Spotting the “small eyes”: using photo-ID methodology to study a wild population of smalleye stingrays (Megatrygon microps) in southern Mozambique

Study sites

Opportunistic photographs of the dorsal surfaces of free-swimming stingrays were collected year-round by local researchers, dive operators and recreational divers in three main regions (Zavora, Tofo Beach/Barra, and the Bazaruto Archipelago National Park) of the Inhambane Province from May 2003 to November 2018 (Fig. 1). Opportunistic photographs were also taken of two individuals landed intentionally or as by-catch during the study period.

Tofo Beach (23°85′S, 35°54′E) and Barra (23°80′S, 35°52′E) lie on the Southern Mozambican coast, on Ponta da Barra peninsula in Inhambane Province. Tofo Beach/Barra dive sites are comprised mostly of rocky reefs with associated corals rising from the sandy seabed. Many of the reefs serve as year-round cleaning stations for elasmobranchs, most notably both species of manta rays (Mobula alfredi and Mobula birostris). The majority of the reefs can be found at depths between 22 and 32 m.

Zavora (24°23′S, 35°43′E) is located on the same coast, 80 km south west of Tofo Beach. Zavora dive sites are very similar to those found in Tofo Beach/Barra with the additional presence of kelp on the deep reefs. Reefs in this region also support year-round cleaning stations for elasmobranchs and are typically situated between 12 and 35 m of depth.

The Bazaruto Archipelago National Park (BANP) is a 1,583 km² area encompassing five sand dune islands situated approximately 15 km from the coast, off the mainland city of Vilankulos (22°00′S, 35°32′E). Unlike the two other study sites, the Bazaruto Archipelago has more diverse tropical reef systems. Deeper reefs are similar to those in the south of the Province only with more coral cover, particularly soft corals. Shallow reefs comprise significantly more hard corals than in the south regions and the inside of the archipelago is mainly tidal sand flats with deep connecting channels to the ocean (South African Association for Marine Biological Research, 2008).

Photographic identification

M. microps are medium-brown in colour, with a longitudinal row of large white spots on either side of the disc at about two-thirds distance from the dorsal midline to pectoral-fin apex (Fig. 2). Additional large white spots are lateral to the eyes and on either side of the mid-disc. Several rows of small white spots are located on either side of the base of the tail and of the disc, at about one third distance from the dorsal midline to pectoral-fin apex (Pierce, White & Marshall, 2008).

In order to investigate the variability of these natural spot patterns and monitor the longevity of markings, an image of the entire dorsal surface of the animal was attempted at every encounter. This region became the standardized identification area (ID-area), as the use of the entire spot patterns area across the flattened dorsal plane allows for efficient visual comparison between individuals with a higher degree of confidence than any one region alone (Fig. 3A).

Photographs of this standardized ID region (ID-photos) were taken from still photographs, with or without flash, or from video screenshots, with or without video lights. Only photographs of good quality were used for this study (147 out of 178 photographs in the study set). These images were in focus, clearly showed the natural spot patterns of the ray, e.g., was not obscured by sediment or associating fish, and had enough resolution to clearly distinguish the details of the spot patterns. However, photographs with only one side of the ID-area visible (partial ID-photos) were still usable for re-sightings: if a partial ID-photo was matched to an initial ID-photo with at least 10 spots clearly visible, the partial ID-photo was accepted as a re-sighting (Figs. 3A and 3B). Once appropriate ID-photos were secured, the sex of the animal was determined through the presence or absence of male reproductive organs (claspers) located on the pelvic fins (Fig. 3C). Scars, bite marks and deformities were also noted to provide a useful secondary verification of an ID, thus bolstering confidence in the accuracy of this validation study.

Matching was performed by visual pairwise comparison (two pictures side by side on the same screen), by one or two researchers. Each photograph was compared at least twice to the entire photo-database. Photographs were matched visually with no ambiguity, e.g., all visible natural markings were matched prior to confirming a re-sighting.

New identifications were cumulatively plotted over time in a discovery curve to show the rate at which newly identified individuals were recruited into the population database (Fig. 4A). Re-sighting events were characterized by the positive identification of a previously identified individual more than 24 h after it was initially seen (e.g., not on consecutive dives) or at a different dive site. Rate of movement was calculated as the minimum time required to travel from one known location to another as confirmed by positive identification of individuals.

Re-sighting events allowed us to reveal some philopatric behaviors. Philopatry is defined as individuals frequently returning to or staying within their home ranges, birthplaces, or other specific localities (Chapman et al., 2015; Flowers et al., 2016). Two types of philopatric behavior were observed in this study: site fidelity, e.g., when the individual makes a long-distance movement away from a defined location and then returns to it (Chapman et al., 2015), and site affinity, e.g., when there is no available evidence to distinguish between site fidelity and residency, e.g., individuals remaining in a defined location for at least 12 months (Couturier et al., 2011).

To examine seasonal trends, the sightings per unit of effort (SPUE), e.g., the number of rays seen per dive and per month, was calculated for each year of the study period and then pooled by calendar month across all years. The means were compared across calendar months (Fig. 4B). Bottom Sea Temperature (BST) was recorded on each research dive with dive computers.

Ethics statement

No approval or permits regarding human or animal ethics were required for this study as this was an analysis of publicly contributed photographs, collected non-intrusively, from recreational scuba divers. All data/photographs collected by authors were collected with special permissions from Bazaruto Archipelago National Park by the ANAC (Administração Nacional de Áreas de Conservação) of Mozambique and the Maritime Administration of Inhambane province.

The use of photo-ID for M. microps

True validation of photo-ID ideally requires double marking (e.g., conventional tagging and photo-ID) of the animals in order to have a reliable and independent confirmation of re-sighted individuals’ identities. Like in many terrestrial (Del Lama et al., 2011; Jackson et al., 2006) and marine studies using photo ID (Bansemer & Bennett, 2011; Domeier & Nasby-Lucas, 2007), this work did not include any conventional double marking techniques to validate patterns on individuals. However, enough data were presented to convincingly demonstrate the accuracy of matches, due namely to the clear and variable spot patterns on this species.

This study suggests adequate variability exists in the natural spot patterns on the dorsal surface of M. microps to reliably study wild populations with photo-ID methodology. These naturally occurring spot patterns were stable and unchanging over periods of up to six years, suggesting they are permanent and fixed, at least after maturity. Based on the stingrays’ disc width (>1m), it is likely that most if not all of the individuals encountered and identified during this study were mature. Further research is needed to confirm if this species undergoes ontogenetic changes in colouration like in many orectolobiform sharks (Dingerkus, 1986). Additional monitoring would be required to confirm the long-term stability of these natural markings over the lifespan of the individual, which is currently unknown. Given the variability and stability of these natural spot patterns and the ease of which they can be captured, we recommend that photo-ID methodology may be a good starting point for future studies on wild populations of M. microps.

Furthermore, photo-ID suitability of this species makes it a good candidate for citizen science contributions. Both videos and still images provided high quality ID-photos. Videos often offered different angles of the animals and their pectoral disc, which proved useful in obtaining the best angle for a high-quality screengrab. Footage and photographs should ideally be taken from above the ray to see the full ID-area. Divers should abide by the following code of conduct: approaching slowly, staying 2–3 m away from the ray, and not chasing the animal. In this particular region, due to the wide-ranging nature of the species, their broad distribution along the coastline and increasing number of identified individuals, future research may explore the feasibility and accuracy of pattern matching algorithms (e.g., I³S or Wildbook®). Those may help to facilitate the development of local/regional databases and the incorporation of citizen science contributions.

Movement patterns and habitat usage

This study extends the known area of occurrence of M. microps in southern Mozambique and suggests that this species may be widespread in the south of the country. Opportunistic collection of photographs from divers helped establish connectivity along a 350 km stretch of coastline in the Inhambane Province, from Zavora in the south to Bazaruto Archipelago in the north. Individuals were found to travel long distances north and south along the coast over short and long periods of time, at least 200 kilometers in as little as 102 days, and 400 kilometers in 535 days which is the longest distance traveled by a dasyatid ray on record. The only similar numbers reported on a dasyatid ray’s movements range from 151 to 258 kilometers, distances traveled by four pelagic stingrays over the course of a 13-day satellite tag recording period (Weidner, Cotton & Kerstetter, 2014). The pelagic stingray however is likely to be more mobile than the smalleye stingray with suspected, but yet to be confirmed, seasonal migrations of hundreds to thousands of kilometers (Mollet, 2002).

M. microps’ ability to travel long distances bolsters the supposition that the species is semi-pelagic (Pierce, White & Marshall, 2008), providing further evidence that it rarely seems to rest on the seabed like most stingrays. Like manta rays, another large and highly mobile species, M. microps also utilizes a wide variety of habitats along the coastline. Furthermore, like manta rays in Inhambane they are most frequently encountered on small deep reefs (20–35 m), particularly those that are surrounded by sand and support cleaning stations for elasmobranch species (Marshall, 2008). Similar to manta rays (Marshall, Dudgeon & Bennett, 2011), some individual M. microps show site affinity to these preferred reefs along the coast.

Both male and female rays were encountered during this study suggesting this species does not segregate by sex. However, no young of the year or even juvenile rays were observed suggesting that either ontogenetic shifts in resource or habitat use is driving the segregation of smaller sizes classes from the adults, which is common among other ray species (Marshall, Dudgeon & Bennett, 2011; Yick, Tracey & White, 2011; Espinoza et al., 2015). Individuals were most often encountered alone, either swimming through or cleaning at deep water reefs in the region, suggesting that cleaning may be an important factor to their visits to these reefs. Visits to cleaning stations have been shown to be linked with social behavior, including mating: solitary, highly mobile species like manta rays appear to use these reefs as meeting points where males can more easily locate and court receptive mates (Jaine et al., 2012; Stevens, 2016). Juvenile M. microps may also have smaller parasite loads and not need to clean as frequently as adults, or they may not be driven to visit these areas for social reasons until they reach maturity.

As with many dorsally flattened rays, late stage or near-term pregnancies were easily distinguishable (Marshall & Bennett, 2010) by distended bellies and particularly backs (Fig. 3D). Only one caught pregnant female smalleye stingray of 206 cm DW was previously examined off the East Indian coast and contained a single, late-term male embryo of 33 cm DW (Nair & Soundararajan, 1976). With visibly pregnant females sighted in all three of the study regions in southern Mozambique, it is likely that this area is regularly used by females to gestate young and parturition may occur in the immediate area. The observed long-distance movement of a near-term pregnant female from Tofo Beach-Barra to Bazaruto Archipelago suggests that M. microps may move into the warmer shallower waters of the archipelago to give birth. However, with no records of juvenile M. microps being landed in artisanal fisheries (Andrea D. Marshall, pers. comm., 2018), it is equally likely that the depth range of this species extends into deeper water than can be accessed by recreational divers and females may be using deeper waters as pupping grounds.

While the discovery curve showed an increasing discovery rate since 2007, the relatively low sightings rate of this species indicates that it is rare, and that the population size off this coastline may be small. Increases in fishing pressure along this coast over the last decade have coincided with dramatic declines in other large ray species, like manta rays and devil rays (Rohner et al., 2013). M. microps are likely to be subject to similar pressures from targeted artisanal fishing and incidental capture in the gill and seine nets used extensively along the coast (Figs. 9A and 9B).

M. microps are large, charismatic, and easy to approach (Figs. 9C and 9D). As such, smalleye stingray encounters are highly sought after by divers (Steven Counsel, pers. comm., 2018). Like manta rays, which are attractions for diving tourism (Venables et al., 2016), it is likely that diving tourism is and will continue to be bolstered in this region by reliable encounters with this rare ray species. While the Bazaruto Archipelago National Park may offer a natural safe haven for these rays, it is recommended that management strategies be examined to increase protections of this lesser known, potentially threatened species throughout its range.