Content uploaded by Christoph A. Rohner
Author content
All content in this area was uploaded by Christoph A. Rohner
Content may be subject to copyright.
MARINE ECOLOGY PROGRESS SERIES
Mar Ecol Prog Ser
Vol. 482: 153–168, 2013
doi: 10.3354/meps10290 Published May 22
INTRODUCTION
Planktivorous elasmobranchs comprise 14 species
of mainly large sharks and rays from 5 genera. This
group includes the 3 focal species of this study: the
whale shark Rhincodon typus, reef manta ray Manta
alfredi and giant manta ray M. birostris. While they
differ in some aspects of their ecology, they all feed
mainly on zooplankton (Stevens 2007, Nakaya et al.
2008, Sims 2008, Cou turier et al. 2012). This reliance
on zooplankton strongly links sightings of planktivo-
rous elasmobranchs to environmental variables, as
© Inter-Research 2013 · www.int-res.com*Email: c.rohner1@uq.edu.au
Trends in sightings and environmental influences
on a coastal aggregation of manta rays and
whale sharks
C. A. Rohner1,2, 3,*, S. J. Pierce1, 4, 5, A. D. Marshall1, 5, S. J. Weeks2, M. B. Bennett6,
A. J. Richardson3, 7, 8
1Manta Ray & Whale Shark Research Centre, Marine Megafauna Foundation, Praia do Tofo, Inhambane, Mozambique
2Biophysical Oceanography Group, School of Geography Planning and Environmental Management,
The University of Queensland, St Lucia, Queensland 4072, Australia
3CSIRO Marine and Atmospheric Research, EcoScience Precinct, Dutton Park, Queensland 4102, Australia
4All Out Africa Research Unit, PO Box 153, Lobamba, Swaziland
5ECOCEAN USA, Praia do Tofo, Inhambane, Mozambique
6School of Biomedical Sciences, The University of Queensland, St Lucia, Queensland 4072, Australia
7Centre for Applications in Natural Resource Mathematics (CARM), School of Mathematics and Physics,
The University of Queensland, St Lucia, Queensland 4072, Australia
8The Ecology Centre, School of Biological Sciences, The University of Queensland, St Lucia, Queensland 4072, Australia
ABSTRACT: Sightings of planktivorous elasmobranchs at their coastal aggregation sites are often
linked to biological, environmental and temporal variables. Many large planktivorous elasmo-
branchs are also globally threatened species, so it is necessary to try and separate population
trends from environmentally driven, short-term fluctuations. We investigated the influence of
environmental variables on sightings of 3 species of planktivorous elasmobranchs off Praia do
Tofo, Mozambique: the reef manta ray Manta alfredi, giant manta ray M. birostris and whale
shark Rhincodon typus. We used 8- (2003 to 2011) and 6-yr (2005 to 2011) logbook data for manta
rays and whale sharks, respectively, and constructed a generalised linear model with animal sight-
ings as the response. Predictors included temporal (year, month, time of day), biological (plankton
categories), oceanographic (water temperature, time from high tide, current direction and strength
and wave height) and celestial (moon illumination) indices. These predictors best fitted reef manta
ray sightings, a coastal species with high residency, but less so for the wider-ranging giant manta
rays and whale sharks. We found a significant decline in the standardised sightings time series for
the reef manta ray (88%) and whale shark (79%), but not for the giant manta ray.
KEY WORDS: GLM · Generalised linear model · Decline · Population trend · Seasonality ·
Environmental variability · Manta alfredi · Manta birostris · Rhincodon typus
Resale or republication not permitted without written consent of the publisher
FREEREE
ACCESSCCESS
Mar Ecol Prog Ser 482: 153–168, 2013
154
the distri bution and abundance of their prey respond
rapidly to environmental changes (Boucher et al.
1987, Richardson 2008). By contrast, sightings of
higher trophic level species are likely to be less
directly influenced by physical factors due to gradual
weakening of the immediate environmental signals
with each step up the food chain.
Manta rays and whale sharks have a tendency to
aggregate on a periodic or seasonal basis at specific
sub-tropical and tropical locations around the world
(Stevens 2007, Kashiwagi et al. 2011). Foraging is the
main driver behind some of these aggregations (Wil-
son et al. 2001, Anderson et al. 2011), although manta
rays also aggregate at ‘cleaning stations’ on reefs
where small fishes remove parasites and dead tissue,
or at inshore areas to mate at specific times of the
year (Marshall & Bennett 2010a,b). These predic table
aggregations have, in some areas, been targeted by
fishers or tourism operators (Pravin 2000, Alava et al.
2002). All 3 species in this study are now globally
listed as Vulnerable on the IUCN Red List of Threat-
ened Species following significant declines in
catches or sightings that have mainly been attributed
to targeted fisheries (Norman 2005, Marshall et al.
2011a,c).
The influence of environmental variability on
sightings of these highly mobile fishes can compli-
cate the interpretation of trend data. For instance, a
decline in whale shark sightings along the eastern
coast of South Africa between 1993 and 1998 (Gifford
2001) was included in the successful proposal for list-
ing the species on Appendix II of the Convention on
International Trade in Endangered Species (CITES
2002). However, with the benefit of hindsight, the
substantial variability in sighting data from this area
in summer months (Cliff et al. 2007) and seasonal
changes in their oceanic distribution (Sequeira et al.
2012) makes it difficult to attribute these earlier data
to a genuine population decline. Even with compara-
tively robust datasets, such as for whale sharks at
Ningaloo Reef in Australia, scientific consensus on
population trends has proven difficult to attain due to
differences in the underlying assumptions of the
models, insufficient data to test which approach is
more robust, and the potential influence of environ-
mental variability (Wilson et al. 2001, Meekan et al.
2006, Bradshaw et al. 2007, 2008, Holmberg et al.
2008, 2009, Sleeman et al. 2010a). In another plank-
tivorous elasmobranch, the basking shark Cetorhi-
nus maximus, a steep decline in catches in a target
fishery off Achill Island in Ireland between 1947 and
1975 was initially attributed to localised stock deple-
tion (Parker & Stott 1965). Re-analysis of these data in
conjunction with concurrent planktonic copepod
abundances at this and other sites led Sims & Reid
(2002) to suggest that the decrease may have
resulted from a shift in prey, and hence basking
shark distribution, rather than strictly due to popula-
tion decline from fishing (Sims & Reid 2002, Sims
2008).
Previous investigations on how environmental
variables affect the sightings of planktivorous elas-
mobranchs have focused primarily on temperature.
Whale shark numbers in pelagic surface waters of
the western Indian Ocean have been shown to
correlate with sea surface temperature (SST)
(Sequeira et al. 2012). In coastal waters, SST and
wind speed have also been correlated with whale
shark numbers in an aerial survey in the Seychelles
(Rowat et al. 2009) and SST, chlorophyll a(chl a)
concentration and bathymetry were important pre-
dictors of whale shark and manta ray abundance
off Ningaloo Reef, Australia (Sleeman et al. 2007).
At Komodo Marine Park, Indonesia, the vast
majority of reef manta rays have been recorded in
water temperatures from 25 to 28°C and when
tidal intensity was highest during new and full
moons (Dewar et al. 2008). Further, large-scale
factors influencing oceano graphic processes, such
as the Southern Oscillation Index and wind shear
at Ningaloo Reef, or monsoon winds in the Mal-
dives, have been linked to shifts in abundance and
distribution of planktivorous elasmobranchs (Wilson
et al. 2001, Sleeman et al. 2010a, Anderson et al.
2011). These and other environmental factors thus
play a role in explaining sighting trends of manta
rays and whale sharks. As planktivorous elasmo-
branchs are often most threatened at their coastal
aggregation sites due to their exposure to shore-
based fishing and boating activities, it is important
to understand drivers of these aggre gations and to
determine whether observed fluc tuations in sight-
ings correspond to short-term environmental vari-
ability or whether they represent long-term trends.
Standardising sightings data and taking into
account environmental covariates are thus poten-
tially important steps in the conservation manage-
ment of these species.
Here, we aimed to (1) estimate trends in standard-
ised sightings over time, and (2) define the relation-
ships of environmental, spatial and temporal vari-
ables on sightings of reef manta rays, giant manta
rays and whale sharks at a coastal aggregation site in
southern Mozambique. We used a generalised linear
model for each species to determine the relative im -
portance of each environmental variable. We hypo -
Rohner et al.: Environmental influences on planktivorous elasmobranchs
thesise that sightings of these plantkivorous fishes
are influenced by specific environmental variables
and that resident reef manta rays are influenced
more strongly by local variables than are transient
giant manta rays and whale sharks.
MATERIALS AND METHODS
Location
The study was conducted off Praia do Tofo
(23.85° S, 35.54° E), Inhambane Province, southern
Mozambique (Fig. 1). All trips were conducted in a
40 km stretch of coastal waters. Manta rays aggre-
gate here mainly for cleaning, and they can spend up
to 8 h d−1 in this activity (Marshall 2009, van Duin -
kerken 2010). We used SCUBA to observe them at
cleaning stations on reefs primarily at Sites 1 and 2.
Less frequently surveyed reefs were combined into
3 ‘northern reefs’, 3 ‘southern reefs’ and 7 ‘central
reefs’. Whale sharks were spotted from boats, with
search effort concentrated in coastal waters within
8 km south of Praia do Tofo (Fig. 1).
Logbook data
Manta rays
Manta ray logbook data were recorded by A. D.
Mar shall. The dataset comprised 855 dives spanning
8 yr (May 2003 to April 2011). Sightings of reef and
giant manta rays were measured as individuals per
dive (‘raw sightings’ henceforth). In terms of sam-
pling effort, data on dive duration (to standardise
each dive over the past 8 yr) was not available. We
thus performed a pilot study to assess the influence
of dive duration on manta ray sightings. The pilot
study included 102 dives over 4 mo in 2012. Dives
had a mean duration of 39.3 min and a relatively
small SD of 7.67 min. We found no significant rela-
tionship be tween number of manta rays (response
variable) and duration of dive (ANOVA; reef manta
ray: F= 1.25, p = 0.265; giant manta ray: F= 3.07,
p = 0.085). This is likely to be because searches
across the reef consistently took the same track over
the study period, that manta rays spend most of the
day at cleaning stations, and that they are large and
conspicuous. Spatial autocorrelation is unlikely to
be a major issue for manta rays because the total
area of cleaning stations where manta rays congre-
gate and divers searched for them is small (0.5 km2).
Manta rays also show some site fidelity to particular
cleaning stations over a period of several hours
(Marshall 2009) and it is unlikely that the same
manta ray will be seen at different sites on closely
spaced dives on the same day. Photographic identi-
fication allowed individual animals to be distin-
guished (Marshall et al. 2011b) and those that were
spotted multiple times in the same dive were
counted once only. Nine predictors were recorded
in the logbook, in addition to sightings of reef and
giant manta rays (Table 1).
To account for temporal influences on manta ray
sightings, year, month and time of day were
recorded for each dive. Dive site was included as a
predictor in the manta ray models because manta
rays were ob served at 5 different sites. This spatial
predictor was used to investigate preferences for
certain reefs. As a general index of food availability,
the major planktonic component during the dive
was visually assessed underwater and grouped into
5 categories: (1) zooplankton (i.e. dense visible zoo-
plankton, such as copepods, pteropods or gelatinous
zooplankton present in the water), (2) phytoplank-
ton (i.e. visibly green water), (3) no obvious plank-
ton, (4) suspended sediments (i.e. stirred up sediment
reducing visibility) and (5) surface slicks (i.e. plank-
ton forming a slick visible only near the surface). As
water temperature has previously been correlated
with planktivorous elasmobranchs (Dewar et al.
2008, Rowat et al. 2009, Sequeira et al. 2012), we in -
cluded in situ temperature (1°C intervals) recorded at
depth on a dive computer. The same computer was
155
Inhambane
Praia do Tofo
10 km50
Northern reefs
Southern reefs
Central reefs
Site 2
Site 1
Whale shark search area
Fig. 1. Study location at Praia do Tofo (23.85°S, 35.54° E) in
southern Mozambique, indicating the dive sites where
manta ray sightings were recorded (thin dashed areas) and
the survey area for whale sharks sightings (thick dashed
area)
Mar Ecol Prog Ser 482: 153–168, 2013
used consistently, which provided the same readings
as a temperature recorder (Onset Hobo Pro v2 with
±0.2°C accuracy) in sporadic assessments. To inves-
tigate the influence of physical environmental con-
ditions on manta ray behaviour, current direction
(N, E, S, W or absent) and strength (absent, light,
medium, strong) were assessed underwater visually
in relation to the reef. Underwater visibility,
visually estimated horizontally at depth in 1 m
increments, was in cluded to account for potential
sightability effects. Moon phase has been shown to
influence biorhythms in other fishes (Colin 1992,
Agenbag et al. 2003, Heyman & Kjerfve 2008) and
hence moon illumination data were downloaded for
the location from NASA’s Horizons website (http://
ssd. jpl. nasa. gov/ horizons.cgi). Tidal phase and
amplitude have also been shown to affect some
large planktivores (Dewar et al. 2008) and tide data
for Inhambane were obtained from Mobile Geo-
graphics (www. mobile geographics.com).
Whale sharks
Whale shark logbook data were recorded by S. J.
Pierce and C. A. Rohner. Observations were made
during 720 trips between January 2005 and April
2011. Whale shark numbers were measured as indi-
viduals sighted per boat trip (‘raw sightings’ hence-
forth). Similar to the methods of Pierce et al. (2010),
searches for whale sharks were conducted from 6.5
to 8.2 m long rigid-hull inflatable vessels, often with
the addition of a raised spotting chair to enlarge the
search area. In terms of sampling effort, most (95%)
trips lasted 2 hours (±15 minutes) because of com-
mercial operational constraints. On each trip, we
performed a zig-zag search pattern between the
beach and 1 km offshore, on the south- and north-
ward passes within the ‘whale shark area’ from
Praia do Tofo to 8 km south (Fig. 1). One ‘spotter’
constantly scanned the surface for whale sharks.
‘Site’ was not included in the whale shark model
because all whale shark observations were made at
one site. Spatial autocorrelation is unlikely to be a
major issue for whale sharks because the area was
small (~8 km2) and was searched in a consistent
way each trip. Individual whale sharks were identi-
fied using under water photographs of the body
region immediately posterior to the gills (Arzou-
manian et al. 2005). Those that were spotted multi-
ple times in the same trip were counted once only,
although there may be some error associated with
aggregations of >10 sharks (2.4 % of observations)
where not all individuals could be photo graphed. To
account for temporal effects, year, month and time
of day were recorded. Moon illumination and tide
data were downloaded as described in the manta
ray dataset methods. However, as in situ observa-
tional plankton data were not recorded for this part
of the study, we used remotely-sensed chl aas a
proxy for zooplankton in the model. Despite a possi-
ble lag time of zooplankton abundance in response
to a phytoplankton bloom (Plourde & Runge 1993,
Flagg et al. 1994), phyto- and zooplankton abun-
dance is often correlated (Hutchinson 1967, Richard-
son & Schoeman 2004, Ware & Thomson 2005) and
chl ahas been used to explain occurrence and
movements of marine animals in various studies
(Zagaglia et al. 2004, Block et al. 2011), including
some on planktivorous elasmobranchs (Sleeman et
al. 2007, Graham et al. 2012). Similarly, satellite-
derived SST was used in the absence of in situ tem-
perature recordings. Chl aand SST data at 1 km
resolution were derived from the Moderate Re -
solution Imaging Spectroradiometer (MODIS; http://
modis. gsfc. nasa. gov). These data were composited
to 5 d means for the region and extracted for the
location off Praia do Tofo (23.893° S, 35.578° E). Due
to cloud contamination, minor temporal interpola-
tion was applied (<5%). Surface conditions can
affect where and how well whale sharks are visible
(Rowat et al. 2009) and we therefore included wave
direction and height and wind direction and speed
in the initial model. These data were acquired from
Buoyweather’s hindcast model at 23.85°S, 35.7° E,
~15 km off Praia do Tofo (www.buoyweather.com).
Model
A negative binomial generalised linear model
(GLM) with natural splines was constructed using
the statistical software R (v.2.13.0; www.r-project.
org), with raw sightings set as the response and the
aforementioned suite of local variables as predictors
(Table 1). Model residuals were visually assessed
for normality and homogeneity of variance and an
un transformed response with a normal error struc-
ture was found to be appropriate. A generalised
additive model (GAM; in the mgcv R-package) was
used to explore the relationships and define the
degrees of freedom (df) to be used for smooth terms
in the GLM. This was conservatively set at df = 3
for all smooth terms, unless the GAM plots indi-
cated that df = 4 or 5 more appropriately defined a
relationship. The best model was conservatively
156
Rohner et al.: Environmental influences on planktivorous elasmobranchs 157
Variable Parameter Explanation Type Units Mean (±SD)
Manta rays
Response RM sightings Number of individual reef manta rays per dive 2.7 (4)
GM sightings Number of individual giant manta rays per dive 0.41 (1.07)
Predictors Plankton category Visual assessment of plankton at depth Categorical PP = phytoplankton, SS = suspended
sediments, ZP = zooplankton,
SU = surface slicks, none
Moon illuminationaProportion of moon disk illuminated Continuous Proportion 0.5 (0.36)
Water temperature Water temperature on the reef (20–30 m depth) Continuous °C 24.68 (2.17)
Current direction Current direction Categorical Levels: N, E, S, W or no current
Current strength Current strength Categorical Levels: no current, light, medium, strong
Tidal rangebDifference between high and low tides on the day Continuous m 1.87 (0.73)
Time from high tidebTime difference between observation & closest high tide Continuous h 0.66 (3.8)
Visibility Estimated underwater visibility Continuous m 10.69 (5.36)
Dive site Reef or group of reefs Categorical Levels: Site 1, Site 2, southern,
northern and central reefs
Month Month of observation Categorical mo
Year Year of observation Categorical yr
Time Time of observation Continuous hh 10.22 (1.67)
Whale sharks
Response WS sightings Number of individual whale sharks per trip 2.28 (3.37)
Predictors Chl acChl aconcentration (5 d mean) Continuous mg m−3 0.22 (0.12)
Moon illuminationaProportion of moon disk illuminated Continuous Proportion 0.48 (0.35)
Wind directiondWind direction Continuous Degrees 115.03 (69.58)
SSTcSea surface temperature (5 d mean) Continuous °C 26.45 (1.83)
Tidal rangebDifference between high and low tides on the day Continuous m 1.89 (0.73)
Time from high tidebTime difference between observation & closest high tide Continuous h –0.38 (4.09)
Wave directiondWave direction Categorical Levels: N, NE, E, SE, S, SW, W, NW
Wave heightdMinimal wave height of the day Continuous ft 5.18 (1.73)
Month Month of observation Categorical mo
Year Year of observation Categorical yr
Time Time of observation Continuous hh 12.52 (0.78)
aNASA Horizons, bMobile Geographics, cNASA’s MODIS satellite, and dBuoyweather
Table 1. Predictors used in the generalised linear model for manta rays and whale sharks. RM: reef manta ray Manta alfredi; GM: giant manta ray M. birostris; WS:
whale shark Rhincodon typus
Mar Ecol Prog Ser 482: 153–168, 2013
assessed using a stepwise Akaike’s Information Cri-
terion (AIC) function with the default penalty per
parameter set at k = 2 and a dropterm χ2function
was performed on the AIC- supported model. The
significance of each predictor on modelled stan-
dardised sightings (‘sightings’ hence forth) in the
final model after the AIC selection process was
assessed using a χ2test (Venables & Ripley 2002). In
the model output figures, the y-axis is a relative
scale, so that a y-value of zero is the mean effect of
the adjusted predictor on the response, a positive y-
value indicates a positive effect on the response,
and a negative y-value indicates a negative effect
on the response. If a horizontal line can be placed
between the 95% confidence limits (dotted lines),
this implies that the relationship between the
response and the predictor is not significant. These
lines tend to diverge near the extremes of the range
for continuous predictors as a consequence of fewer
observations.
Correlations between continuous predictors were
assessed with the Pearson method (cut-off = 0.25),
and there were no significant correlations between
any predictors. Post-hoc Tukey tests were used to
perform multiple comparisons among levels of cate-
gorical predictors. To estimate % change in mega -
fauna sightings through time, rather than using esti-
mates from the GLM for the last and first yr which are
dependent on 2 yr only, we fitted a line through the
annual model estimates. Levels of categorical predic-
tors and values of continuous predictors that resulted
in highest animal sightings were used for each pre-
dictor. A negative exponential fitted best for all spe-
cies. We then estimated % change from the first to
last years from these lines of best fit and used an
ANOVA to calculate significance.
RESULTS
Raw sightings
Raw sightings per unit effort were highest for reef
manta rays (mean ± SE = 2.7 ± 0.14 ind. dive−1), fol-
lowed by whale sharks (2.3 ± 0.13 ind. trip−1) and
giant manta rays (0.4 ± 0.04 ind. dive−1). Inter-
annual variability in raw sightings was high over
the study period for all species (Fig. 2). Effort also
varied over time, with manta ray searches ranging
from 0 to 43 dives mo−1 and whale shark searches
ranging from 0 to 28 trips mo−1 (Fig. 2 insets). We
partially accounted for this bias by standardising
sightings with a GLM.
Overall model
The final GLM for reef manta rays explained
39.9% of the total variance (Table 2) and had an AIC
score of 3226 (model with all predictors: AIC =
3230.0). Ten predictors were retained including year,
month, dive site, water temperature, moon illumina-
tion, time from high tide, plankton categories, cur-
rent direction, current strength and time of day
(Fig. 3). The final GLM for giant manta rays
explained 30.3% of the total variance and had an
AIC score of 1260 (model with all predictors: AIC =
1283.5). Five predictors were retained for giant
manta rays (Fig. 4), including year, month, dive site,
water temperature and moon illumination. The final
GLM for whale sharks explained 23.9% of the vari-
ance and had an AIC score of 2749.7 (model with all
predictors: AIC = 2778.4). Four predictors were
retained for whale sharks (Fig. 5), including year,
month, time from high tide and wave height. The
year factor was the most significant predictor for reef
manta rays (χ2= 116.0, p < 0.001) and whale sharks
(χ2= 112.1, p < 0.001), accounting for 24.6 and 13.7%
of the variance, respectively. The year factor also
explained most (12.2%) of the variance for giant
manta rays, while dive site (9.1%) was the most sig-
nificant predictor (χ2= 55.0, p < 0.001). Other
strongly significant predictors included water tem-
perature (χ2= 29.8, p < 0.001) and dive site (χ2= 22.5
p < 0.001) for reef manta rays, month (χ2= 42.3, p <
0.001) for giant manta rays, and month (χ2= 68.6, p <
0.001) and wave height (χ2= 13.9, p = 0.003) for
whale sharks (Table 2).
Inter-annual trends
Annual standardised sightings from the predictive
model showed a steep decline over time for reef
manta rays (88%, ANOVA, F= 23.52, p = 0.002; 2003
to 2011) and whale sharks (79%, ANOVA, F= 7.39,
p = 0.042; 2005 to 2011), while sightings of giant
manta rays remained relatively stable (ANOVA, F=
1.90, p = 0.21).
Monthly variation
Month was a significant predictor in the GLM for
all 3 species. Reef manta ray sightings fluctuated
over the year (Fig. 3) but a post-hoc Tukey test
showed that sightings were only significantly higher
in January/February compared to April. Giant
158
Rohner et al.: Environmental influences on planktivorous elasmobranchs
manta rays showed a suggestive peak in sightings
during April (Fig. 4), which had significantly more
sightings than January, February, May, August,
October, No vem ber and December. Sightings of
whale sharks also fluctuated over the course of a
year, but showed no clear seasonality. The lower
numbers in November and December were sig -
nificant compared to 5 other months, and the peak
in February was significant compared to 4 other
months (Fig. 5).
Environmental variables
Water temperature and dive site
were strongly significant predictors
(χ2= 29.2 and 23.2, respectively,
both p < 0.001) for reef manta ray
sightings. Warmer waters had a pos-
itive effect on sightings. More ani-
mals were seen at Site 1 than at
Site 2, the northern reefs, or
southern reefs (Tukey test, p < 0.03).
Reef manta rays were sighted less
frequently in northward currents
compared to the predominant south-
ward currents, and also when cur-
rents were absent (p = 0.01 and 0.02,
respectively; Fig. 3). Sighting num-
bers were lower during strong cur-
rents, compared to medium currents
(p = 0.03), but this was marginally
and not significant compared to light
and no currents, respectively. More
reef manta rays were seen when
zooplankton was recorded com -
pared to clear or turbid water condi-
tions (p = 0.003 and 0.03, res -
pectively). The model indicated a
suggestive trend towards decreasing
sightings at ~09:30 h. Reef manta
rays were seen most often during the
second quarter of the lunar cycle
(~20 to 40% of the moon illumi-
nated). Time from high tide was a
significant predictor, but it is difficult
to interpret this relationship.
More giant manta rays were ob -
served at Site 1 than at Site 2 and
the northern reefs (Tukey test, p =
0.001 and 0.03, respectively; Fig. 4).
Water temperature was significant,
but had an inconclusive effect on
giant manta ray numbers. Sightings
were also higher during the first
quarter of the lunar cycle (~0 to 25% of the moon illu-
minated), although this predictor was only margin-
ally significant (χ2= 7.8, p = 0.051).
Wave height was the most significant non-temporal
predictor for whale shark sightings (χ2= 13.9, p =
0.003). More sharks were seen when waves were
small (Fig. 5). Sightings were at a consistently lower
level when waves were >2 m. Time from high tide
was also significant but it was a weak relationship
(χ2= 7.9, p = 0.048).
159
2004 2005 2006 2007 2008 2009 2010 2011
A
No. of reef manta rays dive–1
0
5
10
15
20
25
30
B
No. of giant manta rays dive–1
0
5
10
15
C
No. of whale sharks trip–1
0
5
10
15
20
25
30
35
40
2004 2005 2006 2007 2008 2009 2010 2011
2004 2005 2006 2007 2008 2009 2010 2011
0
10
20
30
40
Dives mo–1
2003 2004 2005 2006 2007 201020092008 2011
Manta ray dives
2003 2004 2005 2006 2007 201020092008 2011
0
5
10
15
20
25
Trip mo–1
Whale shark trips
Fig. 2. Manta alfredi, M. birostris and Rhincodon typus. Raw sightings of indi-
vidual (A) reef manta rays, (B) giant manta rays per dive, and (C) whale sharks
per trip, including dives/trips with no sightings (equal to zero). Monthly effort
is shown for manta rays (combined) and whale sharks in the insets
Mar Ecol Prog Ser 482: 153–168, 2013
DISCUSSION
Overall model
The GLM for reef manta rays retained more vari-
ables and explained more of the variance than the
models for giant manta rays and whale sharks. While
in this study, temporal predictors were significant for
all 3 species and the spatial (dive site) predictor only
used in the manta ray models was significant for
both Manta species, local environmental variables
ex plained sightings of reef manta rays better than for
giant manta rays and whale sharks. This may relate
to ecological differences within these planktivorous
elasmobranchs in terms of the influence local envi-
ronmental variables have on their sightings, but also
with regards to their respective time spent in our
study area and their respective sightability.
Reef manta rays were observed predominantly at
known cleaning stations (Marshall et al. 2011b),
which are also important sites for social behaviours
(Marshall & Bennett 2010b). Although reef manta
rays are regularly seen feeding around these reefs,
they are not feeding areas per se, but rather multi-
purpose areas where cleaning is the main behaviour
seen by observers. Various cleaning stations are
available in the study region and reef manta rays are
likely to select favourable environmental conditions
to clean in, such as warm water. Reef manta rays are
also more strongly resident to the study area than
either the giant manta ray or the whale shark (Mar-
shall 2009). Although coastal movements of ~500 km
have been recorded in eastern Australia (Couturier
et al. 2011), observations from the Hawaiian islands
suggest that relatively small geographic distances
over deep waters are not traversed by this species
(Deakos et al. 2011). High site residency of reef
manta rays has been observed at the current study
site (Marshall et al. 2011b) and elsewhere, including
Komodo Island, Indonesia (Dewar et al. 2008) and
Maui, Hawaii (Deakos et al. 2011). The relatively
high site residency of reef manta rays is likely to be a
major reason for why local variables were more
important drivers of their sightings compared to
sightings of giant manta rays and whale sharks.
In contrast to reef manta rays, giant manta rays and
whale sharks are wider-ranging species. While the
ecology of the giant manta ray remains poorly
known, their seasonal appearance in subtropical and
temperate areas such as northern New Zealand
(Duffy & Abbott 2003) and southern Brazil (Luiz et al.
2009), and their cleaning aggregations at isolated
seamounts surrounded by deep oceanic waters (Mar-
shall et al. 2009) suggest that they travel longer dis-
tances, tolerate cooler waters, and spend more time
in or over deeper water compared to reef manta rays.
Considering their offshore movements and less fre-
quent observation by divers off Praia do Tofo, giant
manta rays are thus likely to be less influenced by
local variables than reef manta rays.
Whale sharks also undertake large-scale oceanic
movements (Eckert & Stewart 2001, Hsu et al. 2007,
Rowat & Gore 2007, Sleeman et al. 2010b), e.g. one
whale shark that was satellite-tagged immediately
adjacent to the current study area travelled ~1200 km
in 87 days (Brunnschweiler et al. 2009). Individual
whale sharks at Praia do Tofo are generally seen for
up to a few days before they move elsewhere (S. J.
Pierce unpubl.). Similar to giant manta rays, their
sightings at coastal sites are therefore likely to be
driven more by broader-scale processes than local
conditions. Other studies support this hypothesis,
160
Predictors Reef manta ray Giant manta ray Whale shark
Variance (%) χ2p Variance (%) χ2p Variance (%) χ2p
Year 24.62 116.0 <0.001 12.16 62.6 < 0.001 13.65 112.1 <0.001
Month 5.60 31.5 0.001 6.30 42.3 < 0.001 7.89 68.6 < 0.001
Dive site 1.73 23.2 < 0.001 9.10 55.0 <0.001 – – –
Water temperature 1.68 29.2 < 0.001 1.60 11.1 0.026 – – –
Moon illumination 1.24 17.8 <0.001 1.13 7.8 0.051 – ––
Time from high tide 0.21 9.6 0.022 – – – 0.80 7.9 0.048
Plankton categories 1.69 16.5 0.003 – – – – – –
Current direction 1.76 15.2 0.004 – – – – – –
Current strength 0.72 7.3 0.063 – –– – – –
Time 0.71 10.5 0.061 – –– – – –
Wave height – – – – – – 1.54 13.9 0.003
Sum 39.9 30.3 23.9
Table 2. Percentage of variance explained by the GLM and significance values of a χ2test performed on the AIC-supported
models, including marginally significant values (p = 0.05−0.1) in italics
Rohner et al.: Environmental influences on planktivorous elasmobranchs 161
0
1
2
−6 −4 −2 0 2 4 6
Time from high tide
−4
−3
−2
−1
0
1
2
Year
2003 2005 2007 2009 2011
Partial (Year)
−4
−3
−2
−1
0
1
2
Month
1 2 3 4 5 6 7 8 9 10 11 12
Partial (Month)
−4
−3
−2
−1
0
1
2
Current direction
Partial (Current direction)
N E S W None
−4
−3
−2
−1
0
1
2
Current strength
Partial (Current strength)
StrongMediumLightNone
−4
−3
−2
−1
0
1
2
Plankton category
None SS PP ZP SU
Partial (Plankton category)
−4
−3
−2
−1
0
1
2
Site 1 Southern
Northern
Central
Site 2
Partial (Dive site)
18 20 22 24 26 28 30
−4
−3
−2
−1
0
1
2
Partial (Water temperature, df = 5)
Water temperature (˚C)
Partial (Time, df = 5)
6 8 10 12 14 16
−4
−3
−2
−1
0
1
2
Time (hh)
−4
−3
−2
−1
Partial (Moon illumination, df = 3)
0.0 0.2 0.4 0.6 0.8 1.0
Moon illumination
Partial (Time from high tide, df = 3)
−4
−3
−2
−1
0
1
2
Fig. 3. Manta alfredi. Reef
manta ray GLM model out-
puts showing the relation-
ship between their sightings
and all significant predic-
tors. The rug plot along the
x-axis indicates sampling
effort and dotted lines mark
the 95% confidence interval.
See Table 1 for details
Mar Ecol Prog Ser 482: 153–168, 2013
with sightings at Ningaloo Reef, Australia, partly
linked to the Southern Oscillation Index that influ-
ences the strength of the Leeuwin Current (Wilson et
al. 2001, Sleeman et al. 2010a).
Declining sightings trends for reef manta rays
and whale sharks
Although sightings of giant manta rays were rela-
tively stable over the study period, there was a signif-
icant decline in reef manta ray (88%) and whale
shark (79%) sightings. By including various temporal
and environmental variables in the GLM, the stan-
dardised sightings trend was adjusted for these vari-
ables. Although other local biophysical predictors —
not included in the GLM — may also influence
sightings, external anthropogenic pressures or lar -
ger-scale oceanographic influences are likely to be
involved. Reef manta rays are actively fished in
southern Mozambique (Marshall et al. 2011b), with
an estimated 20 to 50 individuals killed by fishers
annually along the 50 km of coastline encompassing
the present study area (A. D. Marshall unpubl.). This
reef manta ray population is naturally small, as
demonstrated in the superpopulation estimate of 802
162
0.0 0.2 0.4 0.6 0.8 1.0
−2
−1
0
1
2
Moon illumination
1
2
−2
−1
0
1
2
Year
2003 2005 2007 2009 2011
−2
−1
0
1
2
Month
1 2 3 4 5 6 7 8 9 10 11 12
Partial (Year)
Partial (Month)
−2
−1
0
1
2
18 20 22 24 26 28 30
−2
−1
0
1
2
Dive site
Site 1 Southern Northern Central
Site 2
Partial (Dive site)
Partial (Water temperature, df = 4)
Water temperature (˚C)
Partial (Moon illumination, df = 3)
Fig. 4. Manta birostris. Giant manta ray GLM model outputs
showing the relationship be tween their sightings and all sig-
nificant predictors. The rug plot along the x-axis indicates
sampling effort and dotted lines mark the 95% confidence
interval
Rohner et al.: Environmental influences on planktivorous elasmobranchs
individuals in our study area, with annual population
estimates ranging from 149 to 454 rays between 2003
and 2007 (Marshall et al. 2011b). As this species has
a conservative life history strategy with a low level of
recruitment, fishing pressure is likely to substantially
impact this population (Marshall & Bennett 2010b,
Marshall et al. 2011b). Some of the reefs in the study
area have also been increasingly subjected to
SCUBA diving ecotourism pressure, with up to 11
regular commercial scuba diving operators, most
with multiple boats, active in 2011. Only 4 operators
were active at the start of our study period, and the
steady increase in diver traffic could interrupt clean-
ing activities and otherwise disturb the manta rays.
Cleaning is an important daily activity and reef
manta rays spend particularly long periods at clean-
ing stations in the present study area (Marshall
2009). Approximately 76% of individuals bear bite
wounds of predatory sharks, and removal of dead tis-
sue around injuries is thought to facilitate wound
healing and prevent secondary infection (Marshall &
Bennett 2010a). Giant manta rays are less affected by
bite injuries (~35%; Marshall 2009) and visit cleaning
stations less regularly. As they spend less time at
these reefs, they are likely to be less affected by the
recreational diving activities that focus on these sites.
Further, only one giant manta ray has been observed
to be caught in the local fishery over the past 8 years
(A. D. Marshall pers. obs.), suggesting that fishing
pressure on this species is low in comparison to the
reef manta ray.
A significant decline in whale shark sightings was
also observed. There is no present-day fishery for
whale sharks in Mozambique, although occasional
catches have taken place in the local area during the
study period (S. J. Pierce unpubl.). Tuna purse-seine
fishers spot whale sharks in the Mozambique Chan-
nel (Sequeira et al. 2012) and their accidental catches
could have negative impacts on the sharks (Floch et
al. 2012). Also, some live sharks at Praia do Tofo bear
scars attributable to net entanglement and propeller
strikes (Speed et al. 2008). Given that whale sharks
are capable of long-distance movements, it is plausi-
ble that negative human impacts elsewhere could
affect sightings at Praia do Tofo, as has been sug-
gested for this species at Ningaloo Reef, Australia
(Bradshaw et al. 2007, 2008) and in the Andaman
Sea, Thailand (Theberge & Dearden 2006). Initial
findings from the global whale shark database have
shown little connectivity among distant aggregation
163
−2.0
−1.5
−1.0
−0.5
0.0
0.5
1.0
2005 2006 2007 2008 2009 2010 2011
−2.0
−1.5
−1.0
−0.5
0.0
0.5
1.0
1 2 3 4 5 6 7 8 9 10 11 12
2 4 6 8 10
−2.0
−1.5
−1.0
−0.5
0.0
0.5
1.0
Wave height (ft)
−6 −4 −2 0 2 4 6
−2.0
−1.5
−1.0
−0.5
0.0
0.5
1.0
Year Month
Partial (Year)
Partial (Month)
Partial (Wave height, df = 3)
Time from high tide (h)
Partial (Time from high tide, df = 3)
Fig. 5. Rhincodon typus. Whale shark GLM model outputs showing the relationship between their sightings and all significant
predictors. The rug plot along the x-axis indicates sampling effort and dotted lines mark the 95 % confidence interval
Mar Ecol Prog Ser 482: 153–168, 2013
sites in the Indian Ocean (Brooks et al. 2010, ECO-
CEAN 2012), and no link between sharks from Praia
do Tofo and areas with active whale shark fisheries
(such as Taiwan). However, the limited knowledge of
whale shark movements in the Western Indian
Ocean preclude a full assessment of this possibility
(Gifford et al. 2007, Brunnschweiler et al. 2009,
Brooks et al. 2010). Alternatively, regional- to large-
scale oceanographic parameters, not ac coun ted for in
our models, may also influence local whale shark
sightings, as has been shown for Ningaloo Reef, Aus-
tralia (Wilson et al. 2001, Sleeman et al. 2010a). Of
particular importance may be the meso scale eddies
that dominate the flow of the Mozambique Channel
(de Ruijter et al. 2002). These eddies exhibit high
inter-annual variability (Schouten et al. 2003) and
create strong fronts often associated with high bio-
mass (Lima et al. 2002, Tew Kai & Marsac 2010), sug-
gesting that further investigation of mesoscale eddies
will be important for understanding the strong inter-
annual signal in whale shark sightings at Praia do
Tofo.
Is megafauna visitation seasonal?
The model results suggested that the highest giant
manta ray sightings were in April, although the 95%
confidence limits do overlap with March and July.
Such a pattern may imply an annual migration past
Praia do Tofo during the austral autumn/winter, al-
though additional sightings of giant manta rays were
made throughout the year. Sightings of giant manta
rays at other aggregation sites, such as south-eastern
Brazil (Luiz et al. 2009) and New Zealand (Duffy &
Abbott 2003), similarly are seasonal and indicate that
this species undertakes seasonal movements.
Reef manta rays were sighted throughout the year
without exhibiting a clear seasonal signal. The model
suggested higher sightings in the austral summer,
although numbers from January/February were only
significantly higher than April. More sightings in the
austral summer are likely as this is when they mate
and give birth in the study area (Marshall & Bennett
2010b).
Whale sharks appear seasonally at most of their
coastal aggregation sites around the world, with their
visitation normally coinciding with increased prey
availability (Taylor 1996, Nelson & Eckert 2007, Riley
et al. 2010, de la Parra Venegas et al. 2011, Rowat et
al. 2011). The lack of a seasonal signal in whale shark
sightings in this study indicates a departure from this
general rule. Despite their non-seasonal appearance
at Praia do Tofo, whale shark sightings fluctuated
during the study. This is likely to be a consequence of
their movement patterns and the variable local zoo-
plankton abundance influenced by the dynamic
regional oceanography dominated by event-scale,
non-seasonal eddies (Schouten et al. 2003).
Environmental influences on megafauna sightings
Water temperature was a strongly significant pre-
dictor for reef manta rays, less so for giant manta
rays, and SST was removed from the final model for
whale sharks. Reef manta rays appeared to avoid
cold water (18 to 21°C), while giant manta rays had
an unclear relationship with water temperature. The
distribution of these 2 species overlap in parts of their
range, but giant manta rays extend further into
colder temperate areas (Kashiwagi et al. 2011, Cou-
turier et al. 2012), which could explain why tempera-
ture is not as strong a predictor of their presence as it
is for reef manta rays. Water temperatures on the
reefs off Praia do Tofo fluctuated up to 7.5°C d−1
(C. A. Rohner pers. obs.), with cool-water intrusions
re corded in all seasons. Reef manta rays preferen-
tially se lec ted episodes of warmer water irrespective
of season. Water temperature was not supported for
the whale shark model. Whale sharks can tolerate
cold waters, as they have a wide distribution extend-
ing into areas with cold surface waters (Turnbull &
Randell 2006) and can undertake deep dives into
water as cold as 3.4°C (Brunnschweiler et al. 2009).
In the western Indian Ocean, 90% of whale shark
sightings have been suggested to occur in a narrow
surface temperature band (26.5 to 30°C) (Sequeira et
al. 2012). SST is likely to strongly influence whale
shark sightings offshore, where fronts are a major
driver of the spatial dynamics of predators (Zainud-
din et al. 2008, Tew Kai & Marsac 2010). The lack of
a relationship with SST in our study was either be -
cause such oceanic fronts may have less influence on
coastal surface waters, or because remotely sensed
SST data was inadequate for our purpose.
Site 1 was the preferred reef for both manta ray
species, while Site 2 was also often visited by the reef
manta ray, but not the giant manta ray. Sightings of
the latter were also marginally more common on
southerly reefs than at Site 2. Site 1 and the southern
group are wide reefs that offer a large area for clean-
ing, while Site 2 is narrower and has relatively small
cleaning stations. Giant manta rays, because of their
large body size, may prefer larger reefs that enable
them to remain close to cleaning stations, especially
164
Rohner et al.: Environmental influences on planktivorous elasmobranchs
as they tend to get cleaned in groups, which may
require more space to maneuver and alternate
passes over the cleaning stations.
Minimum wave height impacted the sightings of
whale sharks, with more individuals seen when
waves were small. This may simply reflect an in -
crease in sightability during these more favourable
conditions rather than an actual change in whale
shark presence or behaviour. Sharks may swim at
greater depths when surface waters are rough and
hence may be present, but not seen by surface-based
observers (Rowat et al. 2009). Time from high tide
had a weak effect on sightings of reef manta rays and
whale sharks, with slightly more reef manta rays at
low tide. Moon illumination was a supported predic-
tor for both Manta species, with more sightings made
when less than half the moon was illuminated. These
results are in contrast with reef manta rays in
Lombok, Indonesia, where their numbers increased
during full and new moon when the tidal flux was
highest (Dewar et al. 2008).
Significantly more reef manta rays were seen when
zooplankton dominated the water column compared
to when there was no plankton or when suspended
sediment was present. Sediment-rich water could
negatively affect normal gill ventilation for respira-
tion and filter-feeding, and it is possible that large,
mobile planktivores will avoid such conditions.
Higher numbers of reef manta rays seen at cleaning
stations when zooplankton was locally abundant may
indicate that they preferentially visit cleaning sta-
tions that are in close proximity to good feeding
areas.
Reef manta rays were sighted less frequently dur-
ing northward currents compared to the predomi-
nant south- and eastward currents. Fewer reef manta
rays were also seen in strong currents compared to
medium and light currents, but were not significantly
different from absent currents. We therefore inter-
pret this relationship with caution. It appears that,
given reef manta ray presence at the reefs was gen-
erally related to routine cleaning behaviour, this may
indicate stronger currents are sub-optimal for clean-
ing. This is likely because it would be difficult for the
small cleaner fishes to maintain their position above
the reef in strong currents, or perhaps because reef
manta rays preferentially avoid cleaning when it is
energetically costly to hold their position relative to
the reef. Time of day was marginally significant for
reef manta rays, with a suggestive low in sightings
from 08:30 to 10:00 h. This observation may reflect
the relatively high amount of ecotourism boat traffic
and diver numbers close to the reefs at those times.
Broader applications and implications
for conservation
Praia do Tofo in southern Mozambique is one of the
few coastal hotspots for all 3 species discussed in this
study (Pierce et al. 2010, Marshall et al. 2011b). As
whale sharks and manta rays are globally threat-
ened, it is important to separate environmentally
driven short-term fluctuations from significant long-
term trends. This can be a challenge in these migra-
tory species as sighting data are often from a single
geographical locality within their broader range.
Standardisation of catch data with multivariate mod-
els, which take into account a suite of environmental
variables, has become common practice for fisheries
managers (e.g. Agenbag et al. 2003, Zagaglia et al.
2004, Hazin & Erzini 2008, Zainuddin et al. 2008) and
here we have shown that GLMs are also a useful tool
for examining sightings trends of planktivorous elas-
mobranchs. While, on a broad scale, changing ocean
climate has the potential to shift historical aggrega-
tion sites through influencing environmental condi-
tions and zooplankton composition and density, in
this case sightings of M. birostris showed no signifi-
cant trend in the limited study period. The observed
data suggest a decline in sightings of reef manta rays
and whale sharks, and this is confirmed by the GLM
after taking into account a suite of local predictors in
the model. While larger-scale oceanographic vari-
ables could drive the fluctuating numbers of whale
shark sightings, the decline in sightings of reef manta
rays may represent a true decline driven by mortality
due to fishing or by avoidance of reefs visited by
increasing numbers of dive tourists. However, the
data in the present study do not unambiguously iden-
tify the reasons for an apparent decline in sightings
of reef manta rays, but given that Praia do Tofo is
a global hotspot for these species and also hosts a
burgeoning marine tourism industry, potential
threats such as increased fishing and tourism pres-
sure require immediate attention by scientists and
managers.
Acknowledgements. We thank W. N. Venables for his valu-
able input on the modelling and F. R. A. Jaine for productive
discussions on the manuscript. MODIS satellite data were
provided by NASA’s Ocean Biology Processing Group. Casa
Barry Lodge, Tofo Scuba, Peri-Peri Divers, Ocean Revolu-
tion, Project AWARE International and Fondation Ensemble
supported fieldwork. S.J.P’s work on this study was sup-
ported by the Swiss Shark Foundation, Rufford Small Grants
and private donors. A.D.M’s work was also supported by the
Save Our Seas Foundation. Thanks to the many Marine
Megafauna Association, All Out Africa and Quest Under-
seas staff and volunteers who helped with field work.
165
Mar Ecol Prog Ser 482: 153–168, 2013
LITERATURE CITED
Agenbag JJ, Richardson AJ, Demarcq H, Freon P, Weeks S,
Shillington FA (2003) Estimating environmental prefer-
ences of South African pelagic fish species using catch
size- and remote sensing data. Prog Oceanogr 59: 275−300
Alava MNR, Dolumbalo ERZ, Yaptinchay AA, Trono RB
(2002) Fishery and trade of whale sharks and manta rays
in the Bohol Sea, Philippines. In: Fowler SL, Reed TM,
Dipper FA (eds) Proc Elasmobranch Biodiversity, Con -
servation and Management: Proc Int Sem Work. Shark
Specialist Group, IUCN, Gland, p 132–148
Anderson RC, Adam MS, Goes JI (2011) From monsoons to
mantas: seasonal distribution of Manta alfredi in the
Maldives. Fish Oceanogr 20: 104−113
Arzoumanian Z, Holmberg J, Norman B (2005) An astro-
nomical pattern-matching algorithm for computer-aided
identification of whale sharks, Rhincodon typus. J Appl
Ecol 42: 999−1011
Block BA, Jonsen ID, Jorgensen SJ, Winship AJ and others
(2011) Tracking apex marine predator movements in a
dynamic ocean. Nature 475: 86−90
Boucher J, Ibanez F, Prieur L (1987) Daily and seasonal vari-
ations in the spatial distribution of zooplankton popula-
tions in relation to the physical structure in the Ligurian
Sea front. J Mar Res 45: 133−173
Bradshaw CJA, Mollet HF, Meekan MG (2007) Inferring
population trends for the world’s largest fish from mark-
recapture estimates of survival. J Anim Ecol 76: 480−489
Bradshaw CJA, Fitzpatrick BM, Steinberg CC, Brook BW,
Meekan MG (2008) Decline in whale shark size and
abundance at Ningaloo Reef over the past decade: the
world’s largest fish is getting smaller. Biol Conserv 141:
1894−1905
Brooks K, Rowat D, Pierce SJ, Jouannet D, Vely M (2010)
Seeing spots: photo-identification as a regional tool for
whale shark identification. West Indian Ocean J Mar Sci
9: 185−194
Brunnschweiler JM, Baensch H, Pierce SJ, Sims DW (2009)
Deep-diving behaviour of a whale shark, Rhincodon
typus, during long-distance movement in the western
Indian Ocean. J Fish Biol 74: 706−714
CITES (2002) CITES Appendix II nomination of the whale
shark, Rhincodon typus. Proposal 12.35. CITES, Santiago
Cliff G, Anderson-Reade MD, Aitken AP, Charter GE, Ped-
demors VM (2007) Aerial census of whale sharks (Rhin-
codon typus) on the northern KwaZulu-Natal coast,
South Africa. Fish Res 84: 41−46
Colin PL (1992) Reproduction of the Nassau grouper Epi-
nephelus striatus (Pisces: Serranidae) and its relationship
to environmental conditions. Environ Biol Fishes 34:
357−377
Couturier LIE, Jaine FRA, Townsend KA, Weeks SJ,
Richardson AJ, Bennett MB (2011) Distribution, site
affinity and regional movements of the manta ray, Manta
alfredi (Krefft, 1868), along the east coast of Australia.
Mar Freshw Res 62: 628−637
Couturier LIE, Marshall AD, Jaine FRA, Kashiwagi T and
others (2012) Biology, ecology and conservation of the
Mobulidae. J Fish Biol 80: 1075−1119
Deakos MH, Baker JD, Bejder L (2011) Characteristics of a
manta ray Manta alfredi population off Maui, Hawaii,
and implications for management. Mar Ecol Prog Ser
429: 245−260
de la Parra Venegas R, Hueter RE, Gonzalez Cano J, Tymin-
ski JP and others (2011) An unprecedented aggregation
of whale sharks, Rhincodon typus, in Mexican coastal
waters of the Caribbean Sea. PLoS ONE 6: e18994
de Ruijter WPM, Ridderinkhof H, Lutjeharms JRE, Schouten
MW, Veth C (2002) Observations of the flow in the Mo -
zambique Channel. Geophys Res Lett 29: 140-1–140-3,
doi:10.1029/2001GL013714
Dewar H, Mous P, Domeier M, Muljadi A, Pet J, Whitty J
(2008) Movements and site fidelity of the giant manta
ray, Manta birostris, in the Komodo Marine Park, Indone-
sia. Mar Biol 155:
121−133
Duffy CAJ, Abbott D (2003) Sightings of mobulid rays from
northern New Zealand, with confirmation of the occur-
rence of Manta birostris in New Zealand waters. NZ J
Mar Freshw Res 37: 715−721
Eckert SA, Stewart BS (2001) Telemetry and satellite track-
ing of whale sharks, Rhincodon typus, in the Sea of
Cortez, Mexico, and the north Pacific Ocean. Environ
Biol Fishes 60: 299−308
ECOCEAN (2012) Global whale shark photo-ID library.
www.whaleshark.org
Flagg CN, Wirick CD, Smith SL (1994) The interaction of
phytoplankton, zooplankton and currents from 15
months of continuous data in the Mid-Atlantic Bight.
Deep-Sea Res II 41: 411−435
Floch L, Chassot E, Damiano A, Fonteneau V and others
(2012) Statistics of the French purse-seine fleet targeting
tropical tunas in the Atlantic Ocean (1991–2010). Collec-
tive Volume of Scientific Papers, ICCAT 68: 858–885
Gifford A, Compagno LJV, Levine M (2001) Aerial surveys
of whale sharks (Rhincodon typus) off the east coast of
southern Africa from 1993 to 1998. Report to the Shark
Research Institute, Princeton, NJ, p 53–57
Gifford A, Compagno LJV, Levine M, Antoniou A (2007)
Satellite tracking of whale sharks using tethered tags.
Fish Res 84: 17−24
Graham RT, Witt MJ, Castellanos DW, Remolina F, Maxwell
S, Godley BJ, Hawkes LA (2012) Satellite tracking of
manta rays highlights challenges to their conservation.
PLoS ONE 7: e36834
Hazin H, Erzini K (2008) Assessing swordfish distribution in
the South Atlantic from spatial predictions. Fish Res 90:
45−55
Heyman WD, Kjerfve B (2008) Characterization of transient
multi-species reef fish spawning aggregations at Glad-
den Spit, Belize. Bull Mar Sci 83: 531−551
Holmberg J, Norman B, Arzoumanian Z (2008) Robust, com-
parable population metrics through collaborative photo-
monitoring of whale sharks Rhincodon typus. Ecol Appl
18: 222−233
Holmberg J, Norman B, Arzoumanian Z (2009) Estimating
population size, structure, and residency time for whale
sharks Rhincodon typus through collaborative photo-
identification. Endang Species Res 7: 39−53
Hsu HH, Joung SJ, Liao YY, Liu KM (2007) Satellite tracking
of juvenile whale sharks, Rhincodon typus, in the North-
western Pacific. Fish Res 84: 25−31
Hutchinson GE (1967) A treatise on limnology, Vol 2. Wiley,
New York, NY
Kashiwagi T, Marshall AD, Bennett MB, Ovenden JR (2011)
Habitat segregation and mosaic sympatry of the two
species of manta ray in the Indian and Pacific Oceans:
Manta alfredi and M. birostris. Mar Biodivers Rec 4: e53
166
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
Rohner et al.: Environmental influences on planktivorous elasmobranchs 167
Lima ID, Olson DB, Doney SC (2002) Biological response to
frontal dynamics and mesoscale variability in oligo -
trophic environments: biological production and com -
munity structure. J Geoph Res 107: C8, doi: 10.1029/
2000JC000393
Luiz OJJ, Balboni AP, Kodja G, Andrade M, Marum H (2009)
Seasonal occurences of Manta birostris (Chondrichthyes:
Mobulidae) in southeastern Brazil. Ichthyol Res 56: 96−99
Marshall AD (2009) Biology and population ecology of
Manta birostris in southern Mozambique. PhD thesis,
The University of Queensland, St. Lucia
Marshall AD, Bennett MB (2010a) The frequency and effect
of shark-inflicted bite injuries to the reef manta ray
Manta alfredi. Afr J Mar Sci 32: 573−580
Marshall AD, Bennett MB (2010b) Reproductive ecology of
the reef manta ray Manta alfredi in southern Mozam-
bique. J Fish Biol 77: 169−190
Marshall AD, Compagno LJV, Bennett MB (2009) Rede -
scription of the genus Manta with resurrection of Manta
alfredi (Krefft, 1868) (Chondrichthyes; Myliobatoidei;
Mobulidae). Zootaxa 2301: 1−28
Marshall AD, Bennett MB, Kodja G, Hinojosa-Alvarez S and
others (2011a) Manta birostris. In: IUCN red list of
threatened species version 2012-1. www.iucnredlist.org/
details/198921/0 (accessed 24 Sep 2012)
Marshall AD, Dudgeon CL, Bennett MB (2011b) Size and
structure of a photographically identified population of
manta rays Manta alfredi in southern Mozambique. Mar
Biol 158: 1111−1124
Marshall AD, Kashiwagi T, Bennett MB, Deakos MH
and others (2011c) Manta alfredi. In: IUCN red list of
threatened species version 2012-1. www. iucnredlist.org/
details/195459/0 (accessed 24 Sep 2012)
Meekan MG, Bradshaw CJA, Press M, McLean C, Richards
A, Quasnichka S, Taylor JG (2006) Population size
and structure of whale sharks, Rhincodon typus, at Nin -
galoo Reef, Western Australia. Mar Ecol Prog Ser 319:
275−285
Nakaya K, Matsumoto R, Suda K (2008) Feeding strategy of
the megamouth shark Megachasma pelagios (Lamni-
formes: Megachasmidae). J Fish Biol 73: 17−34
Nelson JD, Eckert SA (2007) Foraging ecology of whale
sharks (Rhincodon typus) within Bahia de Los Angeles,
Baja California Norte, Mexico. Fish Res 84: 47−64
Norman B (2005) Rhincodon typus. In: IUCN Red List of
Threatened Species Version 2012-1. Available at www.
iucnredlist.org/details/19488/0 (accessed 24 Sep 2012)
Parker HW, Stott FC (1965) Age, size and vertebra calcifica-
tion in the basking shark Cetorhinus maximus (Gun-
nerus). Zool Mededeelingen 40: 305−319
Pierce SJ, Méndez-Jiménez A, Collins K, Rosero-Caicedo
M, Monadjem A (2010) Developing a code of conduct
for whale shark interactions in Mozambique. Aquat Con-
serv 20: 782−788
Plourde S, Runge JA (1993) Reproduction of the planktonic
copepod Calanus finmarchicus in the Lower St. Law -
rence Estuary: relation to the cycle of phytoplankton pro-
duction and evidence for a Calanus pump. Mar Ecol Prog
Ser 102: 217−227
Pravin P (2000) Whale shark in the Indian coast — need for
conservation. Curr Sci 79: 310−315
Richardson AJ (2008) In hot water: zooplankton and climate
change. ICES J Mar Sci 65: 279−295
Richardson AJ, Schoeman DS (2004) Climate impact on
plankton ecosystems in the Northeast Atlantic. Science
305: 1609−1612
Riley MJ, Hale MS, Harman A, Rees RG (2010) Analysis of
whale shark Rhincodon typus aggregations near South
Ari Atoll, Maldives Archipelago. Aquat Biol 8: 145−150
Rowat D, Gore M (2007) Regional scale horizontal and local
scale vertical movements of whale sharks in the Indian
Ocean off Seychelles. Fish Res 84: 32−40
Rowat D, Gore M, Meekan MG, Lawler IR, Bradshaw CJA
(2009) Aerial survey as a tool to estimate whale shark
abundance trends. J Exp Mar Biol Ecol 368: 1−8
Rowat D, Brooks K, March A, McCarten C and others (2011)
Long-term membership of whale sharks (Rhincodon
typus) in coastal aggregations in Seychelles and Djibouti.
Mar Freshw Res 62: 621−627
Schouten MW, de Ruijter WPM, van Leeuwen PJ, Ridde -
rinkhof H (2003) Eddies and variability in the Mozam-
bique Channel. Deep-Sea Res II 50: 1987−2003
Sequeira A, Mellin C, Rowat D, Meekan MG, Bradshaw
CJA (2012) Ocean-scale prediction of whale shark distri-
bution. Divers Distrib 18: 504−518
Sims DW (2008) Sieving a living: a review of the biology,
ecology and conservation status of the plankton-feeding
basking shark Cetorhinus maximus. Adv Mar Biol 54:
171–220
Sims DW, Reid PC (2002) Congruent trends in long-term
zooplankton decline in the north-east Atlantic and bask-
ing shark (Cetorhinus maximus) fishery catches off west
Ireland. Fish Oceanogr 11: 59−63
Sleeman JC, Meekan MG, Wilson SG, Jenner CKS and oth-
ers (2007) Biophysical correlates of relative abundances
of marine megafauna at Ningaloo Reef, Western Aus-
tralia. Mar Freshw Res 58: 608−623
Sleeman JC, Meekan MG, Fitzpatrick BJ, Steinberg CR,
Ancel R, Bradshaw CJA (2010a) Oceanographic and
atmospheric phenomena influence the abundance of
whale sharks at Ningaloo Reef, Western Australia. J Exp
Mar Biol Ecol 382: 77−81
Sleeman JC, Meekan MG, Wilson SG, Polovina JJ, Stevens
JD, Boggs GS, Bradshaw CJA (2010b) To go or not to go
with the flow: environmental influences on whale shark
movement patterns. J Exp Mar Biol Ecol 390: 84−98
Speed CW, Meekan MG, Rowat D, Pierce SJ, Marshall AD,
Bradshaw CJA (2008) Scarring patterns and relative
mortality rates of Indian Ocean whale sharks. J Fish Biol
72: 1488−1503
Stevens JD (2007) Whale shark (Rhincodon typus) biology
and ecology: a review of the primary literature. Fish Res
84: 4−9
Taylor JG (1996) Seasonal occurrence, distribution and
movements of the whale shark, Rhincodon typus, at
Ningaloo Reef, Western Australia. Mar Freshw Res 47:
637−642
Tew Kai E, Marsac F (2010) Influence of mesoscale eddies on
spatial structuring of top predators’ communities in the
Mozambique Channel. Prog Oceanogr 86: 214−223
Theberge MM, Dearden P (2006) Detecting a decline in
whale shark, Rhincodon typus, sightings in the Andaman
Sea, Thailand using ecotourist operator-collected data.
Oryx 40: 337−342
Turnbull SD, Randell JE (2006) Rare occurrence of a Rhin-
codon typus (whale shark) in the Bay of Fundy, Canada.
Northeast Nat 13: 57−58
van Duinkerken D (2010) Movements and site fidelity of the
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
➤
Mar Ecol Prog Ser 482: 153–168, 2013
reef manta ray, Manta alfredi, along the coast of southern
Mozambique. MSc thesis, Utrecht University
Venables WN, Ripley BD (2002) Modern applied statistics
with S, 4th edn. Springer, New York, NY
Ware DM, Thomson RE (2005) Bottom-up ecosystem trophic
dynamics determine fish production in the Northeast
Pacific. Science 308: 1280−1284
Wilson SG, Taylor JG, Pearce AF (2001) The seasonal aggre-
gation of whale sharks at Ningaloo Reef, Western Aus-
tralia: currents, migrations and the El Niño/Southern
Oscillation. Environ Biol Fishes 61: 1−11
Zagaglia CR, Lorenzzetti JA, Stech JL (2004) Remote sens-
ing data and longline catches of yellowfin tuna (Thunnus
albacares) in the equatorial Atlantic. Remote Sens Envi-
ron 93: 267−281
Zainuddin M, Saitoh K, Saitoh SI (2008) Albacore (Thun-
nus alalunga) fishing ground in relation to oceano-
graphic conditions in the western North Pacific Ocean
using remotely sensed satellite data. Fish Oceanogr 17:
61−73
168
Editorial responsibility: Matthias Seaman,
Oldendorf/Luhe, Germany
Submitted: February 7, 2012; Accepted: January 28, 2013
Proofs received from author(s): April 29, 2013
➤
➤
➤
➤
