ArticlePDF Available

Shift in baseline chlorophyll a concentration following a three-year Synechococcus bloom in southeastern Florida

Authors:

Abstract and Figures

A picophytoplankton bloom dominated by Synechococcus formed in September 2005 in a series of shallow lagoons between Florida Bay and Biscayne Bay and lasted until May 2008. Chlorophyll a concentrations peaked at >20 μg L⁻¹. The bloom coincided with a massive mortality of sponges and caused massive mortality of the seagrass. However, follow-up analysis to determine if there were any long-term impacts from the bloom on the system is lacking. We used long-term water quality data (chlorophyll a and nutrient concentrations) collected at 13 stations in the affected region over a 20-yr period to compare environmental conditions before (1995-2004) and after (2009-2014) the bloom. We found that after the bloom, baseline chlorophyll a concentration significantly increased 45%, from 0.42 (SE 0.02) to 0.77 (SE 0.04) μg chl a L⁻¹, at the stations most impacted by the bloom. Before-After Control-Impact paired analysis suggested these changes were related to the 3-yr bloom and not a larger, regional scale shift. The increase in chlorophyll a does not appear to be associated with additional changes in water quality, but is potentially due to a reduction in the epibenthic community (e.g., SAV and sponges). Now that the bloom has terminated and the causes of the bloom abated, the system has not returned to its original status, suggesting a lasting impact from the bloom on the ecosystem. © 2018 Rosenstiel School of Marine & Atmospheric Science of the University of Miami.
Content may be subject to copyright.
Bull Mar Sci. 94(1):000–000. 2018
https://doi.org/10.5343/bms.2017.1046
1
Bulletin of Marine Science
© 2018 Rosenstiel School of Marine & Atmospheric Science of
the University of Miami
Shift in baseline chlorophyll a concentration following a
three-year Synechococcus bloom in southeastern Florida
NC Millette 1, 2, 3 *
C Kelble 3
A Linhoss 4
S Ashby 2
L Visser 3, 5
ABS TR AC T.A picophytoplankton bloom dominated
by Synechococcus formed in September 2005 in a series
of shallow lagoons between Florida Bay and Biscayne Bay
and lasted until May 2008. Chlorophyll a concentrations
peaked at >20 µg L−1. e bloom coincided with a massive
mortality of sponges and caused massive mortality of the
seagrass. However, follow-up analysis to determine if there
were any long-term impacts from the bloom on the system is
lacking. We used long-term water quality data (chlorophyll a
and nutrient concentrations) collected at 13 stations in the
affecte d region over a 20-yr period to compa re environmental
conditions before (1995–2004) and after (2009–2014) the
bloom. We found that after the bloom, baseline chlorophyll a
concentration sig nificantly increased 45%, f rom 0.42 (SE 0.02)
to 0.77 (SE 0.04) µg chl a L−1, at the stations most impacted
by the bloom. Before-After Control-Impact paired analysis
suggested these changes were related to the 3-yr bloom and
not a larger, regional scale shift. e increase in chlorophyll
a does not appear to be associated with additional changes
in water quality, but is potentially due to a reduction in the
epibenthic community (e.g., SAV and sponges). Now that the
bloom has terminated and the causes of the bloom abated,
the system has not returned to its original status, suggesting
a lasting impact from the bloom on the ecosystem.
In September 2005, a cya nobacteria bloom of Synechococcus formed in the Manatee
Bay and Barnes Sound region of south Florida (Rudnick et al. 2007, Glibert et al.
2009). e bloom, dominated by three clades of Synechococcus—WH8101, CB0201,
and RS9708—lasted 3 yrs, and spanned from Duck Key in Florida Bay to Card Sound
in southern Biscayne Bay (Fig. 1) (Rudnick et al. 2007, Glibert et al. 2009). A combi-
nation of three hurricanes over a 3-mo period (Katrina, Rita, and Wilma) and major
construction on an adjacent causeway connecting the mainland to the Florida Keys
is hypothesized to have triggered the initiation of the bloom (Rudnick et al. 2007,
Glibert et al. 2009). e main factor limiting phytoplankton growth in this region
1 Mississippi State University,
Starkville, Mississippi 39762.
2 Northern Gulf Institute,
Mississippi State University,
Stennis Space Center, Mississippi
39529.
3 National Oceanic and
Atmospheric Administration,
Atlantic Oceanographic and
Meteorological Laboratory,
Miami, Florida 33149.
4 Agricultural and Biological
Engineering, Mississippi State
University, Starkville, Mississippi
39762.
5 Cooperative Institute for
Marine and Atmospheric
Studies, University of Miami,
Miami, Florida 33149.
* Corresponding author email:
<nicole.millette@noaa.gov>.
Date Submitted: 7 April, 2017.
Date Accepted: 12 October, 2017.
Available Online: 14 November, 2017.
research paper
Fas t Track
publication
Bulletin of Marine Science. Vol 94, No 1. 20182
is phosphorus (Koch et al. 2001, Nielsen et al. 2006) and the three hurricanes and
construction likely released a large concentration of phosphorus through increased
run-off from land and disturbance of the sediment, stimulating rapid phytoplankton
growth (Glibert et al. 2009). For a full description on the initiation, duration, and
termination of this Synechococcus bloom, see Glibert et al. (2009).
e region of Manatee Bay, Blackwater Sound, Barnes Sound, and Long Sound
(Fig. 1) is where the cyanobacteria bloom had the largest and longest impact (Glibert
et al. 2009). e bloom was a dramatic deviation from the normal characteristics of
this region. Typically, this region has lower mean chlorophyll a concentrations (<2 µg
chl a L−1) than the adjacent ecosystem of Florida Bay (Phlips et al. 1999). Diatoms and
dinoflagellates were the dominant phytoplankton groups in this region in the 1990s
(Phlips et al. 1999); however, phytoplankton species data just prior to the bloom were
Figure 1. Location of water quality stations in Manatee Bay, Barnes Sound, Blackwater Sound,
Long Sound, and Biscayne Bay. Stations in Biscayne Bay were used as controls for some analysis.
Millette et al.: Shift in baseline following Synechococcus bloom 3
not available. e highest peak in chlorophyll a concentrations (>20 µg chl a L−1) oc-
curred in Blackwater and Barnes Sound, an order of magnitude higher than average
concentrations (Glibert et al. 2009). It is conceivable that following persistent bloom
conditions for 3 yrs, the study region never fully recovered.
e bloom was eventually terminated around May 2008 (Glibert et al. 2009), but
the lasting impacts of the bloom have never been investigated. ere is some evi-
dence that Synechococcus is toxic, but its toxicity is poorly understood (Cox et al.
2005, Martins et al. 2007, Hamilton et al. 2014). However, there are many ways a
large phytoplankton bloom can negatively impact an ecosystem, regardless of its tox-
icity, as synthesized by Paerl et al. (2001). During the initiation of the bloom, there
was a mass mortality of sponges and following the initiation, high mortality of sea-
grass and benthic macroalgae were recorded (Rudnick et al. 2008). Such mortality
can help to shift the main source of primary production in an ecosystem from the
benthic community to the pelagic community and further fuel the bloom (Glibert
et al. 2009). It was unknown if the system fully shifted back to being dominated by
benthic production after the bloom.
Regime shifts from one stable state to another can occur when an environment ex-
periences a loss of resilience (Folke et al. 2004). e more external stressors added to
a system, the more resilience that is lost and the easier it is to exceed a tipping point
into a new stable state (Folke et al. 2004). e shift from a benthic submersed aquatic
vegetation (SAV) dominant state to a pelagic phytoplankton dominant state is known
to occur in shallow lagoons (Gunderson 2001). Large phytoplankton blooms cause
light limitation for SAV communities, which can result in SAV mortality (Kleppel et
al. 1996, Glibert et al. 2014). us, a phytoplankton bloom can cause a tipping point
to be exceeded and result in hysteresis. Hysteresis occurs when an ecosystem does
not return to its initial stable state after one or more environmental stressors are
reduced, and enters into an alternative stable state. e 3-yr bloom caused massive
SAV mortality (Rudnick et al. 2008), but prior to our study, it was unknown whether
phytoplankton abundances returned to prebloom conditions or a potential regime
shift occurred.
We hypothesized that mean phytoplankton abundances increased following the
bloom, signaling a shift towards a more phytoplankton-dominant state. We investi-
gated whether water quality returned to prebloom conditions in the 5 yrs following
the bloom. Here, we use water quality to refer to nutrient concentrations, specifically
nitrate + nitrite (NOx) and phosphate, NOx:PO4
3− ratio, and chlorophyll a concentra-
tions. Additionally, we looked for any baseline shifts in temperature and salinity.
For shifts detected, we determined if they were related to the bloom or to regional
processes. Lastly, we explored potential mechanistic causes of any shifts; specifically
examining if they were related to other changes in the ecosystem, either in the water
column or benthos.
M
L.—Manatee Bay, Blackwater Sound, Barnes Sound, and Long Sound are
shallow (<3 m detph) oligotrophic lagoons located between Biscayne Bay and Florida
Bay (Fig. 1). e influence of the C-111 canal on this region results in higher inorgan-
ic nitrogen and lower inorganic phosphorus concentrations compared to Florida Bay
Bulletin of Marine Science. Vol 94, No 1. 20184
(Boyer et al. 1997). e low phosphorus concentrations limit phytoplankton growth
and, therefore, chlorophyll a concentrations (Boyer et al. 1997, Glibert et al. 2009).
S.—ere are 16 water qualit y stations throug hout Manatee Bay, Blackwater
Sound, Barnes Sound, and Long Sound that are sampled by three different agen-
cies, the National Oceanic and Atmospheric Administration (NOAA), Miami-Dade
County’s Department of Environmental Resources Management (DERM), and the
South Florida Water Management District (SFWMD). For the purpose of our study,
we selected 13 stations to analyze based on the availability of sufficient chlorophyll a
data before and after the 2005–2008 bloom (Table 1). e parameters collected and
methods used to collect data at each station varied based on which agency sampled
the station; therefore, we analyzed only parameters consistently collected at most
Table 1. The name, location in Florida Bay, agency in charge of collecting data, and the range of years
for which data was available at each station used in the analysis. AOML = Atlantic Oceanographic
and Meteorological Laboratory, NOAA = National Oceanic and Atmospheric Administration,
DERM = Miami-Dade County’s Department of Environmental Resource Management, SFWMD
= South Florida Water Management District. See Figure 1 for station locations.
Station name Region Collection agency Years of data
B11 Barnes Sound AOML/NOAA 2002–2014
B12 Manatee Bay AOML/NOAA 2002–2014
B13 Blackwater Sound AOML/NOAA 2002–2014
BB50 Barnes Sound DERM 1995–2014
BB51 Manatee Bay DERM 1995–2014
FLAB01 Barnes Sound SFWMD 1995–2014
FLAB02 Barnes Sound SFWMD 1995–2014
FLAB03 Manatee Bay SFWMD 1995–2014
FLAB04 Barnes Sound SFWMD 1995–2014
FLAB05 Blackwater Sound SFWMD 1995–2014
FLAB06 Little Blackwater Sound SFWMD 1995–2014
FLAB07 Long Sound SFWMD 1995–2014
FLAB08 Long Sound SFWMD 1995–2014
Table 2. A list of the instruments and laboratory methods used to collect and analyze the data collected by
Atlantic Oceanographic and Meteorological Laboratory (AOML), South Florida Water Management District
(SFWMD), and Department of Environmental Resources Management of Miami-Dade County, Florida
(DERM) and used in the present study. Temperature and salinity were both measured in situ with various
instruments. Chlorophyll a (Chl a), NOx, and PO4
3− concentrations were collected in situ and analyzed in the
laboratory with a range of methods.
Agency Temperature Salinity Chl aNOxPO4
3−
AOML Seabird SBE 21
TSG
Seabird SBE 21
TSG
Filtration extraction
using a 60:40 mixture
of acetone and
dimethyl sulde (Shoaf
and Lium 1976)
EPA Methods
353.4
EPA Methods
365.5
SFWMD Hydrolab
multiparameter
sonde
Hydrolab
multiparameter
sonde
Filtration extraction
using a 90% acetone
(Strickland and Parson
1972)
Alpkem model
RFA 300 (Caccia
and Boyer 2007)
Alpkem model
RFA 300 (Caccia
and Boyer 2007)
DERM Hydrolab Sonde
Model 3 and 4
(until 2006)
YSI 600 xlm
(after 2006)
Hydrolab Sonde
Model 3 and 4
(until 2006)
YSI 600 xlm
(after 2006)
Filtration extraction
using a 90% acetone
(EPA Methods 445.0)
N/A N/A
Millette et al.: Shift in baseline following Synechococcus bloom 5
stations with no apparent bias between different methods: temperature, salinity,
chlorophyll a, nitrate + nitrite, and soluble reactive phosphate (Table 2). Additionally,
for each sampling point we calculated a NOx:PO4
3− ratio using nitrate +nitrite and
soluble reactive phosphate concentrations. DERM did not report nutrient data for
stations BB50 and BB51, but these stations were included because they are near sta-
tions FLAB04 and FLAB03, respectively (Fig. 1), which did report nutrient data.
P-  P.—For our analysis, we defined prebloom as January 1995
to December 2004 and postbloom as January 2009 to December 2014. We began our
analysis in 1995 because there was a shift in the Atlantic multi-decadal oscillation
around 1995 that significantly shifted chlorophyll a in south Florida coastal waters,
including our study domain (Briceño and Boyer 2010). Ten of the 13 stations had data
starting in 1995; the remaining three stations (B11, B12, and B13) had data starting
in 2002 (Table 1). e bloom started in mid–late 2005, and continued until early-mid
2008 (Glibert et al. 2009). We defined the bloom period as January 2005–December
2008 to ensure the prebloom period was before the system experienced any changes
leading up to the 3-yr bloom and that the postbloom period was after the region had
sufficient time to recover following termination of the bloom.
A.—Initially, the data from each month for all stations were averaged to-
gether for preliminary analysis. We then broke the stations into two groups based
on whether chlorophyll a at the individual stations returned to prebloom concentra-
tions or not following the termination of the bloom. Environmental parameters can
be highly variable over space and time, especially in coastal systems. Averaging data
from multiple stations over space and time helps to reduce the variability and iden-
tify the dominant changes throughout the entire sample area. Analyzing each station
individually acknowledges that not every station is responding the same to changes
in the system. en, through grouping stations together based on whether chloro-
phyll a concentrations returned to prebloom concentrations or not, we were able to
compare changes to other environmental factors between the two station groups.
Unequal variance t-tests were used to compare the average pre- and postbloom
values for each environmental factor. A 12-mo moving average was used to trans-
form the data before estimating the rate of change via linear regression for each fac-
tor pre- and postbloom. e moving average was calculated by averaging the 6 mo
prior and 6 mo following each time point. A 12-mo moving average helps remove
seasonal cycles and random components of a data set to facilitate analysis of overall
trends (OECD 2007). e rate of change was estimated using the slope from linear
regression analyses. While postbloom rates of change could be compared between
altered stations and unaltered stations, there were insufficient years of data to detect
an accurate long-term trend (Meals et al. 2011).
Before-After Control-Impact (BACI) analysis was employed to examine whether
changes observed in the region affected by the 3-yr bloom could be attributed to the
bloom vs regional scale changes (Smith 2002). e control group for this analysis
was six water quality stations in southern Biscayne Bay (Fig. 1; SFWMD: BISC113,
BISC122, and BISC123; and NOAA: B7, B8, and B14). Manatee Bay and Barnes Sound
connect with southern Biscayne Bay through Card Sound (Wang et al. 2003) (Fig.
1). ese stations are north of the study region and were selected because they were
close to Manatee Bay and Barnes Sound, had enough data before and after the bloom,
and showed no indication of having been affected by the bloom. e monthly data
at each station were averaged, as described above, to create a single control group.
Bulletin of Marine Science. Vol 94, No 1. 20186
R
M C a.—Before we transformed the data using a 12-mo moving
average, we compared the chlorophyll a concentrations prebloom and postbloom at
each station. Chlorophyll a concentrations at stations B11, B13, BB50, BB51, FLAB01,
FLAB02, FLAB03, FLAB04, and FLAB05 were all significantly higher postbloom
compared to prebloom (P < 0.05, Fig. 2). Chlorophyll a concentrations at stations
B12, FLAB06, FLAB07, and FLAB08 were not significantly different postbloom com-
pared to prebloom (P > 0.05, Fig. 2). us, for all further analyses we combined the
stations into two different groups, stations where there was a significant increase in
chlorophyll a concentrations postbloom (altered group) and stations where there was
not a significant change (unaltered group).
e mean chlorophyll a concentrations for the altered group were 0.42 (SE 0.02) µg
chl a L−1 prebloom compared to 0.77 (SE 0.04) µg chl a L−1 postbloom (Table 3). e
mean chlorophyll a concentrations for the unaltered group 0.63 (SE 0.03) µg chl a L−1
per-bloom compared to 0.67 (SE 0.04) µg chl a L−1 postbloom (Table 3). Before the
bloom, chlorophyll a concentration for the unaltered group was significantly higher
than that for the altered group (unequal variance t-test: P < 0.001, n = 108). Following
the bloom, there was no significant difference in the chlorophyll a concentrations
between the two groups (unequal variance t-test: P = 0.10, n = 60).
Before the bloom, chlorophyll a concentrations at altered and unaltered stations
were stable over the 10-yr period (Fig. 3A, B). After the bloom, chlorophyll a at unal-
tered stations continued to be stable (Fig. 3B). At altered stations, there was a weak
significant increase in chlorophyll a over the 5-yr period (Fig. 4A).
e BACI analysis indicated that the difference between chlorophyll a concen-
trations at altered stations and the control group was significantly higher after the
bloom [+0.31 (SE 0.06) µg chl a L−1] compared to before the bloom [+0.13 (SE 0.01) µg
chl a L−1] (unequal variance t-test: P = 0.004, n = 108, Fig. 4). However, the difference
between chlorophyll a concentrations at unaltered stations and the control group
was not significantly different after the bloom [+0.21 (SE 0.06) µg chl a L−1] compared
to before the bloom [+0.34 (SE 0.03) µg chl a L−1] (unequal variance t-test: P = 0.07,
n = 60, Fig. 4).
C  W Q.—As with the chlorophyll a data, we averaged se-
lected environmental parameters among all stations in the altered group and unal-
tered group for each month from 1995 to 2004 and from 2009 to 2014. We examined
temperature (°C), salinity, nitrate + nitrite (NOx, µmol L−1), phosphate (PO4
3−, µmol
L−1), and the ratio of NOx:PO4
3−.
Temperature was not sign ificantly different between altered and una ltered stations,
and pre- and postbloom (Table 3). At both the unaltered and altered stations, salinity
was significantly higher postbloom compared to prebloom (Table 3). Moreover, sa-
linity at altered stations was significantly higher both pre- and postbloom compared
to the unaltered stations (Table 3). However, salinity increased 28% at unaltered sta-
tions, but only increased 9% at altered stations postbloom compared to prebloom.
NOx concentrations significantly decreased after the bloom at both unaltered
and altered stations (Table 3). After the bloom, NOx concentrations were signi-
cantly higher at altered stations compared to unaltered stations; whereas prior to
the bloom there was no difference (Table 3). ere was no significant change in
Millette et al.: Shift in baseline following Synechococcus bloom 7
Figure 2. A spatial map of interpolated mean chlorophyll a (µg L−1) concentrations at each station
(A) prebloom (1995–2004) and (B) postbloom (2009–2014). The black stars are stations where
chlorophyll a concentrations were signicantly higher (P < 0.05) postbloom compared to pre-
bloom and the black circles are where there was no signicant difference (P > 0.05).
Bulletin of Marine Science. Vol 94, No 1. 20188
PO4
3− concentrations following the bloom at unaltered and altered stations, but
before the bloom, PO4
3− concentrations were significantly higher at unaltered
stations, and following the bloom there was no significant difference between
the two groups of stations (Table 3). At both the unaltered and altered stations,
the NOx:PO4
3− ratio significantly decreased postbloom compared to prebloom.
NOx:PO4
3− was always significantly higher at the altered stations compared to the
unaltered stations (Table 3).
We analyzed the pre- and postbloom rate of change for water quality parameters
that were significantly different after the 3-yr bloom (Fig. 3). Prebloom, salinity was
significantly increasing at unaltered stations and altered stations (Fig. 3). Postbloom,
salinity was still significantly increasing at unaltered stations, but was not signifi-
cantly changing at altered stations (Fig. 3). At unaltered and altered stations, NOx
and NOx:PO4
3− were significantly decreasing pre- and postbloom (Fig. 3).
Using a Pearson’s correlation analysis, we determined whether any of the changes
in salinity, NOx, and, NOx:PO4
3− were correlated for each group of factors (Table 4).
Prebloom, changes in NOx concentrations were significantly negatively correlated
to salinity at both altered and unaltered stations (Table 4). Postbloom, changes
in NOx concentrations were significantly negatively correlated to salinity at unal-
tered stations (Table 4), but not altered stations (Table 4). Both pre- and postbloom,
changes in NOx: PO4
3- ratios were significantly positively correlated with NOx con-
centrations at both altered and unaltered sites, but with much higher r-values post-
bloom (Table 4).
D
Following the 3-yr bloom, chlorophyll a concentrations i n Manatee Bay, Blackwater
Sound, and Barnes Sound did not recover to prebloom concentrations. e mean
chlorophyll a concentration in these three systems was 45% higher in the 5 yrs
(2009–2014) following the bloom compared to the 10 yrs (1995–2004) before the
bloom. In Long Sound chlorophyll a completely recovered to prebloom concentra-
tions, although this region was less affected during the 3-yr bloom compared to the
other three lagoons (Glibert et al. 2009). Based on the environmental factors ana-
lyzed, we found no connection between the changes in environmental parameters
and the increase in mean chlorophyll a concentrations.
Table 3. A comparison of mean values (SE) for water quality factors before the three-year bloom (1995–2004) to after
the bloom (2009–2014) at stations where chlorophyll a was not signicantly different after the bloom (unaltered)
and signicantly different after the bloom (altered). * When the mean value of a factor within the same group was
signicantly different (P < 0.05) after the bloom compared to before the bloom. ¥ When the mean value of a factor
between groups at the same time point was signicantly different (P > 0.05).
Group and time nTemp (°C) Salinity
Chl a
(µg L−1)
NOx
(µmol L−1)
PO4
3−
(µmol L−1) NOx:PO4
3−
Unaltered
Prebloom 108 26.01 (0.36) 16.57 (0.87) 0.63 (0.03) 1.22 (0.07) 0.06 (0.01) 39.84 (6.32)
Postbloom 60 25.86 (0.59) 23.08 (1.06)* 0.67 (0.04) 0.61 (0.06)* 0.05 (0.01) 14.29 (1.42)*
Altered
Prebloom 108 25.84 (0.34) 26.96 (0.50)¥0.42 (0.02)¥1.20 (0.10) 0.04 (0.00)¥139.62 (48.50)¥
Postbloom 60 25.46 (0.48) 29.69 (0.53)*¥0.77 (0.04)* 0.87 (0.10) *¥0.05 (0.00) 24.54 (3.91)*¥
Millette et al.: Shift in baseline following Synechococcus bloom 9
Figure 3. The 12 month moving averages of chlorophyll a (µg L−1) concentrations (A,B), NOx (µmol L−1) concentrations (C,D), salinity (E,F), and NOx:PO4
3−
(G,H) from altered st ations (top) and unaltered stations (bot tom) prebloom (1995–2004) and postbloom (2009–2014). *Refers to a slope in the linear regression
that is signicantly different than zero (P > 0.05).
Bulletin of Marine Science. Vol 94, No 1. 201810
S  M C a B.—In the study region, there is no
evidence that chlorophyll a concentrations were increasing over time prior to the
bloom, but following the bloom there was a clear shift in the baseline with the mean
chlorophyll a concentration increased by 45% at altered stations. e altered stations
were located in the region most affected by the 3-yr bloom, while the unaltered sta-
tions were in a region less impacted by the bloom (Glibert et al. 2009). is is consis-
tent with our conclusion that the shift in baseline chlorophyll a concentrations were
related to the 2005–2008 bloom. If the increase in chlorophyll a concentrations were
unrelated to the bloom, we would have expected chlorophyll a concentrations to
begin to change before the bloom or to see a baseline shift occur at a larger regional
scale, such as in the control group. However, we found no evidence of either occur-
ring in our analysis.
Following the bloom, a weak significant increasing trend in chlorophyll a concen-
trations at altered stations was detected. However, the postbloom time-series only
covers 5 yrs of data and appears to be overly influenced by the beginning and ending
time points. erefore, we have low confidence that this represents a real increasing
trend. is suggests that chlorophyll a concentrations should not continue to shift
away from the new baseline.
Currently, there are no data on phytoplankton species composition following the
bloom. In the 1990s, diatoms and dinoflagellates dominated the phytoplankton bio-
volume throughout the year (Phlips et al. 1999). During the bloom, the phytoplank-
ton species composition was altered and cyanobacteria (Synechococcus) dominated
the phytoplankton biovolume (Glibert et al. 2009). While we have shown that the
mean chlorophyll a concentration at altered stations increased following the bloom,
it is unknown if diatoms and dinoflagellates returned as the dominant groups,
Figure 4. The 12-mo moving average of chlorophyll a concentrations (µg L−1) at control stations
(green circles), altered stations (orange triangles), and unaltered stations (blue ×’s). For altered
and unaltered stations data f rom 2005 to 2008 during the bloom was removed but data during
this range were retained for the control stations.
Millette et al.: Shift in baseline following Synechococcus bloom 11
cyanobacteria continues to dominate, or a new assemblage of phytoplankton has be-
come the most abundant group. Collection of water samples to analyze phytoplank-
ton species composition is needed to fully understand whether there was just a shift
in the baseline of phytoplankton abundances, or a shift in the species composition
occurred as well.
e shift in mean chlorophyll a concentrations throughout Manatee Bay,
Blackwater Sound, and Barnes Sound signals that the 3-yr bloom had a lasting ef-
fect on these systems, but additional data on how phytoplankton species composi-
tion was affected may help determine the full impact. e population dynamics of
phytoplankton are directly linked to large-scale oceanographic phenomena, such as
biogeochemical cycling, fisheries sustainability, and shifts in global climate. Changes
in the phytoplankton population dynamics and community will ultimately affect top
trophic levels (Fredericksen et al. 2006). Protists, the primary grazers of phytoplank-
ton, can be highly selective, choosing their prey based on a range of factors including,
size, shape, and chemical composition (Tillmann 2004). If there were a shift in the
phytoplankton species composition in addition to an increase in abundance, then
the lasting impact of the 2005–2008 bloom would affect the zooplankton population
and higher trophic levels.
C  W Q.—ere was no evidence that changes in water
quality parameters following the bloom caused the shift in chlorophyll a at altered
stations. While certain water quality parameters were significantly different post-
bloom, these parameters were already trending significantly upward or downward
during the 10 yrs prior to the bloom. If the change in chlorophyll a was related to
any of the water quality factors we analyzed, then we would have expected chloro-
phyll a concentrations to be significantly increasing over the 10-yr period before the
bloom. Chlorophyll a concentrations and water quality changed from 1995 to 2014,
although it appears the changes were due to separate reasons.
e changes in salinity, NOx concentrations, and the NOx:PO4
3− ratio from 1995 to
2004 were likely related to efforts taken by SFWMD to redirect the water flow from
the C-111 canal (WRDA 2000). e C-111 canal, the largest source of freshwater into
this region, was built in 1966 to divert water away from the Everglades into Manatee
Bay (McIvor et al. 1994). However, the stated goals of the Comprehensive Everglades
Restoration Plan are to increase freshwater flow back toward the Everglades and
away from the C-111 canal (WRDA 2000). is reduction in freshwater flow out of
Table 4. Results from Pearson correlation analysis that compared the correlations between salinity,
nitrate + nitrite, and NOx:PO4
3−. The analysis was run for each set of factors, at altered and unaltered
stations, prebloom (n = 108) and postbloom (n = 60). We used the moving mean transformed data
for this analysis. * Signicant correlation (P < 0.05).
Altered stations Unaltered stations
NOxNOx:PO4
3− NOxNOx:PO4
3−
Prebloom
Salinity −0.60* −0.41* −0.68* −0.62*
NOx0.53* 0.69*
Postbloom
Salinity 0.05 0.08 −0.50* −0.55*
NOx0.95* 0.97*
Bulletin of Marine Science. Vol 94, No 1. 201812
C-111 began shortly after 2000, and was likely the cause of increasing salinity and de-
creasing NOx. Freshwater river flow is the primary source of NOx (Caccia and Boyer
2007); therefore, a reduction of freshwater inflow from C-111 would simultaneously
increase salinity and decrease NOx concentrations. It appears that the decrease in
NOx:PO4
3− was caused primarily by the decrease in NOx concentrations. us, it is
likely that the changes in the water quality parameters we analyzed were caused by
redirecting the flow of C-111, and not the 2005–2008 bloom.
P C   B S.—We hypothesize that the shift in the
baseline chlorophyll a concentrations after the 2005–2008 Synechococcus bloom was
the result of a reduction in the epibenthic community, specifically SAV and sponges.
Loss of SAV may have increased the concentration of nutrients available to phyto-
plankton (Hunt and Nuttle 2007), and decreased the abundance of epifaunal suspen-
sion-feeders, such as bryozans and amphipods, that can consume high quantities of
phytoplankton (Lemmens et al. 1996, Lisbjerg and Peterson 2000). e loss of spong-
es would also have reduced grazing pressure on phytoplankton (Peterson et al. 2006).
Before the 3-yr bloom, the majority of primary production was produced by epi-
benthic plants, and as a result, pelagic productivity was low (Fourqurean et al. 2002,
Nielsen et al. 2006). Phosphorus is considered the primary limiting nutrient in the
study region, with most of the phosphorus found in the sediment, available to the
benthic community, but not the pelagic community (Koch et al. 2001). After the
initiation of the Synechococcus bloom, massive mortality of the SAV community was
recorded (Rudnick et al. 2008). e large reduction in SAV cover likely caused an
increase in pelagic phosphorus concentrations, further sustaining the bloom (Hunt
and Nuttle 2007, Glibert et al. 2009). is created a feedback loop between the bloom
and SAV, with SAV mortality fueling the bloom by releasing phosphorus that, in turn,
caused additional light limitation and mortality for the SAV population (Glibert et
al. 2009).
Preliminary analysis of SAV coverage data by DERM suggest that total SAV cover-
age in Manatee Bay and Barnes Sound is currently lower compared to 2005 (Avila
et al. 2017). is could mean that more phosphorus may continue to be available
to the pelagic phytoplankton compared to before the bloom and support higher
phytoplankton abundances. Our analysis did not show a significant increase in
PO4
3− concentrations following the bloom, but PO4
3− concentrations were no longer
significantly different between altered and unaltered stations after the bloom. If SAV
coverage was lower near unaltered stations before the bloom compared to altered
stations, this could explain why chlorophyll a and PO4
3− concentrations were signifi-
cantly higher at unaltered stations before the bloom. If SAV coverage did not com-
pletely recover near altered stations following the bloom and is now similar to SAV
coverage near unaltered stations, then that could be why chlorophyll a and PO4
3−
concentrations are now similar throughout the study area. However, further analysis
of SAV coverage data is needed to test this hypothesis.
Almost complete mortality of benthic sponges occurred leading up to the initia-
tion of the Synechococcus bloom, between July and October 2005 (Alleman et al.
2009). e mortality was likely caused by high sedimentation related to the series
of hurricanes—Katrina (August 2005), Rita (September 2005), and Wilma (October
2005)—that passed through southern Florida, and the sponges did not recover by the
end of the bloom (Alleman et al. 2009). Sponge mortality was likely not caused by the
Millette et al.: Shift in baseline following Synechococcus bloom 13
Synechococcus bloom because the bloom initiated in September 2005 (Glibert et al.
2009), after most of the mortality occurred (Alleman et al. 2009). Sponges graze on a
range of phytoplankton (Peterson et al. 2006); therefore, a large loss of sponges would
reduce grazing pressure on phytoplankton and could have contributed to the initial
increase in chlorophyll a concentrations. As with SAV, there is no analysis on how
sponge populations have recovered following the bloom. If benthic sponge coverage
remains near zero, then the continued reduced grazing pressure would allow for an
increase in mean phytoplankton abundance.
Benthic coverage data is collected by DERM (Avila et al. 2017) and Florida’s Fish
and Wildlife Research Institute (Hall and Durako 2016), but further study is need-
ed to address the current status of SAV and sponge coverage compared to before
the bloom. is is necessary to confirm whether a continued loss or reduction of
SAV and sponge populations caused the increase in mean chlorophyll a concentra-
tions following the bloom. Additionally, if data are available, a comparison of ben-
thic sponge and SAV coverage at altered and unaltered stations before and after the
bloom is needed to help understand why chlorophyll a concentrations were higher at
unaltered stations before the bloom, but not significantly different after the bloom.
Loss of benthic communities, typically SAV, has led to alternative stable states in
other coastal systems that have shifted towards a more phytoplankton-dominant
state (Burkholder et al. 2007, Glibert et al. 2014).
Typically, when coastal ecological communities exceed a tipping point into a new
stable state, it is caused by sustained eutrophication over an extended period of time
(Duarte et al. 2009, Wang et al. 2012, Glibert et al. 2014). We suggest, in the present
study region, a tipping point at altered stations apparently was surpassed at some
point over the 3-yr bloom, as compared to a decade or more in typical eutrophied
systems (e.g., Tampa Bay, Maryland and Virginia coastal lagoons, and the Black Sea).
e study region had no symptoms of eutrophication before the bloom, in fact, in-
organic N was decreasing. It is possible the 3-yr bloom was such an extreme event
that it sped up a process which normally takes much longer. Typically, one of the first
symptoms of eutrophication is the increase in chlorophyll a concentrations (Bricker
et al. 1999, Chislock et al. 2013), but before a tipping point occurs, there is a loss of
resilience in a system caused by a range of factors, such as reduction in SAV coverage
(Zhang et al. 2003, Barbier et al. 2011).
e rapid and sustained increase in chlorophyll a concentrations by an order of
magnitude was caused by a large increase in phosphorus concentrations (Glibert et
al. 2009) and could have reduced the resilience through massive mortality of SAV
beds (Rudnick et al. 2008). Once external stressors (here, high inputs of nutrients)
are removed from an ecosystem, it is possible it will fail to return to its original
state because changes that occurred still remain (Duarte et al. 2009, Burkholder and
Glibert 2013). In the case of the 3-yr bloom, once the large increase of phosphorus
that initiated and sustained the bloom returned to normal concentrations, chloro-
phyll a concentrations decreased, but did not completely return to prebloom levels.
is finding suggests that some of the changes to the system caused by the extreme
eutrophication event, such as the loss of SAV, remain. It is possible that by identifying
and restoring these lingering alterations, the system could completely recover to its
prebloom state.
Bulletin of Marine Science. Vol 94, No 1. 201814
S.—Before the 2005–2008 bloom, the system was in a state of benthic
dominance, and then during the bloom the system shifted towards a state of phyto-
plankton dominance (Glibert et al. 2009). After the bloom, chlorophyll a concentra-
tions in the most affected area did not return to prebloom conditions and formed a
new baseline, suggesting that the phytoplankton community is now relatively more
productive compared to before the bloom. During the bloom, there was a large loss
of SAV and sponges (Rudnick et al. 2008, Alleman et al. 2009, Glibert et al. 2009)
and preliminary analysis suggests some of these communities have not completely
recovered (Avila et al. 2017). A reduction in SAV would have made more phosphorus
available for pelagic phytoplankton growth, and a reduction of sponges would reduce
removal of phytoplankton. We hypothesize that this resulted in a hysteresis that did
not allow the system to return to prebloom conditions, even 5 yrs after the bloom
terminated and nutrient concentrations returned to normal. e 3-yr bloom appears
to have had a lasting impact on a majority of the study area, with an increase in the
baseline of mean chlorophyll a concentrations, but more research is needed to un-
derstand how this change impacts higher trophic levels.
A combination of multiple, extreme stressors started the Synechococcus bloom that
lasted for 3 yrs. While it is unlikely that this combination of events, three hurricanes
over a 3-mo period and major road construction, will occur again, the higher base-
line chlorophyll a concentrations at altered stations indicate that the ecosystem may
be more susceptible to large phytoplankton blooms caused by smaller disturbances.
It is hypothesized that the benthic communities have not fully recovered, and that
the loss of SAV and sponge populations has persisted. us, a minor disturbance to
the environment that favored growth of any phytoplankton species could result in
another large phytoplankton bloom.
A
e authors would like to thank DERM and SFWMD for sharing their water quality data.
is research was also funded by a NOAA/Atlantic Oceanographic and Meteorological
Laboratory grant to the Northern Gulf Institute (award number NA160AR4320199).
L C
Alleman R, Chang ML, Doering P. 2009. Management and restoration of coastal ecosystems.
2009 South Florida. Environ Rep. I:1–166.
Avila CL, Varona G, Pierre M, Abdelrahman O, Monty J. 2017. ‘A review of seagrass losses
and algal blooms in Biscayne Bay’. Presented at: Greater Everglades Ecosystem Restoration;
April 19, 2017, Coral Springs, FL. (abstract only).
Barbier EB, Hacker SD, Kennedy C, Koch EW, Stier AC, Silliman BR. 2011. e value
of estuarine and coastal ecosystem services. Ecol Monogr. 81:169–193. https://doi.
org/10.1890/10-1510.1
Boesch D. 2006. Scientific requirements for ecosystem-based management in the restoration
of Chesapeake Bay and Coastal Louisiana. Ecol Eng. 26:6–26. https://doi.org/10.1016/j.
ecoleng.2005.09.004
Boyer JN, Fourqurean JW, Jones RD. 1997. Spatial characterization of water quality in Florida
Bay and Whitewater Bay by multivariate analyses: Zones of similar influence. Estuaries.
20:743–758. https://doi.org/10.2307/1352248
Boyer JN, Fourqurean JW, Jones RD. 1999. Seasonal and long-term trends in the water quality
of Florida Bay (1989–1997). Estuaries. 22:417–430. https://doi.org/10.2307/1353208
Millette et al.: Shift in baseline following Synechococcus bloom 15
Briceño HO, Boyer JN. 2010. Climatic Controls on Phytoplankton Biomass in a Subtropical
Estuary, Florida Bay, USA. Estuaries Coasts. 33:541–553. https://doi.org/10.1007/
s12237-009-9189-1
Bricker SB, Clement CG, Pirhalla DE, Orlando SP, Farrow DRG. 1999. National Estuarine
Eutrophication Assessment: Effects of nutrient enrichment in the nation’s estuaries. NOAA ,
National Ocean Service, Special Projects Office and the National Centers for Coastal Ocean
Science. Silver Spring, MD. 71 p.
Burkholder JM, Glibert PM. 2013. Eutrophication and oligotrophication. Levin S, editor. In:
Encyclopedia of biodiversity, 2nd edition, Volume 3. Academic Press, Waltham, MA. p.
347–371.
Burkholder JM, Tomasko DA, Touchette BW. 2007. Seagrasses and eutrophication. J Exp Mar
Biol Ecol. 350:46–72. https://doi.org/10.1016/j.jembe.2007.06.024
Caccia VG, Boyer JN. 2007. A nutrient loading budget for Biscayne Bay, Florida. Mar Pollut
Bull. 54:994–1008. https://doi.org/10.1016/j.marpolbul.2007.02.009
Chislock MF, Doster E, Zitomer RA, Wilson AE. 2013. Eutrophication: causes, consequences,
and controls in aquatic ecosystems. Nature Education Knowledge. 4:10.
Cox PA, Banack SA, Murch SJ, Rasmussen U, Tien G, Bidigare RR, Metcalf JS, Morrison LF,
Codd GA, Bergman B. 2005. Diverse taxa of cyanobacteria produced beta-N-methylamino-
L-alanine, a neurotoxic amino acid. Proc Natl Acad Sci USA. 102:5074–5078. https://doi.
org/10.1073/pnas.0501526102
Duarte CM, Conley DJ, Carstensen J, Sánchez-Camacho M. 2009. Return to neverland: shifting
baselines affect eutrophication restoration targets. Estuaries Coasts. 32:29–36. https://doi.
org/10.1007/s12237-008-9111-2
Fourqurean JW, Durako MJ, Hall MO, Hefty LN. 2002. Seagrass distribution in south Florida:
a multi-agency coordinated monitoring program. In: Porter, JW, Porter KG, editors. e
Everglades, Florida Bay, and coral reefs of the Florida Keys: an ecosystem sourcebook. CRC
Press, Boca Raton. Pp. 497-522.
Frederiksen M, Edwards M, Richardson AJ, Halliday NC, Wanless S. 2006. From plankton to
top predators: bottom-up controls of a marine food web across four trophic levels. J Anim
Ecol. 75:1259–1268. https://doi.org/10.1111/j.1365-2656.2006.01148.x
Folke C, Carpenter S, Walker B, Scheffer M, Elmqvist T, Gunderson L, Holling CS. 2004.
Regime shifts, resilience, and biodiversity in ecosystem management. Annu Rev Ecol Evol
Syst. 35:557–581. https://doi.org/10.1146/annurev.ecolsys.35.021103.105711
Glibert PM, Heil CA, Rudnick DT, Madden CJ, Boyer JN, Kelly SP. 2009. Florida Bay: water
quality status and trends, history and emerging algal bloom problems. Contrib Mar Sci.
39:5–17.
Glibert PM, Hinkle DC, Sturgis B, Jesien RV. 2014. Eutrophication of a Maryland/Virginia
coastal lagoon: a tipping point, ecosystem changes, and potential causes. Estuaries Coasts.
37:128–146. https://doi.org/10.1007/s12237-013-9630-3
Gunderson LH. 2001. South Florida: the reality of change and the prospects for sustainability:
managing surprising ecosystems in southern Florida. Ecol Econ. 37:371–378. https://doi.
org/10.1016/S0921-8009(01)00179-3
Hall MO, Durako MJ. 2016. South Florida fisheries habitat assessment program (FHAP-SF).
CERP Monitoring and Assessment Plan. Annual Report 5.
Hamilton TJ, Paz-Yepes J, Morrison RA, Palenik B, Tresguerres M. 2014. Exposure to bloom-
like concentrations of two marine Synechococcus cyanobacteria (strains CC9311 and
CC9902) differentially alters fish behavior. Conserv Physiol. 2(1):cou020. https://doi.
org/10.1093/conphys/cou020
Hunt J, Nuttle W. 2007. Florida Bay science program: a synthesis of research on Florida Bay.
Fish and Wildlife Research Institute technical report. Florida Fish and Wildlife Research
Institute, St. Petersburg, FL. p. 157.
Bulletin of Marine Science. Vol 94, No 1. 201816
Kleppel GS. 1996. e state of Florida’s estuaries and future needs in estuarine research. e
Florida Sea Grant College Program, Technical Paper 85, University of Florida, Gainesville,
Florida. p. 43–51.
Koch MS, Benz RE, Rudnick DT. 2001. Solid-phase phosphorus pools in highly organic car-
bonate sediments of northeastern Florida Bay. Estuar Coast Shelf Sci. 52:279–291. https://
doi.org/10.1006/ecss.2000.0751
Lemmens JWTJ, Clapin G, Lavery P, Cary J. 1996. Filtering capacity of seagrass meadows and
other habitats of Cockburn Sound, Western Australia. Mar Ecol Prog Ser. 143:187–200.
https://doi.org/10.3354/meps143187
Lisbjerg D, Petersen JK. 2000. Clearance capacity of Electra bellula (Bryozoa) in seagrass
meadows of Western Australia. J Exp Mar Biol Ecol. 244:285–296. https://doi.org/10.1016/
S0022-0981(99)00147-1
Martins R, Fernandez N, Beiras R, Vasconcelos V. 2007. Toxicity assessment of crude and par-
tially purified extracts of marine Synechocystis and Synechococcus cyanobacterial strains in
marine invertebrates. Toxicon. 50:791–799. https://doi.org/10.1016/j.toxicon.2007.06.020
McIvor CC, Ley JA, Bjork RD. 1994. Changes in the freshwater inflow from the Everglades to
Florida Bay including effects on biota and biotic processes: a review. In: Davis SM, Ogden
JC, editors. Everglades: the ecosystem and its restoration. St. Lucie Press, Delray Beach, FL.
p. 47–84.
Meals DW, Spooner J, Dressing SA, Harcum JB. 2011. Statistical analysis for monotonic
trends, Tech Notes 6, November 2011. Developed for US Environmental Protection
Agency by Tetra Tech, Inc., Fairfax, VA, 23 p. Available online at https://www.epa.gov/
polluted-runoff-nonpoint-source-pollution/nonpoint-source-monitoringtechnical-notes.
Nielsen O, Koch MS, Jensen HS, Madden CJ. 2006. alassia testudinum phosphate uptake
kinetics at low in situ concentrations using a 33P radioisotope technique. Limnol Oceanogr.
51:208–217. https://doi.org/10.4319/lo.2006.51.1.0208
Organization for Economic Co-operation and Development. 2007. Data and metadata report-
ing and presentation handbook. Paris, OECD Publishing.
Orth RJ, Williams MR, Marion SR, Wilcox DJ, Carruthers TJB, Moore KA, Kemp WM,
Dennison WC, Rybicki N, Bergstrom P, et al. 2010. Long-term trends in submerged aquat-
ic vegetation (SAV) in Chesapeake Bay, USA, related to water quality. Estuaries Coasts.
33:1144–1163. https://doi.org/10.1007/s12237-010-9311-4
Paerl HW, Fulton RS 3rd, Moisander PH, Dyble J. 2001. Harmful freshwater algal blooms, with
an emphasis on cyanobacteria. Sci World. 1:76–113. https://doi.org/10.1100/tsw.2001.16
Peterson BJ, Chester CM, Jochem FJ, Fourqurean JW. 2006. Potential role of sponge communi-
ties in controlling phytoplankton blooms in Florida Bay. Mar Ecol Prog Ser. 328:93–103.
https://doi.org/10.3354/meps328093
Phlips EJ, Badylak S, Lynch TC. 1999. Blooms of the picoplanktonic cyanobacterium
Synechococcus in Florida Bay, a subtropical inner-shelf lagoon. Limnol Oceanogr. 44:1166–
1175. https://doi.org/10.4319/lo.1999.44.4.1166
Phlips EJ, Lynch TC, Badylak S. 1995. Chlorophyll a, tripton, color, and light availability in
a shallow tropical innershelf lagoon, Florida Bay, USA. Mar Ecol Prog Ser. 127:223–234.
https://doi.org/10.3354/meps127223
Rudnick DJ, Boyer J, Blair S, Glibert P, Heil P, Jaffe R, Price R, Koch M, Madden CJ. 2008.
Florida Bay biogeochemistry and phytoplankton dynamics: synthesis for ecosystem man-
agement and restoration. 2008 Florida Bay and adjacent marine systems science confer-
ence, Naples, FL. p. 29–33. (abstract only).
Rudnick DJ, Madden CJ, Kelly S, Bennett R, Cunniff K. 2007. Report on algal blooms in east-
ern Florida Bay and southern Biscayne Bay. 2007 South Florida Environmental Report,
Appendix 12-3. 28 p.
Scheffer M, Carpenter SR. 2003. Catastrophic regime shifts in ecosystems: linking theory to
observation. Trends Ecol Evol. 18:648–656. https://doi.org/10.1016/j.tree.2003.09.002
Millette et al.: Shift in baseline following Synechococcus bloom 17
Shoaf WT, Lium BW. 1976. Improved extraction of chlorophyll a and b from algae using di-
methyl sulfoxide. Limnol Oceanogr. 21:926–928. https://doi.org/10.4319/lo.1976.21.6.0926
Smith EP. 2002. BACI design. In : El-Shaarawi AH, Piegorsch WW, editors. Encyclopedia of
enviornmentrics. John Wiley & Sons, Chichester. p. 141–148.
Strickland JDH, Parsons TR. 1972. A practical handbook of seawater analysis. B Fish Res Board
Can. 167:107–112.
Tillmann U. 2004. Interactions between planktonic microalgae and protozoan grazers. J
Eukaryot Microbiol. 51:156–168. https://doi.org/10.1111/j.1550-7408.2004.tb00540.x
Wang JD, Luo J, Ault JS. 2003. Flows, salinity, and some implications for larval transport in
south Biscayne Bay, Florida. Bull Mar Sci. 72:695–723.
Wang R, Dearing JA, Langdon PG, Zhang E, Yang X, Dakos V, Scheffer M. 2012. Flickering
gives early warning signals of a critical transition to a eutrophic lake state. Nature. 492:419–
422. https://doi.org/10.1038/nature11655
WRDA. 2000. Comprehensive Everglades Restoration Plan (CERP). 114 Stat. 2687, Sec. 601 (j),
(h). Public Law No. 106-541. December 11. Available from: http://www.fws.gov/habitat-
conservation/Omnibus/WRDA2000.pdf
Zhang J, Jørgensen SE, Beklioglu M, Ince O. 2003. Hysteresis in vegetation shift—Lake Morgan
prognoses. Ecol Model. 164:227–238. https://doi.org/10.1016/S0304-3800(03)00050-4
B
M
S
... Synechococcus was completely dominant in the surface water of Dongzhen Reservoir [26]. The proliferation of Synechococcus was also recorded in other freshwater lakes [27] and lagoons [45]. Due to the tiny size of this cyanobacterium, little information is available about the ecological risk of its blooms. ...
... Due to the tiny size of this cyanobacterium, little information is available about the ecological risk of its blooms. However, Synechococcus blooms may be lethal to sponges and seagrass, and may have a long-term impact on aquatic ecosystems by increasing baseline chlorophyll a concentrations [45]. ...
Article
Full-text available
Harmful cyanobacterial blooms pose a risk to human health worldwide. To enhance understanding on the bloom-forming mechanism, the spatiotemporal changes in cyanobacterial diversity and composition in two eutrophic lakes (Erhai Lake and Lushui Reservoir) of China were investigated from 2010 to 2011 by high-throughput sequencing of environmental DNA. For each sample, 118 to 260 cpcBA-IGS operational taxonomic units (OTUs) were obtained. Fifty-two abundant OTUs were identified, which made up 95.2% of the total sequences and were clustered into nine cyanobacterial groups. Although the cyanobacterial communities of both lakes were mainly dominated by Microcystis, Erhai Lake had a higher cyanobacterial diversity. The abundance of mixed Nostocales species was lower than that of Microcystis, whereas Phormidium and Synechococcus were opportunistically dominant. The correlation between the occurrence frequency and relative abundance of OTUs was poorly fitted by the Sloan neutral model. Deterministic processes such as phosphorus availability were shown to have significant effects on the cyanobacterial community structure in Erhai Lake. In summary, the Microcystis-dominated cyanobacterial community was mainly affected by the deterministic process. Opportunistically dominant species have the potential to replace Microcystis and form blooms in eutrophic lakes, indicating the necessity to monitor these species for drinking water safety.
... a phytoplankton bloom (dominated by Synechococcus) was initiated in the eastern bay region from a series of hurricanes (Katrina, Rita and Wilma) and by disturbance (mangrove removal, sediment resuspension, etc.) due to construction of the causeway connecting the mainland and Florida Keys ). This bloom resulted in the release of P which sustained the bloom for 3 years after which the system did not return to pre-bloom conditions, but instead formed a new baseline with significantly higher chlorophyll concentrations (Millette et al., 2018). One reason for the sustained increase in chlorophyll is a presumed loss of SAV Millette et al., 2018), driving the system toward a more planktonic condition. ...
... This bloom resulted in the release of P which sustained the bloom for 3 years after which the system did not return to pre-bloom conditions, but instead formed a new baseline with significantly higher chlorophyll concentrations (Millette et al., 2018). One reason for the sustained increase in chlorophyll is a presumed loss of SAV Millette et al., 2018), driving the system toward a more planktonic condition. In 2012 the Comprehensive Everglades Restoration Plan was initiated, bringing increased flow into northern Florida Bay, influencing the productivity of the ICOLLs (Shangguan et al., 2017). ...
Article
Few studies have been carried out on benthic dinoflagellates along the Florida Keys, and little is known about their distribution or toxicity in Florida Bay. Here, the distribution and abundance of benthic dinoflagellates was explored in northern and eastern Florida Bay and along the bay and ocean sides of the Florida Keys. Isolates were brought into culture and their toxicity was tested with oyster larvae bioassays. Seven genera were detected, including Prorocentrum, Coolia, Ostreopsis, Amphidinium, Gambierdiscus, Fukuyoa (all included potentially toxic species) and Sinophysis. In general, distribution increased with water temperature and nutrient availability, especially that of phosphate. This study documented the first record of Coolia santacroce in the Florida Keys. Potential toxic effects of Gambierdiscus caribaeus, the abundance of which exceeded 1000 cells g-1 fw at some sites, were established using oyster larvae as a bioassay organism. These findings suggest a potential risk of ciguatera fish poisoning in this area.
... In 2005-2008, a widespread picocyanobacteria bloom persisted for several years attributed to hurricane activity and a local road widening project (Glibert et al., 2009). This major bloom event led to a regime shift with higher mean Chl-a (0.8 mg m −3 ) observed the five years following the bloom (Millette et al., 2018). ...
... A constant and conservative threshold of Chl-a was used to distinguish between bloom (> 5 mg m −3 ) and non-bloom (< 5 mg m −3 ) conditions despite it being well established that Chl-a bloom thresholds in Florida Bay change significantly over both space and time (Millette et al., 2018;Nelson et al., 2017). Cyanobacteria blooms were further separated from 'other' (non-PC containing) blooms in a variety of ways depending on the in situ dataset. ...
Article
Widespread and persistent Ecosystem Disruptive Algal Blooms dominated by marine picocyanobacteria (Synechococcus) commonly occur in the subtropical lagoonal estuary of Florida Bay (U.S.A). These blooms have been linked to a decline in natural sheet flow over the past century from upstream Everglades National Park. Remote sensing algorithms for monitoring cyanobacteria blooms are highly desired but have been mainly developed for freshwater and coastal systems with minimal bottom reflectance contributions in the past. Examination of in situ optical properties revealed that Synechococcus blooms in Florida Bay exhibit unique spectral absorption and reflectance features that form the basis for algorithm development. Using a large, multi-year match-up dataset (2002–2012; n = 682) consisting of in situ pigment concentrations and Moderate Resolution Imaging Spectroradiometer (MODIS) Rayleigh-corrected reflectance (Rrc(λ)), classification criteria for detecting cyanobacteria blooms with chlorophyll-a concentrations (Chl-a) ~5–40 mg m⁻³ were determined based on a new approach to combine the MODIS Cyanobacteria Index, CIMODIS, and spectral shape around 488 nm, SS(488). The inclusion of SS(488) was required to prevent false positive classifications in seagrass-rich, non-bloom waters with high bottom reflectance contributions. 75% of cyanobacteria blooms were classified accurately based on this modified CI approach with <1% false positives. A strong correlation observed between cyanobacteria bloom in situ Chl-a and CIMODIS (r² = 0.80, n = 32) then allowed cyanobacterial chlorophyll-a concentrations (ChlCI) to be estimated. Model simulations and image-based analyses showed that this technique was insensitive to variable aerosol properties and sensor viewing geometry. Application of the approach to the entire MODIS time-series (2000–present) may help identify factors controlling blooms and system responses to ongoing management efforts aimed at restoring flow to pre-drainage conditions. The method may also provide insights for algorithm development for other lagoonal estuaries that experience similar blooms.
... A number of physical and chemical factors have been suggested as contributors to the success of Synechococcus. These included high residence time [32], warm water temperature [33,34], bio-availability of nutrients, especially phosphorus [35,36]. In our study, high abundance of Synechococcus occurred in Transect 1 in Aug-2 when the water column became vertically well-mixed which the temperature and salinity were uniform across all water layers (Figure 2g and Figure S1). ...
Article
Full-text available
Marine Synechococcus are an ecologically important picocyanobacterial group widely distributed in various oceanic environments. Little is known about the dynamics and distribution of Synechococcus abundance and genotypes during seasonal hypoxia in coastal zones. In this study, an investigation was conducted in a coastal marine ranch along two transects in Muping, Yantai, where hypoxic events (defined here as the dissolved oxygen concentration <3 mg L−1) occurred in the summer of 2015. The hypoxia occurred in the bottom waters from late July and persisted until late August. It was confined at nearshore stations of the two transects, one running across a coastal ranch and the other one outside. During this survey, cell abundance of Synechococcus was determined with flow cytometry, showing great variations ranging from 1 × 104 to 3.0 × 105 cells mL−1, and a bloom of Synechococcus occurred when stratification disappeared and hypoxia faded out outside the ranch. Regression analysis indicated that dissolved oxygen, pH, and inorganic nutrients were the most important abiotic factors in explaining the variation in Synechococcus cell abundance. Diverse genotypes (mostly belonged to the sub-clusters 5.1 and 5.2) were detected using clone library sequencing and terminal restriction fragment length polymorphism analysis of the 16S–23S rRNA internal transcribed spacer region. The richness of genotypes was significantly related to salinity, temperature, silicate, and pH, but not dissolved oxygen. Two environmental factors, temperature and salinity, collectively explained 17% of the variation in Synechococcus genotype assemblage. With the changes in population composition in diverse genotypes, the Synechococcus assemblages survived in the coastal hypoxia event and thrived when hypoxia faded out.
... The delayed recovery of seagrass seascapes can be caused by a regime shift, whereby negative feedback processes create conditions antagonistic to seagrass survival and seascape recoveries, such as sediment resuspension, high turbidity, and anoxic sediments (Carr et al., 2012;Maxwell et al., 2017;McGlathery et al., 2013;Orth et al., 2006). Besides the presence of other macroalgae that impedes the seagrass return, field observations have pointed to an increase in total-suspended-solids, turbidity and chlorophyll-a (Supplementary material Fig. S3); thus, suggesting the presence of negative feedback processes that may hinder the recovery and a transition toward an alternate state within the macroalgal bloom footprint (Millette et al., 2018;van der Heide et al., 2007;Rietkerk, M. et al., 2004). ...
Article
Macroalgal blooms are becoming an increasing problem in coastal regions worldwide and have been associated with a widespread decline of seagrass habitats. It is critical to measure macroalgal bloom (MB) impacts at broad spatial scales since seagrass seascape characteristics can influence feedback processes that regulate the resilience of seagrass ecosystems. We assessed the broad-scale spatial impacts of an MB formed by Anadyomene spp. on the seagrass seascapes in Biscayne Bay (Miami, US) using a multi-scale seascape approach. By integrating field and remote sensing data, our multi-scale approach showed significant reductions in seagrass foliage cover and a seascape structure transformation across the bloom extent. The landscape cover and patch extensiveness declined after the MB peak. Other spatial pattern metrics also showed that the seagrass seascape structure got fragmented. We demonstrated that a persistent MB could transform the structure of seagrass seascapes, hindering the resilience of seagrass habitats.
... The rise in Cyanobacteria in AZ-treated microcosms was coupled to a higher proportion of Synechococcales and a lower proportion of Chroococcales (Fig. 2d), whereas AZ did not affect the relative abundance of orders within the Chlorophyta, the main eukaryotic division (Fig. 2c). Cyanobacteria within the order Synechococcale can form toxic blooms, potentially affecting eukaryotic algal competitors [27][28][29]. ...
Article
Full-text available
Background: Sharp increases in food production worldwide are attributable to agricultural intensification aided by heavy use of agrochemicals. This massive use of pesticides and fertilizers in combination with global climate change has led to collateral damage in freshwater systems, notably an increase in the frequency of harmful cyanobacterial blooms (HCBs). The precise mechanisms and magnitude of effects that pesticides exert on HCBs formation and proliferation have received little research attention and are poorly constrained. Results: We found that azoxystrobin (AZ), a common strobilurin fungicide, can favor cyanobacterial growth through growth inhibition of eukaryotic competitors (Chlorophyta) and possibly by inhibiting cyanobacterial parasites (fungi) as well as pathogenic bacteria and viruses. Meta-transcriptomic analyses identified AZ-responsive genes and biochemical pathways in eukaryotic plankton and bacteria, potentially explaining the microbial effects of AZ. Conclusions: Our study provides novel mechanistic insights into the intertwined effects of a fungicide and eutrophication on microbial planktonic communities and cyanobacterial blooms in a eutrophic freshwater ecosystem. This knowledge may prove useful in mitigating cyanobacteria blooms resulting from agricultural intensification.
... Excessive nutrient loading, water diversions, hurricanes (e.g., Katrina, Rita, and Wilma), and groundwater seepage contaminated with septic effluents (Lapointe and Clark 1992;Corbett et al. 1999) led to extensive eutrophication and other water quality problems (Glibert et al. 2009). This resulted in recurring algal blooms, especially blooms of the cyanobacterium Synechococcus spp., which in some cases led to significant die-offs of sponges and seagrasses in the 1980s and 1990s and more recently from 2005 through 2008 (Millette et al. 2018). ...
Article
The Atlantic Goliath Grouper Epinephelus itajara, a large indigenous tropical reef fish, approached local extinction in U. S. waters by the 1980s as a result of intense fishing pressure. In 1990, federal and state laws intervened to protect this species. The resulting fishery closure, over the intervening years, allowed limited, slow population recovery in Florida waters while populations outside of the United States remained vulnerable (IUCN). The closure led to the blossoming of a dive ecotourism industry catering to local and international divers seeking opportunities to see and photograph these enormous fish. This fundamentally changes the paradigm for Goliath Grouper from a fishery resource to a non‐extractive resource with a commercial value vastly greater than that gained through fishing. While federal and state agencies attempted to re‐establish the fishery, all three stock assessments were rejected. Here, we discuss Goliath Grouper's biology, the controversy surrounding its protection, and the drawbacks of re‐establishing a fishery including: loss of nursery habitat, increasingly destructive episodic events like red tide and cold snaps, and the effects of mercury contamination on survival. Add to this the human health risk of consuming mercury‐contaminated fishes, and the argument supporting re‐opening the fishery evaporates. This article is protected by copyright. All rights reserved.
... While hypersalinity events occur naturally on an annual basis in this region (SFNRC 2016), they have been increasing in Florida Bay since 1987. Further, they are exacerbated by climate extremes and freshwater rediversion (of which the Everglades has a long history) and act synergistically with extreme phytoplankton blooms (Millette et al. 2018) and other events, leading to significant die-offs of essential habitat (Johnson et al. 2018). Major seagrass, macroalgae, and sponge dieoffs occurring in Florida Bay in 1987, in Florida Bay and Biscayne Bay from 2005 to 2008, and again as recently as 2015 are attributed to various combinations of events, including hypersalinity, extreme temperatures, unprecedented phytoplankton blooms, low oxygen levels, and high sulfur conditions in sediments (Hall et al. 2016). ...
Article
Full-text available
Large marine predators occupying the same spatial arena exhibit a variety of temporal and behavioral differences to minimize competition for habitat and prey. Here, we examined two such species in the Florida Everglades, red drum Sciaenops ocellatus and snook Centropomus undecimalis, to evaluate niche separation based on diet and multiple stable isotope (white muscle, δ¹³C, δ¹⁵N, δ³⁴S) analyses. For these two estuarine predators, our results indicated that although dietary niche overlap was broad, different feeding modes (spatial and behavioral) allow niche partitioning. The diet of red drum was dominated by pink shrimp Farfantepenaeus duorarum and other demersal species. For snook, although their diet included significant numbers of pink shrimp, it was largely dominated by pelagic and epibenthic fishes. Mean red drum δ13C signatures (− 10.5 to − 20.8‰) differed significantly between areas and were strongly correlated with both area-specific seagrass concentration and amount of incidental seagrass ingestion. Mean snook δ¹³C signatures were generally depleted (− 20.9 to − 22.4‰) with the exception of one area (− 14.1‰). Red drum and snook diet and mean δ¹⁵N signatures (10.1‰, 10.8‰, respectively) indicated they were both mid-trophic-level consumers. Mean red drum δ³⁴S signatures were significantly depleted (− 0.31‰) in the seagrass-dominated area, but enriched (2.03 to 3.78‰) in the other areas and indicated benthic but no pelagic sources of primary production. Mean snook δ³⁴S signatures varied widely (0 to 20‰) among areas suggesting dependence on benthic (benthic algae and seagrass) and pelagic (phytoplankton) sources of primary production.
Article
Birds have been monitored as indicators of ecosystem health in both the freshwater and marine habitats of south Florida, USA for decades. This study reports on nine years (2010-2018) of monthly systematic surveys of breeding waterbird colonies in Biscayne National Park. Overall, 89% of active nests in the park belonged to Double-crested Cormorants (Phalacrocorax auritus). The annual sum of nest counts within the study area grew by 58% over the course of the study. This growth in the nesting population is driven by a 61% growth in northern colonies. During this same time period, the southern colonies declined to less than half their original size. These opposing trends coincide with differences in habitat quality (salinity, chlorophyll, sea grass density, and/or prey abundance) between the two regions. In addition, Hurricane Irma strongly impacted the nesting Double-crested Cormorants, suggesting a loss of nearly 400 nests, although four months post-storm nesting was back to normal levels. Finally, two colonies appear to have started during the study period in close proximity to recently completed Biscayne Bay Coastal Wetlands restoration projects. One of these colonies supported a maximum of over 350 nests.
Article
Full-text available
Eutrophication is characterized by excessive plant and algal growth due to the increased availability of one or more limiting growth factors needed for photosynthesis (Schindler 2006), such as sunlight, carbon dioxide, and nutrient fertilizers. Eutrophication occurs naturally over centuries as lakes age and are filled in with sediments (Carpenter 1981). However, human activities have accelerated the rate and extent of eutrophication through both point-source discharges and non-point loadings of limiting nutrients, such as nitrogen and phosphorus, into aquatic ecosystems (i.e., cultural eutrophication), with dramatic consequences for drinking water sources, fisheries, and recreational water bodies (Carpenter et al. 1998).
Article
Full-text available
Coastal California experiences large-scale blooms of Synechococcus cyanobacteria, which are predicted to become more prevalent by the end of the 21st century as a result of global climate change. This study investigated whether exposure to bloom-like concentrations of two Synechococcus strains, CC9311 and CC9902, alters fish behaviour. Black perch (Embiotoca jacksoni) were exposed to Synechococcus strain CC9311 or CC9902 (1.5 × 10⁶ cells ml−1) or to control seawater in experimental aquaria for 3 days. Fish movement inside a testing arena was then recorded and analysed using video camera-based motion-tracking software. Compared with control fish, fish exposed to CC9311 demonstrated a significant preference for the dark zone of the tank in the light–dark test, which is an indication of increased anxiety. Furthermore, fish exposed to CC9311 also had a statistically significant decrease in velocity and increase in immobility and they meandered more in comparison to control fish. There was a similar trend in velocity, immobility and meandering in fish exposed to CC9902, but there were no significant differences in behaviour or locomotion between this group and control fish. Identical results were obtained with a second batch of fish. Additionally, in this second trial we also investigated whether fish would recover after a 3 day period in seawater without cyanobacteria. Indeed, there were no longer any significant differences in behaviour among treatments, demonstrating that the sp. CC9311-induced alteration of behaviour is reversible. These results demonstrate that blooms of specific marine Synechococcus strains can induce differential sublethal effects in fish, namely alterations light–dark preference behaviour and motility.
Article
Full-text available
Water quality in the Maryland/Virginia Coastal Bays has been declining for many years from anthropogenic inputs, but conditions appear to have worsened abruptly following a shift from long-term dry to long-term wet conditions in the early 2000s. Annually and regionally averaged total nitrogen concentrations are approximately twofold higher, but ammonium (NH4+) concentrations are up to an order of magnitude higher than in the early 1990s. Averaged nitrate concentrations, however, changed to a lesser degree throughout the time course; water column concentrations remain very low. Total phosphorus has only increased in some bay segments, but increases in phosphate (PO43−) have been more pervasive. There were differences in the year in which large increases in each nutrient were first noted: PO43− in ~2001–2002, followed by NH4+ ~a year later. The effects of a combination of steadily increasing anthropogenic nutrient increases from development, superimposed on nutrient loads from farming and animal operations, and groundwater inputs were accelerated by changes in freshwater flow and associated, negatively reinforcing, biogeochemical responses. Regionally, chlorophyll a concentrations have increased, and submersed aquatic vegetation has decreased. The system is now characterized by sustained summer picoplanktonic algal blooms, both brown tide and cyanobacteria. The retentive nature of this coastal lagoon combined with the reducing nature of the system will make these changes difficult to reverse if the current dual nutrient management practices are not accelerated.
Article
Full-text available
Seventeen sites in Florida Bay were sampled on a monthly basis for 51 months to describe the spatial and temporal patterns of phytoplankton blooms. The study focused on the picoplanktonic cyanobacterium Synechococcus. The greatest frequency and intensity of blooms was observed in the north-central region of Florida Bay, where cellular biovolumes of this species regularly exceeded 10 x 106 μm3 ml-1 and chlorophyll a concentrations were frequently >20 mg m-3. Synechococcus blooms were often restricted to this region of the bay, in part because of the network of shallow mudbanks and islands that restrict water exchange with other regions and outlying waters of the Atlantic Ocean and Gulf of Mexico. The most severe blooms occurred in the summer and fall (May-December). High concentrations of Synechococcus also appeared during the fall in the south-central region of the bay. The appearance of blooms in this region coincided with the onset of seasonal cold fronts, whose strong northerly and northwesterly winds appear to drive bloom-laden water from the north-central region into adjacent parts of the bay. A number of physical and chemical factors appear to contribute to the remarkably high phytoplankton biovolumes observed in the north-central region of Florida Bay. Physical factors include the shallowness and hydrological isolation of the region. The dominance of Synechococcus in the center of the bay may be attributable to several of the unique physicochemical characteristics of this species, including its small size, cyanobacterial metabolism, euryhaline character, buoyancy, and tolerance to high light intensity.
Article
The purpose of impact assessment is to evaluate whether a stressor has changed the environment, which components are adversely affected, and to estimate the magnitude of the effects. The evaluation of impact involves comparative methods. Early approaches to impact assessment involved the use of computer simulation models to predict the impact. Decisions were then made based on the soundness of those predictions. Although the data used to interpret effects are quite varied, the methods for analysis are often quite similar and involve comparison of impact areas with control areas. When information is available prior to the potential impact, the design is often referred to as a Before–After Control-impact (BACI) design. Several variations on the basic design have been proposed and are discussed in this article.