Overfished and under conserved: life-history, ecology and supply chain of the Endangered whitespotted whipray (Maculabatis gerrardi) and sharpnose whipray (Maculabatis macrura) from south-east Asia

Sourcing animals and collecting biological data

Seventy four Maculabatis spp. (animals matching the description of both Maculabatis gerrardi and Maculabatis macrura; the only two species within the genus in Indonesia and Malaysia with distinctive white spots on their dorsal disc, pelvic fins and tail (Last et al. 2016a)), were sourced from a private seafood supplier in Singapore between February 2022 and April 2023. An additional 19 whole animals were sourced from Jurong Fishery Port in Singapore during this time, increasing the total sample size to 93 animals. Animal ethics approval was not necessary as animals were sourced following mortality from commercial fishing gear. For each animal, the import and/or catch location, as well as details of the fishing gear used in their capture were attained from the seafood supplier where possible. Each animal was photographed, sexed, weighed, and their disc width (DW), total length (TL), internasal width, head length, and tail length taken to the nearest mm as described by Last et al. (2016a). Animals were then assigned as either M. gerrardi or M. macrura according to morphological traits (see below) and dissected, and maturity was assessed according to gonad stage (Table 1). The liver, stomach, and a section of thoracic vertebrae were removed and stored frozen until further processing. Two additional vertebrae were also obtained directly from merchants processing large rays at Jurong Fishery Port in March and June 2023. These animals were not collected but their DW, sex and gonad stage was recorded. To supplement data on reproduction, data on gravid females encountered during routine surveys at Jurong fishery port in Singapore were collected; including the size of the mother and the number of embryos (determined by removing embryos through the cloaca).

Determining species

Morphology was used to determine whether animals were M. gerrardi or M. macrura. The morphological characteristics used to distinguish between these two species, per Last et al. (2016a), were ratios of: (1) total length to disc width; (2) internasal width to disc width; (3) head length to disc width; and (4) tail length to disc width (see Supplementary Material S1), although this last measurement is considered least reliable as some animals have cut or amputated tails, either by fishers, traders, or predators. Therefore, if animals matched at least two of the first three criteria for a particular species, they were assigned as that species. If animals matched one or fewer, they were not assigned a species but remained ‘undetermined species of Maculabatis spp.’ In the context of this study, ‘undetermined species’ refers to animals that could be either M. gerrardi or M. macrura; not the other species in the Maculabatis complex.

Stomach content analysis

Stomachs were excised and prey identified to the lowest taxonomic level possible (species, genus, family, or above). Contents suspected to be bait (e.g. portions with straight-edged cuts, attached to hooks) as well as sand/mud or rock (which are not prey but likely incidentally ingested) were excluded from further analysis (Jabado et al. 2015). Although some studies exclude indigestible parts from such analysis (e.g. shells, otoliths, and cephalopod beaks) (Potier et al. 2007; Bornatowski et al. 2014; Dicken et al. 2017), as they are not considered nutritionally valuable, they were included in this study as they were often the only identifiable parts of prey (Buckland et al. 2017). The percent frequency of occurrence (%FO), which is the proportion of individuals with a particular prey item, was calculated. To examine similarity and differences in diet between species, maturity, and sex, stomach contents were analysed using the Bray-Curtis coefficient (20 stress runs) and ADONIS (significance P < 0.05) were performed using the Vegan package (version 2.6-4, J. Oksanen et al., see https://CRAN.R-project.org/package=vegan) on R-studio (ver. 1.2.5042). The software package PRIMER v6 (Clarke and Gorley 2006) was used to analyse similarity percentages (SIMPER) to examine where specific differences occurred.

Vertebral processing and age and growth validation

Vertebrae were sectioned and processed using methods described in Goldman (2005); tissue from vertebrae was removed with a scalpel, and five centra were sectioned, soaked in 5% sodium hypochlorite solution for up to 2 min (to remove remaining tissue), and then thoroughly rinsed with water and dried in an oven at 45–60°C until dry. The two largest centra were chosen and attached to a microscope slide with crystal bond adhesive glue and a heat pad set at 250°C. The centra were then sanded down using fine grain sandpaper (grit size 400CW) submerged in fresh water, until the middle of the centra was reached. The centra were then unglued from the microscope slide, reversed and re-attached, and the other side of the centra sanded down until a ~600-μm section at the middle of the centra remained. The centra were then examined and photographed under a dissecting microscope, and translucent and opaque bands (band pairs) were counted from the birthmark (Fig. 1); identified by a change in the angle of the corpus calcareum (age 0) (Caillet 2015). Annual deposition of band pairs was assumed, based on validation in other tropical whiptail stingrays (Jacobsen and Bennett 2010) and this species living in a constant tropical environment with multi-annual depositions typically attributed to seasonal temperature changes (Smith 2013). Microsoft Powerpoint was used to adjust the saturation of vertebrae images to digitally resolve band pairs, and two readers determined age independently. A third independent reader helped to address discrepancies in the counts of band-pairs between the two initial readers. Each vertebra was also scored according to its readability: 0 = unreadable; 1 = bands visible but difficult to interpret; 2 = bands visible but most bands difficult to interpret; 3 = bands visible but a minority difficult to interpret; and 4 = all bands unambiguous (McAuley et al. 2006). Vertebrae with un-interpretable band counts were excluded from further analysis. The von Bertalanffy growth function (VBGF); (von Bertalanffy 1938), the logistic function, and the Gompertz function (Ratowsky 1990) were used to estimate growth parameters with growth curves generated in R-studio.

Fig. 1.

Vertebral section from a 11-year-old female Maculabatis spp. measuring 786 mm DW.

Interview about the fishery and species

We interviewed a private Singapore-based supplier of 74 of the stingrays through a semi-structured interview about the supply chain and perception of changes in the fishery over time (Supplementary Material S2). The interview was conducted in English, following human ethics guidelines, with no remuneration. Questions focused on: (1) the fisheries harvesting these rays; (2) trends in species harvested; (3) the supply chain; (4) the market; and (5) perceptions regarding future management. Questions were open-ended to encourage the supplier to express their own opinions. This research was undertaken with informed consent of those being interviewed under human ethics application H8683 as well as WWF’s safeguards for engaging with stakeholders (WWF ESSF 2023).

Species composition

Of the 95 Maculabatis spp. in the sample, five were caught in Malaysia and 90 were caught in Indonesia. Eighty five individuals could be identified to species, with the majority of animals identified as the whitespotted whipray (M. gerrardi, n = 45 [20 females, 25 males]), followed by the sharpnose whipray (M. macrura, n = 40 [19 females, 21 males]), and the remainder could not be determined (Maculabatis spp. n = 10 [4 females, 6 males]). Overall, there were more immature (n = 57) than mature (n = 38) individuals in the sample. The size-frequency distribution (Fig. 2) shows that the majority of the sample was dominated by individuals under 700 mm DW (mean DW = 520 mm).

Fig. 2.

Size-frequency distribution of the whitespotted whipray (Maculabatis gerrardi, n = 45), the sharpnose whipray (Maculabatis macrura, n = 40)), and Maculabatis spp. of undetermined species (n = 10) caught from fisheries in Indonesia (n = 89) and Malaysia (n = 5). The smallest individual was an M. gerrardi of 194 mm disc width (DW), and the largest individual was a Maculabatis spp. (undetermined species) of 960 mm disc width (DW). The entire sample had a mean size of 520 mm DW.

Diet analysis

Of the total sample size (n = 95), the stomachs of two individuals were not retrieved, and 12 had empty stomachs, leaving 81 individuals for further dietary analysis. Based on %FO, crustaceans are the most important prey item for Maculabatis spp (Fig. 3). ADONIS revealed no significant difference in proportions of broad prey groups between sex (P = 0.27) or between species (P = 0.93). However, when analysing all Maculabatis spp. together, ADONIS revealed significant differences in the diet of mature and immature individuals (P = 0.002). SIMPER revealed that this difference arose from mature individuals consuming significantly more gastropods (P = 0.001) and fish (P = 0.008) compared to immature individuals, which consumed more crustaceans although the latter was not significant (P = 0.18). A more detailed breakdown of diet can be found in Supplementary Material S3.

Fig. 3.

Percent Frequency Occurance (%FO) of prey for 81 Maculabatis spp., including 42 whitespotted whipray (Maculabatis gerrardi), 32 sharpnose whipray (Maculabatis macrura), and seven undetermined species of Maculabatis spp. Analysis found no significant differences between the diets of M. gerrardi and M. macrura.

Maturity and age-growth analysis

Males matured at smaller sizes than females for both M. gerrardi and M. macrura (Table 2; Fig. 4). Males and females exhibit a similar length-weight relationship, although females in the sample obtained larger sizes and weights (Fig. 5). Of the total sample size (n = 95) the vertebrae of 89 individuals (94%) yielded readable age bands. The oldest agreed age (between the readers) from the study for this species was 15 years old for female undetermined species of Maculabatis spp. (960 mm DW) and a male M. gerrardi (725 mm DW).

Fig. 4.

Proportion by size range for immature and mature whitespotted whipray (Maculabatis gerrardi), sharpnose whipray (Maculabatis macrura), and undetermined species of Maculabatis spp.

Fig. 5.

Length-weight relationship for female (●, n = 33) and male (, n = 30) Maculabatis spp. for which weight was recorded.

MCMC analysis (Supplementary Material S4) revealed that out of several potential growth models, the Logistic model was best when analysing M. gerrardi alone (k-value = 0.65 per year); the Von Bertalanfy was best when analysing M. macrura alone (k-value = 0.18 per year); and the Gompertz model was best performing when analysing all Maculabatis spp. together (k-value = 0.3 per year; Fig. 6). When analysing M. gerrardi alone, males and females had the same growth rate (k = 0.6 per year; Table 2), but males matured earlier (50% age-at-maturity = 5.07; Table 2) than females (50% age-at-maturity = 6.96; Table 2). When analysing M. macrura alone, males grew faster than females (k = 0.58 per year vs k = 0.17 per year; Table 2), but males matured later than females (50% age-at-maturity = 6.36 vs 50% age-at-maturity = 6.00; Table 2). When analysing all Maculabatis spp. together, males grew faster and matured earlier (k-value = 0.48 per year, 50% age-at-maturity = 5.88; Fig. 7; Table 2) than females (k-value = 0.1 per year, 50% age-at-maturity = 7.00; Fig. 7; Table 2). Fig. 8 visualises each individual’s disc width, age, and maturity.

Fig. 6.

Age-growth curve for 89 Maculabatis spp.: whitespotted whipray (Maculabatis gerrardi, n = 43), sharpnose whipray (Maculabatis macrura, n = 38), and undetermined species of Maculabatis spp. (n = 8) using vertebral band counts and the MCMC analysis performed using bayesian (blue line) and frequentist (green line) models. Circles (●) represent individual Maculabatis spp. with light shading indicating the 95% confidence intervals. A maximum age of 25 years was assumed for the genus.

Fig. 7.

Logistic generalised linear models (GLMs) of estimated ages of (a) male and (b) female Maculabatis spp. showing predictions of maturity at a given age. When all Maculabatis spp. are analysed together, the model predicts a 50% age-at-maturity of 5.88 years for males, and a 50% age-at-maturity of 7.00 years for females.

Fig. 8.

Largest ovarian egg diameter by month for mature female (n = 9) Maculabatis spp. for which ova diameter was recorded, showing both gravid (▲) and non-gravid (●) females.

Reproductive analysis

There were two gravid females in this study. One was an M. gerrardi (11 years old, 680 mm DW), which carried two embryos in January (180 mm DW each; 26% the size of the mother). Considering the two embryos were well-developed at 180 mm DW, and the smallest ray provided from the fishery was 194 mm DW, this suggests length at birth likely falls between 180 mm and 194 mm DW for this species. The second gravid female was an undetermined Maculabatis spp. (4 years old, 616 mm DW) at a very early stage (implanted egg) in June. Observed gravid females at Singapore’s fishery ports (Table 3) reveal well-develop embryos in various months throughout the year. In addition, the nine mature Maculabatis spp. in this study for which ova diameter was recorded, showed various ovarian egg diameters across months. These observations suggest reproduction for these species is asynchronous (Fig. 8).

Interview with the supplier

The supplier has fished with their father since 7 years of age, and was involved in the seafood business since 16 years of age (~40+ years in this business).

The fishery and supply chain

The supplier’s Maculabatis spp. are sourced from fisheries that use three primary methods: (1) rawai (bottom longline); (2) gillnets; and (3) 兄弟钩 in Mandarin/Hokien, which translates to ‘brother hooks’ (unbaited hooks are attached every inch along a straight line and dragged along the ocean floor). The latter are specifically for catching bottom-dwelling species and became widely used after the value of stingrays increased in the 1980s. Fishers know where to find Maculabatis spp., and animals are targeted, particularly Maculabatis spp. and Himantura spp. and other stingrays including Pastinachus spp. The bait used on longlines ranges from squid and herring (more costly) to various bycatch (a.k.a. ‘trash’) fish and eel flesh (less costly). Eel flesh is tough and stays on hooks for a long time. The longlines used to catch stingrays can range in length from 200–300 m to 2–3 km. While many longline fisheries leave their gear out overnight, those targeting stingrays tend to check the gear more frequently (i.e. every 3–4 h) so the catch is fresher. Depending on when the stingray gets caught on the line, they may be dead or alive when hauled in. When stingrays are hooked and experience stress, sea lice (copepods) are attracted to them and parasitise them from the inside (copepods were observed on some dissected specimens in this study; on the liver and stomach). The supplier stated that no Maculabatis spp. are released as fishermen retain everything. There is no seasonality to catches of Maculabatis spp., (‘the stingrays are always there’), but during the monsoon seasons boats do not go out as much.

The market for Maculabatis spp.

The market for stingrays in Singapore boomed in the 1980s. The main buyers of Maculabatis spp. are hawker centres (open-air complexes that house many food stalls) and wet markets (the latter of which eventually supply to hawker centres or sell direct to buyers to cook at home). At hawker centres, they are used for the local delicacy ‘BBQ stingray’ or ’Sambal stingray’, known in Bahasa Malaysia as ‘ikan pari bakar’. Maculabatis spp. and Himantura spp. are a preferred species for this dish due to flesh quality and general availability. Maculabatis spp. and Himantura spp. have a high value, selling up to around SGD10 per kg (~USD7) wholesale and up to SGD18–20 per kg (USD13–14) retail. Smaller animals of these species (<6 kg) are preferred for BBQ/Sambal stingray, while larger specimens of these species are used for Asam Pedas (a Malay curry). Stingrays have become a mainstay for the supplier’s business, accounting for 10–15% of business income.

Population trends and management

The supplier reports that Maculabatis spp. are primarily found on mudflats and have suffered a noticeable decline in a short timespan (i.e. during their 45-years in this industry) although it was noted that Himantura spp. ‘disappeared’ before Maculabatis spp. did. The stage where it is no longer ‘economical’ to catch Maculabatis spp. is nearing (i.e. high effort to catch the animal, boats spending too long at sea, travelling farther). Fishermen source from all over the region: including from Banga Belitung to Kalimantan and even Papua; which now has cold rooms so rays can be stored and kept fresh until transfer. When/if Maculabatis spp. is no longer economical to catch, the supplier predicts that fishers will switch to alternative species such as Neotrygon spp. (which is less preferable and currently sells for less than Maculabatis spp. in Singapore). When asked ‘what needed to be done to help [Maculabatis spp.],’ the supplier replied ‘a total fishing ban’ but that this would be detrimental to fishers. While they acknowledged the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES) would help Maculabatis spp., through trade regulation, they mentioned that the species is mainly traded between and consumed in Indonesia, Malaysia and Singapore, and so the trade may be too small-scale for CITES. The supplier also highlighted that any measure to protect Maculabatis spp. should apply to other stingrays, other marine animals (which are also in a terrible state), and the wider environment: ‘we shouldn’t play ‘god’ and protect some and not others…helping [Maculabatis] gerrardi is a starting point, but really everything needs help. Progress is too slow’. They believe only the younger generation will respond to a consumer campaign to stop eating stingray.