dusky dolphin foraging habitat: overlap with aquaculture...

Copyright # 2004 John Wiley & Sons, Ltd. Received 20 June 2002Accepted 12 June 2003

AQUATIC CONSERVATION: MARINE AND FRESHWATER ECOSYSTEMS

Aquatic Conserv.: Mar. Freshw. Ecosyst. 14: 133–149 (2004)

Published online 23 January 2004 in Wiley InterScience(www.interscience.wiley.com). DOI: 10.1002/aqc.602

Dusky dolphin foraging habitat: overlap with aquaculturein New Zealand

TIM M. MARKOWITZ*, APRIL D. HARLIN, BERND W .UURSIG and CYNTHIA J. MCFADDENBehavioral Ecology Laboratory, Marine Mammal Research Program, Department of Wildlife and Fisheries Sciences,

Texas A&M University, Galveston, TX, USA

ABSTRACT

1. Marine farms have the potential to affect dolphin foraging in the coastal environment, yet thisissue has been largely omitted from aquaculture management models. Data on the subject areminimal. This study was conducted to examine potential overlap between dusky dolphin habitat useand New Zealand’s growing green-lipped mussel farming industry.2. Data on dusky dolphin occurrence, distribution, abundance, and behaviour were collected

from small vessels over five successive winters in the Marlborough Sounds, the centre of NewZealand’s mussel farming industry.3. Locations and movements of dolphin groups were recorded at 2min intervals with a global

positioning system receiver to examine the overlap of dusky dolphin use of coastal areas withexisting and proposed marine farms. All cases of dolphins entering the boundaries of mussel farmsand total time spent in farms were recorded. Over 8500 dolphin dorsal fin photographs wereanalysed to develop a catalogue of 421 marked individuals utilizing the area. All instances of dolphinfeeding were noted, and focal group behaviour was recorded at 2min intervals for groups observed51 h.4. Within the Marlborough Sounds, dusky dolphins were most often encountered during the

winter in Admiralty Bay, the area with the greatest density of proposed farming activity in theregion. Mark-recapture data indicate that more than 1000 dusky dolphins used Admiralty Bay overthe course of the 5 year study, with an average of 220 individuals inhabiting the bay on any givenweek during the winters of 1998–2002. As many as 55% of individuals returned to Admiralty Bay inconsecutive winters.5. Overlap of dusky dolphin habitat use with proposed marine farms is high, and dolphins rarely

used areas within the existing farms. If dusky dolphin distribution with respect to farms wererandom, an expected 18�0.5 of 436 groups would be encountered in existing Inner Admiralty Bayfarms; however, no dolphin groups were first encountered in farms.6. In 5 years, only eight of 621 dusky dolphin groups monitored in Admiralty Bay were observed

to enter the boundaries of mussel farms at any point. Dolphins entering mussel farms moved rapidlyup the lanes between rows of lines and floats. Dolphins were observed a total of 14.2min inside farmsversus 147.5 h outside of farms in Admiralty Bay. Correcting for area, dolphins were observedspending significantly less time per survey inside than outside of farms.

*Correspondence to: Tim M. Markowitz, Marine Mammal Research Program, 4700 Avenue U, Building 303, Galveston, TX 77551,USA. E-mail: tim [email protected]

7. Most dusky dolphin groups in Admiralty Bay were observed feeding on small schooling fish,often associated with seabirds and/or fur seals. Movement and diving patterns indicate muchforaging when not actively feeding.8. Regular seasonal migration of dusky dolphins and frequent feeding associations with other

apex predators make management of marine farming a wider socio-economic and ecological issue.Copyright # 2004 John Wiley & Sons, Ltd.

KEY WORDS: Lagenorhynchus obscurus; mussel farming; photo-identification; population estimates; cetacean

behavioural ecology

INTRODUCTION

Resource managers are faced with the task of balancing conservation of aquatic biota with a variety ofhuman activities, including recreation, tourism, commercial fishing and aquaculture. The effects of humanactivities on wild dolphin populations vary depending upon the type of activity, its proximity to dolphinhabitat, and the behaviour and distribution of the animals themselves. Freshwater dolphins, such as theYangtze River dolphin (or baiji, Lipotes vexillifer), often live in direct conflict with people for food andhabitat, leading to especially drastic effects of human industries on these species (Zhou and Zhiang, 1991).Coastal dolphin species, such as the New Zealand Hector’s dolphin (Cephalorhynchus hectori), have beenimpacted by near-shore set net fisheries. Incidental capture in fishing nets set within preferred Hector’sdolphin habitat has resulted in a major decline in the species (Dawson, 1991), and a reduction in the geneticdiversity of some populations (Pichler et al., 1998).

Competition for aquatic resources with humans can impact aquatic mammals, likely exerting a greaterinfluence on their populations than either directed hunting or incidental catch due to fishing (Crespo andHall, 2001). In addition to direct or indirect competition for resources, human-made structures, such asthose associated with aquaculture, may compete with marine mammals for space in the coastalenvironment (W .uursig and Gailey, 2002). However, such effects of aquaculture on wild dolphins are rarelyconsidered in models for managing the environmental impacts of marine farming (e.g. Henderson et al.,2001). The goal of this research is to investigate the possible conflicts between dusky dolphins(Lagenorhynchus obscurus) and an expanding marine farming industry in New Zealand. This studyrepresents one of the first efforts to measure the degree of overlap between dolphin habitat and proposedaquaculture developments, and examines the potential ecological consequences of marine farming activitiesfor these apex predators and associated species.

The IUCN lists the dusky dolphin as a species for which currently available data are insufficient toassess conservation status (Whitehead et al., 2000). In South America, dusky dolphin populationshave been impacted by unsustainable practices, including intentional dolphin harvest (Van Waerebeek,1992), incidental catch (Dans et al., 1997), and reduction of their principal prey species (Manzanilla,1988; McKinnon, 1994). Currently, the impacts of human activities on New Zealand dusky dolphinpopulations, including large-scale ecotourism worth millions of dollars to local revenues, appearminimal (Barr and Slooten, 1998; Fairweather and Simmons, 1998; Brown, 1999). This is in part dueto the careful regulation of human–dolphin interactions by New Zealand resource managementagencies.

Occurring mainly in the continental shelf and slope waters along the coasts of Chile, Peru, Argentina,western South Africa and New Zealand, dusky dolphins adopt different foraging strategies in differenthabitats (W .uursig et al., 1989). In Golfo San Jos!ee, Argentina, the dolphins inhabit a shallow environmentand depend primarily on schooling southern anchovy (Engraulis anchoita), which they hunt cooperativelyin small groups during the day (W .uursig and W .uursig, 1980). In Kaikoura, New Zealand, dusky dolphins

T. M. MARKOWITZ ET AL.134

Copyright # 2004 John Wiley & Sons, Ltd. Aquatic Conserv.: Mar. Freshw. Ecosyst. 14: 133–149 (2004)

gather in large groups and feed at night on lantern fish (family Myctophidae) and squid (Nototodarus sp.and Todaroides sp.) associated with the deep scattering layer. Daylight hours are used for resting,socializing and reproduction (W .uursig et al., 1997).

Water temperature can influence both delphinid thermal energetics and prey availability (Wells et al.,1999). Historic sightings suggest a shift in distribution of New Zealand dusky dolphins to the north inwinter and south in summer (Gaskin, 1968), correlated with seasonal changes in water temperature. Awarmer, shallower and relatively sheltered area roughly 200 km north of Kaikoura, the MarlboroughSounds, represents winter habitat for some dusky dolphins.

As dusky dolphins migrate north from Kaikoura to the Marlborough Sounds they encounter differentaquatic environments, with differences in prey type, abundance and distribution. Many of the mesopelagicprey species utilized by dusky dolphins in deep water off Kaikoura are rare in the warmer, shallowMarlborough Sounds; instead, schooling fish such as the New Zealand pilchard (Sardinops neopilchardus)are locally abundant and may be concentrated in certain areas by tides and currents (Baker, 1972).Consequently, dusky dolphins might be expected to alter their hunting tactics and feed during the day,cooperatively hunting schools of fish in a manner similar to the South American dusky dolphinpopulations.

A major port area and popular holiday spot, the Marlborough Sounds is host to a wide array of humanuses. These include recreational boating, fishing, diving, ferry and other large vessel traffic, and extensivemarine farming. Mussel farming is the largest aquaculture industry in New Zealand, and the largestcommercial enterprise in the Marlborough region. In 2000, green-lipped mussels (Perna canaliculus) weregrown in 520 farms around New Zealand, 455 of which are in the Marlborough Sounds (Gall et al., 2000).With increasing economic demand for mussels and the growth of mussel farming, greater information onthe potential impacts of this lucrative industry on marine ecosystems is required to ensure that it remainssustainable (Smaal, 1991).

Shellfish mariculture can affect dolphin habitat use in a number of ways. Establishment of an oyster farmin Shark Bay, Australia, caused mother–calf pairs of Indian Ocean bottlenose dolphins (Tursiops aduncus)to be excluded from the farm area (Mann, 1999a). Objects at the surface and lines in the water column(Figure 1) may impede dolphin movements, impacting the animals’ ranging and foraging patterns (W .uursigand Gailey, 2002). In addition to the potential for physical obstruction by floats and lines, marine farmscould have less direct effects on dolphins. Increased boat traffic due to aquaculture activity may increasenoise levels that could disturb cetaceans, particularly when cetacean habitat use and marine farming areboth concentrated in the same small areas (Richardson et al., 1995). Mollusc farming can causebiodeposition, faunal changes and the introduction of new species or pathogens to marine ecosystems(Buschmann et al., 1996). Changes in the benthic communities beneath mussel farms, includingincreases in aerobic heterotrophic bacteria (La Rosa et al., 2001) and decreases in meiofauna (Mirtoet al., 2000), result from the biochemical effects of accumulated faeces and pseudofaeces (Grant et al.,1995). Although mussel farming appears to influence water column biochemistry less than fin-fish farming,mussels increase nitrogen levels (La Rosa et al., 2002) and deplete chlorophyll a levels (Grange and Cole,1997; Ogilvie et al., 2000) within and around farms. Although these indirect trophic effects could impactdolphin habitat, we do not evaluate them for the present situation in the Marlborough Sounds. Rather, thepurpose of this study is to examine the relationship between dolphin habitat use and aquaculturedevelopment in the area.

Recently, a further increase in green-lipped mussel farm development has been proposed in theMarlborough Sounds. To assess the potential impacts of such management changes on dusky dolphins,it is first necessary to collect basic information regarding dolphin ecology, including which local areas theyinhabit most, how many of them use these areas, and how they utilize the habitat. This study aims todescribe dusky dolphin occurrence and behaviour, measuring the existing and potential overlap betweendusky dolphin habitat use and marine farming, in the Marlborough Sounds.

DUSKY DOLPHIN FORAGING HABITAT 135


Figure 1. Marine farm lines seeded with green-lipped mussels, possibly a physical obstruction to dolphin movements and/or a visual/acoustic barrier, are shown as (a) a photograph, and (b) a 675 kHz sector-scan sonar image. The space between vertical sections of lines

varies, but is generally 0.2–1m.



METHODS

Research effort

During five successive winters from 1998 to 2002, 455 h of boat-based research on 92 days in theMarlborough Sounds produced focal follow data from 632 dusky dolphin group encounters. All work wasconducted from small, 4–6m inflatable vessels with 25–85 hp outboard motors. During the first threewinters, 1998–2000, observations of dusky dolphin groups were conducted in the Marlborough Soundsduring exploratory non-systematic surveys. Areas examined included Queen Charlotte Sound, the outerPelorus Sound, the greater Admiralty Bay area, and Current Basin. Mean search speed (plus/minus thestandard error (SE)) was 17�4.9 kmh�1 ðn ¼ 23Þ. From 1998 to 2000, 119 h of boat-based research on23 days yielded 43 h of focal follow information on 73 dusky dolphin groups.

During winter 2001, the greater Admiralty Bay and outer Pelorus Sound regions were divided into sixzones and systematically surveyed to compare dusky dolphin use of different areas (Figure 2). From July toSeptember 2001, 118 h of boat-based research conducted on 27 days yielded 52 h of focal follow data on 306dusky dolphin groups. Surveys were conducted by two or more trained observers at mean (�SE) speeds of16�0.4 kmh�1 ðn ¼ 27Þ along transect lines pre-programmed into a Garmin 12X global positioning system(GPS) receiver spaced evenly at 1–1.5 km apart and drawn to within 500m of the shore. In this manner, weconducted 17 surveys of Inner Admiralty Bay covering 405 km over 28 h, 12 surveys of Outer AdmiraltyBay covering 503 km over 28 h, eight surveys of Current Basin covering 149 km over 8 h, three surveys of

Figure 2. Line transect routes for six areas surveyed in the Marlborough Sounds during winter 2001. (a) Inner Admiralty Bay;(b) Outer Admiralty Bay; (c) Current Basin; (d) Forsyth Bay; (e) Waitata Reach; (f) Tawhitinui Reach.



Forsyth Bay covering 67 km over 3 h, four surveys of Waitata Reach covering 195 km over 10 h, and threesurveys of Tawhitinui Reach covering 238 km over 11 h. Queen Charlotte Sound was also surveyed for 12 hover 2 days during winter 2001. In winter 2002, the research effort was focused on the area with highestdusky dolphin sighting rates in the previous four winter seasons, the Inner Admiralty Bay. From June toAugust 2002, 218 h of boat-based research conducted in the Inner Admiralty Bay on 42 days yielded 70 h offocal follow data on 253 dusky dolphin groups. Using the same Inner Admiralty Bay survey route as in2001 (Figure 2; area a), surveys were conducted at a mean (�SE) speed of 14�0.4 kmh�1 ðn ¼ 31Þ.

Search effort in summer was limited to 16 h on 4 days in Queen Charlotte Sound (1999–2000) and 24 h on3 days in Admiralty Bay (2000–2001).

Dolphin location and movements

A dolphin group was defined as all individuals within 10m of at least one other individual at a given time,i.e. the ‘10m chain rule’ (Smolker et al., 1992). Group size was defined as the largest number of individualsseen at the surface at any one time, and changes in group size were noted as they occurred (Mann, 1999b).During focal group follows, the vessel was driven parallel to each group, matching the group heading andspeed at such a distance as to minimize disruption of dolphin movements (W .uursig and Jefferson, 1990).Position and time data recorded with a Garmin 12X GPS receiver at 2min intervals were used to estimatethe location and track the movements of each focal dolphin group. GPS Utility v. 3.40.6 software was usedto download tracks, and to calculate the mean speed and average location of each group encountered. Alloccurrences of dolphins entering the boundaries of marine farms were noted, and total time spent in thefarms was recorded with a stopwatch. During 2002, all instances of dolphins coming within 200m of theexisting farms and the total time spent within this proximity were additionally noted.

Positions of dolphin groups, survey track lines, and both existing and proposed marine farms wereplotted and overlaid in ArcGIS, ArcMap v. 8.2, onto a base map by Eagle Technologies (Wellington, NewZealand). For accuracy, the 44 existing farms in Admiralty Bay were mapped by plotting GPS locationstaken on site at the four corners of each farm. Proposed farms were traced by overlaying the MarlboroughDistrict Council June 2002 resource consent chart onto the base map in ArcGIS.

The observed number of dusky dolphin groups encountered within current farms in Admiralty Bay wascompared with the number that would be expected to occur within the boundaries of the farms given arandom distribution. To accomplish this, random points equal to the observed total number of encountersin Admiralty Bay were generated using the Random Point-In-Polygon Generation Program v.2 (VBAmacro developed for ArcGIS by M. Sawada 2002, http://arcscripts.esri.com/details.asp?dbid=12098).These points were plotted in a polygon drawn around the boundaries of the Inner Admiralty Bay inArcGIS, and the number of random points falling in the farms was tallied. This process was replicated 100times and the resulting values compared with observed dusky dolphin group encounters in Admiralty Baymussel farms.

Photo-identification of individuals

Photo-identification of dusky dolphins using dorsal fin scars was conducted in the manner described byW .uursig and Jefferson (1990). During 1998 and 1999, photographs were taken on 100 to 400 ISO slide filmwith a Nikon N90 camera and 80–200mm and 100–300mm lenses, and later digitized to aid in analysis.During 2000–2002, photographs were captured digitally with a Nikon D1 camera using 100–300mm and80–400mm lenses and stored at high resolution on compact flash media. Of 11 828 dorsal fin photographstaken in the Marlborough Sounds during the five successive winters, 8628 (73%) were suitable for analysisfollowing the criteria of Markowitz et al. (2003). Mark rate, or the percentage of individuals with distinctivemarkings, was estimated by taking photographs of all dolphins at random and counting the number ofphotographs with marked versus unmarked individuals.



Photographs of distinctively marked dorsal fins were catalogued according to the number and location ofnotches and scars. Once non-marked and redundant images were removed, a total of 3788 photographicrecords were compared using the Finscan v. 1.5.4 Computer Assisted Dolphin Photo-Identification System,software that employs string and curve-based matching methods to present most likely identificationmatches in order (Araabi et al., 2000; Hillman et al., 2003). All assessments of individual identity wereconfirmed by eye.

Mark-recapture population estimates were calculated using the POPAN module of SOCPROG v. 1.3(developed in MATLAB by H. Whitehead, programs available at http://is.dal.ca/�whitelab/index.htm).‘Closed’ (Schnabel), ‘mortality’, ‘mortality + trend’ and ‘re-immigration’ models were run (Whitehead,1990; Gowans et al., 2000). The Akaike information criterion (Akaike, 1974) was used to determine themodel that best fit the population for each estimate, and residual differences between expected and observednumber of individuals were plotted and examined to ensure that capture probabilities were notheterogeneous (Gowans et al., 2000). Using a 1 week sampling interval, a population estimate wascalculated for the 5 year period (‘re-immigration’ model) and single-season population estimates(‘mortality’ model) were generated using only data from the best three photo-identification samplingseasons (2000–2002). Based on field estimates of group size, 85%, 97% and 93% of dolphins encounteredwere photographed during the winters of 2000, 2001 and 2002 respectively. A second population estimate,using a 1 year sampling interval, was generated for the 2000–2002 dataset (‘mortality + trend’ model).

Behavioural sampling

For each group, feeding was noted if dolphins were seen apparently ‘pursuing fish or holding fish in theirmouths’ (Acevedo-Guti!eerrez and Parker, 2000). Birds and other marine species associated with feedingwere noted. During the 2002 surveys, all instances of birds and other species feeding without dolphins werealso noted. When possible, the identity of the prey species was also recorded. Only those groups tracked forgreater than 10min were included in analyses of feeding behaviour.

Periods of 20min of behavioural observations were recorded at 2min intervals for focal groups followedfor at least 1 h. Most common behavioural state (defined as ‘travel’, ‘mill’, ‘rest’, and ‘feed’; Shane, 1990),the mode inter-individual distance (up to one body length, between one and three body lengths, or morethan three body lengths) and the number of birds associated with the group were recorded by instantaneoussamples (Altmann, 1974), also called ‘point samples’ (Mann, 1999b), at 2min intervals. In addition to thegroup assessment of behavioural state, specific behavioural events related to possible dolphin foraging weredocumented. All occurrences of clean headfirst re-entry leaps, noisy leaps, and acrobatic (somersault) leaps(W .uursig and W .uursig, 1980) were recorded. When synchronous diving occurred, dive times were recordedwith a stopwatch. Simultaneous bursts of speed by group members were recorded during each interval byone–zero sampling (Martin and Bateson, 1993). These data were compared in SPSS v. 11.0 withinformation collected during 1542 h of boat-based research conducted in a similar manner on 227 days inKaikoura from 1997 to 2000.

RESULTS

Dusky dolphin locations

During all years, dusky dolphin groups wintering in the Marlborough Sounds were most commonlyencountered in Admiralty Bay, where the mean group size was five dolphins. During the winters of 1998–2000, 25.5 h of search effort in Admiralty Bay resulted in 36.5 h of focal work with 62 dusky dolphin groups(2.4 group encounters per hour). By comparison, 33 h of search effort in Queen Charlotte Sound yieldedsightings of seven dusky dolphin groups with focal follows of 4.1 h (0.2 group encounters per hour). Surveys



from French Pass through Current Basin to Okiwi Bay over 23.5 h yielded sightings of another four duskydolphin groups with focal follows of 2.7 h (0.2 group encounters per hour).

These findings are further supported by comparison of encounter rates between areas systematicallysurveyed in 2001 (Table 1). Over 99% of all sightings occurred in the Inner and Outer Admiralty Bay areas,where the encounter rate was at least 16 times that of other locations. No dusky dolphins were encounteredin Queen Charlotte Sound and three dolphin groups were encountered in Pelorus Sound during the winter2001 surveys. The Inner Admiralty Bay area had by far the greatest number of sightings, with significantlymore dolphin groups encountered per kilometre of survey effort than the Outer Admiralty Bay (Mann–Whitney, U ¼ 3276:5, P50:001, median inner bay: 0.86 groups per kilometre; outer bay: 0.26 groups perkilometre). These sightings show a high level of overlap with proposed mussel farms in the MarlboroughSounds (Figure 3). From 2001–2002, encounter rates in the Inner Admiralty Bay dropped from 7.5 to 3.3groups per hour and mean inter-group distance increased from 1.9 to 4.4 km, indicating some inter-annualvariation in dolphin use of the bay.

Although dusky dolphins used the Inner Admiralty Bay extensively, they did not utilize the areaswithin the boundaries of existing marine farms along the edges of the bay as much as adjacent areas andother areas proposed for future farm development in the centre of the bay (Figure 4). Out of 621 duskydolphin group encounters in Admiralty Bay over 5 years, none occurred within existing marine farms.When 466 points (equal to the number of dolphin group encounters on survey effort in 2001–2002,Figure 4) were generated at random in the bay (n ¼ 100 replications), an average of 18 (median: 18;mean�SE=18.5�0.47; range: 8–30) occurred within the boundaries of existing farms. Thus, ifdolphin distribution in Admiralty Bay were random, then an average of 18 of 466 encounters (with aminimum of eight) would be expected to occur within farm boundaries; however, none of 466 encounterswas observed within farm boundaries. Although no groups were encountered in farms, eight groupswere observed to enter farms at some point during focal follows, spending a total of 14.2min in farmsout of 157.7 h focal follows in Admiralty Bay (0.15%). Correcting for area (estimated total Inner AdmiraltyBay area outside farms: 28.5 km2; estimated total area inside farms: 0.85 km2), dolphin groupswere observed spending significantly more time per survey day (Wilcoxon signed ranks test Z ¼ 5:777,P50:001, n ¼ 44) outside farms (median: 4.6; mean�SE=5.0�0.35min km�2) than inside farms (median:0; mean�SE=0.1�0.11min km�2). Areas near farms were used more often than areasinside farms (Figure 5(a)). During 69.7 h of dolphin observation in winter 2002, dolphins were observeda total of 8.1min inside farms (0.19%), but were tracked for 5.6 h (8.0%) within 200m of the farms.Correcting for area (estimated total area within 200m of farms: 7.0 km2, estimated total area in farms:0.85 km2), dolphin groups were observed spending significantly more time per survey day (Wilcoxonsigned ranks test Z ¼ 2:934, P ¼ 0:003, n ¼ 28) within 200m of farms (median: 0.7; mean�SE=2.9�1.07min km�2) than inside farms (median: 0; mean�SE=0.2�0.06min km�2). In all cases,dolphins observed to enter the farms travelled rapidly up the lanes from one end of the farm to the other(Figure 5(b)).

Table 1. Dusky dolphin group sightings by location, winter 2001

Location Groups/survey Dolphins/group Groups/h Inter-group distance (km)

Inner Admiralty Bay 12.5 5 7.5 1.9Outer Admiralty Bay 7.5 5 3.3 5.6Current Basin 0 N/A N/A N/AForsyth Bay 0 N/A N/A N/AWaitata Reach 0.25 6 0.1 N/ATawhitinui Reach 0.5 2 0.2 119



No dusky dolphins were encountered during summer surveys in the Marlborough Sounds; however,residents report sporadic dusky dolphin occurrence throughout the year, and we have confirmed thesesightings with photographs.

Photo-identification information on dusky dolphin abundance and residency

The total number of marked individual dusky dolphins photographed in Admiralty Bay over the five winterseasons from 1998 to 2002 was 421, with an overall estimated mark rate of 76% (SE=2.0%). Discoverycurves for the two seasons of systematic surveys (2001 and 2002) demonstrate that > 100 markedindividuals were photographed in Admiralty Bay each year despite inter-annual variation, and that new

Figure 3. The average location of each dusky dolphin group encountered each day ðn ¼ 306Þ during the winter 2001 surveys of sixzones in the greater Admiralty Bay and outer Pelorous Sound regions is indicated by a plus mark with general location indicated.Boxes show proposed mussel-farm developments and extensions of existing farms traced from the Marlborough District Councilmarine resource consent chart, June 2002. Only those proposed farm developments within the survey areas measuring >250m on a

side are included.



individuals continued to be photographed throughout the winter seasons (Figure 6). Using a 1 weeksampling interval ðn ¼ 28Þ, the estimated total population size of dusky dolphins in Admiralty Bay was1013 (SE=186.7) during the five consecutive winters of 1998–2002, with an estimated mean population sizeof 220 (SE=25.9) during any given week. Single-season estimates were 152 (SE=52.1; 95% confidenceinterval (CI): 113–223) in 2000 (n=3weeks); 272 (SE=17.3; 95% CI: 249–300) in 2001 (n=13weeks); and179 (SE=18.0, 95% CI: 164–198) in 2002 (n=9 weeks). Using each season as a sampling unit, the estimatedtotal population size was 1090 (SE=630.9; 95% CI: 693–1291) dusky dolphins inhabiting Admiralty Bayover three successive winters (2000–2002).

As the photo-identification catalogue grew over the course of this 5 year study, it became clear that atleast some dusky dolphins return to Admiralty Bay in successive winters. In winter 2000, 8% of individualsphotographed at Admiralty Bay in 1998 and 1999 were re-identified in the bay. In winter 2001, 15% ofindividuals photographed in previous seasons were photographically recaptured in Admiralty Bay. Bywinter 2002, 55% of marked individuals photographed in Admiralty Bay had been previously identifiedat the same location during one or more of the previous four winters. Preliminary comparisonswith photographic records of dolphins in Kaikoura demonstrate that at least some ðn ¼ 9Þdolphins photographed in Admiralty Bay during the winter were found in Kaikoura during summer.The photo-identification catalogue for Kaikoura currently comprises approximately 3000 markedindividuals. When it is completed and all photographs have been compared between regions, we expectto confirm that many more individuals residing in Kaikoura during the summer migrate to Admiralty Bayfor the winter.

Figure 4. GPS positions of dolphin group encounters in Admiralty Bay during the winters of 2001 (+ marks, n ¼ 213) and 2002(� marks, n ¼ 253), shown in relation to survey route and the position of 44 marine farms presently situated in the bay.



Dusky dolphin use of Admiralty Bay as a winter foraging habitat

Foraging and feeding were common activities of dusky dolphin groups in Admiralty Bay. Daytime feedingactivity was noted in 72% of groups in Admiralty Bay as opposed to 51% of groups in Kaikoura.Behavioural data from 20min focal-group samples demonstrate that dusky dolphin groups in AdmiraltyBay fed during the daytime significantly more than groups in Kaikoura (Kruskal–Wallis, H ¼ 116:8,P50:001). Dusky dolphin groups in Admiralty Bay spent roughly equal amounts of time feeding, resting,travelling and milling, whereas dusky dolphins in Kaikoura groups spent 99% of their time during daylighthours resting, travelling, and milling (Figure 7). Prey species in Admiralty Bay included the New Zealandpilchard (S. neopilchardus), yellow-eyed mullet (Aldrichetta forsteri) and sprat (Sprattus antipodum). Ahigher proportion of groups encountered during surveys of Admiralty Bay were observed engaged infeeding during 2001 (83%) than in 2002 (59%). If the frequency of observed feeding is related to preyavailability, then reduced feeding opportunities could account for the differences in encounter rate andestimated number of dolphins using the bay during the two winters.

Figure 5. (a) A dusky dolphin leaps in front of a green-lipped mussel farm in Admiralty Bay. The farms consist of rows of lines seededwith mussels strung between floats. At present, farms in the bay are limited to 5200m off shore. (b) Dolphins were observed enteringthe boundaries of farms for brief periods eight times during the 5 year study, swimming rapidly up the ‘lanes’ between lines to the other

side of the farm in all cases.



Figure 7. Daytime activity budgets of dusky dolphin groups in Admiralty Bay versus Kaikoura are shown as percent of instantaneoussamples travelling, milling, resting and feeding (mean values with standard error bars). Within Kaikoura, small adult, nursery and

mating groups are shown separately from large groups comprised of hundreds of individuals.

y = 0.0261x3 - 1.5128x2 + 31.445x - 12.049R2 = 0.9957

y = -0.0747x2 + 7.7874x - 6.0232

R2 = 0.9869

0

50

100

150

200

250

0 5 10 15 20 25 30 35 40Survey #

Cum

ulat

ive

# O

f Ide

ntifi

ed M

arke

d D

olph

ins

Winter 2001

Winter 2002

Figure 6. These discovery curves indicate the cumulative number of distinctively marked individual dusky dolphins photographed inAdmiralty Bay by survey day over the course of the winters of 2001 and 2002.



The estimated speed of group movements in Admiralty Bay was generally low, averaging (�SE)4.1�0.14 km h�1 ðn ¼ 303Þ. These slow swimming speeds in the horizontal plane of the water column maybe related to increased foraging and/or diving activity in the vertical plane. Synchronous diving was morecommonly noted in Admiralty Bay groups (79%) than in Kaikoura groups (16%), and was associated withfeeding in 69% of cases observed in Admiralty Bay. The mean (�SE) duration of group dives from the lastindividual diving to first surfacing was 35�1.9 s ðn ¼ 40Þ, and the mean (�SE) dive duration for focalindividuals was 68�10.7 s ðn ¼ 10Þ.

Dolphins hunting in groups can increase prey encounter rate when searching for fish schools byspreading out over a wider area (Bel’kovich et al., 1991). Inter-individual distance was greater in AdmiraltyBay groups ðn ¼ 24Þ than in Kaikoura groups ðn ¼ 115Þ of comparable size (mean Admiralty Bay bodylength: 51=31%, 1–3=43%, >3=26%; mean Kaikoura body length: 51=67%, 1–3=24%, >3=9%).In Admiralty Bay, the mode body lengths between individuals in small groups was significantly less oftenless than one (Mann–Whitney, U ¼ 273, P50:001) significantly more often one to three (Mann–Whitney,U ¼ 359, P ¼ 0:003) and significantly more often more than 3 (Mann–Whitney, U ¼ 297:5, P50:001) thanin Kaikoura.

Despite low mean swimming speeds, synchronous bursts of rapid movement, often observed whendolphins chased fish schools, were more commonly noted (Kruskal–Wallis, H ¼ 17:97, P ¼ 0:001) inAdmiralty Bay groups (median: 10%, mean�SE: 10.8�2.83% of intervals; n ¼ 24) than in Kaikouragroups (small groups median: 0%; mean�SE: 4.1�1.47% of intervals; n ¼ 115; large groups median: 0%mean�SE: 2.4�0.51% of intervals; n ¼ 149). Leaping also varied between regions: dusky dolphinsengaged in a greater proportion of headfirst re-entry leaps, allowing them to catch a breath and descendagain rapidly as might be expected when herding fish, in Admiralty Bay than in Kaikoura (Kruskal–Wallis,H ¼ 72:45, P50:001). Re-entry leaps comprised 97% of documented aerial behaviour for Admiralty Baygroups ðn ¼ 24Þ; this compares with 15% in large groups ðn ¼ 149Þ, 76% in mating groups ðn ¼ 41Þ, 28% innursery groups ðn ¼ 39Þ and 15% in other small groups ðn ¼ 35Þ in Kaikoura.

Dolphins may use diving seabirds to find food (Figure 8), and vice versa, as the number of birds in thevicinity of a pod was positively correlated with the proportion of intervals dolphins engaged in feedingactivity (linear regression: y ¼ 12:619xþ 0:3419, R2 ¼ 0:8854). Birds most frequently observed feeding withdusky dolphins were Australasian gannets (Sula serrator, 87%), shearwaters (Puffinus spp., 26%), andwhite-fronted terns (Sterna striata, 13%). Feeding aggregations also included New Zealand fur seals(Arctocephalus forsteri) during 9% of observations. During the winter 2002 surveys, 59% of bird groups

Figure 8. A dusky dolphin feeds in Admiralty Bay in association with shearwaters.



observed feeding ðn ¼ 241Þ did so in association with feeding dusky dolphins, and 96% of dolphin groupsobserved feeding ðn ¼ 148Þ versus 52% of dolphin groups not feeding were accompanied by birds.

DISCUSSION

Within the Marlborough Sounds, dusky dolphins are concentrated in Admiralty Bay, the same area withthe greatest proposed increase in marine farming. Admiralty Bay is a small area, the inner bay measuringroughly 5 km wide across the mouth by 7 km long, and yet it is preferred winter foraging habitat forhundreds of dusky dolphins. Behavioural data indicate that dusky dolphins spend roughly 25% of the timein Admiralty Bay actively feeding, and much of the remaining time searching for prey, as indicated byrelatively large distances between individuals, slow swimming speeds, and regular diving. Unusualoceanographic features of the area, most notably the adjacent French Pass with currents at times exceeding12 kmh�1, likely act to increase local productivity and concentrate prey in Admiralty Bay.

The great interest in Admiralty Bay for marine farm development may also be due in part to theconcentration of resources, as well as the lack of human traffic in this remote location. Less than 20 peoplelive in the town of French Pass, and mussel farming is worth millions of dollars in annual revenue.Admiralty Bay may represent a ‘Not In My Back Yard’ space for most human residents and recreationalusers in the Marlborough region; however, hundreds of dolphins remain in the bay through the wintermonths hunting in small groups for schooling fishes.

Although dusky dolphins are able to pass through existing marine farms in Admiralty Bay, they do notuse these areas as much as the surrounding habitat. Comparison of dolphin encounter rates in farms withthose expected in a random distribution indicates that dolphins avoid the areas within farm boundaries.Further, dolphins were observed to spend significantly more time per unit area adjacent to mussel farmsand in the middle of Admiralty Bay than in the farms themselves. These findings indicate that the 44existing Inner Admiralty Bay farms, occupying 51 km2 total area, already influence dolphin distributionand habitat use. The proposed 200m extensions of existing farms and the establishment of larger mid-bayfarms would likely further limit dolphin access to winter foraging areas in Admiralty Bay.

It is beyond the scope of this study to model the potential population level impacts of large-scale changesto this habitat on dusky dolphin survival and reproduction. Genetic analyses indicate that the population ofdusky dolphins in New Zealand is a large and healthy one (Harlin et al., 2003). Nevertheless, a winterforaging community comprising at least hundreds, and over the long term perhaps thousands, of dolphinscould be threatened by such dramatic changes to the Admiralty Bay habitat as have been proposed. Suchpotential effects should be carefully considered when reviewing marine farm applications.

To date, there have been few studies of the effects of aquaculture on wild dolphin populations. Ourresults suggest that dusky dolphin winter foraging could be impacted by proposed aquaculturedevelopments in the Marlborough Sounds. Although more difficult to measure than intentional andincidental take, habitat loss and degradation is one of the major threats to wild cetacean populations(Whitehead et al., 2000). Impacts due to changes in dolphin habitat should be carefully monitored andminimized wherever possible.

Objects, such as floats and lines, at the surface and in the water column could impair dusky dolphinforaging by directly impeding dolphin movements or by acting as visual or acoustic obstructions (Figure 1).When foraging cooperatively, dusky dolphins converge on an area where fish have been found, and beginworking together to encircle and herd fish into a ball, often against or near the surface (W .uursig, 1986). Suchhunting activities are hampered by physical obstacles in their preferred foraging habitat (W .uursig andGailey, 2002). Floats and lines could also disrupt echolocation, as objects placed between an echolocatingdolphin and a small target, such as a fish, can interfere with the dolphin’s ability to detect the target (Au,1993).



Introduction of additional mussel farms could also have indirect trophic effects on dolphins and otherapex predators in the marine environment. Since mussels feed on phytoplankton (Gall et al., 2000), it is notsurprising that mussel farms significantly impact phytoplankton levels (Grange and Cole, 1997). InBeatrix Bay, roughly 16 km east of Admiralty Bay, phytoplankton levels within mussel farms were foundto be significantly lower than outside of farms (Ogilvie et al., 2000). Sediment chemistry and benthiccommunity composition are also influenced by mussel farms, mainly due to the accumulation of faecesand pseudofaeces beneath the farms (Mirto et al., 2000). At present, the consequences of microbioticchanges caused by shellfish aquaculture are unknown for predators at higher trophic levels, includingdolphins.

This study shows how management decisions regarding one relatively small area, such as Admiralty Bay,can potentially have far-reaching demographic, socio-economic and ecological consequences. Some, if notall, of the hundreds of dolphins inhabiting Admiralty Bay during the winter months spend their summers inKaikoura, where they rest, reproduce and interact with tourists, supporting a multi-million dollarecotourism industry during the day. Dusky dolphins switch from a nocturnal feeding strategy in the deep-water habitat off Kaikoura to diurnal feeding when they migrate to the shallower Marlborough Sounds.Considering the prevalence of seabirds and fur seals feeding in association with dusky dolphins, impacts ofaquaculture developments on dolphin foraging are likely to influence the ecology of these other apexpredators as well.

Conservation is not only important when species hover on the edge of extinction. Nor are the onlyimpacts of human activities those that result in dramatic scenes of intentional or unintentional over-harvestof apex predators. Management of coastal areas for mariculture and other uses should be informed bystudies on the use of the habitat by marine animals, and the potential impacts of coastal developments onthem. The question is not whether there will be any further development of a lucrative and generallyenvironmentally friendly marine farming industry in the Marlborough region and elsewhere. The realquestion is whether the potential impacts of aquaculture on areas of particular ecological importance tocreatures living in the coastal environment will be factored into decisions regarding the placement andmanagement of marine farms.

ACKNOWLEDGEMENTS

This work would not have been possible without the assistance of many people who kindly contributedfinancial support, time, and effort. Research conducted in the Marlborough Sounds and in Kaikoura during1998–2000 was funded by: the National Geographic Society Committee for Research and Exploration; InterdisciplinaryResearch Initiative grants from the Research Enhancement Program, Office of the Vice President for Researchand Associate Provost for Graduate Studies, Texas A&M University; and the Earthwatch Institute Center forField Research. Research conducted in the Marlborough Sounds during 2001 and 2002 was accomplished thanksto the financial backing of the New Zealand Department of Conservation and the Marlborough DistrictCouncil. We are also grateful to the Department of Conservation, especially A. Baxter, for logistical support,including use of a towing vehicle for launching at French Pass and a second research vessel for surveys of PelorusSound.Local Marlborough residents were extremely helpful and made us feel very welcome. Most especially, we thank

D. and L. Boulton of French Pass Sea Safaris, L. and Z. Battersby of Dolphin Watch Marlborough, I. andP. Montgomery of Okiwi Bay Lodge, and the teachers and students of the French Pass School for their assistance.Research in the Marlborough Sounds was accomplished with the help of many excellent field staff who kindlyvolunteered their time, including R. Honeycutt, S. Cheeks, R. Constantine, S. Carlyle, B. White, H. Markowitz,J. Markowitz, M. W .uursig, L. Boren, P. Irvine, B. O’Leary, P. Brakes and K. Young. Work in Kaikoura wasaccomplished with the help of many hard working Earthwatch volunteers and full-time field assistants, including C.Littnan, S. Stanley, K. Mazzarella, A. Acevedo, L. McOmber, C. Owens, N. Brown, K. Andrews, and H. Peterson. Weare grateful to Kaikoura residents and businesses for their help, including the staff of Dolphin Encounter, Whale WatchKaikoura, J. Van Berkel and the rest of the good folks at the University of Canterbury Edward Percival Field Station,



J. Hocking at Better Boats, M. Green and S. Howie at Kaikoura Marine, R. Peterson at West End Motors, andS. Johnson.Analysis was accomplished with the volunteer help of Marine Mammal Research Program interns, especially

B. LeBoeuf, K. Manley and H. Williams, and students from the departments of Marine Biology and Wildlife andFisheries Sciences, Texas A&M University. S. Gowans provided invaluable assistance and advice in selecting andcalculating mark-recapture population estimates. We also thank the many faculty, staff, and colleagues at Texas A&MUniversity who contributed helpful ideas and encouragement. T. Brock, W.J. Schrader and two anonymous reviewersprovided excellent and insightful editorial comments and suggestions.

REFERENCES

Acevedo-Guti!eerrez A, Parker N. 2000. Surface behavior of bottlenose dolphins is related to spatial arrangement ofprey. Marine Mammal Science 16: 287–298.

Akaike H. 1974. A new look at the statistical model identification. Institute of Electrical Engineers Transactions onAutomatic Control 19: 716–723.

Araabi BN, Kehtarnavaz N, McKinney T, Hillman G, W .uursig B. 2000. A string matching computer-assisted system fordolphin photoidentification. Annals of Biomedical Engineering 28: 1269–1279.

Altmann J. 1974. Observational study of behavior: sampling methods. Behaviour 49: 227–267.Au WWL. 1993. The Sonar of Dolphins. Springer-Verlag: New York; 277pp.Baker A. 1972. Reproduction, early life history, and age–growth relationships of the New Zealand pilchard, Sardinopsneopilchardus (Steindachner). Fisheries Research Bulletin No. 5. Fisheries Research Division, New Zealand MarineDepartment; 64pp.

Barr K, Slooten L. 1998. Effects of tourism on dusky dolphins at Kaikoura. In Proceedings of the International WhalingCommission, SC/50/WW10; 1–30.

Bel’kovich VM, Ivanova EE, Yefremenkova OV, Kozarovitsky LB, Kharitonov SP. 1991. Searching and huntingbehavior of the bottlenose dolphin (Tursiops truncatus) in the Black Sea. In Dolphin Societies: Discoveries and Puzzles,Pryor K, Norris KS (eds). University of California Press: Berkeley, CA; 38–67.

Brown NC. 1999. Dusky dolphin (Lagenorhynchus obscurus) habitat use, group size and dispersion in Kaikoura, NewZealand. MSc thesis, Ecology, Evolution, and Biostatistics, Auckland University.

Buschmann AH, Lopez DA, Medina A. 1996. A review of the environmental effects and alternative productionstrategies of marine aquaculture in Chile. Aquacultural Engineering 15(6): 397–421.

Crespo EA, Hall M. 2001. Interactions between aquatic mammals and humans in the context of ecosystemmanagement. In Marine Mammals: Biology and Conservation, Evans PGH, Raga JA (eds). Kluwer Academic/Plenum: New York, 463–490.

Dans SL, Crespo EA, Garcia NA, Reyes LA, Pedraza SN, Alonso MK. 1997. Incidental mortality of Patagonian duskydolphins in mid-water trawling: retrospective effects from the early 1980s. Reports of the International WhalingCommission 47: 699–703.

Dawson S. 1991. Incidental catch of Hector’s dolphins in inshore gill-nets. Marine Mammal Science 7(3): 283–295.Fairweather JR, Simmons DG. 1998. Report #8: the economic impact of tourism on Kaikoura. Lincoln UniversityTourism and Education Centre (TREC), Canterbury, New Zealand.

Gall M, Ross A, Zeldis J, Davis J. 2000. Phytoplankton in Pelorus Sound: food for mussels. Water and Atmosphere8(3): 1–16.

Gaskin DE. 1968. Distribution of Delphinidae (Cetacea) in relation to sea surface termperatures off eastern andsouthern New Zealand. New Zealand Journal of Marine and Freshwater Research 2: 527–534.

Gowans S, Whitehead H, Arch JK, Hooker SK. 2000. Population size and residency patterns of northern bottlenosewhales (Hyperoodon ampullatus) using the Gully, Nova Scotia. Journal of Cetacean Research and Management 2: 201–210.

Grange K, Cole R. 1997. Mussel farming impacts. Aquaculture Update 17: 1–3.Grant J, Hatcher A, Scott DB, Pocklington P, Schafer CT, Winters GV. 1995. A multidisciplinary approach toevaluating impacts of shellfish aquaculture on benthic communties. Estuaries 18: 124–144.

Harlin AD, Markowitz TM, Baker CS, W .uursig B, Honeycutt RL. 2003. Genetic structure, diversity, and historicaldemography of New Zealand’s dusky dolphin (Lagenorhynchus obscurus). Journal of Mammalogy 84(2): 702–717.

Henderson A, Gamito S, Karakassis I, Pederson P, Smaal A. 2001. Use of hydrodynamic and benthic models formanaging environmental impacts of marine aquaculture. Journal of Applied Ichthyology 17(4): 163–172.



Hillman GR, W .uursig B, Gailey GA, Kehtarnavaz N, Drobyshevsky A, Araabi BN, Tagare HD, Weller DW. 2003.Computer-assisted photo-identification of individual marine vertebrates: a multi-species system. Aquatic Mammals29: 117–123.

La Rosa T, Mirto S, Marino A, Alonzo V, Maugeri TL, Mazzola A. 2001. Heterotrophic bacteria community andpollution indicators of mussel farm impact in the Gulf of Gaeta (Tyrrhenian Sea).Marine Environmental Research 52:301–321.

La Rosa T, Mirto S, Favaloro E, Savona B, Sara G, Danovaro R, Mazzola A. 2002. Impact on the water columnbiogeochemistry of a Mediterranean mussel and fish farm. Water Research 36(3): 713–721.

Mann J. 1999a. Recent changes in female dolphin ranging in Red Cliff Bay, off Monkey Mia, Shark Bay. Report toWest Australian Department of Fisheries and West Australian Department of Conservation and Land Management,Perth, Australia.

Mann J. 1999b. Behavioral sampling methods for cetaceans: a review and critique. Marine Mammal Science 15: 102–122.

Manzanilla SR. 1988. The 1982–1983 El Ni *nno event recorded in the dentinal growth layers in teeth of Peruvian duskydolphins (Lagenorhynchus obscurus). Canadian Journal of Zoology 67: 2120–2125.

Markowitz TM, Harlin AD, W .uursig B. 2003. Digital photography improves efficiency of individual dolphinidentification. Marine Mammal Science 19: 217–223.

Martin P, Bateson P. 1993. Measuring Behaviour, 2nd edn. Cambridge University Press: Cambridge.McKinnon J. 1994. Feeding habits of the dusky dolphin, Lagenorhynchus obscurus, in the coastal waters of central Peru.Fishery Bulletin 92: 569–578.

Mirto S, La Rosa T, Danovaro R, Mazzola A. 2000. Microbial and meiofaunal response to intensive mussel-farmbiodeposition in coastal sediments of the western Mediterranean. Marine Pollution Bulletin 40(3): 244–252.

Ogilvie S, Ross AH, Schiel DR. 2000. Phytoplankton biomass associated with mussel farms in Beatrix Bay, NewZealand. Aquaculture 181(1–2): 71–80.

Pichler FB, Dawson SM, Slooten E, Baker CS. 1998. Geographic isolation of Hector’s dolphin populations describedby mitochondrial DNA sequences. Conservation Biology 12: 666–682.

Richarson WJ, Greene Jr CR, Malme CI, Thompson DH. 1995.Marine Mammals and Noise, Academic Press: London;576pp.

Shane SH. 1990. Behavior and ecology of the bottlenose dolphin at Sanibel Island, Florida. In The Bottlenose Dolphin,Leatherwood S, Reeves RR (eds). Academic Press: San Diego, CA; 245–265.

Smaal AC. 1991. The ecology and cultivation of mussels: new advances. Aquaculture 94: 245–261.Smolker RA, Richards AF, Connor RC, Pepper JW. 1992. Sex differences in patterns of association among IndianOcean bottlenose dolphins. Behaviour 123: 38–69.

Van Waerebeek K. 1992. Population identity and general biology of the dusky dolphin Lagenorhynchus obscurus (Gray1828) in the southeastern Pacific. PhD thesis, University of Amsterdam.

Wells RS, Boness DJ, Rathbun GB. 1999. Behavior. In Biology of Marine Mammals, Reynolds III JE, Rommel SA(eds). Smithsonian: Washington, DC, 324–422.

Whitehead H. 1990. Mark-recapture estimates with emigration and re-immigration. Biometrics 46: 473–479.Whitehead H, Reeves RR, Tyack PL. 2000. Science and the conservation, protection and management of wildcetaceans. In Cetacean Societies, Mann J, Connor RC, Tyack PL, Whitehead H (eds). University of Chicago Press:Chicago; 308–332.

W .uursig B. 1986. Delphinid foraging strategies. In Dolphin Cognition and Behavior: A Comparative Approach,Schusterman RJ, Thomas JA, Wood FG (eds). Lawrence Erlbaum Associates, Hillsdale, N.J. 347–359.

W .uursig B, Gailey GA. 2002. Marine mammals and aquaculture: conflicts and potential resolutions. In ResponsibleMarine Agriculture, Stickney RR, McVey JP (eds). CAB International Press: New York; 45–59.

W .uursig B, Jefferson T. 1990. Methods of photo-identification for small cetaceans. Reports of the International WhalingCommission 12: 43–52.

W .uursig B, W .uursig M. 1980. Behavior and ecology of the dusky dolphin, Lagenorhynchus obscurus, in the SouthAtlantic. Fisheries Bulletin 77: 871–890.

W .uursig B, W .uursig M, Cirpriano F. 1989. Dolphins in different worlds. Oceanus 32: 71–75.W .uursig B, Cipriano F, Slooten E, Constantine R, Barr K, Yin S. 1997. Dusky dolphins (Lagenorhynchus obscurus) ofNew Zealand: status of present knowledge. Reports of the International Whaling Commission 47: 715–722.

Zhou K, Zhiang X. 1991. Baiji: The Yangtze River Dolphin and Other Endangered Animals of China. Stonewall Press:132pp.



dusky dolphin foraging habitat: overlap with aquaculture...

Documents