Page 50 - Spring 2019
P. 50

HEP“-IHB and Eiflfifinfii‘ Voluntary control of adaptive signal
IDI .
avoided that tail rest pole. He moved his tail up, down, right, :.~§‘ ‘ ' .
and lefta always trying to not have that “thing” touch his tail. By 5 g ( ' ‘so-mo inc
systematic and precise reinforcement of small tail movements, E ' ,. B d me
however, Heptuna finally touched the device with his tail. 3 3‘ ' ma
ee .
. . . . . <_: he. ran
With Heptuna finally positioned correctly, it was possible to 6 ~ he
start the detection training. Whereas the stationing training 1, -'
took weeks, Heptuna was 100% in detection performance in 1 ._—.—.—.—.—.—.—,—.—.—,
just one 100-trial session! To capture outgoing clicks, Mari- 0 an “ " ::Ee,',:,,:::°_,:," ‘n “B M
on Ceruti, a colleague, and W'hit developed a computerized
Syslem lllel Could eelelyze llle eelloleeellml ellek ml“ lllel Figurzs. Atmin afl01 eeheieeetien clicksthatHept1muemittedm the
Hepmne emlllell eompullng lmlll llle overall Peale level and echn detectianphuse nfthe experiment. Each hnriznntulline. smrtingat
Peale frequency elllle emllled ellelee wlllle delng a reel mgel the irettern nfthe figure, 15 im emitted click. The darker the line calms.
lleleellon leek (Cemll end All l983)_ the greater the energy ucmss the frequency band. The eiiek train begins
with Heptimu emitting narrewhund. law-frequency clicks with majnr
D“’lh8 e ‘elele e e°’“P“‘e' ’“°hl‘°'eel l“leP‘“hee °“‘3°lh3 energy in the 30, ta 60—kHz) regian. As the eiiek train eVnIVes(urmmd
ellel“ e“‘l W°“l‘l elel‘ ‘he e"Pe’l‘heh‘e’ ll l“leP‘“he me‘ ‘he eiiek 12) Heptima adds energy in the higher frequencies (at 120 kHz)
e'l‘e'l°h °‘ el‘he' high °’ l°‘” 5°"_’ee le"el °’ high °’ l°W emitting lzimndal energy clicks. The click truin develnps arnund eiiekzo
Peel‘ "e‘1“eheY ehel “'he‘he' ‘he “Shel Wee e°”ee“ Wheh with Heptimu pmducing very widebund clicks with energy ucmss the
‘he e°‘hP“‘e' 5°“h‘le‘l 3‘ hl3h'l’e‘1“eheY P“’e ‘°hev l“leP‘“he frequency spectrum (30 tr no kHz). The click truin ends with Heptimu
‘”°“l‘l eml‘ l°“el elleles el’°"e ‘he e’l‘e’l°h* ehel Wheh ‘he shifting (uraimd click its) ta click: with numiwimnd energy arms: the
‘°’“P“‘e' 5°““‘le‘l “ l°“'e' "e‘1“e“‘Y ‘°“e' he W°“l‘l keel’ 55— tr 110-kHz band. This click truin Iusted just ufiew seeends. Frnm
his clicks below the level criterion. The experimenters also MW"? amipuwlaski (1990)
established a frequency criterion, and when the computer
sounded a fast-pulsed tone, Heptuna was to keep his peak
frequency above a fixed frequency, whereas when the pulses false starts, the bag size, suspension apparatus, and Heptuna
were slow, he kept his peak frequency below a fixed criterion. were under control. The results did not, however, support the
After intensive training. the experimenters managed to de- idea of prey stunning by dolphin clicks (Marten et al., 1988).
velop stimulus control over Heptuna’s click emissions. As a During this set of Experiments, Hepmm had Excdlem Com
full demonstration that Heptuna had learned this complex . .
. . _ _ trol of his head placement, and Whit wanted to take advantage
behavior, mixed tones and pulse rates signaled him to pro- _ , _ _ _ _ _ _
. _ _ of the animals stationing to refine his vertical emission beam
duce high-level, low-frequency clicks and vice versa. Heptu- , _ _ _
. . pattern measurements. Heptunas positioning was a level of
na had learned to change his emitted level and peak frequen- _ . . , .
. . . . improvement in accuracy over W'hits first emitted-beam mea-
cy during an echolocation detection trial and demonstrated _ .
. . . . , surements. For this experiment, the control computer would
conscious control of his echolocation clicks (Figure 5; Moore _ . .

. signal Heptuna to ecliolocate the target (1.3-centimeter-diam-
and Pawloskl’ 1990)' eter solid steel sphere located 6 4 meters in front of the bite
Because Heptuna could produce high source level clicks, plate) and report whether it was present or absent. W'hit used
ab0V€ 200 dB 1'2 1 |lPi\ (at 1 meter). Ken Norris. 0112 Of the six vertical hydrophones to measure Heptuna’s emitted beam
great pioneers of dolphin echolocation studies, thought that for each click emitted. Whit computed Heptuna’s composite
Heptuna could test the prey-stunning theory that he and beam pattern over 2,111 beam measurements and showed that
Bertel Mohl (see the article about Mohl in Acoustics Today thevertical beam was elewted by 5" above the line ofHeptuna’s
by Wahlberg and Au. 2018) had been developing. The hv- teeth. Whit then calculated the vertical beam to be 15° differ-
pothesis was that with their very high intensity clicks, dol- ent from his first measurements. He considered this difference
phins could stun potential prey, making capture much easier. to be attributable to both differences in head anatomy and bet-
Thus, began a truly exciting experiment involving Heptuna, ter control over stationing by Heptuna (Au et al., 1986a).
fish in plastic bags, and suspension devices to hold the bags
in front of the ani.rnal as he produced very high source level Haptuna and "Jawphunaa"
clicks. Bags burst because of bad suspension, sending fresh William (Bill) Evans, a graduate student of Dr. Kenneth Nor-
fish swi.rnming away, with Heptuna giving chase. After many ris, used contact hydrophones in suction cups to measure
43 1 Acuulclcl Tbday 1 Spring 2019







































   48   49   50   51   52