The state of coral reef ecosystems of Navassa Island

Full text

The State of Coral Reef Ecosystems of Navassa Island
117
Navassa Island
The State of Coral Reef Ecosystems of Navassa Island
Margaret W. Miller1, A. Gleason2, D. McClellan1 , G. Piniak4, D. Williams1,2, J.W. Wiener6, A. Gude3, J. Schwagerl5
INTRODUCTION AND SETTING
Navassa is a small (4.64 km2), uninhabited, oceanic island approximately 50 km off the southwest tip of Haiti (Figure
4.1) under the jurisdiction of the U.S. Fish and Wildlife Service. The island is a raised dolomite plateau ringed by vertical
cliffs that descend to a sloping submarine terrace at an approximate depth of 25 m, with coral reef development primarily
on small nearshore ledges and shelves. Navassa’s oceanic position in the Windward Passage exposes it to substantial
physical energy, with the eastern coastline exposed to persistent swells and regular storms and hurricanes. Both geo-
morphology and exposure have resulted in an absence of shallow-water inshore sh nursery habitats (e.g., mangroves,
sandy beaches and seagrasses) that are found on other islands in the region. The local and regional oceanography
around Navassa is poorly characterized, but detailed geology is provided in Miller et al. (in press).
Status of reef resources and threats have been documented by Miller and Gerstner (2002), Miller (2003) and Miller et
al. (2005) from data collected during expeditions in 2000 and 2002. These assessments reported relatively healthy coral
conditions and reef sh assemblages which, though dominated by small planktivores, still compared favorably with other
Caribbean locations. Substantial shing activity by transient Haitians was also reported.
Figure 4.1. A map of Navassa Island showing locations mentioned in the chapter. Map: K. Buja.
1 NOAA Fisheries, Southeast Fisheries Science Center
2 University of Miami, Rosenstiel School for Marine and Atmospheric Sciences
3 U.S. Fish and Wildlife Service, National Wildlife Refuge System Headquarters
4 NOAA-NOS, Center for Coastal Fisheries and Habitat Research
5 U.S. Fish and Wildlife Service, Caribbean Islands National Wildlife Refuge, Boqueron, Puerto Rico
6 Fondation pour la Protection de la Biodiversité Marine, Port-au-Prince, Haiti
75°W
75°W
18°24'N
18°24'N
Land
Water <30 m
Deep Water
0 0.25 0.5 km
"
Northwest
Point
Lulu Bay
Northeast
Point
East
Point
South
Point
West
Pinnacles
Conch
North
North Shelf
L u l u S h e l f
Video
Patch
"

The State of Coral Reef Ecosystems of Navassa Island
118
Navassa Island
ENVIRONMENTAL AND ANTHROPOGENIC STRESSORS
Climate Change and Coral Bleaching
No suitable historic observations are available to determine past occurrence or potential trends in coral bleaching at
Navassa, particularly through the 2005 Caribbean event. Observations in April 2006 indicated that bleaching was not
extensive (G. Piniak, pers. obs.). Miller et al. (2005) suggested that the relatively deep and exposed (i.e., high water
ow) position of most of the coral reefs surrounding Navassa reduced exposure to elevated sea surface temperatures
(SST). However extensive, and in some places, severe coral bleaching was observed at Navassa in November 2006,
when little new bleaching had been reported in the Caribbean and no predictions of bleaching (bleaching alerts) had
been issued based on satellite temperature records. In fact, observed bleaching prevalence was greater at deep sites
(20-30 m) than at shallow (7-10 m) sites. In situ temperature data was collected at a range of depths around Navassa
from April to November 2006 providing potentially useful data that can contribute to the future development of accurate
bleaching predictions for corals in deeper water (20-30 m). Please see the Water Quality and Benthic Habitats sections
of this chapter for more information on elevated SST and coral bleaching.
Diseases
Until 2004, coral diseases at Navassa were rarely seen (Miller et al., 2005). During the November 2004 expedition, how-
ever, a severe coral disease event was observed (Miller and Williams, 2007; see Benthic Habitat section). This disease
event appeared to have developed following hurricanes Charley and Ivan that affected Navassa in 2004. No sampling
for pathogen identication was possible, but disease signs were consistent with a white-plague type disease. Isolated
observations of disease on Acropora palmata (low prevalence) and A. cervicornis (much higher prevalence given this
species’ rarity at Navassa) were also made (Williams and Miller, pers. obs.)
Tropical Storms
Several named storms have passed near
Navassa in recent years (Figure 4.2), includ-
ing Ernesto (2006), Dennis (2005), Charley
(2004), and Ivan (2004). Unfortunately, the
wide spacing of observations makes it dif-
cult to attribute observed reef changes di-
rectly to storms. However, following Charley
and Ivan in 2004, some obvious physical
damage (e.g., toppled hard and soft coral
colonies) and sand movement was ob-
served.
Coastal Development and Runoff
Navassa is uninhabited, except for the tem-
porary presence of transient Haitian shers.
There has been no change in terrestrial ac-
tivity.
Coastal Pollution
No information about coastal pollution
sources from neighboring islands that have
the potential to impact Navassa is avail-
able.
Tourism and Recreation
There is no tourism or recreational use at
Navassa. A Special Use permit from the U.S. Fish and Wildlife Service is required for entry.
Fishing
Despite its status as a National Wildlife Refuge, sheries at Navassa are effectively unmanaged as regulations are not
well publicized and enforcement is not feasible in this remote location. Fishing activities by migrant Haitian artisanal
shermen have been ongoing since at least the 1970s. Miller et al. (2007) perceived an escalation of shing effort based
on observation of the use of novel and more destructive gear types including net shing (rst observed in 2002), which
allowed exploitation and bycatch of previously unexploited species such as queen conch (Strombus gigas) and Hawksbill
sea turtles (Eretmochelys imbricata; Wiener, 2005).
Figure 4.2. Map of Navassa Island showing the path and intensity of major storm
events between 2002-2007. Map: K. Buja. Source: http://maps.csc.noaa.gov/hur-
ricanes/.
IRIS (2001) H1
IRIS (2001) H1
DEAN (200 7) H4
DEAN (200 7) H4
IVAN (2004) H4
IVAN (2004) H4
DENNIS ( 2005) H1
DENNIS ( 2005) H1
DENNIS ( 2005) H2
DENNIS ( 2005) H2
DENNIS ( 2005) H3
DENNIS ( 2005) H3
CHARLEY (2004) H1
CHARLEY (2004) H1
DENNIS ( 2005) H4
DENNIS ( 2005) H4
ERNESTO (2006) H1
ERNESTO (2006) H1
75°W
75°W
18°N
18°N
Category
1
2
3
4
5
0 25 50 km
Hurricanes
2000-2007
Navassa
Cuba
Jamaica
Haiti
Windward
Passage
Anse d'Hainault
!
H

The State of Coral Reef Ecosystems of Navassa Island
119
Navassa Island
The National Oceanic and Atmospheric Administration’s (NOAA) Southeast Fishery Science Center (SEFSC) has re-
cently conducted a sociocultural characterization of Haitian shing communities that exploit the waters surrounding
Navassa (Wiener, 2005; Miller et al., 2007). This study included extensive interviews of shers both on site in Navassa
and in southwest Haiti, as well as limited quantication of landings from three individual boat-trips. Results of the sher
interviews conducted between November 2004 and June 2005 also indicated that capital for boats, traps and fuel was the
primary limitation on current shing effort. Similarly, the harsh living conditions on Navassa were the only factor prevent-
ing permanent settlement of the islands as socioeconomic conditions in Haiti continue to be dismal.
Unexpectedly, the most recent observations in April and November of 2006 revealed a reduction in shing activity when
compared with 2004. A total of 175 xed gear buoys (marking an unknown ratio of traps and nets combined) were
mapped in 2004 (Miller et al., 2007), whereas many fewer traps were being actively shed in April and November 2006
(Table 4.1). Other measures of shing effort appear to have peaked in 2004 and abated in 2006. Particularly notable was
the lack of net shing in 2006. All of the shers present in 2006 were from a single Haitian village, Anse d’Hainault, and
those interviewed indicated that this village had not previously participated in net shing. It is not clear if this apparent
relaxation of shing effort has resulted from a form of self-management, poorer yields (interviewees indicated that the
shing was very poor in 2006) or other external factors such as high fuel prices.
Additionally, Haitian commercial shing operations and international trawlers purportedly from the Dominican Republic
and Jamaica are suspected of targeting pelagic sh species within the Navassa National Wildlife Refuge’s (NNWR) 12
nautical mile territorial sea.
Trade in Coral and Live Reef Species
This threat does not have a major impact on Navassa’s reefs.
Ships, Boats and Groundings
This threat does not have a major impact on Navassa’s reefs.
Marine Debris
Marine debris from recent shing activities and historical uses was described by Miller et al. (2005).
Aquatic Invasive Species
Invasive species have not been observed at Navassa to date.
Security Training Activities
No military activities occur at Navassa.
Offshore Oil and Gas Exploration
No oil and gas exploration activities occur at Navassa.
Table 4.1. Trends in apparent shing effort by transient Haitians on the Navassa shelf. Source: Miller et al., 2004; Wiener, 2005; Piniak
et al., 2006.
MEASURE OF FISHING EFFORT GEAR IN USE
Date of
observation
Duration of
observation (d)
Total # gear
buoys/traps
Mean
boats/day
Mean
shers/day
Traps Hook and
Line
Nets
November 2002 11 NA 2 9.7* X X X
November 2004 13 175 4.4 22 X X X
April 2006 10 7 0.7 2.8 X X -
November 2006 11 34 4 15.9 X X -
* Observations in 2002 were less complete; data are extrapolations based on reported average of ve shers per boat.

The State of Coral Reef Ecosystems of Navassa Island
120
Navassa Island
CORAL REEF ECOSYSTEMS—DATA-GATHERING ACTIVITIES AND RESOURCE CONDITION
Monitoring of the coral reefs of Navassa is now conducted biennially by NOAA-Fisheries and partners, with support from
the NOAA Coral Reef Conservation Program. Cruises took place in November 2002, 2004 and 2006 to conduct under-
water visual censuses of sh, habitat mapping (including single-beam acoustics), and benthic community assessments.
Complementary data sets including multibeam bathymetry, temperature records, additional habitat assessments and
sampling for trophic analysis via stable isotopes (Piniak et al., 2006) were obtained during an additional cruise in April
2006 conducted by NOAA’s National Centers for Coastal Ocean Science (NCCOS), Center for Coastal Fisheries and
Habitat Research (CCFHR). Monitoring locations for both groups are shown in Figure 4.3.
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(!(
!(
!(
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*#*#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*#*
#*
#*
#*#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*#*
#*
#*
#*
#*#*
#*
#*
#*
#*#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
#*
")
")
")
")")
")
")
")
")
75°W
75°W
18°24'N
18°24'N
")2004 Disease Assessment
")Acropora Monitoring Site
")Temperature Logger
")Photo Quadrat Site
#*Rapid Assessment Dive Site
#*Additional Reeffish Visual
Census Sample Site
!(NCCOS-CCFHR
Sampling Site
Land
Water <30 m
Deep Water
0 0.5 1 km
Figure 4.3. Monitoring locations sampled by NOAA/SEFSC and NOAA/NCCOS. Map: K. Buja.

The State of Coral Reef Ecosystems of Navassa Island
121
Navassa Island
BENTHIC HABITATS
Benthic Characterization (SEFSC 2002–2006)
Methods
Given the relatively deep depths and limited sampling effort available for reef assessment at Navassa (about 10 days
every two years), a hybrid sampling approach has been adopted. Standard in situ line intercept transects (15 m transect,
sampled at 15 cm intervals, n=2-4) were used to estimate the percent cover of primary community components (sclerac-
tinian corals, macroalgae, octocorals, sponges) at four, relatively shallow (7-22 m) xed sites every two years. Addition-
ally, haphazard photoquadrats were collected from a distinct set of Rapid Assessment Dive (RAD) sites (22-32 m depth)
distributed throughout the shelf. Photoquadrats (n=4-10) were analyzed using a standard point count method applied by
Coral Point Count (CPCe) software. To enhance comparability between years, data are presented only for reef RAD sites
along the southwest portion of the shelf.
Results and Discussion
Mean percent cover of xed sites and deep-
er RAD sites are shown in Figures 4.5 and
4.6. Macroalgae (predominantly Lobophora
variegata) comprised the dominant benthic
group overall, with values around 40% cov-
er common and values over 70% observed
on occasion (Figures 4.5 and 4.6; mean ±1
SD for 2006 xed sites, 54.4 ±17.3% SD).
Declines in coral cover have occurred, but
only at deeper sites (i.e., Video Patch and
RAD sites) where coral cover had initially
been very high and both 2004 disease and
2006 bleaching were observed at greater
prevalence. Mean coral cover ±1 SD for
southwest shelf RAD sites in 2002 was 39.9
±8.0% SD but had dropped to 11.1 ±6.4%
SD by 2006 (Figure 4.6). Meanwhile, live
coral cover at the shallow xed sites at NW
Point and Lulu Bay have remained fairly
steady in the range of 10-15% and 20-25%,
respectively (Figure 4.5). Coral cover losses
observed through 2006 likely resulted from
disease and hurricane impacts (particularly
during 2004). Ongoing coral mortality is anticipated given the severe bleaching status of corals observed in 2006.
27
28
29
30
31
Apr May Jun Jul Aug Sep Oct Nov
Month
Water Temperature (ºC)
Figure 4.4. Monthly mean (± 1 SD) of hourly temperature readings from ve sites
around Navassa in 2006. Sites are: Conch North (28 m depth), Lulu Bay 7 (15 m
depth), Lulu Bay 6 (26 m depth), West Pinnacles (26 m depth), and Northwest Point
(11 m depth). This period of time preceded the observation of a severe coral bleach-
ing event at Navassa in November 2006. Source: Piniak et al., unpub. data.
Figure 4.5. Percent cover of xed sites sampled via in situ point-intercept transects
over time. NW Point, approximately 10 m; Lulu Bay: 7-10 m; West Pinnacles, 22 m;
Video Patch: approximately 30 m, not sampled in 2000 or 2004. Source: Miller and
Gerstner, 2002; Miller et al., 2003; Miller et al., unpub. data.
0
10
20
30
40
50
60
70
80
90
100
2000
2002
2004
2006
2000
2002
2004
2006
2000
2002
2004
2006
2000
2002
2004
2006
Cover (%)
Sand/rubble
Other Inverts
Turf/bare
CCA
Sponge
Macroalgae
Hard Coral
NWpt Lulu Bay WPinn Video Patch
Since April 2006, temperature has been
regularly measured at Navassa using an ar-
ray of temperature loggers deployed at ve
sites at depths between 11 and 28 m. Hourly
data was retrieved from these sensors in
November 2006 and is summarized in Fig-
ure 4.4. Currently, temperature is the only
water quality parameter being measured at
Navassa.
WATER QUALITY AND OCEANOGRAPHIC CONDITIONS
Conch North
Lulu 7
Lulu 6
W. Pinnacles
NW Point

The State of Coral Reef Ecosystems of Navassa Island
122
Navassa Island
Benthic Characterization (NCCOS 2006)
Methods
The habitat characterizations on the April
2006 cruise focused on the deeper portions
of the inner shelf (30-34 m). Sites were ran-
domly selected from the appropriate depth
range; each site consisted of three replicate
transects deployed in random directions.
A site therefore incorporated a mixture of
habitat types (both reef and non-reef). Due
to differences in techniques, these data are
not strictly comparable to the SEFSC data.
Three 30 m visual sh transects were con-
ducted at each site (data not reported) and
benthic photoquadrats were collected at
each meter along the transect and analyzed
using standard point count methods within
CPCe software.
Results and Discussion
Mean percent cover for the NCCOS sites
are given in Figure 4.7. Macroalgae were
the dominant benthic biota, comprising
36% of the total benthic cover around Na-
vassa. Lobophora variegata was by far
the most abundant macroalga (maximum
34%). Halimeda sp. and Dictyota sp. were
secondary components of the algal com-
munity. Coral cover ranged from 1-7%;
this underestimates typical cover measure-
ments because mixed habitat types (includ-
ing non-reef areas) were surveyed at each
site. The primary components of the coral
communities were the species that make up
the Montastraea annularis species complex
(referred to as Montastraea spp.), Sideras-
trea siderea, Porites astreoides and P. po-
rites. Coral cover was lowest on the eastern
coast, which had high proportions of rock
(36%) and rubble (15%); uncolonized sub-
strate on the north and south coasts was
primarily sand.
0
10
20
30
40
50
60
70
80
90
100
2002 2004 2006
Cover (%)
Substrate
Other Inverts
CCA
Sponge
Macroalgae
Hard Coral
n=3 n=10 n=13
Figure 4.6. Percent cover of southwest coast reef RAD sites (haphazardly selected
each year) as determined from point counts of haphazardly-placed 1 m2 photoquad-
rats (4-10 photoquadrats per site). Algal turfs are poorly resolved from photographs
so they are included with pavement, rubble, and sand called “substrate”. N given
in each bar represents the number of sites (southwest patch reefs only) sampled in
that year. Source: Miller et al., 2003 and Miller et al., unpub. data.
(%)
0
40
50
60
70
90
100
80
10
20
30
Figure 4.7. Community composition characteristics for all sites surveyed at Na-
vassa by NCCOS in April 2006. All sites were surveyed in situ using benthic photo
transects (n=31 photos per 30 m transect, n=3 transects per site). Sites (numbers
along the x-axis) were all 30-34 m depth and stratied by location (southwest, north
or east coasts). Benthic cover types are grouped by NOAA Fisheries categories,
but are not strictly comparable due to differences in methodology. Source: Piniak
et al., in prep.

The State of Coral Reef Ecosystems of Navassa Island
123
Navassa Island
Disease Characterization (SEFSC–2004)
Methods
Haphazardly placed transects (n=3-7)
were sampled at ve sites around the is-
land to examine spatial variation in disease
prevalence upon observation of high coral
disease occurrence in November 2004.
Transect size was either 1 x 7.5 m or 0.5 x
10 m. Each colony within the transect was
scored for species, size category (small <15
cm diameter; medium 15-40 cm; large >40
cm) and disease state was scored as either
“active” disease signs, “recent mortality” or
unaffected. Prevalence of both active and
recently diseased states were expressed as
proportion of total colonies in each disease
state. Prevalence was also calculated for
certain subsets of colonies, namely large
colonies (>40 cm), and Montastraea spp.
for comparison to the coral community as
a whole. Further detail on the methodology
employed can be found in Miller and Wil-
liams (2007).
Results and Discussion
Over 15 species of scleractinians were ob-
served with “white disease” signs (Miller
and Williams, 2007; Figure 4.8). Total preva-
lence (percent) of colonies with active dis-
ease signs at the sites sampled via haphaz-
ard transects ranged from zero at NW Point to over 15% at site A, with an additional 20% of colonies at that site displaying
recent mortality (Table 4.2). Disease prevalence was substantially higher among large colonies and among Montastraea
spp. colonies, with a majority of Montastraea spp. colonies affected by disease at one site (Table 4.2). The ensuing loss
of large colonies is expected to affect coral community structure over a long time span.
Figure 4.8. Photo of pillar coral, Dendrogyra cylindrical, suffering rapid tissue loss
consistent with white-plague type disease in November 2004. White areas of the
colony are recently dead (skeleton); only the gray areas still have live tissue. Photo:
NOAA SEFSC.
Table 4.2. Prevalence of active disease signs and recent mortality consistent with disease. Replicate transects were pooled from each
site to indicate prevalence amongst all colonies compared to large colonies and Montastraea spp. colonies. No colony size information
was collected at Site A. Locations are given in Figure 1. Source: Miller and Williams, 2007.
SITE
TOTAL COLONIES COLONIES >40 cm DIAMETER MONTASTRAEA spp. COLONIES
N% active
disease
% recent
mortality N% active
disease
% recent
mortality N% active
disease
% recent
mortality
A79 15 19 NA NA NA 19 36.8 21.1
B 360 6.9 7 22 31.8 0 44 25 2.3
Video
Patch 267 3.4 2.6 64 14 10.9 28 21.4 7.1
NW Pt 137 0 0 6 0 0 5 0 0
C 300 1.5 0 10 10 0 20 0 0

The State of Coral Reef Ecosystems of Navassa Island
124
Navassa Island
Bleaching Characterization (SEFSC–2006)
Methods
A widespread and fairly severe coral bleaching event was encountered during the November 2006 cruise. Several rapid
assessment techniques were utilized to document the extent (spatial patterns, species affected and severity) of coral
bleaching at Navassa. Using a belt transect (10 x 1 m) at seven sites ranging in depth from 7-27 m, all colonies greater
than 4 cm diameter were identied to species or genus and colonies were ranked by size class and degree of bleaching
(normal, pale, mottled or completely bleached white). These categories were subsequently pooled for the current presen-
tation. Between two and six transects were sampled at each of the seven sites.
In order to get a more representative view, scientists also performed RADs at an additional eight sites ranging in depth
from 27-37 m. In these cases, no transect was laid out and a subset of common hermatypic coral species (limited to Diplo-
ria strigosa, D. labyrinthiformes, Montastraea cavernosa, M. faveolata, M. annularis, M. franksi and Colpophillia natans)
were scored for bleaching state as described above. A haphazard compass heading and a 1 m length PVC pole was used
to delineate an area for sampling. Although the total area sampled (hence colony density) cannot be determined from this
data set, the use of the heading and a 1 m guide minimized bias in ‘choosing’ colonies to record. Bleaching prevalence
(percent of colonies affected) for the sampled species was recorded.
Results and Discussion
Overall prevalence of coral bleaching at the
various sites ranged from approximately
15-78% of colonies when all species were
pooled (Figure 4.9). Shallow sites (<10 m)
were less affected than deeper sites (>20
m) and fringes of bleached colonies that
were overgrown by macroalgae (commonly
Lobophora variegata) were normally pig-
mented. This suggests that the interaction
between the severity of bleaching and dif-
ferences in light levels may be complex.
The most impacted coral taxa were Agari-
cia spp. and Montastraea spp.(M. faveolata
was greatly dominant in this group). Sid-
erastraea siderea, Diploria spp. (predomi-
nantly D. strigosa) and Porites porites were
intermediately affected. Least impacted
were P. astreoides and M. cavernosa.
Qualitatively, the intensity of bleaching ap-
peared to increase over the 11 days of ob-
servation. Some bleached colonies were
clearly undergoing partial mortality, but it
was not possible to differentiate causality
related to bleaching versus disease as both
often co-occurred in colonies (Figure 4.10).
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Lulu
Shelf*
NW Pt* North
Shelf*
Lulu
Base
South
Pt*
West
Pinn*
VidPatch RAD
Sites
Sample Site
Proportion of Colonies Bleached
Figure 4.9. Mean plus ±1 SD bleaching prevalence. Sites with asterisks show no
variance estimate since only two transects were sampled. The error bars shown for
the other sites indicated ±1 SD for n=5 or 6 transects (or six sites for RAD sites).
These means pool all degrees of coral bleaching and all species sampled (see text
for details). The rst three sites are shallow shelf habitats (7-10 m depth). The sec-
ond set of four sites range from 20-27 m depth. The last bar (RAD sites) is the mean
of prevalence scored for a subset of coral species in one rapid assessment dive at
each of six sites ranging from 27-37 m. Source: Miller and Williams, unpub. data.
Figure 4.10. Photos showing bleached coral with partial mortality. Left: Bleached
Siderastraea siderea that appears to have endured multiple recent (estimated 6-18
months) partial mortality events. Note small unbleached conspecic to the left.
Right: Moderately bleached Montastraea spp. with current mortality possibly from
simultaneous disease. Photos: NOAA SEFSC.

The State of Coral Reef Ecosystems of Navassa Island
125
Navassa Island
Status of Acroporids (SEFSC–2002-2006)
Methods
A. palmata habitat is mostly conned to the shallow shelf areas around Lulu Bay and Northwest Point and around the cliff
along much of the southwest and north coasts. Only qualitative observations on Acropora spp. abundance were made in
2002. In 2004, minimal surveys were made along conned sections of the coast. However, in 2006, following the Endan-
gered Species Act listing of these species, targeted abundance sampling and demographic monitoring were established.
We quantied the spatial extent and location of A. palmata colonies along the entire north and southwest coasts using
snorkeler observations and a handheld Global Positioning System to mark the position of each colony encountered. Ap-
proximately 6.8 km of the estimated 9 km of coastline was surveyed for A. palmata. In addition, a total of 77 A. palmata
colonies in ve permanently marked plots (three around Northwest Point and two near Lulu Bay) were tagged, assessed,
photographed, and biopsied for genotyping according to protocols developed and applied in the Florida Keys (Williams et
al., 2006). Future surveys will reveal the recruitment and survivorship of the population at Navassa.
Results and Discussion
In stark contrast to other coral species in the area, the majority of A. palmata colonies observed appeared healthy with
recent mortality observed only occasionally. A total of 1,800 colonies were mapped over 6.8 km of the Navassa coast.
Although rough seas prevented surveys along the east coast of the island, heavy swells along this windward coast seem
to limit coral development, and few Acropora spp. colonies were expected to occur in this area. In contrast, the 1998 expe-
dition to the island (Littler et al., 1999) reported approximately one dozen A. palmata colonies conned to Lulu Bay based
on casual observation. While targeted surveys of A. palmata were not conducted in 1998, it appears that the population
has increased, with our 2006 survey counting more than 100 colonies in Lulu bay, and observations of portions of the wall
that were paved with encrusting A. palmata.
In contrast to A. palmata, A. cervicornis remains extremely rare at Navassa. A total of only ve small colonies were ob-
served in over 250 person dives during the November 2006 cruise. One of these colonies clearly displayed tissue slough-
ing, a sign of disease, as has been observed in the Florida Keys (Williams and Miller, 2005).
Mapping
Directed efforts at mapping Navassa’s
benthic habitats began in 2004 with single
beam acoustic work and benthic commu-
nity characterization by scientists from the
University of Miami’s Rosenstiel School of
Marine and Atmospheric Science (RSMAS).
Multibeam mapping of the Navassa shelf
was conducted in April 2006 by NCCOS-
CCFHR in partnership with Solmar Hydro
from approximately the 20 m contour out to
12 nm (about 22 km) from the island. A digi-
tal elevation model based on Light Detec-
tion and Ranging (LiDAR) data acquired in
1999 and multibeam bathymetry was used
to calculate slope. Figure 4.11 shows the
result when these two output layers were
combined.
Results of a slope calculation performed
using multibeam data from 20–50 m depths
with the Matlab Mapping Toolbox (Version
1.2; Mathworks, Natick, MA), which uses
nite differences to compute the gradient
of a gridded data set are shown in Figure
4.12A. Single beam acoustic data were ac-
quired as points along track lines, classied,
then gridded to 100 m cells using a majority
lter. Figure 4.12B shows the acoustic seabed classication based on the percent of the seabed covered with sediment
(patchiness) and the local variability in depth (relief). Information from all sources (IKONOS satellite imagery, multibeam
bathymetry, benthic community analysis, drop camera and diver observation) were integrated into the habitat map (Figure
4.12C). Details of map construction are given in Miller et al. (in review).
Figure 4.11. Bathymetric map of Navassa Island and the surrounding coastal area.
Source data: Solmar Hydro and NASA.

The State of Coral Reef Ecosystems of Navassa Island
126
Navassa Island
Figure 4.12. A) Slope for Navassa Island computed from NASA’s LiDAR data on land and Simrad XX data from 20-50 m water depth.
Values above 30 degrees were clipped to show detail in the range 0-30 degrees. Maximum slope for this data set was 80 degrees.
Null values due to lack of data are shown in white encircling the island. B) Acoustic seabed classication based on “patchiness”, the
percent of the seabed covered with sediment, and “relief”, the local variability in depth. Data were acquired as points along track lines,
classied, then gridded to 100 m cells using a majority lter. No data is shown as gray. C) Interpreted benthic habitat map based on all
available information sources including bathymetry, IKONOS imagery, benthic community classication, and diver/drop camera obser-
vations. Sources: A) Solmar Hydro, NASA; B and C) A.Gleason, Univ. of Miami/RSMAS.

The State of Coral Reef Ecosystems of Navassa Island
127
Navassa Island
ASSOCIATED BIOLOGICAL COMMUNITIES
Fish Surveys (SEFSC)
Data on reef sh assemblages and other mobile fauna have been collected via a stationary point sampling technique
(Bohnsack and Bannerot, 1986; McClellan and Miller, 2003) referred to as Reefsh Visual Census (RVC). The total num-
ber of samples and summary results are given in Table 4.3. Sites sampled in 2006 included both stratied random sites
(according to habitat map in Figure 4.12C) and targeted RAD sites. In addition to enumerating reef shes, RVC samples
record the presence and abundance of selected mobile macroinvertebrates, including the long-spined sea urchin (Dia-
dema antillarum), queen conch (Strombas gigas) and lobster (Panulirus argus, not reported here). The abundance of
D. antillarum, an important grazer, was also noted in benthic transect sampling at xed sites surveyed by the SEFSC in
2006.
Results and Discussion
There is a clear declining trend in reef sh
biomass (Figure 4.13 and Table 4.3) be-
tween 2002 and 2006 as determined by the
RVC sampling. This trend is most evident
in piscivores, herbivores and planktivores
(the dominant trophic groups in terms of
biomass). Macroinvertivores were the only
group which showed a substantial increase
in 2006 but this increase was due to squir-
relsh only (data not shown), a common
family which are preyed upon by piscivores
as well as Navassa’s human shers. Fish
sizes (mean fork length of individuals >10
cm) also showed a signicant decline be-
tween 2002 and 2004 for grouper, snapper,
triggersh, parrotsh, jack, surgeonsh and
squirrelsh families (Miller et al., 2007).
It should be noted that a more restricted set
of habitats was sampled in 2002, particu-
larly high-relief habitats near shore such as
wall and wall base/boulder habitats. How-
ever, the same declining temporal trends
are evident if relatively depauperate, non-
reef habitats (e.g., sand/rubble) are exclud-
ed from the latter years’ samples (data not
shown). Hence, it is not likely that the ob-
served decline in sh biomass (Table 4.3)
can be explained by differential habitat rep-
resentation.
On the other hand, abundance of D. antil-
larum has increased over the four year in-
terval. The mean density of urchins from the
RVC data (number/sample) increased 400%
between 2002 and 2006 (Table 4.3). Ben-
thic transects indicated a November 2006
D. antillarum density of 0.16 m2 + 0.02 %
SE (n=11 10 m2 transects among six sites).
Although these densities are nowhere near
those that have been shown to correspond
with enhanced coral recruitment (i.e., 2-5
per m2; Carpenter and Edmunds, 2006)],
densities are likely approaching this level
in certain habitats (e.g., nearshore boulder/
calves habitat on night dive; M. Miller, pers. obs.). The marked increase suggests that recovery of Diadema populations
is underway at Navassa. Conch are known to be highly aggregative and we observed no clear temporal trend in their
abundance (Table 4.3). RVC samples in 2004 encountered several conch aggregations and this yielded higher mean and
frequency of occurrence estimates in 2004 (Table 4.3).
2002 2004 2006
# RVC Samples (N) 110 123 150
# census takers 2 3 4
# Fish Species 122 128 139
# Individuals 22,798 41,174 35,633
Density (# indiv/sample) 207 335 238
Total Biomass (g) 1,547,671 1,052,314 1,128,868
Mean Biomass (g/sample) 14,070 8,555 7,526
# Diadema 18 53 99
Mean density (#/sample) 0.16 0.43 0.66
Frequency (proportion of
samples occuring) 0.09 0.20 0.22
# Conch 8 247 65
Mean density (#/sample) 0.07 2.01 0.43
Figure 4.13. Reefsh biomass per sample (mean + 1 SE) by trophic group over a
four year interval. Species included in each trophic group provided in McClellan and
Miller, 2003. Source: Miller et al., 2007; McClellan et al., unpub. data.
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
9,000
10,000
Benthivores
Piscivores
Herbivores
Macroinvertivores
Microinvertivores
Planktivores
Trophic Guild
Mean Fish Mass (g/sample)
2002 n=110
2004 n=123
2006 n=150
Table 4.3. Summary of RVC data, including relative abundance of conch and urchins
(D. antillarum) collected in 2002, 2004 and 2006. Sampling intensity has increased
slightly over the study, including a wider range of habitats. Source: McClellan and
Miller, 2003; McClellan et al., unpub. data.

The State of Coral Reef Ecosystems of Navassa Island
128
Navassa Island
CURRENT CONSERVATION MANAGEMENT ACTIVITIES
Much valuable information has been gathered about the ora, fauna and threats to the ecology of Navassa. Since the
NNWR was established in 1999, it has been faced with a documented increase in threats by foreign nationals, mainly
Haitians, conducting commercial and subsistence shing and hunting activities on the Refuge. Challenges to effective
management are related to the island’s remote location, an absence of local management presence, and an absence
of solid quantitative shery data. Currently, no practicable mechanism exists whereby the NNWR can efciently or eco-
nomically document, manage, or address these threats. Although active management has been limited, work begun by a
Haitian non-governmental organization, the Foundation for the Protection of Marine Biodiversity, is beginning to educate
local shers.
Discussions are now underway for developing a strategy to deal with the unauthorized shing incursions into NNWR via
a collaborative conservation effort with federal agency members of the U.S. Coral Reef Task Force, academic institutions
and non-governmental conservation organizations. The development of a Navassa NWR collaborative conservation ef-
fort will strengthen the National Wildlife Refuge System’s natural resource management efforts. It is foreseeable that a
similar approach can be employed for other remote, insular U.S. possessions, especially National Wildlife Refuges in the
Pacic Ocean.
OVERALL CONCLUSIONS AND RECOMMENDATIONS
It is clear that Navassa reefs, despite their remoteness from many types of local anthropogenic stress, are undergoing
rapid change. Both expanded (but possibly stabilized) shing pressure and disturbances, such as coral bleaching and
disease events, are resulting in rapid loss of live coral cover, including loss of large coral colonies, and reductions in the
size and abundance of reef shes. The jurisdictional/management challenges for Navassa, meanwhile, do not abate. The
occurrence of severe coral disease and bleaching events in this relatively deep (25-30 m) and remote location support
the hypothesis that coral loss in the Caribbean is a regional phenomenon, and effective conservation and management
measures to reverse this trend are not obvious.

The State of Coral Reef Ecosystems of Navassa Island
129
Navassa Island
REFERENCES
Bohnsack, J.A. and S.P. Bannerot. 1986. A stationary visual census technique for quantitatively assessing community structure of coral
reef shes. NOAA NMFS Technical Report 41. Seattle, WA. 15 pp.
Carpenter, R.C. and P.J. Edmunds. 2006. Local and regional scale recovery of Diadema promotes recruitment of scleractinian corals.
Ecol. Lett. 9: 271-280.
Littler, M.M., D.S. Littler, and B.L. Brooks. 1999. The rst oceanographic expedition to Navassa Island, USA: Status of marine plant and
animal communities. Reef Encounter 25: 26-30.
McClellan, D.B. and G.M. Miller. 2003. Reef sh abundance, biomass, species composition and habitat characterization of Navassa
Island. pp. 24-42. In: Miller MW (ed.). Status of Reef Resources of Navassa Island. NOAA Technical Memorandum NMFS SEFSC 501.
Miami, FL. 119 pp. http://www.aoml.noaa.gov/general/lib/tm_501.pdf
Miller, M.W. (ed.). 2003. Status of Reef Resources of Navassa Island. NOAA Technical Memorandum NMFS SEFSC 501. Miami, FL.
119 pp. http://www.aoml.noaa.gov/general/lib/tm_501.pdf
Miller, M.W. and C.L. Gerstner. 2002. Reefs of an uninhabited Caribbean island: shes, benthic habitat, and opportunities to discern
reef shery impact. Biol. Conserv. 106: 37-44.
Miller, M.W., J. Schwagerl, D.B. McClellan, M.J.A. Vermeij, and D.E. Williams. 2005. The state of coral reef ecosystems of Navassa. pp.
135-149. In: J.E. Waddell (ed.). The state of Coral Reef Ecosystems of the United States and Pacic Freely Associated States: 2005.
NOAA Technical Memorandum NOS NCCOS 11. Silver Spring, MD. 522 pp.
Miller, M.W., D.B. McClellan, J.W. Wiener, and B. Stofe. 2007. Apparent rapid sheries escalation at a remote Caribbean island. En-
viron. Conserv. 34: 92-94.
Miller, M.W. and D.E. Williams. 2007. Coral disease outbreak Navassa, a remote Caribbean island. Coral Reefs 26(1): 97-101.
Miller, M.W., R.B. Halley, and A. Gleason. In press. Reef Geology and Biology of Navassa Island. pp. 407-343. In: B. Riegl and R.E.
Dodge (eds.). Coral Reefs of the USA. Coral Reefs of the World, Volume 1. Springer. 806 pp.
Piniak, G.A., C.M. Addison, B.P. Degan, A.V. Uhrin, and T.S. Viehman. 2006. Characterization of Navassa National Wildlife Refuge: A
preliminary report for NF-06-05 (NOAA ship Nancy Foster, April 18-30 2006). NOAA Technical Memorandum NOS NCCOS 38. Silver
Spring, MD. 48 pp.
Wiener, J.W. 2005. Oral history and contemporary assessment of Navassa Island shermen. FoProBiM Report for NOAA National
Marine Fisheries Service. 54 pp. http://www.sefsc.noaa.gov/PDFdocs/Navassa_Fishers_Report_Final_FoProBIM.pdf.
Williams, D.E. and M.W. Miller. 2005. Coral disease outbreak: pattern, prevalence, and transmission in Acropora cervicornis. Mar. Ecol.
Prog. Ser. 301:119-128.
Williams, D.E., M.W. Miller, and K.L. Kramer. 2006. Demographic monitoring protocols for threatened Caribbean Acropora spp. corals.
NOAA Technical Memorandum NMFS SEFSC 543. Miami, FL. 91 pp.