MORBIDITY AND MORTALITY PATTERNS OF INDIAN RIVER LAGOON COMMON BOTTLENOSE DOLPHINS (TURSIOPS TRUNCATUS TRUNCATUS) 2002–2020 Open Access

The examination of health trends and causes of mortality are critical to environmental monitoring and population management (Sleeman et al. 2012; IJsseldijk et al. 2020). Although case studies of individual pathological findings provide valuable information, few studies have systematically examined the cause of morbidity and mortality of marine mammal populations over an extended time period (Cowan et al. 1986; Gerber et al. 1993; Greig et al. 2005; Zagzebski et al. 2006; McFee and Lipscomb 2009; Bogomolni et al. 2010; Ashley et al. 2020). Along the east coast of central Florida, US, a population of common bottlenose dolphins (Tursiops truncatus truncatus) inhabiting the Indian River Lagoon (IRL) estuary system (NOAA 2009) has been the focus of several ecological and health-related studies (Odell and Asper 1990; Bossart et al. 2003; Goldstein et al. 2006; Reif et al. 2006; Durden et al. 2007, 2009, 2019, 2021; Stolen et al. 2007, 2013a; Fire et al. 2020). Dolphins exhibit year-round IRL residency and strong site fidelity (Odell and Asper 1990; Mazzoil et al. 2005) with relatively small linear ranging patterns (mean±standard deviation [SD] 28.1±9.49 km; Durden et al. 2019). Six distinct communities of dolphins inhabit the IRL (Titcomb et al. 2015) which has a mean population size of approximately 1,032 dolphins (Durden et al. 2021). As a long-lived top-level predator, IRL dolphins are exposed to and accumulate persistent pollutants (Stavros et al. 2011) that may increase susceptibility to disease (Fair and Becker 2000). Population studies have documented indicators of diminished health in IRL dolphins including lingual and genital papillomas (Bossart 2007), and skin disease (Paracoccidioidomycosis ceti; Bossart et al. 2017). The IRL dolphin stock has been described as an immunocompromised population (Bossart et al. 2003) and has been subjected to four unusual mortality events (UMEs). These events occurred in 2001 and 2008 (unknown causes), in 2013 (ecological factors), and 2013–2015 (dolphin morbillivirus epidemic; Morris et al. 2015; NOAA 2015). Indian River Lagoon common bottlenose dolphins are listed as a strategic stock because anthropogenic mortality probably exceeds potential biological removal (PBR), which is the maximum number of annual mortalities (excluding natural mortality) that can occur while allowing the stock to reach or maintain an optimal sustainable population level (NOAA 2015). Both direct and indirect anthropogenic threats pose a risk, as the lagoon is extensively used for recreational activities that negatively impact IRL dolphins (Noke and Odell 2002; Durden 2005; Bechdel et al. 2009; Stolen et al. 2013a). The IRL is a microtidal, largely enclosed system that is susceptible to an influx of terrestrial pollutants (Smith 1993, 2001; Woodward-Clyde Consultants 1994). In recent years, the IRL has undergone several large-scale ecosystem changes, including diminishing seagrass coverage associated with nutrient accumulation and declining water quality (Sigua et al. 2000; Morris et al. 2022). As seagrass meadows provide critical habitat to prey consumed by estuarine dolphins (Barros and Wells 1998), these significant changes could further jeopardize the health of the vulnerable IRL dolphin stock. Examining mortality trends is critical to understanding population mortality patterns and to the management of this strategic stock. The objective of this study was to examine morbidity patterns from stranded IRL common bottlenose dolphins from 2002 to 2020.

The IRL is a shallow (majority <1 m; Gilmore 1977) estuarine system, 902 km², along the east coast of central Florida consisting of three interconnected bodies of water, the Indian River, Banana River, and Mosquito Lagoon, which span a linear distance of approximately 250 km, from Ponce de Leon Inlet to Jupiter Inlet (Environmental Protection Agency [EPA] 1996; Fig. 1). The study area extended from Ponce Inlet to Sebastian Inlet (approximately 144 km), including the following subbasins: Mosquito Lagoon, Banana River, northern Indian River, and North-Central Indian River (Woodward-Clyde Consultants 1994; Fig. 1), encompassing 79% of the IRL (731 km²) and 5/6 resident dolphin communities (Titcomb et al. 2015). To evaluate potential geographical differences, probable causes of mortality were evaluated by subbasin, as these present different abiotic and biotic characteristics (Smith 1993; Sigua and Tweedale 2003).

Common bottlenose dolphins that stranded within the study area were examined using established protocols (Geraci and Lounsbury 2005) under a Stranding Agreement with NOAA Fisheries and through section 109h of the US Marine Mammal Protection Act of 1972 (MMPA 1972). Total length (TL) was measured as straight-line length from the tip of the rostrum to the fluke notch (Norris 1961). Sex was determined by external examination and internal gonad examination. Age class was estimated based on total length: adult male (≥246 cm), adult female (≥231 cm), juvenile male (161–245 cm), juvenile female (161–230 cm), calves (≤160 cm; Wells et al. 1987). Age classes were further grouped by maturity (calves and juveniles=immature, adults=mature). If TL was not available, the age class was recorded as unknown. Exceptions were made for carcasses that were witnessed being pushed by a conspecific (epimeletic behavior). In these cases, age class was conservatively estimated as a calf if images indicated the carcass was ≤50% the size of the proximate adult (Mead and Potter 1990) or exhibited definitive young-of-year characteristics (dark color, rostral hairs, presence of fetal lines; Mead and Potter 1990). When available, reproductive maturity assessments were used to refine age class estimates, including ovarian corpora or presence of spermatozoa and associated density of interstitial tissue for adults (Akin et al. 1993). Decomposition state was defined as: code 1 (alive), code 2 (fresh dead), code 3 (moderately decomposed), code 4 (severely decomposed), and code 5 (mummified or skeletal remains; Geraci and Lounsbury 2005). When feasible, fresh dead animals were removed from the stranding location and examined in the laboratory. Carcasses that could not be transported and late code 3, code 4, or code 5 carcasses were typically examined in the field. For live stranded dolphins and those needing intervention, typical data collected included images, total length, sex, human interaction assessment, and blood (for complete blood count and biochemistry analysis). Seasonal patterns for mortality were examined and seasons defined as winter (December–February), spring (March–May), summer (June–August), and fall (September–November; Shane 1990).

Tissues from fresh dead animals, or, as appropriate, moderately decomposed carcasses (early code 3) were examined histologically. Samples from all major organs and suspected lesions were collected (Geraci and Lounsbury 2005), fixed in 10% neutral buffered formalin, embedded in paraffin, sectioned at 3–5 μm, and stained with H&E for examination by light microscopy. Tissue sections typically included skin, skeletal muscle, diaphragm, tongue, larynx, trachea, bronchi, lung, pericardium, heart (all chambers and major vessels), spleen, prescapular lymph node, lung-associated lymph node, pancreatic lymph node, mesenteric lymph nodes, thymus, adrenal, thyroid, pituitary, kidney, bladder, esophagus, all stomach chambers mucosal and submucosal lining, intestine, colon, liver, pancreas, gonads, mammary, eyes, brain, and any significant lesions.

Carcasses were systematically evaluated for evidence of human interaction, such as vessel strike–propeller wounds, entanglements, or ingestion of marine debris (Read and Murray 2000; Moore and Barco 2013). Human interaction presence was scored as “yes,” “no,” or “could not be determined” when a thorough human interaction assessment was precluded. For fresh carcasses (code 2–early code 3), nutritional condition was evaluated based on prominent skeletal structure, postnuchal fat, and epaxial musculature mass, and was scored as robust, ideal, thin, or emaciated (Fair et al. 2006). Carcasses that received a full examination (all major organs present and examined) were further evaluated to identify the probable causes of mortality, which were grouped into the following etiologic categories: degenerative, metabolic, nutritional, inflammatory (infectious and noninfectious), and trauma. Categories represent the probable cause of mortality, but not always the etiology. Probable cause of mortality was determined based on gross and histological findings supplemented by radiology, hematology, urinalysis, bacteriology (aerobic and anaerobic), or virology (PCR) results when available. To simplify analyses, confirmed causes of mortality and probable causes of mortality were combined. Categories were further evaluated by subbasin, year, age class, and sex. Mortality attributed to trauma was further subdivided into anthropogenic, natural, and unknown sources. We calculated summary statistics including mean and SD. Chi-square analysis was conducted in R to evaluate data trends by year, season, age class, sex, and subbasin (R Core Team 2022). Significance was defined as a P<0.05.

From 2002 to 2020, 790 occurrences of common bottlenose dolphins that were stranded (live or dead) or free swimming and warranted intervention (sick or injured; entangled; or out-of-habitat) were reported within the IRL (41.58 ±13.78 SD per year; range: 28–82). Most (89%; 706/790) occurred in the study area, and an additional 84 animals stranded south of the study area (11%). Annual dolphin strandings varied significantly (X²=94.85, df=18, P<0.0001), ranging from 21 to 78 (37.16±13.99 SD) with declared UMEs occurring in 2008 and 2013 (Supplementary Material Fig. S1). Of the known-sex dolphins, 281/490 (57.3%) were male, and 209/490 (42.6%) were female; in 216/706 cases (30.6%) sex could not be determined. Stranding events were significantly more frequent during summer (282/706, 40%; X²=88.62, df=3, P<0.0001), compared to spring (161/706, 23%), fall (142/706, 20%), and winter (121/706, 17%). Of the cases of known age class, adults (225/662, 36%) and calves (215/662, 34%) stranded in nearly equal numbers; with slightly fewer juveniles (192/632, 30%). In 74 cases, age class could not be determined. Age class varied significantly by season (X²=29.80, df=6, P<0.0001). Calves stranded more frequently during fall (Table 1) (X²=6.86, df=2, P=0.03), whereas juvenile strandings were more frequent in winter (X²=9.63, df=2, P=0.008) and less frequent in summer (X²=10.93, df=2, P=0.00424) compared to other age classes. The majority of dolphins were found stranded in the northern Indian River (258/706, 37%) and Banana River (226/706, 32%), followed by Mosquito Lagoon (134/706, 19%) and the North-Central Indian River (88/706, 12%). Evidence of human interaction was present in 19.3% (136/706) of all live and dead stranding events; the majority of these involved fishery interactions (106/136, 77.4%).

We examined 504 dead stranded animals. Observed carcass conditions included moderately decomposed (n=298), advanced decomposition (n=116), fresh dead (n=75), and skeletal or mummified (n=15). A total of 202 additional strandings were excluded, including unrecovered carcasses (n=46), adults exhibiting epimeletic behavior towards a deceased conspecific (n=97), carcasses examined by another network member (n=19), or live animals (n=40) that were either returned to the lagoon (n=29), left onsite (n=5), or taken into rehabilitation (n=6). Probable causes of mortality were evaluated for 392 stranded dolphins that received complete gross examinations (74 fresh dead, 263 moderately decomposed, 55 advanced decomposition). Histological reviews were conducted on 178 dolphins (84 males, 91 females, and 3 unknown sex). The remaining 112 stranded animals received partial examinations or were not intact and were excluded from further analyses.

A probable cause of mortality was evident in 57% (223/392) of cases (Supplementary Material Table S1). In 43% of cases (169/392) the probable cause of mortality was not apparent, as most (83%, 141 cases) lacked histology. For cases in which a probable cause of death was found, inflammatory disease and trauma were identified most commonly (Fig. 2, Supplementary Material Table S2, Fig. S2).

Inflammatory disease accounted for 91/223 (41%) of cases with an evident probable cause of mortality, within which respiratory disease was most common (63/91, 69%) (Supplementary Material Table S1, Fig. S1). Pneumonia accounted for the majority of respiratory disease (43/63, 68%); causative agents included parasitic (n=17), multiagent (including viral, bacterial, fungal, parasitic, and protozoal ciliates; n=9), bacterial (n=8), undetermined etiology (n=7), and fungal (n=2). Mortality in calves was most commonly attributed to inflammatory (67.74%) and metabolic (6.45%) diseases and less frequently to nutritional causes (Supplementary Material Table S3). In juveniles also, the most common suspected cause of mortality was inflammatory disease (Supplementary Material Table S3). Overall, mortality was more frequently attributed to inflammatory disease in immature animals than in mature animals, although the difference was not significant (X²=3.56, df=1, P=0.0592). Inflammatory disease did not vary significantly by season (X²=5.48, df=3, P=0.1398).

Trauma accounted for 36% of findings (80/223), and was primarily related to anthropogenic activities (58%, 46/80), with 26/80 from natural causes and 8/80 being trauma of unknown origin (Supplementary Material Table S2; Fig. 2). Trauma was the leading probable cause of mortality for mature animals. Separating out anthropogenic trauma, this alone was the second most frequent cause of mortality, accounting for 20.63% of cases. Anthropogenic trauma specifically occurred more frequently in immature than in adult animals, although the association was not significant (X²=1.06, df=1, P=0.3032; Table S3). Anthropogenic trauma occurred more frequently in summer, although the difference was not significant (X²=5.3, df=3, P=0.1511). Natural trauma accounted for 12% (n=26) of cases (Fig. 2) and occurred more often in mature than immature animals, although not significantly (X²=2.56, df=1, P=0.1096). Natural trauma included stingray spines (penetrating the abdominal cavity, thoracic cavity, liver, or heart; n=4), prey-associated esophageal obstruction or asphyxiation (n=19), and shark bite lacerations (n=3; Supplementary Material Table S2). The cause of trauma could not be determined in 3% of cases (Fig. 2).

Nutritional issues (starvation or inanition) were the probable cause of mortality in 38 cases (17%) and did not differ significantly between immature (n=20) and mature (n=8) animals (X²=0.02, df=1, P=0.8875). Metabolic disease (hepatic lipidosis) was uncommon (1% of findings) and may reflect a nutritional condition resulting in metabolic derangements. Degenerative disease accounted for 5% of findings (n=12).

No significant associations were found between sex and suspected mortality due to trauma (X²⁼0.12, df=1, P=0.7290), inflammatory disease (X²⁼0.7, df=1, P=0.4028), or nutritional issues (X²⁼0.44, df=1, P=0.5071; Supplementary Material Table S3).

The probable causes of mortality were further evaluated by year and subbasin (Table 2, Supplementary Material Table S4). The percentage of cases in which a probable cause of mortality could not be determined ranged from 18% to 58% annually (42.5±11.0%). Inflammatory disease varied annually (17–80%, 42.96±17.98%), occurring most frequently in 2014 (Table 2). Inflammatory disease occurred significantly more than expected in the northern Indian River compared to other subbasins (X²=13.22, df=3, P=0.0042; Supplementary Material Table S4). Nutritional cases (starvation or inanition) varied annually from 0% to 61% (13.24±17.13%), with more cases in 2013 than in other years (Table 2). Trauma cases ranged from 20% to 67% (37.36±14.11%) annually and were highest in 2002 (Table 2). Compared to other subbasins, trauma occurred more than expected in the Banana River (X²=11.7, df=3, P=0.0085; Supplementary Material Table S4) where anthropogenic trauma was the leading probable cause of mortality (27.42% of cases; Supplementary Material Table S4).

Although mortality causes did not vary seasonally, significantly more events occurred during the summer months, as observed in historical IRL dolphin stranding data (1977–2005; Stolen et al. 2007). Because dolphin calving tends to occur between late spring and early fall (Wells et al. 1987), neonatal mortality may contribute to increased summer stranding events. Furthermore, elevated summer IRL water temperatures (up to 41 C; Robbins and Lisle 2018) may induce thermoregulatory stress and compromise health, as witnessed in dolphins inhabiting Sarasota Bay (Wells et al. 2004). The UMEs in 2001 and 2008 occurred in summer months (NOAA 2015), suggesting that season-specific stressors may influence dolphin mortality. Ecological pressures increase in summer months, as algal blooms are significantly more prevalent (Lopez et al. 2021) and may lead to hypoxic conditions (Hallegraeff 2010) and associated fish kills (Lewis et al. 2021), which may impact prey availability. Further efforts are needed to evaluate how these chronic ecological stressors impact dolphin health.

It was notable that inflammatory disease was considered to be the cause of mortality most frequently, with respiratory disease being the most common ailment. Bottlenose dolphin respiratory anatomy enables a rapid exchange of large volumes of air; however, it may also increase the risk for respiratory infection from exposure to airborne threats at the water’s surface (Simpson and Gardner 1972). Therefore, it is not surprising that pneumonia is one of the most common pathological findings, not only in this study, but in many dolphin health studies (e.g., Sweeney and Ridgway 1975; Baker 1992; Bonar et al. 2007; Gonzalez-Viera et al. 2011; Venn-Watson et al. 2012; Rodrigues et al. 2018). Pneumonia commonly contributed to the mortality of calves and juveniles. Transplacental lungworm infection has been documented in common bottlenose dolphins (Dailey et al. 1991) and lungworms were most common causative agent, in this study, that contributed to pneumonia. The significance of pneumonia and other inflammatory diseases may be underrepresented, because this study evaluated the probable cause of mortality rather than the simple presence of disease.

Anthropogenic trauma played a significant role in IRL dolphin mortality. Findings of anthropogenic mortality were often evident from gross findings, highlighting the importance of a thorough necropsy regardless of decomposition. Interactions with vessels (propeller strikes) accounted for 2.2% of cases. A prior study found that 6% of marked IRL dolphins (n=43) had evidence of vessel strikes (Bechdel et al. 2009); suggesting that, similar to accounts on the west coast of Florida (Wells et al. 2008), nonfatal interactions may occur more frequently than fatal. Fishing gear interactions (entanglement and ingestion) were the most common anthropogenic trauma. Along the west coast of Florida, large-scale ecosystem changes and associated prey depletion have been found to influence dolphin behavior, yielding increased associations with fishing gear (Powell and Wells 2011). Dolphins may come in contact with fisheries while in pursuit of similar target species (Barros 1993), therefore as lagoon fish populations are depleted and ecosystem health declines (Adams et al. 2019; Lewis et al. 2021; Morris et al. 2022), high-risk fishery interaction behaviors (Noke and Odell 2002; Durden 2005) and associated injury or mortality may increase. Trauma was the leading probable cause of mortality for dolphins recovered from the Banana River. Low water exchange rates and long residence times (Smith 1993) result in an accumulation of nutrients (Lapointe et al. 2020), yielding phytoplankton blooms (Phlips et al. 2021), seagrass meadow depletion (Morris et al. 2022), and fish kills in this subbasin (Lewis et al. 2021). Furthermore, all three prior IRL dolphin UMEs have encompassed this subbasin (National Oceanic and Atmospheric Administration [NOAA] 2015), illustrating the chronic stress. As a result, compromised dolphins with comorbidities may engage in risky behavior as they forage in a depleted ecosystem with persistent pressures, potentially increasing trauma from foraging in tandem with fisheries as well as the exploitation of atypical prey. As such, trauma events may mask underlying comorbidities, yielding underrepresentation of coinciding disease.

Natural trauma resulting from interactions with bony fish and elasmobranchs accounted for 12% of events. Penetrating wounds from stingray spines have been widely recognized in cetaceans (Walsh et al. 1988; McLellan et al. 1996; McFee et al. 1997, Duignan et al. 2000; Spanier et al. 2000; McFee and Lipscomb 2009; Silva et al. 2010). We documented four cases of such interactions. The shallow nature of the lagoon, individual foraging preference in shallow waters (≤1 m; Durden et al. 2019), and observations of IRL dolphins playing with stingrays (W. Durden, personal observation), could increase the likelihood of spine penetration. Occurrences were, however, less frequent than documented in other regions where such interactions account for 5% of bottlenose dolphin stranding events (McFee and Lipscomb 2009). Prey-induced laryngeal displacement contributed to 8% of mortalities. Bottlenose dolphins have a unique upper respiratory morphology that makes them more vulnerable to esophageal obstruction and subsequent asphyxiation (Macleod et al. 2007). Previous studies have documented large prey items and those with dorsal spines yielding esophageal obstruction in bottlenose dolphins (Watson and Gee 2005; Mignucci-Giannoni et al. 2009; Byard et al. 2010; Stolen et al. 2013b). Changing environmental conditions may contribute to shifts in prey availability and abundance, potentially influencing dolphin prey selection (Powell and Wells 2011) and increasing risk of prey-related trauma.

Since 2001, the IRL common bottlenose dolphin population has experienced four UMEs, three of which are included in this study (2008, 2013, 2013–2015 morbillivirus event). The etiology of the 2008 UME remaining undetermined. Comparisons of annual mortality revealed higher than average mortality from trauma and metabolic and degenerative disease, but did not identify a cause for this large-scale event. The 2013 UME was the largest event, with 77 dolphin mortalities. The event differed from prior UMEs in that it was preceded by significant seagrass loss (Morris et al. 2022) and involved increased mortality of other species (manatees: Trichechus manatus latirostris and pelicans: Pelecanus occidentalis) (NOAA 2015; Florida Department of Environmental Protection [FDEP] 2016; Landsburg et al. 2022). Annual mortality comparison indicated an increase in nutritional cases (starvation or inanition) in 2013 (more than four times the annual mean). This corresponded with observations of emaciation in free-swimming and deceased dolphins during the event (W. Noke Durden pers. obs.). While a definitive cause was not determined, this UME was suspected to be associated with ecological factors preceding the event including phytoplankton blooms that yielded catastrophic seagrass loss (Morris et al. 2022) and may have impacted prey availability (Lewis et al. 2021). The end of this unprecedented event coincided with the Mid-Atlantic UME (2013–2015) which was caused by a morbillivirus epidemic. During this event more than 1,600 dolphins stranded along the Atlantic coast from New York to Florida (Brevard County) (NOAA 2015). In 2014, morbillivirus was the probable cause of mortality for several dolphins in Mosquito Lagoon based on a positive PCR test (often coinciding with syncytial cells, lymphoid depletion, and verminous, fungal, and/or bacterial pneumonia), yielding the greatest prevalence of inflammatory disease (nearly twice the annual mean). Histopathological presentation was comparable to dolphin morbillivirus events that have occurred in other regions (Kemper et al. 2016; Groch et al. 2020). Morbillivirus mortality in the IRL was relatively low when compared to coastal events (NOAA 2015). This was likely because of decreased exposure in estuarine dolphins (Balmer et al. 2018), preexisting morbillivirus titers in IRL dolphins (9.8%; Bossart et al. 2010), and the low exchange rate between Mosquito Lagoon and the remainder of the IRL (Durden et al. 2021). These retrospective comparisons of mortality, during UME and non-UME years, illustrate the importance of long-term data sets to compare regional mortality trends.

This study, encompassing 19 yr of necropsy findings, provides baseline data that are critical to the conservation and management of IRL dolphins. Nevertheless, our findings represent the minimum occurrence of identified causes of mortality, because decomposition and scavenging often preclude thorough evaluation. Furthermore, dolphin carcass recovery rates in enclosed estuaries may be influenced by winds, carcass buoyancy, distance to shore, as well as human population density and recreational usage (Wells et al. 2015). Prevailing winds vary seasonally in the IRL (Smith 2001), but typically span the lagoon’s longitudinal axis. In Sarasota Bay, only one-third of estuarine common bottlenose dolphin carcasses were estimated to be recovered (Wells et al. 2015), therefore it is imperative to acknowledge that stranding events are negatively biased as recovery rates may be influenced by several factors. Anthropogenic mortality threatens the sustainability of this stock (i.e., exceeds PBR). Likewise, reoccurring mortality events may be an indicator of serious ecological pressures that could lead to the decline of this strategic stock. Therefore, continued outreach efforts are vital to conservation, particularly in areas where anthropogenic trauma is suspected as the leading cause of mortality.

Supplementary material for this article is online at http://dx.doi.org/10.7589/JWD-D-22-00156.

We sincerely thank Dr. Stedman, Dr. Dold, Dr. DiRocco, Dr. Erlacher-Reid, Dr. Croft, Dr. Lindemann, Dr. Staggs, Dr. Gearhart, Dr. Pelton, and Dr. Walsh, SeaWorld Orlando rescue team and vet services staff, and HSWRI volunteers for their support of this study. This work was funded by the John H. Prescott Marine Mammal Rescue Assistance Grant, SeaWorld Busch Gardens Conservation Fund, Discover Florida’s Oceans License Plate, Brevard County Tourism and Development Council, and through the Indian River Lagoon National Estuary Program.