No abstract

To maximise bat retention in urban environments, efficient bat monitoring is needed, but the factors that influence survey results for urban bats are unclear. We used echolocation call detectors (n = 378 detector-nights from November 2005 to October 2006) to assess bat activity among different sites in the Adelaide City parklands, temporal variations in activity (hourly, nightly, monthly), and responses to weather and light (artificial and moon). Bat species did not occur evenly in urban conditions; dark parks were more important for bat diversity and activity (six species groups in the darkest park) than were artificially lit parkland areas (three species groups in the flood-lit park). Gould’s wattled bat (Chalinolobus gouldii) and Mormopterus species 4 (94% of calls) were advantaged in urban parklands, being the only species recorded when lights were on at sports parks, whereas five species groups occurred when the lights were off. Minimal bat activity was recorded in the first 2 h after civil twilight, suggesting that bats may roost outside the city and commute nightly into parklands. Bat activity increased with temperature, with a burst in activity occurring after 7°C. Rainfall (>1 mm/24 h) and moon illumination at midnight did not influence activity. Urban environments should provide diversity to attract a diverse assemblage of bat species. Activity fluctuated among sites, nights, and across the year, indicating that large sample sizes over long periods of time are required to monitor and survey bats reliably with detectors.

The Dilleniaceae is known to produce nectarless flowers pollinated by bees, but the fact that bats ingest Dillenia biflora pollen led us to question pollination assumptions for these trees. We aimed to identify the pollinators of D. biflora, check for nectar presence, and investigate potential for cleistogamy and global prevalence of this pollination system. We examined aspects of the pollination of D. biflora on two Fijian islands using video recordings, direct observations, hand pollination, measurements (flowers, bite marks, nectar), and monitoring. The flowers, receptive for one night, contained copious nectar and had permanently closed globose corollas that required removal by bats for pollination. All the 101 flowers that retained their corolla died and did not produce seeds by cleistogamy. The bat Notopteris macdonaldi was well adapted to corolla removal. Keeping corollas closed until bats manipulate the nectar-rich flowers is a beneficial strategy in high-rainfall environments with many flower parasites. We propose to name a pollination system reliant exclusively on bats “chiropteropisteusis.” From clues in the literature, other species in the geographical range of Dillenia are probably chiropteropisunous. Chiropteropisteusis should be investigated in the Old-World range of Dillenia, many species of which are threatened. The remarkable “fall” of the entire corolla observed by an earlier botanist for several species in the genus is most likely attributable to bats. This discovery has important implications for the conservation of bat-dependent trees and their associated fauna, particularly considering the high level of threat faced by flying-foxes globally.

Bats provide important ecosystem services throughout the South Pacific but many species are threatened. The Fiji archipelago has a rich bat fauna yet few arrangements exist to promote the conservation of this group. From July 2008 to March 2011 we visited important sites identified in previous surveys for bats on Vanua Levu, Taveuni, Viti Levu and the Yasawa Islands. We netted and recorded five of Fiji's six bat species (Pteropus tonganus, Pteropus samoensis nawaiensis, Notopteris macdonaldi, Chaerephon bregullae and the Critically Endangered Mirimiri acrodonta). This was only the sixth recorded capture of M. acrodonta. P. tonganus is widespread in Fiji and exploits many different environments. P. samoensis nawaiensis mostly uses forested areas, and large tracts of forest are required for its long-term protection. Cave protection is vital for the two insectivorous species (C. bregullae and Emballonura semicaudata) and N. macdonaldi. In particular, the Nakanacagi maternity site remains globally significant for C. bregullae. E. semicaudata was not caught or detected during our surveys. It appears to be extinct from most large islands in Fiji and has abandoned the only large cave roost (Yaqeta) observed 10 years ago. This species has disappeared from or declined dramatically on Vanuatu, Samoa, Tonga and Fiji. We recommend that the category of E. semicaudata semicaudata on the IUCN Red List be changed to Critically Endangered. Management priorities should focus on community engagement and education, and improved cave and forest management and protection.

Temporal variation of insect communities in urban environments is poorly known and mechanisms driving these changes are unclear, as are the implications for insectivorous predators. We examined the relationships between season and nocturnal aerial insect biomass and biodiversity, and between temperature and insect biomass in the Adelaide zoological gardens from December 2005 to September 2006. We also compared the effectiveness of two insect trap types and used a bat detector to assess bat activity in relation to insect biomass. During the study, 9,939 insects from 13 orders were collected at the Adelaide zoo with a Malaise trap and a light trap. Mass and diversity of insects were highest during warm months, as was bat activity, and bat activity was positively correlated with insect biomass. Winter-active insects consisted predominantly of Diptera and Lepidoptera, which may provide an important winter food resource for insectivorous bats. The Malaise trap attracted fewer insect orders and biomass than did the light trap, and insects congregated within 6 m of artificial lights, so bats that forage at lights may have an advantage in urban areas. A strong need for the inclusion of urban insects to biodiversity inventories exists in the context of bat conservation.

We used scat and isotope analyses to assess insectivory in Fijian flying foxes (Pteropodidae), seeking insights into niche partitioning of co-occurring bat species with high plant diet overlap. Moth scales were most common in scats of Notopteris macdonaldi (87%; P. tonganus: 62%; Pteropus samoensis: 36%) and may indicate shared resources. The small and highly manoeuvrable N. macdonaldi exploited nectar-rich flowers also favoured by moths (e.g. Barringtonia spp.). Other invertebrate remains were most frequent in scats of P. tonganus (69%). On the basis of scat results and ecological observations, P. tonganus uses a combination of insectivory and a highly varied plant diet to obtain sufficient nutrients. Scats of P. samoensis contained few invertebrate remains, but abundant protein-rich plant species (including Freycinetia spp.), and juveniles seemed to consume moths frequently. Clustered d 15 N and d 13 C for N. macdonaldi and P. samoensis indicated a narrower dietary breadth than that of P. tonganus. P. tonganus juveniles appeared at a significantly higher trophic level than did adults, probably the result of milk consumption and/or higher rates of protein synthesis. The methods used detected little evidence that bats partitioned resources vertically. This study generates hypotheses for the further examination of flying-fox diets.

The response of primary (PF) and secondary (SF) rainforests to cyclones has broad implications for servicing fauna and the resilience of forest functions. We collected fine-scale data on the reproductive phenology of plant communities in Fijian PF and SF in 12 monthly surveys before and after Cyclone Tomas (2010). We generated a resource index from the reproductive loads of 2218 trees and 1150 non-trees (>190 species) and trunk and stem diameter to assess patterns in resource abundance for nectarivores and frugivores (hereafter NF resources). We aimed to determine (i) whether species richness of NF resources differed between forests; (ii) the patterns of resilience of NF resources at community level in both forests after a cyclone; and (iii) the effect of response on NF resources for plant-servicing bats (Pteropodidae). In 12 months preceding the cyclone, NF resources were greater in PF trees; non-tree resources fluctuated and were greater in SF. Lower species richness of NF resources in SF indicated that fewer opportunities exist there for exploitation by a diverse fauna. More resources were available for bats in PF. In 12 months following the cyclone, PF flowers and fruits, and SF fruits specifically used by pteropodid bats decreased for trees. Non-tree resources were especially susceptible to the cyclone. No universal pattern of decline was associated with the cyclone; instead, some NF resources declined and others were resilient or responded rapidly to a post-cyclone environment. Both PF and SF demonstrated resilience at the community level via increased flower survival (PF) and rapid flower production (SF). Reduced species richness of NF resources in SF will compromise future resilience and response to disturbance, including for threatened pteropodid bat species. These findings are critical for long-term management of forests, given predicted increases in cyclone frequency and intensity associated with anthropogenic climate change.

Easily captured species provide powerful opportunities to motivate research and conservation action. Conversely, rare and cryptic species with low capture probability can be difficult to study. We examined the capture success of Fiji’s Pteropodidae (Pteropus tonganus, P. samoensis, Notopteris macdonaldi, and Mirimiri acrodonta) and incorporated capture success and associated observations into an assessment of their suitability as umbrella species for conservation. We caught 252 bats from 2493 mist-net-hours in lowland rainforest near Waisali on Vanua Levu, and 38 bats from 1915 mist-net-hours in cloud forest on Taveuni. Of these captures, 67% were N. macdonaldi, 16% P. tonganus, 18% P. samoensis, and <1% M. acrodonta. All species were difficult to net in cloud forest, where we recorded a female sex bias. Capture success was affected by site, season, and net height; peak capture times varied among species and between sites. N. macdonaldi was recaptured most frequently (8.8%), revealing habitual nightly movements. It has strong potential as an umbrella species for rainforest and cave-dwelling organisms of Fiji, and was readily netted. P. tonganus is an unsuitable umbrella species, P. samoensis has potential as an umbrella species, and M. acrodonta has many attributes of a flagship species. Capture efficiency of N. macdonaldi provides opportunities for monitoring and conservation management activities in Fiji. However, variability in capture success among species means that different approaches should be explored for conserving bats effectively with limited resources.