|
|
|
|
|
|
|
|
Humpback dolphins Sousa chinensis are rare in South Africa; a 1994
census estimated 160 individuals along the KwaZulu-Natal coast. Every year about eight of these dolphins die of suffocation in the shark nets set at popular KwaZulu-Natal beaches. According to the IWC, this represents an unsustainable rate. The Richards Bay Humpback Dolphin project, sponsored by RBM, EWT and NSB, consists of a photo-identification study and a behavioural study. The aims are to examine population dynamics and to test new devices – Pingers - which have been designed to warn the dolphins about the nets.HUMPBACK DOLPHINS SHOWED NO BEHAVIOURAL REACTIONS TO PINGERS IN THE RICHARDS BAY SHARK NETS.
INTRODUCTION – What did we aim to do and why?
Humpback dolphins are caught and killed in shark nets (fishing nets) in KwaZulu-Natal, South Africa at a rate that is probably not sustainable. The Richards Bay shark nets are responsible for more than half of the humpback dolphin deaths, even though this is only one of 40 shark net installations. This makes Richards Bay the perfect place to test if pingers (acoustic warning devices) will keep humpback dolphins out of shark nets. It will take a few years to see if the pingers affect the mortality rate and we decided to investigate the behavioural reactions of the dolphins to the pingers in the meanwhile.
We figured (hypothesised) that if the pingers are effective at reducing the number of humpback dolphins killed in the shark nets at Richards Bay, then we will see a difference in the dolphins’ behaviour.
We tested four specific predictions of hoe the pingers would affect the dolphins:
METHODS – How did we do it?
All experiments need controls for proper comparisons. As our control, we used dummy pingers that looked the same as the proper pingers but didn’t make any sounds. Once a week, the Natal Sharks Board field staff would flip a coin to see whether pingers or controls would be placed in the nets. This was so that so that my expectations did not influence the results. From the end of the North Pier, i monitored the dolphins’ behaviour. If a search was successful (i.e. if dolphins visited the netted area that can be seen from the pier), I noted their time of arrival and depature, their geographic position and their behaviour. I recorded very basic behavioural states: feeding, resting, socialising, travelling and undetermined. Geographic position (where a dolphin surfaced) was recorded at 5-minute intervals. A minimum potential zone of response with a radius of 100m was delimited around the nets and the data of surfacing position within this area was extracted for comparison between pingers and control.
RESULTS – What did we find?
DISCUSSION – What does this mean?
The dolphins did not visit the netted area less often nor did they spend less time there when the pingers were on. The dolphins did not avoid the netted area; they were not excluded from this relatively large area by the pingers. This is can be considered a positive result since concern has been expressed that pingers might exclude dolphins from large areas that are critical to their survival.
The fact that the dolphins were not excluded from a 100m radius from the pingers is a little more worrying, especially since the pingers are spaced at 100m intervals. However, it is possible that the pingers may affect the dolphins’ behaviour over a smaller distance. On the other hand, maybe it was a bad assumption that pingers work by excluding dolphins from a particular area. It doesn’t seem to be a bad assumption, however, since all the other studies of dolphins/porpoises around pingers found that the animals avoided an area immediately around a pinger (ranging from 125m to 640m). The distance that a sound can travel (its active space) is determined by four things: 1) how loud it is; 2) what "things" are in the way that could absorb or scatter it (e.g. air bubbles); 3) background noise; and 4) how good the animal’s hearing is. These four factors that affect the active space of the pinger signal could affect the size of the area that the dolphins avoid. We don’t know over what distance the pingers affect the dolphins. We don’t even know the size of the "danger zone" around the shark nets. All we can say is that the pingers did not cause an avoidance reaction at a distance of 100m.
Lastly, the activity patterns of the dolphins were not changed by the pingers. Whether the pingers were on or not, the dolphins spent most of their time feeding in the area, though occasionally they were observed socialising, travelling. Only once were they seen resting. Even the proportion of time that we couldn’t tell what they were doing (undetermined) was no different.
In conclusion, none of the predictions were supported. The results of this experiment failed to support the hypothesis that the behaviour of the humpback dolphins would be changed when pingers were actively emitting a signal. If they did affect the humpback dolphins in any way, it was at a scale different to those investigated in this study. Further studies are recommended.The behaviour of humpback dolphins Sousa chinensis at the Richards Bay shark nets.
Shanan Atkins, Dr Vic Peddemors and Dr Neville Pillay.
Photo-identification of humpback dolphins in Richards Bay
Mark Keith, Dr Vic Peddemors and Dr Bester.
INTRODUCTION
The Indo-Pacific humpback dolphin Sousa chinensis is an uncommon resident of coastal waters in the Indian and Pacific oceans. In KwaZulu-Natal, the humpback dolphins die in shark nets which are set at popular bathing beaches to catch and kill sharks. On average, the Natal Sharks Board records eight humpback dolphin deaths in these nets each year. The KwaZulu-Natal population size has been estimated at about 165 animals (Durham 1994). This amounts to a population decrease of 4.8% per year. This is probably not a sustainable rate (International Whaling Commission 1994).
Of the 102 recorded humpback dolphin deaths in the nets, 58% occurred in the Richards Bay shark nets. These nets represent less than 5% of the netted area in KwaZulu-Natal. In order to prevent dolphin capture in the shark nets, its causes need to be understood; a study of dolphin behaviour at the shark nets may provide insights.
THE HYPOTHESIS
The behaviour of humpback dolphins at the Richards Bay shark nets is different to their behaviour elsewhere in the Richards Bay area.
METHODS
Searches were conducted…
When dolphins were found…
Behaviour was…
Table 1. Descriptions of the behavioural categories
| Behaviour | Description |
| Travel | Persistent, directional movement. Regular pattern of surfacing and diving. Shallow dive angles. |
| Feed | Localised movement. Extended submersion times (two minutes or more). Dive direction is irregular, varying within and between individuals. Steep dive angles. Some aerial behaviour. |
| Social | Localised movement. Dive direction is unpredictable. Interactions between individuals. Frequent aerial behaviour (jumps, somersaults, lobtailing, spyhopping). |
| Rest | Movement may appear directional but often the animals are circling over the same area. Levels of activity are low. Dive angles are shallow. |
| Undetermined | Behaviour does not fall into any of the above categories. |
RESULTS
Figure 1. The mean proportion of the frequency of each behaviour at the shark nets and elsewhere in the study area
Figure 2. The mean proportion of the duration of each behaviour at the shark nets and elsewhere in the study area
Table 2. Search effort.
| Search frequency | 40 |
|
| # successful searches | 23 |
|
| # observations at the nets | 9 |
|
| Total search time (hours)* | 137 |
|
| Total observation time ** | 47 |
|
| Observation time at the nets | 2.5 |
|
| Average search duration (+ sd) | 3.4 |
1.95 |
| Average observation duration (+ sd) | 2.0 |
1.6 |
| Average duration at nets (+ sd) | 0.3 |
0.4 |
* Includes total observation time
** Includes observation time at the nets
+ SD
DISCUSSION
The behavioural proportions observed "elsewhere" in the study area is very similar to that recorded by Durham (1994) around the northern Tugela Bank, an area that includes Richards Bay, implying that this may be considered "typical" humpback dolphin behaviour. In comparison, behaviour near the nets was not typical and the full behavioural repertoire of this species was not observed near the shark nets.
These data support the hypothesis that the behaviour of humpback dolphins at the shark nets is different to their behaviour elsewhere in the study area. Not only was the full behavioural repertoire not observed near the nets, but there was a significant difference in the proportion of the predominant behaviour (feeding) at and away from the nets. Thus it is possible that the shark nets are situated in a humpback dolphin feeding area. Feeding behaviour may increase the likelihood of capture as the dolphin’s attention is focussed on prey and not the surrounding environment (Goodson et al. 1994). This means that a feeding dolphin is unlikely to detect the nets until it is too late to avoid collision.
Social activity was never observed near the nets. This result is similar to that of a similar study of bottlenose dolphin Tursiops truncatus behaviour at the shark nets (Peddemors 1995). It seems possible that social behaviour does not normally occur in feeding areas. More analysis is required to ascertain if feeding and social behaviour ever occur in the same area.
CONCLUSION
Humpback dolphin behaviour at the Richards Bay shark nets is different to their behaviour elsewhere in the area. It is possible that the nets are placed in a feeding area. The probability of capture may be increased as attention is focussed on prey and not the environment.
Alternatively, it is possible that although feeding behaviour is predominant in this area, capture may occur during rare bouts of other types of behaviour.
Please follow this link for the detailed project proposal.
ACKNOWLEDGEMENTS
Thank you to our sponsors: Richards Bay Minerals and Endangered Wildlife Trust and to the Natal Sharks Board and the NSRI for help with our boat and equipment.
REFERENCES
Durham B. 1994. The distribution of the humpback dolphin (Sousa chinensis) along the Natal Coast, South Africa. Unpublished Masters of Science, University of Natal.
International Whaling Commission. 1994. Report of the Workshop on Mortality of cetaceans in passive fishing nets & traps. La Jolla, California. 22-25 October 1990.
Goodson AD, Klinowska M and Bloom PRS. Enhancing the acoustic detectability of gillnets. Rep. Int. Whal. Commn. (Special Issue 15): 585-595.
Peddemors, V.M. 1995. Aetiology of bottlenose dolphin capture in shark nets off Natal, South Africa. Unpublished Doctor of Philosophy, Faculty of Science, Department of Zoology. University of Port Elizabeth.
Photo-identification of humpback dolphins in Richards Bay
The use of photographic techniques to identify individuals from their natural markings has been well-established for a number of cetacean species. Through examination of photographs, Bigg (1982) found that individual killer whales could be identified from distinct pigmentation patterns and scars on the saddle patch. Photo-identification has the potential to document the lives of individuals in great detail (Bigg, Ellis and Balcomb 1986). This was found to be a highly useful method in the identification of killer whales, since it is a passive method thus not disturbing the whales, as active handling would otherwise be necessary. Furthermore their dorsal fins possess distinctive nicks, gouges, and occasionally some major tissue loss. Similar characteristics are found with humpback dolphins , Sousa chinensis.
Establishing the characteristics of the population
Photographs are taken of the humpback dolphins' dorsal fin and hump, when surfacing to breathe, as perpendicular to the body axis as possible. Appropriate annotation of the photo sequence are noted. Würsig & Jefferson (1990) suggests that rapid annotation can be achieved by taking a picture of a non-dolphin subject subsequent to an important sequence, and to note the event into a field book. Animals are re-photographed each time they change their associations during a follow to allow analysis of group dynamics and sociality patterns. With each dolphin school sighting, proceeding photographing, the film spool number, starting frame are recorded. When all individuals within a group are considered photographed, the number of the end frame are recorded to keep track of individuals and group associations after photographic development.
Photo identification Tracings of the photographed fins are made from the colour slide, for individual identification purposes, after projection and enlargement to a standard size using a slide projector. The photographs are given a quality value (zero being poor quality, with five being excellent), based on image size, focus, light and angle of the dorsal fin as well as exposure of photograph (following Peddemors, 1995). Only photographs in clear focus, and containing an un-obscured dorsal fin and hump, are traced and used for identification purposes. A well-marked individual is one that can be recognised by a matrix of marks which in human-related terms, "form a distinctive face" for the individual (Würsig & Jefferson 1990). When one or two simple identification features are used, one may accidentally identify similar looking animals, as the same individual. Therefore it is essential to use a combination of marks, scratches, nicks and tears in the dorsal fin for a more reliable identikit. Longevity and the variability of marks are of critical importance for compiling the identikit. There seems to be no hard and fast rule on how long marks on dolphins last, but some studies have shown that the dorsal fin markings do last for a considerable time (some up to 7 years; Bigg, 1982), with scratch marks on the body not lasting as long.
The dorsal fin tracings are sorted according to Karczmarski & Cockcroft (in press):
1. Tracings are grouped according to the general shape of the dorsal fin: Normal shape (sickle shaped) (0), or irregular (X). 2. The number and nature of the notches on the body will further separate the tracings: 1. On the leading edge of the fin, 2. On the tip of the dorsal fin, 3. On the trailing edge of the fin and 4. Nicks and notches anterior to the dorsal fin, 5. Notches and nicks posterior to the dorsal fin. 3. The shape of the most prominent notch are used to further categorise tracings. Notches are grouped in "U", "V", "square", "irregularly ragged" or "other" in shape. 4. Scarring on the body are divided into three categories: a) Anterior of the fin, bi) on the dorsal fin, bii) below the dorsal fin, c) posterior of the dorsal fin. 5. The tracings are further divided using any prominent colour variations on the hump or surrounding areas.
A dorsal fin ratio (DFR) for each tracing, as described in Defran, Schultz & Weller (1990) are recorded. The dorsal fin ratio is usually obtained by measuring the distances between the two most prominent nicks or notches visible on the dorsal fin of the particular animal. The distance from the tip of the dorsal to the lowest nick is used to divide the distance between the two prominent notches (see Defran, Schultz & Defran, 1990). Karczmarski & Cockcroft (in press) used a modified DFR as one identification measure for their study in Algoa Bay, South Africa, on humpback dolphins. This modification was executed to include individuals with only one prominent and useable nick/notch. The distance between the top of the fin to the bottom of the notch was measured, and divided by the distance from the top of the fin to the top of the notch. They found the DFR application for humpback dolphins to be limited. Generally all visible marks on an individual were used for identification purposes.
All individuals are catalogued with the data relevant to each sighting to create a photo-identikit. Photographs taken during previous years (Durham, 1994; Sharks Board collection) are included in the catalogue to determine any long-term residency patterns.
Site fidelity
Active acoustic deterrent devices (pingers) are introduced, into the Richards Bay shark net installation, to try and reduce the rate of captures of humpback dolphins in the shark nets. Captured individuals, appear to be foreign individuals "immigrating" into the area (Durham, 1994). An index of residency (Karczmarski & Cockcroft, 1997) are calculated for all identified dolphins, using the identikit and re-sightings, occurring within the area. Residence Index (RI) = S x M/100 Where: RI - residency index S - total number of sightings of an individual, and M - total number of months in which this particular individual was seen.
This is used as an index of identified dolphins' site fidelity to certain areas. The study area is divided into one kilometre strips or zones. With every sighting, the location is noted and the photo-identikit and the photographs for the particular animals are used to plot the animals "affinity" (amount of time spent) for certain areas. Once a dolphin's index of residency (site fidelity) has been established the term "resident" or "foreign" is assigned to the animal in question according to the index obtained. The amount of time the dolphin spends around the nets and which particular animal, either a "resident" or "foreign" individual are made available to the behavioural study.
Home range and movements patterns
Using the site fidelity the home ranges for all identified re-occurring individuals are plotted on a chart of the area. Using both site fidelity and home ranges, the amount of time the dolphin spend around the shark nets in proportion to other areas are used to compare the same figures after pinger installation. This will give insight as to whether the pingers effect the home ranges, movement patterns or site fidelity of the humpback dolphins. Movement patterns are investigated using the amount of resightings and period of time between consecutive sightings. Photographs are taken, of any humpback dolphins caught in the KZN shark nets during the study period. This will allow identification of any possible long distance movements, while also allowing insight into whether the resident or foreign dolphins are being caught in the Richards Bay shark net installation.
Nature of the population
To estimate the population density for humpback dolphins in the Richards Bay area, photo-identification mark-recapture methods are used. A capture (sighting) are considered successful when a recognisable dolphin is photographed. A re-sighting will only occur if a previously recognised and categorised dolphin is re-photographed on another survey. The number of newly identified animals, distinguished over time, are used to plot a "rate of discovery curve". The shape of the "rate of discovery" curve, and sighting frequencies of the individual humpback dolphins are used to determine the nature of the population ("opened" or "closed"). The curve for a closed population will reach an asymptote, when the rate of newly identified animals decrease, since most individuals have been identified, with no new animals being born or immigrating into the population. When the rate of discovery and sighting frequencies do not decrease with time, it can be assumed that the population is open. The population thus has a constant inflow of individuals, either through births or immigration, and losses due to death and emigration.
Estimated size of the population
Once the nature of the population has been established an estimate of the population can be determined. With capture-recapture experiments each population model has assumptions and limitations. Karczmarski & Cockcroft (1997) stresses that all population estimates used in their study were estimates of marked dolphins (usually adults). Estimated group size are used to determine the proportion of individuals identifiable. This is due to the fact that mainly adults are identifiable when using photo-identification methods (juveniles and calves do not generally have markings (Durham, 1994a)). Therefore the estimate for the total number of dolphins (i.e. both identifiable and non-identifiable) in the group or population (P) can be determined, following Karczmarski & Cockcroft (1997), from the number of photographically identifiable individuals using the formula: P = (X/Y)/Z Where X - estimate of the number of adults, Y - proportion of identifiable individuals, Z - mean proportion of adults in a group of humpback dolphins.
Catch trends
Historic catch data of humpback dolphins in Richards Bay, are investigated and analysed for any temporal pattern involving capture rates of dolphins. Since 1980, 68 humpback dolphins have been caught in the Richards Bay shark nets (NSB capture data). Durham (1994a) found a decline in humpback dolphin sightings, within the above mentioned area, with 74% of all humpback dolphin captures occurring in the four northern most installations of the KZN shark nets. Once the specific nature of the Richards Bay population has been assessed, the true impact of the high mortalities on the population, due to the shark nets, become apparent. The acquired information will hopefully enable more accurate management and conservation strategies.
Association patterns Among delphinids, social organisation is best known in killer whales (Orcinus orca) and bottlenose dolphins. Group structure and associations range from stable in the killer whale to, relatively fluid in most other dolphins (Slooten, Dawson & Whitehead, 1993). Humpback dolphins in the Eastern Cape (South Africa), appear to exhibit a highly fluid social structure with casual and short-term affiliations (Karczmarski & Cockcroft, 1997). According to Saayman & Tayler (1979), humpback dolphins near Plettenberg Bay travelled and interacted with different companions in groups of unstable and variable composition. Differences in dispersal, and association patterns between specific age and sex classes within groups, at different times of the year, have also been observed. Lactating females and calves were inclined to be "more" resident in an area, whereas other dolphins and adult males seemed to be more mobile (Durham, 1994a). Levels of association for humpback dolphins are investigated using photographed associations between identified individuals. When using a 200mm lens, at f-stop of approximately 8 or 5.6, Ballance (1990) found, animals separated as little as 2 m would not be in focus. Therefore all photographed animals that are in focus in a frame, are regarded "in close" association (closer than 2 m). These surfacing animals are identified and possibly be used, in analysis of surfacing associations.
Slooten et al. (1993), used cluster analysis (half weight association index), and temporal analysis to determine re-association rates for Hector's dolphins (Cephalorhynchus hectori). The above mentioned authors, used only associations between obvious and clearly identifiable dolphins, being seen for more than two days throughout the study. Karczmarski & Cockcroft (1997) quantified association patterns of individuals using a "Simple Ratio Association Index" or an "Association Index (AI)" AI = J/(A + B) - J Where: AI - Association Index, J - number of joint sightings of individuals A and B, A - total number of sightings of individual A, and B - total number of sightings of individual B. This technique appears more accurate and is used in this study.
Environmental factors influencing population dynamics Collected environmental data are used to correlate individual and group preferences for the duration of the study. This data as well as the other environmental data (tides and sea state) are correlated to the sightings of humpback dolphins within the study area. The individuals identified correlated with the environmental data will hopefully provide better insight into the habitat requirements and preferences of these dolphins within the Tugela Bank region. These information is being compared with other studies and their findings (Durham, 1994a and Karczmarski, 1996).
References
Ballance, L.T. 1990. Residence patterns, group organisation, and surfacing associations of bottlenose dolphins in Kino Bay, Gulf of California, Mexico. In: The Bottlenose Dolphin. Leatherwood, S. & R.R. Reeves (eds), Academic Press, Inc, California.
Barros, N.B. and Cockroft, V.G. 1991. Prey of humpback dolphins (Sousa plumbea) stranded in eastern Cape province, South Africa. Aquatic Mammals 17: 143-136.
Bigg, M.A. 1982. An assessment of killer whale stocks off Vancouver Island, British Columbia. Rep. Int. Whal. Commn. 32: 655-666. Bigg, M.A., Ellis, G.M. & Balcomb, K.C. 1986. The graphic identification of individual cetaceans. Whalewatcher 20: 10-12. Bigg, M.A., Ellis, G.M., Ford, J.K.B. & Balcomb, K.C. 1987. Killer whales: a study of their identification, genealogy and natural history in British Columbia and Washington State. Phantom Press & Publishers Inc., Canada. Bigg, M., Macaskie, I. & Ellis, G. 1983. Photo-identification of individual killer whales. Whalewatcher 17: 3-5.
Cockcroft, V.G. 1989. Biology of the indopacific humpback dolphin (Sousa chinensis) off Natal, South Africa. Paper presented at the Eight Biennial Conference on the Biology of Marine Mammals, Pacific Grove, California, December 1989 (unpublished).
Cockcroft, V.G. 1990. Dolphin catches in the Natal shark nets, 1980 to 1988. S. Afr. J Wildl. Res. 20: 44-51.
Cockcroft, V.G. 1994. Is there common cause for dolphin capture in gillnets? A review of dolphin catches in the sharks nets off Natal, South Africa. Rep. Int. Whal. Commn. (Special Issue 15): 541-547.
Defran, R.H., Schultz, G.M. & Weller, D.W. 1990. A technique for the photographic identification and cataloguing of dorsal fins of the bottlenose dolphin (Tursiops truncatus). Rep. Int. Whal. Commn. (Special Issue 12) 53-55.
Durham, B. 1994a. The distribution of the humpback dolphin (Sousa chinensis) along the Natal Coast, South Africa. Unpublished Masters of Science, University of Natal.
Durham, B. 1994b. Natal's humpback dolphins. Endangered Wildlife 16: 15.
Hammond, P.S. 1986. Estimating the size of naturally marked whale populations using capture-recapture techniques. Rep. Int. Whal. Commn. (Special Issue 8): 253-282.
International Whaling Commission, 1994. Report of the workshop on mortality of cetaceans in passive fishing nets and traps, La Jolla. California, 22-25 October 1990.
Jefferson, T.A., Leatherwood, S. & Webber, M.A. 1993. Marine mammals of the world. FAO identification guide, United Nations Environment Programme. Food and Agriculture Organization of the United Nations, Rome. Karczmarksi, L. 1996a. The Dolphins of Algoa Bay. Endangered Wildlife 24: 4-7. Karczmarksi, L. 1996b. Ecological studies of humpback dolphins Sousa chinensis in the Algoa Bay region, Eastern Cape, South Africa. Unpublished Philosophiae Doctor, in the Faculty of Science, University of Port Elizabeth.
Karczmarski, L. & Cockcroft, V.G. 1997. Socio-ecology and population biology of humpback dolphins (Sousa chinensis) in the Algoa Bay region, South Africa: an overview. Unpublished manuscript presented to the annual scientific meeting of the International Whaling Commission, Bournemouth, United Kingdom. September 1997. SC/49/SM23. xxxvipp. Karczmarski, L. & Cockcroft, V.G. In press. Matrix photo-identification technique applied in studies of free ranging bottlenose and humpback dolphins. Aquatic Mammals. Leatherwood, S. & Show, I.T. 1980. Development of systematic procedures for estimating sizes of populations for bottlenose dolphins. Final contract report to National Marine fisheries service, S.E.F.C., Miami, Florida, 98p.
Peddemors, V.M. 1995. Aetiology of bottlenose dolphin capture in shark nets off Natal, South Africa. Unpublished Doctor of Philosophy, Faculty of Science, Department of Zoology. University of Port Elizabeth.
Peddemors, V.M. 1997. Delphinids of southern Africa: a review. Unpublished manuscript presented to the annual scientific meeting of the International Whaling Commission. Bournemouth, United Kingdom. September 1997. SC/49/SM31.
Pollock, K.H., Nichols, J.D., Brownie, C. & Hines, J.E. 1990. Statistical inference for capture-recapture experiments. Wildl. Monographs. 107:1-97.
Saayman, G.S. & Tayler, C.K. 1979. The socio-ecology of humpback dolphins (Sousa sp.). In: Behaviour of Marine Animals. Win, H.E. & B.L. Olla (eds), Vol 3. Plenum, New York. 438pp.
Slooten, E., Dawson, S.M. & Whitehead, H. 1993. Associations among photographically identified Hector's dolphins. Can. J. Zool. 71:2311-2318.
Urian, K.W. & Wells, R. 1996. Bottlenose dolphin photo-identification workshop: March 21-22, 1996, Charleston, South Carolina; Final Report to the National Marine Fisheries Service, Charleston, SC. NOAA Tech. Mem. NMFS-SEFSC-393, 92p.
Würsig, B. & Jefferson, T.A. 1990. Methods of photo-identification for small cetaceans. Rep. Int. Whal. Commn. (Special Issue 12): 43-50.
