Ecology of the brush-tailed phascogale in jarrah forest of southwestern Australia

This study investigated the ecology of the marsupial brush-tailed phascogale (Phascogale tapoatafa; phascogale) in jarrah forest of Western Australia (WA). The thesis provides a descriptive account of the species' population dynamics, the phascogale’s size and growth, foraging ecology and nesting behaviour. The main work was undertaken between 1992-1995 and information was gathered by capture and by using radiotelemetry. The availability of food sources and refuge sites was examined and the species' immediate response to a logging operation was investigated. On a broader scale, the taxonomy of phascogales in WA was reviewed and their past and present distributions were explored by modelling and examination of museum records. The species has previously been subject to only one detailed study and this was undertaken in the State of Victoria. Both similarities and differences were observed between the phascogales in WA and Victoria. Many of the differences appeared to have a nutritional basis. Phascogales are primarily arboreal insectivores and some of the invertebrates most commonly found in scats and stomachs were not abundant on trees. Nectarivory was very rarely observed (cf. Victoria). While phascogales are generalists in diet, they clearly show preferences for certain foods. The strongly seasonal climate in the southwest (long dry summers and wet winters) and the lack of diversity in tree species and low nectar availability probably limit food for WA phascogales. The dispersed nature of tree invertebrates and the apparent scarcity of some preferred prey, such as beetles, possibly explain the exclusive and often large size of female territories. Phascogales in the study area were smaller and less sexually dimorphic than those in Victoria. Males were 30%, and females 20%, less in weight than those in Victoria. They were similarly smaller in skeletal size. Overall size was also found to vary' between habitat type and between years. In a year of drought phascogales did not achieve typical body size with mature males in that year weighing an average 25% less than usual. The evidence strongly suggests that annual and local availability of food is a major determinant of body size and growth. Such variability in food availability may also be the evolutionary basis behind the litter sizes of WA phascogales. These are smaller than those in Victoria (mode WA = 6, Victoria = 8). The timing of major life-history events was as described for Victorian phascogales, although births occur a little later in the year. All males died at the end of the single annual breeding season (male semelparity) and young took some five to six months to raise to weaning. Some females survived to breed in a second year but the number encountered was low. The toll of lactation probably reduces lifespan and females were often in poor condition at late lactation and maximum maternal effort appears selectively invested in the first litter. Typical of the semelparous species, females initially tried to raise as many young as they have teats. However, there was variation in litter sizes among females. The basis for this was anatomical as females had six, seven or eight teats. Such variation occurred throughout the study area, within litters and appears a State-wide phenomenon. Except in a hybrid Antechinus population, there appears to be no marsupial precedent for intrapopulation and intralitter variation in teat number. In such a strongly selected trait, it is speculated that such variation could only persist if the environment was spatially and/or temporally unstable in terms of food availability. There was no evidence that reduction in teat number was a direct trade-off that improved the chance of surviving to breed on a second occasion, but data were limited. The teat trait is presumably under genetic control and the promiscuous mating behaviour of phascogales may contribute, via multiple paternity, to the intralitter variability observed. In the year of drought litters were significantly female biased. Neither sex were sexually dimorphic until they began foraging for themselves, therefore such bias was unlikely to reduce maternal stress during the drought conditions. The adaptive advantage of the bias was undetermined, but current literature indicates that such bias occurs at conception. However, female offspring probably have better post-weaning survival than males. Examination of refuge requirements showed that natural nest sites were located in tree hollows. The profile of the trees chosen agree with most studies on hollow-nesting species. There was no apparent preference for particular tree species but there was a preference for nesting in older and senescent or dead trees. Females with dependent young showed particular preference for these trees' forms. However, once a tree was used, no tree characteristics measured were predictors of the tree being used on further occasions. Excluding females with young, phascogales typically spent 2-5 days nesting in the one refuge before moving to another. Females moved between alternative refuges more than males and home range maintenance is proposed to account for the difference (males were not territorial cf. females). For both sexes, parasite avoidance might account for the generally low level of nest-site fidelity. Individuals were estimated to use around 27 (males) and 38 (females) different nest-sites during one year of adult life. The parameters of hollows examined showed a preference for using hollows with small entrances. More than predator avoidance, interspecific competition for hollows may explain such selectivity. Following the year of drought, communal nesting was common in autumn and winter. This was in marked contrast to data gathered early in the study and to the nesting behaviour of Victorian phascogales. Communal nesting was probably an energy conserving strategy adopted to compensate for unusually small body size. This behaviour may have been a single year event. However, as phascogales in the area are normally much smaller in body mass than those studied in eastern Australia, they may tend to nest communally in winter or when under conditions of hardship. The immediate response of phascogales to logging was examined. Those affected continued to travel through and feed in the logged areas, which reinforces the concept of high site fidelity among animals. Phascogales were commonly found feeding among ground debris in cut areas indicating flexibility in foraging mode. With rare exceptions they ceased nesting in trees in the logged parts of their territories and confined such nesting to surrounding uncut forest. This suggests that the forestry practice of retaining a select number of hollow-bearing trees/ha may be insufficient to meet the species' refuge requirements in logged areas. Of concern is that currently unlogged sections in logged forest can be cut within 10-20 years yet trees take some 200 years to develop hollows. Additionally, the value of young regrovvth as a food source to this species is questionable. The study highlights the concerns that many scientists have regarding the preservation of hollow-nesting fauna in areas that are impacted by logging. A revision of the species taxonomy (including the subspecies P. t. pirata) indicates that southern WA phascogales warrant subspecies status. Differences in basiacranial features were found between the regional groups in Australia. However, the issue of long-term geographic and reproductive isolation is perhaps the greater argument for suggesting subspecific status for WA phascogales. This is currently being examined using mitochondrial DNA techniques. This will provide not only a DNA profile of similarities and presumed differences between WA and southeastern phascogales, but will give an estimation of the time that the two groups have been separated. A modelling exercise undertaken to examine the potential past distribution of WA phascogales showed that climatic conditions favourable to this species are found throughout the southern part of WA. For many of these areas there are no records of phascogales although a single historical work (1909) indicated that they were present in these areas. At that time they were apparently already extinct in some areas and dying out in others. A survey of the northern jarrah forest and a few other areas (by using nestboxes) failed to detect phascogales. The population examined during this study was therefore quite unusual in density, as phascogales were readily found in nestboxes. A myriad of factors culminating in low feral predation in the area seems the most likely explanation for the high densities observed. The conservation status of phascogales in areas other than the study region requires examination. Predation by exotic predators might account for the species' rarity, but food appears a generally limiting factor for phascogales and processes that alter habitat productivity may significantly impact this species