Measures of Phototaxis and Movement Detection in The Larval Salamander
Department of Psychology, Indiana University of Pennsylvania, Indiana, Pa 15705
and School Of Optometry, Indiana University,
Bloomington, In 47405
Adapted from an article in Physiology & Behavior 50:645-647
key words: phototaxis; movement detection; time-lapse; infrred computerized monitoring; Ambystoma; salamanders; axolotl
Web contact pietsch@indiana.edu
ABSTRACT
In this investigation we describe two precise tests of visual function that integrate quasi-natural
situations with time-lapse video recording and infrared computerized monitoring of activity to assess movement detection and phototaxic tendencies, respectively. Four groups of larvae from A. punctatum, A. tigrinum, A. mexicanum (wild type)
, and a mutant albino axolotl were tested in an alley containing light and dark halves and lined with infrared sensors to monitor their phototaxic response. Wild type A. mexicanum
showed no phototaxic tendency while the other three groups displayed a strong negative phototaxic response. Enucleation of the eyes in mutant albinos eliminated the negative phototaxis.
Visual detection of motion was tested by video recording
the behavior of A. punctatum, A. mexicanum, and the mutant albino axolotl larvae while they explored a large bowl with 6 small vials on the perimeter, one of which contained white worms. A. punctatum
rapidly approached the worm vial and engaged in intense predatory behavior. A. mexicanum
responded to the presence of worms very slowly and rapidly lost interest. Albino axolotls displayed no visual recognition of the worms. The results indicate that visual function can be precisely determined in larval salamanders utilizing behavioral measu
res consistent with the animal's natural tendencies.
INTRODUCTION
The larval salamander offers the physiologically oriented animal
behaviorist favorable opportunities to explore relationships between
behavior and the organization of the nervous system. For example, the
transplantablity of the animal's eye has facilitated investigations into the
effects of altered visual input on avoidance behavior and on other optically
driven reactions (4-9). At the same time, a paucity of reliable and
expedient behavioral measures creates an obstacle to the full realization
of the larval salamander's potentialities as a subject in behavioral
experimentation. In part the problem is a function of the animal's
aquatic lifestyle. In the laboratory aquarium the animals display long
quiescent periods interspersed with short bursts of intense activity,
either spontaneously or reactively. The animals generally habituate rapidly to
almost any perturbation, such as a repetitively tapping the bowl or repeated
touching of the tail. One of our interests has been the time course of
return in visual function following the transplantation of an eye. Our need for
objective measures of visual function, has led to the development of
a computerized infrared monitor to assess phototaxic responses (3) and
the utilization of a time lapse video monitor to capture an animal's
response to movement of a worm (7). Both of the measures, described
below, permit compression of long observation periods into very short blocks
of time without a loss of essential information about the behavior in
question. While these tests were designed for larval salamanders, we
have found them to be appropriate and reliable with other small aquatic
animals such as fishes.
METHOD
Subjects
Phototaxic responses were tested on larvae from A. tigrinum (n=27) and A.
punctatum (n=27) obtained from the field as embryos and A. mexicanum (the
axolotl) (n=30) and a mutant albino of the A. mexicanum (n=30) species
obtained from the Axolotl Colony, Indiana University. The worm detection
test was run on A. punctatum (n=16), A. mexicanum (n=11) and the
mutant albino axolotl (n=20). Twelve of the albino axolotls in the phototaxic
test group were subjected to bilateral enucleation (eyeless) and retested to
assess the effect on the phototaxic response. Five additional albinos served
as anesthetized unoperated controls. In addition, a small number of goldfish,
Carassuis auratus, with and without eyes were used in confidence tests of
phototaxis that will be described in the results. All surgical or
otherwise nociceptive procedures were carried out with animals under
narcosis in 1:5000 MM 222. Operations were conducted under a stereoscopic
microscope, using methods described elsewhere (4,5,6). All of the larvae
were housed individually in 10% fresh Holtfreter's solution and were fed
daily on newly hatched brine shrimp supplemented every other day with
6 white worms Enchytraeus chytra. The larvae were between 25 and 35 mm at
the beginning of testing and were deprived of food 24 hours prior to movement
detection testing.
Apparatus
Phototaxic responses were assessed with a computerized infrared monitor.
The design of the electronic circuit for the infrared sensors, the
computer interface of the monitoring apparatus and the computer program
are described in detail elsewhere (3). The basic monitor consisted of an
alley 18 in. long and 2 in. wide containing black Plexiglas on one half and
white on the other. The walls of the alley contained 17 infrared sensors and
emitters placed equidistant the length of the alley. Half of the alley was
covered and remained in the dark while the other half was illuminated at
26.6 candle/ft square. The infrared sensors were collimated in 1/4 in.
brass tubes 3/4 in. long in order to restrict the field of view of both and
eliminate the response of a sensor to ambient light. This design also limited
the response of the sensor to the emitter directly across from it. The
software used to acquire data from the infrared sensors was written in the
ASYST language (Copyright 1984-1989 ASYST Software Technologies, Inc.).
All animals included in this study were monitored for a period of two
hours each. An animal was placed in the center of the alley, and an IBM
PC XT computer scanned the 17 infrared sensors every 120 ms for a total of
60000 scans. If an animal blocked a sensor on any scan it was counted,
and if it was between sensors the count was attributed to the previously broken
sensor. In this fashion the animal could be located at any point every 120 ms
and the counts cumulated to provide information on the total amount of
time spent at any one of the 17 points between one end of the alley and
the other during the 2 hr sampling period. Movement detection of
worms was assessed with a Panasonic nv-8050 time-lapse video recorder and a
Sony b/w camera. The video display field contained a 20 cm diameter bowl with 6
glass liquid scintillation vials placed equidistantly 6 mm from the wall
around the perimeter of the bowl. Lines drawn on the background beneath the
bowl identified 6 equal sized pie-shaped sectors by number. During a test
period one of the vials, randomly selected, would contain 6 white worms
Enchytraeus chytra obtained from a laboratory stock. Every animal was
introduced into the center of the bowl and videotaped for 2 hrs on a 48 hr
time-lapse mode that permitted playback in real time of 1 hr in 87 sec.
Latency to approach and circling a vial while displaying repeated striking
movements at the unattainable worms was scored during fast-time playback with
a cumulative millisecond timer and converted to real time at a ratio of
87/3600.
RESULTS
Fig. 1 contains a computer generated bar graph of the position of a single
animal during a two hour test period in the infrared monitor (each cell count
is equivalent to 120ms). This is a typical printout from an animal that
possesses a strong negative phototaxic bias. Fig. 2 contains a bar graph
showing the time spent in the dark for four different groups of larvae.
A.mexicanum showed no strong bias toward light or dark, while the A. tigrinum,
A. punctatum, and the mutant albino axolotl displayed strong negative
phototaxic tendencies. An ANOVA comparing the four groups yielded an
F(3,110)=72.4, P<.00001. Multiple comparisons indicated that A.
mexicanum varied significantly (P<.01) from the other three groups. A
comparison of the pre- and post-enucleation test scores of the axolotl albinos
yielded mean dark times of 111.01 +/-3.85 and 58.37 +/-3.98, respectively. The
difference proved to be highly significant, F(1,18)= 90.28, P<.0001.
Mean dark times for controls were 101.7 +/-3.68 and 107.98 +/- 3.56. The
difference was not significant, F(1,8)=1.51, P<.25.
Fig. 3 contains a series of freeze frames illustrating the prey behavior of
a salamander at a vial containing the white worms. The animal will typically
approach the vial, hover and make constant attack movements. After a period
of time the animal leaves the vial, makes several brief returns and then
totally habituates as evidenced by no further interest in the vial for the
remainder of the two hour test period. Latencies to approach the worm vial
and the total length of time at the vial engaging in attack behavior for A.
punctatum and A. mexicanum were (latency) 45.9 sec (hovering) 18.5 min and
(latency) 124.2 sec (hovering) 1.92 min, respectively. ANOVAS yielded an
F(1,25) = 30.28, P<.0001 for latency and an F(1,25) = 123.38,
P<.0001 for hovering. None of the albino mutant axolotls showed any
response to the worm vial and were not included in the statistical
comparisons.
Three eyeless and 6 one-eyed goldfish were tested in the infrared monitor.
The pre-surgical mean dark time and standard error for the one-eyed group
was 106.7 +/-2.5 min and post- surgical was 107.1 +/-3.5 min. The
eyeless group had a pre- surgical mean dark time of 109.4 +/-2.9 min and a
post-surgical mean of 54.8 +/-3.7 min. The lack of overlap in the eyeless group
makes further statistical analyses unneccessary.
DISCUSSION
All of the larvae tested in this experiment, with the exception of
the pigmented axolotl, possessed a negative phototaxic response at the
illumination level employed. The phototaxic responses of salamanders have
been studied primarily in their natural habitat, and while there are some
variations, many appear to be negatively phototaxic (2). Bilateral
enucleation in the albino axolotl destroys the strong negative phototaxis.
Previous testing in our laboratory has indicated that A. opacum larvae and
goldfish also possess a negative phototaxic response under similar
conditions, and extirpation of the eyes destroys the negative phototaxis.
Therefore, it seems highly unlikely that extra-optic photoreception plays
any role in the phototaxic response even though it may influence
circadian rhythms salamander larvae (1).
In addition, preliminary observations with a small number of larvae from A.
punctatum and the mutant albino axolotl indicate that the negative phototaxis
can be eliminated after severing the optic nerves and is reestablished
within two weeks, a time period that would permit only partial regeneration
of the optic connections (10).
There are significant interspecies differences in the response to the
vial containing worms in our measure of movement detection. Mutant albino
larvae, in the absence of all but visual cues, do not respond to the worms.
It is unclear whether their unresponsiveness is due to a lack of detectors in
the retina, an inability to process the information centrally, or to
both central and peripheral deficiencies. We are currently examining that
question. The A. punctatum larvae engage in a rapid and vigorous display
of hovering and attacking the inaccessible worms for approximately 15 to
18 min. This is followed by a marked decline of the response to the worm vial
for the remainder of the two hour test period. The A. mexicanum by comparison
are slow to respond and habituate very rapidly to their inability to capture
their prey.
Phototaxic response and movement detection in salamander larvae have not
been studied extensively in the laboratory. Schneider (8) examined
phototaxis employing a forced choice approach in a T-maze, a choice which
may have been induced by fear. Stone (10) assessed the return of visual
function after transplantation using rubber worm models waved over the bowl
of larvae with transplanted eyes. The two tests we have employed to assess
visual function have a number of advantages over those previously
mentioned. Both the computerized monitor and the time- lapse video permit the
observation, compression, presentation and storage of large amounts of data.
Thus the investigator can avoid having his/her patience and skills as an
observer tested in what would otherwise be a formidable undertaking. A
high degree of precision and reliability can be achieved since the events
are available to as many observers as desired. Data from both sources can be
easily quantified beyond the yes/no level, increasing the number of variables
that can be examined simultaneously. In both tests it is possible to carry
out minute pattern analyses of movement. For example, we can determine
the location of the animal at the dark end of the alley, the amount of time
spent at any point and the frequency of movement. Perhaps the most
valuable aspects of our methods are the exclusion of the experimenter
as a variable and the utilization of quasi-natural situations i.e., live
worms rather than worm models that cannot exhibit normal movements and
free movement and choices in the infrared monitor. Finally, it is worth noting
that we have tested other aquatic creatures, such as fishes, with equal
success, and the infrared monitor and time-lapse system could be easily
modified to monitor the behavior of many terrestrial creatures.
REFERENCES
1. Adler, K. Extraoptic phase shifting of circadian locomotor rhythm in
salamanders. Science, 164:1290-1292, 1969.
2. Duellman, W.E.; Trueb, L. Biology of Amphibians. New York, McGraw-Hill,
1986.
3.Kirkpatrick, T.; Schneider, C.W.; Pavloski, R. A computerized infrared
monitor for following movement in aquatic animals. Behav. Res. Meth. Instrum. Comput. 23: 16-22, 1991.
4. Pietsch, P.; Schneider, C.W. Vision and the camouflage reactions of
Ambystoma larvae: The effects of eye transplants and brain lesions. Brain Res.
340: 37-60, 1985.
5. Pietsch, P.; Schneider, C.W. Transplanted eyes of foreign donors can
reinstate the optically activated skin camouflage reactions in bilaterally
enucleated salamanders (Ambystoma). Brain Behav. Evol. 32:364-370, 1988.
6. Pietsch, P.; Schneider, C.W. Two-eyed salamanders: Does binocularity enhance
the optically evoked skin blanching reactions of Ambystoma larvae? Physiol.
Behav. 48:357-359,1990.
7.Pietsch, P.; Schneider, C.W. Anterior decerebration blocks visual
habituation in the larval salamander (Ambystoma punctatum). Brain Res. Bull.
25:613-615, 1990.
8. Schneider, C.W. Avoidance learning and the response tendencies of the
larval salamander, Ambystoma punctatum, to photic stimulation. Animal Behav.
16:492-495, 1968.
9. Schneider, C.W.; Pietsch, P. The effects of addition and subtraction of
eyes on learning in salamander larvae. Brain Res. 8:271-280, 1968.
10. Stone, L. S.. Heteroplastic transplantation of eyes between the larvae
of two species of Ambystoma. J. Exp. Zool. 55:193-261,1930.
Fig. 1. Computer generated bar graph illustrating the number of times an
infrared sensor was activated by the presence of an animal.
Fig. 2. Mean time in the dark for four species of Ambystoma larvae: albino axolotl
(ald), A. punctatum (punc), A.tigrinum (tig), and A. mexicanum (mex). S.E. (standard error) is indicated at the top of each bar. The scale on the left extends from 0 to 110 minutes, each graduation representing 10 minutes.
Fig. 3. Series of freeze frame pictures illustrating the hovering and attack
behavior of a larvae in the presence of a vial containing worms; the worm vial 4 in this series is at the six-o'clock position in the finger bowl. From left to right: upper row, A, B, C; lower row, D, E. F. F is enlarged in fig. 4.
Fig. 4. Blow-up of F from Fig. 3. Note arrow.
Carl W. Schneider is a Professor Emeritus at Indiana University of Pennsylvania, Indiana, Pennsyvania.
go here for the Eyes and Eyes main menu.
Web contact:pietsch@indiana.edu