Referee 3's review:

**********
N.B. These are the comments made by the referee when reviewing an earlier version of this paper. Prior to publication the manuscript has been revised in light of these comments and to address other editorial requirements.
**********

Our understanding of how bats use echolocation in the field has improved tremendously in recent years because of the use of techniques that allow 3D reconstruction of flight paths with simultaneous recording of echolocation pulses. There are two major ways of doing this - acoustic tomography using time-of-arrival differences at a microphone array, and stereo photo- or video-gramettry. Both methods are used in this paper to describe echolocation call intensity in 11 bat species from 5 families. The methods used are not trivial, and have been used effectively (with some caveats - see below).

Although there have been several recent studies of call intensity in wild bats, the data presented here are especially interesting because they are from tropical species, and include representatives of a range of families other than the Vespertilionidae which have been the focus of previous work. The authors present new data on how call intensity changes in relation to distance from targets (typically 4-7 dB / halving of distance), on extreme intensities produced by trawling bats, and argue that species emitting higher frequencies emit calls at higher intensities to counter the strong effects of atmospheric attenuation at high frequencies. Consequently, all bats seem to have similar detection distances. The intensity methods provided are broadly similar to those recorded from temperate vespertilionid bats.

The data are valuable, and the paper is well written. To be absolutely convincing the authors need to provide more information about the error potential in their measurements however. Some of these details are rather vague (third dimension estimated from spoken commentary, photos, directionality of microphones (p14); Flight paths from both methods correlated nicely (legend to Fig 5). More rigour is needed here - what are the reconstruction errors? Which methods were used in which circumstances for reconstruction? Just how accurately do the two reconstruction methods correlate? How dependent is reconstruction error on distance? In relation to this, why was 10m chosen as a distance cut-off for measurements? What were the directionality criteria for accepting signals for measurement? These details are important given the difficulty of reconstructing position from a 3-microphone array.