Bottlenose dolphin reliance on trawlers in prey-depleted Adriatic Sea regions

The Adriatic Sea is one of the areas most exposed to trawling worldwide. The common bottlenose dolphin Tursiops truncatus —the only cetacean species regularly present in the northern and central portions of the basin—is known to forage and scavenge behind trawlers. We investigated dolphin occurrence behind three types of trawlers off the Italian regions of Veneto (3000 km²) and Marche (2000 km²), using 8 years (2018–2025) of survey data and 17,755 km of navigation effort from independent research boats. A total of 859 trawlers were checked for dolphin presence/absence across 148 days of trawling. Overall, dolphins were observed in the wake of 24.7% of the trawlers: 41.3% of the bottom otter trawlers (n = 375), 35.1% of the midwater pair trawlers (n = 148), and 1.5% of the beam trawlers (n = 336). Dolphin occurrence behind otter trawlers varied significantly between regions: dolphins followed one quarter of the vessels off Veneto (25.9%, n = 259), and three quarters of those off Marche (75.9%, n = 116). In the sampled regions, the large majority of bottlenose dolphins appear to be strikingly reliant on trawl gear, not only specialising in foraging and scavenging behind operating trawlers, but also targeting considerable proportions of the trawl fleets, possibly in response to severe prey depletion and habitat disruption by fisheries. The net effects on dolphin fitness and demography remain unknown. As trawlers have been causing profound damage to Adriatic ecosystems, the promotion of their continued use is inadvisable, not only to protect dolphins but also to conserve marine biodiversity writ large.

The Mediterranean Sea (FAO area 37, including the Black Sea) has one of the highest rates of fishery exploitation in the world: 63% of the fish populations there are fished ‘unsustainably’ (FAO, 2024) and fishing pressure is twice the level considered ‘sustainable’ (FAO, 2022). Within the Mediterranean, the Adriatic Sea is the most heavily overfished area (Colloca et al., 2017; Eigaard et al., 2017; Gissi et al., 2017; Ferrà et al., 2018; Russo et al., 2019), with one of the world’s highest regional intensities of bottom trawling and one of the poorest seabed conditions (Amoroso et al., 2018; Pitcher et al., 2022; Hiddink et al., 2026).

Bottom trawling is a destructive practice that damages marine habitat, depletes marine life, and releases organic carbon from disturbed seabed sediments (Puig et al., 2012; Bearzi et al., 2024b; Zhang et al., 2024), with high costs to society (Millage et al., 2026). In the Adriatic Sea, trawling has helped drive a major regime shift (Fortibuoni et al., 2017a; Sguotti et al., 2022), impairing the functioning and resilience of local food webs (Coll et al., 2007, 2009). Rich three-dimensional habitats have been turned into flattened plains that trawlers continue to exploit, with the support of government subsidies (Skerritt et al., 2020; Gambino et al., 2022; Bearzi et al., 2024b; Millage et al., 2026). In this region, marine biotic communities have experienced alterations and sharp declines (Fortibuoni et al., 2010; Barausse et al., 2011; Lotze et al., 2011). Specifically, Fortibuoni et al. (2017a) documented a decline in biodiversity between 1945 and 2014 that was marked by a steep drop in the mid-1980s, and by a long-running process of ‘fishing-down’ the food web (sensu Pauly et al., 1998). High-order predators were rapidly depleted. For instance, elasmobranchs declined by >94% between 1948 and 2005 (Ferretti et al., 2013). Common dolphins Delphinus delphis, once abundant throughout the Adriatic Sea, virtually disappeared (Bearzi et al., 2004, 2024a; Genov et al., 2020).

Common bottlenose dolphins Tursiops truncatus (hereafter ‘bottlenose dolphins’) around the world have adapted to foraging behind various types of trawlers, with effects that include changes in their demography, distribution, movement patterns, social structure, behaviour, and diet (Bonizzoni et al., 2022). In the Adriatic Sea, bottlenose dolphins have long been known to forage behind trawlers (Bearzi et al., 1999; Genov et al., 2008, 2019; Rako-Gospić et al., 2017; Bonizzoni et al., 2021, 2023; Rudd et al., 2022). Such specialised behaviour includes venturing into a moving trawl net to feed on live fish and other organisms trapped in the net, extracting organisms from the outer mesh, and scavenging on dead, harmed, stunned, or weakened organisms stirred up by the net, as well as catch lost during hauling and discarded catch thrown offboard (Bonizzoni et al., 2022)—similar to instances of scavenging by pinnipeds (Tixier et al., 2021), seabirds (Oro and Ruiz, 1997; Karris et al., 2018), and elasmobranchs (Ferretti et al., 2013; Mitchell et al., 2018). Understanding the extent of bottlenose dolphin occurrence behind trawlers within an area trawled as intensively as the Adriatic Sea can help identify conservation actions meant to protect this high-order marine predator, consistent with the mandate of the European Commission’s Habitats Directive and Marine Strategy Framework Directive (EC, 1992, 2008), and the European Union’s Maritime Spatial Planning (EC, 2014).

Our primary aim here was to investigate the proportion of trawlers followed by dolphins. Building on the initial work of Bonizzoni et al. (2021, 2023) off the region of Veneto, Italy, we used a much larger dataset from this area (2018–2024) and combined it with information from an additional study conducted in 2025 off the region of Marche, Italy. We collected information on the three types of trawlers operating in the Adriatic Sea: bottom otter trawlers (hereafter ‘otter trawlers’), midwater pair trawlers (hereafter ‘midwater trawlers’), and beam trawlers (He et al., 2021).

Study area

Boat surveys to investigate dolphin occurrence in the proximity of trawlers extended up to approximately 22 km (12 nautical miles) offshore and covered a 3000 km 2 area off the region of Veneto (hereafter ‘Veneto’) and a 2000 km 2 area off the region of Marche (hereafter ‘Marche’; Figure 1). Both areas are relatively shallow (maximum depth 33 m in Veneto and 75 m in Marche). Veneto is exposed to high river runoff (Russo and Artegiani, 1996; McKinney, 2007). Italy’s two longest rivers—Po and Adige—flow into this part of the Adriatic and discharge nutrient-rich waters that carry pollutants from hydrographic basins encompassing most of northern Italy. Other relevant rivers in Veneto include Tagliamento, Livenza, Piave and Brenta (Figure 2). Conversely, only a few minor rivers are present in Marche; these rivers have negligible runoff and have little influence on sea salinity and other properties (Appiotti et al., 2014).

Both regions host mariculture facilities for Mediterranean mussel Mytilus galloprovincialis (Pellizzato and Penzo, 2002) located between 2 and 9 km offshore (Figure 2). Fishing and navigation within these areas are banned. In Veneto, fishing and navigation are also prohibited within a radius of 2 km from a large gas terminal situated 13 km offshore (‘Adriatic LNG’; adriaticlng.it) resulting in a no-take area of 12.6 km 2. In Marche, there are some relatively small offshore platforms where navigation and fishing are prohibited within 500-m buffer zones. All along the Italian coast of the Adriatic, trawling is prohibited within 5.5 km (3 nm) of the coast, or inside the 50-m isobath when this depth is reached at a shorter distance from the coast (EC, 2006; Figure 2).

Boat-based surveys

Boat-based visual surveys were conducted from 6–7 m craft with rigid hulls powered by four-stroke outboard engines between April and November, 2018–2025. No surveys were conducted in April–May 2020 (due to a COVID-19 lockdown), and April 2025 (due to logistical constraints). Navigation aimed to cover the study areas as evenly as possible (Figure 1; and see the Supplementary materials), though routes depended on sea and weather conditions, as well as encounters with trawlers and dolphins. Survey effort was discontinued in fog, rain, insufficient visibility or unfavourable sea state conditions (>3 Douglas). Navigation ‘on effort’ was suspended as soon as trawlers and/or dolphins were encountered, and resumed after they were left.

Dolphin occurrence near trawlers was assessed during 148 days of trawling; survey days when trawling was not allowed were excluded from the dataset presented here. Trawling bans varied among years and areas, but normally included Fridays, weekends and national holidays, as well as summer bans of 30–40 days. Survey effort involved 17,755 km of navigation, with a total of 859 fishing trawlers being checked for dolphin presence/absence (hereafter ‘inspected’; Table 1). Some trawlers were subject to multiple inspections on different days across the study period. Most effort occurred in the morning, with early departures to account for the greater trawling activity in those hours.

While observing and tracking dolphins, we attempted to photograph all individuals in the group. Photo-identification in days of trawling was performed as described by Bearzi et al. (2021). A total of 47,261 high-quality, high-resolution photographs of single individuals were included in a catalogue, and matched based on conspicuous dorsal fin markings (Bearzi et al., 2021).

Observation of trawlers and data collection

When a fishing trawler was spotted during navigation, the research boat routinely approached it until it could be determined whether dolphins were present in its wake. A trawler was considered to be fishing when active trawl gear was deployed, either towing, setting or hauling the net. Trawlers that were navigating (travelling) and those dragging their net at the surface for cleaning purposes were not considered. If dolphins were encountered away from trawlers, and during the tracking of those dolphins they happened to approach a fishing trawler, that was not counted as an inspection. During an inspection, two observers carefully observed the sea surface behind and around the fishing trawler. In the case of midwater pair trawlers, the two vessels towing a single net accounted for a single observation. Inspections took place in all months of the study, totalling 28 in April, 126 in May, 193 in June, 157 in July, 3 in August, 164 in September, 177 in October and 11 in November. The low number of inspections in August is due to the summer bans on trawling; in April and November, to unfavourable weather conditions resulting in lower survey effort.

If dolphins were present, their movements behind the fishing trawler were tracked, and additional information on group size, group composition (following Bearzi et al., 1997) and behaviour relative to fishing vessels was collected, with the research boat matching the speed of the trawler while staying within 50 m of the animals, parallel to their route. The position of the research boat was recorded via GPS at 1‐min intervals throughout navigation and tracking, and was used as a proxy for dolphin position. Dolphin and trawl positions were visualized and analysed using ArcMap 10.4 (ESRI, Redlands, CA, USA).

To determine if the proportion of trawlers observed with dolphins varied across gear types, a Chi-square test was used. If the proportions significantly differed, pairwise proportions tests (using the ‘pairwise.prop.test’ function in R) with a Bonferroni p -value correction were used to determine which specific gears dolphins associated with most often. These comparisons were made first across both sites (Veneto and Marche, combined). Then, we repeated comparisons in Veneto, where all three types of trawlers were present (in Marche, nearly all of the observations were of otter trawlers). Further, we used Chi-square tests to determine if the proportion of a given type of trawler followed by dolphins varied between Veneto and Marche, to assess whether dolphin association with trawlers was site-specific. All analyses were conducted in RStudio 467 (R Core Team, 2024). Because of the numerous tests, we used a conservative α = 0.01 to determine statistical significance, in addition to using Bonferroni-corrected p -values for pairwise tests.

Information on trawlers

The registered fleet of trawlers operating in Veneto and Marche in 2023 was reported by the National Fishery Observatory (Anonymous, 2025) to include 156 vessels in Veneto and 157 in Marche (Veneto: 34 beam trawlers, 74 otter trawlers and 24 pairs of midwater trawlers; Marche: 12 beam trawlers, 113 otter trawlers and 16 pairs of midwater trawlers). Detailed information on the spatial and temporal patterns of trawl fishing in the northern and central Adriatic Sea based on Automatic Identification System data (Russo et al., 2020; Coro et al., 2022) indicates that both study areas are trawling hotspots. Analyses of long-term changes in fishery effort and landings in these areas, with emphasis on trawling, were presented in several studies (e.g. Bastardie et al., 2017; Bastari et al., 2017; Fortibuoni et al., 2017a, 2017b; Russo et al., 2019; and see the Introduction), with evidence of a collapse of landings in the mid-1980s (Fortibuoni et al., 2017a) followed by a regime shift caused by overexploitation (Sguotti et al., 2022). Sguotti et al. (2022) pointed out that recovery towards previous baselines may be impossible.

Beam trawlers in the study area normally tow four nets simultaneously, at speeds of up to 13–15 km/h. Gear typically consists of 120-kg box dredges (each about 4-m wide and 30–40-cm high, with 5–7-cm long teeth along the lower leading edge and a net bag to collect the catch); an inclined wooden board acts as a depressor, keeping the gear in contact with the seabed and facilitating penetration of the teeth into the soft sediment (Lucchetti and Sala, 2012). Otter trawlers, operating at speeds of 5.6–7.4 km/h (Sala, 2013), tow a single net that drags the seabed and is kept open by two large boards; the net’s mouth can be 12–22 m wide and 1–4 m high (Lucchetti and Sala, 2012). Midwater trawlers operate in pairs at speeds of 7.4–9.3 km/h; the net’s mouth can be 30–45 m wide and 7–12 m high (Sala, 2013). While midwater trawlers normally target pelagic and epipelagic species, in the shallow waters of the northwestern Adriatic they can fish close to the seabed (Fortuna et al., 2010; Sala et al., 2018). For instance, Casale et al. (2004) reported that benthic taxa were caught in 12% of 496 midwater trawl hauls, indicating net proximity to, or contact with, the seabed. Bonanomi et al. (2018) reported consistently high catches of immature demersal elasmobranchs in midwater trawl gear.

In our study, the type of trawl gear (He et al., 2021) was assessed visually during each inspection. Trawlers were photographed with 20-megapixel digital cameras equipped with 70–200 mm f2.8 AF zoom lenses, and high-resolution photographs were used to complement visual records. Information was included in a dedicated catalogue together with vessel name, plate, Maritime Mobile Service Identity (MMSI), length overall (LOA), and type of gear. The MMSI and LOA of vessels were retrieved from the European Fleet Register (webgate.ec.europa.eu/fleet-europa/index_en); when information was not available there, other online databases were used (either marinetraffic.com or vesselfinder.com). Trawlers were then categorised into small (LOA < 15 m) and large (LOA ≥ 15 m), following regulations that mandate vessels ≥ 15 m to be equipped with VMS and/or AIS (Russo et al., 2019).

Other fishing boats

Fishing boats other than trawlers that were actively engaged in fishing during days of trawling were routinely observed. In Veneto, these included 28 encounters with hydraulic dredges as well as small boats operating traps (82), set-nets (16), and longlines (4); in Marche, they included 12 hydraulic dredges. Unidentified recreational fishing boats 5–8-m long engaged in sport angling were encountered a total of 646 times in Veneto and 12 times in Marche. Sport angling tour boats were encountered 20 times in Veneto and never in Marche.

Information on dolphins

The size of dolphin groups was estimated in the field by means of independent counts performed by two experienced researchers, with groups being defined as dolphins observed in apparent association, moving in the same direction and often, but not always, engaged in the same activity, with members of the focal group usually remaining within approximately 100 m of one another (Bearzi et al., 2021). Dolphin age classes – newborn, calf, juvenile, adult – were assigned following Bearzi et al. (1997). Only the initial size and composition of encountered groups were used in the analyses.

Observation of trawlers

A total of 859 fishing trawlers were checked for dolphin presence: 741 in Veneto and 118 in Marche (Table 1), at distances between 2 and 30 km from the coast (Figure 2). Trawling activity during these inspections included towing (86.2%, n = 740), hauling (12.1%, n = 104), and setting (1.7%, n = 15). Based on vessel identity and field identification of the gear being deployed, we catalogued a total of 223 unique trawlers: 52 beam trawlers (15–27 m long, mean = 21.8 m), 123 otter trawlers (9–29 m long, mean = 17.4 m), 35 midwater trawlers (14–27 m long, mean = 19.2 m), and 13 vessels that switched gear according to fishing day or season. Of the thirteen switchers, 6 vessels of 12–27 m operated either otter or beam trawl gear, and 7 vessels of 15–24 m operated either otter or midwater trawl gear; in the analyses, these vessels were assigned to a trawl category based on the gear they were using at the time of observation. In Veneto, all three types of trawlers were observed; in Marche, only otter and beam trawlers (Table 1). Each individual trawler was observed between 1 and 29 times (mean = 3.8 times, median = 2 times). A total of 150 inspections were of vessels <15 m, 709 of vessels ≥15 metre (Table 2).

Dolphin occurrence near trawlers

Bottlenose dolphins were the only marine mammals encountered across seven years of surveys in Veneto. Similarly, all encounters in Marche were of bottlenose dolphins, except for a few observations of a single well-marked and photo-identified common dolphin that associated with bottlenose dolphins but was not observed following trawlers. Bottlenose dolphin occurrence near trawlers was recorded in both morning and afternoon hours, between 04:41 and 19:49.

Across all gear types, dolphins followed 212 (24.7%) of the 859 trawlers inspected. Of these, 178 (84.0%) were towing, 32 (15.1%) were hauling, and 2 (0.9%) were setting their nets. The proportion of trawlers followed (Figure 3) varied markedly depending on gear (Table 1). Of 336 beam trawlers inspected, only 5 (1.5%) had dolphins in their wake. Of 375 otter trawlers, 155 (41.3%) had dolphins. Of 148 midwater trawlers, 52 (35.1%) had dolphins. Overall, dolphin occurrence varied significantly among the three types of trawlers (n = 859, df = 2, χ 2 = 161.87, p < 0.001). Beam trawlers had significantly fewer dolphin occurrences than both otter and midwater trawlers (pairwise proportions test; both p < 0.001), while the proportions of otter trawlers and midwater trawlers followed by dolphins were not significantly different (pairwise proportions test; p = 0.68). These same trends were observed when focusing exclusively on the Veneto data (n = 741, df = 2, χ 2 = 107.13, p < 0.001), with otter and midwater trawlers each followed more often than beam trawlers (pairwise proportions tests; both p < 0.001), but not different from each other (pairwise proportions test; p = 0.19).

Dolphin occurrence behind otter trawlers varied markedly between regions. In Veneto, 67 of 259 otter trawlers inspected were followed by dolphins (25.9%); in Marche, 88 of 116 (75.9%). This difference was significant (n = 375, df = 1, χ 2 = 80.53, p < 0.001). Since 98.3% of the vessels encountered in Marche were otter trawlers (with the remaining 1.7% being beam trawlers, and no midwater trawlers), we did not compare the frequency of dolphin occurrence behind beam and midwater trawlers between Veneto and Marche.

When considering the two study areas together, the occurrence of dolphins behind otter trawlers differed significantly between large (≥15 m) and small (<15 m) vessels (n = 375, df = 1, χ 2 = 38.66, p < 0.001), with large vessels having significantly more occurrences than small vessels. The same pattern was found in Veneto (n = 259, df = 1, χ 2 = 12.50, p < 0.001) and in Marche (n = 116, df = 1, χ 2 = 7.53, p < 0.001). Vessel size differences in dolphin occurrence for beam and midwater trawlers were not investigated due to small sample size (only 26 and 11 vessels <15 m, respectively; Table 2).

In 212 inspections of trawlers followed by dolphins, vessels were identified based on their name and plate. These trawlers included 125 different vessels: 5 beam trawlers, 83 otter trawlers and 19 pairs of midwater trawlers (i.e. 37 twin vessels towing midwater trawl gear, of which one operated in two different pairs). Of these, 31 otter trawlers and 13 pairs of midwater trawlers were followed by dolphins on more than one occasion. The largest pair of midwater trawlers (both LOA 27.2 m) had the highest number of dolphin occurrences (n = 9), followed by an otter trawler of 16.9 m that had dolphins 7 times.

Photo-identification data were obtained in Veneto during 98 days of trawling (37,581 dorsal fin photographs), and in Marche during 19 days of trawling (9680 dorsal fin photographs). Of 490 marked adult individuals in Veneto, 421 (85.9%) were photographed one or more times behind trawlers; of 308 in Marche, 278 (90.3%).

Dolphin behaviour near trawlers

Dolphin movements in the wake of trawlers were tracked for a total of 86 h 01 min (mean = 25 min, range 1–196 min, SD = 27.305, SE = 1.884), totalling 525 km of tracking (Figure 4). In Veneto, dolphins behind trawlers were tracked between 4.9 and 30.4 km from the coast, in waters 17–33 m deep; in Marche, between 5.6 and 22.5 km from the coast, in waters 14–75 m deep (Figure 4).

The mean estimated group size of dolphins following trawlers was 9.0 (SD = 8.9, SE = 0.6, n = 212, range 1–70). A Kruskal-Wallis test indicated that the group size did not differ significantly among the three types of trawlers (H = 0.144, df = 2, p = 0.93). Table 3 shows the size of dolphin groups according to trawl gear and study area. Adult individuals were present in all 212 groups, while juveniles were present in 109 groups (51.4%), calves in 62 (29.2%) and newborns in 15 (7.1%), indicating that trawlers could be followed by individuals of all age classes. Juveniles, calves and newborns were all routinely observed surfacing and diving in synchrony with the adults above towed nets, performing dives that were as long as those of the adults. The youngest animals, however, did not appear to approach vessels as closely as the adults.

Dolphins foraging behind towing trawlers normally surfaced about 20–200 m behind the vessel’s stern (Figure 5). Precise distances were hard to determine because the animals were often scattered into subgroups, surfacing at variable distances from the vessel. For example, there could be a first subgroup in the immediate vicinity of the vessel (possibly feeding on organisms discarded offboard from the deck), a second subgroup surfacing above the towed net, and a third subgroup behind the codend. For many otter trawlers in Veneto, the codend location was indicated by a codend buoy towed at the surface. In the case of midwater trawlers, dolphins tended to surface above the towed net, halfway between the wake of the two vessels. Dolphin dives in the wake of fishing trawlers normally lasted about 2–3 min, and were followed by ventilation sequences. The longest recorded dives were about 5 min. Dolphin speed sampled opportunistically while following towing vessels averaged 7.7 km h −1 for beam trawlers (n = 103), 6.3 for otter trawlers (n = 2,568), and 7.5 for midwater trawlers (n = 629).

During hauling, when trawlers were motionless or moving slowly backward, dolphins could approach the stern as the net was being retrieved, and surface as close as 1–2 m from the vessel. After the net had been hauled out completely, the dolphins could either abandon the trawler or wait until the net was set again, normally performing non-directional surfacing. In a few cases, midwater trawl fishers discarded entire boxes of epipelagic fish offboard—possibly unwanted catch—after the net had been hauled out. In these cases, dolphins would dive in the area and remain there long after the vessel had left, apparently scavenging on the discards.

Dolphins were never observed foraging near hydraulic dredges, netters or recreational fishing boats engaged in sport angling (see Section 2.5).

What triggers opportunistic foraging and scavenging behind trawlers?

Opportunistic foraging and scavenging in or near fishing gear is a well-known behavioural adaptation of bottlenose dolphins worldwide (Bearzi et al., 2019; Hersh et al., 2025). Specifically, foraging and scavenging behind various types of trawlers has been reported from many areas (Bonizzoni et al., 2022). Understanding the degree of bottlenose dolphin ‘reliance’ on different types of trawlers within a region can help assess the ecological and behavioural change driven by trawl fisheries, and frame it in an appropriate management context.

The Adriatic Sea has long been one of the most intensively trawled areas in Europe (Hiddink et al., 2026), and one with the poorest seabed status, by far, compared to 23 other trawled regions worldwide (Pitcher et al., 2022). Adriatic ecosystems and habitats have been altered dramatically (Fortibuoni et al., 2010, 2017a; Barausse et al., 2011; Lotze et al., 2011; Sguotti et al., 2022), and the natural prey of dolphins and other apex marine predators has been severely depleted—particularly in the heavily-trawled waters in Italy, including those off Veneto and Marche (Russo et al., 2020; Cacciamani et al., 2025). In an area so heavily impacted by trawling and other human activities (Ramírez et al., 2018; and see the Introduction), the adaptation by bottlenose dolphins to foraging and scavenging behind trawlers may simply be a way to take advantage of the resources that remain available to them (Bearzi et al., 2019).

Prey depletion caused by human activities is faced by many large terrestrial carnivores worldwide (Wolf and Ripple, 2016; Parsons et al., 2022). When natural prey is scarce, animals are more likely to exploit ‘human food’. For example, when polar bears’ access to prey is reduced, they become reliant on food in and around human settlements, and often scavenge in garbage dumps, i.e. predictable sources of food (Atwood and Wilder, 2021). Prey depletion and habitat degradation can also induce, or force, marine predators like dolphins to scavenge and forage opportunistically around fishing gear, often exacerbating conflict with humans (Read, 2008; Powell and Wells, 2011; Tixier et al., 2021) and having negative impacts on the animals (Plaza and Lambertucci, 2017; Smith et al., 2023) which can include dietary shifts and population decline (Bearzi et al., 2006, 2008a; Roussouw et al., 2025; Bernard et al., 2026; Pasino et al., 2026).

In the Adriatic Sea and other Mediterranean areas, dolphins were regarded as vermin for centuries and killed by the thousands because of acute conflict with fisheries—particularly between the 18th and mid-20th centuries (Bearzi et al., 2004, 2024a; Meliadò et al., 2020; Petitguyot et al., 2024). Still, interaction with fishing gear may have increased over time, concurrent with prey depletion and habitat degradation, possibly even to a point of being a primary foraging strategy. While such a hypothesis deserves investigation, understanding is bound to be limited by the scarcity of baseline information. In a study conducted between 1990 and 1994 in the northeastern Adriatic Sea, only 10.4% of 202 otter trawlers inspected were followed by dolphins (Bearzi et al., 1999)—apparently indicating a lower rate of interaction with trawlers than that observed in Veneto and Marche in 2018–2025 (Figure 3).

Degree of reliance on trawlers

Long-term, consistent, and deliberate association with a given food source might be interpreted to mean there is a high degree of ‘reliance’ on that source. In the case of bottlenose dolphins, trawlers appear to be dependable, undemanding foraging targets. In Veneto, bottlenose dolphins were able to detect faraway trawlers at distances of at least 2.6 km, and deliberately approached those vessels to forage in their wake (Bearzi and Bonizzoni, 2026). Trawlers (particularly otter and midwater trawlers) were important drivers of dolphin distribution, with dolphins foraging and scavenging behind trawlers during 39.3% of total observation time in trawling days (Bonizzoni et al., 2023). Dolphin adaptation to intensive trawling was illustrated by major differences in spatial distribution between days with and without trawling (Bonizzoni et al., 2023). Preliminary analyses of social structure as related to trawling (Bonizzoni and Bearzi, 2024), together with the results of photo-identification presented above (Section 3.2), suggest that a large majority of bottlenose dolphins in our study areas forage regularly behind trawlers. The size of the population in Veneto was estimated at approximately 600 individuals (Bearzi et al., 2021); in Marche, 436 marked adult individuals were photo-identified in 2025. Preliminary analyses indicate that the size of the two populations combined likely exceeds 1,000 individuals.

Evidence reported above, together with the other results of the present study, suggests that bottlenose dolphin populations in these two Adriatic Sea areas interact with operating trawlers to a remarkable degree. The interaction rate was particularly high in Marche, where more than three quarters of the inspected otter trawlers had dolphins foraging or scavenging in their wake. The threefold difference from Veneto was unexpected, and is difficult to explain, given that the two study areas are only ~150-km apart.

Beam trawlers were followed by dolphins much less (1.5% of vessels) than otter (41.3%) and midwater trawlers (35.1%). In Veneto, beam trawlers were the most numerous vessels, and also the most inspected, but they were the least followed by dolphins. Such a finding likely relates to the fact that those vessels tow narrow metal dredges and flat nets (see Section 2.4) that prevent efficient foraging or scavenging. Also, beam trawlers often target primarily pectinids and flatfish (Pranovi et al., 2001) which are less-favoured prey for dolphins. Conversely, otter and midwater trawls tow nets with large surface areas and wide-open mouths (He et al., 2021; Bearzi et al., 2024b), targeting a variety of species known to be regular prey of bottlenose dolphins in the Mediterranean Sea (Bearzi et al., 2008b).

At least theoretically, foraging and scavenging behind trawlers augments the dolphins’ food intake. In some areas, and for species other than bottlenose dolphins, foraging in association with other types of fishing gear reportedly resulted in increased fitness. For instance, killer whales Orcinus orca interacting with longliners appeared to have higher rates of adult survival (Esteban et al., 2016) and higher calving rates (Tixier et al., 2014; Esteban et al., 2016). In the case of trawling, however, any advantages to individual dolphins that might be gained through opportunistic foraging and scavenging may be offset by ecosystem deterioration and prey depletion caused by towed gear (Puig et al., 2012; Bearzi et al., 2024b), and by damage to dolphin hearing and health that results from chronic exposure to the noise of trawlers and trawl-related pollutants (Bonizzoni et al., 2022; Bearzi et al., 2024b).

Foraging and scavenging behind trawlers is known to affect bottlenose dolphin distribution (Bonizzoni et al., 2023; Labriola et al., 2026), diet (Corkeron et al., 1990; Svane, 2005; Bonizzoni et al., 2022), social organization (Chilvers and Corkeron, 2001; Chilvers et al., 2003; Ansmann et al., 2012; Kovacs et al., 2017; Genov et al., 2019), and communication patterns (Rako-Gospić et al., 2021; McGhee et al., 2026). Additionally, bottlenose dolphins are occasionally injured or killed by trawl gear (Santana-Garcon et al., 2018; Bonizzoni et al., 2022), including in the Adriatic Sea where, however, dolphin mortality in midwater pair trawls appears relatively low (Bonanomi et al., 2022); bycatch in beam and otter trawls in the Adriatic has never been quantified.

While dolphins would still need to forage independently when trawling does not occur, on days of trawling they appeared to forage predominantly on trawl nets, spending a large proportion of their time following a considerable proportion of the operating otter and midwater trawlers (Figure 3; and see Bonizzoni et al., 2023). Consistently, in the Italian waters of the Tyrrhenian Sea, bottlenose dolphins associated with otter trawlers in days of trawling, whereas in days without trawling they occurred in more coastal and shallower waters close to the Tiber River estuary (Labriola et al., 2026), where they appeared to feed on mullet (Mugilidae; Pace and Pedrazzi, 2025).

Cultural transmission

Specialised foraging behaviour is often transmitted culturally (Donaldson et al., 2012; Mann et al., 2012; Allen et al., 2013; Wild and Rendell, 2026). In Veneto and Marche, young animals, including dependent newborns and calves, were often observed diving above towed nets together with the adults, typically in relatively small groups composed of fewer than ten individuals. Genov et al. (2019) also reported groups that included adults, juveniles, calves and newborns following otter trawlers in the north-eastern Adriatic. Dependent newborns and small calves associated with adults that forage and scavenge behind trawlers (also see Delgado-Estrella, 1997; Jefferson, 2000; Siegel et al., 2015; Mevorach et al., 2025; McGhee et al., 2026) presumably would not themselves be feeding in the proximity of nets. However, by observing the behaviour of adults, immature animals are likely to learn this behaviour and become specialised foragers and scavengers themselves as they mature.

Cultural transmission of foraging and scavenging behind trawlers can reinforce the segregation of dolphins that do from those that do not associate with trawlers (Kovacs et al., 2017; Chilvers and Corkeron, 2001; Chilvers et al., 2003; McGhee et al., 2026). In northern Adriatic waters off Slovenia and Friuli Venezia Giulia (Italy), adjacent to Veneto, the dolphins that regularly interacted with trawlers belonged to a separate social cluster (Genov et al., 2019). The long-term demographic consequences of such social segregation are unknown. However, when dietary specialisation is culture-driven, odontocete communities can become fragmented and more vulnerable (Whitehead and Ford, 2018; Hersh et al., 2025). In the case of bottlenose dolphins, their fluid social system may allow a fragmented population to recover its integrity following a near-complete trawling ban, as found in eastern Australia (Ansmann et al., 2012). Still, culture-driven foraging specialisation can lead to ‘evolutionary traps’ (Schlaepfer et al., 2002) when, for example, a foraging niche experiences reductions or depletion, and the behavioural trait targeting the niche becomes maladaptive (Wild and Rendell, 2026). Therefore, dolphin adaptation to human encroachment is often a double-edged sword (Bearzi et al., 2019), with any short-term foraging benefits masking long-term detrimental effects (Whitehead and Ford, 2018).

Management and conservation implications

Measures to deter dolphins or exclude them from trawl nets, either to reduce perceived depredation or to mitigate bycatch in trawl gear, have included the use of acoustic devices, physical barriers and exclusion devices. However, attempts to determine whether acoustic devices deter bottlenose dolphins from approaching trawl gear have been inconclusive (Santana-Garcon et al., 2018; Bonizzoni et al., 2022; Dolman et al., 2022). Similarly, physical barriers and exclusion devices mounted on trawl nets have not been shown to reduce the incidental mortality of dolphins consistently (Hamilton and Baker, 2019).

As mentioned earlier, the Adriatic Sea has been exposed to one of the highest intensities of trawling worldwide (Pitcher et al., 2022; and see Hiddink et al., 2026 for EU waters), but there is virtually no published information on the economic and other impacts of bottlenose dolphins on Adriatic trawl fisheries, in terms of e.g. damage to gear, loss of work time, damage to, or removal of, organisms caught in trawl gear, or measurable changes in catch quantity and composition. As noted in Section 4.3, evidence of bycatch of bottlenose dolphins in trawl gear is also scant. In this ecologically degraded Mediterranean region, sensible approaches to protect dolphins while maintaining ecosystem function and allowing sustainable fishing, might include spatial restrictions on trawl fisheries (Bastari et al., 2016; Bastardie et al., 2017; Armelloni et al., 2021) and the adoption of less destructive fishing methods (Bearzi et al., 2024b). Such approaches would be more consistent and compliant with the commitments made under international agreements such as the Convention on Biological Diversity (CBD), the Convention on Migratory Species (CMS), and the Barcelona Convention for the Protection of the Marine Environment and the Coastal Region of the Mediterranean.

Marine conservation action within the European Union (EU) is required by, inter alia, Council Directives 92/43/EEC (the ‘Habitats Directive’; EC, 1992), 2008/56/EC (the ‘Marine Strategy Framework Directive’; EC, 2008) and 2014/89/EU (the ‘Maritime Spatial Planning Directive’; EC, 2014). Until now, however, implementation of Council Directives has been, at best, an uphill process (e.g. Mackelworth et al., 2011), with regulatory frameworks operating at different scales with uncoordinated objectives (Gissi et al., 2018). Similarly, international agreements such as the CBD, the CMS, the Barcelona Convention, and the Agreement on the Conservation of Cetaceans of the Black Sea, Mediterranean Sea and Contiguous Atlantic (ACCOBAMS; accobams.org) prescribe measures that tend to be ignored by the very nations that agreed to them. As noted by Bearzi et al. (2024a), “adopted resolutions often clash with the broad economic interests involved in the human uses of the marine environment, and these interests have invariably prevailed”.

Counter-intuitively, trawling continues to be allowed even within many marine protected areas (Dureuil et al., 2018; Claudet et al., 2020; Millage et al., 2026). In 2023, the European Commission proposed an Action Plan to phase out all bottom trawling within MPAs (mapamed.org) and Natura 2000 sites (natura2000.eea.europa.eu), consistent with existing obligations and requirements, and called on EU Member States to implement national measures accordingly (EC, 2023). Member States with strong fishery lobbies expressed opposition, and weakened any undertaking regarded as detrimental to bottom trawl fisheries. Such opposition to even the most basic measures of protection is in sharp contrast to the declared global urgency of increasing and expanding MPAs and other area-based conservation measures—consistent with the United Nations commitment to ‘protect’ 30% of global oceans by 2030 (UN, 2023). Within the EU, opposition to the phasing out of bottom trawling within MPAs also defies the EU Biodiversity Strategy’s goal to protect 30% of the EU’s seas, as well as EU directives requiring Member States to take measures to protect the seabed to achieve ‘good environmental status’ of EU waters (Directive 2008/56/EC) and to achieve or maintain the ‘favourable conservation status’ of certain seabed habitats in marine Natura 2000 sites.

Notably, bottom trawling in the Mediterranean is either restricted or banned within Fisheries Restricted Areas (FRAs) created under GFCM (fao.org/gfcm/data/maps/fras/en/). In 2005, the GFCM established a large FRA (1,730,000 km 2) encompassing all Mediterranean and Black Sea seabed deeper than 1,000 m (thus including a portion of the southern Adriatic Sea), where the use of trawl nets and towed dredges is prohibited to protect deep-sea benthic habitats. Of ten other FRAs established since 2006, the Jabuka/Pomo Pit FRA (established in 2017 in the Adriatic Sea) is a compelling example of how spatial conservation management can benefit both the marine environment and fisheries. However, while compliance was high within the no-trawl area, fishers maintained similar overall effort by trawling elsewhere—likely to mitigate the economic consequences of the closure to them (Elahi et al., 2018).

The northern Adriatic Sea has been classified as an Important Marine Mammal Area for bottlenose dolphins (Notarbartolo di Sciara et al., 2016; IMMA, 2017; IUCN, 2018). From a scientific perspective, the creation of appropriately large no-take areas in the northern Adriatic Sea would open up interesting research avenues. For example, it would enable researchers to investigate whether, over time, bottlenose dolphins reduce their time spent foraging and scavenging in the wake of trawlers, and redirect their effort to catching and consuming prey ‘the natural way’. Ecosystem recovery might also allow other cetaceans to return—notably common dolphins, that were historically abundant in this area but have virtually vanished (Bearzi et al., 2024a).

Limitations of this study

One of the main limitations of this study is the uneven allocation of effort between Veneto and Marche—where surveys spanned 7 years and 1 year, respectively. Ideally, comparisons should be drawn between areas surveyed more evenly, and simultaneously rather than sequentially, as was done in this study. Despite only one year of sampling, however, the comparatively high number of otter trawler inspections in Marche (n = 116, compared to 259 in Veneto) allowed us to conduct robust analyses. The difference in the proportion of otter trawlers followed by dolphins was so striking (roughly one quarter in Veneto versus three quarters in Marche) that it is unlikely to have been a result of sampling bias. Still, this result would require validation.

Bottlenose dolphins in the Adriatic Sea regions of Veneto and Marche appear to have adapted to foraging and scavenging behind operating trawlers, following considerable proportions of the fishing fleets during days of trawling. In the Adriatic, such behaviour has been reported since the early days of trawling (Bonizzoni et al., 2022), and it is presently unclear whether the intensity of opportunistic foraging and scavenging increased as natural prey became depleted. The long-term demographic consequences for the dolphin population(s) remain largely unknown. Conversely, the negative effects of trawling on Adriatic marine habitats and biodiversity have been documented extensively.