Degree Name

Doctor of Philosophy


School of Biological Sciences


In harvested forests, stand structure varies widely with logging history, yet little is known about how differences in forest structure affect flight activity in insectivorous bats (Microchiroptera). In this project, conducted in harvested eucalypt regrowth forests on the south coast of New South Wales, Australia, I assessed the vertical distribution of insectivorous bats and their potential aerial prey in relation to structural clutter. I also assessed various methodological issues relevant to forest-bat research.

Bat echolocation activity and insect abundance were each compared in three vertical strata (understorey, subcanopy and canopy), two locations (forest and track) and three time periods (evening, night and dawn), in old and young regrowth sites. Mean bat activity in the subcanopy and canopy was up to 11 times that in the understorey of forests, a pattern opposite to that of insect abundance. However, vegetation was more cluttered and bat activity lower in the upper strata of young, than old, regrowth. Mean activity on the track was 2-5 times higher than in the forest, particularly at understorey level where clutter levels differed most. Time of night had little effect on bat activity. The negative response of bat guilds to increased clutter was strongest in the open-space guild and weakest in the edge-space guild with the highest frequency calls, as predicted from echolocation call design.

Although half of the young regrowth sites I used had been silviculturally thinned or selectively harvested 4-9 years previously, there was no significant effect of thinning, stratum or their interaction, on the activity of any bat group examined. Thinned regrowth was less cluttered for some vegetation components, but more for others. Therefore, thinning may not have been sufficient to remove clutter-induced restrictions on bat activity.

Predator evasion by insects may influence their spatial distribution in forests and, thus, their potential availability to bats. For example, atympanate (earless) moths may escape aerial predation passively using cryptic flight behaviour in proximity to clutter, while tympanate (eared) moths can detect approaching bats and employ evasive flight behaviours. In my insect samples, both eared and earless macromoths (moths with single wingspan ≥4.5 mm) were significantly more abundant and taxonomically rich in the cluttered understorey than the upper strata of old and young regrowth forests. Earless macromoths had greater relative abundance, and were more diverse, in both the understorey and the foliage-rich canopy than the subcanopy, supporting the possibility of passive predator avoidance. However, there was little evidence that frequent activity by echolocating bats in the upper strata of old regrowth forest altered the vertical stratification of eared moths, via evasive flight responses. Combined, macromoths were less abundant, in both absolute and relative terms, where bats were most active. This should confer anti-predation benefits, whether macromoths are hunted preferentially or opportunistically. The potential availability of flying prey for forest bats clearly varies with prey group and vertical stratum.

There are likely to be levels of clutter at which even small edge-space bats using highfrequency echolocation are restricted in flight. I assessed the flight patterns of Vespadelus vulturnus in large circular field arenas at forest sites with different logging histories. Light-tagged individuals used similar, direct, patterns of flight in the two oldest classes of regrowth, selecting both subcanopy and understorey strata over the canopy. On average, they flew in two strata and changed flight stratum every three or four observations. However, flight patterns differed in the youngest regrowth age class, which was either unthinned and very densely cluttered, or recently thinned and very open in structure. In both unthinned and thinned regrowth, only the understorey was selected and many bats returned to lower strata after an initial ascent. However, in unthinned regrowth, bats also differed by using three strata and changing flight stratum every two observations. Here, control of flight height may have been difficult whilst negotiating the densely spaced stems. The tendency for bats to fly close to the ground in thinned regrowth suggests that these sites did not provide equivalent flight conditions or habitat for V. vulturnus as older regrowth.

I recommend that forest with tall open structure and forest tracks should be preserved in timber production landscapes to facilitate flight and aerial foraging by bats. More research is required to determine whether thinning practices can be refined to attract more use by bats. Forest-bat ecologists should refine their sampling methods to encompass the vertical and horizontal variation in bat activity, clutter and potential prey availability apparent within and among forest management units. I also found that using attractant insect light traps can artificially increase bat echolocation activity, although at different rates for different guilds. Therefore, light traps and bat recording equipment should be spatially separated in habitat-use studies, as was done in my work. However, species identification was enhanced in recordings of bats close to light traps, and this method could be a novel and useful inclusion in presence-absence bat surveys.