ArticlePDF Available

Septic systems contribute to nutrient pollution and harmful algal blooms in the St. Lucie Estuary, Southeast Florida, USA

Authors:

Abstract and Figures

Nutrient enrichment is a significant global-scale driver of change in coastal waters, contributing to an array of problems in coastal ecosystems. The St. Lucie Estuary (SLE) in southeast Florida has received national attention as a result of its poor water quality (elevated nutrient concentrations and fecal bacteria counts), recurring toxic Microcystis aeruginosa blooms, and its proximity to the northern boundary of tropical coral species in the United States. The SLE has an artificially large watershed comprised of a network of drainage canals, one of which (C-44) is used to lower the water level in Lake Okeechobee. Public attention has primarily been directed at nutrient inputs originating from the lake, but recent concern over the importance of local watershed impacts prompted a one-year watershed study designed to investigate the interactions between on-site sewage treatment and disposal systems (OSTDS or septic systems), groundwaters, and surface waters in the SLE and nearshore reefs. Results provided multiple lines of evidence of OSTDS contamination of the SLE and its watershed: 1) dissolved nutrients in groundwaters and surface waters were most concentrated adjacent to two older (pre-1978) residential communities and the primary canals, and 2) sucralose was present in groundwater at residential sites (up to 32.0 mg/L) and adjacent surface waters (up to 5.5 mg/L), and 3) d 15 N values in surface water (+7.5 o / oo), macroalgae (+4.4 o / oo) and phytoplankton (+5.0 o / oo) were within the published range (>+3 o / oo) for sewage N and similar to values in OSTDS-contaminated groundwaters. Measured d 15 N values in M. aeruginosa became increasingly enriched during transport from the C-44 canal (5.8 o / oo) into the mid-estuary (8.0 o / oo), indicating uptake and growth on sewage N sources within the urbanized estuary. Consequently, there is a need to reduce N and P loading, as well as fecal loading, from the SLE watershed via septic-to-sewer conversion projects and to minimize the frequency and intensity of the releases from Lake Okeechobee to the SLE via additional water storage north of the lake. These enhancements would improve water quality in both the SLE and Lake Okeechobee, reduce the occurrence of toxic harmful algal blooms in the linked systems, and improve overall ecosystem health in the SLE and downstream reefs.
Content may be subject to copyright.
Research
paper
Septic
systems
contribute
to
nutrient
pollution
and
harmful
algal
blooms
in
the
St.
Lucie
Estuary,
Southeast
Florida,
USA
Brian
E.
Lapointe*,
Laura
W.
Herren,
Armelle
L.
Paule
Harbor
Branch
Oceanographic
Institute
at
Florida
Atlantic
University,
Marine
Ecosystem
Health
Program,
5600
US
1
North,
Fort
Pierce,
FL,
34946,
USA
A
R
T
I
C
L
E
I
N
F
O
Article
history:
Received
14
July
2017
Received
in
revised
form
27
September
2017
Accepted
27
September
2017
Available
online
xxx
Keywords:
Harmful
algal
blooms
Septic
system
Eutrophication
Stable
isotopes
Sucralose
Macroalgae
A
B
S
T
R
A
C
T
Nutrient
enrichment
is
a
signicant
global-scale
driver
of
change
in
coastal
waters,
contributing
to
an
array
of
problems
in
coastal
ecosystems.
The
St.
Lucie
Estuary
(SLE)
in
southeast
Florida
has
received
national
attention
as
a
result
of
its
poor
water
quality
(elevated
nutrient
concentrations
and
fecal
bacteria
counts),
recurring
toxic
Microcystis
aeruginosa
blooms,
and
its
proximity
to
the
northern
boundary
of
tropical
coral
species
in
the
United
States.
The
SLE
has
an
articially
large
watershed
comprised
of
a
network
of
drainage
canals,
one
of
which
(C-44)
is
used
to
lower
the
water
level
in
Lake
Okeechobee.
Public
attention
has
primarily
been
directed
at
nutrient
inputs
originating
from
the
lake,
but
recent
concern
over
the
importance
of
local
watershed
impacts
prompted
a
one-year
watershed
study
designed
to
investigate
the
interactions
between
on-site
sewage
treatment
and
disposal
systems
(OSTDS
or
septic
systems),
groundwaters,
and
surface
waters
in
the
SLE
and
nearshore
reefs.
Results
provided
multiple
lines
of
evidence
of
OSTDS
contamination
of
the
SLE
and
its
watershed:
1)
dissolved
nutrients
in
groundwaters
and
surface
waters
were
most
concentrated
adjacent
to
two
older
(pre-1978)
residential
communities
and
the
primary
canals,
and
2)
sucralose
was
present
in
groundwater
at
residential
sites
(up
to
32.0
m
g/L)
and
adjacent
surface
waters
(up
to
5.5
m
g/L),
and
3)
d
15
N
values
in
surface
water
(+7.5
o
/
oo
),
macroalgae
(+4.4
o
/
oo
)
and
phytoplankton
(+5.0
o
/
oo
)
were
within
the
published
range
(>+3
o
/
oo
)
for
sewage
N
and
similar
to
values
in
OSTDS-contaminated
groundwaters.
Measured
d
15
N
values
in
M.
aeruginosa
became
increasingly
enriched
during
transport
from
the
C-44
canal
(!5.8
o
/
oo
)
into
the
mid-
estuary
(!8.0
o
/
oo
),
indicating
uptake
and
growth
on
sewage
N
sources
within
the
urbanized
estuary.
Consequently,
there
is
a
need
to
reduce
N
and
P
loading,
as
well
as
fecal
loading,
from
the
SLE
watershed
via
septic-to-sewer
conversion
projects
and
to
minimize
the
frequency
and
intensity
of
the
releases
from
Lake
Okeechobee
to
the
SLE
via
additional
water
storage
north
of
the
lake.
These
enhancements
would
improve
water
quality
in
both
the
SLE
and
Lake
Okeechobee,
reduce
the
occurrence
of
toxic
harmful
algal
blooms
in
the
linked
systems,
and
improve
overall
ecosystem
health
in
the
SLE
and
downstream
reefs.
©
2017
Elsevier
B.V.
All
rights
reserved.
1.
Introduction
Despite
their
ability
to
provide
invaluable
ecological
services
to
human
populations,
coastal
and
estuarine
ecosystems
are
being
degraded
on
a
global
scale.
Humans
have
signicantly
increased
the
concentrations
of
nitrogen
(N)
and
phosphorus
(P)
in
fresh-
waters
owing
into
the
coastal
zone
(Nixon,
1995;
Vitousek
et
al.,
1997;
MEA,
2005),
exacerbating
eutrophication,
harmful
algal
blooms
(HABs),
and
subsequent
habitat
loss
(NRC,
2000;
Glibert
et
al.,
2005;
Bricker
et
al.,
2007;
Heisler
et
al.,
2008).
The
complexity
of
this
problem
was
exemplied
by
Rothenberger
et
al.
(2009)
who
showed
that,
while
hog
farming
practices
were
an
important
contributor
of
eutrophic
and
unsafe
conditions
in
the
Neuse
River
(i.e.
toxic
Pesteria
blooms
(Burkholder
and
Glasgow,
2001)),
wastewater
treatment
plants
(WWTPs)
and
population
growth
were
also
signicant
nutrient
contributors.
Similarly,
increasing
nutrient
inputs
from
urban,
agricultural,
and
industrial
sources
have
synergistically
promoted
blooms
of
the
potentially
toxic
cyanobacterium
Microcystis
aeruginosa
on
a
global
scale
(Paerl
and
Otten,
2013;
Li
et
al.,
2017;
Liyanage
et
al.,
2016;
Preece
et
al.,
2017).
Some
of
the
most
chronic
blooms
have
occurred
in
Lake
Erie
(Wynne
and
Stumpf,
2015),
San
Francisco
Estuary
(Lehman
et
al.,
2015),
Cape
Fear
River
(Isaacs
et
al.,
2014,
Polera,
2016),
Patos
Lagoon
Estuary,
Brazil
(Yunes
et
al.,
1996),
and
Lake
Taihu,
China
(Chen
et
al.,
2003).
Findings
from
these
and
other
impacted
areas
showed
that
both
growth
and
toxicity
of
non-nitrogen
xing
(Paerl
et
al.,
2011)
*
Corresponding
author.
E-mail
address:
blapoin1@fau.edu
(B.E.
Lapointe).
https://doi.org/10.1016/j.hal.2017.09.005
1568-9883/©
2017
Elsevier
B.V.
All
rights
reserved.
Harmful
Algae
70
(2017)
122
Contents
lists
available
at
ScienceDirect
Harmful
Algae
journal
homepa
ge:
www.elsev
ier.com/locate/ha
l
M.
aeruginosa
are
directly
linked
to
eutrophication
and
exemplify
the
need
for
simultaneous
reduction
of
N
and
P
inputs
to
freshwater
and
estuarine
systems
(Conley
et
al.,
2009;
Ma
et
al.,
2014;
Gobler
et
al.,
2016).
In
estuarine
waters
with
salinity
>10
this
species
experiences
a
decrease
in
cellular
growth
and
abundance
and
an
increase
in
cell
mortality
ultimately
prompting
toxin
release
(Warhurst,
2014;
Preece
et
al.,
2017).
Until
recently,
high
P
inputs
alone
associated
with
low
N:P
ratios
(<44:1)
have
been
thought
to
promote
growth
of
M.
aeruginosa
blooms
(Downing
et
al.,
2005;
Horst,
2014;
Horst
et
al.
2014;
Parrish,
2014).
Recent
work
by
Lehman
et
al.
(2015)
and
a
review
by
Gobler
et
al.
(2016)
also
clearly
demonstrated
the
importance
of
inorganic
N,
especially
ammonium
(NH
4+
),
in
the
formation
of
Microcystis
blooms.
Contrary
to
previous
consensus,
Gobler
et
al.
(2016)
reports
that
Microcystis
has
multiple
physiological
adaptations
that
allow
bloom
formation
in
inorganic
P
!
depleted
waters.
Like
growth,
production
of
the
N-rich
hepatotoxin
microcystin
is
also
driven
by
N:P
ratios,
but
primarily
as
they
relate
to
N
assimilation
(Downing
et
al.,
2005;
Gobler
et
al.,
2016).
Because
microcystin
is
N-rich,
toxic
strains
of
Microcystis
require
more
N
than
non-toxic
strains
(Davis
et
al.,
2010).
Downing
et
al.
(2005)
show
that
N:P
ratios
between
8
and
51
promoted
the
highest
microcystin
content.
The
direct
relationship
between
microcystin
levels
and
dissolved
reactive
N
concentrations,
both
NH
4+
(Donald
et
al.,
2011)
and
nitrate
(NO
3!;
Horst
et
al.,
2014),
provided
evidence
of
wastewater
(also
referred
to
as
sewage)
inputs
during
these
blooms.
Like
wastewater
itself,
elevated
microcystin
levels
exacerbated
by
wastewater
have
the
potential
to
impact
both
human
and
ecosystem
health
(Rastogi
et
al.,
2014).
In
southeast
Florida,
the
St.
Lucie
Estuary
(SLE)
has
received
national
attention
and
has
been
the
subject
of
litigation
for
chronic
human
health
impacts
and
severely
degraded
ecosystem
health.
The
system
is
exceptional
in
its
anthropogenic
complexity,
reoccurrences
of
economically
and
ecologically
devastating
M.
aeruginosa
blooms,
frequent
health
advisories
for
high
fecal
bacteria
counts,
and
proximity
to
the
northern
extent
of
tropical
coral
species
along
the
east
coast
of
the
United
States
(Fig.
1A,
B).
Through
a
partnership
between
the
U.S.
Environmental
Protection
Agency
and
the
Florida
Department
of
Environmental
Protection
(FDEP),
the
SLE
has
been
identied
as
an
impaired
waterbody
and
Total
Maximum
Daily
Loads
have
been
established
for
total
nitrogen
(TN),
total
phosphorus
(TP),
dissolved
oxygen
(DO),
and
fecal
coliforms
(Parmer
et
al.,
2008;
White
and
Turner,
2012).
The
SLE
receives
freshwater
inputs
from
an
articially
large
watershed
as
the
result
of
a
network
of
canals
constructed
in
the
early
to
mid
1900s
to
alleviate
ooding
and
increase
development
potential
Fig.
1.
Ecosystem
responses
to
eutrophication
in
the
St.
Lucie
Estuary
(SLE)
and
nearshore
reefs:
(A,
B)
Microcystis
aeruginosa
in
the
SLE
with
40x
magnication
scale
bar
in
micrometers
(
m
m)
and
(C)
Clionid
sponge,
(D)
Codium
intertextum,
(E)
Dictyota
spp.,
(F)
four
common
species
of
sea
urchins
(Diadema
antillarum,
Tripneustes
ventricosus,
Echinometra
viridis,
Eucidaris
tribuloides)
along
the
nearshore
reefs.
Photo
credits:
(A,
B)
James
Sullivan,
(D)
Brian
Lapointe,
(C,
E,
F)
Laura
Herren.
2
B.E.
Lapointe
et
al.
/
Harmful
Algae
70
(2017)
122
(FDEP,
2009;
SFWMD,
FDEP,
FDACS,
2009).
Hydrological
alter-
ations
began
in
1924
when
the
South
Fork
of
the
SLE
was
connected
to
Lake
Okeechobee
via
the
C-44
canal
to
reduce
water
levels
in
the
lake
(Fig.
2;
Blake,
1980).
Massive
fresh
water
releases
from
Lake
Okeechobee
and
the
C-44
watershed
in
2005,
2013,
and
2016
lowered
the
salinity
in
the
estuary,
seeded
the
system
with
M.
aeruginosa,
and
ultimately
resulted
in
the
formation
of
three
unprecedented
M.
aeruginosa
blooms
that
extended
from
the
estuary
downstream
to
the
nearshore
reefs.
Consistent
with
the
literature,
the
TDN:TDP
ratios
during
each
of
these
blooms
were
<33
(Lapointe
et
al.,
2012;
Lapointe,
unpublished
data;
FDEP,
unpublished
data).
While
emphasis
has
been
placed
on
the
nutrient
inputs
from
Lake
Okeechobee
and
the
subsequent
ecological
impacts
the
additional
load
brings,
Lapointe
et
al.
(2012)
suggested
that
there
were
sufcient
local
nutrient
loads
from
the
SLE
watershed
itself
to
support
bloom
development
and
toxicity.
To
this
point,
d
15
N
values
of
M.
aeruginosa
samples
collected
in
the
mid-estuary
during
the
2013
and
2016
blooms
were
highly
enriched
(+8.6
o
/
oo
and
+7.0
o
/
oo
,
respectively)
compared
to
samples
from
the
C-44
canal
(<
6.0
o
/
oo
),
indicating
wastewater
N
as
a
primary
N
source
fueling
the
blooms
in
the
SLE.
Furthermore,
the
2005,
2013,
and
2016
blooms
were
conrmed
to
be
comprised
of
toxic
strains
of
M.
aeruginosa
(Ross
et
al.,
2006;
Phlips
et
al.,
2012;
Oehrle
et
al.,
2017),
ultimately
suggesting
that
there
were
high
enough
concentrations
of
nitrate
in
the
SLE
to
promote
toxin
production.
During
the
2016
bloom,
Oehrle
et
al.
(2017)
documented
that
most
(>85%)
of
the
total
microcystins
were
microcystin-LR,
a
form
that
was
found
at
concentrations
as
high
as
4500
mg/L.
The
World
Health
Organiza-
tion
drinking
water
and
recreational
water
contact
standards
are
set
at
1
mg/L
and
20
mg/L,
respectively.
Their
research
also
showed
that
toxin
concentrations
increased
from
the
C-44
(i.e.
seed
algae
from
Lake
Okeechobee
and
the
canal)
to
the
Middle
Estuary
where
high
concentrations
of
nitrate
from
the
South
Fork
and
ammonium
and
phosphorus
from
the
North
Fork
(Lapointe
et
al.,
2012)
converge
to
enrich
the
bloom.
With
such
high
nutrient
availability,
Microcystis
has
the
potential
to
double
its
biomass
in
approxi-
mately
two
days
(Nicklisch
and
Kohl,
1983;
Li
et
al.,
2014).
This
combination
of
a
toxic
gradient
and
nutrient
availability
provides
additional
evidence
that
signicant
bloom
development
has
been
occurring
in
the
SLE
itself
rather
than
the
upstream
seed
sources.
In
addition
to
phytoplankton-based
HABs,
degradation
of
nearshore
reefs
by
macroalgal
HABs
likewise
result
from
inputs
of
land-based
sources
of
N
and
P
(Lapointe
et
al.,
2005;
Littler
et
al.,
2006;
Lapointe
and
Bedford,
2010;
Lapointe
et
al.,
2011).
Along
the
east
coast,
the
Florida
Reef
Tract
extends
from
the
Florida
Keys
north
to
St.
Lucie
Inlet
in
Martin
County.
While
reefs
supporting
tropical
coral
species
are
exclusively
found
south
of
the
inlet,
Fig.
2.
Martin
County
watershed
to
reef
project
site
map
with
subsets
showing
the
Old
Palm
City
(OPC)
and
Golden
Gates
Estates
(GG)
residential
communities.
Analytes
measured
at
each
site
are
represented
by
shapes
as
indicated
by
the
key.
DN
=
dissolved
nutrients,
AI
=
aqueous
isotopes,
PF
=
phytoplankton,
SAV
=
submerged
aquatic
vegetation,
SUC
=
sucralose.
Martin
County
Department
of
Health
enterococcus
monitoring
stations
are
italicized;
LP
=
Leighton
Park,
RB
=
Roosevelt
Bridge,
SP
=
Sandsprit
Park,
and
SS
=
Stuart
Sandbar.
B.E.
Lapointe
et
al.
/
Harmful
Algae
70
(2017)
122
3
Sabellariid
wormrock
(Phragmatopoma
lapidosa)
reefs
span
both
north
and
south.
The
nearshore
reefs
adjacent
to
St.
Lucie
Inlet
are
exposed
to
eutrophic
water
discharged
from
the
Indian
River
Lagoon
(IRL),
SLE,
and
periodic
freshwater
releases
from
Lake
Okeechobee
located
west
of
the
SLE
(Lapointe
et
al.,
2012,
2015b).
Smith
(2016)
concluded
that
during
an
ebbing
tide
most
(!91%)
of
the
water
owing
out
of
the
SLE
continues
east
through
the
St.
Lucie
Inlet
ultimately
bathing
nearshore
reefs
while
the
other
!9%
ows
north
in
the
IRL
towards
Fort
Pierce
Inlet.
Land-based
discharges
to
these
nearshore
reefs
likely
contribute
to
biological
indicators
of
stress,
including
unusually
high
abundances
of
species
indicative
of
high
nutrient
environments
such
as
clinoid
(boring)
sponges,
macroalgal
blooms
(Codium
intertextum
and
Dictyota
spp.),
and
sea
urchins
(Herren
and
Monty,
2006;
Fig.
1C
F).
At
these
reef
sites,
corals
also
show
negative
physiological
responses
associated
with
changes
in
water
chemistry
and
light
associated
with
prolonged
releases
from
Lake
Okeechobee
(Beal
et
al.,
2012).
Thus,
to
improve
conditions
along
these
biodiverse
nearshore
reefs,
there
is
a
pressing
need
to
identify
and
subsequently
manage
upstream
nutrient
sources
(Lapointe
et
al.,
2012).
The
health
of
the
SLE
and
the
underlying
causes
of
its
impairment,
including
reoccurrence
of
toxic
HABs,
high
fecal
bacteria,
and
degradation
of
downstream
nearshore
reefs,
have
been
debated
for
decades.
One
emerging
issue
is
the
potential
nutrient
loading
associated
with
the
application
of
biosolids
(domestic
wastewater
residuals;
Tetra
Tech,
2017).
No
Class
AA
biosolids
are
produced
and
no
Class
B
biosolids
are
applied
in
Martin
County,
however,
the
practice
does
occur
in
other
counties
(but
not
signicantly)
along
the
IRL
(FDEP,
2014).
This
has
raised
concerns
as
a
potential
nutrient
source
supporting
HABs
and
introduction
of
pathogens
and
chemical
contaminants
into
surface
waters.
While
not
without
risk,
when
properly
treated
and
applied
this
method
of
recycling
waste
material
has
been
deemed
safe
for
both
humans
and
the
environment
(AMS,
2011).
There
has
also
been
an
increasing
interest
in
Florida
to
understand
the
often
overlooked
role
of
on-site
sewage
disposal
systems
(OSTDS;
septic
systems
and
shallow
injection
wells)
in
enrichment
and
microbial
contamination
of
shallow
groundwaters
and
adjacent
surface
waters
via
submarine
groundwater
discharge
(SGD;
Lapointe
et
al.,
1990,
2012;
Lapointe
and
Krupa,
1995a,b;
Paul
et
al.,
1995a,b;
Grifn
et
al.,
1999;
Tarnowski,
2014).
In
addition
to
SGD,
SLE
water
quality
is
also
affected
by
tidal
creeks
and
primary
canals
(C-44,
C-23,
and
C-24;
Fig.
2)
and,
in
turn,
the
sediments
that
can
sequester
incoming
N
and
P
from
these
sources
(Howes
et
al.,
2008;
Havens
et
al.,
2016).
The
C-23
and
C-24
deliver
water
solely
from
the
SLE
watershed.
Conversely,
inputs
from
the
C-44
originate
from
both
the
watershed
(C-44
basin)
and,
periodically,
Lake
Okeechobee
via
freshwater
releases
up
to
10,000
cfs
(Doering,
1996;
Lapointe
et
al.,
2012).
While
Lapointe
et
al.
(2012)
previously
documented
the
deleterious
effects
of
prolonged,
high-volume
releases
to
the
SLE,
the
20052006
study
simultaneously
indicated
local
septic
system
contributions.
Furthermore,
tidal
creeks
were
found
to
be
a
signicant
source
of
fecal
coliform,
total
coliform,
and
enterococcus
bacteria;
both
showing
high
to
low
count
gradients
from
upstream
(residential
areas)
to
downstream
(SLE).
The
human
fecal
source
marker
qPCR
Bacteroidales
HF183
was
also
documented
at
multiple
sites
throughout
the
SLE
during
a
2014
microbial
source
tracking
study
performed
by
FDEP
(2015d).
The
combination
of
sources
interacts
to
exacerbate
the
chronically
poor
water
quality
conditions
and
susceptibility
of
the
SLE
and
downstream
nearshore
reefs
to
HABs.
The
ability
to
distinguish
between
water
quality
impacts
from
the
groundwater
and
stormwater
runoff
derived
from
the
SLE
watershed
versus
impacts
directly
related
to
periodic
discharges
from
Lake
Okeechobee
is
an
important
water
management
issue
in
this
region.
Recent
water
and
nutrient
budgets
for
the
St.
Lucie
Estuary
indicate
that,
between
Water
Year
1997
and
2015,
about
30%
of
the
N
came
from
Lake
Okeechobee,
compared
to
70%
from
the
St.
Lucie
River
watershed
and
tidal
basin
(Zheng
et
al.,
2016).
Although
more
attention
has
been
given
to
the
contribution
of
freshwater
releases
from
Lake
Okeechobee
to
this
system,
it
is
also
important
to
understand
the
signicance
of
nutrients
(e.g.,
atmosphere,
fertilizers,
wastewater)
from
local
watersheds
to
achieve
nutrient
mitigation
for
the
SLE
and
the
downstream
ecosystems
(Badruzza-
man
et
al.,
2012).
When
combined,
multiple
analytical
approaches
can
provide
corroborative
evidence
of
eutrophication
and
nutrient
sources.
While
dissolved
nutrient
concentration
data
may
suggest
biologi-
cal
thresholds
for
nutrient
pollution,
articial
substances
con-
sumed
by
humans
and
evident
in
waste
streams
and
receiving
waters,
and
stable
isotope
analyses
of
water
and
algal
tissue
are
reliable
protocols
for
tracking
nutrient
sources.
For
example,
analysis
of
water
samples
for
the
articial
sweetener
sucralose
provides
an
indicator
of
human
wastewater.
Sucralose
is
not
broken
down
by
any
treatment
process
(including
the
body)
and
is
transported
conservatively
through
WWTPs
and
OSTDS
(Oppen-
heimer
et
al.,
2011).
Meanwhile,
aqueous
stable
isotopes
of
N,
both
d
15
N-NH
4+
and
d
15
N-NO
3"
,
can
be
utilized
to
discriminate
different
sources
of
dissolved
inorganic
N.
Previously,
Lapointe
et
al.
(2015a)
documented
stormwater
runoff
from
urban
and
agricultural
land
uses
within
the
SLE
watershed
with
depleted
mean
d
15
N
values
within
the
accepted
range
for
inorganic
fertilizers
("2
to
+2
o
/
oo
).
Similarly,
macroalgae
and
phytoplankton
are
commonly
used
indicator
organisms
for
assessment
of
the
relative
importance
of
N
sources
and
algal
tissue
nutrient
contents
may
indicate
the
degree
of
N
versus
P
limitation.
Macroalgae
are
especially
ideal
bio-
observatories
for
assessing
nutrient
availability
as
they
are
typically
attached
to
the
benthos
and
integrate
nutrient
availability
over
temporal
scales
of
days
to
weeks
(Lapointe,
1985).
Documen-
tation
of
stable
nitrogen
isotope
(d
15
N)
ratios
in
macroalgal
tissue
has
been
widely
used
to
discriminate
among
natural
(upwelling,
N-xation)
and
anthropogenic
(wastewater,
fertilizer)
nutrient
sources
(Risk
et
al.,
2009).
The
published
ranges
are
generally
as
follows:
natural
N-xation
(0
o
/
oo
;
Heaton,
1986;
France
et
al.,
1998),
offshore
upwelled
nitrate
(!+2.0
o
/
oo
;
Knapp
et
al.,
2008),
atmospheric
N
("3
o
/
oo
to
+1
o
/
oo
;
Paerl
and
Fogel,
1994),
and
synthetic
fertilizer
N
("2
o
/
oo
to
+2
o
/
oo
;
Bateman
and
Kelly,
2007).
All
of
these
N
sources
are
depleted
relative
to
enriched
values
of
+3
o
/
oo
to
+19
o
/
oo
for
human
wastewater
(Heaton,
1986;
Costanzo
et
al.,
2001)
and
+10
o
/
oo
to
+20
o
/
oo
for
livestock
waste
(Kreitler,
1975,
1979;
Heaton,
1986).
These
livestock
waste
values
depend
on
if
the
efuent
is
nitried,
or
not,
as
values
can
be
much
lower.
Nitrogen
in
OSTDS
efuent
is
primarily
in
the
form
of
ammonium
(Bicki
et
al.,
1984;
Lapointe
et
al.,
1990;
Valiela
et
al.,
1997)
with
d
15
N
values
of
+4
"
5
o
/
oo
(Lapointe
and
Krupa,
1995a,b;
Hinkle
et
al.,
2008;
Katz
et
al.,
2010),
but
through
ammonia
volatilization
and
microbial
processing
values
can
become
more
enriched
(i.e.
treated
wastewater).
Because
it
is
difcult
to
discern
overlapping
signatures
(i.e.
#+10
o
/
oo
)
one
must
consider
land
use
in
the
adjacent
watershed.
Accordingly,
enriched
macroalgae
d
15
N
values
>+3
o
/
oo
have
been
reported
in
a
wide
variety
of
sewage-polluted
coastal
waters,
including
Floridas
densely-developed
IRL
(Lapointe
et
al.,
2015b),
nearshore
reefs
off
urban
areas
of
east-
central
Florida
(Barile,
2004),
and
coastal
urban
areas
of
southeast
(Lapointe
et
al.,
2005)
and
southwest
(Lapointe
and
Bedford,
2007)
Florida.
In
addition
to
stable
isotope
analyses,
measurement
of
C:N:P
content
in
macroalgae
and
phytoplankton
provides
a
measure
of
nutrient
quantity
and
stoichiometry
that
is
useful
in
assessing
the
relative
importance
of
N-
versus
P-limitation
(Atkinson
and
Smith,
1983;
Lapointe
et
al.,
1992).
This
is
particularly
appropriate
for
4
B.E.
Lapointe
et
al.
/
Harmful
Algae
70
(2017)
122
assessing
OSTDS
groundwater-borne
sewage
pollution
that
can
deliver
nutrient
pollution
at
high
N:P
ratios
as
a
result
of
selective
adsorption
of
P
onto
soil
particles
(Bicki
et
al.,
1984;
Lapointe
et
al.,
1990;
Weiskel
and
Howes,
1992).
In
dense
residential
communities
relying
primarily
on
OSTDS,
high
cumulative
P
inputs
to
groundwater
can
supersaturate
the
soil
and
reduce
its
ability
to
selectively
adsorb
P
(Bicki
et
al.,
1984).
When
this
occurs,
groundwater
and,
ultimately,
surface
water
P
concentrations
become
higher,
thereby
lowering
the
N:P
ratio
and
increasing
M.
aeruginosa
bloom
potential
(Horst,
2014;
Horst
et
al.,
2014;
Parrish,
2014).
A
one-year
watershed
to
reef
study
was
designed
to
document
sources
of
nutrients
causing
eutrophication
in
the
SLE,
the
periodic
M.
aeruginosa
blooms
seeded
by
Lake
Okeechobee
discharges,
and
the
subsequent
downstream
decline
of
the
nearshore
Sabellariid
wormrock
and
coral
reefs.
The
study
was
a
comprehensive
analysis
of
the
interactions
of
septic
systems,
groundwaters,
and
surface
waters
that
included:
1)
groundwater
sampling
in
two
residential
areas
identied
as
high
priority
septic
to
sewer
conversion
sites
and
undeveloped
reference
sites
for
dissolved
nutrients,
aqueous
N
isotopes,
and
sucralose;
2)
surface
water
sampling
in
the
C-44
and
C-23
canals,
throughout
the
SLE,
and
nearshore
reefs
for
dissolved
nutrients
and
sucralose;
and
3)
collection
and
analysis
of
macro-
phytes
and
phytoplankton
for
stable
isotope
(d
13
C
and
d
15
N)
analysis
and
algal
tissue
elemental
composition
(C:N:P).
2.
Materials
and
methods
Located
on
Floridas
east
coast,
the
SLE
is
the
largest
tributary
to
the
251
km
long
IRL.
The
upper
SLE
includes
the
lower-salinity
North
and
South
forks
which
converge
to
form
the
higher-salinity
Middle
Estuary.
The
Middle
Estuary
ows
east
through
the
Lower
Estuary
to
the
IRL
and
out
to
tide
through
St.
Lucie
Inlet
(Fig.
2).
There
are
three
primary
canals
(C-44,
C-23,
and
C-24)
that
drain
directly
into
the
SLE
that
have
ultimately
increased
the
natural
extent
of
the
SLE
watershed.
Four
SLE
watershed
to
reef
sampling
events
were
performed
during
ebbing
tides
in
2015;
April
710,
17,
20
(Dry
1),
May
1114
(Dry
2),
August
57
(Wet
1),
September
21,
23,
24
(Wet
2).
During
each
event,
11
groundwater,
18
surface
water,
and
8
macroalgae
sampling
stations
within
primary
canals,
the
SLE,
and
nearshore
reefs
were
visited,
each
with
unique
parameters
of
interest
(Fig.
2).
Upstream
to
downstream
sampling
networks
were
created
within
and
adjacent
to
two
older
(pre-1978)
residential
communities
previously
identied
as
high-priority
septic
to
sewer
conversion
sites
(Keene,
2015;
Lapointe
et
al.,
2016).
In
Old
Palm
City
(OPC),
xed
ground
and
surface
water
sites
were
selected
along
All
American
Ditch
(OPC1-3)
and
the
culvert-connected
tidal
creek
draining
into
the
South
Fork
(OPC4-5;
Fig.
2).
In
Golden
Gates
Estates
(GG),
xed
ground
and
surface
water
sites
were
selected
in
and
adjacent
to
the
community
retention
pond
(GG1-2)
down-
stream
to
the
conuence
of
Willoughby
Creek
and
the
Lower
Estuary
(GG5;
Fig.
2).
Sites
were
also
selected
upstream
of
the
water
control
structures
on
the
C-44
and
C-23
(C44W
and
C23W,
respectively)
and
on
the
nearshore
reefs
north
(BTR)
and
south
(SLR-N
and
SLR-S)
of
St.
Lucie
Inlet.
To
provide
multiple
lines
of
evidence,
several
analyses
were
conducted
on
samples
collected
during
each
of
the
four
sampling
events
(Table
1).
The
University
of
Georgias
Center
for
Applied
Isotope
Studies
Stable
Isotope
Ecology
Laboratory
(UGA-SIEL)
in
Athens,
Georgia
conducted
all
analyses
except
sucralose.
Concentrations
of
this
articial
sweetener
were
analyzed
by
Florida
Department
of
Environmental
Protections
Central
Laboratory
in
Tallahassee,
Florida.
2.1.
Freshwater
inputs
and
enterococcus
bacteria
counts
Freshwater
inputs
to
the
SLE
system,
including
rainfall
and
discharges
through
the
C-44,
C-23,
and
C-24
canals
and
bacterial
counts
at
four
sites
within
the
study
area
were
followed
throughout
the
study.
2.1.1.
Rainfall
Rainfall
data
were
downloaded
from
the
National
Oceanic
and
Atmospheric
Administration
National
Centers
for
Environmental
Information
(http://www.ncdc.noaa.gov/data-access)
for
the
du-
ration
of
the
project
(January
1September
30,
2015).
The
station
(GHCND:US1FLMT0018,
STUART
1.0
ESE
FL
US)
was
centrally
located
(27.1883,
!80.2279)
within
the
study
area;
just
north
of
the
airport
in
Stuart,
Florida.
To
ensure
a
complete
dataset,
missing
data
were
obtained
from
U.S.
Climate
Datas
station
Stuart
1
s
(http://www.usclimatedata.com/climate/stuart/orida/united-
states/us0468)
located
1
km
west
of
the
above
station
(27.1897,
!80.2397).
Daily
total
precipitation
(mm)
was
plotted
to
indicate
seasonal
rainwater
inputs
relative
to
sampling
events.
2.1.2.
Canal
discharges
Discharge
(ow)
rates
from
the
water
control
structures
nearest
to
the
SLE
along
the
C-44,
C-23,
and
C-24
canals
were
downloaded
from
South
Florida
Water
Management
Districts
(SFWMD)
online
database
DBHYDRO.
Flow
data
for
the
S-80
structure
(Key:
DJ238)
at
the
conuence
of
the
C-44
canal
and
the
South
Fork
were
monitored
by
the
U.S.
Army
Corps
of
Engineers
(USACE),
data
for
the
S-48
structure
(Key:
JM106)
on
eastern
end
of
the
C-23
canal
near
the
conuence
of
the
north
and
south
forks
of
the
system
were
monitored
by
the
SFWMD,
and
data
for
the
S-49
structure
(Key:
JW223)
along
the
east
end
of
the
C-24
canal
that
empties
into
the
North
Fork
were
also
monitored
by
SFWMD.
Data
were
obtained
for
January
1September
30,
2015
and
separated
into
Dry
(January
1
to
May
31,
2015)
versus
Wet
(June
1
to
September
30,
2015)
seasons.
The
USACE
and
SFWMD
were
informed
of
the
study
and
requests
were
made
by
Martin
County
to
stop
releases
during
the
sampling
events.
2.1.3.
Fecal
bacteria
distribution
and
abundance
Martin
County
Department
of
Health
(DOH)
provided
entero-
coccus
bacteria
count
data
collected
between
January
1
and
September
30,
2015.
Counts
(number
of
colony-forming
units
[cfu]/100
mL
river
water)
were
reported
for
four
monitoring
stations:
1)
Roosevelt
Bridge,
2)
Sandsprit
Park,
3)
Leighton
Park,
and
4)
the
Stuart
sandbar
(near
Sailsh
Point
and
St.
Lucie
Inlet).
Table
1
Analytes
measured
during
the
2015
Martin
County
watershed
to
reef
study.
General
Analysis
Media
Collected
(#
Stations)
Analytes
dissolved
nutrients
ground
(11),
surface
(18)
water
ammonium,
nitrate,
phosphate,
total
nitrogen,
total
phosphorus
stable
aqueous
isotopes
groundwater
(11)
d
15
N-ammonium,
d
15
N-nitrate
dissolved
articial
substances
ground
(11),
surface
(8)
water
sucralose
stable
tissue
isotope
macroalgae
(8),
phytoplankton
(18)
d
13
C,
d
15
N
elemental
tissue
composition
macroalgae
(8),
phytoplankton
(18)
C:N,
C:P,
N:P
B.E.
Lapointe
et
al.
/
Harmful
Algae
70
(2017)
122
5
The
water
quality
scale
based
on
count
data
set
by
Florida
DOH
includes
the
following
categories,
good
(0
to
35
cfu/100
mL),
moderate
(36
to
70
cfu/100
mL),
and
poor
(!71
cfu/100
mL).
2.2.
Groundwater
To
clarify
the
potential
effects
of
OSTDS
on
surface
waters
in
the
SLE
and
downstream
nearshore
reefs,
nine
wells
were
installed
at
two
residential
sites;
OPC
and
GG
as
described
by
Lapointe
and
Herren
(2016);
Fig.
2).
The
OPC
well
cluster
consisted
of
two
shallow
(3.7
m),
one
intermediate
(7.4
m),
and
one
deep
(17.5
m)
well
behind
a
single-family
residential
home
along
All
American
Ditch
(Fig.
2).
The
GG
well
complex
included
three
shallow
(3.7
m),
one
intermediate
(7.3
m),
and
one
deep
(17.4
m)
well
behind
a
duplex
adjacent
to
the
community
retention
pond
(Fig.
2).
Two
existing
reference
or
control
wells,
one
managed
by
SFWMD
(PCP-
C)
and
the
other
by
the
Martin
County
Utilities
Department
(W4B),
were
also
incorporated
into
the
study
to
investigate
anthropogenic
effects
of
residential
septic
systems
on
water
quality.
PCP-C
(9.0
m)
was
located
at
Pendarvis
Park
along
the
South
Fork
and
W4
B
(14.9
m)
was
installed
along
Jensen
Beach
Boulevard
adjacent
to
the
Savannas
Preserve
State
Park
in
Jensen
Beach
(Fig.
2).
Because
of
the
high
degree
of
mixing
between
the
surface,
intermediate,
and
deep
wells
(especially
the
two
former),
results
were
pooled
and
presented
by
community
(OPC,
GG)
and
reference
(Pendarvis
Park,
Jensen
Beach
Boulevard).
Protocols
for
well
purging
and
stabilization
prior
to
sample
collection
were
outlined
in
Lapointe
and
Herren
(2016).
After
stabilization
criteria
were
met,
ground-
water
samples
were
collected
using
the
following
procedures.
2.2.1.
Groundwater
dissolved
nutrients
and
aqueous
N
isotopes
Dissolved
nutrient
and
aqueous
N
isotopes
samples
were
collected
from
the
same
pump
after
a
0.45
mm
high
capacity
cartridge
lter
was
tted
to
the
discharge
tube
of
the
pump.
A
new
cartridge
lter
was
used
for
each
well.
From
each
well,
three
1
L
replicates
for
both
forms
of
N
aqueous
isotopes
(d
15
N-NH
4+
and
d
15
N-NO
3"
)
were
collected
in
a
total
of
six
acid
washed
high-density
polyethylene
(HDPE)
bottles.
Samples
were
placed
in
wet
ice
lled
insulated
coolers
to
chill,
and
then
maintained
at
temperatures
below
4
#
C
through
shipping.
The
coolers,
containing
completed
chain-of-custody
forms,
were
sealed
and
shipped
to
the
UGA-SIEL
for
analysis.
Once
received,
the
samples
were
immediately
frozen.
Prior
to
analysis,
the
samples
were
thawed,
homogenized,
and
$100
mL
of
sample
was
removed
from
three
of
the
six
1
L
bottles
for
dissolved
nutrients
analysis.
The
remaining
water
was
analyzed
for
aqueous
N
isotopes.
For
dissolved
nutrients,
the
$100
mL
subsamples
were
divided
for
either
persulfate
digestion
(TN/TP)
or
direct
analyses
(NO
x
,
NH
4+
,
and
soluble
reactive
phosphorus
[SRP
or
PO
43"
]).
To
digest
TN/TP,
5
mL
of
sample
were
digested
with
1
mL
persulfate
reagent,
autoclaved
until
all
N
was
oxidized
to
nitrate
and
all
P
was
oxidized
to
orthophosphate
per
Koroleff
(1983)
methods
as
modied
by
Qualls
(1989)
(UGA-SIEL,
2015a).
Once
digested,
all
nutrient
forms
(NH
4+
,
NO
x
,
SRP,
TN,
and
TP)
were
analyzed
on
an
Alpkem
300
series
nutrient
autoanalyzer
using
EPA
standard
methods
(4500-
NH
3
G,
4500-NO
3"
F,
and
4500-P
F).
To
analyze
aqueous
N
isotopes
,
UGA-SIEL
ran
the
water
samples
through
ammonia
diffusion,
which
involved
increasing
the
pH
of
the
dissolved
sample,
converting
the
ammonium
to
gaseous
ammonia,
which
is
captured
on
an
acidied
lter
in
the
bottle
headspace.
Nitrate-specic
N
was
quantied
by
rst
boiling-off
the
volatile
ammonia,
adding
a
reducing
agent
to
convert
oxidized
N
to
NH
4+
,
then
proceeding
with
the
standard
diffusion
and
ammonia
capture
on
an
acidied
lter.
The
lter
was
then
analyzed
as
a
typical
solid
sample
on
a
Carlo
Erba
Isotope
Ratio
Mass
Spectrometer
(IRMS)
for
d
15
N-NH
4+
and
d
15
N-NO
3"
.
2.2.2.
Groundwater
sucralose
A
single
1
L
amber
glass
bottle
provided
by
the
FDEP
Central
Laboratory
was
lled
directly
from
the
peristaltic
pump
used
to
purge
each
of
the
11
wells.
In
addition
to
these
11
unique
samples,
associated
eld
blank
and
duplicate
samples
were
collected
during
each
of
the
four
sampling
events.
All
samples
were
placed
in
insulated
coolers
with
wet
ice
to
chill
and
then
maintained
at
temperatures
below
4
#
C
through
shipping.
The
coolers,
with
accompanying
chain
of
custody
forms,
were
returned
overnight
to
the
FDEP
Central
Laboratory
the
same
day
that
the
samples
were
collected.
At
the
laboratory,
water
samples
were
ltered
through
a
0.75
mm
glass
ber
(GF/F)
lter.
A
250
mL
aliquot
of
ltered
sample
was
then
passed
through
a
graphitized
carbon-based,
solid-phase
extraction
(SPE)
column.
After
extraction,
the
absorbed
analytes
were
eluted
from
the
SPE
column
with
a
mixture
of
80%
methylene
chloride:
20%
methanol.
The
extract
was
reduced
to
near
dryness
and
brought
to
a
nal
volume
of
1
mL
with
10%
acetonitrile
in
deionized
water.
Analytical
standards
were
prepared
in
the
same
nal
solvent
mixture.
Each
extract
was
analyzed
by
high
performance
liquid
chromatography/tandem
mass
spectrometry
in
the
negative
ion
mode
for
the
determination
of
sucralose
(method
detection
limits
of
0.01
mg/L).
This
analytical
procedure
is
based
on
U.S.
Geological
Survey
(USGS)
method
O-206001
(Furlong
et
al.,
2001)
and
its
details
are
described
in
FDEP
standard
operating
procedure
LC-001/LC011
2
located
on
their
website
(http://www.dep.state..us/labs/library/lab_sops.htm).
2.3.
Surface
water
and
algal
tissue
Surface
water
and
algal
tissue
samples
were
collected
by
Harbor
Branch
Oceanographic
Institute
at
Florida
Atlantic
University
(HBOI-FAU)
and,
with
the
exception
of
the
sucralose
samples,
processed
the
same
day
according
to
the
parameter
of
interest
at
the
HBOI-FAU
HAB
Laboratory.
2.3.1.
Surface
water
dissolved
nutrients
Surface
water
samples
were
collected
in
triplicate
just
below
the
surface
into
acid-washed
250
mL
HDPE
bottles
and
covered
with
ice
in
a
dark
cooler
until
return
to
the
laboratory
for
processing.
The
samples
were
ltered
(0.7
mm
GF/F
lters)
and
frozen
until
analysis.
At
UGA-SIEL,
samples
were
thawed,
homogenized,
and
subsampled
for
either
persulfate
digestion
(TN/TP)
or
direct
analyses
(NO
x
,
NH
4+
,
and
PO
43"
)
as
mentioned
for
groundwater
in
2.2.1.
The
resulting
data
were
used
to
characterize
ambient
dissolved
inorganic
N
(DIN),
TN
and
TP
concentrations,
DIN:SRP
ratios,
and
total
dissolved
N
(TDN):total
dissolved
P
(TDP)
ratios
at
the
18
surface
water
collection
sites.
Calibrated
YSI
Model
1030
and
ProODO
hand-held
meters
were
used
to
document
salinity,
temperature,
DO,
and
pH
at
the
time
each
water
sample
was
collected.
2.3.2.
Surface
water
sucralose
Unique
surface
water
samples
were
collected
for
sucralose
at
four
sites
in
both
OPC
and
GG.
Each
of
the
samples
were
collected
in
1
L
amber
bottles
and
kept
on
ice
throughout
overnight
shipment
to
the
FDEP
Central
Laboratory.
Processing
of
the
surface
water
samples
was
identical
to
the
groundwater
samples
described
above
in
Section
2.2.2.
2.3.3.
Macroalgal
tissue
d
13
C,
d
15
N
and
C:N:P
Triplicate
samples
of
macroalgae
were
collected
at
eight
submerged
aquatic
vegetation
sampling
stations.
Because
of
the
inconsistent
presence
of
macroalgae
in
some
sections
of
the
SLE,
especially
in
the
upper
reaches,
cages
were
deployed
for
4
weeks
to
hold
replicates
of
the
red
alga
Gracilaria
tikvahiae
provided
by
the
Marine
Botany
Laboratory
at
HBOI-FAU.
Once
the
algae
were
either
6
B.E.
Lapointe
et
al.
/
Harmful
Algae
70
(2017)
122
collected
from
the
natural
environment
or
the
deployed
cages,
the
samples
were
cleaned
of
epiphytes
and
debris,
<