and an intact fat layer. No
consistent, significant gross or histologic changes were present in
the examined tissues and a cause of death could not be determined
histologically. None of the terrapins exhibited any overt signs of
physical trauma. A subset (n ¼ 4) of the terrapins had small to
moderate amounts of red-tinged, clear pulmonary fluid; in other
terrapins, the lungs were aerated and free of fluid. Other findings,
including mononuclear infiltrates in the renal interstitium (n ¼ 3)
and mild parenchymal hemorrhage (n ¼ 1) were considered mild
and incidental to the mortality event. The degree of postmortem
autolysis and freeze thaw in a number of carcasses limited histo-
logic tissue interpretation.
Although not previously documented in terrapins, amphibian-
like ranaviruses have been associated with mortality in other
chelonian species (Marschang et al., 1999; Sim et al., 2016;Winzeler
et al., 2015). Combined choanal/cloacal swabs collected from 27
terrapins were all negative for the major capsid protein of
amphibian-like ranaviruses by qPCR.
3.3. Paralytic shellfish toxins in diamondback terrapin tissue
Initial analyses of terrapin gastrointestinal contents revealed the
qualitative presence of PST via mouse bioassay (Darcie Couture,
Research Access International, ME, USA; pers. comm.). Subsequent
analyses quantified PST in various organs (kidney, heart, liver,
gastrointestinal contents) of diamondback terrapins collected over
the duration of the reported mortality event (Table 1). Individuals

Table 3
PST analog concentrations (mg 100 g
1) and toxicity equivalents (mg STX eq 100 g
1)
GI ¼ Gastrointestinal.
Project ID #
Organ
PST concentration in tissue (and eq
mg 100 g
1 (mg STX eq. 100 g
1)
GTX5
NEO
SB1
Heart
CU1
Kidney
CU1
Liver
CU2
Heart
8.4 (7.8)
CU5
GI contents
3.6 (0.2)
CU6
GI contents
4.2 (0.3)

that were collected and frozen immediately after death (CU1, CU2,
SB1) had the highest levels and most consistent presence of PST
(Table 1). Consistent with Alexandrium cultures isolated from the
New York region (Hattenrath-Lehmann et al., 2015), C2 was the
most abundant congener measured in terrapin tissue (mg 100 g
1;
Table 3).
Heart tissues from these individuals had the highest PST con-
centrations with individual CU2 and SB1 having 12.5 (consisting of
NEO and STX) and 1.2 mg STX eq 100g
1 (consisting of C2),
respectively (Tables 1 and 3). Individual CU1 contained 1.2 mg STX
eq 100g
1 liver tissue and 1.0 mg STX eq 100g
1 kidney tissue
consisting of C2 (Tables 1 and 3). Individuals in advanced states of
decay had either low levels of PST within gastrointestinal contents
(0.2e0.3 mg STX eq 100g
1) consisting of GTX5 or no detectable
levels of PST in any tissue (Tables 1 and 3).
3.4. High throughput sequencing of diamondback terrapin
intestinal contents
Samples sequenced with primer sets targeting the mitochon-
drial cytochrome c oxidase subunit I gene (COI) generated 440,776
paired-end reads with an amplicon size of ~300bp. After quality
filtering and joining reads a total of 388,897 reads clustered at 99%
identity into 259 OTUs. The majority (257) of OTUs were ‘unas-
signed’ meaning that they did not match any sequences in the
custom database of fish and local shellfish species that may
represent potential PST vectors. The two OTUs that did match the
custom database were both Geukensia demissa (ribbed mussel; 99%
identity). G. demissa sequences were amplified from all three
samples (CU5, CU6, and Del3; Table 1) representing <1% of total
reads. Paralytic shellfish toxins were present in the intestinal con-
tents of all but one (Del3) of the samples sequenced (Table 1).
4. Discussion
Since the middle of the 20th century, harmful algal blooms have
become more pervasive along coastlines world-wide (Anderson
et al., 2008, 2012, 2002; Hoagland et al., 2002; Sunda et al., 2006;
Van Dolah, 2000). Concurrently, diamondback terrapin pop-
ulations, a species that is limited to shallow brackishwater marshes
and mangrove swamps of the U.S. Atlantic and Gulf Coasts, have
become threatened or endangered in parts of its range due to
overharvesting, habitat loss, road mortality, and by-catch (Bishop,
1983; Dorcas et al., 2007; Grosse et al., 2011; Ner and Burke,
2008; Roosenburg et al., 1997; Schaffer et al., 2008; Szerlag and
McRobert, 2006). While turtle deaths have been attributed to
HABs in general and PST in particular in some tropical regions
(Amaya et al., 2014; Licea et al., 2008; Maclean, 1975), such occur-
rences have not been reported within temperate estuaries. This
study documented the concurrent outbreak of Alexandrium blooms,
PST accumulation in bivalves, terrapin mortality, PST accumulation

in the organs of diamondback terrapins collected from the Flanders Bay region.
. toxicity)
Total toxicity
mg STX eq. 100 g
1
STX
C2
12.5 (1.2)
1.2
10.3 (1)
1
12.1 (1.2)
1.2
4.7 (4.7)
12.5
0.2
0.3