Observations on predation of the flatback turtle Natator depressus by the saltwater crocodile Crocodylus porosus at a major rookery in northern Australia

Introduction

Predation constitutes a significant ecological pressure across all life stages of sea turtles. While adult and large juvenile sea turtles typically exhibit low predation rates due to their substantial size and effective predator-avoidance strategies (Heithaus 2013), incubating eggs and newly emerged hatchlings on sandy beaches remain highly susceptible to a diverse array of predators, e.g. ghost crabs, goannas, foxes and birds (Lei and Booth 2017; Avenant et al. 2024; King et al. 2024). Among these, the saltwater crocodile (Crocodylus porosus) is hypothesised to represent a notable predator of nesting female turtles and their early life stages. The geographical distribution spans northern Australia to the east coast of India (Patro and Padhi 2019; Fukuda et al. 2022; Lloyd-Jones et al. 2023) and thus C. porosus exhibiting considerable range overlap with several sea turtle species, particularly at mainland and island nesting beaches. The recovery of C. porosus populations in northern Australia, following legislative protection since the 1970s, has restored some populations to near pre-exploitation levels after extensive hunting for their skins between the 1940s and 1970s (Fukuda et al. 2021). This resurgence potentially amplifies predation pressure on co-occurring sea turtle populations.

The predatory efficacy of C. porosus is underpinned by a suite of physiological and behavioural adaptations, including acute sensory capabilities (vision, olfaction, gustation, and hearing), streamlined morphological characteristics facilitating concealment in aquatic environments, and exceptional agility for short-distance pursuits (Caldicott et al. 2005). The dietary breadth of C. porosus is considerable, with adults preying on large vertebrates such as water buffalo, pigs, birds, turtles, and snakes (Adame et al. 2018), and juveniles primarily consuming crustaceans and insects (Isberg 2007). The spatial and temporal predictability of sea turtle nesting may render turtles a significant food resource for crocodiles in regions of sympatry. Nesting turtles, characterised by slow, awkward terrestrial locomotion and a trance-like state during oviposition, are particularly vulnerable to predation by large crocodiles. Conversely, juvenile crocodiles are more likely to target hatchlings during their emergence or excavate nests to consume eggs (Whiting and Whiting 2011).

Although predation of sea turtles by C. porosus, the American crocodile (C. acutus), and the American alligator (Alligator mississippiensis) has been documented (Ortiz et al. 1997; Nifong et al. 2011; Whiting and Whiting 2011), such observations remain infrequent, and the broader ecological impact of crocodilians on sea turtle populations is inadequately elucidated. We aim to provide a detailed examination of predation by C. porosus on the hatchling and adult life stages of the flatback turtle (Natator depressus), an Australian endemic, at a major nesting rookery in northern Australia. Through this investigation, a more comprehensive understanding of crocodilian-sea turtle interactions is sought, with implications for the conservation of N. depressus in the context of recovering crocodile populations.

Materials and methods

Study area

The investigation was conducted at Cape Domett (14°48.1′S, 128°24.5′E), situated at the entrance to Cambridge Gulf, ~110 km north-northwest of Kununurra in the East Kimberley region of tropical northern Australia (Fig. 1a). Characterised by a tidal range exceeding 7 m during spring tides, the area features an extensive intertidal zone and relatively turbid waters. The primary nesting beach at Cape Domett, spanning 1.9 km (Fig. 1b, c), supports high-density nesting of the flatback turtle (N. depressus), a genetically distinct stock comprising ~3250 nesting females (Fitzsimmons et al. 2020). Nesting occurs year-round, with peak activity in August–September, and no other turtle species have been recorded at this site (Whiting et al. 2008). Flatback turtles are medium-sized sea turtles, with a curved carapace length of ~90 cm and a body mass ranging from 75 to 85 kg (Limpus 2007). The region also serves as critical habitat for the saltwater crocodile (C. porosus), with seasonal creeks, tidal mudflats, floodplains, mangrove wetlands, and sandy beaches facilitating their presence within Cambridge Gulf (Cresswell et al. 2011; Semeniuk et al. 2011). C. porosus is recognised as a predator of sea turtles (Whiting and Whiting 2011).

Fig. 1.

(a) Map of the study area, illustrating the location of the Natator depressus rookery at Cape Domett, northern Australia. (b, c) Panoramic photographs of the ~1.9 km nesting beach where observations were conducted. Map adapted from NASA (2010). Photographs by C. Avenant (b, c).

Results

Observations at Cape Domett revealed frequent presence of saltwater crocodiles (C. porosus), with individuals observed drifting in shallow waters or basking on the beach throughout diurnal periods. Data collected through beach patrols, drone surveys, and infrared videography documented predation and scavenging events involving both hatchling and adult flatback turtles (N. depressus). These methods provided comprehensive evidence of crocodile activity, including direct predation during daytime, scavenging incidents, and nocturnal hatchling predation by juvenile crocodiles.

Over 35 days of observations across three nesting seasons (7, 14, and 14 days in 2021–2023), morning beach patrols recorded an average of 2.4–4.6 crocodile tracks per day, 0.1–0.9 crocodiles basking on the shore, and 0.8–1.1 crocodiles in adjacent waters (Table 1). A notable drone survey on a single morning in 2023, conducted independently of track counts and post-nesting surveys, identified 11 crocodiles (~2.5–5 m in length) basking along the 1.9 km nesting beach, indicating significant crocodile presence.

Five distinct predation and scavenging events were documented (summarised in Table 2).

Ambush Predation (10 August 2022): At sunrise, a large crocodile (~5 m) was observed drifting 10 m offshore, evidently targeting a female turtle attempting to nest. No additional turtle tracks were recorded, suggesting a solitary nesting attempt. As the turtle crawled toward the water following an unsuccessful nesting effort, the crocodile approached the shoreline, positioning itself behind breaking waves (Fig. 2a). Upon the turtle’s entry into the water, the crocodile executed a lateral lunge, seizing the turtle’s front left flipper. Minimal resistance was observed, likely due to the turtle’s exhaustion. The crocodile then swam with the turtle beyond a rocky promontory, rendering the outcome (consumption or escape) indeterminate.

Fig. 2.

(a) Moments preceding an attack by a ~5 m saltwater crocodile (Crocodylus porosus) on a female flatback turtle (Natator depressus) reentering the water following an unsuccessful nesting attempt. (b, c) A ~5 m crocodile adjacent to a turtle carcass. (d) Turtle carcass exhibiting a crushed carapace and multiple puncture marks from crocodile teeth. (e) Tracks and slide marks in the sand indicating the site of a crocodile attack on a turtle. (f) Evidence of the turtle being dragged toward the water. Photographs by J. Gee (a–f).

Carcass Observation (3 August 2022): In the afternoon, a deceased turtle exhibiting a crushed head, carapace, and puncture marks was recorded at low tide (Fig. 2d). A large crocodile (>5 m) was positioned adjacent to the carcass, holding the turtle’s head in its jaws (Fig. 2b). The crocodile retreated to the water upon observer approach, leaving the carcass (Fig. 2c), which was absent by the following morning, likely scavenged or removed by tidal action.

Terrestrial Attack (7 August 2023): Evidence of a terrestrial predation event was identified through converging crocodile and turtle tracks near the beach crest. Disturbed sand, carapace fragments, and turtle entrails indicated a violent encounter (Fig. 2e). Tracks returning to the ocean, accompanied by drag marks from the turtle’s carapace, suggested the crocodile transported the turtle seaward (Fig. 2f). It is unclear whether the crocodile located the turtle by following the track or intercepted the turtle while crawling back to the water.

Prolonged Scavenging (9–14 August 2023): A turtle carcass, initially observed without visible injuries below the high tide line on 9 August, was monitored over 6 days. Dingo (Canis lupus dingo) tracks were noted on 10 August, but no feeding occurred (Fig. 3a). A crocodile (3.5 m in length) was observed adjacent to the carcass by 13 August. It had apparently dismembered the carcass, removing the head and rear flippers, and was recorded exerting force on the posterior section of the carapace, resulting in an audible crunch (see Fig. 3b, c). Subsequent inspections revealed progressive degradation, with only the headless anterior half remaining by 14 August, surrounded by crocodile slide marks (Fig. 3d, e). On 15 August, the same crocodile was observed performing ‘death rolls’ to dislodge flesh, with the carcass reduced to a fragmented carapace attached to the plastron (Fig. 3f). Two additional crocodiles (3 m) were noted nearby, one in the shallows and one at the waterline.

Fig. 3.

Sequential documentation of a saltwater crocodile (Crocodylus porosus) scavenging a flatback turtle (Natator depressus) carcass over several days. (a) Carcass initially observed without visible external damage. (b, c) Carcass being consumed by a large crocodile days later. (d) Approximately three-quarters of the carcass remaining the following day. (e) Half the carcass remaining, surrounded by crocodile tracks in the sand. (f) Large crocodile performing a ‘death roll’ on the remains, 6 days after initial observation. Photographs by C. Avenant (a–f).

Hatchling Predation (12–14 August 2023): Infrared videography on 12 August captured a juvenile crocodile (<2 m) preying on hatchlings emerging post-midnight, concurrent with night heron activity. The crocodile consumed at least five hatchlings over 1.5 h, moving intermittently while herons maintained their distance (Fig. 4a). On 13 August, a subadult crocodile (~2.5 m) was observed pursuing hatchlings, capturing at least two with head-lifting motions to facilitate swallowing (Fig. 4b). Impressions in wet sand marked termination points of hatchling tracks at the crocodile’s jaws (Fig. 4c). Similar predation by a comparably sized crocodile was noted on another morning, with tracks leading to high-density nesting areas (Fig. 4d). Drone imagery on 14 August revealed extensive crocodile tracks across the beach, indicative of hatchling-targeted hunting, though the number of contributing crocodiles remains uncertain (Fig. 5a, b). It remains uncertain whether the tracks were produced by a single crocodile or multiple crocodiles.

Fig. 4.

(a) Nocturnal predation of emerging flatback turtle (Natator depressus) hatchlings by a juvenile saltwater crocodile (Crocodylus porosus) and night herons. (b) A crocodile pursuing hatchlings near the water’s edge. (c, d) Converging hatchling tracks and crocodile slide marks on the beach, indicative of predation activity. Photographs by C. Avenant (a) and J. Gee (b–d).

Fig. 5.

(a, b) Multiple tracks illustrating the search patterns of juvenile saltwater crocodiles (Crocodylus porosus) at the high tide mark, above which flatback turtles (Natator depressus) typically nest, suggestive of hunting activity targeting emerged hatchlings. Photographs by C. Avenant (a, b).

Discussion

The predation of flatback turtles (N. depressus) by saltwater crocodiles (C. porosus) at Cape Domett, as documented in this investigation, provides a comprehensive account of predator-prey dynamics at a high-density flatback turtle nesting beach in northern Australia. These findings contribute to and extend the limited body of literature on crocodilian predation of sea turtles (Table 3; Hirth et al. 1993; Ortiz et al. 1997; Sutherland and Sutherland 2003; Leis 2008; Whiting et al. 2008; Nifong et al. 2011; Whiting and Whiting 2011; Koeyers et al. 2015). The observed frequency of adult turtle predation (approximately one event per week) and hatchling predation during emergence events at Cape Domett aligns with patterns reported at other Australian sites, such as Crab Island and Cape van Diemen, where C. porosus similarly targets N. depressus and L. olivacea adults (Sutherland and Sutherland 2003; Leis 2008; Whiting and Whiting 2011). The integration of direct observations, unmanned aerial vehicle (drone) surveys, and infrared videography elucidated specific behaviours, including stalking, scavenging with ‘death rolls,’ and juvenile crocodile predation on hatchlings triggered by concurrent night heron activity. These observations underscore the opportunistic and ambush strategies of C. porosus (Caldicott et al. 2005), and highlight the heightened vulnerability of turtles during nesting and hatchling emergence phases.

Crocodilian predation on sea turtles appears to be geographically widespread, encompassing C. porosus, C. acutus, and A. mississippiensis across regions including Australia, Papua New Guinea, Costa Rica, and the United States (Table 3). Adult turtles are frequently targeted at the water’s edge or during nesting, with evidence such as punctured carapaces, severed limbs, and intersecting tracks corroborating predation events (Ortiz et al. 1997; Sutherland and Sutherland 2003; Nifong et al. 2011). Scavenging, as observed at Cape Domett, is also documented for A. mississippiensis (Nifong et al. 2011), suggesting that crocodilians exploit both live and deceased turtles. Hatchling predation, consistent with findings at Crab Island and Cape van Diemen (Leis 2008; Whiting and Whiting 2011), was frequently recorded at Cape Domett, whereas egg predation, reported solely at Cape van Diemen (Whiting and Whiting 2011), was not observed, potentially due to detection limitations or lower prevalence.

The recovery of C. porosus populations since the cessation of large-scale hunting in the 1970s, facilitated by protective legislation and expansive habitats, has likely intensified predation pressure at Cape Domett, where crocodiles are abundant alongside dense N. depressus nesting (Webb et al. 1978; Fukuda et al. 2021). This trend is consistent with other Australian turtle rookeries (Sutherland and Sutherland 2003; Whiting and Whiting 2011), whereas predation by C. acutus in Costa Rica (approximately two adults per month; Ortiz et al. 1997) and A. mississippiensis in the United States (Nifong et al. 2011) appears less frequent, possibly attributable to lower crocodilian densities or less sea turtle nesting activity.

The ecological implications of crocodilian predation for the Cape Domett N. depressus stock remain unclear due to insufficient quantification of predation rates across egg, hatchling, and adult life stages. While the loss of approximately one adult per week may exert minimal impact on the substantial Cape Domett population, smaller N. depressus rookeries with lower nesting densities could be more susceptible. Hatchling predation, observed during emergence events, likely affects a disproportionate number of individuals given their inherently low survival rates (Limpus 2007). Furthermore, predation by multiple predator species has been widely observed on sea turtle nesting beaches globally (Barton and Roth 2008; King et al. 2024), with nocturnal predation by night herons documented at Cape Domett, alongside evidence of egg predation by dingoes (C. Avenant, unpubl. data), suggesting cumulative pressures and complicating the predation dynamics.

The methodological framework employed in this study, combining beach patrols, infrared videography, and drone surveys, provides a robust approach for predation research, surpassing reliance on indirect evidence such as tracks or remains. Infrared cameras proved particularly effective in capturing unbiased nocturnal behaviours without disturbing wildlife, consistent with prior applications (Avenant et al. 2024; Avenant 2025). Drone surveys facilitated aerial documentation of crocodile tracks and size estimations, reinforcing their utility in behavioural and population studies (Ezat et al. 2018; Thapa et al. 2018). Future research integrating thermal drone imaging (Sellés-Ríos et al. 2022) and satellite tracking (Read et al. 2007; Mascarenhas-Junior et al. 2023) of crocodiles is recommended to further clarify movement patterns and predation strategies, thereby enhancing understanding of the frequency and spatial distribution of crocodile-turtle interactions.