EFFECTS OF INCREASED FOOD DISPERSAL AND RANDOM FEEDING TIME/PLACE ON STEREOTYPED BEHAVIORS IN OTTERS AT ADELAIDE ZOO

BY LIANNE HAWKE, PETER LAUER, DAVID BARTHOLOMEUSZ AND ZETA STEEN

Introduction

Despite the best intentions of zoos, there are problems concerned with keeping wild animals in captivity. One of these problems is stereotypy. Stereotyped behavior patterns are considered symptomatic of boredom and unhappiness in the animal caused by captivity. These behaviors can take the form of pacing, weaving and self-mutilation. Stereotypy in captive animals is of great concern to zoos, as the condition indicates that some essential environmental or social element is missing (Markowitz et al., 1978; Mellen et al., 1981; Markowitz and LaForse, 1987).

Stereotyped behavior is not necessarily abnormal. Stereotypies occur both in captive wild animals maintained in zoos and in domestic animals kept in high-density housing. Although researchers recognize that abnormal stereotypy occurs only in monotonous, barren and confined environments, they do not agree on the underlying theoretical causes of such stereotypy, i.e. whether the animal is trying to reduce the tension created by the environment or whether it is trying to alleviate the boredom by creating its own stimulation. Caged environments are extremely stressful for the animal. This is particularly true for captive wild animals in zoos. Stereotypy is a behavioral response by these animals used to reduce stress (Repp et al., 1988).

Myers (1978) ran a behavioral enrichment program with a rich diversity of animals kept in zoos and aquariums. He reduced the stereotyped pacing in a pair of jaguars by training them to rear up and strike a paddle, which triggered a feeding mechanism delivering a small piece of horsemeat. This not only reduced their pacing but also made the exhibit more attractive to the visiting public. Similarly, Steen (1995) found that food acquisition devices appeared to work well in promoting more natural percentages of foraging behavior in two species of tamarin (Saguinus oedipus and Leontopithecus r. rosalia). Shepherdson, Brownback and James (1989) conducted an experiment on the slender-tailed meerkat (Suricata suricatta). They created a mealworm dispenser out of a plastic tube by drilling holes in it and filling it with sawdust and mealworms. The mealworms dispensed themselves by moving around until they fell out of the holes. This caused the meerkats to spend significantly more time foraging. There was also a subsequent reduction in the frequency of stereotypy and sleeping (Shepherdson et al., 1989).

Although otters are not threatened with extinction they still need to be protected ( Butler , 1994). Major threats to the survival of otters in the wild are all related to human impact. As people continue to spray pesticides and dispose of chemicals into the environment, waterways are polluted, thus affecting the otter's food supply and water quality (Lekagul and McNeely, 1988). The other major threat to otters is habitat destruction, including clearance of vegetation near suitable calm streams and lakes which otters need to feed and breed (Foster-Turley and Engfer, 1988).

There are four genera and 13 species of otter, distributed over many parts of the world in habitats which include mountain streams, lakes and rivers, estuaries and coastal areas. Otters range in size from the giant otter (Pteronura brasiliensis) and sea otter (Enhydra lutris) through the river otters (Lutra spp.) to the small oriental small-clawed otter (Aonyx cinerea), which reaches only 90 cm in length (Lekagul and McNeely, 1988). A. cinerea can be found from the Philippines through to Indonesia , south-eastern China , South-East Asia and westwards through the Himalayan foothills of Bangladesh ; it may also be seen in zoos throughout the world. The reduced claws in A. cinerea are accompanied by extreme delicacy of the sense of touch, and freedom of movement due to reduced webbing between the digits (Lekagul and McNeely, 1988). This species eats utilizing its front paws and is the most dexterous of the otters. Their diet is based mainly upon molluscs, crabs and other shellfish which they dig out of mud banks and river beds (Kruuk et al., 1994). As the metabolic rate of otters is very high and food is digested and defecated within one hour of ingestion (Lekagul and McNeely, 1988), otters become hungry again quite soon after being fed. If captive otters are fed only once a day they will spend the rest of the day pacing up and down calling and begging for food (Mellowship, 1990).

Otters are social animals, living in groups of several pairs and their families (Furuyu, 1977; Wilson, 1987). The intelligence of the otter is expressed in its curiosity and its need for constant stimulation. Unlike other mammals where play behavior is largely confined to the young, otters remain playful throughout their lives. They need a rich environment containing such things as hollow logs, tree stumps, bushes, saplings and most importantly areas of sand, gravel and grass for grooming (Foster-Turley and Markowitz, 1982; Partridge, 1991). The daily activity pattern of the oriental small-clawed otter is not well known since there has been little study of this animal in the wild (Lekagul and McNeely, 1988).

The aim of the present study was to enrich the environment and document time usage of two captive otters. A catapult (see Fig. 1) was designed to launch whitebait and cockles into the otter enclosure at random times and places throughout the day. It was intended that the otters would not be able to predict where and when they would be fed, that their food would be more dispersed, and hence increases in foraging and decreases in stereotypy (e.g. begging from keepers) would be evident.

Methods

The study was carried out in Section 4 of the South East Asian Rainforest at Adelaide Zoological Gardens , South Australia . This section has a ring lake, two islands (occupied by two siamang gibbons) and perimeter ground. The otters have full 24-hour access to the site. Observations were taken from the walkway and viewing platforms, which were approximately three meters above the exhibit.

Since tattooing, tagging, banding and coat dying were all deemed impractical, small but distinct morphological differences were used to differentiate the otters. One animal was defined as `Large' and was broader in the shoulder, longer and heavier than the other otter, `Small', who had a lighter colored coat and throat.

Data collection

Data collection involved evaluating the behavior patterns of the otters before and after the feeding regime was altered. The instantaneous scan sampling technique was used to sample behaviors (Altman, 1974). This involved two observers (each observer assigned to Otter 1 or Otter 2) scanning the enclosure once every minute, immediately noting the behavior shown by each otter and recording it on the standard data sheets. Therefore, 60 observations were recorded each hour, for each otter.

Behavioral classifications were defined as follows:

Figure 1. Sketch plan of a catapult used in random feeding of otters.

Enrichment apparatus

Two identical catapults were constructed for this project. The frame was constructed with treated pine, and clear plastic bowls were used to hold food. The design had to be sturdy to withstand the force exerted on it by the springs and to tolerate constant handling by curious schoolchildren.

Experimental design

Baseline data were collected, by three observers, over a period of approximately two-and-a-half weeks. Both otters were observed at the same time by two different observers. Each one-hour time period from 7.00 a.m. to 5.00 p.m. was repeated three times, and by the end of the collection of baseline data each otter had been observed for 30 hours.

After the collection of baseline data the feeding regime was altered by introducing the catapult treatment. It was anticipated that the catapults would reduce human impact as much as possible and disperse the food in a random manner, at a random time and random location, so as to increase the amount of foraging. Otters were fed the same mass of whitebait and cockles as they had been before the catapult was introduced. Six suitable sites for launching the food into the enclosure were chosen and numbered 1 to 6. A dice was rolled to determine at what time and place the catapults were to be launched. The catapults were launched six times a day, with the normal ration of 900 grams being divided into 6 ´ 150 g. Half of each 150-gram portion was launched by one catapult and half by the other immediately afterwards in the same position.

During the five days of enrichment data collection, the otters were fed solely by the catapults and not by the keepers. Data was collected similar to baseline data, with 30 hours of observations for each otter. New behaviors were observed and recorded when seen, and added to the behavioral repertoire.

Results

Means of each behavior for each hour and for both otters were calculated for both baseline and treatment data. Data were analyzed using Mann-Whitney U tests.

Data for each otter were combined and mean percentages calculated for all behaviors for the baseline and treatment period. Figure 2 shows that foraging time did not increase after the catapult was introduced, with otters spending 6% of their daytime foraging in both feeding regimes. Sleep increased by 30% after the treatment while rest decreased by 18%. Moving around the enclosure decreased by 4%, while the two stereotype categories (general stereotyped behavior, and observing) both showed decreases, with the former decreasing by approximately 5%, and the latter by approximately 3%. The other categories on their own showed no great difference between the two feeding regimes.

A Mann-Whitney U test was used to test whether there were significant differences in the behavioral time budget of the otters before and during the altered feeding regime.

Table 1. Mann-Whitney results for differences in behavior between feeding regimes (* = significant, NS = not significant).

Some behaviors have been grouped together as well as being analyzed separately. The table indicates that the differences between sleep, rest, observing, moving, stereotyped, active (swim + move + rub), inactive (rest + sleep), and stereotyped (stereotyped + observing), were significant, while the remaining behaviors were not.

Figures 3 and 4 show that the otters spent a large proportion of each hour sleeping or resting (inactive). This did not change with the altered feeding regime. Activity generally followed the same pattern throughout the day, as did inactivity. Feeding behavior was more constant after the treatment, constituting approximately 8% of each hour. Stereotyped behavior decreased from approximately 8% to approximately 1% of the total behavioral budget of the animals.

New behaviors were continually documented during collection of baseline and treatment data. The baseline data were best fit with a log curve (Rsq = 0.9842), while the treatment data were best fit with a linear relation (Rsq = 0.9748). Figure 5 shows that before the treatment, the behavioral repertoire of the otters was exhausted after 25 hours. During the treatment, however, although the behavioral repertoire was presented at a slower rate, it showed no sign of tending toward exhaustion.

Discussion

The aim of the study was to increase foraging time with the use of a catapult, in order to reduce the level of stereotypy. The otters did not show an increase in foraging time, yet stereotyped behaviors were significantly reduced. This may be because their foraging time was increased, but not at the time in which data were collected. It was observed that they were not finding the cockles launched from the catapult during the day. Often the morning feed would remain untouched for the rest of the day, but would be gone by the next morning. It was postulated that the otters might have been foraging at night, in which case foraging may have been greater during the treatment than when they were being fed by the keepers. If this is what was happening, then results are positive, since in the wild otters are considered to be nocturnal, traveling by night up and down streams hunting, and resting in riverside holts by day. A healthy otter is supposed to be always active during its waking hours, and spends roughly equal time asleep and awake (Stephens, 1954).

The increased sleeping and reduced moving and swimming during the day may indicate that the otters were foraging at night, and so conserving their energy during the day. A reduction in stereotyped behavior such as begging may have occurred because nocturnal foraging meant the otters depended less on the keepers for food. Before the catapult apparatus was introduced, both otters exhibited begging behavior when the keepers were in sight, and they would beg at the doors to the rooms, which were frequented by keepers. A greatly reduced exhibition of begging and calling for the keepers was observed while feeding with the catapult, which is likely to be due to the reduced human aspect of the feeding mechanism. As the otters were not being directly fed by the keepers, they stopped begging for food. The catapult was effective as it fed the otters from random locations (so the otters did not know where to find the food), at random times (so they did not know when the food was coming), and it dispersed the food widely, propelling it into reeds and water, instead of throwing it directly to them.

According to Harris (1968), captive otters in more stimulating environments swim more often. Our results, however, showed that swimming decreased with the catapult treatment. These data conflict with Harris, suggesting that time spent swimming, by itself, may be a poor indicator of increased enrichment. In addition, since the otters in this study may have been foraging at night, swimming at night may have increased, without being recorded. Future studies might record data over a 24-hour period to test this conjecture.

The stereotyped behaviors that were observed were limited, since the enclosure for these otters was extremely naturalistic. Mellowship (1990) observed constant bouts of bouncing behavior interspersed with a single, non-repetitive, pace or run along the edge of the otters' small enclosure. The otters in that study also exhibited stereotyped `fossicking' (searching) behavior, due to the frustration of being subjected to an arbitrary feeding routine and an inability to forage for their own food. The stereotyped behaviors which Mellowship observed were begging and barking for food, as well as a ritualized form of bipedal observing behavior which has been observed in many species of otter (Mellen et al., 1981; Mellowship, 1990; Butler, 1994).

The otters in the present study are not naive subjects; their stereotypy has been inadvertently reinforced for a long period of time simply by the keepers feeding them. Skinner (1948) proposed that `stereotyped behaviors develop in an organism in anticipation of food.' He labeled these responses `superstitious', as they emerged through `accidental juxtapositions of reward and response' (Skinner, 1948). It was also noted that zoo visitors fed the otters, effectively reinforcing this behavior.

Since there was only one group of otters available to study, a control was not carried out. This obvious flaw in design could not be avoided. Although it cannot be maintained that the enrichment apparatus decreased stereotypy (since data did not show an increase in foraging), the altered feeding regime was associated with a significant reduction in stereotyped behavior in the subjects. This study showed significant results, which if investigated further, might prove useful when considering the role of more naturalistic feeding regimes in the reduction of stereotypic behavior in captive animals.

Future studies might obtain more useful and powerful information by: collecting data over 24-hour time periods; recording data in a range of exhibits (from barren to naturalistic); and comparing the behaviors of captive-born to wild-born animals under the same conditions. Such research might provide zoo keepers and biologists alike with a better understanding of how captivity affects wild animals, and ways in which increased enrichment might be affordably offered to captive animals.

Acknowledgments

We would like to thank the staff at the Adelaide Zoological Gardens. Thanks also to Dr Michael Schwarz and Prof. Stephen Hall for advice on experimental design and statistics. Thanks to Peter Lauer for making the catapults.

References

Altman, J. (1974): Observational study of behavior: sampling methods. Behavior 49: 227–267.

Butler , J. (1994): Cape clawless otter conservation and a trout river in Zimbabwe : a case study. Oryx 28: 277–282.

Foster-Turley, P., and Engfer, S. (1988): Species survival plan for the Asian small-clawed otter. International Zoo Yearbook 27: 79–84.

Foster-Turley, P., and Markowitz, H. (1982): Behavioral enrichment study: small-clawed river otters. Zoo Biology 1: 29–43.

Furuyu, Y. (1977): Otters in Padas Bay , Sabah , Malaysia . Journal of the Mammal Society of Japan 7: 39–43.

Harris, C.J. (1968): Otters: A Study of the Recent Lutrinae. Weidenfeld and Nicolson, London .

Kruuk, H., Kanchanasaka, B., O'Sullivan, S., and Wanghongsa, S. (1994): Niche separation in three sympatric otters, Lutra perspicillata, L. lutra and Aonyx cinerea in Huai Kha Khaeng , Thailand . Biological Conservation 69: 115–120.

Lekagul, B., and McNeely, J.A. (1988): Mammals of Thailand (2nd ed.). Darnsutha Press , Thailand .

Markowitz, H., and LaForse, S.: Artificial prey as behavioral enrichment for felines. Applied Animal Behavior Science 18: 31–43.

Markowitz, H., Schmidt, M., and Moody, A.: Behavioral engineering and animal health in the zoo. International Zoo Yearbook 18: 190–194.

Mellen, J., Stevens, V.J., and Markowitz, H. (1981): Environmental enrichment for servals, Indian elephants and Canadian otters at Washington Park Zoo, Portland . International Zoo Yearbook 21: 196–201.

Mellowship, R.: Stereotypy and the reinforcement of play behavior in Asian small clawed otters (Aonyx cinereus). Unpublished Honors thesis, Adelaide University collection.

Myers, W.A. (1978): Applying behavioral knowledge to the display of captive animals. In Captive Wild Animals (eds. H. Markowitz and V. Stevens), pp. 133–159. Nelson-Hall, Chicago.

Partridge, J.: The Asian short-clawed otter: some husbandry notes. International Zoo News 38 (1): 25–29.

Repp, A.C., Felce, D., and Barton, L.E. (1988): Basing the treatment of stereotypic and self injurious behaviors on hypotheses of their causes. Journal of Applied Behavior Analysis. 21 (3): 281–289.

Shepherdson, D., Brownback, T., and James, A. (1989): A mealworm dispenser for the slender-tailed meerkat Suricata suricatta at London Zoo. International Zoo Yearbook 28: 268–271.

Skinner, B.F. (1948): `Superstition' in the pigeon. Journal of Experimental Psychology 38: 167–172.

Steen, Z. (1995): Effects of enriched food acquisition on activity budgets of two tamarin species at Adelaide Zoo. International Zoo News 42 (5): 284–298.

Stephens, M.N. (1954): The natural history of the otter. (A report to the Otter Committee, quoted in Mellowship op. cit.)