Content uploaded by Whitney Hoot
Author content
All content in this area was uploaded by Whitney Hoot on Jul 18, 2019
Content may be subject to copyright.
REPORT
Successive bleaching events cause mass coral mortality in Guam,
Micronesia
L. J. Raymundo
1
•D. Burdick
1
•W. C. Hoot
2
•R. M. Miller
1
•V. Brown
3
•
T. Reynolds
1
•J. Gault
1
•J. Idechong
1
•J. Fifer
1,4
•A. Williams
1
Received: 15 October 2018 / Accepted: 17 June 2019
Springer-Verlag GmbH Germany, part of Springer Nature 2019
Abstract The reefs of Guam, a high island in the Western
Pacific, were impacted by an unprecedented succession of
extreme environmental events beginning in 2013. Elevated
SSTs induced severe island-wide bleaching in 2013, 2014,
2016, and 2017. Additionally, a major ENSO event triggered
extreme low tides beginning in 2014 and extending through
2015, causing additional coral mortality from subaerial
exposure on shallow reef flat platforms. Here, we present the
results of preliminary analyses of environmental and bio-
logical data collected during each of these events. Accu-
mulated heat stress in 2013 was the highest since satellite
measurements began, but this record was exceeded in 2017.
Overall, live coral cover declined by 37% at shallow reef flat
sites along the western coast, and by 34% at shallow seaward
slope sites around the island. Staghorn Acropora commu-
nities lost an estimated 36% live coral cover by 2017.
Shallow seaward slope communities along the eastern
windward coast were particularly devastated, with an esti-
mated 60% of live coral cover lost between 2013 and 2017.
Preliminary evidence suggests that some coral species are at
high risk of extirpation from Guam’s waters. In light of
predictions of the near-future onset of severe annual
bleaching, and the possibility that the events of 2013–2017
may signal the early arrival of these conditions, the persis-
tence of Guam’s current reef assemblages is in question.
Here, we present detailed documentation of ongoing changes
to community structure and the status of vulnerable reef taxa,
as well as a critical assessment of our response protocol,
which evolved annually as bleaching events unfolded. Such
documentation and analysis are critical to formulating
effective management strategies for the conservation of
remaining reef diversity and function.
Keywords Guam Mariana Islands Bleaching mortality
Rapid response
Introduction
Small islands are likely to be disproportionately impacted by
climate change-related stressors, as their high reef-to-land
area and heavy dependence on shallow marine ecosystems
increase their vulnerability to the decline and loss of these
ecosystems. Many such islands have experienced gradual
declines in health, diversity, and productivity in recent
decades, principally from local anthropogenic stressors.
However, recent climate change-related shifts in sea surface
temperature have added a global stressor to this list, with
sudden and devastating consequences in some areas. The
U.S. Territory of Guam (13280N, 144460E), the southern-
most island in the Mariana Archipelago, lies just outside the
Indo-Pacific center of reef biodiversity (Roberts et al. 2002)
and hosts approximately 350 species of shallow-water scle-
ractinian corals (Randall 2003). Like many small islands in
Topic Editor Morgan S. Pratchett
Electronic supplementary material The online version of this
article (https://doi.org/10.1007/s00338-019-01836-2) contains sup-
plementary material, which is available to authorized users.
&L. J. Raymundo
ljraymundo@gmail.com
1
Marine Laboratory, University of Guam, Mangilao, GU,
USA
2
Bureau of Statistics and Plans, Government of Guam,
Haga
˚tn
˜a, GU, USA
3
NOAA Fisheries, Field Office Guam, Tamuning, GU, USA
4
Department of Biology, Boston University, Boston, MA,
USA
123
Coral Reefs
https://doi.org/10.1007/s00338-019-01836-2
the tropical Pacific, the condition of Guam’s coral reefs has
gradually declined since the 1960s from repeated Acan-
thaster planci outbreaks, worsening water quality, and high
fishing pressure from a growing population (Chesher 1969;
Randall and Holloman 1974; Colgan 1987; Burdick et al.
2008; Caballes 2009; McNeil et al. 2015). The recent
superimposition of warming sea surface temperatures and
other unpredicted stressors onto these already stressed
communities had devastating effects on Guam’s reefs. These
events triggered an evaluation of our current approach to reef
management and conservation, and highlighted an urgent
need to develop a strategy for coping with climate-related
change. Our experiences, and the evaluation of our response
strategy, provide important documentation of climate
change impacts to small island ecosystems.
Prior to 2013, coral reefs in Guam had been mildly
affected by anomalous sea surface warming, relative to other
sites in the Western Pacific. Paulay and Benayahu (1999)
described the first recorded bleaching event in Guam in 1994
as affecting reefs island-wide but resulting in little mortality.
While the authors did not believe the event was associated
with unusually high sea surface temperatures at the time of
the study, a later review of Pathfinder sea surface tempera-
ture measurements indicated that the bleaching threshold
used for Guam for the period 1985–2003 (29.9 C) was
exceeded in 1994 (Burdick et al. 2008). Birkeland et al.
(2000) and Richmond et al. (2002) mentioned coral
bleaching in association with the historic 1997–1998 El Nin
˜o
Southern Oscillation (ENSO) event, but mortality was lim-
ited and the overall impact was considered mild compared to
the significant impacts observed in nearby Palau (Bruno et al.
2001) and at other reef locations around world (Wilkinson
2000). Burdick et al. (2008) reported bleaching on Guam’s
reefs in 2006 and 2007, but mortality was limited primarily to
Acropora along the reef margin.
Although the frequency and severity of mass coral
bleaching events on Guam’s reefs has increased in recent
decades, impacts were limited. However, a severe bleaching
event in 2013, and subsequent events in 2014, 2016, and 2017,
elevated concern among managers and researchers regarding
the high potential for these events to drive a rapid and sig-
nificant lossin coral cover, change in speciescomposition, and
decline in overall reef condition. Concern was also raised
regarding the potential for the onset of near-annual frequency
of heat stressevents occurring earlier than had previously been
predicted for the region (Donner et al. 2005;Donner2009;van
Hooidonk et al. 2016). In addition to the repeated heat stress
events, a year-long ENSO-related period of extreme low tides
began in late 2014 and extended into 2015, resulting in further
coral mortality on Guam’s shallow reef flat platforms. Here,
we present summaries of these annual events, our assessment
of the current state of reefs in Guam, and an analysis of the
methods we utilized to respond to these events, which will be
incorporated into future management protocols.
Materials and methods
The severity, scale, and repetitive nature of these bleaching
events necessitated the development of multiple survey
approaches, to maximize the quality of data collected given
limited time and resources. These surveys and our assess-
ment of their utility are described in Table S1. An intensive,
island-wide survey approach was formulated by the authors
in 2013, based on qualitative reconnaissance assessments of
a subset of sites where bleaching was first detected, and
informed by a draft bleaching response plan previously
developed by the Guam Bureau of Statistics and Plans (BSP)
and the National Oceanic and Atmospheric Administration
(NOAA) Fisheries Guam Field Office, with support from the
NOAA Coral Reef Conservation Program. However, time
and resources were not available to achieve the same scale
of sampling effort for subsequent years, and thus, a scaled-
down effort was adopted for a smaller number of sites
during these events. Surveys were conducted by a team of
managers and scientists from government agencies (BSP,
NOAA Fisheries, Guam Environmental Protection Agency,
National Parks Service) and academe (University of Guam).
In addition, an existing reef flat long-term monitoring pro-
gram tracked bleaching events at five monitored sites
throughout the entire period. Shallow staghorn Acropora
communities were particularly vulnerable and were initially
spot surveyed for bleaching impacts in 2014, and later
quantitatively surveyed for the extent of loss in 2014–2015
and again in 2017. In 2016, a rapid reef flat site assessment
protocol was developed and tested for ‘‘canary sites’’—those
sites intended to serve as part of an early warning system to
guide decisions on further actions. Figure 1provides the
locations along the coast of Guam where these different
surveys were conducted. Prevalence of bleaching and
bleaching mortality were calculated for all species encoun-
tered in these surveys as:
#of colonies or %of coral cover with bleaching or bleaching mortality
#of colonies counted 100
Coral Reefs
123
Seaward slope surveys
Shallow seaward slope surveys were undertaken in 2013,
2015, 2016, and 2017 using generally the same protocols;
details of modifications are described where appropriate.
All sites were at approximately 5 m depth and included
reef fronts of fringing platform reefs, shallow reaches of
apron reefs, and shallow veneering communities at the cliff
base along the northeastern coast; surveys did not include
sheltered environments such as lagoons. Sites were located
in the field using a handheld GPS receiver, and divers
entered the water at a safe distance immediately seaward of
the site and navigated to the target depth. Surveys were
conducted along (n= 3) 25 m (2013 and 2015) or 30 m
(2016 and 2017) transects placed along the target depth
contour, clockwise around the island during all survey
periods. Benthic photo-transect surveys were conducted
using a compact point-and-shoot camera (Canon Pow-
erShot SD940 IS in 2013, Canon PowerShot S120 in 2015,
and Sony Cybershot RX100 in 2016 and 2017) to obtain an
image every meter along each transect at 1 m above the
substrate. Semiquantitative coral community bleaching
assessments carried out in 2017 were conducted by one
diver (either DB or LJR) swimming off-transect (in the
general vicinity) for 20 min and recording the bleaching
condition of each haphazardly encountered coral. Corals
were identified to the lowest taxon possible and scored
according to the following bleaching severity categories:
normal (normal pigmentation), partial colony paling (such
as on surfaces directly exposed to insolation), whole colony
paling, partial bleaching (on surfaces directly exposed to
insolation), whole colony bleaching, partial bleaching-as-
sociated mortality (on surfaces directly exposed to insola-
tion), and whole colony bleaching-associated mortality.
These in situ observations were augmented by photo-doc-
umentation of additional colonies not assessed on-site;
bleaching condition was assessed from these photographs
using the same categories presented above and combined
with the in situ observations into a single database.
In 2013, benthic photo-transect surveys were carried out
between October and December at 46 sites around Guam in
response to observations of bleaching at multiple sites
around the island in August. Forty-one of these sites were
among the 52 sites visited during a NOAA Pacific Islands
Fisheries Science Center (PIFSC) reef fish community
assessment in 2011 (see Williams et al. 2012). The NOAA
PIFSC sites were generated using a depth-stratified random
sampling design. We located these sites using waypoints
provided by NOAA PIFSC, keeping our survey sites to 5 m
depth. The locations of seven additional sites were ran-
domly generated along the northeast coast using ArcGIS,
as the NOAA PIFSC surveys underrepresented this portion
of the coastline.
In 2015, benthic photo-transect surveys were carried out
between June and September at a randomized subset
(n= 17) of the 2013 survey sites. The objective of these
surveys was to assess cumulative impacts of 2013 and
2014, and to establish a new baseline of coral cover, spe-
cies composition, and condition against which recovery or
future impacts could be measured. Coral quadrat surveys
were conducted along the same transects as the benthic
photo-transect surveys, to assess the size and condition of
coral colonies within (n= 6) 0.25 m
2
quadrats placed
every 5 m along each transect. Colony size was visually
estimated and binned (B10 cm, 11–30 cm, 31–60 cm,
61–100 cm, 101–200 cm,[200 cm). Health impacts were
recorded and the percentage of colony affected was esti-
mated. Partial and full-colony mortality were assessed and
ascribed to bleaching-associated mortality from the 2013
and 2014 events based on expert assessment of the pattern
and estimated timing of the mortality. Partial mortality was
characterized as low (B10%), medium (10–50%), or high
([50%).
Benthic photo-transects were carried out at 20 sites
around the island between July 2016 and January 2017,
including 17 of the sites surveyed in 2013 and four new
sites, in conjunction with a NOAA Saltonstall-Kennedy
Grant-funded reef resilience assessment led by Dr. Jeffrey
Maynard and DB. The resilience assessment sites were
selected from the 2013 bleaching assessment sites in a non-
random manner to achieve an even distribution of sites
around the island, with priority placed on those sites that
had been surveyed in both 2013 and 2015. Persistent poor
water conditions prevented surveys at five of the sites along
the east coast.
In 2017, benthic photo-transect surveys and semiquan-
titative coral community bleaching assessments were car-
ried out at 12 sites in October in response to observations
of bleaching at multiple sites around the island. Several
sites, including most sites on the windward eastern coast,
could not be accessed due to hazardous water conditions.
Images from photo-transect surveys were color-cor-
rected using Adobe Photoshop Lightroom 6.0 and analyzed
for percent cover using CPCe 4.1. The benthic feature at
each randomly generated point (16 points per image) was
identified to the lowest possible taxonomic level. Points
that fell on a portion of a coral colony that exhibited non-
normal pigmentation or exhibited bleaching-associated
mortality were classified as pale, bleached, or recently dead
in order to generate estimates of the percent of bleaching-
impacted coral cover.
Reef flat long-term monitoring
A long-term monitoring program for reef flat platforms
along Guam’s western coastline was established in 2009.
Coral Reefs
123
Coral Reefs
123
Five reef flats are monitored three to four times per year,
along (n= 3) permanent 20 m 91 m belt transects per
site. Two of these sites (Luminao and Piti Bomb Holes) are
dominated by Porites; Tanguisson, Tumon Bay and West
Agan
˜a are dominated by large staghorn Acropora thickets
and Pavona spp. The line-intercept method was used to
characterize benthic composition; live hard coral cover
data from this data set were analyzed to examine changes
over time between 2012 (prior to bleaching onset) and
2017. At one site (West Agan
˜a), transect markers were lost
in a storm in 2013 and redeployed in approximately the
same positions; thus, only data from 2014–2017 were used
in this analysis.
Assessment of staghorn Acropora loss
The areal extent of all known staghorn Acropora popula-
tions (n= 21) around Guam was previously determined
using ArcGIS heads-up digitization of a 2011 Worldview-2
satellite image mosaic, and ground-truthed by in-water
surveys carried out between 2009 and 2013 (described in
Raymundo et al. 2017). A rapid, semiquantitative assess-
ment of the extent of mortality and condition of these
communities was undertaken from November 2014 to
February 2015, after anecdotal reports of severe bleaching
in 2014 (Fig. 1). Surveys were carried out while snorkel-
ing, as these communities are primarily restricted to shal-
low reef flats and lagoonal patch reefs. Surveys involved
visual estimation of percent and patterns of mortality for all
populations, verification of species composition, and geo-
referenced photo-documentation to further develop the
spatial data layer compiled prior to bleaching. Post-
bleaching areal extent was calculated by multiplying per-
cent mortality estimates by pre-bleaching areal extent.
Coral loss was then calculated as the difference between
pre- and post-bleaching area values, with a margin of 10%
variation from the mean to account for uncertainty that is
inherent in visual estimates (Raymundo et al. 2017). An
assessment of 13 individual staghorn Acropora colonies
tagged for reproductive activity in Tumon Bay Marine
Preserve was undertaken in July 2014, when bleaching was
observed within this population. Colonies were individu-
ally inspected for percent of the colony affected and
bleaching severity (pale, bleached, partially dead).
Repeated mortality events following the completion of
surveys in 2015 prompted a second set of surveys of the
same 21 populations in 2017. All populations except those
in Apra Harbor were assessed from February to May, prior
to bleaching season. Surveys at Apra Harbor sites surveys
took place in October, during the height of the bleaching
event. These surveys scored coral condition as live or dead
at 16 points within replicate 0.25 m
2
quadrats placed on
staghorn thickets. Quadrats were placed every 1–2 m along
a visually estimated transect that bisected the thicket; one
or more additional transects perpendicular to the first were
also assessed for larger thickets. Percent mortality was then
calculated as:
#of points of dead skeleton or rubble
16 100
and average mortality per thicket was then calculated from
the replicate quadrats. Additional data on coral condition
collected within quadrats included species composition,
recovery via recruitment or resheeting, disease, predation,
percent rubble (a sign that the thicket was breaking down),
and recruitment of other species onto dead skeleton. As
above, total coral loss per site was calculated as the pre-
bleaching areal extent (assuming 100% coral cover within
thickets) minus percent mortality per thicket, and expressed
in ha.
Rapid reef flat site assessments
Eight shallow reef flat sites were selected in 2016 for rapid
assessment of bleaching severity during the bleaching
event (Fig. 1). Sites were selected based on accessibility
and species composition. The selected coral communities
were predicted to respond quickly to bleaching and thus
provide a rapid means of tracking the scope and severity of
the bleaching event as it progressed.
Reef flat sites were surveyed on multiple occasions
between August and December 2016 and between
September and November 2017, corresponding to warming
events during each of these years. Snorkelers conducted
20-minute timed swims along (n= 3) parallel 1 m-wide
belt transects, at 1–1.5 m depth. The start and end coor-
dinates for each transect were recorded using a handheld
Garmin GPS attached to a float. Survey area was calculated
by multiplying the transect length by 1 m width. Transect
length and the distance between transects were dependent
on the spatial extent of target coral communities, and thus
varied across sites. Re-surveys started at the same coordi-
nate and compass heading. In 2016, two sites were sur-
veyed twice and two were surveyed three times; the
remaining four sites were each surveyed four times over
the five-month period, for a total of 26 surveys. In 2017, all
sites but one were surveyed twice, for a total of 15 surveys.
The total reef area surveyed in 2016 was 791 m
2
, with a
mean of 98 m
2
±62 m
2
surveyed per site. In 2017, we
surveyed a total reef area of 916 m
2
, with a mean of
116 m
2
±54 m
2
per site. All coral colonies within each
bFig. 1 Map of Guam, showing location of survey sites for the
shallow seaward slope, rapid reef flat sites, and staghorn mortality
assessments, and reef flat long-term monitoring bleaching surveys
Coral Reefs
123
belt were identified to species, where possible, and char-
acterized by the severity of bleaching (no bleaching, partial
colony paling, partial colony bleaching, whole colony
paling, whole colony bleaching, and partial or whole col-
ony mortality). Bleaching prevalence was calculated as:
#of points of dead skeleton or rubble
16 100
A bleaching mortality index (BMI) (McClanahan et al.
2004) was calculated for all genera with more than 5
colonies counted across all surveys accomplished in
September for each year (2016 and 2017), using the
formula:
BMI ¼0c1 þ1c2 þ2c3 þ3c4ðÞ
3
Bleaching severity categories used in our study were
pooled to fit into the four bleaching categories used in the
index: c1 = unbleached; c2 = moderate (partial colony
paling; partial colony bleaching; whole colony paling);
c3 = severe (whole colony bleaching; partial colony mor-
tality); c4 = dead (whole colony mortality).
Environmental parameter data
Satellite-derived sea surface temperature (SST) and degree
heating weeks (DHW) data, bleaching alerts, and predic-
tions were accessed via NOAA Coral Reef Watch (2017).
Temperatures and sea level were also monitored using the
NOAA Tide Gauge in Apra Harbor (NOAA CO-OPS
2018a,b). Temperature and wave data from wave buoys
located near Ipan, eastern Guam, and Ritidian, northern
Guam, were monitored at PACIOOS (2018), and quality-
controlled datasets were obtained for analysis from NOAA
(2018a,b). Tropical cyclone records between 2013 and
2017, for all systems within 200 nm of Guam, were gath-
ered from NOAA Digital Coast Historic Hurricane Tracks
Viewer (NOAA 2019). Reef flat temperatures have been
monitored at three reef flat long-term monitoring sites since
2009, using Onset
Hobo pendant loggers installed at
approximately 1 m depth. Additional loggers were instal-
led at three rapid reef flat assessment sites at the same
depth during 2016 and 2017 bleaching events.
Statistical analyses
Seaward slope percent coral cover and percent bleaching-
impacted coral cover values were square-root-transformed
and pooled at the site level. Values were tested for nor-
mality using a Shapiro–Wilk test and for homoscedasticity
using the modified Levene equal-variance test. The R
package Partiallyoverlapping, which was developed for
comparing samples with a mix of paired and unpaired
observations based on method presented in Derrick et al.
(2017), was used to perform two-sample comparisons for
sites between each possible sample year combination.
Comparisons between windward and leeward sites within
the same sampling year were made using an equal-variance
ttest or Aspin–Welch unequal-variance test for normally
distributed data, or a Mann–Whitney U test for non-normal
data.
Reef flat long-term monitoring live hard coral cover data
were square-root-transformed to meet the assumptions of
normality and homoscedasticity and were examined via a
two-way ANOVA using DataDesk V8.0.3 software
, with
year and site as predictors.
Results
Synopsis of events in 2013
Environmental parameters
Satellite-derived SST first exceeded the predicted coral
bleaching threshold for Guam (30 C) on 1 June. NOAA
CRW issued a Bleaching Watch (0 C\HotSpot \1C)
on 11 June, and SST remained below 30 C until 14 July.
Temperatures reached Alert Level 1 levels (4 BDHW
B8) on 13 August, and Alert Level 2 (C8DHW)on3
September; Alert Level 2 status continued through most of
October. A maximum SST of 31.5 C was recorded on 31
August (Fig. 2, 2013). Maximum recorded in situ temper-
atures of 31.5 C were recorded from the Ipan buoy
(southeastern Guam; Fig. 1) in both August and Septem-
ber, 32.6 C from the Ritidian buoy (northwestern Guam;
Fig. 1) in September, and 34 C from the Luminao (central
western Guam; Fig. 1) reef flat logger in July. Accumu-
lated heat stress reached a peak of 12 DHW in early
October and did not fully dissipate until late December.
Temperatures remained above 29 C through 19 Decem-
ber, with brief declines as low-pressure systems passed
nearby. A pair of cyclones in mid-October caused south-
west wave heights in excess of 5 m and resulted in a brief
hiatus from high temperatures (Fig. 2, 2013). Wave heights
on western exposures were variable; however, wave
heights on eastern exposures stayed below 2 m through
most of the bleaching period.
Seaward slope surveys
The island-wide mean percentage of shallow (5 m depth)
seaward slope coral cover that was pale or bleached in
2013 was 20% ±16%; an additional 11% ±9% of coral
cover exhibited bleaching-associated mortality, for a total
of 32% ±19% of coral cover that was impacted by the
Coral Reefs
123
bleaching event (Fig. 3). The percentage of pale or
bleached coral cover was greater for eastern windward sites
(31 ±17%) than for western leeward sites (11 ±7%)
(U= 87, Z= 3.75, p\0.001), but there was no significant
difference between the percentage of coral cover with
bleaching-associated mortality at eastern (10 ±7%) versus
western sites (12 ±10%)(t(43) = 0.48, p\0.001)
(Fig. 3).
Synopsis of events in 2014
Environmental parameters
The onset of the sea surface temperature anomaly occurred
early in the year, less than six months after dissipation of
the previous event, with satellite-derived SST first
exceeding the 30 C bleaching threshold on 25 May, and a
maximum SST of 31 C recorded on 20 June (Fig. 2,
2014). Satellite-derived temperatures remained between
29 C and 30 C through December; however, tempera-
tures recorded by in situ loggers exceeded 31 C periodi-
cally between June and September. Accumulated heat
stress reached a peak of 9 DHW in mid-September and
dissipated by mid-December. Periodic low-pressure sys-
tems were associated with increased wave heights and
decreased temperatures in July, August, and October
(Fig. 2, 2014). As the El Nin
˜o event developed in 2014, sea
levels began to decrease around Guam, resulting in extreme
low tides beginning in August (Figure S1, 2014).
Prior to 2015, the NOAA CRW automated operational
Satellite Bleaching Alert email system was based on
measurements obtained at a single stationary 50 km satel-
lite grid cell. In 2014, this alert system issued a Bleaching
Watch for Guam in mid-May, as SST first reached the
maximum monthly mean within the grid cell. A Bleaching
Warning was issued between mid-June and early July,
corresponding to SST values between 30.0 and 30.6 C.
Bleaching Watch conditions continued through late
November, but accumulated heat stress never exceeded 3
DHW. NOAA CRW has since migrated to a higher reso-
lution product that utilizes a grid of 5 km cells extending
across a larger area around each jurisdiction. The retro-
spective SST measurements presented above and associ-
ated CRW alerts based on the 5 km product indicated that
the 2014 temperature anomaly was of significantly greater
magnitude and duration than that recorded by the 50 km
product. The 5 km product signaled the onset of Bleaching
Watch conditions in early April. Alert Level 1 conditions
were reached on 29 June, and after brief declines in SST,
Alert Level 2 conditions persisted from late August
through September. Bleaching Watch conditions did not
dissipate until the end of December.
Reef flat bleaching observations
Anecdotal observations and photo-documentation by LJR,
DB, and VB confirmed that paling, bleaching, and
bleaching-associated mortality affected multiple coral
species at several shallow reef flat sites within six months
Fig. 2 2013–2017 time series of running 7-d mean sea surface
temperature (SST) from the NOAA CRW virtual station for Guam
plotted against annual maximum monthly mean SST (AMMM),
bleaching threshold (BT), and degree heating weeks (DHW) for
2013–2017. Diamonds indicate observed bleaching events; circles
represent cyclone events, and triangles are periods of extreme low
tides
Coral Reefs
123
of the previous bleaching event. Nearshore staghorn
Acropora communities were particularly affected, with LR
and VB observing bleaching and associated mortality of A.
muricata,A. cf. intermedia, and A. cf. pulchra between
May and July; 62% of tagged Acropora cf. pulchra and A.
cf. intermedia colonies were affected.
Synopsis of events in 2015
Environmental parameters
Guam experienced frequent weather disturbances in 2015,
including close approaches by six cyclones. Three others
passed close enough to affect weather conditions in Guam
(Fig. 2, 2015). Decreased SST and high wave heights were
associated with these events. Satellite-derived sea surface
temperature reached 29 C in May and exceeded the 30 C
bleaching threshold briefly in July and August, reaching a
peak of 30.8 C on 31 July; conditions warranted Bleach-
ing Warning status for a total of eleven days (Fig. 2, 2015).
Data collection at the buoys was inconsistent in 2015 due
to impacts from disturbances; data are not presented here
for this year. With the onset of a strong ENSO event, sea
level decreased by 0.35 m between late 2014 and 2015.
Mean low water declined to -0.15 m in December 2015
(Fig. S1, 2015), and extreme low tide values in excess of -
0.1 m were recorded every month in 2015, with peak
values in excess of -0.3 m below mean sea level in October
and November. These ENSO-associated extreme low tide
events repeatedly subaerially exposed shallow reef flat
coral communities throughout the year, causing mortality
of the top several inches of exposed tissue. Partial mortality
was particularly pronounced during the summer months,
triggered by a combination of exposure for several after-
noon hours on consecutive days and doldrum-like wind
conditions resulting in reduced water circulation (Fig. S2).
Fig. 3 Box plots of percentage of bleaching-impacted coral cover
values from shallow (5 m) seaward slope benthic photo-transects
surveys in 2013, 2015, 2016, and 2017. The percent of coral cover
exhibiting paling or bleaching is presented for all sites island-wide
(a), eastern windward sites (b), and western leeward sites; the percent
of coral cover exhibiting bleaching-associated mortality is presented
for all sites island-wide (d), eastern sites (e), and western sites (f); and
the percentage of coral cover exhibiting paling, bleaching, or
bleaching-associated mortality is presented for all sites island-wide
(g), eastern sites (h), and western sites (i). Data were obtained from a
total of 46 sites (21 east, 25 west) in 2013, 17 sites (9 east, 8 west) in
2015, 19 sites (7 east, 12 west) in 2016, and 11 sites (3 east, 8 west) in
2017
Coral Reefs
123
Seaward slope surveys
Seaward slope surveys took place from mid-June to mid-
September, when signs of heat stress would be expected.
Coral quadrat and photo-transect surveys yielded similar
bleaching prevalence values (pale and bleached colonies),
with island-wide means of 3 ±3% and 3 ±2% (n= 3851
colonies censused), respectively. Bleaching was not reported
by the authors or other observers (such as Guam’s Eyes of
The Reef participants) at other reef areas during the period
coinciding with the temperature anomaly. The results of the
quadrat surveys indicated that an average of 13 ±9% of
observed colonies exhibited partial to full-colony mortality
attributable to the 2013 and 2014 events. The prevalence of
mortality attributed to the 2013 and 2014 events was similar
for the eastern windward and western leeward sites
(14 ±12% and 12 ±4%, respectively) (Fig. 3).
Staghorn Acropora populations
Prior to 2013, staghorn populations covered a total of 33.3 ha
and were composed of monospecific, or occasionally mixed-
species, stands. Eight species had been identified from
Guam: Acropora cf. pulchra, A. cf. intermedia, A. muricata,
and A. aspera existed in extensive thickets up to 7.8 ha in
area, while A. vaughani, A. virgata, A. austera, and A. teres
were rarer, and limited to individual colonies or small
thickets of \12 m
2
at several sites. In 2015, two species, A.
aspera and A. virgata, were observed in one site each, and a
third species, A. teres, was observed at two sites. Acropora
vaughani was not observed in 2015 and has not been seen
since, which suggests that it may be extirpated from Guam’s
waters. Of the remaining three staghorn species, A. cf. pul-
chra was the most common, found in 16 of the 21 sites
surveyed (Raymundo et al. 2017).
Surveys in 2015 documented cumulative mortality from
both elevated SSTs and the onset of extreme low tides
across 2013 and 2014; thus, it was not possible to attribute
mortality within these populations to specific years or
events. Cumulative mortality from the 2013–2015 events,
which was first reported in Raymundo et al. (2017), was
estimated at 53% ±10%. One extensive population suf-
fered complete mortality, and live coral cover at eight
others was estimated to have decreased by 70% or more.
Synopsis of events in 2016
Environmental parameters
Conditions around Guam warranted Bleaching Watch sta-
tus by 18 May. Sea surface temperature reached 30 C, and
a Bleaching Warning was issued on 19 July. A maximum
SST of 30.9 C was recorded on 25 July (Fig. 2, 2016).
In situ temperatures above 30 C were recorded at the Ipan
buoy starting on 30 May, and the maximum reef flat
temperature recorded was 35.6 C, in July, from Tumon
Bay (West-central Guam). The Ritidian buoy was offline
until 16 June but recorded in situ temperatures above 30 C
beginning on 28 June and maximum temperatures above
32 C in July and August (Fig. 2, 2016). Conditions
reached Alert Level 1 for seventeen days starting on 27
August, and accumulated heat stress peaked at 5.5 DHW
between 9 September and 10 October. Sea surface tem-
perature decreased after tropical disturbances and mon-
soons in August and September, but remained elevated
through 19 December.
Seaward slope surveys
In contrast to observations at reef flat sites, heat stress-
associated bleaching was very low along the shallow sea-
ward slope, with the percent of pale or bleached coral cover
at 0.1 ±0.5%, and no bleaching-associated mortality
(Fig. 3). Similar to the 2015 bleaching site re-surveys, the
2016 reef resilience surveys began prior to the onset of the
temperature anomaly but continued into months during
which bleaching would be expected.
Reef flat rapid site assessments
A total of 13,640 observations were made on corals during
26 surveys at eight reef flat sites, representing 18 coral
genera and at least 37 species. Bleaching prevalence across
all sites and survey dates was 46% ±17% (mean ±SD).
Average bleaching prevalence across sites between August
and September was 53% ±15%, and from October to
December mean prevalence dropped to 39% ±16% as
water temperatures cooled. The most severe bleaching
impacts captured by these surveys were reported at Agat
(Fig. 1), where a maximum temperature of 35.3 C was
recorded in July; impacts at this site included 70% of
colonies impacted by paling, bleaching, and mortality in
August, 78% impacted in September, 56% in October, and
58% in December (refer to Table 1). Significantly, this site
also exhibited the complete loss of an extensive staghorn
bed, documented in the 2015 staghorn surveys (see 2015
Synopsis). Goniastrea, Acropora, and Isopora showed the
highest BMIs, but the two most common genera, Porites
and Leptastrea, showed relatively low BMIs (Table 2).
Synopsis of events in 2017
Environmental parameters
Sea surface temperature exceeded 29 C on 30 April and
30 C on 8 June. Water conditions were calm and no
Coral Reefs
123
Table 1 Mean bleaching prevalence (±SD) and mean bleaching mortality prevalence (±SD) for each coral taxon assessed in 2017
Taxon Survey Reef
zone
No. sites,
west
No. sites,
east
Total no.
colonies
Bleaching
prev.
Bleaching
mortality prev.
Scleractinian taxa
Acanthastrea echinata 1 s 5 2 68 19 ±29 0.9 ±3
Acropora abrotanoides 1 s 5 4 87 8 ±16 45 ±46
Acropora aspera 2 f 1 0 121 Na 61 ±0
Acropora austera 2 f, l, s 5 0 45 13 ±23 51 ±15
Acropora cf. azurea 1 s 7 1 29 12 ±25 36 ±40
Acropora cerealis 1 s 4 2 26 25 ±37 17 ±32
Acropora cophodactyla 1s2 0 50 15±38
Acropora digitifera 1 s 5 2 11 25 ±43 14 ±33
Acropora globiceps 1 s 6 3 64 21 ±29 43 ±40
Acropora humilis 1 s 6 2 103 15 ±19 44 ±36
Acropora cf. intermedia 2 f, l 5 0 141 18 ±36 54 ±23
Acropora latistella 1s2 0 315±38 0
Acropora monticulosa 1s1 0 18±28 0
Acropora muricata 2 f, l 7 0 104 Na 59 ±13
3 f 1 0 86 84 ±015±0
Acropora cf. nasuta 1s2 1 510±29 13 ±32
Acropora cf. pulchra 2 f, l 15 1 853 Na 56 ±13
3 f 3 0 140 60 ±35 13 ±12
Acropora secale 1 s 6 2 109 12 ±23 40 ±44
Acropora surculosa 1 s 8 4 128 51 ±34 29 ±31
Acropora tenuis 1 s 5 1 14 15 ±32 31 ±44
Acropora teres 2f1 0 5Na 44±0
Acropora valida 1 s 5 2 25 23 ±35 23 ±44
Acropora verweyi 1 s 2 3 59 7 ±14 18 ±35
Acropora virgata 2l1 0 3Na 92±0
Acropora sp. ‘‘quelchi’’ 1 s 1 2 3 8 ±28 0
Acropora sp. ‘‘wardii’’ 1 s 5 1 8 31 ±48 15 ±38
Acropora sp. 1 1 s 0 1 3 0 8 ±28
Acropora spp. caespitose 1 s 7 4 174 23 ±33 40 ±42
Acropora spp. corymbose 1 s 4 2 80 4 ±740±45
Acropora spp. other 1 s 3 2 30 0.4 ±238±50
Astrea curta 1 s 6 3 52 45 ±40 5 ±11
Astreopora elliptica 1 s 2 1 10 13 ±32 3 ±10
Astreopora gracilis 1s1 0 38±28 0
Astreopora listeri 1 s 4 3 28 26 ±43 8 ±27
Astreopora myriopthalma 1 s 8 4 163 35 ±25 0
Astreopora ocellata 1 s 2 0 20 13 ±32 0
Astreopora randalli 1 s 4 2 17 33 ±43 0
Astreopora spp. 1 s 6 2 202 36 ±40 0.3 ±1
Coscinaraea columna 1s4 1 539±51 0
Cycloseris sp. 1 s 1 0 1 8 ±28 0
Cyphastrea chalcidicum 1 s 5 3 23 20 ±31 0
Cyphastrea serailia 1 s 8 3 33 19 ±31 0
Cyphastrea spp. 1 s 3 2 48 13 ±22 2 ±6
Diploastrea heliopora 1 s 7 2 79 8 ±13 0
Dipsastrea danae 1 s 4 2 35 23 ±37 3 ±9
Coral Reefs
123
Table 1 continued
Taxon Survey Reef
zone
No. sites,
west
No. sites,
east
Total no.
colonies
Bleaching
prev.
Bleaching
mortality prev.
Dipsastrea favus 1 s 9 3 144 70 ±30 0.3 ±1
Dipsastrea helianthoides 1 s 3 3 48 25 ±39 13 ±28
Dipsastrea matthaii 1s1 0 10 0
3f1 0 10±00
Dipsastrea pallida 1 s 6 3 99 63 ±46 0.4 ±1
Dipsastrea sp. 1 1 s 3 0 6 0 0
Dipsastrea spp. 1 s 7 4 511 20 ±38 0
Echinophyllia echinata 1s1 0 18±28 0
Echinopora pacificus 1s3 2 627±44 12 ±30
Euphyllia glabrescens 1s1 0 10 0
Favites flexuosa 1 s 5 2 18 37 ±47 8 ±28
Favites rotundata 1 s 0 2 14 3 ±11 0
Favites russelli 1 s 6 2 65 47 ±41 0
Favites spp. 1 s 3 0 10 20 ±38 0
Fungia fungites 1 s 2 0 14 10 ±29 0
Fungia granulosa 1s1 0 10 0
Fungia paumotensis 1s1 0 28±28 0
Fungia scutaria 1 s 3 0 10 12 ±30 0
Galaxea fascicularis 1 s 9 4 322 9 ±13 1 ±4
Gardineroseris planulata 1 s 6 1 22 54 ±52 0
Goniastrea edwardsii 1 s 9 2 250 56 ±31 17 ±25
3f1 0 333±00
Goniastrea minuta 1s1 0 10 8±28
Goniastrea pectinata 1 s 7 2 50 43 ±48 24 ±43
Goniastrea retiformis 1 s 9 3 351 50 ±28 16 ±20
3 f 4 2 17 57 ±49 0
Goniastrea stelligera 1 s 8 4 172 35 ±33 25 ±35
Goniopora fruticosa 1 s 5 1 41 1 ±40
Goniopora minor 1s2 0 60 0
Goniopora somaliensis 1s1 0 28±28 0
Goniopora tenuidens 1s1 0 10 0
Goniopora spp. 1 s 7 2 42 8 ±20 4 ±14
Herpolitha limax 1s6 0 731±48 0
Hydnophora microconos 1 s 7 4 93 66 ±41 2 ±5
Isopora palifera 1 s 0 1 42 1 ±42±7
3 f 1 1 197 55 ±01±0
Leptastrea pruinosa 1s1 1 30 0
Leptastrea purpurea 1 s 9 4 415 13 ±15 0 ±1
3 f 6 2 2060 15 ±18 0 ±0
Leptastrea spp. 1 s 4 2 23 7 ±18 0
Leptastrea transversa 1 s 4 3 16 22 ±38 0
Leptoria phrygia 1 s 8 4 436 74 ±27 3 ±5
Lobophyllia hemprichii 1 s 8 2 46 77 ±44 0 ±1
Lobophyllia sp. 1 s 1 0 8 3 ±10 1 ±4
Merulina ampliata 1s4 1 919±38 8 ±28
Montipora caliculata 1s3 0 523±44 0
Montipora danae 1s1 0 18±28 0
Coral Reefs
123
Table 1 continued
Taxon Survey Reef
zone
No. sites,
west
No. sites,
east
Total no.
colonies
Bleaching
prev.
Bleaching
mortality prev.
Montipora foveolata 1 s 5 2 48 35 ±42 17 ±29
Montipora grisea 1s3 0 40 19±38
Montipora hoffmeisteri 1 s 5 2 47 36 ±40 17 ±26
Montipora informis 1s1 0 33±90
Montipora monasteriata 1s1 0 44±14 4 ±14
Montipora planiuscula 1s1 0 10 0
Montipora tuberculosa 1s4 1 78±19 12 ±30
Montipora verrucosa 1 s 8 4 138 42 ±36 50 ±36
Montipora spp. 1 s 9 3 370 51 ±33 35 ±33
Oulophyllia crispa 1 s 6 4 24 68 ±47 2 ±6
Pachyseris speciosa 1s2 0 315±37 0
Pavona bipartita 1s2 0 68±19 4 ±14
Pavona chiriquiensis 1 s 5 2 45 24 ±30 0
Pavona decussata 3 f 2 1 214 6 ±80
Pavona cf. diffluens 1s1 0 18±28 0
Pavona divaricata 3 f 1 1 39 63 ±52 0
Pavona duerdeni 1 s 6 4 43 66 ±46 11 ±28
Pavona maldivensis 1s3 1 821±38 10 ±24
Pavona sp. 1 ‘‘white collines’’ 1 s 1 0 1 8 ±28 0
Pavona spp. 1 s 7 2 24 0 0
Pavona varians 1 s 8 2 46 28 ±32 0
Pavona venosa 1 s 4 0 15 4 ±14 0
3f1 0 10±00
Platygyra daedalea 1 s 8 4 146 55 ±35 17 ±28
Platygyra pini 1 s 9 4 242 ## ±21 10 ±17
3 f 2 0 21 39 ±56 0
Plesiastrea versipora 1s1 1 28±28 0
Pocillopora ankeli 1 s 6 1 30 38 ±45 10 ±17
3f1 0 813±00
Pocillopora damicornis 1 s 4 1 26 8 ±28 4 ±16
3 f 6 2 497 60 ±24 0 ±0
Pocillopora danae 1s1 0 18±28 0
Pocillopora elegans 1 s 4 2 24 29 ±43 9 ±22
Pocillopora grandis 1 s 7 4 98 41 ±48 28 ±32
Pocillopora cf. ligulata 1 s 5 3 21 41 ±48 4 ±14
Pocillopora meandrina 1 s 8 4 184 31 ±25 39 ±32
Pocillopora setchelli 1s1 1 91±48±28
Pocillopora verrucosa 1 s 9 4 304 41 ±16 32 ±24
3f2 0 575±33 0
Pocillopora woodjonesi 1s0 1 20 8±28
Pocillopora spp. 1 s 7 2 119 24 ±28 30 ±42
Porites annae 1 s 4 0 33 0 ±10
3 f 4 2 38 41 ±44 0
Porites australiensis 1s5 0 94±14 0
3f1 0 4##±00
Porites cylindrica 1s2 0 20 0
3 f 2 0 476 37 ±25 0 ±0
Coral Reefs
123
Table 1 continued
Taxon Survey Reef
zone
No. sites,
west
No. sites,
east
Total no.
colonies
Bleaching
prev.
Bleaching
mortality prev.
Porites deformis 1 s 8 0 56 9 ±28 0
Porites densa 1s2 1 93±90
Porites lichen 1 s 5 3 47 24 ±27 21 ±26
Porites lobata 1s1 0 20 0
Porites lutea 1 s 4 0 15 10 ±29 3 ±6
Porites sp. 1 1 s 1 0 2 4 ±14 0
Porites spp. massive 1 s 9 4 1099 35 ±10 12 ±8
3 f 5 2 645 49 ±26 0.1 ±0.3
Porites spp. submassive 1 s 1 2 20 13 ±29 3 ±9
Porites spp. other 1 s 3 1 12 3 ±98±28
Porites cf. myrmiodonensis 1s3 0 68±28 9 ±28
3 f 1 0 58 55 ±029±0
Porites rus 1 s 9 2 312 2 ±30
3 f 2 0 87 63 ±52 4 ±6
Porites vaughani 3f1 0 367±00
Psammocora contigua 1 s 4 0 10 18 ±37 0
3 f 2 1 15 0 ±00
Psammocora haimeana 1s3 1 527±44 4 ±14
Psammocora nierstraszi 1 s 5 1 20 29 ±46 0
Psammocora profundacella 1 s 4 2 12 35 ±47 0
Psammocora stellata 3f1 0 1##±00
Psammocora sp. 1 1 s 1 0 1 0 0
Psammocora spp. 1 s 6 1 35 26 ±34 1 ±4
Sandalolitha dentata 1s1 0 18±28 0
Scapophyllia cylindrica 1s0 1 10 8±28
Stylocoeniella armata 1s3 0 512±30 0 ±0
3f1 0 1##±00
Stylocoeniella guentheri 1s1 0 20 0
Stylophora sp. ‘‘mordax’’ 1 s 5 4 133 8 ±15 61 ±44
3f1 0 30±0 100 ±0
Turbinaria reniformis 1s1 1 38±28 0
Turbinaria stellulata 1s1 1 412±30 0
Non-scleractinian taxa
Sinularia spp. 3 f 4 0 239 86 ±14 0
Heliopora coerulea 1 s 3 3 63 25 ±32 0
3 f 3 0 40 20 ±17 0
Millepora dichotoma 1 s 1 0 11 2 ±80
Millepora platyphylla 1 s 8 4 122 53 ±41 0
Millepora tuberosa 1s1 0 18±28 0
Mean bleaching prevalence was calculated as the percentage of total observed colonies recorded as pale or bleached, averaged across all survey
sites; colony assessments were pooled within sites. Staghorn mortality assessments were not conducted during a bleaching events, and thus,
bleaching prevalence is not included here (these cases are denoted as ‘‘Na’’). Mean bleaching mortality prevalence was calculated as the
percentage of total observed colonies with partial to full-colony bleaching mortality, averaged across all survey sites. Survey methods are
denoted as 1 = off-transect coral condition assessment during island-wide bleaching assessments (5 m and 12 m data pooled), 2 = reef flat
staghorn mortality assessment, and 3 = rapid reef flat site assessment. Reef zones are denoted as l = lagoon patch reef, f = reef flat platform, and
s = seaward slope
Coral Reefs
123
storms or significant monsoons or cyclones developed
(Fig. 2, 2017). Both satellite-derived and buoy tempera-
tures exceeded 31 C in August. Maximum in situ reef flat
temperatures of 34.4 C and 34.8 C were recorded from
Tumon Bay in June and August, and 35 C in Agat in June.
Conditions warranted Alert Level 1 on 6 August, Alert
Level 2 on 22 August, and remained at Alert Level 2 status
for 57 d. Temperatures remained above 29 C through the
end of the year, and bleached corals were observed at
depths in excess of 30 m in October. Accumulated heat
stress reached a peak of 13 DHW in mid-October,
exceeding the previous record high of 12 DHW in 2013
(Fig. 2, 2017).
Seaward slope surveys
The impacts of the 2017 bleaching event on shallow sea-
ward slope communities exceeded those observed in 2013,
with an island-wide mean of 33 ±12% pale or bleached
coral cover and an additional 15 ±17% of coral cover
exhibiting bleaching-associated mortality (Fig. 3). In total,
48 ±17% of coral cover was impacted by the 2017
bleaching event, compared to the 32 ±19% of coral cover
impacted by the 2013 bleaching event (t(39.2) = 2.61,
p= 0.013). In contrast to the pattern observed in 2013,
percentages of pale, bleached, or recently killed coral cover
at the seaward slope sites during the 2017 bleaching event
were similar for both the eastern windward (52 ±21%)
and western leeward coral communities (46 ±17;
t(9) = 2.26, p= 0.66) (Fig. 3). We recognize, however,
that the limited number of windward sites (n= 3; Fig. 1)
likely affected this comparison. Observations from semi-
quantitative, off-transect surveys revealed prevalence of
pale, bleached, and/or bleaching mortality was 61 ±12%,
slightly greater than the severity calculated from photo-
transects (t(11.7) = 2.19, p\0.001).
Semiquantitative, off-transect surveys also revealed that
92% of all surveyed coral taxa and 98% of coral genera
exhibited paling, bleaching, or bleaching-associated mor-
tality (summarized in Table 1). The five coral genera with
the highest percentages of normally pigmented colonies
were as follows: Galaxea (94%, n= 322), Goniopora
(92%, n= 92), Leptastrea (88%, n= 457), Diploastrea
(81%, n= 80), and Cyphastrea (64%, n= 104). (This
excludes Euphyllia, which was represented by a single
unbleached colony seen in one site.) In contrast, the five
coral genera with the highest percentage of full-colony
bleaching-associated mortality include Acropora (55%,
n= 967), Stylophora (35%, n= 133), Montipora (23%,
n= 62), and Millepora (21%, n= 134), and Isopora (19%,
n= 42). The genus Porites, dominated by massive species
at these depths, also had a relatively high proportion of
normally pigmented colonies (64%, n= 1627), but this was
lower than expected, given that Porites is generally thought
to be bleaching resistant.
Staghorn Acropora populations
All 21 staghorn populations re-assessed in 2017 showed
reductions live cover ranging from 29 to 100%. Four
populations were devoid of any living tissue and consisted
of standing dead skeleton or rubble piles, three others
showed [70% dead skeleton within the existing areal
extent, while seven populations showed [50% live cover
(Fig. 4). Overall, surveys revealed a reduction in total area
of live coral cover from 33.3 ha prior to 2013 to an
Table 2 Bleaching mortality
index (after McClanahan et al.
2004) calculated for genera
surveyed at eight rapid reef flat
assessment sites in 2016 and
2017
Genus 2016 2017
BMI n% of total BMI n% of total
Goniastrea 25.93 9 0.2 18.33 20 0.4
Platygyra 0.00 0 0.0 23.81 21 0.4
Psammocora 10.71 28 0.6 3.51 19 0.4
Heliopora 4.27 39 0.8 7.50 40 0.8
Acropora 24.65 169 3.5 36.26 273 5.5
Pavona 3.57 196 4.1 4.10 260 5.2
Sinularia 15.71 227 4.7 30.82 239 4.8
Isopora 24.16 229 4.8 18.95 197 3.9
Pocillopora 15.03 408 8.5 15.20 544 10.9
Porites 9.57 1337 28.0 18.73 1315 26.4
Leptastrea 5.56 2139 44.7 2.20 2060 41.3
Total colony count 4781 4988
N= total number of colonies of each genus assessed during the September surveys in each year; % of total
is the percent contribution of that number to the total population of colonies counted
Coral Reefs
123
estimated live cover of 21.3 ha, a loss of 36%. Recovery,
via resheeting over dead skeleton or larval recruitment onto
dead skeleton by other species, was observed in seven
populations with extensive thickets, though only one
staghorn recruit was seen in one site. Populations with high
mortality subjected to physical disturbance (high wave
energy or high human use) were reduced to rubble and
showed no recovery.
Rapid reef flat site assessments
A total of 9670 observations were made during 15 surveys
at the eight sites, representing 18 coral genera and at least
35 species. Mean bleaching prevalence across surveyed
reef flat sites for all surveys was 35 ±4%. Mean bleaching
prevalence across sites for September was 40 ±7%, and
for October to November was 29 ±5%. As in 2016, the
most severe bleaching was observed in Agat; bleaching
prevalence at this site was 74% in late September. BMIs
were higher in Acropora, Heliopora, Porites, and Sinularia
than those calculated for 2016, but lower in Psammocora,
Goniastrea, and Isopora (Table 2). Other genera showed
similar responses between years.
Cumulative impacts to coral cover
Reef flat communities
Live coral cover on monitored shallow reef flats declined
by 36% between 2012 (pre-bleaching) and 2017
(F= 6.135; p= 0.023) (Table 3). Sites dominated by
staghorn Acropora showed substantial cumulative loss
(between 43 and 80%), which represented a statistically
significant overall decline at two sites (West Agan
˜a and
Tanguisson: F= 17.07; p=\0.0001) (Fig. 4). In contrast,
coral cover at sites dominated by Porites did not decline
significantly and at one site, the Piti Bomb Holes Marine
Preserve, cover increased slightly (by 4%). Interestingly,
Tumon, located in the heart of the tourism district and
Guam’s most prominent marine preserve, exhibited the
widest annual fluctuation in coral cover, showing a total
estimated loss of 49%, but recovering between bleaching
events in 2015 and 2017, with a net gain of 24% during
these two years. The extreme low tide episodes recurring
throughout 2015 contributed to the mortality rate, despite
the fact that 2015 was not a significant bleaching year.
Staghorn Acropora were particularly impacted by subaerial
exposure; the monitored West Agan
˜a site declined from 29
to 7% live coral cover across the January, May, and
December survey periods, with the majority of the loss
from the site’s primary staghorn thicket (Table 3, Fig. 4).
Shallow seaward slope communities
Mean coral cover at the shallow seaward slope sites
declined from 25 ±13 to 18 ±8%, (t(36.7) = 2.66,
p= 0.012) between 2013 and 2015, and from 18 ±8to
13 ±8% between 2015 and 2016 (t(20.1) = 2.35,
p= 0.029) (Fig. 5). No statistically significant difference
in seaward slope coral cover was detected between 2016
and 2017 (t(15.3) = 1.39, p\0.183). When considering
the entire 2013–2017 period, island-wide mean coral cover
at the shallow seaward slope sites declined by 34% (from
25 ±13 to 17 ±9%, t(38.9) = 2.24, p= 0.031). The
Fig. 4 Change in estimated
areal population size in 18
surveyed staghorn Acropora
populations around Guam. The
graph depicts the areal extent of
the sites surveyed prior to 2013
compared to the estimated
extent in 2017 calculated from
bleaching mortality. Populations
with an original size \100 m
2
are excluded here (n=3)
Coral Reefs
123
decline in coral cover as a result of the 2013 and 2014
events was greatest at the eastern windward sites (-45%,
from 29 ±13 to 16 ±7%, t(16) = 3.57, p= 0.003), while
no significant difference in coral cover was detected at the
western leeward sites over the same time period
(t(19.4) = 0.68, p= 0.51). The disparity observed in
impacts to coral cover between eastern windward and
western leeward sites between 2013 and 2017 was even
greater than that observed between 2013 and 2015, with a
59% decline (from 29 ±13 to 12 ±1%, t(16.6) = 6.05,
p\0.001) at the eastern sites and no significant difference
in coral cover observed at western sites (t(20.4) = 0.9,
p= 0.377). However, it should be noted that the small
sample size of eastern seaward slope sites surveyed in 2017
may not be representative of the full extent of the wind-
ward side of the island.
Staghorn Acropora populations
All known staghorn populations were assessed in 2015 and
2017. Total live coral cover loss was estimated at 53% in
2015 and 36% in 2017, based on the total 33.3 ha areal
extent measured prior to 2013. All populations experienced
bleaching-induced mortality in similar patterns. Large
thickets (C0.3 ha in size; 11 out of the 21 sites) showed
high mortality in the center of the stands, with remaining
live tissue limited to thicket margins. In sites that had
experienced high storm-driven wave energy, thickets with
high mortality were reduced to rubble, with no signs of
recovery. Outbreaks of a rapidly progressing white syn-
drome were observed in Tumon in 2016 (18.7% preva-
lence), and in three sites in 2017 (Tumon, Tanguisson, and
Apra Harbor; 13.4%, 10.2%, and 24% prevalence,
respectively) during the bleaching season, which caused
additional mortality. Resheeting over dead skeleton was
observed by 2017 in thickets with large central dead
patches, which accounted for lower total mortality esti-
mated in 2017. However, three species, Acropora aspera,
A. virgata, and A. teres, were all reduced to a single stand,
with estimated mean mortalities of 62% ±37%,
92% ±10%, and 44% ±23%, respectively. One species,
Acropora vaughani, possessed no live tissue in 2017,
though isolated clumps were observed in Apra Harbor in
2015; it is likely extirpated from Guam.
Evidence of species replacements within staghorn
thickets was observed at three sites with high mortality but
remaining intact structure. Recruits of several common reef
flat species: Porites cylindrica, Pavona decussata, P.
divaricata, Pocillopora damicornis, Psammocora con-
tigua, and Leptastrea purpurea recruited onto standing
dead skeletal structure and had begun consolidating it
during the 2017 surveys.
Discussion
A recent study of coral cores by Cybulski (2016) noted
dominance by Acropora on Guam reefs for the previous
500 yrs, with a human disturbance-driven shift to Pocillo-
poridae 100 yrs ago. The author noted no evidence of
significant bleaching-related mortality within the 500 yr
period, indicating that the magnitude of recent heat stress
impacts is unprecedented over at least 500 yrs. The level of
heat stress associated with the 2013 sea surface tempera-
ture anomaly around the island was the highest since
satellite measurements began, but this record was exceeded
in 2017. Lower-magnitude, but still historically significant,
heat stress events occurred in two of the three intervening
years. Guam was impacted by warming events two years
prior to the 2015-16 ENSO event that affected other
countries in the region, and this effect was prolonged
through 2017, during the subsequent La Nin
˜a. The ENSO
Table 3 Change in live coral cover within five monitored reef flats along western Guam, 2012–2017
Year Tanguisson %
Change
Tumon %
Change
West
Agan
˜a
%
Change
Piti %
Change
Luminao %
Change
2012 29.3 Na 52.8 Na Na 29.9 Na 35.2 Na
2013 20.0 29.3 45.4 27.4 Na 28.8 21.1 40.6 5.4
2014 13.8 26.2 26.7 218.7 29.3 Na 31.8 3.0 27.1 213.5
2015 13.6 20.2 40.7 14.0 29.3 0.0 28.0 23.8 24.7 22.4
2016 8.9 24.7 16.7 224.0 16.5 212.8 30.5 2.5 27.2 2.5
2017 5.8 23.1 27.0 10.3 16.6 0.1 34.0 3.5 28.3 20.2
Net change per site 223.5 21.5 212.7 4.1 28.3
Total % change per site 280.2 248.9 243.3 13.7 219.6
N= 3 transects per site. Mean % cover taken during the last quarter surveys for each year presented here. Numbers in bold represent net losses in
cover between years
Mean % change across sites: -35.7%
Coral Reefs
123
event itself triggered repeated extreme tide episodes that
killed exposed corals, which were then subjected to further
warming in the subsequent year.
Differential responses to events
The past five years of repeated, anomalous environmental
events triggered profound and sudden change in the
structure, and likely the function, of Guam’s reefs. Coral
cover on monitored reef flats along the western coast
declined 37% by 2017. Staghorn Acropora communities
were particularly devastated; three experienced complete
mortality. Nearly a third of coral cover was lost island-
wide along the shallow seaward slope between 2013 and
2017, with approximately 60% of coral cover lost along the
eastern windward coast. An earlier analysis of 2013 data
concluded that the difference in bleaching prevalence
observed between windward and leeward sites was, in part,
attributable to the greater proportion of bleaching-suscep-
tible taxa (primarily acroporids) within shallow windward
coral communities (Reynolds 2016).
The decline in shallow seaward slope coral cover
observed between 2015 and 2016 did not appear directly
associated with thermal stress-driven mortality, as bleach-
ing prevalence at seaward slope sites during this period was
very low and no mortality was reported. The cause of this
decline could not be determined from this preliminary
analysis, but observations by LR, DB, and WH suggest that
some of the mortality may have been attributed to elevated
coral disease (white syndrome) and corallivorous snail
(mainly Drupella) predation. The minimal bleaching
prevalence at seaward slope sites in 2016 was in contrast to
significant bleaching impacts recorded at reef flat sites that
same year. Thus, our data suggest differential responses to
heat stress between reef flat and seaward slope coral
communities. Differences in environment (such as water
circulation), species-specific responses of dominant taxa,
and a latent effect of the 2015 extreme low tide events on
the bleaching susceptibility of reef flat corals could have
contributed to the mortality differences we observed
between these distinct reef zones. No significant decline in
coral cover was observed at shallow seaward slope sites
between 2016 and 2017, despite the more severe thermal
stress experienced in 2017. However, the 2017 bleaching
response surveys were carried out while the event was
ongoing; quantitative surveys were not conducted follow-
ing the complete dissipation of thermal stress in late 2017.
The higher percentage of bleaching-impacted coral cover
recorded during the 2017 event, and qualitative observa-
tions of catastrophic mortality of shallow-water Acropora
species at several seaward slope sites in 2018, suggest a
loss in coral cover at shallow seaward slopes comparable to
the 2013 event.
A critical methods analysis
This assessment of recent bleaching-associated impacts to
Guam’s coral reefs relied upon quantitative or semiquan-
titative datasets generated by five separate survey types that
involved a total of seven individual survey methods
(summarized in Table S1). The sampling approaches
included that used by an existing long-term monitoring
Fig. 5 Box plots of percent coral cover values from shallow (5 m)
seaward slope benthic photo-transect surveys between 2013 and 2017
for all sites island-wide, eastern windward sites, and western leeward
sites. Data were obtained from a total of 46 sites (21 east, 25 west) in
2013, 17 sites (9 east, 8 west) in 2015, 19 sites (7 east, 12 west) in
2016, and 11 sites (3 east, 8 west) in 2017
Coral Reefs
123
program of benthic cover and coral size/condition data at
high priority reef flat sites, and four that were developed
specifically to assess bleaching severity and bleaching-as-
sociated impacts to benthic cover during and after the
events documented above. These survey methodologies
were developed and implemented in succession, as the
Guam Coral Reef Response Team adaptively allocated
limited field survey capacity based on need, and as
informed by personal observations and preliminary analy-
sis of existing data.
The implementation of different survey methods to
document the extent and severity of coral bleaching events
on Guam resulted from:
(1) The need to take advantage of data produced by an
existing survey effort though this program was not
specifically designed to assess bleaching impacts as
they are occurring;
(2) The need for methods appropriate for particular reef
zones or communities;
(3) Limited personnel and resources available to carry
out surveys; and
(4) The timing of the surveys, also related to resource
availability.
Multiple sampling approaches and survey methods are
often necessary to understand the impacts of acute distur-
bances, such as coral bleaching events, across distinct coral
reef communities and at different spatial and temporal
scales. However, the integration of multiple datasets into a
single, comprehensive analysis requires an understanding
of the limits of each methodology and the potential biases
that manifest in the generated datasets. A critical analysis
of these methods, and the datasets they generate, can
inform changes to an approach to assessing the impacts of
acute disturbances, with the aim of maximizing data
accuracy and comparability. Here, we present a qualitative
analysis of the sampling approaches and survey methods
used in the present study, with recommendations for
improving comparability of datasets generated by different
methodologies across different reef communities, as well
as suggestions for further evaluating the comparability of
these data. Our rationale explaining the order of imple-
mentation of survey methods is as follows:
2013 The island-wide randomized approach targeting
the seaward slope maximized overlap with randomly gen-
erated sites surveyed by NOAA PIRSC in 2011. Recon-
naissance indicated this zone was most severely impacted
by bleaching stress; significant mortality had not yet been
observed at the reef flat/staghorn areas being monitored
when this decision was made.
2014 The NOAA CRW automated system did not catch
the warming event in time to allow for team mobilization;
as stated above, the 2014 bleaching event occurred six
months after the 2013 event. Documentation of 2014
events was thus limited to bleaching reconnaissance and
limited personal observations. Subsequent improvements
in the accuracy of the NOAA CRW early warning system
in 2015 greatly increased our capacity to mobilize for
future events.
2015 Resources were available for a subset of the island-
wide seaward slope sites surveyed in 2013; these were
prioritized to assess the cumulative impacts of the 2013 and
2014 events and to establish a new baseline against which
recovery and future impacts could be evaluated. Observa-
tions of the rapid onset of bleaching mortality among a
small number of staghorn communities in 2014 triggered
the island-wide staghorn coral mortality assessment in
2015 to address this major knowledge gap. The geographic
scale of the effort and limited staff availability necessitated
the adoption of a rapid assessment protocol that took
advantage of existing geospatial data developed by the
Guam Long-Term Monitoring Program.
2016 As Co-PI on a NOAA Saltonstall-Kennedy reef
resilience assessment project, DB helped develop a sam-
pling strategy that maximized overlap between resilience
assessment sites and seaward slope sites surveyed in 2013
and 2015. Resilience assessment field surveys conducted at
the beginning of the bleaching season were fortuitous in
that the data collected at both the seaward slope and lower
depths provided a good record of the limited impact of the
2016 temperature anomaly on these communities. In con-
trast, significant bleaching and mortality was observed at
rapid reef flat canary sites. Though semiquantitative, data
generated allowed BMI values to be derived, BMI was not
initially incorporated into our protocol but it provided a
tractable and rapid post hoc assessment of genera at risk,
and will be utilized as part of our standard protocol in the
future.
2017 Significant bleaching had not been observed at the
reef flat canary sites when Response Team members
observed severe bleaching along the seaward slope. Those
observations and NOAA CRW predictions informed the
timing of the re-survey of sites previously visited in 2013,
2015, and 2016. Additional mortality observed at several
staghorn sites as a result of extreme low tides in 2015, and
bleaching and disease in 2016, necessitated the re-survey of
all major staghorn communities around the island. A more
quantitative approach to assessing staghorn condition was
implemented at this time.
We identify the following specific issues with our
multiple survey bleaching response approach and highlight
changes we are incorporating to address these issues.
•Two metrics were used to assess bleaching prevalence
from photo-quadrats: percentage of total colonies and
percentage of total coral cover. A preliminary analysis
Coral Reefs
123
of data collected using both methods in 2013 suggests
that these values were significantly different for the
same sites. These differences are likely related to the
sensitivity of each measure to community composition
and size structure. We intend to quantitatively evaluate
these data sets to determine the nature and consistency
of these differences across taxa and population size
distributions, as they potentially influence our interpre-
tation of the data and subsequent management
decisions.
•The two methods used to assess staghorn mortality in
2015 versus 2017 were not comparable quantitatively.
The 2015 method allowed for a very rapid assessment
of a large number of sites in a very short period, which
was the intention of the surveys. However, the 2017
assessments provided a more quantitative reference
point for future assessments of the condition of
remaining staghorn beds. An effort to re-map the areal
extent of these beds is planned, but because satellite-
based photo-imagery cannot distinguish live versus
dead coral cover, these surveys will require in-water
ground-truthing.
•The use of quasi-permanent transect locations based on
GPS coordinates introduced additional variance into
key parameters such as species composition, and
undoubtedly lowered statistical power.
•The ability of the canary sites to detect island-wide
bleaching across reef zones was called into question in
2017. NOAA CRW products showed that thermal stress
for Guam peaked in October, when mean bleaching
prevalence at the canary sites was 29%. However, BMI
calculated after the events revealed differential generic
responses between 2016 and 2017, with three key
genera, Acropora, Porites, and Sinularia, showing
much more severe responses in 2017. Further, October
seaward slope assessments estimated bleaching and
bleaching mortality to be 48%, with observations of
bleaching to 40 m. These widely varying responses
between reef zones suggested that reef flat rapid
assessments may not adequately inform bleaching
response efforts for seaward slope sites. We plan to
incorporate the use of the BMI as part of our rapid
response protocol, as its utility as a means of identi-
fying taxa at particular risk and comparing taxon
performance between reef zones makes it a valuable
addition to our toolbox. We also intend to establish
permanent transects at these sites for future assessments
and to add shore-accessible seaward slope sites to
canary sites for rapid reconnaissance.
•The Guam Coral Bleaching Response Plan (Hoot and
Burdick 2017) was designed to leverage existing
mechanisms, such as projections provided by NOAA
CRW and temperature data from in situ loggers
deployed in long-term monitoring programs. In general,
in situ loggers recorded maximum temperatures of
1–2 C higher than satellite-derived SSTs. Thus, these
loggers may provide an effective local early warning
system at scales relevant to individual coral communi-
ties, in conjunction with CRW alerts.
A schematic of our response protocol is presented in
Fig. 6as a decision tree to guide activities according to
both event severity and resource availability. This protocol,
the bleaching response plan, and the formation of an
interagency Guam Coral Reef Response Team have
allowed local managers and researchers to proactively plan
for events before they occur and mobilize rapidly, thus
increasing effective resource allocation and improving
consistency among response efforts. However, maintaining
flexibility is key to success, given limited resources. As
overall improvements to our bleaching response, we will
(a) align reef flat, canary, and staghorn sites where possible
and appropriate; (b) align semiquantitative bleaching con-
dition assessments for all surveys and zones, to the extent
possible (i.e., standardize coral condition categories, sur-
vey area, and survey duration); (c) conduct power analyses
prior to surveys to determine appropriate sample sizes;
(d) conduct rapid calibration protocols between methods
and among personnel and discuss calibration results, to
better standardize surveys; and (d) coordinate data collec-
tion efforts for other projects to maximize site overlap and
timing to extent possible.
In summary, the use of multiple survey methods is often
necessary in order to adequately sample different reef
communities within distinct reef environments, and to take
advantage of data that may not have been collected for the
same purpose. Our results show that these different reef
communities can respond to an acute disturbance in sig-
nificantly different ways, and data collected for a single
reef community type, however broadly distributed it may
be within a biogeographic region, may not be representa-
tive of other reef communities and thus may not provide an
adequate measure of the totality of impacts to the reef areas
within that region.
The results of Jokiel et al. (2015) indicate that estimates
of broad-level parameters generated from different survey
methods, such as coral cover, appear to be relatively con-
sistent across most methods. However, their results also
indicate that measures of diversity can differ significantly
between methods. While the impacts of recent bleaching
events on the diversity of corals on Guam’s reefs were not
assessed in this study, further analysis will include an
examination of impacts to diversity, and thus should con-
sider known biases in estimates generated by different
sampling approaches.
Coral Reefs
123
Species loss and the trajectory of change
The possible extirpation of one species, Acropora vaugh-
ani, was noted in our surveys, and three others, A. aspera,
A. teres, and A. virgata, are reduced to single stands.
Wallace (1999) synonymized Acropora virgata with A.
formosa (now A. muricata) but Guam colonies, which are
identical to Dana’s A. virgata, are morphologically distinct
from A. muricata. This taxonomic quandary is illustrative
of the difficulty in ascertaining scleractinian coral species
boundaries, and the challenge this presents for coral species
conservation. Our analyses and recent post-bleaching
observations suggest that other species may also be at high
risk. Three previously common species: Stylophora pistil-
lata f. mordax and the caespitose Acropora cf. azurea and
A. verweyi have been greatly reduced across multiple sites.
Stylophora pistillata f. mordax declined precipitously in
2017, and few living colonies were observed during mul-
tiple visits to various sites in 2018. Acropora cf. azurea and
A. verweyi, typically found at 1–4 m depth along exposed
seaward slopes, were locally abundant prior to 2013 (refer
to Fig. 7a, d, f). While bleaching impacts recorded during
the event were not exceptional for these two species,
observations in 2018 at previously surveyed sites, and at
other reef areas where these species were locally abundant,
suggest catastrophic ([95%) mortality. This apparent
discrepancy may be a result of the relatively low number of
colonies encountered during surveys, which were con-
ducted at depths slightly greater than their preferred range,
or a sharp increase in mortality following survey comple-
tion. Observations of significant mortality among colonies
of the major structure-providing species, Acropora abro-
tanoides (Fig. 7d), as well as other species, such as Acro-
pora monticulosa and Acropora palmerae, once common
features of Guam’s reef front and shallow slope zone, have
also raised concern about the long-term viability of these
species. Thus, ‘‘safety in numbers’’ (Birkeland et al.
2013a,b) may not be providing a refuge from extirpation
for these key, highly susceptible species. Montipora ver-
rucosa and other encrusting Montipora species also
exhibited high rates of mortality, although field observa-
tions of mortality early in bleaching events, and the diffi-
culty in detecting dead Montipora colonies in benthic
photo-transect images, suggest that actual mortality rates
Fig. 6 A flowchart developed by the Guam Coral Reef Response Team as a decision tree for coral bleaching response, included in the Guam
Coral Bleaching Response Plan (Hoot and Burdick 2017)
Coral Reefs
123
among Montipora species were likely greater than what we
report here.
The trajectory and character of change in Guam reef
communities that has been initiated by the onset of repe-
ated, severe coral bleaching events will be a focus of future
analyses. For example, dead staghorn communities that
have already been reduced to rubble at several sites may
persist in this flattened state, as rubble is an
unstable recruitment substrate (Raymundo et al. 2007;
Birkeland et al. 2013a,b). The flattening of these stands
may, in turn, reduce the ability of reef flat platforms to
dissipate wave energy, particularly during storms (Ferrario
et al. 2014; van Beukering et al. 2007), and will likely
impact the populations of fish species that utilize staghorn
Acropora thickets for one or more phases of their life
history. Colonies that remained secured to existing
Fig. 7 Extent of bleaching within the different coral community
types on Guam. aMixed Acropora–Pocillopora eastern exposure
shallow seaward slope community bleaching in 2013. bStaghorn
Acropora pulchra thicket bleaching on a reef flat in 2016. cStaghorn
Acropora muricata bleaching in Apra Harbor in 2017. dEastern
exposure Acropora abrotanoides community bleaching in 2017.
eWestern exposure Porites community in 2017, showing bleached
massive Porites sp. fBleaching mortality of mixed Acropora–
Pocillopora eastern exposure shallow seaward slope community in
2018. Photo credits: D. Burdick, W. Hoot, L. Raymundo
Coral Reefs
123
substrate began resheeting within months after warming
subsided, via an apparent ‘‘phoenix effect’’ (Dias-Pulido
2009; Roff et al. 2014). This phenomenon involves the
survival of residual tissue deep in the skeleton, thus pro-
viding a tissue ‘‘reservoir’’ for regrowth. Observations
suggest resheeting of dead skeleton was largely responsible
for the increase in coral cover in the Tumon Bay and West
Agan
˜a thickets from 2016 to 2017, despite the severe
temperature anomaly in 2017 (Table 2). This effect was
enhanced at sites that were well flushed, as increased water
movement likely mitigated warming impacts by improving
coral resilience to heat stress (Nakamura and van Woesik
2001; Fifer 2018). However, recruitment of other species
onto dead staghorn skeleton was also common, suggesting
that colonies that do not resheet may be replaced by
developing communities comprised of Pavona, Pocillo-
pora, and Porites.
Coral diversity declines northward along the Mariana
Arc, a pattern primarily driven by the general northward
decrease in island size and differences in habitat type and
availability between the older, primarily carbonate islands
in the south and the younger, volcanically active islands in
the north (Richmond et al. 2008, Brainard et al. 2012). A
recent analysis of larval transport pathways in the region
concluded that Guam populations of coral taxa with pelagic
larval durations \20 d are primarily self-seeding, and that
there is a northward bias in larval transport within the
archipelago (Kendall and Poti 2015). The implications of
these findings, in light of recent coral bleaching-associated
impacts to Guam’s coral communities, are twofold:
Recovery following mass coral mortality events via larval
import for these taxa is likely to be negligible on Guam,
and as vulnerable coral populations decline on the high
diversity coral reefs of Guam, their ability to act as a larval
source for the northern reefs of the Mariana Archipelago
will decrease in the coming decades.
The future of Guam’s reefs
Ocean warming events of unprecedented frequency and
magnitude resulted in significant declines in coral cover at
reef flat and shallow seaward slope sites around Guam. The
variability in response to thermal stress between leeward
and windward communities appeared driven in part by
differences in the proportion of bleaching-susceptible taxa,
while variation in the responses of reef flat and seaward
slope communities may be driven by differences in com-
munity composition and the environmental regimes of
these distinct reef zones. Other environmental drivers of
bleaching patterns, such as cyclonic activity, wave height,
the timing and severity of anomalous warming, have not
yet been examined, but will be considered in future anal-
yses. Other possible drivers of the declines we observed,
such as disease, predation, and local stressors, must also be
considered.
The full impact to the diversity, structure, and function
of Guam’s coral reefs from the mortality documented in
this study, and implications for the future of Guam’s coral
reef ecosystem, remains unresolved. However, projections
of increasing bleaching frequency raise concern that coral
recovery will not keep pace with mortality. Should this
occur, the result will be a net loss in cover until a currently
undefined threshold is reached, beyond which recovery
may not be possible at timescales relevant to present
human communities. Van Hooidonk et al. (2016) predicted
that annual severe bleaching in the Mariana Islands could
begin by the early 2020s, but the events documented here
suggest that even this alarming estimation may have
overestimated the time that Guam’s shallow-water corals
have to acclimate and adapt to rapidly warming ocean
temperatures. Detailed documentation of ongoing changes
to community structure, key ecological processes, and the
status of vulnerable reef taxa is critical to formulating
effective management strategies for the conservation of
remaining reef diversity and function. Guam is likely a
sentinel for the type of near-future change that awaits other
small islands throughout the global range of coral reefs.
The lessons we learned from documenting and responding
to this experience may thus provide guidance and insight
for scientists and managers who are challenged by similar
impacts.
Acknowledgements M. Gawel of the National Parks Service, J. Cruz
and A. Leon-Guerrero of Guam EPA helped with initial development
of survey protocols; C. Tebeek, and A. Hershberger helped with
photo-transect surveys. Agat reef flat temperature data were provided
by the National Parks Service. The authors also acknowledge boat
captains T. Genereux and J. Miller. Funding for the Guam Long-term
Coral Reef Monitoring Program, which contributed substantial sup-
port to this study, is provided through a recurring NOAA Coral Reef
Conservation Program’s State and Territorial Coral Reef Conserva-
tion Cooperative Agreement; Grant Nos. NA13NOS4820012,
NA15NOS4820039, and NA17NOS4820038.
Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of
interest.
References
Birkeland CE, Green A, Fenner D, Squair C, Dahl AL (2013a)
Substratum stability and coral reef resilience: Insights from
90 years of disturbances to reefs in American Samoa. Microne-
sia 11:1–15
Birkeland CE, Craig P, Davis G, Edward A, Golbuu Y, Higgins J,
Gutierrez J, Idechong N, Maragos J, Miller K, Paulay G,
Richmond R, Tafileichig A, Turgeon D (2000) In: Wilkinson C
Coral Reefs
123
(ed) Status of the coral reefs of the world: 2000. Australian
Institute of Marine Science, Cape Ferguson. Pp. 199–217
Birkeland C, Miller MW, Piniak GA, Eakin CM, Weijerman M,
Elhany PM, Dunlap M, Brainard RE (2013b) Safety in numbers?
Abundance may not safeguard corals from increasing carbon
dioxide. Bioscience 63:967–974
Brainard RE, Asher J, Blyth-Skyrme V, Coccagna EF, Dennis K,
Donovan MK, Gove JM, Kenyon J, Looney EE, Miller JE,
Timmers MA, Vargas-Angel B, Vroom PS, Vetter O, Zgliczyn-
ski B, Acoba T, DesRochers A, Dunlap MJ, Franklin EC, Fisher-
Pool PI, Braun CL, Richards BL, Schopmeyer SA, Schroeder
RE, Toperoff A, Weijerman M, Williams I, Withal RD (2012)
Coral reef ecosystem monitoring report of the Mariana
Archipelago: 2003–2007. Pacific Islands Fisheries Science
Center Special Publication SP-12-01, 1019 pp
Bruno J, Siddon C, Whitman J, Colin P, Toscano M (2001) El Nin
˜o
related coral bleaching in Palau, Western Caroline Islands. Coral
Reefs 20(2):127–136
Burdick D, Brown V, Asher J, Caballes C, Gawel M, Goldman L,
Hall A, Kenyon J, Leberer T, Lundbald E, McIlwain J, Miller J,
Minton D, Nadon M, Pioppi N, Raymundo L, Richards B,
Schroeder R, Schupp P, Smith E, Zgliczynski B (2008) Status of
the coral reef ecosystems of Guam. Bureau of Statistics and
Plans, Guam Coastal Management Program. iv ?76 pp
Caballes CF (2009) The Role of Chemical Signals on the feeding
Behavior of the Crown-of-Thorns Seastar, Acanthaster planci
(Linnaeus, 1758). M.Sc. Thesis. 164 pp
Chesher RH (1969) Destruction of the Pacific corals by the sea star
Acanthaster planci. Science 165:280–283
Colgan MW (1987) Coral reef recovery on Guam (Micronesia) after
catastrophic predation by Acanthaster planci. Ecology
68(6):1592–1605
Cybulski J (2016) Push-core Sampling in Micronesia: Using Paleoe-
cological Data to Reconstruct Guam’s Coral Reef Community.
M.S. Thesis, American University. 90 pp
Derrick B, Russ B, Toher D, White P (2017) Test statistics for the
comparison of means for two samples which include both paired
observations and independent observations. Journal of Modern
Applied Statistical Methods 16(1):137–157
Diaz-Pulido G, McCook LJ, Dove S, Berkelmans R, Roff G, Kline
DI, Weeks S, Evans RD, Williamson DH, Hoegh-Guldberg O
(2009) Doom and Boom on a Resilient Reef: climate change,
algal overgrowth and coral recovery. PLoS One 4(4):e5239.
https://doi.org/10.1371/journal.pone.0005239
Donner SD (2009) Coping with commitment: Project thermal stress
on coral reefs under different future scenarios. PLoS One
4(6):e5712. https://doi.org/10.1371/journal.pone.0005712
Donner SD, Skirving WJ, Little CM, Oppenheimer M, Hoegh-
Gulberg O (2005) Global assessment of coral bleaching and
required rates of adaptation under climate change. Glob Chang
Biol 11:2251–2265
Ferrario F, Beck MW, Storlazzi CD, Micheli F, Shepard CC, Airoldi
L (2014) The effectiveness of coral reefs for coastal hazard risk
reduction and adaptation. Nat Commun 5:3794. https://doi.org/
10.1038/ncomms4794
Fifer J (2018) Examining Gene Expression of Heat-Stressed Staghorn
Coral Under Different Flow Environments. Graduate Program in
Biology, University of Guam. M.S. Thesis. 144 pp
Hoot WC, Burdick D (2017) Guam Coral Bleaching Response Plan.
Bureau of Statistics and Plans. 56 pp
Jokiel PL, Rogers KS, Brown EK, Kenyon JC, Aeby G, Smith WR,
Ferrell F (2015) Comparison of methods used to estimate coral
cover in the Hawaiian Islands. PeerJ 3:e954. https://doi.org/10.
7717/peerj.954
Kendall MS, Poti M (2015) Transport pathways of marine larvae
around the Mariana Archipelago. Tech. Memorandum NOS
NCCOS 193. 130 pp
MacNeil MA, Graham NAJ, Cinner JE, Wilson SK, Williams ID,
Maina J, Newman S, Friedlander AM, Jupiter S, Polunin NVC,
McClanahan TR (2015) Recovery potential of the world’s coral
reef fishes. Nature 520:341–344
McClanahan T, Baird AH, Marshall PA, Toscano MA (2004)
Comparing bleaching and mortality responses of hard corals
between southern Kenya and the Great Barrier Reef, Australia.
Mar. Pollut. Bull. 48:327–335
Nakamura T, van Woesik R (2001) Water-flow rates and passive
diffusion partially explain differential survival of corals during
the 1998 bleaching event. Mar Ecol Prog Ser 212:301–304
NOAA Coral Reef Watch (2017) updated daily. NOAA Coral Reef
Watch Version 3.0 Daily Global 5-km Satellite Virtual Station
Time Series Data for Guam, Jan 01, 2013-March 30, 2018.
College Park, Maryland, http://coralreefwatch.noaagov/vs/index.
phpf. Accessed 27 Sep 2018
NOAA Center for Operational Oceanographic Products and Services
(2013) updated daily. Daily sea level and temperature data at the
Apra Harbor, Guam Station ID: 1630000. https://tidesandcur
rents.noaa.gov/stationhome.html?id=163000. Accessed 05 Oct
2018
NOAA National Data Buoy Center (2018a) Historical Data for
Station 52200—Ipan, Guam. 2013-2017. https://www.ndbc.
noaa.gov/station_history.php?station=52200. Accessed 05 Oct
2018
NOAA National Data Buoy Center (2018b) Historical Data for
Station 52202—Ritidian, Guam. 2013-2017. https://www.ndbc.
noaa.gov/station_history.php?station=52200. Accessed 05 Oct
2018
NOAA Office for Coastal Management (2019) Digital Coast Histor-
ical Hurricane Tracks Viewer. https://coast.noaa.gov/hurricanes/.
Accessed 28 Apr 2019
PacIOOS (2018) Water Temperature Buoy Observations for Guam.
www.pacioos.hawaii.edu/water-category/buoy
Paulay G, Benayahu Y (1999) Patterns and consequences of coral
bleaching in Micronesia (Majuro and Guam) in 1992–1994.
Micronesica 31:109–124
Randall RH (2003) An annotated checklist of hydrozoan and
scleractinian corals collected from Guam and other Mariana
Islands. Micronesica 35–36:121–137
Randall RH, Holloman J (1974) Coastal Survey of Guam. UOGML
Tech. Rep. No. 14. 404 pp
Raymundo LJ, Maypa AP, Gomez EDD, Cadiz P (2007) Can
dynamite-blasted reefs recover? A novel, low-tech approach to
stimulating natural recovery in fish and coral populations. Mar
Pollut Bull 54:1009–1019
Raymundo LJ, Burdick D, Lapacek VA, Miller R, Brown V (2017)
Anomalous temperatures and extreme tides: Guam staghorn
Acropora succumb to a double threat. Mar Ecol Prog Ser
564:47–55
Reynolds T (2016) Environmental Regimes Predict the Spatial
Distribution of Coral Assemblages and Climate-Induced Bleach-
ing Patterns Around Guam. Graduate Program in Biology,
University of Guam. M.S. Thesis. 68 pp
Richmond RH, Houk P, Trianni M, Wolanski E, Davis G, Bonito V,
Paul VJ (2008) Aspects of the biology and ecological function-
ing of coral reefs in Guam and the Commonwealth of the
northern Mariana Islands. In: Riegl BM and Dodge RE (eds.)
Coral Reefs of the world Vol. I. Coral reefs USA. SpringerVer-
lag, Berlin. Pp. 719–739
Richmond R, Kelty R, Craig P, Emaurois C, Green A, Birkeland C,
Davis G, Edward A, Golbuu Y, Gutierrez J, Houk P, Idechong N,
Maragos J, Paulay G, Starmer J, Tafileichig A, Trianni M,
Coral Reefs
123
Vander Velde N (2002) In: Wilkinson C (ed.) Status of the Coral
Reefs of the World: 2002. Pp. 217–235
Roberts C, McClean CJ, Veron JEN, Hawkins JP, Allen GR,
McAllister DE, Mittermeier CG, Schueler FW, Spalding M,
Wells F, Vynne C, Werner TB (2002) Marine Biodiversity
Hotspots and Conservation Priorities for Tropical Reefs. Science
295:1280–1284
Roff G, Bejarano S, Bozec YM, Nugues M, Steneck RS, Mumby PJ
(2014) Porites and the Phoenix effect: unprecedented recovery
after a mass coral bleaching event at Rangiroa Atoll, French
Polynesia. Mar Biol 161:1385–1393
Van Beukering P, Haider W, Longland M, Cesar H, Sablan J,
Shjegstad S, Beardmore B, Liu Y, Garces GO (2007) The
economic value of Guam’s coral reefs. Univ Guam Mar Lab
Tech Rep 116:102
van Hooidonk R, Maynard J, Tamelander J, Gove J, Ahmadia G,
Raymundo L, Williams G, Heron SFSFSF, Planes S (2016)
Local-scale projections of coral reef futures and implications of
the Paris Agreement. Sci Rep 6:1–8
Wallace C (1999) Staghorn corals of the world: A revision of the
genus Acropora (Scleractinia; Astrocoeniina; Acroporidae)
worldwide, with emphasis on morphology, phylogeny and
biogeography. CSIRO Publishing. xviii, 422 pp
Wilkinson C (ed) (2000) Status of the Coral Reefs of the World: 2000.
Australian Institute of Marine Science, Townsville, p 361
Williams I, Zamzow J, Lino K, Ferguson M, Donham E (2012) Status
of coral reef fish assemblages and benthic condition around
Guam: A report based on underwater visual surveys in Guam and
the Mariana Archipelago, April-June 2011. U.S Dep. Commer.,
NOAA Tech. Memo NOAA-TM-NMFS-PIFSC-33,
22 pp. ?Appendices
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
Coral Reefs
123
A preview of this full-text is provided by Springer Nature.
Content available from Coral Reefs
This content is subject to copyright. Terms and conditions apply.