A mark-recapture study of the social organisation of the honey possum Tarsipes rostratus in the Fitzgerald River National Park, Western Australia

The honey possum Tarsipes rostratus is a small (7-12g) diprotodont marsupial found only in south-western Australia, mainly in sandplain heathlands. Tarsipes, the only non-flying mammal that feeds exclusively on nectar and pollen, shows considerable morphological and physiological adaptation to this specialised diet. Their anatomy, diet and reproduction have been comprehensively described but with the exception of one field study near Albany (Wooller et al. 1981) and a detailed study of the behaviour of captive animals (Russell 1986), little has been published on the spatial and temporal patterns of activity of honey possums. A series of mark-recapture studies were started in the Fitzgerald River National Park in 1984 and this thesis describes some of the results. From February 1984 to April 1996, honey possums were caught in grids of 10x10 pitfall traps in the Fitzgerald River National Park on the south coast of Western Australia. The 72 trapping sessions resulted in 8 044 captures of honey possums from a total of 110 859 trapnights in vegetation long unburnt, an overall trapping success of 7.3%. Overall, highest capture rates of honey possums were in winter when nectar levels were high and lowest in spring when nectar was most scarce. Significantly more males than females were caught but this was not due to an unequal sex ratio but rather because males were more active. Females carrying pouch young were caught in all seasons, although autumn usually had fewest breeding females. Free-living young were also caught in each season. The maximum life span recorded for honey possums in the field was about two years, but very few individuals were recaptured after only one year. The distance moved by honey possums between successive captures varied greatly but most moved less than 30m. Significantly more males (30%) than females (22%) moved more than 30m within a three day trapping session; however, the difference between the sexes was not significant between trapping sessions one to three months apart. The distances moved by honey possums did not differ significantly either seasonally or over increasing intervals between trapping sessions. Home ranges, estimated by the minimum area method for individuals caught six or more times, varied from 1 50m2 to 3494m2, with a mean of 1 277m2 (0.1 3ha) for 30 males, and from 112m2 to 2212m2, with a mean of 701m2 (0.07ha) for 20 females. Males had significantly larger home ranges than females but neither sex appeared to be territorial. Honey possums are sexually dimorphic in size, with many more females than males reaching a large size. On the basis of home range overlap, it is suggested that honey possums may have a promiscuous or polyandrous mating system. A promiscuous mating system is more likely, with males travelling longer distances looking for females as well as food. Interception lines were erected around one grid in order to detect dispersal from this grid. Very little evidence of dispersal was obtained, so that disappearance of marked individuals from the population was probably due to their death. There was some suggestion that all honey possums move farther in spring, when nectar is scarce in the Fitzgerald River National Park, but the low numbers in spring make this difficult to confirm. Of 23 male and 18 female honey possums tracked using the spool-and-line technique, 9 males and 9 females could be followed for at least 10m. On average, about 75% of each of these trails were along runways on the ground. Eleven individuals were tracked to feeding sites on Banksia, Dryandra or Calothamnus flowers. Over 116 hours of field observations, only 30 honey possums were observed, usually only fleetingly. However, five of these were observed feeding for a total of ten minutes. Honey possums appear to be trap-line foragers, utilising a specific foraging path over sequential nights. Comparison of the social organisation of the honey possum with other small mammals and other small nectarivores suggests that Tarsipes rostratus is widely separated from phylogenetically related small mammals and has no real ecological counterpart anywhere in the world