Thursday, October 22, 2009

Simulations Galore

Virtual unreality

Regardless of all the intrigue and discussion about the double knocks I recorded on March 18, the event did not have a major impact on our strategy for the rest of the season. If they were real double knocks that occurred in response to the simulator, what did this mean? It meant that the MIMDKWTFII had executed a series of double knocks in response to a loud anthropogenic banging noise in the woods between Rhodes Lake and Willow Flat. Well, that would make the third instance of a possible double knock response to a loud human sound at Moss Island, and the second one consisting of a series of double knocks within the same small area. It would make the sixth (I think; may have miscounted) occurrence overall of double knock(s) in this area. Nothing new there; just reinforcement of the previous patterns and strategies. The one thing I did feel it provided was some measure of vindication, both at the personal level and the project level, that I and we were in fact accurately distinguishing the "real" candidate double knocks from the background sounds. But the goal remained the same and remained unattained: Get a visual on the mystery double knocker, whatever it is.

It did perhaps suggest that a targeted cavity survey within this area might be worthwhile. Given the early time of the 3/18 double knocks (4 minutes before sunrise) if it was in fact a cavity roosting bird making these sounds then it might have a roost cavity nearby. Scott and Melinda were not available in the immediate future for more field work, and I was the least experienced (and skilled) cavity searcher of the crowd. I have generally considered shotgun cavity searches a waste of time and so have never really participated in them. In late March I did try my hand at it in the area from which the 3/18 double knocks seemed to have emanated. After a day I decided I was a lousy cavity searcher and dropped it, returning to the earlier approach of just sitting quietly with camera at the ready. Of course, I saw and heard nothing. Three weeks had elapsed between the last two possible MIMDKWFTII detections in the area, no reason to expect I'd get lucky any quicker this time.

In early April, Scott and Melinda were able to get back in the field. We concentrated on the same area, essentially bounded by Rhodes Lake Road on the south, a line drawn directly north from the Turnaround on the west, Willow Flat on the north and east, and the ATV trail that crosses the Rhodes Lake sill on the southeast. Melinda and I stationed ourselves variously within this area while Scott moved through it doing a double knock simulation series every hour or half hour. After Scott had to leave, Melinda and I remained an additional day and a half with me operating the simulator while she stationed herself aboard kayak in Rhodes Lake. All told we put in three days of this routine, covering April 7-9th.

After the confusion around the 3/18 event, we instituted some improvements to our field protocol for when the double knock simulator was in use. We made sure that all listeners were in radio contact with the operator of the simulator, and he would announce one minute and 10 seconds before the beginning of each series. I always turned my video camera on when each simuation began, and kept it running for about 10 minutes afterwards. When I was the operator, I made a point of getting a time stamp on the video as well. The communication, time stamping, and recording of the simulations should be adequate to resolve any future uncertainties about whether a sound was simulated or not.

As usual, however, all this cross checking and documenting was never put to the test, because none of the three of us heard, saw, or recorded anything worthy of note during this time. It did yield an abundance of recordings of simulated double knocks from various distances, however, as well as a more personal experience with how well the sound can be heard. We also got much more skilled at operating the simulator; Melinda commented that now the majority of the simulated double knocks generated by either Scott or me sounded like good approximations to the real Campephilus sound; earlier it had been a small minority. I noted that, though it was sometimes faintly audible at 1000m, most of the time the simulator was difficult or impossible to hear beyond about 500m. Leafout was under way as well, which was doubtless affecting the propagation of sound.

Here are three sets of simulated double knocks recorded from increasing distances. These are all from after we got practiced at using the device so they are much more consistent in timing, loudness, and quality than what we had going back in March. In each case I have edited the recording to place the double knocks at 1 second intervals. First, a set by me from only a few meters away (click the sonogram for the full image):

You can also see that the double knock simulator is hardly a high-tech device; it's a plywood box, open on two sides, that you tie to the tree and whack with a couple of wooden rods. Looking at the sonogram, beyond the overall loudness note the large amount of accoustic energy up in the high frequencies as well as the long trailing echos. The simulated knocks are almost like mini gunshots. Also note (this will come up later) how sharp, clean, and perfectly vertical the line is that marks the beginning of each double knock.

Next, a series by Scott from about 200m away:

Though to the ear they still sound fairly clean and doubled, the sonograms show a great deal of "mushiness." In some cases the individual knocks are not even clearly resolved from each other. Most of the higher frequency sound has faded away to near invisibility. Another feature that shows here is that in most cases (I see it best on #2, 3, and 5) the second knock is missing the lowest tones, being concentrated more in the upper part of the range of pitch that has survived the attenuation at this distance.

Finally, a set by Scott from about 500m away:

The double knocks are still audible, though the doubleness is not always so distinct. On the sonogram they are nearly gone, however. There are just faint indistinct smudges at about the right frequency; the crisp onset of the sound and the separation between the two knocks are both pretty well obliterated.

After this excursion in early April, I was mostly on my own for the rest of 2009. Melinda and Scott had other business to attend to as the nesting season approached; my own field time reduced as well. I did not use the simulator in the field much anymore. After our saturation effort in the "hot zone" yielded nothing further we decided to give it a rest. As Melinda said when she and I were packing it in after the last morning of whacking on trees, "Elvis has left the forest." Or at least he seemed to be done with his performances for the time being.

For the rest of April and in to May, I focused on spending more time in the few areas adjoining the "hot zone" where our effort had been lacking. These were places of difficult access along the northern and eastern fringes of the WMA, close to the Obion. At the very least I wanted to be sure I had seen the whole place and spent some token effort within earshot of every hectare.

Other posts in this series:


At 6:50 AM, Blogger fangsheath said...

Cavity inventories quickly become impractical as leafout proceeds. Even partial leafout greatly obscures cavities (e.g., late Feb in southern La.). Also, subjective impressions of entrance size are highly influenced by cavity elevation. In my opinion, measurement is essential (and easily accomplished remotely with camera and rangefinder). It does require some man-hours to cover significant areas though.


Post a Comment

<< Home

Site Meter