Dolphins, Whales and Research, OH MY!

Abstracts submitted to the 22nd Biennial Conference on the Biology of Marine Mammals based in whole or in part on Cascadia Research Collective’s Hawai‘i research program Worldwide phylogeography of the rough-toothed dolphin (Steno bredanensis) described using mitogenomes and nuclear introns Renee Albertson, Oregon State University Alana Alexander, University of Kansas, Karen Martien, NOAA Fisheries, Southwest Fisheries Science Center , Robin Baird, Cascadia Research Collective, Marc Oremus, WWF, World Wildlife Fund, Susana J. Caballero-Gaitan, Michael Poole, Marine Mammal Research Program, Robert Brownell, Deborah Duffield, Portland State University, C. Scott Baker, Oregon State university

Abstract:

Rough-toothed dolphins have a worldwide circumtropical to warm-temperate distribution in open- ocean and slope waters. Despite this large geographic distribution, the global phylogeography and the divergence between the Pacific and Atlantic Oceans has not been well described. Given their circumtropical distribution, continental landmasses have acted as barriers to dispersal. Therefore, we investigated the potential for ocean basins to lead to phylogeographic structure, and further evaluated population level structure among oceans, using mitogenomes (n=24), the mtDNA control region (n=351), and six nuclear introns (n=35) for representative samples from the Pacific, Indian and Atlantic Oceans. Phylogenetic analyses of the mitogenome revealed four clades including an Atlantic clade sister to a larger cosmopolitan clade containing individuals from all three oceans. Using the first mitogenome substitution rate estimated specifically for rough-toothed dolphins, we dated the divergence time between the endemic Atlantic clade and the cosmopolitan clade to one million years ago. Although the phylogenetic reconstruction did not reflect reciprocal monophyly by ocean basin, the mitogenome population level analyses revealed significant genetic differentiation among all oceans (global ΦST = 0.368, p<0.001). Nuclear introns indicate significant differentiation between the Atlantic and the other two oceans (FST=0.150, p<0.001 Pacific and FST=0.135, p<0.001 Indian), but no significant differentiation was found between the Pacific and Indian Oceans (FST=0.0091, p=0.441). Within each ocean, local isolated populations were also documented. Overall, these results suggest rough-toothed dolphins have maintained low levels of gene flow across the Indo-Pacific on an evolutionary time scale, and that dolphins in the Atlantic have been isolated sufficiently from conspecifics in the other oceans to be on an independent evolutionary trajectory.

Abstract:

Pelagic false killer whales (PFKW) are killed or seriously injured in the Hawai‘i-based deep-set longline fishery more than any other cetacean population, with bycatch regularly exceeding allowable levels. Knowledge of the movements and habitat use of this population is limited, but could inform efforts to mitigate bycatch. We use satellite tag data from five PFKW from three groups to assess longline fishery interactions and resource selection. In 2013, two whales (154d and 14d) were tagged in the Northwestern Hawaiian Islands, and a group of three (123d, 15d and 12d) off Hawai‘i Island. Tracks ranging from west of Midway to >1900km east of Hawai‘i Island were

Longline fishery interactions and resource selection of satellite-tagged pelagic false killer

whales in the North Pacific

David Anderson, Cascadia Research Collective

Robin Baird, Cascadia Research Collective

Amanda Bradford, Pacific Islands Fisheries Science Center, NMFS, NOAA

Daniel Webster, Cascadia Research Collective

1

compared to 4182 set positions made by the Hawai‘i-based deep-set longline fleet during the tracking period. Information on longline effort from other countries was not available. In one case, three PFKW changed direction and rapidly approached a set from almost 100km when the vessel was hauling gear and remained within the area being fished during the following set, before moving away during the third set. On some occasions (n=3), PFKW potentially reacted to the gear haul, but either stopped short of approaching the set or quickly transited the area of the set without stopping. Other times (n=2), PFKW did not appear to react to sets that were within 30km. Resource selection analysis was conducted on the three PFKW with the longest independent tracks, using a switching state space model to determine representative daily positions. We assessed positions as a function of the nearest longline set, as well as 25 oceanographic, atmospheric and lunar variables. The best fit model was chlorophyll-a concentration, distance to the nearest longline set, and significant wave height. Understanding the behavior of PFKW in response to both longline vessels and the natural environment is vital for ongoing efforts aimed at reducing bycatch.

Abstract:

In Hawai‘i, the diet of endangered main Hawaiian Islands false killer whales (FKWs) includes pelagic and nearshore game fish, overlapping almost entirely with fishery catches. Interactions with nearshore fisheries are one of the greatest threats to this population, with at least 23% of individuals having injuries suggesting fishery interactions. Meetings with stakeholders have primarily brought fishermen to the table that fish off Kona, the area within the islands with the highest levels of fisheries effort/catch. At-sea observer or electronic monitoring programs to assess interactions and bycatch have been discussed, but there are more than 3,000 fishing license holders in Hawai‘i, and prioritizing monitoring locations is needed. To determine the areas of greatest overlap between FKWs and fisheries, we developed an index using satellite tag (n=28) and fisheries catch data (from 1994-2014). The index reflects the probability of a fishermen encountering FKWs relative to fish catch by area, which should reflect the likelihood of depredation and potential bycatch. Probabilities for each area were scaled relative to Kona (=1). Of the 97 fisheries statistical areas that overlapped with FKWs, nine areas had 52% of the catch, and FKWs spent just 7.4% of their time in those areas. Areas east of O‘ahu and north of Moloka‘i and Maui had overlap index values ranging from 124 to 782 times that of Kona, while areas off the north end of Hawai‘i Island had index values ranging from 65 to 2,887 times that of Kona. Our results suggest that management agencies need to bring fishermen who fish in these high overlap index areas to the table, in order to inform discussions of depredation, bycatch and potential solutions. Efforts to reduce and monitor fisheries interactions, including outreach efforts with fishermen, and possible observer or electronic monitoring programs, should concentrate on these areas with high index values.

Abundance Estimates for Management of Endangered False Killer Whales in the Main Hawaiian Islands
Amanda Bradford, Pacific Islands Fisheries Science Center, NMFS, NOAA
Robin Baird, Cascadia Research Collective

Sabre Mahaffy, Cascadia Research Collective
Antoinette Gorgone, Cascadia Research Collective
Daniel McSweeney, Wild Whale Research Foundation
Tori Cullins, Wild Dolphin Foundation
Daniel Webster, Cascadia Research Collective
Alexandre Zerbini, Marine Mammal Laboratory, Alaska Fisheries Science Center

2

Abstract:

Data limitations frequently result when monitoring endangered populations that are rare, cryptic, or inaccessible. Appropriately using the best available data to meet management mandates for these populations is a common conservation challenge. False killer whales (Pseudorca crassidens) occur as three distinct populations in Hawaiian waters, including a population resident to the main Hawaiian Islands (MHI) that is endangered under the U.S. Endangered Species Act and strategic under the U.S. Marine Mammal Protection Act. A longitudinal photo-identification dataset representing 574 yearly encounters (i.e., identifications compiled at an annual scale) of 171 distinctive individuals and open mark-recapture methods were used to estimate annual MHI false killer whale abundance as needed for management of this population. The data are from dedicated and opportunistic surveys conducted from 2000-2015 around the MHI and reflect unquantified spatiotemporal biases imposed by constraints on sampling (e.g., rough sea conditions on windward sides of islands). Accounting for temporal variation and especially social group affiliation was important in modeling capture probability. Sensitivity analyses found the resulting time series of 16 abundance estimates is robust to some forms of sampling variability and bias. However, because the study area was partially sampled each year, the annual abundance estimates apply only to the portion of the population using the sampled area and may underestimate true population abundance. Nonetheless, the resulting estimates and supporting evidence indicate the MHI false killer whale population is relatively small; for example, only 167 (SE = 23, 95% CI = 128-218) individuals were estimated to have used the sampled area in 2015. Until data are available to estimate or overcome existing sampling biases, the present estimation framework offers a tool for using data that have been regularly collected each year to produce annual abundance estimates that are improvements over current inputs to management.

Abstract:

Rough-toothed and common bottlenose dolphins are known to be resident to the western main Hawaiian Islands, an area that overlaps with extensive Navy activity. To estimate home ranges for both species and overlap with Navy training areas we used location data from LIMPET satellite tags deployed on 17 rough-toothed and 13 bottlenose dolphins from 2011-2016. Kernel density interpolation with barriers was used to produce 50% and 95% home range estimates for each species, individual, and season within species. Polygon areas were related to total tag attachment duration, but median tag durations were short (rough-toothed=14.8 days; bottlenose=12.5 days), making it difficult to determine if the data asymptotes. Both species had similar median tag durations but differed in the areas of their core ranges (rough-toothed dolphin – 754 km2; bottlenose dolphin – 280 km2) and by location. Rough-toothed dolphin core range was in deeper water around Kaua‘i and Ni‘ihau with some movement around O‘ahu, and a larger range in winter months. Bottlenose dolphin core range was centered in <1,000m deep water around Kaua‘i, utilizing a larger area during winter and fall and a smaller area in summer. Median tag durations for bottlenose were similar in fall and winter and longer in summer, suggesting the range expansion was not an artifact of tag durations. Although both species inhabit overlapping geographical areas, differences in depth preference suggest that their home ranges have limited overlap. Individual rough-toothed dolphins generally moved widely within their range, with one moving as far as O‘ahu, while individual bottlenose dolphin movements were more limited, with the exception of one individual that moved as far as O‘ahu. The core areas of both species overlapped extensively with the Pacific Missile Range Facility, the primary Navy training area in Hawai‘i. Identifying habitat and seasonal variations will inform conservation and management of these populations.

Rough-toothed and bottlenose dolphin home ranges in the western main Hawaiian

archipelago

Megan Lewis, Cascadia Research Collective

David Anderson, Cascadia Research Collective

Daniel Webster, Cascadia Research Collective

Robin Baird, Cascadia Research Collective

3

Abstract:

False killer whales are highly social, known for maintaining strong, long-term bonds and engaging in cultural behaviors including prey-sharing and mass stranding that may make them vulnerable to anthropogenic impacts. An endangered main Hawaiian Islands population includes three recognized social clusters (C1, C2, C3). Although they appear longitudinally stable, all three are interconnected and some individuals are regularly associated with multiple clusters. Such individuals may have a strong impact on the health of the population by disseminating information between clusters or acting as a vector for diseases. To understand their influence on fine-scale social structure, as well as assess peripheral individuals that do not link clusters, we examined sighting histories (2000-2016) and associations of socially ambiguous individuals using photo-identification and social network analysis. We then simulated targeted removals of each ambiguous individual, imposed restrictions on association strength and recorded the impact on the social network. Depending on the restrictions used, either four or six peripheral clusters were identified using eigenvector-based modularity, two or three of which were mostly younger individuals seen 1-2 times and considered sample size artifacts. One of the remaining peripheral clusters was robust to change: at association strengths ≥0.3, all links between it and other clusters dissolved, warranting designation of a new cluster (C4). Links among established clusters were extensive but weak: almost half of all association links within the network disappeared when restricted to ≥0.3, leaving one individual linking C1-C2 and three individuals linking C2-C3. Links between C1 and C3 either disappeared entirely, with the two clusters linked instead through two remaining peripheral clusters, or were joined by a small number of individuals that clustered with either C1 or C3. These individuals may provide valuable lines of communication among clusters and help minimize inbreeding, although additional encounters may be required to elucidate their cluster affiliation.

Abstract:

Environmental barriers, social structure, and the divergence of foraging tactics have driven the development of strong genetic differentiation within many marine species, including the largest delphinids commonly referred to as ‘blackfish’ (subfamily Globicephalinae). The melon-headed whale (MHWs, Peponocephala electra) is a poorly known member of the blackfish family for which genetic structuring is largely unknown. MHWs are globally distributed in tropical and subtropical waters, and can form resident populations around oceanic islands. They frequently mass strand, suggesting strong social cohesion within groups. In order to explore potential drivers of the genetic population structure of this oceanic species, we used mitochondrial sequence data (mtDNA) and genotypes from 14 nuclear microsatellite loci (nDNA) from 232 MHWs sampled throughout their range. Based on their ecological and behavioral characteristics, we hypothesized that MHWs would exhibit strong regional genetic differentiation, similar to that observed in other members of the Globicephalinae subfamily. Instead we found only moderate differentiation (median mtDNA ΦST =

4

0.204, median nDNA FST = 0.012) among populations both within and between ocean basins. Our results suggest that populations of MHWs that are resident to oceanic islands maintain a higher level of genetic connectivity than is seen in most other blackfish. MHWs may be more behaviorally similar to smaller delphinids (particularly the spinner dolphin, Stenella longirostris), which are known to form coastal and island-associated resident populations that maintain genetic connectivity either through occasional long-distance dispersal or gene flow with larger pelagic populations. Our results suggest that differences in social organization may drive different patterns of population structure in social odontocetes.

Abstract:

Changes in the availability of light, both solar and lunar, are likely to have a strong influence on the behavior of vertically migrating animals, and the behavior of predators that forage on them. However, it is not clear what influence, if any, that lunar cycles or seasonal changes in light level have on the diving behavior of deep divers. In Hawai‘i, short-finned pilot whales regularly dive >800 m to forage. We used information from depth-transmitting satellite tags deployed on 34 short-finned pilot whales in Hawai‘i over an eight-year period to examine the influence of lunar and seasonal cycles on their movements and diving behavior. During a full moon, maximum dive depth was 47.8% deeper and dive duration was 22.5% longer, compared to during a new moon. This change appeared to be driven primarily by an increase (~20%) in the depth of deep dives (dives > 200 m) at night over the lunar cycle, and an increase in the proportion of deeper daytime dives during a full moon. Dive bouts occurred on average 11.4 km further offshore during the full moon compared to a new moon. During summer (longer day length), dives were deeper (33.9%) and longer (22.1%) compared to winter. Although other environmental variables (such as sea surface temperature and net primary productivity) also varied between seasons, the changes in these parameters did not appear to be associated with a change in diving behavior. This suggests that a change in light level, both solar and lunar, is likely to have an influence on the depth of the prey targeted by pilot whales, which in turn influences the diving behavior of the whales. The potential for such variations in foraging behavior across long temporal scales should be taken into consideration in studies measuring diving behavior over shorter time scales.

Abstract:

Social structure can increase genomic diversity and population structure within many marine and terrestrial species. In the marine environment, where long-distance vision is of limited value and there are few boundaries to dispersal, vocal repertoire may be an important mechanism to

5

maintaining social structure. Short-finned pilot whales form stable social groups for periods of a decade or more, and have highly variable vocal repertoires. Because of these traits, we expect to see genetic structure driven by the retention of individuals within family units, and genetic structure driven by an effect of social structure on mate selection. Furthermore, we hypothesize that vocal repertoire is an important mechanism used by short-finned pilot whales to maintain group segregation, and therefore plays an important role in driving social and genetic structure. Here, we focus on short-finned pilot whales in the Hawaiian Islands, which have a well-documented hierarchical social structure of resident island communities, which comprise clusters of individuals that often associate, and in turn comprise one or more social units that spend the majority of their time in association. We used 123 mitochondrial sequences (mtDNA), and 49 nuclear SNPs (nuDNA) from 106 short-finned pilot whales, and collected recordings from 26 encounters throughout the Hawaiian Islands. Our genetic data indicate a reciprocal link between social structure and genetic structure in this population. Our ability to differentiate among clusters on the basis of vocal repertoire indicates that it is an important mechanism maintaining social structure, and therefore helps stabilize the reciprocal relationship between social and genetic structure in this population. Socially-driven population structure, as seen in Hawaiian pilot whales, can increase genomic, ecological and cultural diversity, as well as ecological resilience, but could also indicate a greater degree of vulnerability to anthropogenic threats.

Abstract:

Cookiecutter sharks (Isistius spp.) are small pelagic squaloid sharks found throughout tropical and sub-tropical waters that are known to feed opportunistically on a range of prey including animals much larger than themselves. Short-finned pilot whales are resident to the island of Hawai‘i and are often observed with fresh and healed bites from these sharks. As cookiecutters are pelagic, observing their behaviour directly can be problematic, but studying their bite patterns allows us to infer the dynamics of their predatory behaviour. We used a photo-identification catalogue of 405 resident whales (representing 5,871 identifications from 365 encounters from 2003-2012) to look at prevalence of fresh bites and examine possible seasonal variation in shark behaviour. For each bite we noted: status (fresh, healed or scared), location on body, and mean sea surface temperature. The mean proportion of the body visible for documenting bites was 22.2% (SD=10.0), thus many bites and scars were likely missed. We recorded 9,293 bites of all states, and bites were documented on 396 of 405 whales (97.8%). Fresh bites were recorded from 161 identifications, healing bites were recorded from 490 identifications, and scars were recorded from 2,779 identifications. Bites were most frequently documented on the head (31.7% of all bites), even though this and the dorsal fin had the smallest surface area, followed by the lateral sides (31.1%) and peduncle (26.6%), while the dorsal fin had the lowest percentage of bites (10.9%). The presence of fresh bites varied with day of the year and were negatively correlated with sea surface temperature. Peaks in fresh bite probability were at approximately day 115, 200 and 290 of the year. The peaks in bite probability may be an indicator of fluctuations in other potential prey species or movements of the whales into areas where they may be more likely to be bitten.

Seasonal prevalence of cookiecutter shark (Isistius brasiliensis) bites on short-finned pilot

whales (Globicephala macrorhynchus) in Hawai‘i

Natasha Walker-Milne

Yannis Papastamatiou, Florida International University

Robin Baird, Cascadia Research Collective

Sabre Mahaffy, Cascadia Research Collective

Defining key habitats: an analysis of short-finned pilot whale communities in Hawai‘i using

satellite tag data

Elise Walters, Cascadia Research Collective

Robin Baird, Cascadia Research Collective

Daniel Webster, Cascadia Research Collective

David Anderson, Cascadia Research Collective

6

Sabre Mahaffy, Cascadia Research Collective

Gregory Schorr, Marine Ecology & Telemetry Research

Russel Andrews, University of Alaska Fairbanks and Marine Ecology and Telemetry Research

Abstract:

While only a single stock of short-finned pilot whales is recognized in Hawai‘i, photo-identification, social network analyses, and genetics have all suggested that there are at least two populations; an open-ocean (pelagic) population and an island-associated (insular) population in which there are western, central and eastern communities. Potential anthropogenic impacts in Hawai‘i may vary depending on both population and community. Off Hawai‘i Island, pilot whales are increasingly the focus of whale watching and swim-with tours and there are high levels of fishing activity, whereas off Kaua‘i they are frequently exposed to naval sonar. Establishing the key habitat and hotspots for different communities within this population could help discern which animals are threatened by each of these factors and inform management decisions to reduce impacts. From 2006-2016, 109 LIMPET satellite tags were deployed on pilot whales, with 86 tagged once and 8 tagged multiple times. Kernel density estimations and minimum convex polygons were used to determine the range of individuals and communities. Core areas for the western (5,243 km2) and eastern (5,945 km2) insular communities were similar, but the central community appears to have a smaller (1,378 km2) core area. The core area of the pelagic population (n=8) was estimated at 34,445 km2, although more information is needed to characterize their range. In addition, 14% of the satellite tag records from the western insular community were within the boundaries of the Pacific Missile Range Facility, highlighting the overlap with their home range off Kaua‘i. Our results show that the three insular communities of short-finned pilot whales utilize different habitats within the archipelago, and though the pelagic animals overlap occasionally, their broad range is primarily in open-ocean waters. Based on these findings, each of the insular communities and the pelagic population are subject to different anthropogenic influences and should be managed accordingly.

Diet of Cuvier’s beaked whales (Ziphius cavirostris) in the North Pacific and a comparison

with their diet world-wide

Kristi West, University of Hawaii at Manoa

William Walker, National Marine Mammal Laboratory

Robin Baird, Cascadia Research Collective

Jim Mead

Paul Collins

Abstract:

Cuvier’s beaked whales (Ziphius cavirostris) are distributed world-wide and are the most common cetacean to strand coincident with navy sonars. They are known for their extreme diving capabilities but diet information, fundamental to understanding foraging at depth, is limited from most regions. We report on 11,441 prey items from stomach contents of 16 stranded or by-caught specimens collected from 1976-2016 across the North Pacific. Overall diet included cephalopods, fish and crustaceans, but was dominated by cephalopods. Thirty-seven cephalopod species representing 16 families contributed 98.0% by number and 87.7% by mass. The Gonatidae (26.4% by number; 40.4% by mass), Octopoteuthidae (27.0% by number; 20.2% by mass) and Cranchiidae (27.2% by number; 10.7% by mass) families were dominant. Of these, the species Gonatopsis

borealis represented 36.0% of the prey by mass and 13.2% by number with prey size (dorsal mantle length) ranging between 73 and 346 mm. Taonius borealis contributed 21.3% by number and 8.7% by mass with prey sizes between 107 and 443 mm. The majority of prey items (7,997) were from an adult male stranded in California which contained 20 species from 10 families of cephalopods and fishes. Regional variation was suggested by a higher incidence of crustaceans from whales in the western Pacific and fishes in the eastern Pacific, primarily represented by the giant grenadier. World- wide the most important cephalopod families in the diet of Cuvier’s beaked whales are Cranchiidae, Gonatidae, Histioteuthidae, Octopoteuthidae, Ommastrephidae, Onychoteuthidae, Pholidoteuthidae and Mastigoteuthidae, with Cranchiidae comprising important prey in all locations. While Gonatidae,

7

Octopoteuthidae and Cranchiidae are the dominant prey in the North Pacific, Histioteuthidae and Cranchiidae are most important in the North Atlantic. Knowledge of diet composition can be used to understand how whales utilize their habitat, in calculating nutritional requirements, and may also help to define the locations of important foraging grounds.

Abstract:

Species misidentification is a common problem for some groups of cetaceans, and may lead to inaccurate descriptions of behavior, habitat use, group size, abundance, or acoustic characteristics, among other things. Four tropical species of “blackfish” overlap broadly in their ranges and are often confused at sea owing to their similar gray-black coloration, bulbous head without a beak, and falcate dorsal fin: pygmy killer whales, melon-headed whales, short-finned pilot whales, and false killer whales. Adult pilot whales have a distinct fin size and shape, but misidentification can occur prior to maturity before distinguishable characteristics develop. Some blackfish species can be identified through full-body documentation of unique cape or mouthline pigmentation, but no method currently exists to determine species from nondescript photographs using dorsal fin or body dimensions. This study identified ten two-dimensional, lateral measurements obtainable from typical photos at sea that quantify the distance between reference points along the fin’s contour and the dorsal ridge between the blowhole and fin origin, which can be used to discriminate among species. We measured 285 age-classified, known individuals photographed around the main Hawaiian Islands spanning 31 years of sighting history. Results showed that both mature and immature pilot whales had significantly higher ratios of fin base to dorsal ridge length than other blackfish, and mature animals had significantly greater horizontal distance between reference points than other species. False killer whales had significantly smaller fin base to dorsal ridge length ratios than other blackfish, and significantly shorter fin height to base length than melon-headed and pygmy killer whales. Conversely, melon-headed whales were distinguishable from pygmy killer whales by a significantly wider dorsal fin at multiple reference points. This study established a reliable method for discriminating between nondescript blackfish encountered at sea, which will improve the accuracy of species identification for sightings supported by photographic documentation.

How to tell them apart? Blackfish Species Discrimination using Fin and Body Morphometrics

Obtainable from Photos at Sea

Shelby Yahn

Robin Baird, Cascadia Research Collective

Sabre Mahaffy, Cascadia Research Collective

Daniel Webster, Cascadia Research Collective

8