The State of Qatar is a 11,586 km² peninsula located halfway along the western shore of the Gulf between latitudes 24.27 and 26.10° N and longitudes 50.45 and 51.40° E. It is bordered by Saudi Arabia to the south, and shares maritime borders with Bahrain, Iran, and the United Arab Emirates. Qatar is characterized by its hot, arid climate in the summer and relatively cool winters. Its coastal water is hypersaline, resulting from lack of freshwater input, excessive evaporation rates with an annual average of 2200 mm, and sparse rainfall, creating an extreme environment (Abu Sukar et al., 2007). Our study site is in the northwest of Qatar, south of Ras Eshairij and north of Dukhan; an area called Umm AlMaa. This site was selected based on historical data (Preen, 2004), which recognized the area between Qatar and Bahrain as one of the most significant dugong habitats in the region. Furthermore, initial data, indicated multiple observation of substantial dugong herds during winter, detected through boats and helicopter surveys (Marshall et al., 2018), along with various sightings of significant groups reported by local fishermen. Based on these published observations, and anecdotal information, it was evident that dugongs gathered between Qatar and Bahrain from December to March.

Transect surveys were conducted by boats from the northern Hawar Islands to Ras Eshairij, during 2019-2020. The transect lines were predetermined, evenly spaced out at intervals of 2 km and input into a GPS device. Surveys proceeded along intended parallel transect lines that are perpendicular to the coastline and extended out to the 10m bathymetric contour of the Bahraini border (Figure 1). Two center console boats were used, each measuring 28 feet and able to accommodate 6 individuals. Similar to their Australian counterparts, dugongs in the Arabian Gulf are wary of people and frequently hide when loud or abrupt noises happen. As a result, the speed of the boat was continually checked to stay under 8 km/hr, usually fluctuating between 5.5 and 7.4 km/hr, with 2-minute pauses taken every 5 minutes to reduce engine noise as much as possible. Each observer employed the naked eye and Nikon 10x50 marine binoculars to scan directly from front to either port or starboard, depending on the side of the boat they were on. When dugongs were sighted, the GPS coordinates, the angle from the transect line, and an estimated distance to the group were noted. The boat remained at no less than half a kilometer while ideally keeping a speed close to that of the herd. Modern, innovative state-of-the-art UAVs (hexacopters) equipped with cutting-edge side-by-side infrared and high-definition cameras were deployed to capture images of entire groups and assess the counts of individuals and cow-calf pairs.

In 2019, three boat-based drone surveys were conducted between January and February, during which a large dugong herd was recorded only once, on the 7th of February (Table 1). In 2020, survey effort was increased to eleven surveys covering a wider time range including 9 surveys from February to March, one in August and one in November, to investigate if dugong aggregations happen outside the previously noted peak period. Surveys took place in the morning to prevent strong sun glare typically at its peak around noon and when wind speed was less than 6 knots. Moreover, to enhance water visibility, all surveys were conducted during moderate tide periods when the water’s turbidity was low, and the sea appeared calm and glassy. Each of these factors was considered to enhance the chances of spotting when the images were captured.

Boat-based drone surveys were conducted with a DJI Mavic 2 Pro, featuring a 1-inch CMOS sensor and 28mm focal length. The drone utilized was equipped with a built in, high-resolution camera, recording a 4k video at 3840x2160 pixels. To reduce motion blur, the camera was adjusted to shutter speed priority with automatic exposure enabled. Additionally, flight information, such as GPS coordinates and altitude, was captured and noted. The UAV was deployed from height of the boat and steadily climbed, recording video during its journey. The entire herd became completely visible in the video frame at an elevation of about 80 meters. For five minutes, the UAV flew straight above the dugong group, recording continuous video with the camera directed downward. The drone’s time, location, and altitude were recorded during each flight using data from the GPS stream. Video recordings were examined manually to select frames that showcased the entire gathering in the footage. In 2019, merely 4 high-resolution images were obtained for examination and the groups count. In 2020, 32 high-resolution images of the dugong aggregation were acquired, processed and used for image analysis. These screenshots were not extracted at regular intervals; instead, they were deliberately taken only when the entire herd was visible within a single video frame.

All images and videos gathered during the drone surveys were analyzed on a 27-inch, LCD screen. To improve the detection of dugongs in drone images and enable more accurate group counts, multiple image adjustments were implemented. Color correction was applied in the images to enhance analysis and reduce the likelihood of misidentifying dugongs (Figures 2A, B). A magenta-purple color filter was created by modifying the white balance, effectively neutralizing the prevalent blue-green tones of the water. Saturation and contrast were enhanced to highlight the dugongs’ natural hues and enhance their separation from the background (Figure 2). Additionally, localized brightness adjustments were applied to brighten the central area with the highest density. These adjustments together enhance the subject visibility and counting accuracy.

Dugong individuals as well as proportion of dependent calves were counted manually in each image (Figure 3B). Mother-calf pairs were identified as 1) calves are significantly smaller than the mother, 2) the two dugongs swim in echelon position, in proximity of each other, 3) synchronous behavior in surfacing and direction and duration of association (Figures 3C, D). While calves are often undercounted in boat-based drone surveys, the accuracy of the count was greatly enhanced by the assurance that they were observed in clear water and good weather conditions Consequently, no adjustments were implemented for changing weather conditions. (Marsh et al., 1984) claims that vertical boat-based drone images of dugongs taken in ideal weather conditions that improve water visibility, yield the most precise calf counts. To facilitate counting, a grid and alphanumeric rows and columns were superimposed on each image (Figures 2B, 3A, B). Counts were noted on an analogous datasheet for each image. The criteria for counting dugongs (indicated with a dot) included: 1) dugongs surfacing in the water, 2) dugongs that were underwater showing a distinct elongated shape, with visible tail, 3) dugongs with part of their snout or tail visible above the surface, and 4) a submerged dugong exhibiting a clear contrast in color relative to the surroundings. The identification process was made easier by having several chronologically ordered images, with each second of the video depicted by one image. This approach allowed for tracking individuals over time to avoid false positives, like vague “dugong-like” forms, densely packed groups where individual identification was challenging, or erroneous benthic features. All images of the dugong aggregation were initially reviewed and annotated by a primary observer, who recorded the dugong counts and conducted a second review to verify their initial annotations. To ensure accuracy and consistency, the images were then independently analyzed by a second observer with extensive field experience in dugong identification and behavior.

Environmental parameters including temperature, salinity, dissolved oxygen (DO) and pH were measured in the dugong site. CTD-Diver^® (manufactured by Van Essen Instruments, model DI273) was used to measure temperatures daily, with 6 hours reporting intervals. One device was deployed in the dugong location at the west of Qatar (Umm AlMaa) while another one was deployed in the east of Qatar at Alad Algarbai (Figure 1), to compare temperatures at the two locations. The CTD-Divers^® were installed on-site for two years, from 2020 to 2022, recording monthly mean sea surface temperatures (SST°C). Additional environmental parameters, including salinity (ppt), dissolved oxygen (mg/L), and pH, were measured in the survey location using the EXO2 Sonde—a multiparameter water quality device. Measurements were taken during winter and summer at nine stations along the transect line. Data of wind speed (Km/hours) for the month of February in 2019 and 2020 was obtained from the Weather underground (2025), which was accessed on [July 31, 2025]. This data was used to understand the prevailing wind condition during the study period.

Our surveys were conducted using vertical drone imagery, high-resolution cameras, and calm, clear waters ≈3.5m in depth, which enabled strong visual contrast between dugongs and the surrounding seafloor. Using the dugongs counts from two observers’ protocol in a set of frames, two observers conducted independent counting using colored marking on images. Following the counting, numbers of shared detections (nA) was determined and in addition to detections by observer B not seen by observer A (nB). Then with the assumption that the total (nA+nB) approximates the total number of animal present, the estimate of observer A’s detection probability (p_d) will be:

To address the uncertainty in dugong availability, we utilized the depth-dependent framework suggested by Hagihara et al. (2018). While the Environmental Conditions Index (ECI1) assumes availability = 1.0 when the water is clear with a visible bottom, our survey circumstances did not consistently satisfy this criterion. Even though a significant portion of the bottom had favorable visibility, depths at the aggregation location often ranged from 3.5 to 5 m, exhibiting clear water but diminished bottom clarity. Consequently, we utilized the availability detection probabilities reported by Hagihara et al. (2018) for ECI3 conditions (p≈0.8-0.98), which represent environmental conditions and depth range (<5m) present in our study.

Data of measured environmental parameters for the study area are showing in Table 1, all values are reported as mean ± SD. Temperatures in the study area peaked during the summer months (35.5 ± 0.46 °C in August), while the lowest mean monthly temperature (18.4 ± 1.04 °C) was in January (Figure 4A). Temperatures exhibited slightly higher values in the west during winter and slightly lower values in summer months compared to the eastern region (Figure 4A). For example, the average monthly temperature in January in the dugong location (18.52 ± 1.25) was slightly higher than the average monthly temperature in the east of Qatar (18.37 ± 1.03). Similarly, in February, the average temperature in the gathering site (19.2 ± 1.1) was slightly higher than the average values in Alad Algarbia (18.9 ± 1.0). Notably, the year’s lowest temperatures, between January and February, aligned with the largest dugong aggregation in the northwest of Qatar. A comprehensive assessment of key environmental parameters in the northwest region of Qatar, analyzing a total of 463 samples (N = 463) for water parameters indicated that the average salinity was 48.1 ± 1.1 ‰, dissolved oxygen (DO) levels averaged 6.0 ± 0.7 mg/L and the pH levels had an average of 7.9 ± 0.3 (Table 1). Additionally, the average depth recorded across sampling sites was 3.5 ± 1.4 meters, providing insight into the shallow nature of the dugong gathering area. These findings offer a baseline for understanding the environmental characteristics of this region.

From 14 surveys conducted in 2019 and 2020, only two encounters for the large gathering were identified while in the other encounters, a small number was found. In agreement with reports from local fishermen and research by (Marshall et al., 2018; Preen, 2004), dugongs were found approximately 5km off the coast near the region from Ras Eshairij to Dukhan. They moved collectively as a single, large herd across the area between latitudes 25.97 and 25.88° N and longitudes 50.97 and 50.89°E. The dugong herd displayed no noticeable behavioral changes in response to drone operations, even when the drone approached within 5 meters above them. Their swimming behavior remained steady throughout the observation period, and no indication of disturbance or avoidance actions were observed. Dugongs were moving very close to each other, alternating surfacing and diving. Their movements were in very large aggregations and during the diving, we reported scattered distinctive feeding plumes indicator of active foraging behavior.

There were no large marine fauna other than dugongs observed in the area. The double observer approach yielded detection probability of 0.96, consequently, we assumed an observer bias of <0.05.

Boat based drone surveys conducted in 2019 and 2020 recorded maximum observed dugongs of 1108 and 1209 respectively (Table 1). Corrected herd size estimates for dugongs were derived from the maximum observed counts after correcting for detection probability (0.96) and availability (range 0.80-0.98) estimated to herd sizes from 1113 to 1443in 2019 and from 994 to 1574 in 2020 (Table 1). We applied a triangular distribution using the upper (1574) and lower (994) bounds and a calculated mode of 1176, resulting a bias adjusted mean estimate of 1248 ± 122 dugongs in the two years (Table 1). The proportion of calves in 2019 was approximately 5.8% whereas the proportion of calves in 2020 were almost doubled, 10.2% (Table 1). Analysis for frequency distribution of dugong raw counts from the analyzed 36 images is shown in Figure 5. About 72% of the images (n=26) contained between 1100 and 1209 dugongs (1100> n < 1209). Our methodology does not allow for definitive determination of the sex of individual dugongs. Nonetheless, it is reasonable to assume that a dugong accompanied by a calf is a female.

The current research documented the largest single aggregation of dugongs recorded globally in the waters of northwest of Qatar between 2019 to 2020, with a maximum observed number of 1209 and an estimated herd size of 1248 ± 122, after correction for detection and availability biases. The observation was conducted employing UAVs in a shallow water, with a visible seabed (Pollock et al., 2006). Notably, Marshall et al. (2018) had previously observed 508 dugongs in the exact same area, suggesting the group may have comprised as many as 600 individuals. While this was already considered as an exceptionally large sighting, our results demonstrated aggregations more than twice that size. This discovery is twice the previously noted largest single group of 700 dugongs near Hawar Island in Bahrain based on UAV surveys (Khamis et al., 2023). It is also two times the figures reported by Preen (2004), which documented groups of 674 and 508 individuals in that region. On the other hand, our figures are considerably higher than a large group reported by Hodgson (2009) that consisted of about 50 individuals in a manned aerial survey, which estimated the total population size in this study to be 1164 individuals after applying correction factors (Table 2). The gathering of such large numbers of dugongs are exceptionally rare even in Australia. Recent studies reported aggregations of ≤700 but not above thousand individuals at a single time, highlighting the significance of this area for conservation efforts. Dugongs’ aggregation of >1000 dugongs in a shallow marginal extreme basin in the northwest of Qatar is remarkable. The largest actual sighting of a single dugong group that was not in the Arabian Gulf, was about 300 dugongs in Moreton Bay, Queensland (Lanyon, 2003).

Surveys on marine mammals usually use different tools including UAVs, manned aerial surveys and boat-based observations, each of these methods has its limitations. It is quite possible that these reported dugong gatherings in the Gulf are a part of one Gulf population. Although dugongs traverse territorial boundaries, there is no data that sheds light on whether the Qatar-Bahrain and UAE dugongs comprise one or more populations, or if these groups migrate to and from UAE along the southern coast. Wider and more coordinated efforts are crucial to understand elements driving these gatherings, including comparative evaluations of biotic and abiotic aspects such as seagrass biomass seasonality at the dugong aggregation location versus nearby areas and the hydrodynamics of the open water during the gathering time. Moreover, studies to investigate the genetic and demographic connectivity to other groups in the Gulf region, like those noted in Bahrain and UAE are essential for making conservation decisions.

Our surveys indicated that mother-calf pairs comprised 5.8% and 10.2% in 2019 and 2020, respectively. The observed difference in calves’ numbers between years is probably due to a reduced sample size in 2019 (4 images only) when compared to 2020 (n=32), or it can be due to reduced human activity in the coastal waters during COVID time which would lower the number of calves lost in boat strikes and fishing nets. The ratio of mother-calf pairs in our research is consistent with findings from other studies including ratios reported by Marshall et al. (2018) 9.9% and 5.8-5.4% of two groups reported in February and December (2015), respectively. The ratios of the calves in the study by Khamis et al. (2023) of 6.4% were close to the ratios we reported in 2019. However, Preen (2004) found higher ratios reaching 15.7% of the dugong population in the Arabian Gulf during winter. The variation in calf ratios between different studies likely reflects differences in the geographic coverage and the duration over which each study was conducted, with Preen’s research spanning a broader area and longer timeframe. Variation can also arise from differences in the criteria used to classify an animal as a calf, as well as observer experience and survey methods. Our findings are consistent with calf proportions reported in other regions, which generally range from around 1.3% to 14.8% in areas like New Caledonia, Moreton Bay, Torres Strait, Shark Bay, and Cape Bedford (Garrigue et al., 2008; Hodgson et al., 2013; Lanyon, 2003; Marsh and Saalfeld, 1990, 1991). Higher calf ratios have also been recorded from other regions including the Solomon Islands (17.3%), near Shark Bayand Ningaloo Reef (21.4%) and in Johor, Malaysia (24%) (Bass, 2010; Hodgson, 2007; Ponnampalam et al., 2015).

There is no evidence that dugongs gather in large numbers for social hierarchy; instead, they probably come together to share resources like seagrass, find shelter, or seek suitable water temperatures (Marsh et al., 2011). This location in the northwest of Qatar encompasses a vast area of seagrass beds, ranging from 500 to 1,000 km², and provides a rich feeding ground for the Gulf’s dugongs (Erftemeijer and Shuail, 2012). A recent study found that a significant portion of this expanse contains dense seagrass beds (Butler et al., 2020), making it one of the largest seagrass habitats along the coastal waters of Qatar. It is considered the second-largest seagrass habitat in the Gulf after the UAE, making up about 1% of the total areas of seagrass around the world (Al-Mansoori and Das, 2024). It was established that dugongs actively feed on mixed seagrass habitats that include Halodule uninervis, Halophila stipulacea, and Halophila ovalis, as well as macroalgae (Marshall et al., 2018). These three opportunistic seagrass species thrive in the extreme environment of the Arabian Gulf due to their high tolerance for salinity and temperature fluctuations (Erftemeijer and Shuail, 2012). As pioneering species, they not only endure these harsh conditions but also exhibit rapid recovery, making them the most dominant seagrasses in the region. Additionally, they are among the most globally recognized primary food sources for dugongs (Al-Abdulrazzak and Pauly, 2017; Aragones, 1994; Bayliss and Freeland, 1989; D’Souza and Patankar, 2009; Lanyon, 2003; Ponnampalam et al., 2015). More specifically, the existence of Halophila uninervis was linked to a notable rise in dugong numbers in Australia (Said et al., 2025). Dugongs probably prefer these seagrass species due to their low fiber content, derived from low biomass strands. They are also abundant in readily available nutrients like starch and nitrogen, and they grow quickly (Marsh, 2018; Tol et al., 2016; Yamamuro et al., 2004).

Studies on the seasonal trends of seagrass in the Gulf are scarce, and their results are contradictory. Research along the coastlines of Saudi Arabia and Qatar indicated that seagrass experiences senescence or dieback during winter due to the powerful northwesterly Shamal wind that pushes surface cooler water toward Qatar eastern shores especially in February (Thoppil and Hogan, 2010), that disturb and break apart seagrass rhizomes, resulting in heightened water turbidity (Whitehead, 2015). On the other hand, research along the UAE coastline has demonstrated that seagrass vanishes from shallow areas (<5m) in summer, due to high temperatures (>34 °C) leading to leaf loss and forcing dugongs to seek food in deeper waters (Al-Mansoori and Das, 2024). Further studies showed no clear seasonal pattern in seagrass biomass and failed to link this pattern to changes in environmental elements such as temperature or wind patterns (Price and Coles, 1992). Preliminary findings from our ongoing research (not yet published) indicated that seagrass density in northwestern Qatar was high from November to March, with a peak biomass of approximately 55 g dry wt/m². This seasonal increase corresponds with the documented notable dugong aggregations in the area. In contrast, seagrass biomass decreased to around 17 g dry wt/m² at the onset of early summer (unpublished data, n.d). Previous studies have connected dugong aggregations to the peak growth seasons of seagrass (Aragones, 1994). Our novel observation of seagrass biomass in the aggregation site is similar to a maximum seagrass biomass of 54.2 g dry wt.m² reported from a recognized dugong feeding area in Australia, highlighting the significance of this location in Qatar as a feeding ground during this seasonal period (Westlake et al., 2022).

Hence, to conserve and manage the large dugong herds in the Gulf, additional research is needed to understand seasonal seagrass dynamics and their relationship with water temperature and physicochemical conditions, especially given that the Gulf supports one of the largest numbers of dugongs reported outside Australia.

In addition to the food availability indicated by seagrass biomass, other factors such as water temperature and water stability influence the presence and numbers of dugongs (Budiarsa et al., 2021). Dugongs possess slow metabolic rates and limited thermoregulation, making them sensitive to temperature changes (Heinsohn et al., 1977) and capable of large-scale movements when sea surface temperatures decline (Deutsch et al., 2022). In northwest Qatar, aggregations coincided with the coldest months (January–February), when coastal waters reached 18–20 °C. Dugongs generally prefer warmer waters, often avoiding temperatures below 18 °C (Marsh et al., 1994; Preen, 2004). For example, dugongs in New South Wales and Queensland were active at 18.9–24.6 °C and moved away from colder waters (~17–18 °C) (Allen et al., 2004; Sheppard et al., 2006). Similar patterns were reported in Moreton Bay (Lanyon, 2003), Japan (Nishiwaki et al., 1979), and Shark Bay (Anderson, 1986; Marsh et al., 1994). In Qatar, winter sea temperatures rarely fall below 18–21 °C due to the Gulf’s shallow morphology. This region exhibits a strong correlation between air and water temperatures, likely due to limited thermal inertia (Vasou et al., 2024). Comparable results from Egypt also show dugongs inhabiting waters from 22 °C in winter to 29 °C in summer (Hanafy et al., 2006). Seasonal shifts in distribution have been noted at the species’ northern range limits (Zeh et al., 2018). Preen (2004) observed summer groups near the UAE and a winter herd between Qatar and Bahrain, previously attributed to freshwater springs. However, this explanation is unlikely, as there have been no in-situ measurements to test this theory. Recent hydrogeological studies indicated that all land-based springs in the area have dried up and only a limited number of submarine springs remain; they are insufficient to support dugong aggregations (Rausch and Dirks, 2024).

The largest single aggregation of dugongs, globally, was observed in the northwest of Qatar after strong wind gusts in the open Gulf as shown in Figure 4B, during which wind speed exceeded 35 km/h. Nevertheless, no surveys were conducted following the peak wind event of February of 2019, during which wind speed surpassed 40 km/h between 17^th and 19^th of the same month. In the winter months especially from December to February, the Gulf experiences intense shamal winds driven by cold fronts, with speeds over 50 km/h and gusts exceeding 100 km/h (Thoppil and Hogan, 2010), peaking in strength during February (Owlad et al., 2022). These winds not only transport cold air leading to a decrease in surface temperature that can attain 15 °C (Burt and Paparella, 2023), but they also initiate downwelling or subsidence of surface water across the shallow southern banks (Al-Thani et al., 2023). This can generate relatively strong waves reaching up to 2.7 m in the southern Gulf (Kamranzad et al., 2013). Under these intense winter winds, the low temperatures of the open Gulf waters render some areas excessively cold for dugongs to access seagrasses, like in Ras Tanura near Saudi Arabia, prompting dugongs to relocate to warmer regions (Hodgson, 2009). Earlier studies indicated that wind speed and direction and the associated effects on the water dynamics might affect the distribution and the foraging behavior of the dugongs rendering seagrass unreachable and promoting movement to sheltered habitats during rough conditions (Budiarsa et al., 2021; Ng et al., 2022). Dugongs are generally sluggish and are unable to sustain activity for longer than a few minutes (Anderson, 1981). They maintain their optimal energy balance between the energy spent during dives and the energy gained while foraging by residing in shallow and sheltered waters (Lefebvre, 2023). Their dives turn energetically less efficient in deeper and rougher waters. Waves and currents caused by wind interfere with the feeding habits of dugongs, increasing the energy expenditure required and complicating the process of diving and reaching seagrass meadows. Dugongs are observed to reduce their travel distances in the winter months as a strategy for adjusting moving costs and displaying thermoregulatory behavior (Zeh et al., 2018). The resulting interference with their ability to feed effectively, likely compels them to relocate to calmer, more stable waters. The gathering site on the western shores of Qatar encounters significantly reduced wave heights, due to the shallow depths and the safeguarding topography of the Qatar and Bahrain peninsulas in the central Gulf (Aboobacker et al., 2021). Moreover, the current study noted that this area tends to have slightly cooler temperatures in summer and slightly warmer during January to March (Figure 4A). Besides, dugongs tend to exhibit less predictable migratory patterns, which are frequently influenced by periodic seagrass loss caused by extreme temperature events, leading them to undertake occasional long-range foraging excursions (Deutsch et al., 2022). We suggest that the plentiful seagrass along Qatar’s western coastline during winter, coupled with the protective impact of the land during winter shamal winds and the comparatively mild and warmer environment, creates ideal circumstances for the substantial aggregation of dugongs noted in February. The west of Qatar also has minimal human interference, creating an ideal habitat for these marine mammals. Therefore, it is likely that female dugongs not only seek this area because it supports open-sea grazing, but because it also serves as a crucial nursery ground, where they can give birth and nurture their calves without significant threats. It has been evidenced in literature that dugongs seek the seclusion of sheltered backwaters for calving, which further emphasizes the importance of the northwest of Qatar as a nursery area (Marsh et al., 1984).

The average salinity in the study area was 48.1 ± 1.1 ‰, reflecting a hypersaline condition. Preen (2004) indicated that dugong presence in the region does not appear to be constrained by high salinity. For instance, the salinity was determined to be 70‰ when dugongs were discovered in the southern Gulf of Salwa, Qatar, whereas it ranged from 50‰ to 60‰ where they were found in the northern Gulf of Salwa. Additionally, dugongs were discovered in water with salinities between 40‰ and 50‰ along the northern Saudi Arabian coast (Preen, 2004).