High calf mortality in bottlenose dolphins in the Bay of Islands, New Zealand–a local unit in decline

Photo-identification

Boat-based, photo-identification surveys were conducted in the Bay of Islands (35.14°S, 174.06°E; Fig. 1), New Zealand from 1994 to 2006. The study area is an open embayment of approximately 244 km², including large estuaries with varying hydrological conditions ranging from estuarine to oceanic (Booth 1974). Opportunistic and directed photo-identification surveys were conducted from tour boats and an independent research vessel. For each encounter, basic data were collected including time of encounter, group size, age-class composition, and individual photo-identification (Tezanos-Pinto et al. 2013). Attempts were made to collect photo-identifications of each individual in the group without bias towards distinctively marked dolphins (Würsig and Jefferson 1990). Individual identification photographs were taken using a Canon EOS 35 mm camera with 200 and 400 ASA color film and a Canon digital camera equipped with a 200–400 mm lens.

Nicks and cuts on the trailing edge of the dorsal fin were the only feature used for identification purposes. Analysis of photo-identification data followed Tezanos-Pinto et al. (2013). Briefly, all dorsal fin photographs were classified into four categories of quality (excellent, good, fair, and poor) according to focus, size of the dorsal fin relative to the frame, exposure (contrast between dorsal fin and background), and relative angle to the dolphin. Only excellent and good quality photographs were matched to the Bay of Islands catalog and included in the analyses. Photographs of unique individuals were graded according to a scale including three levels of distinctiveness ranging from one (small marks) to three (large marks). Before adding a new individual into the catalog images were checked by three experienced researchers and all photographs were cross-matched to the rest of the catalog.

Group Composition

Four age size-classes were considered based on visual assessment of body sizes as compared to average adult size: neonate (<3 mo old and present visible dorso-ventral fetal folds, paler coloration, and swimming in an “infant position”; Mann and Smuts 1998); calf (no fetal folds, estimated 3 mo old to 3 yr old, measures half the size of an adult, swimming in an “infant position”; Mann and Smuts 1998); juvenile (3 yr old to approximately 9 yr old, not sexually mature, measured 2/3 the size of an adult; Kasuya et al. 1997, Wells 2000); and adult (sexually mature dolphin, usually more >9 yr of age, measured ca. 3–3.5 m in length). For the purposes of our study, neonates and calves <1 yr old are referred to as “young of the year.”

Sex Identification, Reproductive and Nonreproductive Females

The sex of individuals was determined by observations of mother-calf association, molecular markers, or visual examination of the genital slit. For molecular sexing, total genomic DNA was isolated from biopsy samples (Tezanos-Pinto 2009, Tezanos-Pinto and Baker 2012). The sex of biopsy sampled and tissue-sampled beachcast dolphins was identified by amplification of a fragment of the sry gene multiplexed with fragments of the ZFY/ZFX genes as a positive control (Gilson et al. 1998).

Photo-identification Database

Photo-identification data collected from 1994 to 2006 included individual identification photographs, information on group size, composition, and mother-calf associations. This large data set provides substantial information for the estimation of reproductive parameters. Nonetheless, different data sets were used to provide the best estimate for the different parameters in order to minimize potential bias and maximize the use of the data. Calving interval was estimated using a subset of the 1994–2006 data set consisting only of those calves for which fate could be documented by analysis of the encounter history of the mother.

Female Reproductive Parameters

Individually identified adults were assumed to be reproductively mature females when they were consistently associated with a calf during two or more subsequent and independent encounters. In most cases, this was confirmed by individual identification photographs of the mother swimming in close association with the calf (Mann and Smuts 1998). Reproductive females (i.e., sexually mature females) were defined as those known to have given birth to viable calves during the study period. nonreproductive females were defined as adults positively sexed as females but were never sighted in close association with a calf during the 12 yr study period.

Estimation of the annual calving rate was restricted to sightings of mother-calf pairs from 1994 to 2006 that were resighted over two consecutive encounters and there was an estimated age of the calf (see below). A binomial variance was used to calculate the confidence interval for the parameter with 95% limits (Baker et al. 1987, Wells and Scott 1990).

The annual calving rate was estimated as the total number of young of the year divided by the total number of reproductive females sighted during that year (Baker et al. 1987). This assumes that the chances of sighting a female are not affected by the presence or absence of its calf (Baker et al. 1987). However, we conducted fine-scale boat surveys and it is possible that some females are more prone to approach a boat than others or tend to use the area more frequently (i.e., frequent users vs. transient animals; Tezanos-Pinto 2009), resulting in higher capture probabilities. Conversely, some females may have shifted to other locations or may be less prone to boat approaches, resulting in lower capture probabilities. Only a single female dolphin was observed giving birth; therefore for almost all data points the first sighting of a female accompanied by a neonate was used as an indication of date of birth. It is acknowledged that this is a minimum date of birth and is referred to as the “approximate date of birth.”

Calving interval was estimated as the time elapsed between subsequent births; this was estimated as the period of time the female was first sighted accompanied by a calf to the date of first sighting of the female accompanied by her next calf. Previous research suggest that most births in the Bay of Islands occur during spring and summer (Constantine 2002, Tezanos-Pinto 2009); therefore calving intervals were estimated from those females with consecutive births and a minimum of two sightings during the reproductive season.

Apparent Calf Mortality

Calf mortality was calculated as the number of the young of year (<1 yr old) that were inferred to have died, divided by the total number of young of the year assigned to individually identified mothers (Wells and Scott 1990). Second-year calf mortality was calculated as the proportion of calves that were inferred to have died before reaching two years of age of life, divided the total number of 2 yr old calves assigned to individually identified mothers. Dolphins in the Bay of Islands have changed their patterns of residency and habitat use over time, with an apparent shift in home range (Tezanos-Pinto 2009, Hartel et al. 2014). In a stable population that had not undergone a shift in home range, it would be expected that all calves would have equal survival probabilities. However, for animals that have previously been resighted in the study area for a number of years prior to going missing, this assumption is unrealistic. These dolphins may have either relocated or been displaced to a different habitat that may, or may not offer the same chance of survival, or else they have died. For this reason, we estimated “apparent calf mortality” (mortality potentially confounded with emigration) and provide two scenarios: (1) calves of mothers not resighted in the Bay of Islands over subsequent years have all died and (2) calves of mothers not resighted in the Bay of Islands over subsequent years have all survived.

If the mother was sighted without the calf and the calf was <1.5 yr old, the calf was assumed to have died. This is because the minimum weaning age in the bottlenose dolphin has been estimated at 18–20 mo (Smolker et al. 1992, Wells and Scott 1999). A calf that was associated with the same mother over this period was assumed to be the same calf, providing its estimated age was concordant with previous sightings. To avoid potential errors caused by uncertainties regarding a calf's year of birth or age, only data from known neonates or very young calves in a given year were used in the analysis. The approximate minimum age a calf survived was estimated by considering the interval between the first sighting of the mother-calf pair to the last sighting of the pair.

Effort

From March 1994 until May 2006, a total of 704 surveys were conducted, during which 5,577 photo-identification sightings were collected from 625 groups of dolphins (Table 1). Surveys were conducted in all seasons; a total of 74% of surveys (n = 519) were conducted during the main reproductive season (September to April). To evaluate whether effort was comparable across years, a survey ratio was calculated as the proportion of surveys conducted per reproductive season over the total number of surveys. Effort per reproductive season varied across years, but was comparable between periods with consistent effort (mean survey ratio 1997–1999 = 0.74, mean survey ratio 2003–2006 = 0.80; Table 1).

Reproductive and Nonreproductive Females

From 1994 to 2005, a total of 64 individually identified dolphins were sexed as females based on associations with calves (69%), molecular sexing (16%), both molecular sexing and association with calves (9%), or observation of the genital slit (6%). Of these, 53 females were observed over a cumulative total of 1,342 resightings, of which 27% (n = 368) included a neonate or calf (Table 2). Overall, 66% of these females (n = 35) were sighted with one calf, 23% were sighted with two calves (n = 12), and 11% (n = 6) were sighted with three different calves over the study period (Table 2). When a mother-calf pair was not resighted in the Bay of Islands in consecutive years (i.e., ID 80, 296, and 325; Table 2), the data were excluded for the estimation of calving rate.

Out of the 64 dolphins sexed as females, 10 were never sighted in association with a neonate or calf during the study period (1994–2006) despite extensive sighting histories in some cases (e.g., ID 193, 454; Table 2). One female (ID 229) was sighted in association with a calf once, but this association could not be confirmed by further sightings. These eleven dolphins were sighted over a cumulative total of 350 encounters and were sexed by direct observation of the genital slit (n = 3, ID131, 309, and 339; Table 3), molecular sexing (n = 7), and observation of the carcass after stranding and molecular sexing (ID 300; Table 3).

Female Reproductive Rate and Calving Interval

Calving rates were evaluated using a subset of the data that consisted of years with intense effort and consistent methodologies (1997–1999 and 2003–2005). We excluded data collected during 2000 because there were no suitable records regarding the approximate age of calves; further mother-calf pairs were not recorded in detail as per intensive surveys conducted during 1997–1999 and 2003–2006. Annually, calving rates during 1997–1999 ranged from 0.11 to 0.19 young of the year/reproductive female/year, with an overall average of 0.13 young of the year/reproductive female/year; 95% binomial CL = 0.07–0.21). For 2003–2005, annual calving rates ranged from 0.21 to 0.32 young of the year/reproductive female/year with an overall average of 0.25 young of the year/reproductive female/year; 95% binomial CL = 0.16–0.35; Table 1).

The calving interval could be inferred for seven females that gave birth to a cumulative total of 15 calves from 1994 to 2006 (Table 4). Based on this, the average calving interval was estimated at 4.3 yr (95% CL = 3.2–5.4, SD = 1.45, median = 4.42).

Apparent Calf Mortality

Since 1994, 53 young of the year dolphins were observed in close association with their mothers in two or more independent encounters. Excluding two calves born in 2006 and one in November 2005 (no data were available after this period), there were 50 calves born in the Bay of Islands during 1994–2005. Seventeen calves were presumed dead within their first year of life based on subsequent sightings of their mothers without the calf, whereas 24 were assumed to have survived based on subsequent sightings of mother-calf pairs. In addition, there were nine mothers that were not resighted in the area afterwards and therefore, the fate of their calves was unknown (Table S1). Apparent calf mortality ranged from 0.34 (assuming that the calves from undocumented mothers all survived; 95% CI = 0.21–0.48) to 0.52 (assuming that calves from undocumented mothers all died; 95% CI = 0.37–0.66) for the first year of a calf's life.

The fate of 19 calves could be further documented into their second year of life based on resightings of mother-calf pairs; of these five died before reaching two years of age. There were 12 additional calves whose mothers were not resighted (undocumented mothers). Calf apparent mortality for the second year ranged from 0.15 (assuming that the calves from undocumented mothers all survived; 95% CI = 0.02–0.24) to 0.59 (assuming that calves from undocumented mothers all died; 95% CI = 0.41–0.75). Overall, calf mortality during the first two years, considering those mothers with unknown fate, was estimated at 0.44 (95% CI = 0.30–0.59; Table S1).

Female Reproductive Rates

Calf mortality and female reproductive rates both contribute to a population rate of increase. We used two different estimates of calving rate; one based on sequential sightings of individual dolphins and a second based on the frequencies of calves born to known females. The average calving rate observed here of 0.25 young of the year/mature female/year during 2003–2005, suggests a crude birth rate of 6.25%, assuming that half the population are females (Tezanos-Pinto 2009) and that half are reproductive females (0.25 × 0.50 × 0.50 × 100). This is lower than values reported in other populations found in similar habitats that used similar age class definitions (range 7.15%–11%; Würsig 1978, Hansen 1990). However, the calving rate obtained here of 0.25 does not take into account calf mortality. This estimate suggests that on average, a female gives birth only once every 4 yr, which is consistent with our observed calving interval (4.3 yr).

Estimates of calving rate varied considerably between periods. Fluctuations in calving rates among dolphins, however, are not uncommon. For example, Hasse and Schneider (2001) reported a 100% increase in calving rate from one year to the next. This was attributed to several females reaching sexual maturity at the same time. Herzing (1997) found a similar pattern with Atlantic spotted dolphins (Stenella frontalis) and suggested that years with higher births may reflect several females attaining sexual maturity simultaneously. It is likely that fluctuations observed between periods (1997–1999 and 2003–2005) in our study, may reflect a cohort effect among mature females. A similar finding was recently reported in Doubtful Sound, where periodic fluctuations in calving rates were attributed to a group of females synchronizing their births and being more successful at rearing calves to weaning age than other groups of females (Henderson et al. 2014).

Calving intervals of female bottlenose dolphins in the Bay of Islands are similar to those reported in other regions (mean = 4.3 yr, SD = 1.45). For instance, in Sarasota Bay, female bottlenose dolphins with surviving calves have a minimum calving interval of 2 yr, but 3–6 yr intervals are more common (Wells and Scott 1999). In Shark Bay, the median interbirth interval for female bottlenose dolphins is 4.1 yr (Connor et al. 1996). Longer periods of parental care could be explained by differences in nutrition levels, body size and the time it takes a calf to become proficient at catching prey (Connor et al. 1996). We acknowledge, however, that longer calving intervals are less likely to be observed in long-lived mammals, since its detection will depend on the duration of the study (Barlow and Clapham 1997) and female resighting rates. Therefore, calving intervals estimated here are likely to underestimate the true values.

Nonreproductive Females

Unlike most prior studies, there were known females (from observation and molecular markers) that were never sighted with a calf. Of the 64 known females, nine were never sighted with a calf and one female was sighted with a calf only once. Four of these nine females had a minimum age of 16–18 yr. It is possible that these females had a calf that died before being sighted or that miscarried. However, three of these females were regularly encountered and photographed without calves (sighting frequency ranged between 14 and 62 sightings over a 12 yr period). Such nonreproductive females have been reported elsewhere. In Mikura Island, Japan, two adult females, out of a photo-identification catalog of 169 individuals, were never sighted with calves over an 8 yr period (Kogi et al. 2004). Similarly, in Doubtful Sound four females known to be more than 12 yr old have not calved yet (Henderson et al. 2014). Nevertheless, nine out of 64 females seems a high proportion and would contribute to lower overall potential reproductive output of the local Bay of Islands dolphins.

Bottlenose Dolphins in the Bay of Islands Have High Calf Mortality

Overall, 34%–52% of calves observed in the Bay of Islands were inferred to have died before their first year of life and, of those surviving their first year, another 15%–59% died before reaching their second year of life. These values are higher than those reported elsewhere (first year mortality range 13%–39%, second year mortality range 10%–38%; Table 5). The high calf mortality observed in the Bay of Islands is comparable to that of bottlenose dolphins in captivity (first year mortality = 39%; DeMaster and Drevenak 1988) and dolphins subjected to anthropogenic disturbance such as human provision feeding of wild dolphins (i.e., human provisioning first year mortality = 56%; Mann et al. 2000).

Biases in Calf Mortality and Female Reproductive Rates

It is acknowledged that our estimates may include some biases. Calf mortality estimates presented here are likely to be underestimates of true rates given that some calves could have been born and died before being observed. Similarly, calving rates could have been underestimated as some births may not have been observed. The requirement of two independent and consecutive sightings to confirm mother-calf associations could have resulted in an underestimation of reproductive parameters by having to remove single sightings (as occurs with transient animals). Conversely, calf mortality could have been overestimated if a calf was overlooked or temporarily separated from its mother while the photo-identification survey was underway and the pair later moved away from the area. In saying this, the majority of female dolphins were frequently resighted in the bay over the duration of this study, so we are confident that our results have captured most of the reproductive events for these dolphins. The requirement of two sightings for the assignment of mother-calf pairs can be challenging when individuals are infrequent or occasional visitors; but this step is necessary to avoid incorrect assignments. Although the mortality rate was assessed indirectly from calves that disappeared (as few direct observations of dead calves could be made), the methodology employed here is consistent with other long-term studies, resulting in reasonable comparisons (e.g., Wells and Scott 1990, Steiner and Bossley 2008, Henderson et al. 2014).

Potential Causes of Low Female Reproductive Rates and High Calf Mortality

Potential causes for low female fecundity are difficult to evaluate in wild populations of dolphins. One possibility is: diseases such as brucellosis (or Mediterranean fever). Bottlenose dolphins infected with Brucella have been known to abort their fetuses as a result of the infection (Miller et al. 1999). Analyses of blood samples collected from an adult female Hector's dolphin (Cephalorhynchus hectori) in Akaroa Peninsula (South Island) indicated she was infected with Brucella (Duignan 2004), confirming that this disease exists in other cetaceans in New Zealand waters. Improved sampling of beachcast and live bottlenose dolphins is needed to investigate this possibility. Another possible explanation for lower reproduction is exposure to behavioral disturbance due to dolphin-watch tourism. Elsewhere, there is evidence that the cumulative amount of time a female is exposed to dolphin-watching tour boats may affect reproduction (Bejder 2005). Given the high level of dolphin-watching tour boats in the Bay of Islands and the negative effect these have on the dolphin's behavior (Constantine 2001, Constantine et al. 2004) it is necessary to investigate the potential effects of tour-boats on female reproduction.

Causes of high calf mortality in the Bay of Islands are also uncertain. Known predators of bottlenose dolphins are large sharks and killer whales (Orcinus orca); however, dolphins in the Bay of Islands bear very few scars from sharks or killer whales suggesting that attacks are rare (or always fatal). It is acknowledged that true rates of predation are uncertain; however, with little direct evidence, it is unlikely that predation alone will explain the high calf mortality observed in the Bay of Islands.

It is also possible that females using the Bay of Islands do not feed as successfully as females in other areas. However, there is no obvious evidence of emaciation to suggest poor nutrition and, dolphins in the Bay of Islands are known to have a broad diet (Constantine 2002). If predation and poor condition of the mothers are not the direct cause of calf mortality, it would appear there are some other factors affecting the survival of the calves. Direct impacts on calf mortality in other populations include undetected entanglements, injuries from fishing gear or boat strikes (Wells et al. 2008). Injuries from all these sources have been observed on dolphins in the Bay of Islands and other areas throughout the range of the population (Dwyer et al. 2014a), but these are unlikely to be a major contributor to the mortality rates reported in our study.

Conclusion

Female reproductive rates estimated here provide baseline information on bottlenose dolphins using the Bay of Islands. The decline in abundance (Tezanos-Pinto et al. 2013) and high calf mortality are of concern and highlight the need for continued monitoring of this population. Further, other areas along the range of the northeastern North Island bottlenose dolphin population should be monitored given the shift in home ranges observed in the Bay of Islands (Tezanos-Pinto 2009, Hartel et al. 2014) and studies showing new areas with high dolphin encounter rates that were previously under-reported (Berghan et al. 2008; Dwyer et al., 2014b). We also recommend improved documentation of stranded and beachcast dolphins and retrieving of carcasses to determine causes of mortality. The current system administered by the Department of Conservation in New Zealand allows for reporting but little priority has been given to retrieving bottlenose dolphin carcasses and conducting comprehensive necropsies.

Given the high calf mortality reported here and studies showing impact on bottlenose dolphin reproductive success, management should attempt to minimize sources of anthropogenic impact, wherever possible. This could include a better enforcement on the ban on swimming with groups of dolphins containing calves, limiting tour boats interactions around groups of mother and calves, restricting the occurrence of potentially hazardous events during calving season (such as speed boat races), and extending the moratorium on further permits for commercial swim-watch operations in the Bay of Islands.