Visually guided behavior in salamanders

There is an ongoing effort to answer such questions as: how much information the salamander's retinal ganglion cells transmit to the brain, which information, and by what code. In addition to relating the spike train to the signal it encodes, one would like to relate the spike train to the information decoded from it .

Therefore it is of interest to know what aspects of the stimulus the salamander ultimately sees. Behavioral experiments place a lower bound on the information encoded in the spike trains. Any stimulus feature that the animal can discriminate (as demonstrated in behavioral experiments) must necessarily be encoded in the spike trains.

Summary of results of some published behavioral experiments

The behavioral assay

The best assay is the animal turning its head towards the stimulus, which is the first response in the prey-capture sequence. Terrestrial, aquatic, and larval animals all do this, and they do it at higher frequency and lower stimulus threshold than the subsequent prey-capture behaviors. The animals might detect, but fail to respond to, a stimulus the animal regards as not prey-like. Thus the thresholds could be estimated too conservatively. In these experiments the stimuli are projected onto a translucent paper or frosted glass on one side of the animal's tank. With the exception of the flicker, animals only respond to moving stimuli.

For results quoted, "minimum" and "maximum" values are the minimum or maximum that were shown to elicit a response above baseline; if no stimulus was used that was small or large enough that response was lost, the value is expressed as <= x or >= x. "Optimum" values are the values that elicited the strongest (highest frequency of) response. Except where indicated, several species from both Triturus and Salamandra were used, and if species varied, the range of results across species is shown.

Size: Minimum <= 1 deg, optimum 4-10 deg, maximum >= 20 deg

Four degrees of visual angle corresponds to a 5mm diameter spot 3cm away from the salamander. Data shown were measured only in young S. salamandra larvae, because animals that have already caught prey will only react to spots the size of the prey they've caught. When spot is larger than animal's head it turns towards the edge of the spot, i.e. it responds to the moving contour not the spot per se. The range of optima reflects a range of equal performance.

Contrast: Minimum 0.010 to 0.015, optimum >=0.1, no maximum.

Contrast is defined as (Is - Ib) / (Is + Ib) where Is is intensity of the stimulus and Ib the intensity of the background. The range of minima reflect the range obtained across species of salamanders.

Velocity: Minimum 0.5 to 1.0 deg/sec, optimum 5-20 deg/sec to 10-30 deg/sec, maximum 40 deg/sec to 110 deg/sec.

The range of minima, maxima, and optimal-range values indicate the considerable species differences. Generally the land salamanders were slower than the aquatic salamanders; larvae for either land or aquatic species had maximum velocity threshold of 60-70 deg/sec. (The retina experiments of the Meister lab are done on the aquatic larval stage of a land salamander).

Flicker: Minimum and maximum not reported, optimum 2-10 Hz.

I have not as yet looked up the experiment.

Acuity: (Minumum) 0.1 to 0.15 deg

Acuity was measured with drifting bar grating, visual angle given is width of 1 bar. Range shown applies to all species except that for the aquatic salamanders, this range applies in air only at close (3cm) distances; they are nearsighted in air due to lens optics.

Color: Minumum <= 457nm, optima are < 518 and > 555, maximum >= 635nm

Ability to respond to contour of color against equiluminant gray was found for all wavelengths tested: 457, 518, 555, 584, 635. Response was worst for greens (but may be due to prey selection criteria).

Relationship of species used

Salamanders and newts are urodeles (order Caudata) as opposed to the other two kinds of amphibians, anurans (frogs and toads) or apoda (wierd things). There are lots of families within Caudata, only a few of which are listed below. I give the family and genus, species if I know it, and common names. All the studies summarized here were done on Salamandridae and most involved comparisons of members of the Salamandra (terrestrial) and Triturus (aquatic) genii. The Meister lab retina experiments are done on larval tiger salamanders, and another favored retina preparation is mudpuppy.

Ambystomidae Ambystoma tigrinum (tiger salamander)

Ambystomidae Ambystoma mexicanum (axlotl)

Salamandridae Salamandra salamandra (fire salamander)

Salamandridae Triturus ________ (newts)

Proteidae Necturus __________ (mudpuppy)

References

(Specific locations refer to the Harvard library system.)

The Amphibian Visual System, ed. Fite, Academic Press, NY, 1976.

You can find this in the MCZ library QL 669.2 .A46. The data summarized above were gleaned from this book and from my attempt to read the primary sources for the data, which were mainly papers by Himstedt published in the 1960's, written in German.

Visual Behavior in Salamanders, ed. Gerhard Roth. Springer-Verlag 1987.

MCZ Library QL 668.C2 R68 1987. Contains most or all of the information in the previous, older reference and much more.

The Amphibian Visual System ed. Ingle, Goodale, Mansfield. MIT Press 1982.

Prey Selection in Salamanders

Neural Mechanisms of Prey Recognition: An example in amphibians, by Gerhard Roth, in Predator-Prey Relationships, ed. Feder and Lauder. U. Chicago Press 1986.

Discussion bridges from neuroethology to ecology.

Back to Pam Reinagel's home page
Back to Reinagel Lab