All known contemporary or recent occurrences are in areas of rugged limestone karst with largely intact forest cover. The documented elevational range of c. 500–2,400 m spans the habitat categories of Hill Forest, Lower Montane and Upper Montane Forests. The dentition and other features of the species indicate that it is frugivorous. The individual shot near Taman Village was visiting a fig tree (Ficus sp.) in an old garden area (Flannery 1995).

The two confirmed cave roosts are both associated with precipitous sinkholes that afford some protection from local hunters. At Luplupwintem an adult female was collected carrying a dependent young in May 1992. In PNG the closely related Bare-backed Fruit Bat (Dobsonia moluccensis) also favours inaccessible roost sites in areas where there is a risk of human hunting. However, in northern Australia (Churchill 1998) and in parts of PNG that are remote from human habitation (K.P. Aplin unpublished) they roost in many different situations, including large and small sinkholes, caves, fissures and rock overhangs, tree hollows, and in variable-sized groups ranging from a dozen or so individuals up to congregations of many thousands. There is some landowner testimony to suggest that large congregations are seasonal phenomena, perhaps associated with breeding activity. However, females with newborn young or in advanced stages of pregnancy have also been observed using small caves and fissures, even when larger systems are present in the same local area. Accordingly, it seems likely that Dobsonia moluccensis (and by extension, Aproteles bulmerae) are not restricted by physiological factors to any particular type of roost but select roost sites based on various criteria that includes the risk of predation.

All pteropodids including Dobsonia moluccensis (Dwyer 1975, Churchill 1998) give birth to single young. Gestation period in Australian populations of Dobsonia is c. 5 months and weaning occurs at around 5–6 months of age. Sexual maturity is attained in the second year. Similar parameters can be safely assumed for Aproteles bulmerae.

Nothing is known about patterns of movement in either Aproteles bulmerae or Dobsonia moluccensis. Other similar-sized species of Pteropus are known to travel at speeds of 40 km/hr and to travel several 10s of kilometres to visit preferred food resources. However, many of the relevant studies have been carried out in highly modified landscapes where resources are fragmented and scattered, and the same pattern of dispersal may not apply in areas of continuous forest cover.

This species is endemic to New Guinea and is recorded only from the eastern half of the island (Papua New Guinea). Living animals have been recorded at two sites in two widely separated localities, and recent skeletal remains have been found in two other areas. The known elevational range of the records now is from c. 500 m to 2,400 m a.s.l. and includes areas of Hill Forest and Montane Forest. All known areas of current or recent occurrence are in areas of rugged limestone karst.

It was first described in 1977 from 12,000 year old fossil remains from Eastern Highlands Province of PNG. Unbeknown to the describer, living animals had been collected in 1975 by an anthropologist working in Western Province of PNG. These came from a sinkhole called Luplupwintem, close to the edge of the Hindenburg Wall and at an elevation of 2,400 m a.s.l. The first visit to Luplupwintem by biologists found only a few resident bats which had been visited shortly before by landowners armed with ropes and shotguns. Subsequent visits throughout the 1980s failed to record any bats. Tim Flannery of the Australian Museum and Lester Seri of the Papua New Guinea Wildlife Division (Flannery and Seri 1993) visited the cave in 1992 and 1993 and estimated the population to be 137 and 160 individuals in successive years. In the 1980s and 1990s the species was recorded from two sites in the Star Mountains region but the species is now no longer present.

In 1984 an individual was shot at night over a native garden at Taman Village, in the Telefomin valley (Flannery 1995). This locality is c. 32 km from Luplupwintem which seems an excessive overnight foraging distance from this site. If another, closer roost site is present in the Telefomin area, its location is not yet known.  

In 1995 fresh skeletal remains were collected in Eastern Highlands Province—these were killed by a local hunter at the locality of Huavegemu Cave near the village of Herowana. Subsequently, skeletal remains of two more individuals were recorded in the same area which lies within Crater Mountain Wildlife Management Area (CMWMA) (D.D. Wright pers. comm.).

In 2005-2006 University of Papua New Guinea student Kore Tau examined ten bat-occupied and four unoccupied caves and sinkholes in two distinct localities within CMWMA, twenty kilometers apart (surrounding Herowana and Supa). Among the ten roosting sites, living Aproteles bulmerae was confirmed as present (distinguished by both the lack of incisors and the presence of second digit claw; Bonaccorso 1998) in only one sinkhole near Herowana. This sinkhole is connected to the cave from which hunted remains had come (Wright et al. 1995). Tau was able to net six individual A. bulmerae at this site, which also hosted roosting Dobsonia. Because the two species co-occurred a count could not be used to estimate the size of the Aproteles colony. At a second nearby sinkhole two possible Aproteles were netted but for these only the lack of incisors was noted, without examination for the developed claw. An analysis of environmental variables to find predictors for Dobsonia/Aproteles occupation indicated that difficulty of human access to the cave/sinkhole was the only solid predictor.

In 2011–2012 Kore Tau and Enock Kale of the PNG Institute of Biological Research searched and netted caves and sinkholes, and interviewed local hunters, at several sites in PNG including the Crater Supa area and sites near Mt. Karimui in Chimbu Province, Mekil-Mt. Stolle in Sandaun Province, and the Huon Peninsula in Morobe Province. They found no evidence of A. bulmerae on these trips.

Between 2011 and 2015 wildlife consultant Ken Aplin found the remains of Aproteles bulmerae in two additional areas of remote limestone karst in southern Papua New Guinea. One of these areas is in Western Province in the headwaters of the Fly River, with a local elevational range of 500–1150 m a.s.l. The other is in Gulf Province in the vicinity of the middle Purari River, with local elevational range of 100–550 m a.s.l. In each of these areas Aproteles remains were found among bones and teeth accumulated in caves by roosting owls (probably the Sooty Owl Tyto tenebricosa). In each case, the remains were judged to be recent, based on geomorphic criteria and the presence in two sites of a small number of recently regurgitated casts (these did not contain Aproteles remains). Two owl roosts were sampled in each of the areas and all roosts produced remains of Aproteles. At the two sites in Gulf Province the Aproteles remains are exclusively unerupted teeth derived from very young bats; the owls were most likely predating young bats left unattended at a nearby maternity roosts (the hunting range of the owls might extend a few kilometres from the roost). The two sites in Western Province contained the remains of subadult bats. In both areas the caves are remote from higher elevation country, hence the Aproteles were residing in areas of Hill Forest habitat. Both areas are characterised by moderately to extremely rugged karst terrain and support large tracts of undisturbed forest. Because of their remoteness it is unlikely that these populations have been subjected to any hunting activity in recent times. Unless they have declined for some natural reason, it is likely that they are extant. The known extant distribution of Aproteles, if limited to the areas immediately around the confirmed sites of Luplupwintem and the two Crater Mountain caves, is probably around 100–200 km².

However, the new records in the Fly and Purari catchments suggest a potentially broader distribution in the hinterland of the Gulf of Papua. Much of this country is extremely rugged karst terrain lying at 400–600 m a.sl., and it is very thinly populated, heavily forested and biologically poorly known. Similar low to mid-elevation karst terrain extends west into the Indonesian province of Papua. The contemporary occurrence of Aproteles in this region requires confirmation.

The Crater Mountain sites are within the Crater Mountain Wildlife Management Area and there is local community awareness of the conservation significance of this species.

The Luplupwintem site is currently unprotected and the majority of regional landowners seem to be unaware of the conservation significance of this species. Among Wopkaimin people there also seems to be a lack of awareness of the distinction between Aproteles bulmerae and the morphologically similar Dobsonia moluccensis. A failure to differentiate between these species presents a challenge to the effective conservation of the rarer species.

Areas of suitable habitat in adjacent areas of Papua Province, Indonesia, or to the east of Luplupwintem should be surveyed to locate any remaining colonies of this species. The status of the species in the Upper Fly River and Middle Purari River catchments needs to be reassessed and other areas of similar, low-elevation limestone karst in the hinterland of the Gulf of Papua also needs to be surveyed. Initial detection of the species through the examination of owl roost remains is an effective means of survey for the species in previously unsurveyed localities, especially where the location of potential roost caves is not known to landowners.

Efforts should be made to capture and non-invasively sample a small number of individuals at each of the known sites so that genetic methods can be applied to determine the extent of contemporary and past gene flow between the Luplupwintem and Crater Mountain populations, and to derive a long-term perspective on the demographic history of the species.
If the species is found to be genuinely restricted to only a few isolated populations, a captive-breeding and reintroduction programme should be considered. Research is ongoing to assess habitat requirements of this species (D.D. Wright pers. comm.).