Shelter

The availability of shelter is the second most important habitat feature which influences the abundance of western ringtail possums within the sites. Our ringtails are active during the night and need to rest safely during daylight hours. While dreys are the dominant shelter for coastal populations, inland populations use tree hollows as their preferred resting place. (Jones et al, 1994a)

Ringtail possums are natural drey builders. Dreys are nest-like structures made from twigs and leaves preferably of the peppermint tree because those stay pliable and soft for quite a long period and can keep the drey functional for years.
Dreys can mostly be found in the crown of peppermint trees, but also occasionally in eucalypts such as jarrah, tuart or marri, in banksias, New Zealand Christmas trees, melaleuca and even non-native vegetation.

Drey building seems to mainly occur in areas dominated by peppermint (Jones et al, 1994a, Clarke, 2011) and the vegetation structure determines the choice of rest site to a far higher degree than physiological necessities such as for instance protection from heat stress. (Wayne et al, 2005)
Hollows which provide the best protection from high temperatures take a long time to form and are usually in short supply.

Research into shelter use and nesting behaviour of the common ringtail possum agrees with the importance of vegetation structure but argues that this is a more complex issue than recognised and that it is an individual trait – animals are either primarily hollow users or drey users – and that there are differences in usage between the genders. (Lindenmayer et al, 2008)
Judy Clarke also found significant differences not only between locations but between genders. Females in her research used dreys more than males and adjusted the size of the drey according to need. The largest most elaborate structures that can provide substantial protection usually accommodate a female and her offspring. (Clarke, 2011)

There is a wide variety of dreys and probably in drey building skills. They range from a simple platform of twigs with no protection from above to dense structures resembling the form of a football, with one or sometimes two entrance/exit holes. The latter keep the animals remarkably dry even in heavy rain. (Thompson and Thompson, 2009)
The drey’s height in the tree is also highly variable but the majority seems to be in the middle strata of the vegetation, well protected from predators from the ground and from above. (Jones et al, 1994b)

The lack of insolation in dreys compared to tree hollows potentially increases the incidence of mortality in hot summers. However, the ability to construct a drey alleviates the need to compete for hollows with other arboreal animals such as brushtail possums. The use of dreys could also be beneficial as the vibration of branches gives warning of the approach of native predators such as pythons. It is also harder to escape from a hollow if a predator emerges.

In urban and semi-urban, peppermint dominated areas mere tree forks are sometimes used for resting or if more protection (e.g. from ravens) is necessary the fronds of palm trees provide hidden shelter. In our backyards any thick vegetation such as lilly-pilly, ivory-covered trees or hedges, spiky lantana thickets or even Brazilian/Japanese pepper trees (Schinus terebinthifolius), which are actually a threat to coastal Australian ecosystems, are potentially used by ringtails.

During hot summers dense understorey like sword grass, hollow logs on the ground, a clump of sedges, reeds, blackberry thickets, disused rabbit warrens and even debris can accommodate a ringtail. (Harewood, 2008)

At her woodland release sites Judy Clarke observed the frequent use of grass trees and witches brooms (a dense mass in a woody plant resembling a broom, caused by a disease) for day sheltering. She attributes the high use of grass trees to their insulative quality (Swinburn et al, 2007) and the need to alleviate the risk of heat stress. (Clarke, 2011)

The number of different rest sites used by one ringtail varies widely in the literature - 3 to 15 locations. 
The number of dreys is usually given as 3 to 7 used by a single individual. 
The use of a high number of rest sites could make a catch harder for potential predators as repetitive behaviour could otherwise be observed and exploited (Clarke, 2011). 

Drey occupancy also seems to vary widely with different habitats. Barbara Jones determined an occupancy rate between 8 and 45%. (Ellis et al, 1992)
Dreys are also used by other ringtails, not just the builder.
While common ringtails share a drey with their partner and offspring, in western ringtails only mothers share with their offspring.

According to a Honours study there is a 30% probability that possums will use a drey on days with no rain and that this probability decreases with increasing rainfall. The interesting question here would of course be, where the animals rest if not in their dreys. The author concludes that they would then move into man-made possum boxes. Translocated ringtails should therefore be provided with boxes. (Harring-Harris, 2014)
Those findings are in stark contrast to another Honours study looking into the usage of possum boxes by translocated animals which found no advantage in providing them. (Moore, 2007)

Researchers acknowledge that it is hard or sometimes impossible to identify whether a drey is used by animals. Harring-Harris (Harring-Harris, 2014) does not specify how she developed her assumption and whether she actually checked dreys in dry and wet conditions. No sample size, frequency or method of checking are given. She only states that she used an information theoretic approach and explains the relevant calculations. It is very hard to reliably estimate occupancy of boxes but checking high numbers of dense structures in the canopy of trees particularly in heavy rain seems close to impossible to me.  
In contrast to this thesis, ringtail possums were observed that stopped feeding and retreated to their shelter when rain became heavy. (Thompson and Thompson, 2009)
My own experience would tell me that the animals will wait out the rain in their (well made) dreys and not be seen entering or exiting. However this is solely based on observation without any scientific backing.

I fail to see the logic in the conclusion that if wild animals do not use their own structure, they will use our man-made structures. We cannot expect a ringtail to recognise the advantages of a wooden box in comparison to a woven organic structure. The very limited research and my personal observations hint at a low uptake rate of boxes in forests or woodland areas by wild ringtails. Hand-reared and animals that endured a long period in captivity are however likely to behave differently. (also see 'Field experiences with nest boxes')

Our man-made structures are inherently inferior options (Gleeson and Gleeson, 2012) to protecting and retaining trees with natural hollows or that contain dreys.