SAVI’S WARBLER (Locustella luscinoides) - Locustelle luscinoïde
SUMMARY
A shy bird of the reedbeds, difficult to see and difficult to hear as it sounds like an insect.
So now things get a little tricky - here we have a bird that sounds more like an insect than some insects do !
Savi's Warbler is a summer resident from the end of March until the end of September (although some birds seem to leave as early as the end of July). It favours wetland areas with mature reed beds and so occurs only spasmodically across Switzerland. Good populations can be found in the wetlands on the south eastern shores of Lac Neuchâtel where I made my recordings.
Not an easy bird to see deep in the reeds, it is also well disguised by its song which is a rapid, mechanical buzzing noise just like that of a grasshopper.
It will sit at mid-level up a reed stalk making this sound with its beak open, slowly turning its head from side to side to project the sound around. This effect can be heard in the recording, when the open beak was pointing directly at the microphone the sound becomes very loud, so I am guessing that the throat, mouth and beak shape act as a sort of avian megaphone to amplify and project the call:
60 seconds of Savi’s Warbler singing, upper (blue) shows sound intensity, lower (grey) shows sound frequency.
This graph shows three sequences of singing, 0 - 10 secs, 21 - 37 secs and 42 secs to the end.
It has two parts. The upper (blue) part shows the intensity of the sound in the recording against time in seconds; the further the blue area extends from the centre line the louder is the sound entering the microphone. You can see how the blue area gets wider and narrower as the bird turns its head towards and away from my microphone. The lower (grey) part shows the sound frequency against time; the dominant frequency is around 4 Khz.
To better understand what is going on here I have taken a "slice" of the analysis at 52 seconds (the red oval) and magnified it in the diagram below:
0.2 seconds of Savi’s Warbler song.
This graph shows what was happening from 52.0 seconds to 52.2 seconds in the recording - i.e. the total time shown across the bottom axis is only one-fifth of a second. During that brief snatch in time, if we count the blue "beats", we can tell that the bird made about 20 separate "clicks" - this is about 100 per second, or 6,000 in a minute. In the recording you just listened to, the longest call sequence lasted 34 seconds so the bird must have made about 3,400 separate clicks just in that sample alone. Extrapolating that to 12 hours of singing per day for a 2-month breeding season just makes my head hurt !
There is something else interesting going on in this diagram however.
You can see that in each one-tenth of a second there are two "beats" of sound, the first being a bit shorter in duration and less intense, the second being a bit louder and longer. Next, if you look at the frequency diagram below each beat the first sound seems to have a wider range of frequencies in it - 3 Khz to about 6.5 Khz, whilst the second is from about 3.2 Khz to 5 Khz. So it seems as though there may possibly be two separate sounds or sound sources here.
Diagram of a bird syrinx (from www.wysinfo.com)
Now, humans have one "voice box" - the larynx (Adam’s apple), situated above the point where the two bronchii (the tubes that connect our lungs to our mouth) meet, and we push air from our lungs over vocal chords in the larynx to make human speech. But birds have two voice boxes in a complex structure called the syrinx, and this is situated just below where the two bronchii meet and so can receive two independent air flows. They don’t have vocal chords but complex tympanic membranes, and a structure called the pessulus, which vibrate as air passes over them. The two parts of the syrinx have independent sets of muscles and nervous connections to the brain. Hence two voice boxes which can be operated independently. This "two-voice" structure accounts for a lot of the complexity in bird song. It has also been shown that the two syrinxes can play different roles in the final sound that comes out.
So, to get back to our bird, what I think could be happening in the diagram is that the Savi's Warbler first produces a sound from one side of the syrinx and then from the other, and by working them in alternation it can produce these incredibly rapid vibrating sounds. Impossible to know for sure without some complex monitoring apparatus.
But of course producing the sound is only one half of communication in birds - hearing and interpreting it is the other half. We know that birds have much better hearing acuity than humans, being able to discern both frequencies and speeds much better than we can. So whilst to our pathetic human ears this bird is simply making a buzzing sound like this:
the chances are very high that another Savi's Warbler is hearing much more detail. If we slow down that area around 50 secs above by a factor of 10 (which lowers the pitch by the same amount) it may be that another Savi's Warbler may be hearing something like this where the double beats are a bit clearer to us:
But here we can still only take a guess.
So why "Savi's Warbler" and not "buzzing bird" ? Well it is a member of the warbler family and is named after an Italian geologist and ornithologist, Paolo Savi, Professor of Natural History at Pisa University, who first described the bird in 1824 !
