Bottlenose Dolphin (Tursiops truncatus) Survival in Captivity

Bottlenose Dolphin (Tursiops truncatus) Survival in Captivity Abstract Bottlenose Dolphins (Tursiops truncatus) were first placed in captivity in the U.S. in 1953, and have become familiar to the public due to their wide exposure in marine parks, movies and television shows. Dolphins have shown to have successful breeding in captivity, but despite this breeding success there is growing concern for cetacean welfare in captivity, with claims of poor survival fueling the concern. This claim has brought to question whether whales and dolphins should continue to be kept in captivity A series of tests including Kaplan-Meier and Cox proportional hazards analyses were employed on 1,568 captive bottlenose dolphins based on data retrieved from the Marine Mammal Inventory Report (MMIR) to determine how sex, facility location (U.S. vs. foreign), captive-born vs. wild-capture and pre-and post-1985 affects survival. Bottlenose dolphins housed in foreign facilities (29.6 years) demonstrated a significantly higher median survival than those in U.S. facilities (18.9 years), as did dolphins that entered captivity post 1/1/1985 (23.4 years) vs. those entering captivity prior to 1/1/1985 (9.6 years). Facility location and pre- vs. post-1985 were good predictors of captive bottlenose dolphin hazard rate. Introduction Many different species of cetaceans and other marine mammals are maintained in captivity for many different purposes. Among the marine mammals used for entertainment, killer whales (Orcinus orca) were first placed into captivity in the U.S. in 1961, although serious health problems related to captivity have been reported (e.g., St. Leger et al 2011, Jett and Ventre 2013). Bottlenose dolphins (Tursiops truncatus) were first placed into captivity in the U.S in 1953 (NOAA). Since 1953 problems with the health and others regarding captivity has been reported (e.g., St. Leger et al 2011, Jett and Ventre 2013). Though candida yeast is not thought to be an important source of morbidity or mortality among wild cetaceans (John Jett pers. Comm.) candida infection has led to the deaths of several various species of cetaceans in captivity including the long finned pilot whale (Globicephala melas), harbor porpoise (Phocoena phocoena), Atlantic bottlenose dolphin (Tursiops truncatus) and killer whale (Orcinus orca) (Dunn et al. 1982). Among bottlenose dolphins, chronic and severe stressors may result in degenerative health, both mental and physical, and may lead to prolonged illness or death (Carter 1982). Clinical symptoms of stress among mammals include weight loss, prone to infection, decrease of circulating white blood cells (lymphocytes and eosinophils), increase in neutrophils, and antisocial behavior (Fowler, 1978; Thomson and Geraci, 1986; Dierauf, 1990; Sapolsky, 1994). The deaths of killer whale trainers and a park guest in 1991, 2009 and 2010 by captive killer whales have led to a heightened criticism regarding the maintenance and welfare of cetaceans in captivity (Parsons 2012). These criticisms often focus on health impacts, psychological stressors, other negative consequences of captivity and poor survival. Although Bottlenose dolphins face many of the same stressors in captivity that have been described in killer whales such as changes in group dynamics, resource competition, and unstable dominance hierarchies, (Waples K.A, Gales N.J. 2002) the theme park industry has not endured the same level of criticism regarding captivity-induced problems among dolphins (Carter 1982). With growing concern regarding captive animals in general, the welfare of captive animals has gained greater focus, with supporting literature for many different species of animals. As seen in Mason (2010), many different animals including, giraffes (Giraffa cameleopardalis), African and Asian Elephants (Loxodonta africana and Elephas maximus), cheetahs (Acinonyx jubatus), Humbolt penguins (Spheniscus humboldti) and many others are highly susceptible to stress in captivity and are difficult to maintain and breed. Both acute and chronic stress is reflected in their infant mortality rates, shorter lifespans, and health vulnerability (Mason 2010). For the zoo species discussed above, survival rates are worse than those animals from protected populations in wild habitat (Mason 2010). In one specific example, African elephants had a mortality rate that is 2.8 times higher (P<0.01) in captivity than the population of the Amboseli females that died from natural causes (i.e. by a cause other than poaching) (Clubb et al. 2008). In another example, Weiss et al. (2011) analyzed the survival rates of orang-utans in captivity within the context of subjective evaluations of well-being by their keepers. Researchers noted that there was a higher rate in animals that keepers had judged as “happier” (p=0.012). This study suggests that the wellbeing of the animal may be correlated with its survival, and that some individual animals may be more suited or adaptable to a captive setting than others. Frank Robson, dolphin trainer and research scientist at Napier Dolphinarium expressed concern that almost every disease that captive dolphins contract, has a strong causal link to psycho-physiological factors (Carter 1982). In order to test this, Robson analyzed data from dolphins that were accidently captured in trawl nets and seemed to be physically healthy. He separated dolphins into three different classifications which included sudden death, survival followed by death after a month and dolphins that died from respiratory problems after being held for varying periods of time. He determined that a dolphin’s inability to deal with emotional and physical stress disorders, attributable to captivity, was responsible for pneumonia and other respiratory problems that ultimately led to their deaths (Carter 1982). Captive cetaceans in the United States are tracked by the National Marine Fisheries Service, and overseen by National Oceanic and Atmospheric Administration (NOAA) as mandated by the Marine Mammal Protection Act. The Marine Mammal Protection Act of 1972 (MMPA) and other similar legislation in foreign countries, has determined that marine mammals provide aesthetic and recreational resources, and therefore, the legislation is meant to protect numerous species of marine mammals from harm from human activity (Alker 1996). NOAA fisheries maintain documentation of whales, dolphins, porpoises and other marine mammals that are held in captivity for public display and other purposes (NOAA 2013). Within the requirements of the MMPA, any U.S. facility that displays captive cetaceans, including facilities in foreign countries that send or receive animals from the U.S. are required to submit details regarding the individual animal to the NOAA. These data includes, holding institution, sex, age, date entering captivity (wild-capture or birth), animal imports, disposition of the animal (death, transfer or release) and date upon death. These data are maintained within the Marine Mammal Inventory Report (MMIR) by NOAA Fisheries. There have been several studies that have used the data within the MMIR to evaluate the annual survival rates (ASR) among captive marine mammals through time, between sexes, between captive-born and wild-capture and among different facilities (DeMaster and Drevenak 1988, and Small and DeMaster 1995, Jett and Ventre 2015). Using the same ASR approach as DeMaster and Drevenak (1988) Small and DeMaster (1995) analyzed life data for 864 captive bottlenose dolphins. The authors found that there was a statistically significant difference (P<0.025) in the survival rates between individual facilities holding captive bottlenose dolphins. Secondly, they determined that the survival rates of the captive bottlenose dolphins were significantly less for the calves born in captivity than the survival of non-calves in captivity. Small and DeMaster (1995) used the ASR analysis based on data retrieved from the MMIR to determine that among known age captive bottlenose dolphins, the difference between calf and non-calf survival rates were statistically significant. They also found that there were no statistically significant differences in survival between distinct year classes of captive-born non-calves compared to wild-born non-calf bottlenose dolphins. Lastly they determined that ASR’s were also significantly different among various institutions holding non-calf bottlenose dolphins. Jett and Ventre (2015) evaluated captive killer whale survival using ASR, Kaplan-Meier Regression and Cox Proportional Hazards. The authors analyzed the difference in survival rates based on several covariates including, sex, facility location (U.S vs. foreign) and captive-born vs. wild-capture. In addition, as 1985 marked a milestone in the ability to keep a captive-born killer whale alive for greater than one year (Asper et al. 1988) they evaluated survival between pre-1985 and post-1985 whale cohorts. They found that whales held in U.S. facilities had higher survival rates than those held in foreign institutions (P=0.01), they found that whales captured from the wild had higher survival than captive-born whales (P=0.019) and they found that whales that entered captivity prior to 1/1/1985 had lower survival rates than those that entered post 1/1/1985 (P=0.004) although survival failed to differ once age upon arrival estimates were added (P=0.361), They found that there were not statistically significant differences between males and females (P=0.280). They also determined that survival of captive killer whales, in recent years, has become asymptotic, meaning that captive killer whale care has generally improved the survival of the whales in captivity overtime, but that survival is now stabilizing and is no longer improving. Lastly they determined that captive animals were not living to reach the age milestones that are typically found in those animals found in the wild. In comparing survival in captivity to that of the wild, survival to age milestones can be beneficial in determining whether the captive species reach the essential milestones such as sexual maturity and menopause, which can be found in their wild counterparts. Therefore, the purpose of this study is, to follow up on Jett and Ventre’s (2015) analyses. However, rather than analyzing the survival rate of captive killer whales, I will be analyzing the survival rate of captive bottlenose dolphins using Kaplan-Meier (KM) and Cox Proportional hazards models. The Cox proportional hazards model includes evaluations among sex, U.S. vs. foreign facility, captive-born vs. wild capture and pre- vs. post- 1985. The Kaplan-Meier model will allow me to estimate the survival of these animals and will generate survival curves to determine at what age in captivity survival either deteriorates or improves over time. This information could better inform facility management and animal care practices as well as to determine if standards of care for these animals should to be better regulated for foreign institutions. Materials and Methods Data of 1,568 Bottlenose dolphins received from the MMIR was analyzed with SPSS (v. 19; IBM 2009). We employed the KM survival estimate to evaluate the following questions regarding bottlenose dolphins in captivity: 1. What is the overall survival estimate of bottlenose dolphins in captivity? 2. Are there differences in survival estimates based on sex? 3. Are there differences in survival estimates between facilities? 4. Are there differences in survival estimates between captive-born and wild-capture dolphins? 5. Are there differences in survival estimates between pre- and post- 1985? 6. How do the following covariates affect captive bottlenose dolphin hazard rate? a. Sex b. Facility location (US vs. foreign) c. Captive-born vs. Wild-capture d. Pre- and post- 1985 Upon receiving data from the MMIR, various data entries were removed from the data set due to lack of necessary information for analysis. Records were deleted if they failed to include captivity date, custody date, facility location, or disposition. Additional rows of information were then added to the data set, including, year (based off custody date), total age (disposition date-captivity date/365) and total duration in captivity (DOC) (disposition date-custody date/365). For animals that were still alive at the time research was conducted, a date of 1/1/2015 was used to calculate DOC. Coding was also done for sex, location of facility (U.S. or foreign), captive-born or wild captured, and pre- and post- 1985. We calculated KM survival estimates for all animals in the MMIR (2/27/1953-1/1/2015). This allowed for the computing of the probabilities of experiencing death given those animals that had died, as well as those that were still alive at the time of the study’s conclusion (Nuss and Warneke 2010). The KM analyses generate survival statistics based on the number of animals at risk and the number of animals that had died (i.e., positive for an event) or which were still alive at the end of the study (i.e., negative for an event) (Hosmer and Lemeshow 1999). In addition to generating overall survival estimates and survival curves for each individual cohort, we utilized the Mantel- Cox log-rank analysis at the .05 level of significance to test the null hypothesis of no survival differences between sexes, animals housed in foreign vs. U.S. facilities, captive-born and wild-capture, and pre- and post- 1/1/1985. We cite median survival for the analyses that provided the 95% confidence interval. For analyses in which mortality rate didn’t fall below 50%, we reported mean survival for 95% confidence interval rather than median survival. We tested individual cohorts within the multivariate model at the .05 level of significance, and then tested for interaction effects based on the results. The hazard ratio is reported to better understand how the variables affect the chance of death that each animal faces. The significant hazard ratios were converted to odds of death occurring given each covariate, by: odds= HR/ (1+HR) (Spotswood et al. 2004). Results Data for 1,568 captive bottlenose dolphins were examined. Of the 1,568 captive dolphins, 707 (45%) were males, and 861 (55%) were females (Figure 1a). One-thousand two-hundred and twenty-six (78%) were housed in U.S. facilities, whereas 342 (22%) were housed in foreign facilities (Figure 1b). Captive-born animals consisted of 716 (46%), while wild-capture totaled 857 (54%) (Figure 1c). Of the 1,568 dolphins held in captivity, 969 (62%) had died by 1/1/2015 while only 596 (38%) were still alive (Figure 1d). There were 770 (49%) dolphins that entered captivity prior to 1/1/1985 and 798 (51%) entered post 1/1/1985 (Figure1). 0 100100100 200200200 300300300 400400400 500500500 600600600 700700700 800800800 900900900 1000100010001000 malesmales femalesfemales females Number of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose Dolphins Sex of Bottlenose Dolphins in Captivity Sex of Bottlenose Dolphins in CaptivitySex of Bottlenose Dolphins in CaptivitySex of Bottlenose Dolphins in CaptivitySex of Bottlenose Dolphins in CaptivitySex of Bottlenose Dolphins in CaptivitySex of Bottlenose Dolphins in CaptivitySex of Bottlenose Dolphins in CaptivitySex of Bottlenose Dolphins in CaptivitySex of Bottlenose Dolphins in Captivity Sex of Bottlenose Dolphins in CaptivitySex of Bottlenose Dolphins in Captivity Sex of Bottlenose Dolphins in Captivity Sex of Bottlenose Dolphins in Captivity Sex of Bottlenose Dolphins in Captivity Sex of Bottlenose Dolphins in CaptivitySex of Bottlenose Dolphins in Captivity Sex of Bottlenose Dolphins in Captivity Sex of Bottlenose Dolphins in Captivity Sex of Bottlenose Dolphins in Captivity 55% 45% 0 200200200 400400400 600600600 800800800 1000100010001000 1200120012001200 1400140014001400 US Facilities US FacilitiesUS FacilitiesUS Facilities US Facilities US Facilities Foreign FacilitiesForeign FacilitiesForeign FacilitiesForeign Facilities Foreign FacilitiesForeign Facilities Foreign Facilities Foreign Facilities Foreign Facilities Number of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose Dolphins Location of FacilitiesLocation of FacilitiesLocation of Facilities Location of FacilitiesLocation of Facilities Location of Facilities Location of FacilitiesLocation of FacilitiesLocation of FacilitiesLocation of FacilitiesLocation of Facilities Location of Facilities 78% 22% 600600600 650650650 700700700 750750750 800800800 850850850 900900900 captive captivecaptive captive wild capture wild capturewild capturewild capture wild capturewild capture wild capturewild capture Number of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose Dolphins Origin of Bottlenose Dolphins in CaptivityOrigin of Bottlenose Dolphins in Captivity Origin of Bottlenose Dolphins in Captivity Origin of Bottlenose Dolphins in CaptivityOrigin of Bottlenose Dolphins in CaptivityOrigin of Bottlenose Dolphins in CaptivityOrigin of Bottlenose Dolphins in CaptivityOrigin of Bottlenose Dolphins in CaptivityOrigin of Bottlenose Dolphins in CaptivityOrigin of Bottlenose Dolphins in Captivity Origin of Bottlenose Dolphins in CaptivityOrigin of Bottlenose Dolphins in Captivity Origin of Bottlenose Dolphins in Captivity Origin of Bottlenose Dolphins in Captivity Origin of Bottlenose Dolphins in Captivity Origin of Bottlenose Dolphins in CaptivityOrigin of Bottlenose Dolphins in Captivity Origin of Bottlenose Dolphins in Captivity Origin of Bottlenose Dolphins in Captivity Origin of Bottlenose Dolphins in Captivity 46% 54% 0 200200200 400400400 600600600 800800800 1000100010001000 1200120012001200 Dead Alive AliveAlive Number of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose Dolphins Disposition of Bottlenose Dolphins in Disposition of Bottlenose Dolphins in Disposition of Bottlenose Dolphins in Disposition of Bottlenose Dolphins in Disposition of Bottlenose Dolphins in Disposition of Bottlenose Dolphins in Disposition of Bottlenose Dolphins in Disposition of Bottlenose Dolphins in Disposition of Bottlenose Dolphins in Disposition of Bottlenose Dolphins in Disposition of Bottlenose Dolphins in Disposition of Bottlenose Dolphins in Disposition of Bottlenose Dolphins in Disposition of Bottlenose Dolphins in Disposition of Bottlenose Dolphins in Disposition of Bottlenose Dolphins in Disposition of Bottlenose Dolphins in Disposition of Bottlenose Dolphins in Disposition of Bottlenose Dolphins in Disposition of Bottlenose Dolphins in Disposition of Bottlenose Dolphins in Captivity Captivity Captivity Captivity 62% 38% Figure 1: Descriptive Statistics of the bottlenose dolphins in captivity from 1953 to 1/1/2015. (a) each sex, (b) U.S. and foreign facilities, (c) Wild-Capture and Captive-Born, (d) Disposition of the dolphins in captivity, (e) Pre-1985 and post-1985 d. c. b. a. 750750750 760760760 770770770 780780780 790790790 800800800 810810810 Pre-1985 Pre-1985Pre-1985Pre-1985Pre-1985Pre-1985 Post-1985 Post-1985Post-1985Post-1985Post-1985Post-1985Post-1985Post-1985 Number of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose Dolphins Captivity Date Captivity Date Captivity Date Captivity DateCaptivity DateCaptivity Date Captivity DateCaptivity Date c. e. 49% 51% When separated into ten year increments, it is clear that up until 1985, the main source for bringing dolphins into captivity was through wild-capture. After 1985 the main source for dolphin populations in captivity was through captive breeding programs (Figure 2). Figure 2: Distribution of captive-born vs. wild-captured Bottlenose dolphins from 1953 to 1/1/2015. The values embedded in the graph represent the number of animals who entered captivity either through wild-capture or captive-birth for each time category. Overall, Male and Female, and U.S. and Foreign Survival The Overall KM median survival estimate was 13.0 years (95% CI: 11.4-14.6 yrs.)(Figure 3a) This survival curve indicates a steep decline initially upon entering captivity. 8 256 331 243 18 1 2 7 96 248 188 175 0 50 100100100 150150150 200200200 250250250 300300300 350350350 1953-19631953-19631953-19631953-19631953-19631953-19631953-19631953-19631953-1963 1964-19731964-19731964-19731964-19731964-19731964-19731964-19731964-19731964-1973 1974-19831974-19831974-19831974-19831974-19831974-19831974-19831974-19831974-1983 1984-19931984-19931984-19931984-19931984-19931984-19931984-19931984-19931984-1993 1994-20031994-20031994-20031994-20031994-20031994-20031994-20031994-20031994-2003 2004-20142004-20142004-20142004-20142004-20142004-20142004-20142004-20142004-2014 Number of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose DolphinsNumber of Bottlenose Dolphins Years in 10 year increments Years in 10 year increments Years in 10 year incrementsYears in 10 year incrementsYears in 10 year incrementsYears in 10 year increments Years in 10 year increments Years in 10 year increments Years in 10 year increments Years in 10 year increments wild capture wild capturewild capturewild capture wild capturewild capture wild capturewild capture captive born captive borncaptive born captive born captive borncaptive borncaptive borncaptive bornSurvival deteriorates drastically up until year 8.0 and then begins to display a relatively constant decline throughout time. KM Survival curves for each sex are shown in Figure 3b. Overall survival estimate for females was 12.0 years (95% CI: 9.7-14.2 yrs.) and 14.1 years (95% CI: 11.1-17.1 yrs.) for males. Utilizing the Mantel-Cox log rank test, survival estimates were not significantly different (ᵪ2=2.265; P=.13). Both male and female survival curves show a steep decline in their survival and remain overlapped until approximately 3.0 years at which point male and female survival diverge with males experiencing higher survival than that of the females. Both survival curves remain parallel past the 3.0 year mark. However females experience a decline in survival around 17.0 years, while males experience an increase. This also occurs around the 30.0 year mark until about 37.0 years. KM Survival curves for dolphins held in U.S. and foreign facilities are shown in Figure 3c. The mean survival between dolphins housed in U.S. facilities and foreign facilities (18.9; 95% CI: 17.4.-20.4 yrs.) and (29.6; 95% CI: 27.2- 32.0 yrs.) respectively was significantly different (P=0.000). Dolphins housed in U.S. facilities displayed a much more drastic decline in survival from the time they entered captivity until year two. Decline of survival begins to slow at this time; however, U.S. facilities continue to demonstrate lower survival than those housed in foreign facilities. From year 20-43 there were no recorded deaths that occurred, therefore displaying that the likelihood of an animal dying during this age bracket is slim. After the age of 43, survival declines, though the cumulative survival never falls below 50%. KM Survival curves for wild-capture vs. captive-born dolphins are shown in Figure 3d. The median survival for wild-captured dolphins was 11.7 years (95% CI: 9.9-13.5 yrs.) and 15.8 years (95% CI: 10.7-21.0 yrs.) for captive-born dolphins. Utilizing the Mantel-Cox log rank test, survival estimates were not significantly different (ᵪ2=2.32; P=.13). Captive-born dolphins display a drastic decline in survival up until year 3.0 and continue to have a relatively constant decrease throughout time. Whereas, the decline in survival for wild-capture animals is not as drastic, the steep decline lasts much longer than those that were captive-born. After year 15.0 the decline in survival for wild capture, becomes relatively constant. KM Survival curves for Pre-vs. Post- 1985 can be seen in Figure 3e. The difference in median survival between dolphins entering captivity prior to 1/1/1985 and those entering post-1985 (9.6 yrs.; 95% CI: 8.3-11.0 yrs.) and (23.4 yrs.; 95% CI not possible due to too many cases being censored) respectively, was significant (P=0.000). Both pre- and post- 1985 display a drastic decline in survival until year 2.0. From year 2.0-6.0 there is a slight increase in survival for Pre- 1985. Post-1985 continues to show decline in survival until year 5.0 when Pre- and post- cross and both continue to decrease in survival at different rates. Figure 3: Kaplan Meier survival curves representing the number of T. truncates alive over time for all dolphins (a), each sex (b), U.S. and foreign facilities (c), U.S. and Foreign Facilities (d), Wild-Capture and Captive-Born (e), Pre and Post 1985. Horizontal line indicates 50% survival. Tick marks on each curve represent a specific censored animal, with some overlapping. Asterisk indicates significant differences between survival estimates. e. d. c. b. a. Post-1985 Pre-1985 Wild-Capture Captive-Born U.S. Foreign Females Males p=.000* p=.000* Cox Proportional Hazards Regression We first evaluated the data using the Omnibus Tests of Model Coefficients. All covariates were entered into the model with results demonstrating that the data were a good fit to the Cox Model (χ2=97.10; P=0.000). All predictors were entered simultaneously. Results of the analysis indicates that facility location (U.S. vs. foreign) (P=0.000; hazard ratio=.523; CI of hazard ratio: .439-.624), and pre- vs. post- 1/1/1985 (P=0.000, hazard ratio=.648, CI of hazard ratio: .555-.757) were significant predictors of the hazard rate. The interaction between sex and wild-capture vs. captive born were not significant. Holding all other covariates constant, bottlenose dolphins held in foreign facilities face a 52.4% decrease in hazard ratio and a 34.3% chance of death on any given day, compared to dolphins housed in facilities in the U.S. Similarly, dolphins that entered captivity after 1/1/1985 face a 65.1% decrease in hazard ratio and 39.3% chance of death on any given day, compared to those that entered captivity prior to 1/1/1985. Discussion Employing KM analyses, we found that the overall median survival estimate over the 62 year history of maintaining bottlenose dolphins in captivity was 13.0 years, with 12.0 years for females and 14.1 years for males. Contrary to the higher survival that females in many mammal species typically exhibit both in captivity and in the wild (Trivers 1985; Smith 1989), our data displayed that in the survival estimates based on sex, males displayed a higher survival rate than that of females in captivity; however, there was not a statistically significant difference. When studying bottlenose dolphins on the west coast of Florida, Irvine et al. (1981) found that the oldest living female examined was estimated to be 46 years old while the oldest male examined was 34 years old. The findings of Irvine et al. display what is typically found in cetaceans whereas the female displays longer lives than males. Typically this is seen due to the ability for males to reproduce throughout their entire life, therefore maintaining a costly trait (Hill 1991). Females exhibiting higher survival may also be due to the length of maternal care required for her offspring, especially in cetaceans which display strong maternal care and extended post-menopausal periods. The findings of males in captivity enjoying higher survival than females may also be due to the inability for female dolphins in captivity to escape the sexually aggressive male dolphins when confined in their small tanks. Male dolphins when sexually active can be very persistent and spend a large amount of time bothering or harassing the females. This may add additional stress to the female dolphins, which therefore may be adversely affecting their health and survival (John Jett pers. comm.). Even though female and male survival failed to differ statistically, future research should further examine why females demonstrate poorer survival than males in captivity. Future research should also examine how artificial impregnation may impact the survival of young captive females, since this has become widely popular practice in oceanaria, and aquatic parks today. The overall median survival for whales held in foreign facilities (29.6 yrs.; 95% CI: 27.2-32.0 yrs.) differed significantly from the overall median survival of those housed in U.S. facilities (18.9 yrs.; 95% CI: 17.4- 20.4 yrs.) (P=0.000). With the difference in regulatory oversight and zoo and aquarium regulation between U.S. and foreign countries (Cooper 2003), we might logically expect to see higher survival among U.S.-based dolphins. Further, U.S. based facilities housed 87% of the dolphins who entered captivity prior to 1970, thus suggesting a longer and more established care history for U.S.-based facilities. Based on our data, there are 3.5 times more bottlenose dolphins housed in U.S. facilities than those housed in foreign facilities. This finding may be due to the large difference in the number of dolphins housed in U.S. facilities (1,226 individuals) in comparison to those in foreign facilities (342 individuals). Also, As of 2015, the U.S. had a total of 74 different facilities housing 1,226 bottlenose dolphins, while there were only 49 foreign facilities that were housing 342 bottlenose dolphins. This allows us to determine that on average, each U.S. facility housed 16.5 dolphins, while on average each foreign facility is housed 6.9 dolphins. While the United States may have increased technology and greater regulation placed on the facility and care specifications, the foreign facilities have approximately half the amount of dolphins that they have to care for and maintain, which would allow more health care to be available for each individual, leading to a higher survival rate. With regards to each facility having fewer dolphins, there may also be less social stress that the dolphins are exposed to in captive environments with dolphins from various pods and families. For captive animals, being confined to a limited environment and among incompatible social groupings can escalate and social encounters can become more intensive due to limited means of escaping the situation (Hedinger 1964). Thus to further examine why animals in foreign facilities enjoyed higher survival than those in U.S. facilities, we stratified the pre- and post- 1985 cohorts. After stratification, foreign survival was still higher than U.S. survival as was wild- captured versus captive-born survival The proportion of captive-born to wild-captured bottlenose dolphins has increased throughout the years. However, median survival between captive-born and wild-capture was not statistically significant between the two (ᵪ2=2.32; P=.13). While the captive-born individuals have a higher decline in survival during the beginning years of life to about 10.0 years of age, the survival curve suggests that survival for captive-born animals improves over that of wild-captured animals beyond 10.0 years, the median survival of these. There are several variables that could lead to higher survival in captive-born animals than that of wild-capture. Firstly, the additional survival of the captive-born animals could be due to the fact that wild-captured animals come in to captivity at some age above zero, so they are already at a disadvantage age-wise. Due to this disadvantage, it would be suggested that captive-born animals would live longer due to having less years already accumulated when entering captivity. When looking at the survival of wild-captured dolphins, the stress that is involved in the capture and transfer into captivity could be a major factor in the reduction of survival once in captivity. Because of this additional stress, these animals require a period of time that Small and DeMaster (1995) termed the “period of acclimation” which is a period of time used to allow the dolphins to become acclimated their new habitat. In Small and DeMaster’s study, when looking at the survival of the wild-captured dolphins, once the dolphin had entered captivity, they were given an acclimation period of 3 days in order to eliminate deaths that were caused by stress in order to present a more accurate display of overall dolphin survival in captivity. Future research should attempt to facture in age-upon-arrival estimates for wild-captured dolphins. When looking at differences in survival based on pre- and post- 1985, there was a statistically significant difference between the survival of dolphins kept in captivity prior to 1985, and those kept in captivity post-1985 (18.9 yrs.; 95% CI: 17.4-20.5 yrs.) and (17.6 yrs.; 95% CI: 16.7-18.6 yrs.) respectively, was significant (P=0.000). We utilized 1985 as a landmark in time for estimating higher survival because, in 1985 veterinary care and pool designs for maintaining these animals in oceanaria were sufficient enough to keep the first baby killer whale alive for more than a year in captivity. The KM survival curves we generated allow us to determine points in time in which bottlenose dolphin survival deteriorates in captivity. With this information we are able to inform animal care personnel those ages in which there is a higher probability of death and survival in order to better health management practices. For example, the survival curve for captive-born dolphins displays a drastic decline in survival up until age six at which time, survival continues to slightly decline. Given these results, oceanaria management should place special precaution and care to the juvenile dolphins that fall within this age category. One way to utilize these results to increase survival of these age cohorts in captivity could be to attempt to reduce stressful situations that dolphins within these parameters are exposed to, such as mother calf separations, transportation between facilities, etc. We made no attempt to perform survival comparisons between captive bottlenose dolphin populations and wild bottlenose dolphin populations due to the difficulty of comparison between the two. According to Irvine et al. (1981) and Venn-Watson et al. (2015), median survival may vary between differing populations. Wells et al. (2013) states that the mean age for a closely monitored free-ranging population in Sarasota, Florida is 19.9 years while results of another study of free-ranging dolphins in the Indian River Lagoon, Florida indicates that mortality rate increases significantly for dolphins after the age of 15 and few dolphins survive 30-35 years (Stolen et al. 2003). Lastly, Sergeant et al (1973) determines that free-ranging dolphins from North-East, Florida are estimated to live a mean of 25 years. In order to be able to accurately compare the two populations (captive bottlenose dolphins vs. wild bottlenose dolphins) more comprehensive research using Kaplan-Meier and Cox Proportional Hazards should be done. Based on the knowledge on free-ranging bottlenose dolphin age provided by Wells, Stolen and Sergeant et al, Comparing the statistically inappropriate mean (rather than median) survival of captive-born dolphins in our study (25.9 years) with estimates of wild bottlenose dolphin survival (e.g., Sergeant et al.,1973, Stolen et al., 2005, Wells et al. 2013) the two appear not to differ substantially. However due to the censoring of animals still alive at the conclusion of most survival studies, mean survival values are rarely cited in the literature. Indeed. In the present study, mean survival (25.9 years) among captive born dolphins is a much higher estimate than the more statistically appropriate median survival (15.8 years). Further research between wild and captive survival using the KM model applied to both is needed to better understand how survival between the two actually differ. Literature Cited Alker, S.C. 1996. Marine Mammal Protection Act: Refocusing the Approach to Conservation, The UCLA Law Review, Vol. 44, 2:527-578 Carter, N. (1982). Effects of psycho-physiological stress on captive dolphins. International Journal for the Study of Animal Problems 3:193-198. Clubb, R., M.Rowcliffe, P.Lee, K.Mar, C.Moss and G.Mason 2008. Comprised survivorship in zoo elephants. Science 322:1649 Clubb, R., M.Rowcliffe, P.Lee, K.Mar, C.Moss and G.Mason. 2009. Fecundity and population viability in female zoo elephants: problems and possible solutions. Animal Welfare 18:237-247 DeMaster, D.P., and J.K. Drevanak. 1988. Survivorship patterns in three species of captive cetaceans. Marine Mammal Science 4:297-311. Dierauf LA. 1990. Stress in marine mammals. In: Dierauf LA, editor. Handbook of marine mammal medicine: health, disease and rehabilitation. Boca Raton: CRC Press. p 295–301. Dunn, J. Lawrence, John D. Buck, and Stephen Spotte. "Candidiasis in captive cetaceans." Journal of the American Veterinary Medical Association181.11 (1982): 1316- 1321. Fowler ME. 1978. Stress. In: Fowler ME, editor. Zoo and wild animal medicine. Philadelphia: WB Saunders. p 33–4. Hill, G.E. 1991. Plumage coloration is a sexually selected indicator of male quality. Nature, Land. 350, 337-339 Hosmer, D.W., and S. Lemeshow. 1999. Applied survival analysis. Wiley, New York, NY. Irvine, A. B., M. D. Scott, R. S. Wells and J. H. Kaufmann. 1981. Movements and activities of the Atlantic bottlenose dolphin, Tursiops truncatus, near Sarasota, Florida. Fish. Bull. U.S. 79:671-688. Jett, J., and J.Ventre. (2015) Captive killer whale (Orcinus orca) survival. Marine Mammal Science, 31:1362-1377 National Oceanic and Atmospheric Administration. Accessed 3/21/15: http://www.nmfs.noaa.gov/pr/permits/inventory.htm. Nuss, K., and M. Warneke. 2010. Life span, reproductive output, and reproductive opportunity in captive Geoeldi’s monkeys (Callimico goeldii). Zoo Biology 29:1-15 Promislow, Daniel EL. "Costs of sexual selection in natural populations of mammals." Proceedings of the Royal Society of London B: Biological Sciences 247.1320 (1992): 203-210. Sapolsky R. 1994. Why zebras don’t get ulcers: a guide to stress, stress related diseases and coping. New York: WH Freeman. 367 p. Sergeant DE, Caldwell DK, Caldwell MC. Age, growth, and maturity of bottlenosed dolphin (Tursiops truncatus) from Northeast Florida. J Fish Res Board Can 1973;30:1009–1011. Small, J. & DeMaster, D.P. (1995). Survival of Five Species of Captive Marine Mammals. Marine Mammal. Science 11: 209-226. Smith, D.W.E. 1989. Is greater female longevity a general finding among animals? Biol. Rev.64, 1-12 Stolen MK, Barlow J. A model life table for bottlenose dolphins (Tursiops truncatus) from the Indian River Lagoon System, Florida, U.S.A. Mar Mamm Sci. 2003;19:630–649. Spottswood, L., S.L. Spruance, J.E. Reid, M. Grace and M. Samore. 2004. Hazard ratio in clinical trials. Antimicrobial Agents and Chemotherapy. Thomson CA, Geraci JR. 1986. Cortisol, aldosterone, and leucocytes in the stress response of bottlenose dolphins, Tursiops truncatus. Can JFish Aqu Sci 433:1010–6. Trivers, R.L. 1985. Social Evolution. Menlo Park, California: Benjamin Cummings. Venn-Watson S., Jensen E, Smith CR, Xitco M, Ridgeway SH (2015) Annual Survival, Mortality and longevity of bottlenose dolphins (Tursiops trancatus) at the U.S. Navy Mammal Program 2004-2013. J Am Vet Med 246: 893-898 Waples K.A, Gales N.J.. 2002. Evaluating and Minimizing Social Stress in the Care of Captive Bottlenose Dolphins (Tursiops aducus). Zoo Biology 21:5-26 Wells RS, McHugh KA, Douglas DC, et al. Evaluation of potential protective factors against metabolic syndrome in bottlenose dolphins: feeding and activity patterns of dolphins in Sarasota Bay Florida Frontiers Endocrinol 2013;4:139