Observations on Big Brown Bats Eptesicus fuscus, 16 March 2005, Rocheport, Missouri, USA

On 16 March 2005, Kim Livengood and I walked along the Katy Trail south of Rocheport, about 30 km west of Columbia, MO, looking for bats in the hour and a half after sunset. It was still, with no clouds and a bright moon a little less than half illuminated. It was coldish - about 5 degrees C. The Katy Trail runs along the edge of the Missouri River in this area, following a disused railway. At frequent intervals along the trail, it passes right beside substantial bluffs - vertical cliffs where the river has cut through a series of parallel ridges. These bluffs are typically about 50 m high and 100 m wide, and they face west adjacent to the river, so they can be expected to get a lot of exposure to the afternoon sun. They seemed ideal places to look for bats early in the season when conditions are still generally cold.

At every bluff, there was a lot of activity from Big Brown Bats Eptesicus fuscus. What struck me most was the frequent detection of calls which sounded quite unlike typical Big Brown Bat echolocation calls. From an Anabat detector, these calls sounded like brief, explosive trills  - quite unlike search phase echolocation calls in all their immense variability, and also quite different to feeding buzzes. These trills were heard at every bluff we visited, and were being heard almost the whole time we were detecting Big Brown Bats.

I have heard these calls before, but very rarely. I have sometimes seen brief sequences of calls from Big Brown Bats which I thought showed some analogy to the so-called directive calls of Pallid Bats Antrozous pallidus. More recently, I have seen several  examples of these trills in a dataset collected from a passive detector set up on the edge of the Meramec River at Meramec State Park, also in Missouri. This monitoring station has been running since late November 2004, and there have been occasional examples of trills in late November, late December, January and more commonly in February and March.

In addition to the trills, we noticed behaviors which I have not seen before in this species. At several bluffs, I saw Big Brown Bats circling repeatedly in front of the cliffs, their trajectories taking them perhaps 10 m from the cliff at the farthest point and much less than a metre at the closest approach. On at least 6 occasions, we saw a bat land on the rock surface. Most of these landings were brief, lasting about one second. On one occasion, the bat then scurried for a few cm before flying off. On two occasions, the bats scurried several cm up into crevices in the rocks and were lost from sight. I don't know to what extent these behaviors might have been influenced by my spotlight.

Trills were recorded from some of the circling bats, and they were also recorded from both bats of a pair involved in a chase - where one bat followed closely behind the other for about 30 seconds. In most situations though, it was difficult to be sure a trill came from a particular bat, because there were nearly always more than one bat present at a time and there was no obvious behavior associated with the trill which would make it clear where the sound originated.

Descriptions of the trills

Each trill consisted of two phases, the first consisting of three to five calls of relatively low slope and fairly consistent frequency. The second phase is more like a feeding buzz, with the times between pulses shorter than in the first phase, but more consistent than in a feeding buzz, without the tendency to tighten up to very short intervals towards the end. The second phase was generally hard to detect and often not seen at all. The following description therefore applies to the first phase only.

The first phase pulses varied considerably in both frequency and shape with some calls much steeper than others. Characteristic Frequency (Fc) values varied from about 19 to 45 kHz, but this is confused by the fact that some of these were detected on the first harmonic while others were principally detected on the second. It seems clear that calls with Fc around 20 kHz are on the first harmonic, while calls with Fc around 40 kHz are on the second harmonic, but it is less clear what is happening with some calls which have Fc about 30 kHz. There is one instance (shown below) of calls showing third harmonics, and quite likely another showing fourth harmonics. Times between pulses varied from 28 to 42 milliseconds, but most were between 30 and 35 ms. Durations varied from 4.2 to 11.3 ms, but most were between 7 and 10 ms, with shorter values at least mostly due to missed parts of calls. Numbers of pulses in the first phase varied from 3 to 5, but most showed 4 or 5 and it is quite possible that the cases of three were due to calls being missed because the bat was too far away. In one file, I recorded a total of 11 trills, five from one bat and 6 from another bat, all in a period of just over 9 seconds. 

Below I provide seven examples of Big Brown Bat call sequences, the first two of typical search-phase and hunting sequences and the later five of trills. All these sequences were recorded on the Katy Trail on 16 March 2005. All are displayed at a time resolution of 25 milliseconds per 40 pixels, which corresponds to F7 in Analook. The frequency scales are all the same, and logarithmic. Note that I use the terms 'call' and 'pulse' as synonymous. 

Fig 1) Typical Eptesicus fuscus low-clutter search-phase, approach phase and feeding buzz, followed by more search-phase calls. These are the sorts of calls you would normally see when the bat is hunting in an open situation, eg. over the forest canopy or along a wide roadway. This sequence has been compressed, so that all the time between calls has been removed, allowing more calls to be seen at a useful resolution.

Fig 2) Typical search-phase calls of Eptesicus fuscus when flying in moderate clutter. These types of calls would be seen when the bat was flying closer to trees or other objects than in Fig 1.  This sequence has also been compressed. The time and frequency scales are the same as for Fig 1.

Fig 3) A trill. Note the four relatively long duration pulses in the first phase of the trill, followed by seven pulses of much shorter duration, with the last pulse of the second phase revealed by only a single dot on the screen. This and the following sequences have not been compressed, so the time scale accurately represents the time between pulses as well as the time within a pulse. The times between pulses are much less than for normal search-phase echolocation calls. Unlike feeding buzzes, there is no lead-in through an attack phase, and the first phase of the trill consists of relatively similar pulses, not becoming progressively shorter in duration as in a typical attack phase.

Fig 4) A second trill, showing another manifestation where the frequency is much lower. This sequence shows pulses where the lowest slopes occur just below 20 kHz, an octave lower  than the equivalent parts of the calls in Fig 3. This suggests that Fig 3 might be showing pulses on their second harmonics, while this sequence shows pulses on their first harmonics. If so, then the differences in frequency would be just a result of viewing different harmonics, and the pulses of the two sequences might actually be very similar. The lower tails visible in Fig 3 could still be present but unseen, because Anabat detectors with standard microphones have a reduced sensitivity below 20 kHz. If tails the same as in Fig 3 were present, they would have dropped to below 14 kHz on the first harmonic.

Fig 5) Another trill, showing first-phase pulses of much lower slope than in the previous examples. Note how the final parts of most of these pulses were detected at frequencies double those of the earlier portions. In the first pulse, the frequency appears to drop from 49 kHz down to 27.5 kHz, then jump up to 55 kHz and then drop to 42 kHz. So two harmonics are being displayed, the higher being twice the frequency of the lower. This pattern suggests that the displayed harmonics are the first and second. If you took the first pulse in the sequence, and reduced the portion on the second harmonic down to the first harmonic, you can see how the shape of the call as a whole would much more closely fit the shape of the third pulse, where the second harmonic didn't appear for nearly so long.

Fig 6) Yet another trill, showing a general form similar to that in Fig 3, but displaying two harmonics, the higher being twice the frequency of the lower. However, in this case, the frequencies of the lower harmonic are much higher than in Fig 5. One possibility is that these are actually the second and fourth harmonics.

Fig 7) An extreme example in which three harmonics are clearly discernible. In the second pulse, the downsweep starts at 46 kHz and drops to 32 kHz on the first harmonic before jumping up to the second harmonic, which sweeps down from 58 to 40 kHz. It then jumps up again to the third harmonic at 60 kHz and drops to 55 kHz. The pulses in the second phase of this sequence are showing the first and second harmonics. Note that the frequencies present in these calls are quite similar to those in Figs 3, 4 and 6 if we assume that Figs 3 and 4 are on the second and first harmonics respectively, and that Fig 6 is showing the second and fourth harmonics. All these sequences would then have shown first-phase pulses with characteristic frequencies of about 20 kHz on the first harmonic, had it been visible. Although the highest harmonics being displayed in Fig 7 and Fig 6 are of quite similar frequencies, it must be appreciated that in Fig 6 the highest harmonic is showing the very bottom of the steep downsweep at the end of the call, while in Fig 7, the highest harmonic is showing during the flattest part of the call. These observations are quite consistent with Figs 6 and 7 showing four different harmonics between them.

What's it all about?

So what are these calls, and what is going on with all this social activity at this time of year? To me the whole performance looked like courtship displays. Is it feasible that Big Brown Bats could be mating in mid March? If so, it might be reasonable to regard these trills as songs. They resemble other bat songs I have seen, being stereotyped sequences of pulses that clearly differ from echolocation calls. Any ideas?

Back to: Bat Notes contents,  Anabat Contents,  Home