Journal of Marine Systems 164 (2016) 85100

Contents lists available at ScienceDirect

Journal of Marine Systems

journal homepage: www.elsevier.com/locate/jmarsys

Occurrence of energetic extreme oceanic events in the Colombian

Caribbean coasts and some approaches to assess their impact
on ecosystems
G. Bernal a,, A.F. Osorio a, L. Urrego b, D. Pelez a, E. Molina b, S. Zea c, R.D. Montoya d, N. Villegas e
a
Grupo de investigacin OCEANICOS, Departamento de Geociencias y Medio Ambiente, Facultad de Minas, Universidad Nacional de Colombia, Cra 80 No 65-223, AA 1027 Medelln, Colombia
b
Grupo de investigacin OCEANICOS, Departamento de Ciencias Forestales, Facultad de Ciencias Agrarias, Universidad Nacional de Colombia, Calle 59 A N 63-20, Medelln, Colombia
c
Instituto de Estudios en Ciencias del Mar-CECIMAR, Universidad Nacional de Colombia, Sede Caribe, c/o INVEMAR, Calle 25 2-55, Rodadero SurPlaya Salguero, Santa Marta 470006, Colombia
d
Grupo de Investigacin en Ingeniera Civil-GICI y Grupo de Investigacin en Calidad de Agua y, Modelacin Hdrica-GICAMH, Programa de Ingeniera Civil, Universidad de Medelln,
Carrera 87 N 10 30-65, Medelln, Colombia
e
Grupo de investigacin en oceanologa, CENIT, Departamento de Geociencias, Facultad de Ciencias, Universidad Nacional de Colombia, Carrera 45 # 26-85, Edicio Manuel Ancizar, Bogot,
Colombia

a r t i c l e i n f o a b s t r a c t

Article history: Above-normal meteorological and oceanographic conditions that generate damage on coastal ecosystems and
Received 19 February 2016 associated human communities are called extreme oceanic events. Accurate data are needed to predict their oc-
Received in revised form 26 July 2016 currence and to understand their effects. We analyzed available data from four localities in the Colombian Carib-
Accepted 14 August 2016
bean to study the effect of wave-related extreme events (hurricanes, surges) in three coastal ecosystems, i.e.,
Available online 18 August 2016
mangroves, beaches, and reefs. Three localities were continental (Portete Bay mangroves at the Guajira Peninsu-
Keywords:
la, Bocagrande Public Beach at Cartagena City, Tayrona Natural Park reefs near Santa Marta City), and one was
Coastal ecosystems oceanic (Old Providence Island reefs in the San Andres and Old Providence Archipelago, SW Caribbean). We gath-
Colombian Caribbean ered data on ocean surface winds (19782011) for the four locations, then modeled signicant wave heights,
Wind extremes then identied extreme events, and nally tried to identify effects on the ecosystems, directly or from published
Wave extremes literature. Wave-related extreme surges were also compiled from Colombian press news (19702008). Modeled
wave maximums (N5 m signicant wave height) and press-reported events coincided with hurricanes, extreme
dry season, mid-summer drought and northern hemisphere winter cold fronts, with neither a relationship to
ENSO events, nor a temporal trend of increase, excepting Portete Bay, with a marked increase after 1995. Changes
in Portete Bay mangroves were analyzed from aerial photographs before and after Tropical Storm Cesar (1996). In
the 38 years before Cesar there was mangrove inland colonization, with some loss associated to beach erosion, while
during the 8 years following the storm there were localized retreats and important changes in vegetation composi-
tion related to the falling of large trees and subsequent recolonization by species that are faster colonizers, and
changes in soil composition brought about by inundation. Cartagena's Bocagrande Beach was followed between
2009 and 2011 by video, and two events of strong retreat were observed in 2010, one associated to the arriving
of cold fronts in March, and the other to the passing of Hurricane Tomas in NovemberDecember. Together, they
produced N 90 m beach retreat. We identied modeled wave maximums during Hurricane Lenny (1999) at Santa
Marta city, and hurricane Beta (2005) at Old Providence Island, both of which, according to the literature, had
transient minor effects on local coral reefs, which had been more affected by diseases and bleaching.
2016 Elsevier B.V. All rights reserved.

1. Introduction
An extreme oceanic event occurs when the meteorological and
oceanographic conditions exceed their normal levels (typically, and
Corresponding author.
E-mail addresses: gbernal@unal.edu.co (G. Bernal), afosorioar@unal.edu.co
(A.F. Osorio), leurrego@unal.edu.co (L. Urrego), dspelaez@gmail.com (D. Pelez),
sezeas@unal.edu.co (S. Zea), rmontoya@udem.edu.co (R.D. Montoya),
nlvillegasb@unal.edu.co (N. Villegas).
depending on the variable, the 90th, 95th or 99th percentile, from a
probability density function estimated from existing observations); it
is not predictable and has an impact on marine ecosystems and on the
coast. Among the most recognized events are sea temperature rises
above the seasonal maximums, sediment outpourings, storms, tropical
cyclones (called hurricanes in the Atlantic), and storm surges. Long-
term climatic conditions (from El Nio, Southern Oscillation, ENSO, to
global climate change) are linked to the frequency and intensity of ex-
treme oceanic events, but in some regions this relationship is not fully
understood (IPCC, 2012).

http://dx.doi.org/10.1016/j.jmarsys.2016.08.007
0924-7963/ 2016 Elsevier B.V. All rights reserved.
86 G. Bernal et al. / Journal of Marine Systems 164 (2016) 85100
The nature and severity of the impacts of extreme events on the eco-
systems depend not only on the events themselves, but also on the ex-
posure and vulnerability of the ecosystems. In turn, the events affect
future ecosystem vulnerability by changing their resilience and adapta-
tion capacity (IPCC, 2012). Coastal ecosystems such as reefs, sea grass
beds, mangroves, and beaches, are directly exposed to the impacts of
extreme oceanic events. Although they may have a degree of adaptation
to them, with current human pressure and global climate change they
may be seriously threatened (Hughes, 1994; Wilkinson, 2008). Evi-
dence for the change in the extremes in global and local climate is avail-
able only after about 1950. Thus, condence in observed trends depends
on the amount and quality of data, and on the availability of studies an-
alyzing them, which vary greatly for different regions and types of
events.
To face the social and ecological challenge that involves new trends
in extremes over a region, it is necessary to understand the association
between physical processes (threats) and ecosystem responses (vulner-
ability). Identifying and understanding the extremes of physical pro-
cesses are a complex task, requiring knowledge on the average
conditions and observations lasting long enough to record events with
large recurrence times. Similarly, ecosystems are complex and in order
to understand their responses and vulnerability, it is also required to
know their average behavior. Furthermore, knowledge about the char-
acteristics and impact on marine ecosystems of one forcing is not
enough; it is necessary to address synergistic effects (The MerMex
Group, 2011).
In an attempt to improve the understanding of the energetic ex-
treme oceanic events and their impact on the ecosystems, this paper il-
lustrates some cases of ecosystems affected by storm surges and
hurricanes in the Colombian Caribbean at different timescales. It begins
with a brief overview of extreme oceanic events and their impacts on
mangroves, beaches, and reefs. Then, an analysis of the occurrence of
physical extremes over the Colombian Basin is presented. Finally, their
possible inuence on the ecosystems is examined in four locations
where some evidence is available. Three continental and one oceanic lo-
cation were studied; in each, a particular ecosystem where data are
available was examined. Continental: mangroves in Portete Bay in La
Guajira Peninsula, beaches in Cartagena City, and reefs at Tayrona Na-
tional Natural Park near de city of Santa Marta. For the oceanic location,
the reefs of Old Providence in the Colombian San Andres Archipelago in
Fig. 1. Location of the cases presented. A: San Andres and Old Providence Archipelago reefs in the SW Caribbean (Old Providence Island, PR). B: Cartagena City beaches (Bocagrande, BG). C:
Santa Marta area reefs (Tayrona National Natural Park, TY). D. Guajira Peninsula mangroves (Portete Bay, PO).
the SW Caribbean were chosen (Fig. 1). Some ecological cases (man-
groves, beaches) contain new evidences and data, but others (reefs)
were taken from published information. Each case begins with a short
area description, followed by an explanation of the methodology, and
ends with results and discussion. Final remarks and recommendations
are made in the conclusions section.
2. Overview of extreme oceanic events and impact on
coastal ecosystems
For the near future, a considerable increase in the frequency and
magnitude of weather extremes has been predicted over a broad
range of ecosystems. In particular, the maximum wind speed, the inten-
sity and the duration of hurricanes have increased during the last few
decades (Jentsch and Beierkuhnlein, 2008; Webster et al., 2005;
Emanuel, 2005).
In addition, Webster et al. (2005) found no global trend in the total
number of tropical storms and hurricanes between 1970 and 2004.
Only the North Atlantic showed a statistically signicant increase in
hurricanes, which started in 1995. However, they found that the stron-
gest hurricanes (categories 4 and 5) have almost doubled in number
and proportion in all of oceanic basins in the last decade compared to
the 1970s. For the Caribbean Sea, this behavior has been reported by
Montoya (2013), who showed an important trend in the number of Cat-
egory 4 and 5 hurricanes (+4.2 hurricanes/century) for the last decades
(19792011). Nevertheless, this increase was not accompanied by an
increase in the intensity of the most intense hurricanes: The maximum
intensity has remained static over the past 35 years.
However, an index of the potential destructiveness of tropical cy-
clones based on the total dissipation of power integrated over its life-
time (Emanuel, 2005) has increased markedly since the mid-1970s.
The record of net cyclone power dissipation is highly correlated with
tropical sea surface temperature (SST), following multi-decadal oscilla-
tions in the North Atlantic and North Pacic, and reecting global
warming. But only part of the observed increase in tropical cyclone
power dissipation is directly due to increased SSTs; the rest can be ex-
plained by changes in other factors such as vertical wind shear or tem-
perature distribution of the upper ocean. There is some indication that
sub-surface temperatures have also been increasing, thereby reducing
the negative feedback from storm-induced mixing (Emanuel, 2005).
 G. Bernal et al. / Journal of Marine Systems 164 (2016) 85100
The probability that hurricanes directly affect the Colombian
Caribbean coasts is low, except for the oceanic San Andres and Old
Providence Archipelago (Ortiz-Royero, 2007). The peak of the hurri-
cane season for the Atlantic Basin (including the Caribbean) is from
mid-August to late October. Sixty of these storms passed through
the Colombian Caribbean between 1900 and 2010. Some of the
hurricanes that have inuenced the area and whose effects on the en-
vironment have been recorded qualitatively or quantitatively are: Joan
(October 1988), Mitch (November 1998), and Lenny (November
1999). The extreme waves or episodic coastal inundation events
(storm surges, known locally as mares de leva or elevated seas)
are mainly originated either by the sporadic passage of hurricanes at
some distance from the coast, cold fronts, or low pressures (storm
surges). Cold fronts and storm surges usually have intra-annual peri-
odicity. Cold fronts occur during the main dry season (November to
May) with 6 events on average per year, and 2010 was characterized
by the highest number of these events (20) in the area (Ortiz-Royero
et al., 2013). However, it is still needed more knowledge about the oc-
currence of cold fronts, storm surges or mixing dynamics in the Co-
lombian Caribbean.
The ecological effect of extreme events on ecosystems is one of the
main knowledge gaps in ecology. Over the last years, ndings on the
ecological effects of extreme weather events have emerged, such
as the effects on diversity, productivity, reproduction, phenology,
nutrient cycling, and community resistance to invasions (Jentsch and
Beierkuhnlein, 2008 and references therein). A major challenge in
their study lies in the nature of these phenomena. Extreme weather
events are relatively rare, with effects out of proportion compared to
their short duration, and are also difcult to predict and include in cur-
rent numerical climate models. A comprehensive and consistent data-
base on extreme weather events with high temporal and spatial
resolution spanning long time scales is not available. Thus, ecologists
have started to perform experimental climate change research by com-
puter simulation of extreme weather events and their effects on ecolog-
ical processes, biotic interactions, and ecosystem functions (Jentsch and
Beierkuhnlein, 2008).
2.1. Mangroves
Mangrove forest growth requires a combination of environmental
conditions, the most relevant being geomorphology, tides and uvial
ooding, and interstitial water salinity (Krauss et al., 2008). During the
last few decades, eustatic sea level rises in the Caribbean have caused
both mangrove expansion and contraction. Expansions depend on uvi-
al sediment inputs and coastal plain availability; contractions occur
mainly through coastal erosion and anthropogenic disturbances. Never-
theless, other processes related to global change, particularly extreme
events, such as storm surges and hurricanes, have also affected man-
grove areas and have changed not only soil composition but also vege-
tation structure and composition (Lugo, 2008).
While immediate post hurricane effects are represented by exten-
sive areas with fallen and uprooted trees, mainly the largest ones, longer
time scale responses are evidenced when natural regeneration leads to
changes in composition and structure (Lugo, 2008). Soil nutrient avail-
ability may be beneted by the massive litter fall from defoliation of
the forest canopy, which may cause not only increased diversity, but
also primary productivity (Tanner et al., 1991). It has been recognized
that the forest canopy of areas that were previously exposed to hurri-
canes are dominated by Laguncularia racemosa and Avicennia germinans,
owing to their capacity to re-sprout from dormant epicormic buds; the
understory is dominated by a dense carpet of seedlings of Rhizophora
mangle (Baldwin et al., 2001). Knowledge on mangrove species
autoecology and analysis of coastal geomorphology changes are in fact
useful tools to understand the effects of storm and hurricanes in forest
composition.
87
2.2. Beaches
Beaches respond to extreme wave or episodic coastal inundation
events changing their morphology, and normally eroding. Signicant
cross-shore movements of sand and a reshaping of the beach face ac-
company large swell events. Subsequent beach recovery begins quickly
when weak to moderate wave conditions favor onshore sediment trans-
port (Dail et al., 2000). Shoreline responses to a wave/storm event on
wave dominated coasts are important, but studies are rare, due to the
difculty of capturing both pre and post-event conditions when the
events themselves are rare and difcult to predict. Another difculty is
obtaining robust data to adequately characterize the morphological re-
sponse of the beaches over time (Adams et al., 2008; Barnard and
Warrick, 2010; Anderson et al., 2010). Shoreline recoveries from
storm induced erosion can last from one day to years. In fact, the vulner-
ability of a beach to storms depends on the balance between storm fre-
quency and recovery rates (Coco et al., 2014).
Coastal zone managers are increasingly seeking accurate quantita-
tive predictions of beach erosion hazards within a probabilistic frame-
work. This replaces the approach in which only a benchmark event
(usually the largest measured historical event) was applied to assess
beach erosion and inundation. One reason is the dependency of erosion
volumes on storm duration. Shorter duration events having the same
peak wave height result in less beach erosion. Other reason is the merg-
ing of independent meteorological events into one beach erosion event.
For example, two events of equal magnitude occurring within a few
days generate more erosion than if separated by many months, in
which the beach is able to recover from the rst erosion event
(Callaghan et al., 2008). In the case of developed (or anthropogenic)
coasts, human activities dominate over natural processes in shaping
the coastline and thus, the adaptive capacity (resilience) of the system
is dependent on the socio-economic system in which it occurs, and
not only on environmental variables, as in natural coasts (Cooper and
Lemckert, 2012).
2.3. Reefs
Coral reefs are dynamic systems with non-linear behavior, critical
thresholds, and multiple successional phases or stable states
(Dudgeon et al., 2010). Although regime shifts in coral reefs are intui-
tively associated with large stochastic events, such as hurricanes, dis-
eases, and mass bleaching, this is not necessarily the case. Slow
degradation of key processes could cause coral reefs to become progres-
sively vulnerable to even small incremental changes (Nystrm et al.,
2008).
Cyclones have been documented as intermittent, severe stress
events on coral reefs. Major effects on coral reefs include debris and sed-
iment from land that may smother corals, physical breaking of corals,
and repositioning of coral reef superstructures (Freeman et al., 2012).
In general, the most drastic effect of wave-induced hydrodynamic forces
is the breaking of the colony at the upstream end; either at a branch or
at the colony's base. Massive coral colonies may be held in place by their
weight, although they are susceptible to toppling (Denny and Gaylord,
2010).
The damage that hydrodynamic forces can exert on coral reefs can
be short-term or long-lasting, depending on the situation. When chron-
ic human stressors are at a low level, reefs can recover. These short-term
disturbancerecovery trends are part of the dynamics of functional coral
reefs, which have been subject to them, especially to hurricanes, for mil-
lions of years (Jackson et al., 2014). For example, at some localities reef
crests routinely lose their coral cover every decade or so because of cy-
clones, yet they have retained their ability to recover quickly and show
no propensity to undergo a long-term change. In contrast, many coral
reefs have been slowly pushed close to a threshold by chronic human
impacts, and commonly fail to recover from pulses of coral mortality
(Hughes et al., 2010).
88 G. Bernal et al. / Journal of Marine Systems 164 (2016) 85100
Besides hydrodynamic forces, heavy rainfall brings towards reefs
greater quantities of fresh or brackish waters (lowering salinity), sedi-
ments, turbidity, and water enriched in nutrients and contaminants
washed from the land (Fabricius, 2005), which on occasion may have ef-
fects on reproduction and growth in corals (Van Woesik, 1991; Coles
and Jokiel, 1992; Fabricius et al., 2007). Also, the associated turbidity
temporarily decreases light availability, lowering the photosynthetic
yield of zooxanthellae and thus coral calcication rates; although reef-
associated algae may be favored by the higher nutrient input, their pro-
ductivity may not be increased owing to the shadowing effect of turbid-
ity and their being uprooted by the waves (Hubbard, 1997; Restrepo
and Alvarado, 2011; Martnez et al., 2012).
Caribbean coral reefs have been the most extensively degraded in re-
cent decades, due to a complex sequence of chronic and acute distur-
bances and to widespread recruitment failure of corals. Coral cover
has declined by about 8090% since the late 1970s or 1980s at most
Caribbean locations, whereas the abundance of macro algae and other
weedy species (Hughes et al., 2010; Jackson et al., 2014) and of sponges
(Pawlik, 2011) has increased sharply. Reef corals including Acropora
spp. were depleted at sites across the Caribbean as early as the 19th cen-
tury, and the proportion of sites dominated by Acropora corals in partic-
ular had declined before the 1980s (Cramer et al., 2012). The most
recent account compiling data of coral cover throughout the Caribbean
(120 m in depth) shows a decline from about an overall mean of 34.8%
in 1970 to a current (2012) 16.8% (range 2.853.1%) (Jackson et al.,
2014). Branching species of Acropora and Porites corals have been re-
placed by non-branching and foliose species of Agaricia and by massive
Porites (Cendales et al., 2002). This recent and widespread collapse of
Caribbean coral communities has been convincingly attributed to vari-
ous combinations of proximate causes such as coral and urchin disease,
coral bleaching, overgrowth by seaweeds, sedimentation, etc., but ulti-
mate causes of human origin that have been established are overpopu-
lation (including increase in tourism), overshing, coastal pollution,
ocean warming and introduced pathogens (Jackson et al., 2014).
Studies have shown that the recruitment of non-branching corals in
the Caribbean declined signicantly in years following hurricanes and
storms (Knowlton et al., 1981; Crabbe, 2012; Mallela and Crabbe,
2009; Crabbe et al., 2002). Storm surges can, for example, aid excavating
sponges in the colonization of massive corals, by throwing against them
sponge-colonized coral branches (Lpez-Victoria and Zea, 2004). Like-
wise, a decline in architectural complexity (roughness) of Caribbean
reefs has occurred since the 1980s, related mainly to hurricanes break-
ing apart branching coral stands that had previously been massively
killed by diseases (Alvarez-Filip et al., 2011). In the Caribbean, before
1984, coral cover was independent from the probability of hurricane
damage, but not afterwards, apparently in relation to degree of
overshing, especially of parrotsh. This seems to have indirectly affect-
ed somehow the capacity of Caribbean reefs to recover from hurricane
damage (Jackson et al., 2014).
3. Extreme events in the Colombian Caribbean
The occurrence of extreme events across the Colombian Caribbean
was studied by time series analysis of winds and waves at the selected
locations. Additionally, documentary records about episodic coastal in-
undation events (locally called mar de leva) were considered.
Wind data were obtained from the Cross-Calibrated Multi-Platform
(CCMP) Ocean Surface Wind (ftp://podaac-ftp.jpl.nasa.gov/allData/
ccmp/L3.0/k/). This database was chosen after checking that it records
wind extremes adequately. Six-hourly time series from 1988 to 2011
were analyzed for the four locations of interest. Winds exceeding posi-
tively the 99% (q99 percentile) of the data were considered extremes
(IPCC, 2013), and each event was chosen according to the peak over
threshold method (Leadbetter, 1991). Extreme events were counted
monthly and annually. We found a marked difference in the variance be-
tween winds before and after 1999, veried with a Levene test of change
of variance. The critical value of the F-distribution was F(0.05,1,35,062) =
3.8417, and the computed Levene's statistics were 747, 124, 18 and
474 for Bocagrande, Portete Bay, Old Providence and Tayrona, respec-
tively. Therefore, the null hypothesis that the variances are equal was
rejected, which mean that the variances of the two segments of all series
are statistically different. This change of the variance coincides with a
change in the satellite platforms and the database used as a starting es-
timate of the wind eld in the integration method, hence we calculated
extremes separately for the series before and after 1999 (Fig. 2).
Winds were stronger off Portete Bay (PO), but more variable off
Bocagrande Beach in Cartagena (BG), where the extreme threshold
(q99) was higher. The number of extreme wind events per year with
the effect of change of variance corrected (Fig. 3) does not show an in-
crease in the two decades of data. Over the coastal locations, N15 events
per year occurred in 19911992, 1997, 2002 and 2004, all being the
most signicant El Nio years on record. Over Old Providence in the
SW Caribbean (PR), this number of events occurred only in 1989
(La Nia year) and 2004 (El Nio year).
The number of extreme wind cases per month during Normal, El
Nio (warmer phase in the E. Pacic), and La Nia (colder phase in
the E. Pacic) conditions (according to the series length and the Oceanic
El Nio Index, ONI) are shown in Fig. 4. Wind extremes occurred espe-
cially during the main dry season (from December to March) and the
midsummer drought (JuneAugust), except for Bocagrande, where the
midsummer drought was not outstanding. At Portete Bay the extremes
were more frequent under El Nio conditions. At Tayrona Park and
Bocagrande this was true for most of the months. But at Old Providence
they were more frequent during La Nia. These results correspond to
those of Fig. 3, where more extreme events occurred during El Nio
years in PO, TY and BG. There was not a signicant temporal trend for
the annual wind maximum and mean (Fig. 5).
Clearly, frequency of extreme winds over the Colombian Caribbean
does not increase during the hurricane season. Moon et al. (2003),
Zhuo et al. (2008), Wright et al. (2001), Xu et al. (2007), and Montoya
et al. (2013) researched the spatial behavior of the winds and waves
during hurricane conditions, showing that the most energetic winds
and waves occur in the right forward quadrant of the hurricane transla-
tion, and less energetic wind and waves occur in the left backward
quadrant. Additionally far from the eye of the hurricane the winds are
weakened due to the large pressure gradients generated near the eye
of the hurricane (Montoya et al., 2013). That means, based on the
most probable hurricane track behavior in the Caribbean Sea (east-
west direction), that the most energetic wind and waves are located to-
ward the north of the hurricane translation and far from the Colombian
Caribbean Coast. Thus, the extreme wind conditions along the coast of
the Colombian Caribbean Sea may be associated with the annual evolu-
tion of the Caribbean Low Level Jet (CLLJ, Wang, 2007) or the cold fronts.
Time series of signicant wave heights for the studied localities were
modeled using the third generation spectral model WAVEWATCH III
(Tolman, 2002, 2009). Even though NCEP/NCAR Reanalysis I winds do
not register extremes adequately, given their low spatial and temporal
resolution, they can be improved for calculating wave elds employing
a blended methodology. In fact, 10 m winds from NCEP/NCAR Reanaly-
sis I, corrected with the blended methodology proposed by Montoya
et al. (2013) for hurricane conditions, were employed. The trajectory
of hurricanes in the Caribbean and information about them were ob-
tained from the National Hurricane Center, NHC (http://www.nhc.
noaa.gov). The radius of maximum winds Rmax was included in the se-
ries by the parametric representation of the hurricane wind structure
proposed by Willoughby (2004). Details of the model implementation
and wind blending methodology are presented in Montoya et al.
(2013). Again, events exceeding positively 99% of the data were consid-
ered extremes (IPCC, 2013). For average conditions, the correction
methodology proposed by Montoya and Osorio (2014) was employed.
In this methodology, the Environmental PredictionNational Center
for Atmospheric Research (NCEPNCAR, Reanalysis I) winds have
 G. Bernal et al. / Journal of Marine Systems 164 (2016) 85100 89

Fig. 2. Six-hourly wind series for the four regions of interest, including the threshold of the extremes (q99) and the exceeding cases (black dots, representing individual cases). Obtained
from the Cross-Calibrated Multi-Platform (CCMP) Ocean Surface Wind. PO = Portete Bay, TY = Tayrona reefs, BG= Bocagrande Beach, PR = Old Providence reefs. = mean, 2 =
variance, q99 = 99 percentile extreme threshold. Owing to changes in satellite platforms and database in 1999 (vertical dashed line), statistics were computed separately before and after.

been corrected, employing the vector correlation and triple collocation From the models, the mean wave height, the variance, and the ex-
theories combined with information from different accurate sources treme wave height threshold were higher for the oceanic locality, PR
from scatterometer and altimeter data. (Fig. 6). In the continental locations, mean wave height was higher at

Fig. 3. Total extreme wind events per year at the four locations. Data and conventions as in Fig. 2.
90 G. Bernal et al. / Journal of Marine Systems 164 (2016) 85100

Fig. 4. Monthly percentage of extreme winds (number indicated above the bars) separated according to the relative occurrence in Normal, El Nio (warm) and La Nia (cold) conditions at
each locality. Numbers above the bars indicate the number of events. Data and conventions as in Fig. 2.

Tayrona and lower at Bocagrande, but the variance and the extreme
wave height threshold were similar for both of these locations.
Table 1 shows the events exceeding 5 m signicant wave height at
the four locations of interest. In PO, maximum waves coincide with Hur-
ricanes Cesar (1996), Lenny (1999) and Humberto (2007), all of them
after 1995. In TY, N 5 m waves occurred during Hurricanes Joan (1988)
and Lenny (1999). In BG, the highest waves happened during Hurricane
Emily (1993), and in PR, during Hurricane Cesar (1996). However,
similar to winds, in general the longer events coincide with the dry sea-
son and/or cold fronts. The number of events of waves N 5 m per year
(not shown) is very variable at the four locations, but the years with
the most events were 1984, 1988, 1996, 2004 and 2009. Seasonally, a
higher number of high wave events occur in the dry season; only in
PR there was a minor increase of these events during the midsummer
drought. No relationship was found between the number of N5 m
wave height events and ENSO conditions. The only location with a

Fig. 5. Annual mean and maximum (Max.) of the wind speed anomalies (individual values minus yearly mean), with their linear trend (box) calculated with least square method. Data and
conventions as in Fig. 2.
 G. Bernal et al. / Journal of Marine Systems 164 (2016) 85100 91

Fig. 6. Hourly signicant wave height series for the four locations of interest, including the threshold of the extremes and cases exceeding them (black dots). Results from the
WAVEWATCH III model as explained in the text. Conventions as in Fig. 2.

signicant temporal trend towards an increase in the wave annual max-
imums was Portete Bay (Fig. 7). High maximums coincided with the
events identied in Table 1, and marked a change after 1995.
A review of Colombian newspapers from 1970 to 2008 was per-
formed to record the published cases of episodic coastal inundation
events and/or storm surges (mares de leva). Most events were related
to winds, hurricanes, storms, and cold fronts (Fig. 8a). Seasonally, they
were more common between November and March (the dry season).
During the midsummer drought (locally known as veranillo de San
Juan) there was also an increase in the number of cases (Fig. 8b).
Press reports started in 1983 and during 1996 a peak in the reports
was evident, with the passing of hurricane Cesar (Fig. 9).
The four hurricanes that will be mentioned for the study cases ana-
lyzed below (in relation to ecosystems) are shown in Fig. 9. Hurricane
Cesar (1996), Category 4, was remarkable because it passed very close
to La Guajira Peninsula, a region known for its low sea surface temper-
ature where hurricanes die. It had a stronger impact on Portete Bay
and Old Providence, the northernmost localities. Hurricane Lenny
(1999), Category 4, had an unprecedented west-to-east track (the
direction contrary to all Atlantic hurricanes) and affected the entire
Colombian coast, but most importantly Portete Bay and Tayrona
where maximum wave heights occurred. Hurricane Beta (2005),
Category 3, was also anomalous as it was compact but intense and con-
ned to the SW Caribbean, very close to San Andres and Old Providence,
where its impacts were more noticeable. Hurricane Tomas (2010), cat-
egory 2, was not among the most impacting, but occurred in combina-
tion with cold front events in the Caribbean.
In the next sections, considerations about the different ecosystems'
responses to extreme events are made in the context of the four loca-
tions chosen as study cases. For mangroves and beaches, the effects of

Table 1
Events of waves exceeding 5 m in height in the four locations (conventions as in Fig. 2) with a description of the physical phenomenon associated. WH = wave height reached; L =
duration of the event (hours).

Event PO TY BG PR

Month-Year WH L WH L WH L WH L Physical phenomenon

03-1979 5.2 73 5.2 90 Dry seasonCold front

08-1980 5.1 12 Hurricane Allen
12-1984 5.5 127 Hurricane Lili
10-1988 5.2 17 6 11 5.4 33 Hurricane Joan
08-1993 5.3 10 6.6 10 Hurricane Emily
07-1996 7.3 13 5.4 7 7.8 19 Hurricane Cesar
03-1997 5 70 Dry seasonCold front
07-1997 5.1 25 Hurricane Denny
11-1999 7.5 24 6.1 18 Hurricane Lenny
01-2000 5.2 34 5 52 Dry seasonCold front
02-2001 5.2 180 5.5 179 5.6 163 Dry seasonCold front
03-2004 5.7 143 Dry seasonCold front
09-2004 5.8 17 Hurricane Ivan
07-2005 6.2 12 Hurricane Emily
09-2007 7.4 10 6.2 12 Hurricane Humberto
03-2009 5.2 114 Dry seasonCold front
92 G. Bernal et al. / Journal of Marine Systems 164 (2016) 85100

Fig. 7. Annual mean and maximum (Max.) of the wave anomalies (individual values minus yearly mean), with their linear trend (box) calculated with least square method. Data as in
Fig. 6, conventions as in Fig. 2.

extreme events were studied directly by us, while for coral reefs we
used cases whose effects were described in the literature, relating
them to the associated wave magnitudes found in our models.
4. Case studies
4.1. Portete Bay mangroves
Molina (2009) analyzed the coastal geomorphology and vegetation
changes along an extensive area covered by mangroves (5 km2) and
coastal vegetation, along Portete Bay (80 km2), located at the northeast-
ern Colombian Caribbean. The area is predominantly exposed to north-
eastern trade winds. The mean annual precipitation (300 mm) is
concentrated in two short rainy seasons, from March to June and from
September to November. The mean annual air temperature is 27 C
(1528 C). Upwelling along the Guajira coast causes sea surface salinity
increases and sea surface temperature decreases (INVEMAR, 2005). The
geomorphology is characterized by mountains (up to 900masl), small
hills, highly eroded foothills and large coastal plains. The littoral is char-
acterized by sand bars that enclose salt lagoons. Fringe mangroves and

Fig. 8. Number of episodic coastal inundation and/or storm surges (mares de leva) events reported by the Colombian newspapers (19702008), classied by the possible cause to which
each was attributed: (a) intensity and month of the year, (b) and number of reports per year (19702008). Causes are according to the press (in the case of waves, even though they can be
associated with other causes, they indicate that the report only argued high waves as a cause for the event). LPS: Low pressure systems. Intensity was classied as C1: Sea entry restrictions,
problems with navigation; C2: Damage to the beaches, oods, paralysis on boats, coastal erosion; C3: Floods, damage to beaches, damage to infrastructure, housing destruction, loss of life.
 G. Bernal et al. / Journal of Marine Systems 164 (2016) 85100 93

Fig. 9. U.S. National Hurricane Center's best tracks of the hurricanes relevant for the cases studied here. The markers are plotted each 6 h. Black stars mark the locations of the sites of
interest.

halophitic vegetation dominate lagoon shores and mudats (IGAC,

1975; IDEAM, 1998; Snchez-Pez et al., 1997). Gullies or sporadic
water courses shed the littoral with fresh water mainly during the
rainy seasons (IAVH, 1998).
Three mangrove types were recognized in the study area (Molina,
2009): (1) fringe mangroves dominated by R. mangle, (2) basin man-
groves dominated by A. germinans, and (3) mixed forests composed of
these two species and of L. racemosa. Basin mangroves were dominated
mainly by small trees of A. germinans with a few pool-fringing trees of R.
mangle. In mixed mangroves, the dominant trees of L. racemosa and A.
germinans were larger than those of R. mangle that were present.
While trunk diametric distributions of fringe and mixed mangroves
showed a unimodal shape biased to the left, in the mixed mangroves
it was a truncated inverted J shape (Molina, 2009), which may imply
a recovered forest. However a truncated J shape diametric distribution
also could characterize a forest where big trees were cut or fallen.
Spatialtemporal changes in geomorphology and vegetation were
reconstructed based on eld sampling and digital analysis of
georeferenced aerial photographs, from 1957 (1:60,000), 1995
(1:50,000), and 2003 (1:50,000), and using algebraic maps of ArcGis
9.2 (Esri, Inc., Redlands, CA). Vegetation changes were related to pro-
cesses such as sediment accumulation and erosion associated to marine
processes that facilitated vegetation colonization and/or loss.
Main changes in mangrove vegetation are shown in Fig. 10. Since the
rst period (19571995) encompasses 38 years and the second (1995
2003) only 8 years, the analyzed time spans were considered as two
time scales. Although environmental changes during the longer period
(19571995) favored mangrove growth, the coastal erosion rates
were also the highest, and there was forest loss associated to beach ero-
sion related to sea level rise, and stronger and higher number of ex-
treme wind events recorded in Portete bay in 19911992, a strong El
Nio period. In contrast, inland mangrove colonization took place, espe-
cially by the A. germinans-dominated basin forest and in a lower propor-
tion of fringe stands dominated by R. mangle and of mixed mangroves.
Overall, mangrove losses were rather low during this period.
During the 19952003 period, although the shortest, major changes
occurred in mangrove vegetation composition (Fig. 10). A large area Fig. 10. Top: Changes in mangrove extensions in Portete Bay. Gain corresponds to the area
covered mainly by Avicennia-dominated basin mangrove was lost prob- that was occupied by any other cover different from mangrove forest in a former period,
ably due to erosion of the substrate and in a lesser degree to the conver- but was covered by mangroves in the second one. Loss corresponds to areas that were
sion to Rhizophora and mixed mangrove stands on new substrates covered by mangroves in the rst period and shifted this vegetation cover in the second.
Net gain or loss was obtained from the difference between them. Bottom: Changes in
created by sediment transportation and deposition. Nevertheless, this the dominant species that led to variations in extensions of the physiographic types of
Avicennia-dominated basin mangrove was also the most successful col- mangroves and their conversion into another mangrove type (as percentage of the total
onizer, not only in new areas (70%) but also through the conversion studied area).
94 G. Bernal et al. / Journal of Marine Systems 164 (2016) 85100

from other mangrove types (30%). Conversely, mixed mangrove origi-
nated mainly from conversion of the other two mangrove types (87%).
Sixty percent of the Rhizophora-dominated forest was originated from
colonization of new areas and the rest from conversion of other two
mangrove types. A combination of high evaporation and low precipita-
tion that characterizes the region, and the sedimentation processes and
natural disturbances due to strongest events after 1995, were the possi-
ble reasons for the Avicennia dominance in mangrove stands during this
period and the present time (Molina, 2009).
Despite the high daily wind speed that predominates in the area
and that was regular prior to 1995, it was during the second
period (19952003) that two severe weather events hit PO, tropical
storm Cesar (1996, which later turned into hurricane category) and
hurricane Lenny (1999, which severely intensied the NE trade winds
for a few days). These are the two highest wind and wave events
that occurred during this time span (Table 1). The increase of extreme
wind and wave events after 1995 and these two particular events
may explain, at least partly, the vegetation changes found. In fact, the

Fig. 11. Location of areas where major changes in costal processes took place between 1995 and 2003.
 G. Bernal et al. / Journal of Marine Systems 164 (2016) 85100 95

present mangrove type distribution exhibits the ngerprints of
such events.
The largest changes in the shortest period were related to the in-
crease in A. germinans-dominated forest probably due to the resistance
of this species to wind damage, on one hand, and its vegetative capacity
to colonize new sandy substrata and resprout from fallen trees, on the
other. This could also be the cause of the expansion of mixed mangroves
upon the other two mangrove types, especially in sites where the salin-
ity was about 20% less than in A. germinans dominated mangroves, a sit-
uation that favors L. racemosa colonization. Indeed, L. racemosa is not
only a pioneer species, recognized by the capacity to invade disturbed
forests (Beneld et al., 2005), but also because, as A. germinans, it can re-
sprout from fallen trees (Baldwin et al., 2001). High P concentrations
(55 ppm) found in the surface soils of Portete Bay mixed mangroves
(Molina, 2009) may also be an evidence of hurricane sediment deposits
(Castaeda-Moya et al., 2009).
The small mean size of trees in the A. germinans-dominated stands
may also be evidence of post-hurricane remnant forests, since large
trees are more prone to damage by strong winds and resprouting is
more frequent among small trees (Baldwin et al., 1995; Kovacs et al.,
2004; Lugo, 2008). The unimodal shape of the diametric distributions
of both mixed and R. mangle-dominated forest may also reect this pre-
vious disturbance.
Additionally to vegetation structure and composition, spatial distri-
bution of mixed and Avicennia-dominated mangrove stands also has
the imprint of these events. The most affected forests were located in
the northern and north-eastern sides of the area, the direction of the
storm-accentuated trade winds. This was especially true for those
stands that were not protected by other vegetation and physical barriers
(Fig. 11).
4.2. Cartagena beaches
Bocagrande Beach at Cartagena City was monitored continuously
between 2009 and 2011 using a video camera system as part of the pro-
ject HORUS (National University of Colombia and Cantabria University,
Spain, http://www.horusvideo.com/). During this period, some extreme
events produced a strong retreat of the Bocagrande Beach, particularly
in 2010.
A requirement for quantifying the information contained in an
image (the shoreline position) is the knowledge of the photogrammet-
ric transformation between a 2D image (pixel) and 3D world coordi-
nates (meters). One approach is to constrain at least one system-
related spatial dimension of the world coordinate (X, Y, Z), thereby
yielding a unique solution for the image to ground transformation.
The most common approach is to assume the Z coordinate according

Fig. 12. Shoreline and wave climate evolution between June 2009 and March 2011. Beach width of Bocagrande Beach (Cartagena City), measured with the camera coastal monitoring
system HORUS (gray point) and trend of accretion and erosion (black line) (upper panel). Signicant wave height (medium panel). Wave direction (lower panel). The events and the
cold fronts described in the text are marked.
96 G. Bernal et al. / Journal of Marine Systems 164 (2016) 85100
to hydrodynamic conditions; in our case the Z level was assumed to be
the tidal level (even though in Cartagena it is small, around 40 cm). In
this way, based in the HORUS system, the Pinhole model or the photo-
grammetric transformation was solved. The resolution of the HORUS
system for this camera in the study area was around 30 cm/pixel
(Osorio et al., 2012; Prez et al., 2013).
After the estimation of shoreline position and its temporal evolution,
the extreme events analyzed during this period (Fig. 12) were not out-
standing regarding wave height but nevertheless they produced a 50 m
beach width retreat between Jan 2010 and Jan 2011. The normal sea-
sonal variability of the beach width, without extreme events, was
around 1020 m, comprising episodic erosion (January to March) and
recovery (April to November) phases. But for this 50 m retreat, the
most likely explanation could be that the March 2010 cold front events
eroded the beach (around 25 m) and affected its resilience (see rst
black box between Jan. and Apr. 2010), so that when the Tropical De-
pression of Oct. 1 2010 (after Tropical Storm Nicole) and a month later
(No 23, 2010) Hurricane Tomas (category 2) arrived, the beach had
not completely recovered, and these events produced a very high im-
pact (another + 25 m retreat). The direction of approaching waves
(Fig. 12) also could have played a key role in this case, which will be ex-
plained below.
Morphological changes ranged from 20 m to 30 m of erosion of
beach width (see rst black box in Fig. 12. between Jan and Apr 2010)
Fig. 13. Photos of Bocagrande Beach in July 2010 (top) and Dec 2010 (bottom), after the seasons of consecutive extreme events.
and were due to direct action of waves with normal components over
the beach; in particular, the erosion of the upper beach and the deposi-
tion of the rst subaquatic bar. Between 50 m and 100 m offshore, the
alongshore sediment transport (the direction will be explained later)
also generated erosion and leaked sediment from the system. However,
a single extreme event could not be relevant in the erosion, because the
resilience of the system was historically able to recover in the following
months (April to October). However, the real effect was due to the accu-
mulation of atypical extreme events (hurricanes and cold fronts) from
November to March (during 20092010 and 20102011). The erosion
increased dramatically and the system was not able to recover from
the impact of a the tropical depression of Oct 1, 2010, which produced
erosion of about 10 m, followed by Hurricane Tomas a month later
(Nov 35, 2010), which added another 10 m. Thereafter, the system im-
balance was exacerbated by several consecutive extreme events associ-
ated to cold fronts escalating the erosion to 30 m (in Feb. 2011).
According to Fig. 12, the rate of accretion was 5 m/month in 2009
(JulNov) against 3 m/month in 2010 (AprJul). This contrasts with
the rate of erosion of 8 m/month in 20092010 against 10 m/month
in 20102011. Fig. 13 depicts Bocagrande Beach in July 2010 and Dec.
2010, after the seasons of consecutive extreme events.
The genesis of this extreme erosion was mainly associated with the
particular 2010 Atlantic Hurricane Season: the Accumulated Cyclone
Energy (ACE) value (Bell, 2000) was 166.3 104 kt2, which
 G. Bernal et al. / Journal of Marine Systems 164 (2016) 85100
corresponds to 190% of the 19502000 median value (Bell et al., 2011).
This places 2010 as the tenth most active season since 1950. NOAA clas-
sies the season as above normal, as dened in http://www.cpc.ncep.
noaa.gov/products/out-looks/background_information.shtml.
The Atlantic storm tracks during 2010 were generally divided into
two clusters. One cluster comprised eight storms that formed over the
eastern tropical Atlantic. The second cluster of storm tracks during
2010 consisted of eleven systems that formed over or near the Caribbe-
an Sea. The regional atmospheric conditions during 2010 showed links
to a combination of three climate factors. The rst was the active Atlan-
tic phase of the tropical multi-decadal signal, which increased Atlantic
hurricane activity since 1995 (Bell and Chelliah, 2006). The second
was the transition between El Nio and La Nia; strong El Nio condi-
tions prevailed over the tropical Pacic from January to March 2010,
which swiftly transitioned to moderate La Nia conditions developed
during July 2010 and then to strong events during ASO (AugustSep
October). Its impacts intensied the tropical multi-decadal signal,
resulting in an extensive area of weak vertical wind shear and upper-
level easterlies in the Caribbean Sea (Blunden et al., 2011). The third
was record warm SSTs (Sea Surface Temperatures) in the Main Devel-
opment Region (MDR), which covers the Caribbean Sea and tropical
Atlantic Ocean (9.5N and 21.5N). These rst two atmospheric condi-
tions, in association with record warm SSTs, originated early in the
year a dramatic weakening of the mean anticyclonic gyre over the cen-
tral Atlantic, and record weak trade winds. This explains many Atlantic
named storms formed from the African easterly. This is a classic scenario
for tropical storm formation and amplication, and it is seen routinely
during active seasons and high-activity areas (Bell et al., 2011).
Fig. 14. Wave time series in Bocagrande (Cartagena) between September and November 2010. Signicant wave heightHs (m) and peak periodTp (sec) (upper panel). Wave direction
(lower panel).
97
These storms during the second semester of 2010 are responsible for
particular wave patterns: 1) the west component of the waves opposite
to the normal condition of waves dominated by trade winds: Fig. 14 de-
picts SW and W dominant wave direction post-Tropical storm Nicole;
2) two high peak periods responsible for high swells developed during
extreme events (Tropical depression after Tropical Storm Nicole and
Hurricane Tomas); and 3) strong swell (during Tomas) matching a
quick change in wave direction (from NE to S and SW). As a response
to the storm condition, these three particular wave patterns produce
alongshore sediment transport eastward, and after the higher wave
height (Hs over 3 m), they produce offshore sediment transport from
the upper beach to the outer-bar or outside the systems. This means
that this sediment never goes back to the beach.
4.3. Santa Marta (Tayrona) reefs
Since 1980, the strongest wave events impacting the coast of Santa
Marta, where Tayrona Park reefs are located, occurred during the pas-
sage of hurricanes Joan (1988) and Lenny (1999), with signicant
wave heights reaching or surpassing 6 m; ve other extreme events
with signicant wave height N5 m (up to 5.4 m) hit the local reefs
from 1979 to 2009 (Table 1). Thanks to the Colombian Coral Reefs
Monitoring System (SIMAC), there is published information on the ef-
fects of hurricane Lenny (November 1999) on Tayrona reefs.
Comparing SIMAC data from December 1998 and December 1999
plus additional observations, Rodrguez-Ramrez and Garzn-Ferreira
(2003) could ascertain the effect of Lenny on Tayrona reefs (see also
Garzn-Ferreira and Rodrguez-Ramrez, 2010). According to these
98 G. Bernal et al. / Journal of Marine Systems 164 (2016) 85100
authors, Lenny only caused a slight coral cover reduction, from 34% to
31%, especially affecting the shallow-dwelling elkhorn coral Acropora
palmata. The algae increased 10% as a result of fragmentation and
overturning of coral colonies. Sedimentation rates (also measured
under SIMAC) almost doubled, from 1.24 in the months previous to
Lenny to 2.28 mg cm2 day1 during Lenny, due to high resuspension
by waves. In subsequent years (until at least 2004, with two more
events with signicant wave heights N5 m, Table 1), A. palmata did
not recover, but there was a partial rebound of overall coral cover in
monitored Tayrona shallow plots, indicating that strong surge effects
were transient at the community level; overall, coral cover at Tayrona
reefs was stable, at least during the period 19982004 (Rodrguez-
Ramrez et al., 2010), despite Lenny and the two other signicant
wave heights above 5 m in 2000 and 2001. Most coral loss in the
Santa Marta area and elsewhere in the Colombian Caribbean occurred
from the late 1970s into the early 1990s as a result of multiple
interacting natural and human factors (review in Rodrguez-Ramrez
et al., 2010). Indeed, today at Tayrona reefs there is a dynamic of con-
stant coral recovery which counters low-level chronic loss from
bleaching (Rodrguez-Ramrez et al., 2008), diseases (Gil-Agudelo and
Garzn-Ferreira, 2001), predation (Santodomingo et al., 2002), and
stress from continental runoff (Vega-Sequeda et al., 2011; Ardila,
2014). However, this scenario of slow loss and recovery is being punctu-
ated by strong episodic losses from extreme events such as the 2010
bleaching event which was closely followed by a storm surge and
a cold spell (from upwelling) that prevented the recovery from
bleaching that was already underway and produced further mortality
(Vega-Sequeda et al., 2011; Esteban, 2012; Romero-Rodrguez et al.,
2014).
4.4. Old Providencia reefs
During the study period, Old Providence had several events with sig-
nicant wave heights N 5 m, ve from hurricanes and three from cold
fronts (Table 1). The island possesses a barrier reef that surrounds it
on the NE, E and SE sides, which protects the lagoon and land from
the predominant waves from the NE (Geister, 1992); cold fronts usually
only increase predominant wave height and thus would not impact the
shore signicantly. However, during hurricane season storms, waves
from other directions strongly affect unprotected shores, including its
western terrace reef formations. Hurricane Beta (October 2005) had a
trajectory very close to the San Andres and Old Providence Archipelago.
It made landfall from the North and crossed from East to West. Although
according to our models, Beta only had a 2.37 m signicant wave height
(thus not presented in Table 1), we included it as a case study because it
is the only hurricane passing through the San Andrs and Old
Providence Archipelago with published ecological information on its
impact. In this case, damage was insignicant for coral reefs, sea grasses,
beaches and mangroves, but terrestrial vegetation and infrastructure on
the island were severely damaged; in the western terrace formations,
only 10% of corals were affected, with fragmentation reaching 9% and
bleaching 1% (Rodrguez-Ramrez and Reyes-Nivia, 2008). SIMAC mon-
itoring data at nearby San Andres Island (western terraces) did not
show a signicant change in coral cover between 2005 and 2006
(Ardila, 2014), although coral overturning and breakage and abraded
surfaces were evident in 19921994, which was attributed to the pass-
ing of Hurricane Joan in 1988 (Zea et al., 1998), with wave heights above
5 m (Table 1). Overall, assuming a similar fate in Old Providence reefs,
most coral decline in San Andrs occurred before 1988, and coral
cover has remained relatively stable from 1988 to 2004 (Rodrguez-
Ramrez et al., 2010) with a slight but signicant trend towards declin-
ing when analyzed from 1988 to 2011 (Ardila, 2014), more attributable
to chronic or episodic loss from bleaching (Romero-Rodrguez et al.,
2014), diseases (Snchez et al., 2010), sewage (Chaves-Fonnegra et al.,
2007), and biotic interactions (Lpez-Victoria et al., 2006), than from
wave extremes.
5. Conclusions
Extreme events of winds over the Colombian Caribbean are related
to seasonality of the Caribbean Low Level Jet (CLLJ), being more frequent
during the dry seasons (December to March and June to July). They do
not increase during the hurricane season, probably because the strong
winds produced by hurricanes are located mainly north of the CLLJ
core. In front of PO there are more wind extremes during El Nio condi-
tions, while in PR there are more of them during La Nia. There is no sig-
nicant trend for the annual maximum and mean of wind over the
studied localities.
In turn, extreme events of wave heights over the Colombian Caribbe-
an are seasonally related to CLLJ and winds, but to the hurricanes too,
due to the swell produced north of the CLLJ by the associated extreme
winds. No relationship between extreme events of waves and ENSO
was found. Only the PO wave series show a trend towards increasing
its annual maximums.
Press reports of cases of episodic coastal inundation events and/or
storm surges coincide with the seasonality of extreme events of winds
and waves, increasing during the main dry season, the midsummer
drought and hurricane season (OctoberNovember). Main causes have
been attributed to winds, hurricanes, and cold fronts.
The ecological responses of three ecosystems (a mangrove, a beach,
and two reefs) help to understand the complex interrelationship be-
tween the ecosystem and the oceanic energetic extreme events. These
responses document cases ranging from long response times, as in man-
groves, to more adapted ecosystems like reefs, up to very fast response
times with synergetic events, as in BG Beach.
As recognized in other mangroves (Doyle et al., 2010), present man-
grove distribution, structure, and composition in Portete Bay are the re-
sult of the combined action of sea level rise and extreme events, such as
hurricanes and storms that determine not only ooding levels, but also
salinity, sediments deposition, and erosion.
The 2010 season was the tenth most active since 1950, having being
classied by NOAA as above normal. The regional climatic patterns
(Active Atlantic phase of the tropical multi-decadal signal and transition
in July 2010 between strong-Nio to moderate-Nia (after strength
during ASO), in association with record warm SSTs, was the classic sce-
nario for tropical storm formation and amplication. During the second
semester of 2010, these storms were responsible for particular wave
patterns: 1) the marked west component of the waves, 2) two high-
peak periods responsible for high swells developed during extreme
events (OctNov, 2010), 3) strong swell (during Tomas). These pro-
duced in Bocagrande beach (Cartagena) a rst erosion (20 m), and after-
wards, a system imbalance with accelerated extreme erosion associated
to cold fronts (30 m in Feb. 2011). This example shows how a combina-
tion of extreme events could generate a great loss of coastal areas and
could affect infrastructure and coastal activities (e.g. tourism, among
others). This means that this should be taken into account by coastal
managers in future developing plans, and that they should plan in the
context of extreme climate variability.
The case studies on Colombian coral reefs show a relatively small im-
pact of extreme waves. Most reef degradation had occurred before the
studied period, and a long-term impact is not evident in the data, as
coral cover remained stably low or is slowly decreasing, more as an
effect of chronic or episodic human-related stressors. The small impact
detected above a background of degradation from other causes does
not allow us to determine whether normal reef recovery from
hurricanes is being jeopardized by the currently low health status of
Colombian Caribbean coral reefs.
Acknowledgements
The authors wish to thank Colciencias, grant number 1118-569-
34826 and the National University of Colombia, grant number 17944,
for supporting the project EVENTOS OCENICOS EXTREMOS EN
 G. Bernal et al. / Journal of Marine Systems 164 (2016) 85100
ECOSISTEMAS COSTEROS INSULARES DEL PACFICO Y CARIBE
COLOMBIANOS; the National University of Colombia for the time of-
fered by Professor Andrs Osorio during his sabbatical year to help nish
this paper; several students from National University of Colombia:
Oscar Zapata, Mara Alejandra Piedrahta, Deisy Alejandra Romero, and
Johann Kamil Delgado for their collaboration; and the cited institutions
which have made available the oceanographic data used, especially
NOAA and NASA. Sven Zea's work is contribution 433 of CECIMAR.
References
Adams, P.N., Inman, D.L., Graham, N.E., 2008. Southern California deep-water wave cli-
mate: characterization and application to coastal processes. J. Coast. Res. 24 (4),
10221035. http://dx.doi.org/10.2112/07-0831.1.
Alvarez-Filip, L., Gill, J.A., Dulvy, N.K., Perry, A.L., Watkinson, A., M., C.I., 2011. Drivers of
region-wide declines in architectural complexity on Caribbean reefs. Coral Reefs 30,
10511060. http://dx.doi.org/10.1007/s00338-011-0795-6.
Anderson, T.R., Frazer, L.N., Fletcher, C.H., 2010. Transient and persistent shoreline
change from a storm. Geophys. Res. Lett. 37, L08401. http://dx.doi.org/10.1029/
2009GL042252.
Ardila, M.E., 2014. Ocupacin de espacio por esponjas y otros componentes ssiles en
arrecifes del Caribe colombiano en relacin con la materia orgnica en suspensin
(M.Sc. Thesis M.Sc. Biology) Universidad Nacional de Colombia, Bogot (68 pp.).
Baldwin, A., Egnotovich, M., Ford, M., Platt, W., 2001. Regeneration in fringe mangrove
forests damaged by Hurricane Andrew. Plant Ecol. 157, 149162.
Baldwin, A., Platt, W.J., Gathen, K.L., Lessmann, J.M., Rauch, T.J., 1995. Hurricane damage
and regeneration in fringe mangrove forests of Southeast Florida, USA. J. Coast. Res.
Spec. Issue 21, 169183.
Barnard, P.L., Warrick, J.A., 2010. Dramatic beach and nearshore morphological
changes due to extreme ooding at a wave-dominated river mouth. Mar. Geol. 271,
131148.
Bell, G.D., 2000. The 1999 North Atlantic Hurricane Season: a climate perspective. State of
the Climate in 1999. Bull. Am. Meteorol. Soc. 81, S19S23.
Bell, G.D., Chelliah, M., 2006. Leading tropical modes associated with interannual
and multi-decadal uctuations in North Atlantic hurricane activity. J. Clim. 19,
590612.
Bell, G.D., Blake, E., Kimberlain, T., Landsea, C., Schemm, J., Pasch, R., Goldenberg, S., 2011.
The 2010 North Atlantic Hurricane Season A Climate Perspective. NOAA (http://
www.cpc.ncep.noaa.gov/products/expert_assessment/hurrsummary_2010.pdf).
Beneld, S., Guzmn, H., Mair, J., 2005. Temporal mangrove dynamics in relation to coastal
development in Pacic Panama. J. Environ. Manag. 76 (3), 263276.
Blunden, J., Arndt, D.S., Baringer, M.O. (Eds.), 2011. State of the Climate in 2010. Bull.
Amer. Meteor. Soc., Special Supplement to the Bulletin of the American Meteorolog-
ical Society 92 (6), pp. S1S266.
Callaghan, D.P., Nielsen, P., Short, A., Ranasinghe, R., 2008. Statistical simulation of wave
climate and extreme beach erosion. Coast. Eng. 55, 375390.
Castaeda-Moya, E., Twilley, R.R., Rivera-Monroy, V., Zhang, K., Davis III, S.E., Ross, M.,
2009. Sediment and nutrient deposition associated with Hurricane Wilma in man-
groves of the Florida coastal everglades. Estuar. Coasts 33, 4558.
Cendales, M.H., Zea, S., Daz, J.M., 2002. Geomorfologa y unidades ecolgicas del complejo
arrecifal de Islas del Rosario e Isla Bar (Mar Caribe, Colombia). Rev. Acad. Colomb.
Cienc. Exctas Fs. Nat. 26 (101), 497510.
Chaves-Fonnegra, A., Zea, S., Gmez, M.L., 2007. Abundance of the excavating sponge
Cliona delitrix in relation to sewage discharge at San Andrs island, SW Caribbean, Co-
lombia. Bol. Investig. Mar. Costeras 36, 6378.
Coco, G., Senechal, N., Rejas, A., Bryanc, K.R., Capo, S., Parisot, J.P., Brown, J.A., MacMahan,
J.H.M., 2014. Beach response to a sequence of extreme storms. Geomorphology 204,
493501.
Coles, S.L., Jokiel, P.L., 1992. Effects of salinity on coral reefs. In: Connell, D.W., Hawker,
D.W. (Eds.), Pollution in Tropical Aquatic Systems. CRC Press Inc., London,
pp. 147166 (252 pp.).
Cooper, J.A.G., Lemckert, C., 2012. Extreme sea-level rise and adaptation options for coast-
al resort cities: a qualitative assessment from the Gold Coast, Australia. Ocean Coast.
Manag. 64, 114.
Crabbe, M.J.C., Mendes, J.M., Warner, G.F., 2002. Lack of recruitment of non-branching
corals in discovery bay is linked to severe storms. Bull. Mar. Sci. 70 (3), 939945.
Crabbe, M.J.C., 2012. The inuence of extreme climate events on models of coral colony
recruitment and survival in the Caribbean. Am. J. Clim. Chang. 1, 3340.
Cramer, K.L., Jackson, J.B.C., Angioletti, C.V., Leonard-Pingel, J., Guilderson, T.P., 2012. An-
thropogenic mortality on coral reefs in Caribbean Panama predates coral disease
and bleaching. Ecol. Lett. http://dx.doi.org/10.1111/j.1461-0248.2012.01768.x.
Dail, H.J., Merrield, M.A., Bevis, M., 2000. Steep beach morphology changes due to ener-
getic wave forcing. Mar. Geol. 162, 443458.
Denny, M.W., Gaylord, B., 2010. Marine ecomechanics. Annu. Rev. Mar. Sci. 2, 89114.
http://dx.doi.org/10.1146/annurev-marine-120308-081011.
Doyle, T.W., Krauss, K.W., Conner, W.H., From, A.S., 2010. Predicting the retreat and mi-
gration of tidal forests along the northern Gulf of Mexico under sea-level rise. For.
Ecol. Manag. 259, 770777.
Dudgeon, S.R., Aronson, R.B., Bruno, J.F., Precht, W.F., 2010. Phase shifts and stable states
on coral reefs. Mar. Ecol. Prog. Ser. 413, 201216.
Emanuel, K., 2005. Increasing destructiveness of tropical cyclones over the past 30 years.
Nature 436, 686688.
99
Esteban, L.V., 2012. Seguimiento del efecto del blanqueamiento masivo de 201011 del
coral Acropora palmata en el rea de Santa Marta (NE Colombia) (B.Sc. degree work
report, Biology) Universidad Nacional de Colombia, Bogot (23 pp.).
Fabricius, K., 2005. Effects of terrestrial runoff on the ecology of corals and coral reefs: re-
view and synthesis. Mar. Pollut. Bull. 50, 125146.
Fabricius, K., Hoegh-Guldberg, O., Johnson, J., McCook, L., Lough, J., 2007. Vulnerability of
coral reefs of the Great Barrier Reef to climate change. In: Johnson, J., P.A., M. (Eds.),
Climate change and the Great Barrier Reef. Chapter 17. Great Barrier Marine Park Au-
thority and Australian Greenhouse Ofce, Australia, pp. 515554 (801 pp.).
Freeman, L.A., Miller, A.J., Norris, R.D., Smith, J.E., 2012. Classication of remote Pacic
coral reefs by physical oceanographic environment. J. Geophys. Res. 117, C02007.
http://dx.doi.org/10.1029/2011JC007099.
Garzn-Ferreira, J., Rodrguez-Ramrez, A., 2010. SIMAC: development and implementa-
tion of a coral reef monitoring network in Colombia. Rev. Biol. Trop. 58 (Suppl. 1),
6780.
Geister, J., 1992. Modern reef development and Cenozoic evolution of an oceanic island/
reef complex: Isla de Providencia (Western Caribbean Sea, Colombia). Facies 27,
170.
Gil-Agudelo, D.L., Garzn-Ferreira, J., 2001. Spatial and seasonal variation of the Dark
Spots Disease in coral communities of the Santa Marta area (Colombian Caribbean).
Bull. Mar. Sci. 69 (2), 619629.
Hubbard, D., 1997. Reefs as dynamic systems. Pp. 4367. In: Birkeland, C. (Ed.), Life and
Death of Coral Reefs. Chapman and Hall, New York (536 pp.).
Hughes, T.P., 1994. Catastrophes, phase shifts, and large scale degradation of a Caribbean
coral reef. Science 265 (5178), 15471551.
Hughes, T.P., Graham, N.A.J., Jackson, J.B.C., Mumby, P.J., Steneck, R.S., 2010. Rising to the
challenge of sustaining coral reef resilience. Trends Ecol. Evol. 25 (11), 633642.
IAVH, Instituto Alexander Von Humboldt, 1998. El Bosque seco Tropical (Bs-T) en
Colombia. Programa de Inventario de la Biodiversidad. Grupo de Exploraciones y
Monitoreo Ambiental GEMA, Bogot, D.C. (24 pp.).
IDEAM. Instituto de hidrologa, meteorologa y estudios ambientales, 1998b.
Geomorfologa del litoral caribe, morfo-dinmica y amenazas naturales. Santa f de
Bogot, Colombia.
IGAC, 1975. Estudio social aplicado de la alta y media Guajira. Ministerio de Hacienda y
crdito Pblico. Instituto Agustn Codazzi, Bogot (128 pp.).
INVEMAR, Instituto de Investigaciones Marinas y Costeras Jos Benito Vives de Andris,
2005o. Capacity building to improve adaptability to sea level rise in two vulnerable
points of the Colombian coastal areas (TumacoPacic coast and Cartagena
Caribbean coast) with special emphasis on human populations under poverty condi-
tions. Santa Marta DTHC, Colombia.
IPCC, 2012. Managing the risks of extreme events and disasters to advance climate change
adaptation. In: Field, C.B., Barros, V., Stocker, T.F., Qin, D., Dokken, D.J., Ebi, K.L.,
Mastrandrea, M.D., Mach, K.J., Plattner, G.-K., Allen, S.K., Tignor, M., Midgley, P.M.
(Eds.), A Special Report of Working Groups I and II of the Intergovernmental Panel
on Climate Change. Cambridge University Press, Cambridge, UK, and New York, NY,
USA (582 pp.).
IPCC, 2013. Climate change 2013: the physical science basis. In: Stocker, T.F., Qin, D.,
Plattner, G.-K., Tignor, M., Allen, S.K., Boschung, J., Nauels, A., Xia, Y., Bex, V.,
Midgley, P.M. (Eds.), Contribution of Working Group I to the Fifth Assessment Report
of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cam-
bridge, United Kingdom and New York, NY, USA (1535 pp.).
Jackson, J.B.C., Donovan, M.K., Cramer, K.L., Lam, V.V. (Eds.), 2014. Status and Trends of Ca-
ribbean Coral Reefs: 19702012. Global Coral Reef Monitoring Network, IUCN, Gland,
Switzerland (304 pp.).
Jentsch, A., Beierkuhnlein, C., 2008. Research frontiers in climate change: effects of ex-
treme meteorological events on ecosystems. C. R. Geoscience 340, 621628.
Knowlton, N., Lang, J.C., Rooney, M.C., Clifford, P., 1981. Evidence for delayed mortality in
hurricane-damaged Jamaican staghorn corals. Nature 294, 251252.
Kovacs, J.M., Malczewski, J., Flores-Verdugo, F., 2004. Examining local ecological knowl-
edge of hurricane impacts in a mangrove forest using an analytical hierarchy process
(AHP) approach. J. Coast. Res. 20, 792800.
Krauss, K.W., Lovelock, C.E., McKee, K.L., Lpez-Hoffman, L., Ewe, S.M.L., Sousa, W.P., 2008.
Environmental drivers in mangrove establishment and early development: a review.
Aquat. Bot. 89, 105127.
Leadbetter, M., 1991. On a basis for peaks over threshold modeling. Stat. Probab. Lett. 12
(4), 357362.
Lpez-Victoria, M., Zea, S., 2004. Storm-mediated coral colonization by an excavating Ca-
ribbean sponge. Clim. Res. 26, 251256.
Lpez-Victoria, M., Zea, S., Weil, E., 2006. Competition for space between encrusting exca-
vating Caribbean sponges and other coral reef organisms. Mar. Ecol. Prog. Ser. 312,
113121.
Lugo, A.E., 2008. Visible and invisible effects of hurricanes on forest ecosystems: an inter-
national review. Austral Ecology 33, 368398.
Mallela, J., Crabbe, M.J.C., 2009. Hurricanes and coral bleaching linked to changes in coral
recruitment in Tobago. Mar. Environ. Res. 68 (4), 158162. http://dx.doi.org/10.1016/
j.marenvres.2009.06.001.
Martnez, J., Smith, C., Richmond, R., 2012. Invasive algal mats degrade coral reef physical
habitat quality. Estuar. Coast. Shelf Sci. http://dx.doi.org/10.1016/j.ecss.2011.12.022.
Molina, E.C., 2009. Dinmica de los manglares de Baha Portete, alta Guajira, a escala de
paisaje y su relacin con variables climticas asociadas al cambio climtico global y
regional. Magster en Bosques y Conservacin Ambiental. Universidad Nacional de
Colombia, Sede Medelln (70 pp.).
Montoya, R.D., 2013. Study of the interactions between large scale climate phenomena
and waves in the Caribbean Sea: methodologies to correct wind speed during aver-
age and hurricane conditions for wave climate analysis (Ph.D. thesis) Univerisidad
Nacional de Colombia (340 pp.).
100 G. Bernal et al. / Journal of Marine Systems 164 (2016) 85100
Montoya, R.D., Osorio, A.F., 2014. Methodology to correct wind speed during average
wind conditions: application to the Caribbean Sea. J. Atmos. Ocean. Technol. 31,
19221945.
Montoya, R.D., Osorio, A.F., Ortiz-Royero, Ocampo-Torres, F.J., 2013. A wave parameters
and directional spectrum analysis for extreme winds. Ocean Eng. 67, 100118.
Moon, I.-J., Ginis, I., Hara, T., Tolman, H., Wright, C.W., Walsh, E.J., 2003. Numerical simu-
lation of sea surface directional wave spectra under hurricane wind forcing. J. Phys.
Oceanogr. 33, 16801706.
Nystrm, M., Graham, N.A.J., Lokrantz, J., Norstrm, A.V., 2008. Capturing the cornerstones
of coral reef resilience: linking theory to practice. Coral Reefs 27, 795809. http://dx.
doi.org/10.1007/s00338-008-0426-z.
Ortiz-Royero, J.C., 2007. Hurricanes and Tropical Storms in the Colombian Caribbean Sea
since 1900. Colombia, Scientic Bulletin. Center Oceanographic and Hydrographic Re-
search, Cartagena, pp. 5460 (ISSN: 0120-0542. CIOH Invemar v.25 colciencias fasc.
In spanish).
Ortiz-Royero, J.C., Otero, L.J., Restrepo, J.C., Ruiz, J., Cadena, M., 2013. Cold fronts in the Co-
lombian Caribbean Sea and their relationship to extreme wave events. Nat. Hazards
Earth Syst. Sci. 13, 27972804.
Osorio, A.F., Medina, R., Gonzalez, M., 2012. An algorithm for the measurement of shore-
line and intertidal beach proles using video imagery: PSDM. Comput. Geosci. (ISSN:
0098-3004).
Pawlik, J.R., 2011. The hemical ecology of sponges on Caribbean reefs: natural products
shape natural systems. Bioscience 61 (11), 888898.
Prez, J.C., Ortiz, C.A., Osorio, A.F., Meja, C.E., Medina, R., 2013. Environmental applications
of camera images calibrated by means of the LevenbergMarquadt method. Comput.
Geosci. 51, 7482 (ISSN: 0098-3004).
Restrepo, J., Alvarado, E., 2011. Los sedimentos de ro Magdalena y su impacto sobre los
arrecifes coralinos de las islas del Rosario: una prioridad de investigacin nacional.
In: Zarza-Gonzlez, E. (Ed.), El entorno ambiental del Parque Nacional Natural Corales
del Rosario y San Bernardo. Parques Nacionales Naturales de Colombia, Cartagena,
pp. 7894 (416 pp.).
Rodrguez-Ramrez, A., Garzn-Ferreira, J., 2003. Monitoreo de arrecifes coralinos, pastos
marinos y manglares en la Baha de Chengue (Caribe Colombiano): 19931999.
INVEMAR, Serie de Publicaciones Especiales No.8, Santa Marta, Colombia (170 pp.).
Rodrguez-Ramrez, A., Reyes-Nivia, M.C., 2008. Evaluacin rpida de los efectos del
huracn Beta en la Isla de Providencia (Caribe colombiano). Bol. Investig. Mar.
Costeras 37 (1), 217224.
Rodrguez-Ramrez, A., Bastidas, C., Rodrguez, S., Leo, Z., Kikuchi, R., Oliveira, M., Gil, D.,
Garzn-Ferreira, J., Reyes-Nivia, M.C., Navas-Camacho, R., Santodomingo, N., Daz-
Pulido, G., Venera-Ponton, D., Florez-Leiva, L., Rangel-Campo, A., Orozco, C.,
Mrquez, J.C., Zea, S., Lpez-Victoria, M., Snchez, J.A., Hurtado, M.C., 2008. The effects
of coral bleaching in southern tropical America: Brazil, Columbia, and Venezuela. In:
Wilkinson, C., Souter, D. (Eds.), Status of Caribbean Coral Reefs after the Bleaching
and Hurricanes of 2005. Global Coral Reef Monitoring Network, and Reef and
Rainforest Research Centre, Ownswille, pp. 111119 (152 pp.).
Rodrguez-Ramrez, A., Reyes-Nivia, M.C., Zea, S., Navas-Camacho, R., Garzn-Ferreira, J.,
Bejarano, S., Herrn, P., Orozco, C., 2010. Recent dynamics and condition of coral
reefs in the Colombian Caribbean. Rev. Biol. Trop. 58 (Suppl. 1), 107131.
Romero-Rodrguez, D., Bernal, G., Zea, S., 2014. Variables ambientales durante el
blanqueamiento coralino en el Caribe colombiano. Rev. Acad. Colomb. Cienc. Exctas
Fs. Nat. 38 (149), 347355.
Snchez, J.A., Herrera, S., Navas-Camacho, R., Herrn, P., Pizarro, V., Acosta, A., Castillo, P.,
Montoya, P., Orozco, C., 2010. White plague-like coral disease in remote reefs of the
Western Caribbean. Rev. Biol. Trop. 58, 145154.
Snchez-Pez, H., lvarez, L.R., Pinto, N.F., Snchez, A.A., Pino, R.J., Garca, H.I., Acosta, P.M.,
1997. Diagnstico y zonicacin preliminar de los manglares del Caribe de Colombia.
Ministerio de Medio Ambiente, Santa Fe de Bogot D.C., Colombia (511 pp.).
Santodomingo, N., Rodriguez-Ramirez, A., Garzon Ferreira, J., 2002. Territorios del pez
Stegastes planifrons en formaciones coralinas del Parque Nacional Natural Tayrona,
Caribe Colombiano: un panorama general. Bol. Investig. Mar. Costeras 31, 6584.
Tanner, E.V.J., Kapos, V., Healey, J.R., 1991. Hurricane effects on forest ecosystems in the
Caribbean. Biotropica 23, 513521.
The MerMex Group, 2011. Marine ecosystems' responses to climatic and anthropogenic
forcings in the Mediterranean. Prog. Oceanogr. 91, 97166.
Tolman, H.L., 2002. User manual and system documentation of WAVEWATCH-III version
2.22. NOAA / NWS / NCEP / MMAB.
Tolman, H.L., 2009. User manual and system documentation of WAVEWATCH III version
3.14. NOAA/NWS/NCEP/MMAB Technical Note 276 (194 pp.).
Van Woesik, R., 1991. Immediate impact of the January 1991 oods on the coral assem-
blages of the Keppel Islands. Great Barrier Reef Mar. Park Authority Res. Publ. 23
(30 pp.).
Vega-Sequeda, J., Navas-Camacho, R., Gmez-Campo, K., Lpez-Londoo, T., Duque, D.L.,
2011. Estado del conocimiento de los arrecifes coralinos. En: INVEMAR. Informe del
Estado de los Ambientes Marinos y Costeros en Colombia: Ao 2010. Ser. Publ. Perid.
8, 88119.
Wang, C., 2007. Variability of the Caribbean low-level jet and its relations to climate. Clim.
Dyn. 29, 411422. http://dx.doi.org/10.1007/s00382-007-0243-z.
Webster, P.J., Holland, G.J., Curry, J.A., Chang, H.R., 2005. Changes in tropical cyclone num-
ber, duration, and intensity in a warming environment. Science 309 (5742),
18441846.
Wilkinson, C., 2008. Status of coral reefs of the world: 2008. Global Coral Reef Monitoring
Network and Reef and Rainforest Research Centre, Townsville, Australia (296 pp.).
Willoughby, H.E., 2004. Parametric representation of the primary vortex. Part I: observa-
tions and evaluation of the Holland (1980) Model. Mon. Weather Rev. 30333048.
Wright, C.W., Walsh, E.J., Vandemark, D., Krabill, W.B., Garcia, A.W., Houston, S.H., Powell,
M.D., Black, P.G., Marks, F.D., 2001. Hurricane directional wave spectrum spatial var-
iation in the open ocean. J. Phys. Oceanogr. 31, 24722488.
Xu, F., Perrie, W., Toulany, B., Smith, P.C., 2007. Windgenerated waves in Hurricane Juan.
Ocean Model. 16, 188205.
Zea, S., Geister, J., Garzn-Ferreira, J., Daz, J.M., 1998. Biotic changes in the reef complex of
San Andrs Island (Southwestern Caribbean Sea, Colombia) occurring over nearly
three decades. Atoll Res. Bull. 456, 130.
Zhuo, L., Wang, A., Guo, P., 2008. Numerical simulation of sea surface directional wave
spectra under typhoon wind forcing Export. J. Hydrodyn. Ser. B 20 (6), 776783.