Advertisement

Coral Reefs

, Volume 33, Issue 2, pp 329–342 | Cite as

Population dynamics of the reef manta ray Manta alfredi in eastern Australia

  • L. I. E. CouturierEmail author
  • C. L. Dudgeon
  • K. H. Pollock
  • F. R. A. Jaine
  • M. B. Bennett
  • K. A. Townsend
  • S. J. Weeks
  • A. J. Richardson
Open Access
Report

Abstract

The reef manta ray Manta alfredi aggregates at several sites along the east coast of Australia. Photographic identification and mark–recapture methods were used to report on the site affinity, size and structure of this population of M. alfredi. A total of 716 individuals were identified in 1982–2012, including 636 at Lady Elliot Island (LEI), southern Great Barrier Reef. Over 60 % of individuals identified were resighted at least once during the study period. Multiple resightings within and among years imply a high degree of site affinity by individuals to aggregation sites. One individual was sighted 11 times at LEI over a 30-yr period. The sex ratio of this population was significantly biased towards females (1.2:1 female-to-male ratio), and females were more commonly resighted than males. Robust design population models were used to estimate the population size of the winter aggregation at LEI over a 4-yr period. The models estimated up to 456 (95 % CI 399–535) M. alfredi individuals in the population within one winter season and a high annual apparent survival. This study demonstrated that waters around LEI form a key aggregation site for a large portion of the M. alfredi population in east Australian waters.

Keywords

Program MARK Photographic identification Aggregation Abundance Survival Site affinity 

Introduction

Effective assessment of a species’ status and conservation requires detailed information on its biology, ecology and threats (e.g., IUCN 2001). Reliable abundance estimates are essential to the study of population dynamics and to underpin conservation biology (Caughley and Gunn 1996; He and Gaston 2000). Obtaining these estimates is often problematic, especially for wide-ranging species that are difficult to observe and sample. These challenges are acute for large oceanic animals such as elasmobranchs that are able to travel vast distances and remain submerged. Knowledge on their population dynamics is often limited due to a lack of information on their habitat use within the geographical areas they occupy and traverse (Stevens 2010). For such species, predictable aggregations at specific sites provide unique opportunities to assess population sizes, distribution patterns and, potentially, movement patterns of these otherwise elusive fishes (e.g., Dudgeon et al. 2008; Bansemer and Bennett 2009; Holmberg et al. 2009).

The reef manta ray Manta alfredi has a circumglobal distribution in tropical and subtropical waters and is resident in coastal areas (Marshall et al. 2009). Individuals exhibit affinities for particular sites over many years where they often form predictable seasonal aggregations (e.g., Dewar et al. 2008; Marshall et al. 2011a). These aggregations leave the species vulnerable to targeted fisheries. The rising demand for mobulid products in Asia has led to increasing targeted fisheries for Manta and Mobula species in several parts of the world (Couturier et al. 2012), and local M. alfredi populations have declined in some fished areas (Marshall et al. 2011b; Rohner et al. 2013). The species is vulnerable to localise fishing pressure because of its conservative life history strategy (i.e. slow growth, late age at maturity and low fecundity) and because connectivity between geographically distinct subpopulations is likely to be limited (Couturier et al. 2012). Despite growing scientific interest in M. alfredi and the species being listed as Vulnerable on the IUCN Red List of Threatened Species (Marshall et al. 2011b) and on Appendix II of the CITES (CITES 2013), information on the status of the global population, as well as many local subpopulations, is limited.

Photo-identification (photo-ID) of individuals, using natural markings and/or scarring patterns on the body, provides an effective, minimally invasive method of collecting sight–resight (capture–recapture) data for population modelling (Marshall and Pierce 2012). Individual M. alfredi can be identified from the unique skin pigmentation patterns on their ventral surface (e.g., Marshall et al. 2011a). These markings are present from birth (Marshall et al. 2008) and remain unchanged for >30 yrs (Marshall et al. 2011b). The use of photo-ID techniques on M. alfredi has already provided information on the ecology, population structure and behaviour of the species (e.g., Marshall and Bennett 2010a; Deakos 2012). It has also enabled population size estimates at several key aggregation sites. Deakos et al. (2011) estimated that up to 230 individuals were resident off Maui Island, Hawaii, within a 3-month sampling period. Marshall et al. (2011a) estimated that the annual population of M. alfredi off Tofo beach, Mozambique, was 149–454 individuals, with a superpopulation (i.e. total number of individuals in the population over the study period, assuming no mortality) of 890 individuals in 2003–2007. A population size of 537 individuals was estimated around North Male Atoll in the Maldives (Kitchen-Wheeler et al. 2012). Assessment of population dynamics through a combination of mark–recapture modelling and photo-ID data can provide valuable information for the conservation and management of a species. Reliable population estimates require appropriate sampling regimes (Pollock et al. 1990), which should be considered prior to data collection and meet appropriate model assumptions.

Manta alfredi individuals travel seasonally and aggregate at several tropical and subtropical coral and rocky reefs along the east coast of Australia (Couturier et al. 2011). Here, we use photo-ID to investigate the population structure (size distribution and sex ratio), resighting rate, movements between aggregation sites and individual longevity within the M. alfredi population in eastern Australia. We applied mark–recapture methods over a 4-yr intensive survey at the key aggregation site of Lady Elliot Island (LEI) reef to estimate the population size, survival rate and emigration of both males and females. We used Pollock’s robust design (Pollock et al. 1990) as this model allows for temporary emigration and heterogeneity in capture probabilities and thus provides the best estimates of abundance for each surveyed period. This is the first study to use this modelling approach to estimate the population size of M. alfredi during a seasonal aggregation.

Methods

Study site

Photographs of the ventral surface of manta rays were collected year round by the authors, dive instructors and recreational divers at manta ray aggregation sites along the east coast of Australia in 2007–2012 (Fig. 1). The primary sampling sites were Heron Island, LEI, North Stradbroke Island (NSI), Byron Bay and the Solitary Islands (Fig. 1). All sites have shallow coral or rocky reefs (5–25 m depth) where manta rays are commonly observed near the surface and around cleaning stations (Couturier et al. 2011). Most of the data collection was opportunistic, and thus, sampling effort was unequal across years and sites (Table 1). Photographs taken before 2007 were obtained for LEI and NSI sites.
Fig. 1

Locations of monitored sites in eastern Australia

Table 1

Summary of analyses applied to photo-identification data

Analysis

Site

Data

No. manta individuals

Effort type

Observers

Period

Resightings

EA (all sites)

Sight-resight, ID, sex

716

Opportunistic + surveys

Authors + community

1982–June 2012

Site affinity

LEI

Sight-resight, ID, sex, location

636

Opportunistic + surveys

Authors + community

1982–June 2012

Movements among sites

EA (all sites)

Sight-resight, ID, location

96

Opportunistic + surveys

Authors + community

1982–June 2012

Sex ratio and maturity

EA (all sites)

Sex, sighting, ID, location

716

Opportunistic + surveys

Authors + community

1982–June 2012

Sex ratio and maturity

LEI

Sex, sighting, ID, location

636

Opportunistic + surveys

Authors + community

1982–June 2012

Body size

LEI, NSI

ID, laser photogrammetry

75

Opportunistic

Authors

April 2010–June 2012

Minimum population size

LEI

ID, sight-resight, sex, location, date

621

Opportunistic + surveys

Authors + community

June 2009–June 2012

Robust design: population survival, temporary emigration, abundance estimate

LEI

ID, sight-resight, sex, location, date

430

Surveys

Authors

June 2009, June 2010, June 2011, June 2012

EA East Australia, LEI Lady Elliot Island, NSI North Stradbroke Island

Population size estimates were generated for LEI (24°07′S 152°42′E), the southernmost coral cay of the Great Barrier Reef, where M. alfredi is sighted year round with a peak aggregation during cooler months (Couturier et al. 2011; Jaine et al. 2012). Specific photo-ID surveys were conducted at this site during June of each year between 2009 and 2012 to meet sampling design requirements for the application of population models (Table 1). The primary dive site, Lighthouse Bommie (9–15 m depth), is located off the western side of the island and has a sandy substrate with several large scattered coral bommies of ~2–8 m maximum width and 0.5–2.5 m height, spread across an area of about 100 m × 50 m. This dive site is readily accessible and is a key cleaning station for M. alfredi (Jaine et al. 2012).

Photo-ID and laser photogrammetry

Photo-ID procedures followed those in Couturier et al. (2011). Population characteristics extracted from the database included longevity, mean number of sightings, site affinity, sex ratio and size structure among different sites (Table 1).

Two parallel laser pointers were mounted 20 cm (2010) or 50 cm apart (2011–2012) on an underwater camera housing using a custom-made aluminium frame (based on Deakos 2010). Projected laser beams were visible on the body of the photographed manta ray, allowing extrapolation of its size (Fig. 2). Size estimations were determined using only images in which the photographed surface of the ray was near perpendicular to the axis of the laser projections and camera (Fig. 2).
Fig. 2

Photographs showing projected green laser spots, 50 cm apart, on the ventral side of an individual M. alfredi

Disc length (DL) was measured from photographs using Image J.1.45s (Java 1.6.0_20). For comparison, conventional disc width (DW) measurements (Francis 2006) were estimated using the equation of Deakos (2010).
$$ {\text{DW}}({\text{mm}}) = 1.9576 \times {\text{DL}}({\text{mm}}) + 469.13 $$

Since parallax may still be present in some photos, individuals were separated into four 0.5-m size classes ranging from smallest (2.5 m DW) to largest (4.5 m DW) individuals.

The sex of individuals was determined by the presence or absence of claspers. Male maturity was assessed visually with individuals classified as ‘immature’, ‘subadult’ and ‘mature’ based on length and apparent thickness of the claspers, and observable clasper gland structure (Marshall and Bennett 2010a). Maturity of individual females could only be confirmed if a pregnancy was observed, and was presumed when a female was seen engaged in a courtship train or had reproductive mating scars on the left pectoral fin (Marshall and Bennett 2010a). Courtship behaviour, pregnancies and mating scars were monitored opportunistically throughout the study period through direct observations, video sequences and photographs. Female maturity was also assessed based on disc width estimates observed in Hawaii (Deakos 2012), where females ≥3.5 m DW were considered mature. Sex ratio data were analysed using a binomial test with a significant level of p < 0.05.

Population size estimate at LEI

Sampling design

Intensive photographic surveys were conducted by the authors during the peak aggregation of M. alfredi at LEI in 4 primary periods: June 2009, 2010, 2011 and 2012 (Table 1). Each primary period comprised 2 weeks of data collection, with surveys conducted twice daily for 50–60 min on SCUBA. Of the 104 dives at LEI (25 in 2009, 25 in 2010, 27 in 2011 and 27 in 2012), 101 were at Lighthouse Bommie. Each survey had one or two teams of 2–4 divers swimming a standard circuit. Differences in number of divers per dive were not considered to affect sampling success as manta rays are large and conspicuous, and all divers present at the same dive site saw the same individuals. Each dive team was allocated to opposite ends of the dive site to minimise possible effects of divers on manta ray behaviour. Similarly, recreational divers were briefed before each dive to minimise their impact on manta ray behaviour. All daily data were pooled to obtain the total number of identified individual rays per day.

Robust design

Annual population sizes of M. alfredi aggregating at LEI in winter from 2009 to 2012 were estimated using Pollock’s robust design (RD) (Pollock et al. 1990; Kendall et al. 1995, 1997; see Electronic Supplementary Material, ESM) applied in program MARKv6.1 (White and Burnham 1999). The four winter seasons were designated as primary sampling periods separated by 1-yr time intervals (i.e. June 2009, June 2010, June 2011 and June 2012). During each of the primary periods, the population was sampled in secondary sampling periods consisting in daily sampling during the 2 weeks (see ESM). Days with ≤2 individuals ‘captured’ within secondary periods were removed from the data set as small sample size limits the ability of the model to assess temporary emigration and abundance. Several assumptions are inherent in the application of the robust design model to this species: (1) all manta rays possess unique markings that do not change over time; (2) survival rate among primary periods is equal for all manta rays of each sex; (3) the M. alfredi population is closed from additions (i.e. immigration and birth) and deletions (i.e. emigration and death) within each primary period.

Annual apparent survival φ between primary periods was modelled as constant over time φ(·), varying annually φ(t) and with sex effects φ(sex), φ(sex + t). The effect of temporary emigration on abundance estimates was assessed using the Markovian model γ′ and γ″ and the random model γ (γ′ = γ″) (see ESM). The temporary emigration estimate is the probability of individuals present in the population being unavailable for capture in a certain period (Kendall et al. 1997). The influence of temporary emigration for both types of model was examined as time varying (t), constant over time (·) and with and without sex effect (sex), (sex + t). No temporary emigration γ(0) models were included in the candidate model set (Kendall 2012). Due to the negligible effect of photo-ID techniques on manta ray behaviour, capture p and recapture c probabilities were assumed to be equal at all time (p = c) and were modelled as constant (·) or time varying within secondary periods (t), with and without sex effects (sex), (sex + t). Some parameters can be poorly estimated near the probability boundaries of 0 and 1 due to data sparseness. Data cloning procedures were applied to selected models to help identify parameters that did not appear to be estimated as values were close to one of the boundaries (Cooch and White 2012). Akaike’s information criterion for small sample sizes (AICc) was used to assess model support, where the smaller AICc value indicates better model fit to the data (Burnham and Anderson 2004). Abundance estimates and standard errors were averaged across models adjusted using normalised Akaike weights (White et al. 2001).

Results

Occurrence and resightings

A total of 716 M. alfredi individuals were identified along the east Australian coastline out of 2,168 reported encounters for which there was a photographic image suitable for identification purposes between 1982 and 2012. Of these, 636 individuals out of 1,828 encounters were sighted at least once at LEI, including 82 individuals also sighted at another location (i.e. Byron Bay, NSI and/or Heron Island). There were 80 individuals that were only sighted at locations other than LEI (i.e. Osprey Reef, Whitsunday Islands, Heron Island, Wolf Rock, NSI, Byron Bay and Solitary Islands; Fig. 1). Of the 716 individuals identified, 63 % were resighted at least once. The maximum number of sightings for the same individual was 20 (all at LEI between 2007 and 2012). Of the 636 individuals identified at LEI, 66 % were resighted at least once within the study period and 62 % were resighted at least once at that same site. A maximum of 32 different individuals were identified within one dive at LEI. The longest period between first and last sighting events was 30 yrs for a male photographed at LEI in 1982 (visibly mature at the time) and resighted 10 times at the same site in 2007–2012 (Table 2; Fig. 3).
Table 2

Sighting records for M. alfredi individuals photographed prior to 2007

Manta ID

Sex

First sighting

Site

Maturity status

Last sighting

Site

Maturity status

Years between 1st and last sighting

No. of resighting

Locations resighted

#002

F

02/2005

NSI

Unknown

17/06/2012

LEI

Unknown

7

6

LEI

#012

F

07/04/2004

LEI

Unknown

29/06/2012

LEI

Unknown

8

19

LEI

#069

F

11/03/2003

NSI

Unknown

1/02/2012

NSI

Unknown

9

4

LEI and NSI

#084

M

30/09/2006

NSI

Immature

26/06/2012

LEI

Mature

6

8

LEI and NSI

#134

M

1982

LEI

Mature

20/06/2012

LEI

Mature

30

10

LEI

#274

M

11/03/2003

NSI

Mature

29/06/2012

LEI

Mature

9

11

LEI and NSI

#320

F

07/09/2004

LEI

Unknown

26/06/2012

LEI

Unknown

8

2

LEI

#430

M

04/04/1993

NSI

Immature

3/03/2012

NSI

Mature

19

2

LEI and NSI

LEI Lady Elliot Island, NSI North Stradbroke Island, M Male, F Female

Fig. 3

Photo-ID of individual male #134 at LEI in 1982 (month unknown, photographed by Peter Ross Allen, Aqua-Photo Pty Ltd) and June 2012 (last sighting). Numbers and arrows show different matching marking sets used to identify this individual; these remained unchanged over time

Overall, 621 individuals were identified at LEI between June 2009 and June 2012 including 430 in the June surveys of 2009, 2010, 2011 and 2012 combined. A minimum of 110 and a maximum of 244 individuals were sighted in any one June survey (Fig. 4). Of the 430 individuals, 62 % were only seen in a single survey, 16 % were sighted in at least two consecutive surveys, while 22 % were seen at least twice but were absent in one or two of the surveys.
Fig. 4

Number of female (grey bars) and male (black bars) M. alfredi sighted during intensive June surveys in 2009–2012. Hatched areas represent the number of individuals sighted only once within the sampling period for each sex

Sex ratio and maturity

Of the 716 individuals identified across all sites, 377 (53 %) were females, 302 (42 %) were males, with a female-biased sex ratio of 1.2:1 (p < 0.05; proportion female = 0.56, 95 % CI 0.52–0.59), and 37 (5 %) could not be sexed. Of all males identified, 74 % were mature, 6 % were subadults and 20 % were immature. Considering pregnancies, presence of mating scars and observations of courtship behaviour as indicators of maturity, 18 % of the females identified were mature and 10 % confirmed to be pregnant at least once. Out of the 22 observations of courtship behaviour at LEI, 13 were in June–August, 5 in October–November and 1 in March. At NSI, 9 courtship behaviours were observed from October to March. One occurrence of courtship behaviour was reported at Osprey Reef in June 2009. A total of 16 identified females bore mating scars on their left pectoral fin, of which 4 had fresh red abrasions indicative of recent mating: 2 were seen in September 2010 at LEI, 1 in June 2012 at LEI and 1 in December 2008 at NSI. Mating scars were not observed on the right pectoral fin.

Of the 450 resightings across all sites, 262 (58 %) were females, 177 (39 %) were males, and 11 (2 %) could not be sexed. Individual females were significantly more likely to be resighted than males, with 69 % of the total number of females resighted at least once in contrast to 59 % of all males (χ 2 = 16.46, df = 1, p < 0.05). The mean number of sightings per individual was 3.4 for females and 2.8 for males. For resighted males, 73 % were mature, 5 % were subadults, and 21 % were immature.

Of the 636 individuals identified at LEI between 2007 and 2012, 340 (53 %) were females, and 269 (42 %) males, with a female-biased sex ratio of 1.3:1 (p < 0.05; proportion females = 0.56, 95 % CI 0.52–0.60), and 27 (4 %) could not be sexed. Of the males identified at LEI, 74 % were mature, 7 % were subadults and 19 % were immature. Of the 395 individuals resighted at LEI, 60 % were females and 38 % males. Females were resighted more than males at LEI, with 69 % of the females resighted at least once in contrast to 57 % of the males identified (χ 2 = 18.19, df = 1, p > 0.05). The mean number of sightings per individual at LEI was 3.3 for females and 2.7 for males.

Of the 621 individuals identified at LEI between June 2009 and June 2012, 331 (53 %) were females, 265 (43 %) were males and 25 (4 %) could not be sexed. Out of the 430 individuals sighted at LEI during the four intensive surveys, 252 (59 %) were females and 178 (41 %) were males (Fig. 4). Females were more likely to be resighted within and among primary periods than males (Figs. 4, 5).
Fig. 5

Proportion of female (grey bars, n = 252) and male (black bars, n = 178) M. alfredi identified at LEI over the 4-yr intensive survey (June months of 2009–2012) with different sighting intervals among sampling years

Size distribution

Disc width estimates of 75 M. alfredi were pooled for 2010–2012 (54 females and 21 males). Most individuals were 3–4 m DW (n = 62), 5 individuals were <3 m DW (2 females, 3 males: all immature), and 8 were >4 m DW (all females) (Fig. 6). The vast majority of males (81 %) were 3–3.5 m DW, with 16 of 17 males classified as mature within this size class. Only one mature male was 3.5–4 m DW. The majority of females (56 %) were 3.5–4 m DW. Based on the assumption that M. alfredi in Australia reaches maturity at ~3.5 m DW (Deakos 2012), 70 % of the females would be mature.
Fig. 6

Distribution of disc width for female (grey bars, n = 54) and male (black bars, n = 21) M. alfredi pooled for 2010–2012

Movements

A total of 96 (13 % of 716 individuals) M. alfredi were sighted at more than one site along the east coast of Australia, including 83 individuals seen at two different sites and 13 at three different locations (Table 3). One manta ray was sighted at both LEI and North-West Solitary Island, in the Solitary Islands Marine Park, 650 km apart, within a 6-month period.
Table 3

Number of individuals sighted at more than one location

 

LEI and NSI

LEI and Byron Bay

LEI and HI

NSI and Byron Bay

LEI and North-West Solitary Island

LEI, NSI and Byron Bay

HI, LEI and NSI

Total

56

12

1

12

1

12

1

Male

21

5

0

5

 

6

 

Female

33

5

1

7

1

6

1

Unsexed

2

2

     

LEI Lady Elliot Island, NSI North Stradbroke Island, HI Heron Island

Population modelling and abundance estimates

A total of 15 out of the 17 RD selected models demonstrated information-theoretic support (Table 4). Models including constant apparent survival were best supported (Table 4: sum of Akaike weights for φ(·) = 0.66), and this parameter was estimated close to the upper boundary [1] in the three most parsimonious models. Data cloning procedures showed that this parameter was being estimated by the models but could not be maximised away from the boundary due to data sparseness (low capture), and thus, no meaningful standard error and confidence interval were reported. Models incorporating random temporary emigration were better supported than Markovian models (Table 4: sum of Akaike weights for γ = 0.82), and thus, the probability for an individual to be absent at a certain period was independent of its presence or absence at the previous sampling period. Random temporary emigration varying between sexes had more support than constant (·), time varying (t) or no temporary emigration (0) [Table 4: sum of Akaike weights for γ(sex) = 0.73]. The best-fit model indicated that females were more likely to be temporarily emigrant than males between primary periods (γ female = 0.32 ± 0.06, γ male = 0). Temporary emigration parameters for males could not be maximised away from the [0] boundary due to data sparseness.
Table 4

Model selection for the robust design (n = 17) models used to estimate population size, survival and capture probability parameters for females and males

Model

AICc

ΔAICc

AICc weights

Model likelihood

No. of parameters

φ(·)γ(sex)p = c(sex + t)

1,111.391

0.000

0.551

1.000

65

φ(sex)γ(sex)p = c(sex + t)

1,113.745

2.354

0.170

0.308

66

φ(·)γ″(sex)γ′(sex)p = c(sex + t)

1,115.942

4.550

0.057

0.103

67

φ(sex + t)γ(0)p = c(sex + t)

1,116.347

4.955

0.046

0.084

66

φ(·)γ(·)p = c(sex + t)

1,116.459

5.068

0.044

0.079

64

φ(sex)γ(·)p = c(sex + t)

1,117.452

6.060

0.027

0.048

65

φ(sex + t)γ″(sex + t)γ′(sex + t)p = c(sex + t)

1,117.891

6.499

0.021

0.039

68

φ(sex + t)γ(sex + t)p = c(sex + t)

1,118.298

6.906

0.017

0.032

68

φ(sex)γ″(sex)γ′(sex)p = c(sex + t)

1,118.301

6.910

0.017

0.032

68

φ(sex)γ(0)p = c(sex + t)

1,118.336

6.944

0.017

0.031

64

φ(sex + t)γ(sex)p = c(sex + t)

1,118.340

6.948

0.017

0.031

68

φ(·)γ(0)p = c(sex + t)

1,120.735

9.343

0.005

0.009

63

φ(sex)γ(t)p = c(sex + t)

1,121.684

10.293

0.003

0.006

67

φ(sex + t)γ″(sex)γ′(sex)p = c(sex + t)

1,122.862

11.470

0.002

0.003

70

φ(·)γ(sex)p = c(t)

1,140.838

29.447

0.000

0.000

61

φ(·)γ(·)p = c(t)

1,146.287

34.895

0.000

0.000

55

φ(·)γ(sex)p = c(sex)

1,199.858

88.467

0.000

0.000

20

Time-varying capture probability with a sex effect was supported by all 15 informative RD models (Table 4). This is attributed to the high variation in M. alfredi sightings between secondary samples (Fig. 7). These models also strongly supported differences in capture probabilities between males and females (Table 4: sum of Akaike weight = 1). Although both sexes followed the same trend within each primary period, females had higher capture probability than males at all times, with differences between female and male probability values varying between 0.005 and 0.11 (Fig. 7).
Fig. 7

Capture probabilities for secondary sampling periods for female (grey line) and male (black line) M. alfredi at LEI, taken from the best-fit model φ(·)γ(sex)p = c(sex + t). Standard errors are shown

Little variation in abundance estimates was attributed to model selection for males and females for primary periods (Table 5). Weighted abundance estimates showed an increase in population size and that females were more abundant than males during the first and second primary periods, and then lower during the third and fourth periods, although 95 % CI for male estimates encompassed the female values (Table 5).
Table 5

Population size estimates of female and male M. alfredi; weighted average across 17 robust design (RD), overall population size estimates from best-fit model φ (·) γ (·) p = c(t) and total number of individuals identified at LEI using photo-ID between June 2009 and 2012

Sex

Method

Year

Weighted average

Uncond. SE

95 % CI

% variation

Female

RD

2009

140

15

110–169

0

  

2010

183

38

109–257

10.48

  

2011

229

36

158–300

15.05

  

2012

230

17

196–264

5.43

 

Photo-ID

2009–2012

331

   

Male

RD

2009

121

21

80–163

0

  

2010

150

31

90–211

3.04

  

2011

264

60

147–382

3.37

  

2012

301

45

214–389

6.17

 

Photo-ID

2009–2012

265

   

Overall

RD

2009

256

24

219–314

 
  

2010

321

46

248–432

 
  

2011

454

58

361–589

 
  

2012

456

34

399–535

 
 

Photo-ID

2009–2012

621

   

Uncond. SE standard error estimate that is unconditional on a particular model, CI confidence interval for the weighted average estimate based on the logit transformation, % variation variation in the estimate attributable to the model uncertainty

Separate models excluding sex differentiation were run to obtain overall population abundance estimates for each June survey. The best-fit model had most of the AIC weighting [φ(·)γ(·)p = c(t); Akaike weight = 0.95] and abundance estimates from this model varied between 256 and 456 individuals for the June surveys (Tables 4, 5).

Discussion

Seasonal aggregations of M. alfredi have been documented across their range (e.g., Dewar et al. 2008; Anderson et al. 2011; Deakos et al. 2011; Marshall et al. 2011a). Although these predictable aggregations are not likely to represent entire regional populations, they nonetheless provide unique opportunities to investigate subpopulation dynamics. Using photo-ID, we have provided detailed information on the population dynamics of M. alfredi in eastern Australia, as well as the first population size estimates for this species in Australia. Females were sighted more frequently, and site visitation patterns varied between sexes. As individuals use multiple aggregation sites within east Australian waters and adequate sampling effort could not be achieved at all monitored sites (Couturier et al. 2011), it was not possible to estimate the total population size of M. alfredi for the entire area. The boundaries separating M. alfredi subpopulations and the interconnectivity with neighbouring regions are currently unknown. The focus for assessing population size was thus on manta rays that use waters around LEI, the most important known aggregation site in eastern Australia.

Photo-ID validation

The availability of photographs of M. alfredi from the 1980s provides supporting evidence on the longevity of M. alfredi (Marshall et al. 2011b), with one individual photographed 30 yrs apart. Moreover, this photographic record indicates that retention of ventral body surface pigmentation extends over long period of times, including for melanistic (i.e. dark-coloured skin) manta rays.

Site affinity and movements

Over 88 % of rays recorded between Osprey Reef and South Solitary Island were sighted at least once at LEI, and individuals revisited this same site multiple times over long periods. Dive sites at LEI are accessed almost daily, which may help explain the large numbers of manta rays sighted and resighted at LEI between 2007 and 2012 (Couturier et al. 2011; Jaine et al. 2012). Nonetheless, occurrence of manta rays comparable to those seen at LEI, e.g., over 30 individuals sighted in one dive or 80 rays seen feeding at the surface (Jaine et al. 2012), has not been observed or reported at any other location in eastern Australia. It is possible, however, that there are similar aggregations along the coastline that have yet to be identified. M. alfredi presence at LEI may be related to seasonal food availability in the area. The island is located near the continental shelf edge where the Capricorn Eddy supplies nutrient-rich waters to the neighbouring reefs via upwelling (Weeks et al. 2010). This oceanographic process could be the source of a pulse in zooplankton productivity within this region (Jaine et al. 2012). Findings of the current study, together with those from previous research (Couturier et al. 2011; Jaine et al. 2012), demonstrate that waters off LEI provide an important seasonal habitat for what appears to be a large proportion of the M. alfredi population in eastern Australia.

Over 66 % of identified individuals were seen more than once at LEI. These results are comparable with those of Hawaii, where over 70 % of identified rays revisited the same site within the 5-yr study period (Deakos et al. 2011). M. alfredi showed greater site affinity at LEI than in Mozambique and the Maldives. Over 40 % of identified individuals in Mozambique were resighted at least once in the study area over a 4-yr period (Marshall et al. 2011a), and 36 % of the identified individuals at North Male Atoll revisited the same site over a 9-yr period (Kitchen-Wheeler et al. 2012). M. alfredi exhibit site affinity for several locations within a certain range, with individuals travelling seasonally up to 270 km in the Maldives (Anderson et al. 2011), 400 km in Japan (Marshall et al. 2011b) and up to 650 km in eastern Australia (this study). These recurrent movements indicate that subpopulations occupy large areas that include several key aggregation sites. Long-term resighting records of individuals at these key sites combined with strong site affinity suggest that M. alfredi subpopulations are unlikely to overlap with other geographically distant subpopulations (e.g., Australia and the Maldives). Interestingly, no population overlap was detected between Maui and Hawai’i Islands, two aggregation sites for M. alfredi only 49 km apart, even though both sites were intensively monitored for over 10 yrs (Deakos et al. 2011). These sites are separated by a 2,000-m deep channel, which suggests movements of individuals between subpopulations might be restricted by bathymetric features and/or regional circulation (Deakos et al. 2011). The possibility exists, however, that geographically adjacent subpopulations to the present study area have a degree of connectivity, and this should be assessed through analysis of manta ray image databases from different localities, such as waters off Fiji, New Caledonia and western Australia, and by the application of molecular genetics approaches (e.g., Dudgeon et al. 2012; Kashiwagi et al. 2012).

Population structure

Size range and size at maturity (3.0–3.5 m) for males in eastern Australia are in agreement with estimates generated from Hawaiian to Mozambican reef manta ray populations (Marshall and Bennett 2010a; Deakos 2012). It is not possible to determine female sexual maturity without an indicator of mating activity, and thus, only 18 % of the identified females were considered mature. However, direct size measurements showed that 70 % of the females examined were larger than the size at maturity (≥3.5 m) reported by Deakos (2012) and within the size range (3.0–4.5 m) estimated for mature females by Kitchen-Wheeler et al. (2012). This might not be representative of the whole population but suggests that the majority of females are likely to be mature.

Females were significantly more prevalent than males at LEI (and in eastern Australian waters as a whole, although these data are strongly influenced by the LEI sightings) with a 1.3:1 female:male ratio. Although more pronounced than in eastern Australia, female-biased sex ratios were also observed in the Maldives (1.8:1) (Kitchen-Wheeler et al. 2012) and Mozambique (3.5:1) (Marshall et al. 2011a). By contrast, the M. alfredi population at Maui Island had no significant bias (Deakos et al. 2011). Reasons behind sex-biased habitat use in manta rays are unclear but could be related to behavioural strategies. A strong female-biased sex ratio in Mozambique, in addition to higher site affinity by females, suggests that this area is a refuge habitat for females and may be an important breeding and/or pupping site for M. alfredi (Marshall et al. 2011a). Molecular genetic analyses on several elasmobranch species suggest greater levels of philopatry in females than males (e.g., Schrey and Heist 2003; Blower et al. 2012). Male M. alfredi may also roam more than females, which return more regularly to a natal or pupping site. Courtship behaviours and mating scars observed at LEI suggest that this site is important for social interaction and mating activities. However, the lack of small M. alfredi (<2 m DW) indicates that females are unlikely to give birth at this site.

The smallest free-swimming M. alfredi reported in the literature measured 1.2–1.5 m DW (Marshall and Bennett 2010a). We found only one individual <2 m DW at NSI in March 2011 over our 4-yr study. Few small individuals were also reported in Hawaii, Mozambique and the Maldives (Deakos et al. 2011; Marshall et al. 2011a; Kitchen-Wheeler et al. 2012). The rare occurrence of small individuals (<2 m) at major aggregation sites may reflect the low reproductive rate of the species. It may also suggest size-based segregation in M. alfredi populations, with different habitats used by neonates and young-of-year. Many other elasmobranch species give birth in nursery areas where food resources are plentiful and neonate survival is thought to be enhanced due to lower predation pressure (Feldheim et al. 2002; Heupel et al. 2007; Bansemer and Bennett 2011), but it is unknown whether this is the case in manta rays.

Population dynamics from mark–recapture models

Survival

Although annual apparent survival of M. alfredi at LEI could not be assessed robustly, all models estimated this parameter to be near 1 in both sexes, suggesting little mortality and/or permanent emigration of individuals between years. This is biologically plausible as M. alfredi is not commercially fished in Australia and probably suffers low natural predation rates once mature. The high survival rate between years was thus not surprising as the sampled population comprised mostly large and mature individuals exhibiting strong site affinity for LEI. High survival estimates are also strongly supported by the long-term photo-ID sighting records showing that some individuals were regularly resighted at LEI over at least 6 yrs and up to 30 yrs. Higher rates of mortality likely occur at neonate and early juvenile life stage, as is common for many elasmobranchs (Cortés 2004). Given their apparent longevity and their expected low natural mortality, this 4-yr study represents a relatively short period in the lifespan of a reef manta ray. High survival rates between years were also found in the Maui subpopulation where M. alfredi is fully protected against commercial fishing and appears to have low exposure to predation pressure (Deakos et al. 2011). The annual apparent survival of the Mozambique population was estimated to vary between 0.6 and 0.7, which may be due to the local subpopulation sustaining a high fishing mortality (Marshall et al. 2011a). Further, >75 % of individuals identified at this location bore shark-inflicted injuries indicative of high predatory pressure on this population (Marshall and Bennett 2010b), especially when compared with the Hawaiian and east Australian subpopulations where 33 % (Deakos et al. 2011) and 23 % (LIE Couturier pers obs) of individuals have scars that result from shark predatory interactions,.

Abundance and temporary emigration

A minimum of 621 individuals were sighted at LEI in June 2009–June 2012, which is likely an underestimate of the true population size of individuals using LEI waters. In addition, the rare occurrence of individuals <2.5 m DW means that the sampled population excluded most immature individuals and only represented a portion of the true population. The largest annual number of M. alfredi estimated to visit LEI during winter was 456 individuals (95 % CI 399–535) for our last survey period in 2012, suggesting that not all individuals present in the sampled population use this habitat in winter. Limitations in the interpretation of RD abundance estimates, with regards to model assumptions, are discussed in the ESM. The total number of females identified between 2009 and 2012 was larger than the RD estimates in any given year, which suggests that the subpopulation of females visiting in winter represents only a portion of the available population. By contrast, the total number of males identified and the abundance estimate from RD models were similar in the last 2 yrs, which suggests that individual males are more likely to revisit the site every year. Temporary emigration estimates from the best-fit RD model indicated that females were more likely to be temporary emigrants than males (see ESM for temporary emigration estimate limitations). This further supports the assumption that not all the available female population visit LEI in winter and individual females have different visitation intervals than males. Courtship behaviours of M. alfredi were regularly observed in winter at LEI, suggesting mating occurs during these seasonal aggregations. Although M. alfredi can produce offspring every year, this species may have a 1- or 2-yr resting period between pregnancies (Marshall and Bennett 2010a). This biennial (or triennial) reproductive periodicity may explain why not all females visit LEI each winter, and individual females may have different visitation intervals. In Carcharias taurus, an elasmobranch with a biennial (or longer) reproductive periodicity, movement patterns are dependent upon whether a female is reproductively active or resting (Bansemer and Bennett 2011).

The RD models indicated an annual increase in abundance for both sexes from 2009 to 2012, which may indicate genuine growth of the subpopulation as M. alfredi is not exploited in Australian waters. Variations in abundance among years may also be influenced by fluctuations of broader environmental parameters that would affect seasonal visitation patterns at LEI (Jaine et al. 2012). However, this population increase pattern could also be, in part, an artefact due to year-on-year improvement in the ability of observers to recognise whether an individual manta ray had already been photographed within a single dive. With increased experience in the field, observers were less likely to miss an individual ray when many rays were on site at one time. The effect of these different factors can only be assessed by extending the study across multiple years and controlling for sampling effort.

Comparison of trends in population sizes from this study with those from other subpopulations of M. alfredi is constrained by differences in sampling, modelling approaches and different environmental conditions. Deakos et al. (2011) reported an increase in the estimated population at a single site across several opportunistic sampling periods over 5 yrs, with the exception of the last period monitored. By contrast, Marshall et al. (2011a) showed a decrease in reef manta ray population size in Mozambique over the last three surveyed years. This decline may be linked to local anthropogenic pressures (Marshall et al. 2011a) or natural predation (Marshall and Bennett 2010a). Population estimates for M. alfredi in the Maldives from Kitchen-Wheeler et al. (2012) are difficult to interpret as the sampling design and analysis appear to violate assumptions of the models used, and the model selection process was not reported. Together, these studies suggest that subpopulations of M. alfredi generally number in the hundreds within defined areas across years. Rapid removal of individuals through fisheries or habitat loss at such aggregation sites may have a strong impact on the survival of these subpopulations due to slow fecundity and limited immigration in M. alfredi.

Capture probability

Probability of capture within and between each primary period showed strong variation between sampling days. It is unlikely that these results are biased by trap-dependent behaviour of individuals as photo-ID is a minimally invasive technique that generally does not interfere with the ray’s activity (LIE Couturier pers ob). Within the primary period, variation in capture probability could reflect changes in the local environment as the daily abundance of individuals at LEI is influenced by temperature, wind speed, tide, local productivity and moon phase (Jaine et al. 2012). Differences in capture probability among primary periods could also be influenced by broader atmospheric and oceanographic processes associated with the El Niño Southern Oscillation, which influences the oceanography of the southern Great Barrier Reef (Weeks et al. 2010; Redondo-Rodriguez et al. 2012).

Aggregation sites as key habitats

Lady Elliot Island is an important aggregation site for M. alfredi in eastern Australia and provides a unique opportunity to study its population dynamics across seasons and among years. We showed that ~456 individuals visited this site within one winter season by application of a robust sampling design. The role that the environment plays in supporting the M. alfredi subpopulation at LEI is not fully understood. However, this site supports a substantial seasonal aggregation, which is likely to be a consequence of regional productivity events triggered by oceanic circulation patterns (Jaine et al. 2012). We showed here that this aggregation may also be linked with the reproductive ecology of the species and that a high proportion of the surveyed population was associated with this site for an extended period of time. Investigation of residency and site fidelity across seasons, as well as movement patterns outside the study area, will provide greater information on the role of LEI as a critical habitat for M. alfredi.

This study highlights the importance of aggregation sites as critical habitat for reef manta ray populations over extended periods. It also presents a robust sampling design that could be replicated at other aggregation sites to monitor local subpopulations. This is relevant to management, as localised anthropogenic pressures such as coastal development, unmanaged tourism and/or fisheries can have a direct impact on manta ray visitation patterns or population depletion (Marshall et al. 2011b). Considering the relatively low population size and high site affinity estimated for all monitored aggregations (i.e. Hawaii, Mozambique, the Maldives and LEI), it is appropriate that manta rays at these sites are protected from overexploitation and disturbances.

Notes

Acknowledgments

We thank our colleagues and numerous dive associates who contributed photographs and information on manta ray sightings. We are grateful to S. McGrellis, C. Rohner, M. Atkinson, C. Garraway, R. Cheseldene-Culley, P. Gartrell, C. Gillies, A. Donnelly and Earthwatch volunteers for their assistance in sample collection and photo-ID. We thank K. Burgess, C. Bustamante and T. Kashiwagi for their comments on the manuscript. This study was supported by the Australian Research Council Linkage Grant (LP110100712), Earthwatch Australia, Sea World Research and Rescue Foundation Inc. and Sibelco Pty Ltd. Field work was supported by Lady Elliot Island Eco Resort, Manta Lodge and Scuba Centre and Sundive Byron Bay, and was conducted under Great Barrier Reef Marine Park permit (G09/29853.1), Marine Parks permit (QS2008/CVL1440a) and Ethics approval (SBMS/071/08/SEAWORLD).

Supplementary material

338_2014_1126_MOESM1_ESM.doc (34 kb)
Supplementary material 1 (DOC 34 kb)

References

  1. Anderson RC, Adam MS, Goes JI (2011) From monsoons to mantas: seasonal distribution of Manta alfredi in the Maldives. Fish Oceanogr 20:104–113CrossRefGoogle Scholar
  2. Bansemer CS, Bennett MB (2009) Reproductive periodicity, localised movements and behavioural segregation of pregnant Carcharias taurus at Wolf Rock, southeast Queensland, Australia. Mar Ecol Prog Ser 374:215–227CrossRefGoogle Scholar
  3. Bansemer CS, Bennett MB (2011) Sex-and maturity-based differences in movement and migration patterns of grey nurse shark, Carcharias taurus, along the eastern coast of Australia. Mar Freshw Res 62:596–606CrossRefGoogle Scholar
  4. Blower DC, Pandolfi JM, Bruce BD, Gomez-Cabrera MdC, Ovenden JR (2012) Population genetics of Australian white sharks reveals fine-scale spatial structure, transoceanic dispersal events and low effective population sizes. Mar Ecol Prog Ser 455:229–244CrossRefGoogle Scholar
  5. Burnham KP, Anderson DR (2004) Multimodel inference understanding AIC and BIC in model selection. Sociological Meth Res 33:261–304CrossRefGoogle Scholar
  6. Caughley G, Gunn A (1996) Conservation biology in theory and practice. Blackwell Science, Cambridge, MAGoogle Scholar
  7. CITES (2013) Amendments to Appendices I and II of the Convention on International Trade in Endangered Species of Wild Fauna and Flora. www.cites.org
  8. Cooch EG, White GC (2012) Appendix F: Parameter identifiability by data cloning. In: Cooch EG, White GC (eds) Program Mark: ‘A Gentle Introduction’, pp 975–995Google Scholar
  9. Cortés E (2004) Life history patterns, demography, and population dynamics. In: Carrier JC, Musick JA, Heithaus MR (eds) Biology of sharks and their relatives. CRC Press, Boca Raton, FL, pp 449–469CrossRefGoogle Scholar
  10. 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 62:628–637CrossRefGoogle Scholar
  11. Couturier LIE, Marshall AD, Jaine FRA, Kashiwagi T, Pierce SJ, Townsend KA, Weeks SJ, Bennett MB, Richardson AJ (2012) Biology, ecology and conservation of the Mobulidae. J Fish Biol 80:1075–1119PubMedCrossRefGoogle Scholar
  12. Deakos MH (2010) Paired-laser photogrammetry as a simple and accurate system for measuring the body size of free-ranging manta rays Manta alfredi. Aquat Biol 10:1–10CrossRefGoogle Scholar
  13. Deakos MH (2012) The reproductive ecology of resident manta rays (Manta alfredi) off Maui, Hawaii, with an emphasis on body size. Environ Biol Fish 94:443–456CrossRefGoogle Scholar
  14. 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 429:245–260CrossRefGoogle Scholar
  15. 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, Indonesia. Mar Biol 155:121–133CrossRefGoogle Scholar
  16. Dudgeon CL, Noad MJ, Lanyon JM (2008) Abundance and demography of a seasonal aggregation of zebra sharks Stegostoma fasciatum. Mar Ecol Prog Ser 368:269–281CrossRefGoogle Scholar
  17. Dudgeon CL, Blower DC, Broderick D, Giles JL, Holmes BJ, Kashiwagi T, Krück NC, Morgan JAT, Tillett BJ, Ovenden JR (2012) A review of the application of molecular genetics for fisheries management and conservation of sharks and rays. J Fish Biol 80:1789–1843PubMedCrossRefGoogle Scholar
  18. Feldheim KA, Gruber SH, Ashley MV (2002) The breeding biology of lemon sharks at a tropical nursery lagoon. Proc R Soc Lond B Biol Sci 269:1655–1661CrossRefGoogle Scholar
  19. Francis MP (2006) Morphometric minefields—towards a measurement standard for chondrichthyan fishes. Environ Biol Fish 77:407–421CrossRefGoogle Scholar
  20. He F, Gaston KJ (2000) Estimating species abundance from occurrence. Am Nat 156:553–559CrossRefGoogle Scholar
  21. Heupel MR, Carlson JK, Simpfendorfer CA (2007) Shark nursery areas: concepts, definition, characterization and assumptions. Mar Ecol Prog Ser 337:287–297CrossRefGoogle Scholar
  22. Holmberg J, Norman B, Arzoumanian Z (2009) Estimating population size, structure, and residency time for whale sharks Rhincodon typus through collaborative photo-identification. Endangered Species Res 7:39–53CrossRefGoogle Scholar
  23. IUCN (2001) IUCN red list categories and criteria: version 3.1 (Prepared by the IUCN Species Survival Commission). In IUCN Red List of Threatened Species. Version 2011. 1. www.iucnredlist.org/technical-documents/categories-and-criteria/2001-categories-criteria
  24. Jaine FRA, Couturier LIE, Weeks SJ, Townsend KA, Bennett MB, Fiora K, Richardson AJ (2012) When giants turn up: sighting trends, environmental influences and habitat use of the manta ray Manta alfredi at a coral reef. PLoS ONE 7:e46170PubMedCentralPubMedCrossRefGoogle Scholar
  25. Kashiwagi T, Marshall AD, Bennett MB, Ovenden JR (2012) The genetic signature of recent speciation in manta rays (Manta alfredi and M. birostris). Mol Phylogenet Evol 64:212–218PubMedCrossRefGoogle Scholar
  26. Kendall WL (2012) Chapter 15: The ‘Robust Design’. In: Cooch EG, White GC (eds) Program Mark: ‘A Gentle Introduction’, pp 567–617Google Scholar
  27. Kendall WL, Pollock KH, Brownie C (1995) A likelihood-based approach to capture-recapture estimation of demographic parameters under the robust design. Biometrics 51:293–308PubMedCrossRefGoogle Scholar
  28. Kendall WL, Nichols JD, Hines JE (1997) Estimating temporary emigration using capture-recapture data with Pollock’s robust design. Ecology 78:563–578Google Scholar
  29. Kitchen-Wheeler AM, Ari C, Edwards AJ (2012) Population estimates of Alfred mantas (Manta alfredi) in central Maldives atolls: North Male, Ari and Baa. Environ Biol Fish 93:557–575CrossRefGoogle Scholar
  30. Marshall AD, Bennett MB (2010a) Reproductive ecology of the reef manta ray Manta alfredi in southern Mozambique. J Fish Biol 77:169–190PubMedCrossRefGoogle Scholar
  31. Marshall AD, Bennett MB (2010b) The frequency and effect of shark-inflicted bite injuries to the reef manta ray Manta alfredi. Afr J Mar Sci 32:573–580CrossRefGoogle Scholar
  32. Marshall AD, Pierce SJ (2012) The use and abuse of photographic identification in sharks and rays. J Fish Biol 80:1361–1379PubMedCrossRefGoogle Scholar
  33. Marshall AD, Pierce SJ, Bennett MB (2008) Morphological measurements of manta rays (Manta birostris) with a description of a foetus from the east coast of Southern Africa. Zootaxa 1717:24–30Google Scholar
  34. Marshall AD, Compagno LJV, Bennett MB (2009) Redescription of the genus Manta with resurrection of Manta alfredi (Krefft, 1868) (Chondrichthyes; Myliobatoidei; Mobulidae). Zootaxa 2301:1–28Google Scholar
  35. Marshall AD, Dudgeon CL, Bennett MB (2011a) Size and structure of a photographically identified population of manta rays Manta alfredi in southern Mozambique. Mar Biol 158:1111–1124CrossRefGoogle Scholar
  36. Marshall AD, Kashiwagi T, Bennett MB, Deakos MH, Stevens G, McGregor F, Clark T, Ishihara H, Sato K (2011b) Manta alfredi. IUCN 2011. IUCN Red List of Threatened Species. Version 2011.1. 1 www.iucnlist.org
  37. Pollock KH, Nichols JD, Brownie C, Hines JE (1990) Statistical inference for capture-recapture experiments. Wildl Monogr 107:3–97Google Scholar
  38. Redondo-Rodriguez A, Weeks SJ, Berkelmans R, Hoegh-Guldberg O, Lough JM (2012) Climate variability of the Great Barrier Reef in relation to the tropical Pacific and El Niño-Southern Oscillation. Mar Freshw Res 63:34–47CrossRefGoogle Scholar
  39. Rohner C, Pierce S, Marshall A, Weeks S, Bennett M, Richardson A (2013) Trends in sightings and environmental influences on a coastal aggregation of manta rays and whale shark. Mar Ecol Prog Ser 482:153–168CrossRefGoogle Scholar
  40. Schrey AW, Heist EJ (2003) Microsatellite analysis of population structure in the shortfin mako (Isurus oxyrinchus). Can J Fish Aquat Sci 60:670–675CrossRefGoogle Scholar
  41. Stevens JD (2010) Epipelagic oceanic elasmobranchs. In: Carrier JC, Musick JA, Heithaus MR (eds) Sharks and their relatives II: Biodiversity, adaptive physiology, and conservation CRC press, Boca Raton, FA, pp 3–36Google Scholar
  42. Weeks SJ, Bakun A, Steinberg CR, Brinkman R, Hoegh-Guldberg O (2010) The Capricorn Eddy: a prominent driver of the ecology and future of the southern Great Barrier Reef. Coral Reefs 29:975–985CrossRefGoogle Scholar
  43. White GC, Burnham KP (1999) Program MARK: survival estimation from populations of marked animals. Bird Study 46:120–139CrossRefGoogle Scholar
  44. White GC, Burnham KP, Anderson DR (2001) Advanced features of program MARK. Proceedings of the second international wildlife management congress: Wildlife, land and people: priorities for the 21st century, pp 368–377Google Scholar

Copyright information

© The Author(s) 2014

Open AccessThis article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.

Authors and Affiliations

  • L. I. E. Couturier
    • 1
    • 2
    Email author
  • C. L. Dudgeon
    • 3
  • K. H. Pollock
    • 4
  • F. R. A. Jaine
    • 2
    • 5
  • M. B. Bennett
    • 1
  • K. A. Townsend
    • 6
    • 7
  • S. J. Weeks
    • 5
  • A. J. Richardson
    • 2
    • 8
  1. 1.School of Biomedical SciencesThe University of QueenslandSt. LuciaAustralia
  2. 2.Climate Adaptation FlagshipCSIRO Marine and Atmospheric ResearchDutton ParkAustralia
  3. 3.School of Veterinary ScienceUniversity of QueenslandGattonAustralia
  4. 4.Department of BiologyNorth Carolina State UniversityRaleighUSA
  5. 5.Biophysical Oceanography Group, School of Geography, Planning and Environmental ManagementThe University of QueenslandSt. LuciaAustralia
  6. 6.School of Biological SciencesThe University of QueenslandSt. LuciaAustralia
  7. 7.Moreton Bay Research StationThe University of QueenslandDunwich, North Stradbroke IslandAustralia
  8. 8.Centre for Applications in Natural Resource MathematicsThe University of QueenslandSt. LuciaAustralia

Personalised recommendations