ArticlePDF Available

Visual adaptations in a diurnal rodent, Octodon degus


Abstract and Figures

The degu (Octodon degus) is a diurnal rodent, native to Chile. Basic features of vision and visual organization in this species were examined in a series of anatomical, electrophysiological and behavioral experiments. The lens of the degu eye selectively absorbs short-wavelength light and shows a progressive increase in optical density as a function of age. Electroretinograms recorded using a flicker-photometric procedure reveal three spectral mechanisms: a rod with peak sensitivity of about 500 nm and two types of cone having respective spectral peaks of about 362 nm and 507 nm. Opsin antibody labeling was used to determine the retinal distributions of the three receptor types. A total of about one-third of the approximately 9 million photoreceptors of the degu retina are cones with the two types (507 nm/362 nm) represented in a ratio of about 13:1. The contributions to vision of all three receptor types were examined in a series of behavioral experiments. A consistent feature of both the electrophysiological and behavioral results is that relatively high levels of light adaptation are required to effect the full transition from rod-based to cone-based vision. In behavioral tests degus were shown to be able to make color discriminations between ultraviolet and visible lights.
Content may be subject to copyright.
Visual adaptations in a diurnal rodent,
Octodon degus
Received: 19 December 2002 / Revised: 18 February 2003 / Accepted: 5 March 2003 / Published online: 5 April 2003
Ó Springer-Verlag 2003
Abstract The degu (Octodon degus) is a diurnal rodent,
native to Chile. Basic features of vision and visual or-
ganization in this species were examined in a series of
anatomical, electrophysiological and behavioral experi-
ments. The lens of the degu eye selectively absorbs short-
wavelength light and shows a progressive increase in
optical density as a function of age. Electroretinograms
recorded using a flicker-photometric procedure reveal
three spectral mechanisms: a rod with peak sensitivity of
about 500 nm and two types of cone having respective
spectral peaks of about 362 nm and 507 nm. Opsin an-
tibody labeling was used to determine the retinal dis-
tributions of the three receptor types. A total of about
one-third of the approximately 9 million photoreceptors
of the degu retina are cones with the two types (507 nm/
362 nm) represented in a ratio of about 13:1. The
contributions to visi on of all three receptor types were
examined in a series of behavioral experiments. A con-
sistent feature of both the electrophysiological and be-
havioral results is that relatively high levels of light
adaptation are required to effect the full transition from
rod-based to cone-based vision. In behavioral tests de-
gus were shown to be able to make color discriminations
between ultraviolet and visible lights.
Keywords Cones Æ Photopigments Æ Color vision Æ
Electroretinogram Æ Octodon degus Æ Ultraviolet
Abbreviations ERG electroretinogram Æ M middle-
wavelength-sensitive Æ S short-wavelength-sensitive Æ
UV ultraviolet
Rodents comprise the most abundant of the mammalian
orders, collectively occupying an unusually broad span of
terrestrial habitats. Basic features of vision and the visual
system have been studied for a number of species from
Sciurognathi, the larger of the two suborders of rodent
(Macdonald 2001). Animals from this group vary from
highly diurnal to highly noct urnal, and among them are a
number of common laboratory species. All appear to have
duplex retinas, but there is striking variation in the relative
representation of rods and cones that correlates with the
predominant visual lifestyle of the animal. For example,
cones comprise approximately 85% of all photoreceptors
in a strongly diurnal ground squirrel Spermophilus
beecheyi (Kryger et al. 1998), while making up only about
1% of the recep tors in the nocturnal rat Rattus norvegicus
(LaVail 1976; Szel and Rohlich 1992). Rods in all of these
rodents have peak sensitivity in the vicinity of 500 nm, but
there are significant variations in the nature of the cone
pigment complement. Three general cone patterns have
been documented in different rodents (Jacobs 1993;
Ahnelt and Kolb 2000). Some species have middle-
wavelength-sensitive (M) cones paired with short-wave-
length-sensitive (S) cones; others have M cones in addition
to cones with maximum absorption in the ultraviolet (UV)
wavelengths, while the retinas of still other species contain
a population of M cones but appear to lack cones con-
taining either S or UV pigments. In addition, there can
also be significant vari ations in the spatial distribution
and relative representation of these cone types and, in
some rodents, the two types of cone pigment may be
co-expressed in the same receptor (Szel et al. 1996, 2000).
In contrast to relatively large number of studies of
basic visual adaptations in sciurognathic rodents, there
is only limited information about rodents of the other
sub-order Hystricognathi. Rodents of this group are
distributed across portions of the Americas, Africa and
Asia, but are especially abundant in Central and South
America. They too are extremely diverse in visual
J Comp Physiol A (2003) 189: 347–361
DOI 10.1007/s00359-003-0408-0
G. H. Jacobs Æ J. B. Calderone Æ J. A. Fenwick
K. Krogh Æ G. A. Williams
G. H. Jacobs (&) Æ J. B. Calderone Æ J. A. Fenwick
K. Krogh Æ G. A. Williams
Neuroscience Research Institute and Department of Psychology,
University of California, Santa Barbara, CA 93106, USA
Fax: +1-805-8932005
lifestyles spanning the gamut from nocturnal to diurnal
and including a number of fossorial spec ies (Macdonald
2001). Because of its status as a common laboratory
animal, visual adaptations have been examined for the
guinea pig (Cavia porcellus), one species from this sub-
order. These animals, described as showing crepuscular
activity rhythms in their native habitats (King 1956),
have a duplex retina containing a significant number of
cones (Peichl and Gonzalez-Soriano 1994). The guinea
pig has two spectral types of cone (Jacobs and Deegan
1994), S and M, and about 10% of all these cones may
co-express the two photopigments to some degree (Parry
and Bowmaker 2002). As noted, both of these arrange-
ments have also been seen in sciurognathic species. An
additional species from this group, the degu (Octodon
degus), has recently emerged as a popular laboratory
subject for studies of the photic control of activity
rhythms (e.g., Lee and Labyak 1997; Jiao et al. 1999).
This rodent, native to Chile, is described as diurnal
(Fulk 1976) in its natural environment, although labo-
ratory studies index activity rhythms that could be
considered crepuscular in nature (Garcia-Allegue et al.
1999) To expand our understanding of basic visual ad-
aptations among rodents from this understudied group,
and to provide baselines for predicting the effectiveness
of light on its behavior, we have undertaken a series of
investigations of Octodon degus that includ ed (1) mea-
surements of the absorption properties of the lens, (2) an
assessment of photoreceptor distributions by opsin an-
tibody labeling, (3) identification of the spectral prop-
erties of these receptors by means of electrophysiological
measurements, and (4) an examination of how these
adaptations relate to basic visual capacities through a
series of behavioral experiments. A preliminary version
of this research has been reporte d in an abstract (C ald-
erone et al. 2001).
Materials and methods
Adult degus (Octodon degus) of both sexes were used. These ani-
mals, all derived from laboratory-bred lines, were housed either
singly or in small groups under standard colony conditions. During
the light phase of a 12:12 h daily cycle the ambient illumination in
the colony room measured at the position of the cages averaged
150 lx. With the exception of animals engaged in behavioral ex-
periments, all others were given free access to food and water.
Those behavioral subjects were fed daily immediately following test
sessions in an amount sufficient to maintain their weight at 90% or
more of their free-feeding weights.
Lens transmission measurements
Transmission measurements were made using an Ocean Optics
pulsed xenon light source (PX-1) and spectrometer (USB 2000 UV-
VIS). Animals were euthanized by exposure to halothane and the
lenses were rapidly dissected from the eyes. The extracted lens was
immersed in mineral oil (refractive index 1.47) and placed in a
custom-made cell that was illuminated from below though a UV-
transmissive quartz window. The lens was centered over the win-
dow on an aluminum mask that had an aperture of 1.6 mm. The
transmitted beam had a maximum spread of ±11°. The detector,
which was positioned above, was a UV-VIS cosine receptor (UR12,
Ancal) that was coupled to the spectrometer by a 600-lm quartz
fiber. Transmission measurements were collected at 10-nm intervals
over the spectral range from 300 to 700 nm.
Degus were euthanized by exposure to halothane. The eyes were
rapidly enucleated and placed in chilled 4% paraformaldehyde. The
retinas were dissected from the eyecups and postfixed for an addi-
tional period of 3 h. Following this, they were rinsed in PBS, bathed
in 0.3% hydrogen peroxide, rinsed, and placed in 2% BSA, 10%
NGS solution for 3 h. The retinas were then exposed to polyclonal
antibodies specific to either UV/S(JH455; 1:100,000 dilution) or M/L
(JH492; 1:30,000 dilution) opsins (Chu et al. 1994) or to a mono-
clonal antibody specific to rod opsin (rho402; 1:500 dilution). The
retinas were incubated in the primary antibodies for 72 h, followed
by 24-h incubation in secondary antibody and, finally, exposed to
avidin–biotin–peroxidase complex for an additional 24 h. The reti-
nas were subsequently flat mounted and photoreceptor counts were
taken at 1-mm intervals across the tissue.
For confocal analysis, retinas were blocked overnight at 4°Cin
normal donkey serum (1:20) containing 0.1 mol l
PBS, 0.5%
BSA (Sigma, St. Louis, Mo., USA), 0.1% Triton X-100 (LabChem,
Pittsburgh, Pa., USA), and 0.1% sodium azide (Sigma), together
referred to as PBTA. Next, retinas were incubated for 24 h at 4°C
on a rotator in a combination of biotinylated peanut agglutinin
(PNA) lectin (1:200; Vector Laboratories, Burlingame, Calif.,
USA) to label cone matrix sheaths and rod opsin antibody (1:500).
The retinas were then rinsed in PBTA and incubated in streptavidin
conjugated to the fluorochrome Texas Red (1:200; Vector Labo-
ratories) and donkey anti-mouse IgG conjugated to the fluoro-
chrome Cy2 (1:200, Jackson ImmunoResearch Laboratories, West
Grove, Pa., USA) for 24 h at 4°C on a rotator. Finally, the retinas
were whole-mounted photoreceptor side up in 5% n-propyl gallate
in glycerol and viewed on a laser scanning confocal microscope
(model 1024; Bio-Rad, Hercules, Calif., USA). All solutions were
made in PBTA.
In preparation for recording, animals were anesthetized with an
IM injection of a mixture of xylazine hydrochloride (8 mg kg
and ketamine hydrochloride (42 mg kg
) and the pupil of the test
eye was dilated with a topical application of atropine sulfate
(0.04%). The animal was placed in a head holder that allowed for
alignment of the eye with an optical system. Electroretinograms
(ERGs) were differentially recorded using Burian-Allen style
contact-lens electrodes. Except as indicated below, all the re-
cording was done in a room equipped with overhead fluorescent
lighting that provided an ambient illumination of about 100 lx at
the position of the test eye.
Spectral properties of the degu eye were assessed with ERG
flicker photometry. We have frequently used this technique and have
provided details elsewhere (Jacobs and Neitz 1987; Jacobs et al.
1996). In this procedure, temporally alternating stimuli derived from
two light sources are presented to the eye in Maxwellian view (59°
spot) in the form of an interleaved train of pulses each of which
operates on a 25% duty cycle. One of these lights serves as a test light,
the other as a reference light. The amplified ERG signal is windowed
with a sinusoid that is set to the frequency of the stimulus train with
the position of the window adjusted to maximize its correlation with
the ERG signal (Jacobs et al. 1996). ERGs elicited by the two lights
are electronically compared and, over presentations, the test light is
adjusted in intensity (by positioning a circular neutral density wedge)
until the ERG it generates is the same as that produced by the ref-
erence light. The comparison of the responses to test and reference
lights were based on averages of the last 50 of 70 responses elicited
from each source. The intensity equations, each of which defines one
point on a spectral sensitivity function, typically vary by less than
0.04 log units in repeated measurement (Jacobs et al. 1996). Repeti-
tion of this procedure for a range of test wavelengths permits a
characterization of the spectral sensitivity of the mechanism(s) that
are responsive to the lights. Variations in the details of the test situ-
ation (e.g., adaptation state, temporal frequency) are used to selec-
tively access signals from different spectral generators. Some of the
ERG experiments did not involve spectral sensitivity measurements.
In those cases stimuli were from the test light alone, but in all other
regards the stimulus and recording arrangements were identical to
those already described.
Stimuli were derived from a three-beam optical system featuring
three, optically-superimposed beams. The test light came from a
monochromator (15 nm bandpass) equipped with either a 100-W
tungsten-halide lamp or a 150-W xenon lamp. The reference light
and a third beam that could be used for accessory adaptation
originated from 50-W tungsten-halide lamps. Stimulus timing was
controlled through the use of high-speed electromechanical shut-
ters. The ERG experiments were as follows.
Flicker rate sensitivity
This experiment examined the sensitivity of the degu ERG to lights
presented at differing temporal rates. A test light of 500 nm was
used with flicker rates taken at stepped intervals of 4 Hz over the
range of 4–48 Hz. At each tested rate, the intensity of the test light
was adjusted to yield an ERG having average amplitudes of 6.3 lV.
Each threshold was measured twice and these results were subse-
quently averaged.
Scotopic spectral sensitivity
Animals were dark adapted for a period of 30 min and the sub-
sequent recording was made in a darkened room. The pulse rate of
the photometer was 5 Hz and the reference light was achromatic
(2850 K; 1.19 log td). Photometric equations were made following
the above procedures for test lights taken at 10-nm intervals from
440 nm to 610 nm. Two scans across this spectral range were
completed and the results were averaged.
Full-spectrum photopic sensitivity
ERG spectral sensitivity functions were obtained under photopic
test conditions for test wavelengths from 360 nm to 620 nm. The
pulse rate was 31.25 Hz and the achromatic reference light
provided a retinal illuminance of 2.7 log td.
Photopic short-wavelength spectral sensitivity
To selectively assess contributions from cones sensitive to short-
wavelength lights, spectral sensitivity measurements were made
using a 31.25-Hz stimulus. Several other features of the test situa-
tion were designed to maximize the contribution to the ERG signal
from short-wavelength mechanisms. These included (1) concurrent
retinal illumination by long-wavelength light that was produced by
placing a Schott long-pass glass filter (GG475, 3 mm) having 50%
transmission at 475 nm into the adapting beam (yielding a corneal
irradiance of 17.2 log photons/s/sr), and (2) a short-wavelength
reference light (370 nm; 15.3 log photons/s/sr). With these condi-
tions in force, spectral sensitivity measurements were made for test
wavelengths over the range of 370–420 nm.
M-cone spectral sensitivity
To define the spectral sensitivity of the M cones, flicker photo-
metric ERG spectral sensitivity measurements were made with the
following test conditions: flicker rate 31.25 Hz; reference light
achromatic (3.97 log td). Spectral sensitivity measurements were
made at 10-nm intervals over the spectral range from 430 nm to
590 nm, all of which were obtained and treated as described
Increment thresholds of cone mechanisms measured
for long-wavelength adaptation
Increment thresholds were determined for two test lights (390 nm
and 510 nm) that were flickered at 31.25 Hz. Thresholds were de-
fined as the intensity of the test light eliciting an ERG having av-
eraged amplitude of 3.2 lV. These thresholds were measured on
steady, long-wavelength backgrounds that were produced by using
a 475-nm long-pass filter. Separate threshold measurements were
made for both test wavelengths on a series of such backgrounds
that were successively varied in intensity from 13.19 to 17.17 log
photons/s/sr. The initial value was low enough that it produced no
change in sensitivity of the eye from that measured in the absence
of accessory adaptation. Thresholds were determined for a total of
14 different adaptation intensities.
Visual capacities were assessed using a three-alternative, forced-
choice discrimination test. The apparatus used and the general
procedures were as described earlier (Jacobs 1983; Jacobs et al.
1999). In brief, lights were transprojected onto three circular and
translucent test panels that were mounted in a line on the wall of a
small test chamber (diameter and center–center distance of pan-
els=2.5 cm). The lights came from 100-W or 150-W tungsten-
halide lamps and from an Instruments SA grating monochromator
(half-energy passband=16 nm) equipped with a 75-W xenon lamp.
The former lights were used to diffusely and equally illuminate the
three test panels (background lights). With the use of an automated
mirror system, light originating from the monochromator (the test
light) was directed toward any of the three panels and, depending
on the goal of the experiment, it either replaced one of the three
background lights or was added to one of the three. Except for one
experiment (noted below) the test chamber was ambiently illumi-
nated (70 lx) by fluorescent tubes that were mounted in the ceiling
of the test chamber.
Operant conditioning procedures were used to train degus to
indicate which of the three panels was illuminated by the test light
by touching that panel. Correct choices were reinforced by the
automatic delivery of 20 mg food pellets. Over trials the location of
the test light was varied randomly across the three panels. The
nature of the illumination difference between the positive and the
two negative panels was systematically varied across trials and test
sessions to allow measurement of detection thresholds. Test trials
were marked for the subject by the occurrence of a concurrent
cueing tone which terminated when the animal responded or after a
period of 15 s without a response. A noncorrection procedure was
employed. The intertrial duration was varied somewhat across
subjects to accommodate individual behavioral idiosyncrasies while
having an average duration of about 6 s. To suppress adventitious
responding a penalty time was assessed such that the onset of test
trial was delayed by 5 s following any responses made during the
intertrial interval. Subjects typically completed 300–700 trials per
daily test session. The behavioral tests involved measurements of
both increment sensitivity and of spectral discrimination. The
specific experiments were as follows.
Scotopic spectral sensitivity
To assess scotopic spectral sensitivity the overhead lights in the test
chamber were turned off and the background lights were filtered to
provide dim (0.4 cd m
), achromatic (5500 K) panel illumination.
Animals were placed in this darkened chamber for 10–15 min prior
to the onset of testing. Sensitivity was measured for test lights over
the range 420–600 nm taken at 10 nm intervals. In preliminary tests
we established for each subject the range of test light intensities
required at each wavelength to yield discrimination performance
varying from about 80–90% correct down to chance (33%). This
range typically encompassed about 2 log units of test light intensity.
In the experiment proper, test lights of each wavelength were varied
in steps of 0.3 log unit. Each intensity/wavelength combination was
presented as a block of 3 trials. The ordering of test wavelengths
was randomized. This process was continued across test sessions
until results for 50–70 trials had been accumulated at each wave-
length/intensity combination. From these cumulated results, psy-
chometric functions were derived by plotting mean percentage
correct responses as a function of stimulus intensity and then fitting
these points to a logistic function having asymptotes of 100% and
33% correct with the variance and mean as free parameters. The
function yielding the best fit (least squares) to the data array was
determined, and threshold taken as the stimulus intensity required
for performance at a 99% level of confidence.
Effects of light adaptation on spectral sensitivity
Spectral sensitivity measurements were also made under conditions
of light adaptation. In each of these cases the test chamber was
ambiently illuminated. In a series of experiments, spectral sensi-
tivity measurements were made on a variety of different achromatic
(6600 K) backgrounds having luminance values that varied be-
tween 13 cd m
and 514 cd m
. In all other regards, including the
details of data analysis, the procedures used here were identical to
those just described for the scotopic test conditions.
Increment thresholds measured with long-wavelength adaptation
This experiment was similar in nature to an ERG experiment de-
scribed above. In this case thresholds were measured using the
standard psychophysical procedures also described above (Scotopic
spectral sensitivity) for four test lights (wavelengths of 380, 390, 500
and 510 nm) presented on steady long-wavelength backgrounds.
These background lights were produced by passing light from the
tungsten halide lamp through a 475-nm long-pass filter. In suc-
cessive experiments, increment thresholds were measured on this
long wavelength background for a variety of background intensi-
ties that were varied over a range from 11.51 to 14.51 log photons/
s/sr. All the other experimental details were as described above.
Spectral discriminations
This test was explicitly designed to see if degus could discriminate
between UV and visible lights in the absence of any consistent
brightness cues. To do this we first established brightness matches
between different spectral stimuli using a technique earlier em-
ployed in a study of mouse vision (Jacobs et al. 1999). Animals
were initially trained to select the more luminant (‘‘brighter’’) of
three lights, i.e., the task employed for all of the spectral sensitivity
measurements described above. Once that skill was well estab-
lished, brightness matches between different spectral stimuli were
obtained. To accomplish this, the test arrangement was altered so
that the monochromatic test light now replaced one of the back-
ground lights on each test trial rather than being added to it. Over
trials the intensity of the test light was progressively reduced in
steps of 0.3 log unit. When the test light appeared much brighter
the animal selected it consistently, just as they had been trained to
do, but as the test light intensity was reduced the percentage of
correct selections declined and eventually reversed, i.e., the subject
now avoided the test light, presumably because the comparison
panels appeared brighter. The point at which the animal’s perfor-
mance passed through the level of chance (33% correct) is taken to
define a brightness equation between the test light and the two
background lights. In this fashion, brightness equations were ob-
tained for a 500-nm test light versus both 500-nm and 370-nm
backgrounds and between a 370-nm test light and 500-nm and
370-nm background lights. With these equations in hand and from
reference to the spectral sensitivity functions already measured
predicted brightness matches could be derived for subsequent tests
of spectral discrimination.
Animals were next tested to see if they could learn to discrim-
inate a 370-nm test light from 500-nm lights. Over trials the test
light was presented at intensity values calculated to be equally
bright to the 370-nm light and at intensity values that were ran-
domly varied over a range of ±0.4 log unit in steps of 0.1 log unit
from the point of brightness equation. Once the animal acquired
this discrimination the wavelength of the test light was progres-
sively changed in steps of 5 or 10 nm toward the spectral location
of the 500 nm light. At each spectral point the intensity of the test
light was varied in the fashion just described. In this way a total of
50–70 trials were accumulated at each of the wavelength/intensity
combinations. In a second stage of this experiment the direction of
the discrimination was reversed, i.e., the test light was set to 500 nm
and the comparison light was 370 nm. Controls for brightness cues
were instituted in the same fashion as before. Once this second
discrimination was acquired the test light was again stepped in
wavelength, but this time toward the shorter wavelengths. As be-
fore, at each test wavelength the intensity of the test light was
varied in steps of 0.1 log unit over a range encompassing ±0.4 log
unit around the location of the brightness match.
Lens transmission
Transmission measurements were obtained from five
lenses of three adult degus (mean age ca. 3 years). The
resulting values were normalized to have 100% trans-
mission at 700 nm and then averaged across animals.
These values are plotted as solid triangles (±1 SD) in
Fig. 1. The degu lens shows only modest absorption
throughout the middle to long wavelengths but then
progressively increases in optical density at wavelengths
shorter than about 430 nm to a point where transmis-
sion became immeasurably low at wavelength values of
Fig. 1 Lens transmission measurements. All values have been
normalized to 100% at 700 nm. Shown are mean values (±1 SD)
for five adult degu lenses (solid triangles), six adult rat lenses (filled
circles), and two ground squirrel lenses (open circles). The data
points are connected by straight lines
about 310–320 nm. To assess the effectiveness of the
measurement procedure, and to provide a comparative
context, lens transmission measuremen ts were also mad e
with the same procedures for adult rats, Rattus
norvegicus (six lenses from four animals), and for two
lenses taken from a California ground squirrel
(Spermophilus beecheyi) of approximately 6 months of
age. The averaged transmission values obtained from rat
(filled circles) and ground squirrel lenses (open circles)
are very similar to those previously obtained from these
two species with quite different procedures (Yolton et al.
1974, Gorgels and van Norren 1992) thus validating the
adequacy of our measurement technique. From a com-
parative perspective there is significant variation among
these species in the magnitude of short-wavelength fil-
tering in the ocular lens; the ground squirrel lens absorbs
short-wavelength light strongly, the rat lens is quite
transmissive down to very short wavelengths, while the
absorption of short-wavelength light in the degu lens lies
intermediate to that of these other two so that, for ex-
ample, the spectral locations where lens transmissivity
drops to 50% of its maximum value are 343 nm,
378 nm, and 464 nm for the rat, degu, and ground
squirrel lenses, respectively.
It has long been known that the human lens becomes
progressively less transmissive with age and that this
effect is particularly pronounced at short wavelengths
(Weale 1988; Xu et al. 1997). Similar changes have also
been documented for nocturnal species like the golden
hamster Mesocricetus auratus (Zhang et al. 1998) and for
diurnal animals such as tree squirrels Sciurus carolinensis
(Zigman et al. 1985). To see if this is also true for the
degu lens, we addi tionally measured lenses obtained
from two young degus (age 7 months) and from an older
animal (5 years). The mean absorption curves ob-
tained from animals of the three ages are summarized in
Fig. 2. As for all the species just mentioned, the degu
lens also shows a progressive increase in optical density
as a function of age. For example, transmission at
400 nm declines from about 80% of its maximum in the
very young animals to about 40% in the aged animal.
Opsin labeling
Two classes of cone were identified by opsin labeling in
the degu retina. The antibody JH492, specific for M/L
cones, labeled a substantial pop ulation of receptors,
while JH455 labeled a smaller number of UV/S cones.
The density distribution of the M/L cones is shown for
two degu retinas in Fig. 3. These cones reach a peak
density of more than 55,000 mm
in an area superior to
the optic nerve head then decrease to fewer than
15,000 mm
in the periphery. This decrease is more
gradual along the horizontal axis than in the vertical
direction. Figure 4 shows analogous density maps for
the labeled UV/S cones. These receptors have highest
density in a region slig htly inferior to the optic nerve
head where they peak at 5,400–7,000 cells mm
the case for the M/L cones, the UV cones decrease in
density more rapidly toward the superior and inferior
periphery than in the orthogonal direct ions. From these
counts it is estimated that there are about 2.9 million
M/L cones (average of two retinas) and about 221,000
UV/S cones (average of three retinas) yielding a ratio of
about 13:1 (M/L:UV /S).
Similar counts were conducted for cells labeled with
an antibody specific to rod opsin. The density distribu-
tion map for rods obtained from one animal appears in
Fig. 5 (left). Rods reach a peak density of about
110,000 mm
in the temporal retina in a region slightly
inferior to the optic nerve head. Rod density declines
modestly across the degu retina never falling below
about 50,000 mm
. Summing the derived densities
across the retina yields an estimate of 6.5 million rods.
These counts indicate that as many as one-third of all
the photoreceptors in the degu retina are cones. The
relatively high representation of cones in this retina is
illustrated in the micrograph (Fig. 5, right) taken from a
location where degu rods and cones are approximately
equally frequent.
Temporal sensitivity of the ERG
The stimuli for all of the ERG experiments were flick-
ering lights. Accordingly, we first examined the sensi-
tivity of the degu ERG as a function of flicker frequency.
Figure 6 shows the average sensitivity of the funda-
mental component of the ERG response recorded at
each of 12 flicker rates for four animals. The ERG was
maximally sensitivity to a pulse rate of about 12 Hz.
Sensitivity declined from this point toward the slower
rates and there is also a more gradual decline in sensi-
tivity for rates faster than 12 Hz. By extrapolation from
the highest frequency tested, it appears that with these
Fig. 2 Degu lens transmission as a function of age. Measurements
were made for animals of age 7 months (closed circles), age 3 years
(triangles), and at an age of approximately 5 years (open circles).
Other details as in Fig. 1
test conditions ERG signals recorded from the degu
faithfully follow temporal change up to at least about
50 Hz.
ERG spectral sensitivity measurements
The opsin labeling experiments indicate that the degu
has three classes of photoreceptor. As outlined above,
we conducted a series of ERG experiments to delineate
their spectral properties. Spectral sensitivity functions
obtained from two degus under scotopic test conditions
appear in Fig. 7 (top). The results for the two are very
similar. The sensitivity values in that figure have been
corrected for lens absorption (Fig. 1) and then best fitted
with a visual pigment template (Govardovskii et al.
2000). For the latter step, the template (continuous
curve in Fig. 7) was shifted along the wavelength axis in
steps of 0.1 nm to determine the spectral location pro-
viding the best least squares fit to the data array. The
spectral peak so defined, 500.2 nm, is typical for rod
photopigments measured in many other mammalian
Under conditions of light adaptation the degu ERG
shows maximum sensitivity to lights that are flickered at
about 12 Hz (Fig. 6). Spectral sensitivity curves were
obtained using ERG flicker photometry for three ani-
mals tested with 12.5-Hz stimulation. The averaged re-
sults are plotted as the bottom function of Fig. 7 where
they have been best fit to a photopigment template in the
fashion just described. The spectral peak of this curve
(500.6 nm) is virtually identical to that recorded under
scotopic test conditions. The two functions of Fig. 7
have been positioned on the sensitivity axis to accurately
reflect the absolute difference in sensitivity for the two
test conditions. Although the differences in the stimulus
conditions between the two experiments (state of light
adaptation and flicker rate) produced an average
threshold elevation of about 2.8 log units, there was no
corresponding shift in spectral sensitivity. This suggested
Fig. 3 Spatial distribution of
(M) cones in the degu retina.
The cones were identified by
their labeling to the opsin
antibody JH492. The isodensity
contours were derived from
counts made at approximately
95 sites. The plots represent
results obtained from two
animals. Orientations: D dorsal,
T temporal, N nasal. Here and
in Figs. 4 and 5 the scale
bar=1 mm
Fig. 4 Spatial distribution of
ultraviolet (UV) cones in the
degu retina. The cones were
identified by their labeling to
the opsin antibody JH455.
Other details are as for Fig. 3
either that the M/L cones of the degu have the same
spectral sensitivity as that of the rods or that they
require more stringent adaptation conditions for their
ERG signatures to emerge.
To evaluate these possibilities, and since the degu
ERG responds quite robustly to fast flicker (Fig.6), we
next measured spectral sensitivity functions for three
animals using 31.25-Hz stimuli, in this case testing
wavelengths over the full range from 360 nm to 620 nm.
The results (Fig. 8, top) have two noteworthy features.
First, the function shows a secondary peak in the UV
centered at 360–370 nm. This could indicate either the
presence of a UV pigment or reflect contributions to
short wavelength sensitivity from beta-band absorption
by pigments with their main peaks in the visible spec-
trum (Jacobs 1992). The magnitude of the elevation in
the UV suggested the former alternative as the more
likely. A second feature is that the main peak of this
curve is shifted toward longer wavelengths by 5–7 nm
relative to the spectral sensitivity functions that appear
in Fig. 7, and thus its seems likely that this spectral
sensitivity function reflects contributions from both UV
and M cones. To examine this we conducted a more
extensive evaluation of spectral sensitivity over the
Fig. 5 Left: spatial distribution
of rods in the degu retina. Rods
were identified by their labeling
to an antibody specific to rod
opsin (rho402). The plot
represents results obtained from
one retina. Other details are as
for Fig. 3. Right: laser scanning
confocal micrograph
illustrating the relative densities
of rods (labeled with rod opsin
antibody shown in green) and
cones (labeled with PNA and
shown in red) in the degu retina.
At this retinal location rods and
cones are about equally
Fig. 6 The temporal sensitivity of the degu electroretinogram
(ERG). At each pulse rate the intensity of a 500-nm test light
was adjusted until it produced an ERG having an averaged
amplitude of 6.3 lV. The plotted values represent mean thresholds
recorded for four animals (error bars=±1 SD)
Fig. 7 Spectral-sensitivity functions for the degu ERG obtained
under conditions of dark adaptation (top, 5-Hz flicker) and light
adaptation (bottom, 12.5-Hz flicker). Results are shown for two
animals tested under dark adaptation (open circles and closed
triangles) and three animals examined under light adaptation where
the solid circles represent mean sensitivity (±1 SD). The vertical
separation of the two sets of data accurately reflects the difference in
sensitivity recorded for the two test conditions. The sensitivity
values have been corrected for absorption by the degu lens (middle
curve, Fig. 1) and these have been best fitted to a visual pigment
absorption function (see text). The best-fit curves so derived have
wavelength peak values of 500.2 nm (top) and 500.6 nm (bottom)
visible region of the spectrum. The results for six addi-
tional animals are shown in the lower right of Fig. 8.
Note that there is only modest variability across these
subjects and that a fitted pigment template having a
spectral peak of 507.2 nm gives an excellent fit to the
data. Individual fits similarly made for the data from six
subjects yie lded a mean peak value of 507.3 nm with
small between-animal dispersion (SD=1.48 nm).
The data to the lower left of Fig . 8 represent the
mean spectral sensitivity values measured for three
subjects over a spectral range of 370–420 nm under test
conditions (defined above) that would be expected to
favor contribution from a UV-sensitive photopigment
while at the same time minimizing the influence of pig-
ments maximally sensitive in the visible spectrum. The
results indicate clearly the presence in the degu retina of
a UV cone photopigment. To generate the template fit to
these data we reasoned that, despite intense long-wave-
length adaptation, there might still be some contribution
to short-wavelength responses from stimulation of the
M cones. Accordingly, we employed a fitting procedure
based on the assumption that the sensitivity values re-
flected a linear summation of signals from the M cone
and from the UV cone. To derive this fit the spectral
position of the M cone was set to 507.2 and an iterative
search was performed to find the spectral location of a
second template pigment that provided the best account
of the entire data array. The resulting curve has a peak
of 362 nm. These measurements indicate that the two
classes of degu cone have peaks of about 362 and
507.2 nm and, accordingly, the full spectral sensitivity
function at the top of Fig. 8 has been best fit by linearly
summing these two pigment templates. That fit required
an 84.7% contribution from the M pigment and 15.3%
contribution from the UV pigment. In sum, these ERG
spectral-sensitivity measurements provide estimates of
the spectral sensitivities for the three classes of photo-
receptor present in the degu retina.
Increment thresholds measured for UV- and M-cone
The previous experiment showed that the two cone
mechanisms in the degu retina can be differentially in-
fluenced by chromatic adaptation. To explore the nature
of this process increment thresholds for 390-nm and 510-
nm test lights were measured on a series of long-wave-
length backgrounds produced by passing light through a
475-nm long-pass filter. The test lights were progres-
sively incremented in intensity and flickered at the rate
previously shown adequate for accessing cone signals
(31.25 Hz). Figure 9 summarizes the results in the form
of increment-threshold functions for the two test lights.
Thresholds for the 510 -nm light (open circles) increased
linearly (slope=0.71, r
=0.991) across the full intensity
range of long-wavelength adaptation. Results for the
390-nm test are quite different over a roughly 2.5-log-
unit range, starting from the lowest adaptation level, the
390-nm thresholds also showed a linear increase, but
with a substantially lower slope (slope=0.324; r
0.927). Above a point corresponding to an adaptation
intensity of about 15.7 log photons/s/sr there are no
further increases in the thresholds measured for 390 nm
test lights (the line fit to the second segment has a slope
of 0.004).
For one animal we also followed the course of the
recovery of sensitivity following intense ch romatic ad-
aptation. To accomplish this, an averaged response was
first obtained for a 510-nm light (31.25 Hz; 15.21 pho-
ton/s/sr) prior to the initiation of chromatic adaptation.
The eye was then steadily exposed to a long-wavelength
adaptation light the intensity of which is indicated by the
small arrow on the axis in Fig. 9. The response was
again measured using the 510-nm test light. This adap-
tation was sufficient to reduce the amplitude of the
response to about 7% of its pre-adaptation value.
The adaptation light was extinguished and averaged
responses to the test light were obtained at varying in-
tervals following light offset. The inset in Fig. 9 shows
that the recovery of the response to its pre-adaptation
level was effectively complete in about 3 min. The ex-
ponential recovery curve fitted to those recovery data
has a time constant of 93 s. Recovery times measured
Fig. 8 ERG spectral-sensitivity functions recorded from degus
using 31.25-Hz flicker. The function at the top shows mean values
for three animals. Plotted at the bottom right are mean values for
six animals while the function at the bottom left similarly reflects
values for three animals. In all cases the error bars are ±1 SD. All
three spectral-sensitivity functions have been corrected for lens
absorption and best fit (continuous lines) with visual pigment
templates. The reference light and adaptation conditions used are
specified in the text, as are the details of the template fitting
procedures. The function to the lower right represents an ERG
estimate of the spectral properties of degu M cones
=507.2 nm), while analysis of the lower left function infers
contributions from a UV cone having peak sensitivity at about
362 nm. The full spectral sensitivity function at the top was fit by
linear summation of the curves for the UV (15.3%) and M (84.7%)
cone pigments
following adaptation are heavily dependent on experi-
mental parameters, but this value is quite similar to
those obtained for recovery of human cone sensitivity
following a full bleach as assessed with either ERG
(Paupoo et al. 2000) or densitometric measurements
(Hollins and Alpern 1973).
Behavioral measurements of spectral sensitivity
Complete spectral-sensitivity functions were obtained
from two animals tested under scotopic conditions
(darkened test chamber, background panel luminance=
0.4 cd m
). The threshold values for the two animals
were closely similar. In Fig. 10 the spectral-sensitivity
values have been best fit to a visual pigment template
using the same procedures as for the ERG spectral-
sensitivity measurements. The spectral peak of that
curve is 500.2 nm, a value very close to that obtained
from the ERG measurements of scotopic spectral sen-
sitivity (Fig. 7, top). Spectral-sensitivity functions were
next measured for three light-adapted degus (ambient
illumination of 70 lx and a background lumi-
nance=13 cd m
). The resulting spectral sensitivity
functions are shown at the bottom of Fig. 10. Over the
spectral range 430–590 nm the data are well fit by a
visual pigment template having a peak value of
500.5 nm. Thus, although the change from dark to light
adaptation produced an overall drop in sensitivity of
about 2.2 log units there was no significant chan ge in
spectral sensitivity, suggesting that rods control led vi-
sual behavior over the spectral range tested for both of
these lighting conditions.
To assess what adaptation conditions might be re-
quired to allow the emergence of M cone contributions,
complete spectral sensitivity functions were determined
on four additional background levels (equivalent to 1.78,
2.1, 2.37 and 2.71 log cd m
). Two or three animals were
tested at each of the first three of these levels and a single
animal was examined at the highest level. In each case
spectral sensitivity was measured as described above and
these data were subsequently best fit to a visual pigment
template. Figure 11 shows the derived peak of this curve
plotted as a function of the luminance of the background
light. The results indicate a continuous shift toward the
longer wavelengths from the spectral position character-
istic of degu rods to that of degu M cones. Although the
total shift in spectral sensitivity is not very large across the
adaptation range, it is quite reliable (note the variation
among subjects.) From this experiment one can conclude
that in a behavioral test situation a steady light level
corresponding to a luminance of about 2.3 log cd m
required to fully shift degu spectral sensitivity from
control by rods to control by cones.
With this fact in mind, complete increment-threshold
spectral-sensitivity functions were determined for two
Fig. 9 ERG increment threshold functions obtained for 390 nm
(solid circles) and 510 nm (open circles) test lights in the presence of
long wavelength adaptation. The data points are mean values
(±1 SD) for nine (510 nm) and five animals (390 nm), respectively.
Threshold elevation is specified relative to values obtained in
absence of any chromatic adaptation. In this figure and in Fig. 13B,
the specification of the intensity of the adapting light was obtained
by first making spectroradiometric readings across the visible
spectrum at 10-nm intervals from 470 nm to 700 nm. These values
were then weighted according to the spectral-sensitivity curve for
the degus cone (k
=507 nm). The statistics for the fitted linear
regressions are given in the text. The inset at the upper left shows
the time-course of ERG recovery following the offset of chromatic
adaptation. The intensity of this adaptation light is indicated by the
arrow on the x-axis. The plotted values on the recovery curve are
means for four measurements made at each point. The fitted
recovery curve (solid line) is an exponential having a time constant
of 93 s (r
Fig. 10 Degu spectral-sensitivity functions derived from behavioral
discrimination tests. Top: mean increment threshold values
(±1 SD) obtained from two animals with scotopic test conditions
(darkened test chamber; background panel luminance=0.4
cd m
). Bottom: thresholds similarly measured for three animals
(means ±1 SD) under conditions of ambient illumination with
background panel luminance of 13 cd m
. The vertical separation
of the two curves reflects the magnitude of the threshold differences
seen for the two test conditions. The functions have been corrected
for lens absorption and best fit (continuous lines) with photopig-
ment templates
degus over the spectral range 370–570 nm with the
background luminance set 2.4 log cd m
. The resulting
spectral-sensitivity fun ction (Fig. 12) displays a strik-
ingly elevated sensitivity in the UV with a secondary
(and much lower) peak at around 500 nm. The contin-
uous line through the data points represents the best-
fitting linear summation of two visual pigment templates
having respective peak values of 362 nm (97%) and
507 nm (3%). Although the general shape of the func-
tion is similar to that derived from ERG measurements,
the relative contributions of the UV and M cone
mechanisms are greatly different indeed, the two are
nearly reversed in proportion for the two types of
Chromatic adaptation and increment threshold
Increment thresholds were measured for two UV test
lights (380 and 390 nm) and two test lights positioned
near the peak of M-cone sensi tivity (500 and 510 nm).
The results from each pair were then averaged together.
Three subjects were tested on a series of adapting
backgrounds spanning the range 0.63–2.66 log cd m
Thresholds for both UV and visible test lights increased
as the background light level was raised. The result (for
a total of 33 independently determined thresholds) is
summarized in Fig. 13A which plots the elevation in
threshold for the UV test lights as a function of the
elevation in threshold seen for the visible test lights. The
continuous line is the best fittin g linear regression
(slope=0.64; r
=0.94; P<0.01). The broken line in-
dicates the relationship that would be expected if the two
sets of thresholds reflected the adaptation of a uni-
variant mechanism. That prediction differs significantly
(P<0.01) from the relationship that was observed. The
maximum intensities of background lights that could
be explored in this experiment were limited. We were
subsequently able to extend somewhat the upper limit of
available background light by increasing the power of
the background lamp. With this new arrangement, but
otherwise with all the same procedures, the adaptation
experiment was re-run on one of the subjects whose data
are shown in Fig. 13A. The results for this animal are
shown in Fig. 13B where the thresholds for both sets of
test lights are now plotted as a function of background
irradiance. As before, both thresholds increased mono-
tonically as a function of the level of light adaptation
although they increased with very different slopes
(0.86 versus 0.47). Both of these experiments show that
long-wavelength adap tation has a differential effect on
thresholds measured for UV and visible test lights and
that, as the earlier comparison of ER G and behavioral
spectral-sensitivity functions suggested, the degu UV
and M mechanisms do not behave according to strict
The effects of long-wavelength chromatic adapta-
tion on sensitivity to UV and visible lights were ex-
amined in parallel measurements made with the ERG
and with behavioral discriminations (Figs. 9 and 13B).
In comparison to the ERG results, the slopes for in-
crement-threshold functions were somewhat steeper for
behavioral measurements, implying that there is some
additional sensitivity adjustment in the degu visual sys-
tem that must be accomplished somewhere central to the
outer retina. An additional difference between the two
sets of measurements is that ERG thresholds for the UV
test light became independent of background intensity
for measurements made with long-wavelength back-
grounds having an intensity in excess of about 15.7 log
photons/s/sr. Presumably at that value and above the
sensitivity of the degu M cones is now sufficiently de-
pressed that its stimulation by the UV test light provided
no further detectable contribution to overall threshold.
The adaptation lights have been equivalently scaled in
Fig. 11 The dependence of the location of the peak the increment-
threshold spectral-sensitivity functions on the luminance level of
the background light. Using techniques that are described in the
text, spectral sensitivity functions were measured in a behavioral
test situation for a variety of background light levels. The plotted
values represent the spectral peaks of the best-fitting photopigment
absorption function obtained at each of these background levels
(mean±1 SD)
Fig. 12 Increment-threshold spectral-sensitivity function measured
with a background luminance of 2.4 log cd m
. The values are
means obtained from two degus. The values were corrected for lens
absorption and best fit (continuous line) by linearly summing two
photopigment templates having respective peak values of 362 nm
(97%) and 507 nm (3%)
Figs. 9 and 13B and the comparison of the two figures
implies that the adapting light levels required to reach
that point in the ER G measurements were not available
in the behavioral test situation.
Spectral discriminations
These experiments were designed to see if degus could
make color discriminations. Figure 14 illustrates the
brightness matching procedure that served as a pre-
liminary to the formal color tests. In this case a UV test
light appeared on one panel and the other two panels
were illuminated with 500-nm light. The animal had
been previously trained to pick the ‘‘brighter’’ of lights
presented and, because of that, he selected the panel
illuminated with the UV illuminated panel nearly 100%
of the time. Over presentation blocks of three trials each
the luminance of the UV light was systematically re-
duced and that whole process was repeated a number of
times. As can be seen, as the test light was dimmed the
percentage of trials on which it was selected progres-
sively declined to chance; at still lower luminances the
subject consistently avoided that light, presumably
because the other two test panels now appeared the
brighter. The intensities of the two lights at the level of
chance performance (vertical line) were taken to define a
brightness equation.
Once brightness equations were established, degus
were trained to determine if they could use spectral
cues alone to discriminate the two lights; in effect, to
then ignore luminance-related cues. For this phase, and
for all further color tests, the test and comparison
lights were presented at intensities calculated to be
equally bright and at values deviating in steps of 0.1
log units over a range of ±0.5 log units around the
equation value with the ordering of intensity values
randomized. The triangles in Fig. 14 show the perfor-
mance achieved by this same subject at the conclusion
of the training regimen. The animal quite clearly has
learned to discriminate the 500 nm light from the UV
light in the absence of an y consistent brightness dif-
ferences, i.e., to make a color discrimination. It was
apparently not trivial for these rodents to learn to ig-
nore brightness differences as evidenced by the fact that
the three subjects tested required between 12 and 21
Fig. 13 A Effects of chromatic adaptation on visual thresholds in
the degu. Increment thresholds for UV (380 nm, 390 nm) and
middle-wavelength (500 nm, 510 nm) test lights on long-wave-
length backgrounds of various intensities (range 0.63–2.66 log
cd m
). Each plotted point is a comparison of the elevation in
threshold for the averaged 500-nm and 510-nm stimuli and for the
averaged 380-nm and 390-nm stimuli at a single background light
intensity. The straight line is the best-fit linear regression
(slope=0.64; r
=0.94; P<0.01). The broken line shows the
relationship that would be expected if the UV and M cone
mechanisms had adapted in a univariant fashion. B Complete
increment-threshold functions obtained for a single animal. The
stimuli and adapting backgrounds were of the same nature as for
the data at the left. The intensity of the adaptation light is specified
as described in the legend of Fig. 9 with the additional assumption
that in this test circumstance the subject had an effective pupil
diameter of 4 mm. The best fit lines for the 500-/510-nm test lights
has a slope of 0.86 (r
=0.989), that for the UV test lights has a
slope value of 0.47 (r
Fig. 14 Results from a spectral discrimination test run on a degu.
The UV test light (peak 370 nm; half-bandwidth 30 nm) was
presented at various intensity levels (given as attenuations from
the maximum available light at that wavelength). A 500-nm
comparison light was used. The solid circles summarize the initial
performance (mean number of trials/intensity=360) of the animal
on this problem and they illustrate the strong salience of the
brightness differences between the 370-nm and 500-nm lights. The
point of predicted equal brightness between the two lights is
indicated by the vertical dashed line. The triangles show the average
asymptotic performance (50 trials/point) following extensive
training on the task. That successful performance is independent
of the relative brightness of the two lights shows the animal is
capable of making color discriminations
test sessions (a total of 6,400 to 13,000 trials) to reach
this stage. In any case, once this discrimination was
established the test light was progressively shifted in
steps of 5 or 10 nm toward the spectral location of the
comparison light. Figure 15 summarizes the results
obtained from three subjects (filled circles), all of whom
performed quite similarly. Plotted are the perfo rmances
achieved at the intensities calculated to make the test
lights equally bright to the comparison lights. Each
animal successfully discriminated the two lights until
the test light reached about 420 nm; for those wave-
lengths and long er values the discrimination failed. In a
second phase of this experiment the direct ion of the
discrimination was reversed, i.e., the test light was
500 nm and the comparison lights were set to 370 nm.
Brightness equations, training procedures and all
details were otherwise the same. Following an initial
training period all three subjects also mastered this
discrimination and Fig. 15 further shows the results
(open triangles) that were obt ained from shifting the
spectral position of the test light progressively toward
the shorter wavelengths. The result is an effective mir-
ror image of the first test in that the degus failed to
make the discrimination for test lights shorter than
about 425 nm. degus seem clearly capable of making
color discriminations.
These experiments reveal a number of aspects of the
visual adaptations achieved by O. degus. These can
be summarized succinctly as follows. The degu retina
contains a substantial number of cones (upwards of 3
million) of two spectral types, M and UV, present in
relative proportion of ab out 13:1; that ratio is similar to
that recorded for populations of M/L:S/UV cones in
many other mammals. The spectral absorption proper-
ties of both the two cone pigments (k
values of about
362 nm and 507 nm) and the rod pigment (k
of ca.
500 nm) are also quite similar to those earlier docu-
mented for common laboratory rodents such as mouse
(Jacobs et al. 1991; Sun et al. 1997; Yokoyama et al.
1998; Lyubarsky et al. 1999) and rat (Jacobs et al. 2001).
As indexed both by signals recorded from the degu outer
retina and from direct behavioral evaluations, the fa-
miliar transition from rod to cone-based vision can be
produced throug h the agency of increasing levels of light
adaptation. Relatively high levels of light adaptation are
required to fully effect this shift. Finally, like oth er
mammals with two cla sses of cone, this rodent can
apparently make some color discriminations.
Diurnal adaptations in the degu
Several features of our results reinforce the conclusion
that degus are adapted for a diurnal lifestyle. Most
obviously, their retinas contain a significant number
of cones reaching a peak density of more than 50,000
. By comparison, the peak co ne density in the
nocturnal rat is only about 7,000 mm
(Jacobs et al.
2001) while the crepuscular rabbit peaks at about
12,000 mm
(Juliusson et al. 1994). On the other side of
the coin, ground squirrels, one of the most consistently
diurnal of all mammals, show local cone density peaks at
values not greatly different than that for the degu
(Kryger et al. 1998). The cone/rod ratio of the degu is
also comparatively quite high, well above those for
nocturnal mammals though still far short of the cone/
rod ratios seen in retinas of strongly diurnal animals
such as ground squirrels (Kryger et al. 1998) and tree
shrews (Muller and Peichl 1989). The functional mea-
sures also indicate the presence of robust cone based
vision, e.g., degus continue to show behavioral dis-
criminations under comparatively high levels of light
adaptation while the ERG faithfully follows pulsed in-
puts presented at high temporal frequencies. Finally,
there is a clear attenuation of short-wavelength light by
filtering in the degu lens. Although varying greatly in
pattern and extent, lens pigmentation is classically ap-
parent in a wide range of diurnal animals while typically
absent from the lenses of nocturnal species (Dougla s and
Marshall 1999). On all of these grounds our measure-
ments support the idea that O. degus maintains clear
visual adaptations for diurnal life.
It is apparent both from ERG measurements and
behavioral discriminations that relatively high light
levels are required to complete the shift from rod-based
to cone-based vision in the degu. That point can perhaps
be most clearly appreciated through comparisons of
measurements made on degus and on nocturnal rodents.
For example, ERG spectral sensitivities recorded for
Fig. 15 Wavelength discrimination by the degu. The plotted points
are the means performances (±1 SD) achieved by three subjects
for various test wavelengths versus the 370-nm comparison light
(open triangles) and versus a 510-nm comparison light (filled
circles). These performance data are for lights calculated to be of
equal brightness for the degu. The horizontal dashed line indicates
chance performance for the number of test trials that were run
12.5-Hz flicker are still clearly rod dominated in the degu
(Fig. 7, bottom) but with the iden tical stimulus condi-
tions spectra recorded for the rat have been shifted to
cone control (Jacobs et al. 2001). Similarly, in behavioral
tests adapting light levels that are quite sufficient to
produce photopic spectral-sensitivity curves in both rats
and mice (Jacobs et al. 1999, 2001) support only scotopic
spectral sensitivity in the degu (Fig. 11). These results
seem somewhat surprising in light of the relatively large
cone representation in the degu. Although an explana-
tion for this seeming disparity is prese ntly lacking, this
observation does suggest that studies of photic control
of activity rhythms, as well as any direct studies of vi-
sion, should be pursued in cognizance of these require-
ments for cone vision. For example, ambient light levels
of 200–300 lx have previously been used to define the
‘‘light periods’’ in light-entrainment studies (Garcia-
Allegue et al. 1999). Our results indicate such light levels
are probably insufficient to produce a complete shifting
from rod to cone based vision and are, therefore, not
fully photopic for the degu.
Sensitivity to UV light and color vision
Although specific sensitivity to ultraviolet light has long
been known to be present in animals from a wide range
of taxa, it is only relatively recently that UV-selective
pigments were detected in mammalian retinas (Jacobs
1992). Among mammals, UV pigments are apparently
only found in rodents and, as noted above, their pres-
ence across this order seems far from universal. As yet,
the understanding of rodent UV pigments offers few
generalizations, either as to the phylogeny of these
pigments or to their functional roles. Heretofore, one
consistency was that UV pigments had only been doc-
umented for nocturnal rodents while their short-wave-
length alternatives, the S-cone pigments, were seen in the
retinas of strongly diurnal animals, such as various sci-
urids, or in crepuscular species like the guinea pig. The
degu, however, seems clearly diurnal and its retina
contains a population of UV cones. Evidently, there is
nothing inevitable about the linkage between rodent UV
cones and circadian lifestyle.
Only two of the four families of vertebrate cone op-
sins are represented in mammals where these two yield
the entire complement of mammalian M/L and UV/S
cone pigments, respectively (Bowmaker 1998). The
spectral peaks of mammalian UV/S pigments span a
range of some 80 nm and a variety of differences in
opsin structure have recen tly been proposed as respon-
sible for shifting spectral absorption across this range
(Yokoyama and Shi 2000; Hunt et al. 2001; Fasick et al.
2002). From a comparative analysis of the amino acid
sequences of vertebrate UV/S pigments, Hunt et al.
(2001) concluded that UV pigments represent the an-
cestral condition. According to them, UV pigments were
subsequently replaced in some mammals through an
accumulation of amino acid substitutions that shifted
the peaks of these pigments toward longer wavelengths
where they thus became the classical mammalian S-co ne
pigments. In this view, of all mammals only a subset of
the rodents has retained UV pigments. We know too
little about rodent photopigments at the present to say
whether that group includes the majority of rodents or
not. The current study does make clear that very dis-
tantly related rodents (octodonts and murids) share their
UV pigments and that the presence of UV or S pigments
cannot be predicted from any obvious consideration of
visual behaviors.
There has been much discussion of the utility of UV
vision (Goldsmith 1994; Hunt et al. 2001; Kevan et al.
2001; Honkavaara et al. 2002). To a considerable extent
this has been stimulated by early observations docu-
menting how bees utilize the UV coloration patterns of
flowers as guides to nectar locations and the subsequent
belief that similar compulsive linkages may be found in
other animals. But whereas specific roles for UV vision
can be assigned, or at least reasonably suggested, for
some cases, in many instances no unique visual function
appears clearly linked to an animal’s UV vision. Thus
far that certainly seems to be the case for UV sensitivity
in rodents. In the absence of a specific linkage to a visual
behavior, it is usually pointed out that, at the least,
the presence of UV cones expands sensitivity to short-
wavelength light beyond that available to retinas con-
taining only S cones. How this might specifically
enhance degu vision is unclear, but is worth noting that
in many natural daylight environments there is relatively
abundant UV light (Endler 1993) and that there are
specific increases in the ratio of 360-nm:520-nm light
coming from the sky during crepuscular periods (Hut
et al. 2000). The results from behavioral measurements
document that degus have relatively high sensitivity to
UV light (Fig. 12) and thus, potentially, UV light seems
likely to provide the basis for useful vision in natural
The presence of a UV pigment in conjunction with an
M pigment also sets the stage for a color vision capacity
and, with test conditions that obviate the use of lumi-
nance-related cues, degus are indeed able to successfully
discriminate between spectrally distinct stimuli (Fig. 15).
By definition, these are color discriminations. The ef-
fective nature of this capacity is that these rodents can
discriminate a range of short wavelengths from a range
of longer wavelengths. There is a rather abrupt transi-
tion between these two zones at around 420 nm. Al-
though the narrow band monochromatic lights used in
this test are unlike most natural stimuli, the results
suggest that, based on spectral differ ences alone, light
sources that are relatively UV rich (e.g., direct skylight),
or objects that are UV reflective, are potentially dis-
criminable from stimuli whose spectral signatures slope
upward toward the longer wavelengths.
Whether and how degus might exploit color vision in
their normal environments remains an open question.
There are two cautionary observations about degu color
vision coming from our experiments. First, as noted
above, it required considerable training before our
subjects succeeded at making cle ar color discrimina-
tions. Second, we observed that once the animals had
shown success at the initial discrimination (UV light
positive versus visible light negative) then, when the di-
rection of the discrimination was reversed (visible light
now positive) the animals again required a similarly long
training period before succeeding at this second dis-
crimination. One could interpret this as implying that
the animal had learned an absolute discrimination (e.g.,
to simply select stimuli having longer wavelength con-
tent) rather than lea rning to pick the panel having a
different color. Both of these observations may simply
reflect the artificial nature of the task or perhaps they
imply that color differences do not normally have much
salience for degus. In any case, the experiments show
that degus formally have dichromatic color vision,
although its utility remains to be discovered.
Acknowledgements We thank T.M. Lee and B.A. Tate for pro-
viding animals and J.A. Endler for loaning us equipment. This
research was supported by a grant from the National Eye Institute
(EY02052). All animal care and experimental procedures were in
accordance with institutional animal care and use guidelines and
with the Principles of animal care, publication No.86-23, revised
1985 of the National Institutes of Health.
Ahnelt PK, Kolb H (2000) The mammalian photoreceptor mosaic-
adaptive design. Prog Ret Eye Res 19:711–770
Bowmaker JK (1998) Evolution of colour vision in vertebrates. Eye
Calderone JB, Fenwick JA, Jacobs GH (2001) Photoreceptors,
photopigments and visual capacities of a diurnal rodent,
Octodon degus. Invest Ophthal Vis Sci 42:S178
Chu MI, Zack DJ, Wang Y, Nathans J (1994) Murine and bovine
blue cone pigment genes: cloning and characterization of the S
family of visual pigments. Genomics 21:440–443
Douglas RH, Marshall NJ (1999) A review of vertebrate and in-
vertebrate ocular filters. In: SN Archer, MBA Djamgoz, ER
Loew, JC Partridge, S Vallerga (eds) Adaptive mechanisms in
the ecology of vision. Kluwer, Dordrecht, pp 95–162
Endler JA (1993) The color of light in forests and its implications.
Ecol Monogr 63:1–27
Fasick JI, Applebury ML, Oprian DD (2002) Spectral tuning in the
mammalian short-wavelength sensitive cone pigments.
Biochemistry 41:6860–6865
Fulk GW (1976) Notes on the activity, reproduction and social
behaviour of Octodon degus. J Mammal 57:495–505
Garcia-Allegue R, Lax P, Madariaga AM, Madrid JA (1999)
Locomotor and feeding activity rhythms in a light-entrained
diurnal rodent, Octodon degus. Am J Physiol 267:R523–R531
Goldsmith TH (1994) Ultraviolet receptors and color vision:
evolutionary implications and a dissonance of paradigms. Vis
Res 34:1479–1487
Gorgels TGMF, Norren D van (1992) Spectral transmittance of the
rat lens. Vision Res 32:1509–1512
Govardovskii VI, Fyhrquist N, Reuter T, Kuzmin DG, Donner K
(2000) In search of the visual pigment template. Vis Neurosci
Hollins M, Alpern M (1973) Dark adaptation and visual pigment
regeneration. J Gen Physiol 62:430–477
Honkavaara J, Koivula M, Korpimaki E, Siitari H, Viitala J (2002)
Ultraviolet vision and foraging in terrestrial vertebrates. Oikos
Hunt DM, Wilkie SE, Bowmaker JK, Poopalasundaram S (2001)
Vision in the ultraviolet. Cell Mol Life Sci 58:1583–1598
Hunt S, Cuthill IC, Bennett ATD, Church SC, Partridge JC (2001)
Is the ultraviolet waveband a special communication channel in
avian mate choice? J Exp Biol 204:2499–2507
Hut RA, Scheper A, Daan S (2000) Can the circadian system of a
diurnal and a nocturnal rodent entrain to ultraviolet light?
J Comp Physiol A 186:707–715
Jacobs GH (1983) Within-species variations in visual capacity
among squirrel monkeys (Saimiri sciureus): sensitivity differ-
ences. Vision Res 23:239–248
Jacobs GH (1992) Ultraviolet vision in vertebrates. Am Zool
Jacobs GH (1993) The distribution and nature of colour vision
among the mammals. Biol Rev 68:413–471
Jacobs GH, Neitz J (1987) Inheritance of color vision in a New
World monkey (Saimiri sciureus). Proc Natl Acad Sci USA
Jacobs GH, Neitz J, Deegan JF II (1991) Retinal receptors in ro-
dents maximally sensitive to ultraviolet light. Nature 353:655–
Jacobs GH, Deegan JF II (1994) Spectral sensitivity, photopig-
ments and color vision of the guinea pig (Cavia porcellus).
Behav Neurosci 108:993–1004
Jacobs GH, Neitz J, Krogh K (1996) Electroretinogram flicker
photometry and its applications. J Opt Soc Am A 13:641–648
Jacobs GH, Fenwick JC, Calderone JB, Deeb SS (1999) Human
cone pigment expressed in transgenic mice yields altered vision.
J Neurosci 19:3258–3265
Jacobs GH, Fenwick JA, Williams GA (2001) Cone-based vision of
rats for ultraviolet and visible lights. J Exp Biol 204:2439–2446
Jiao Y-Y, Lee TM, Rusak B (1999) Photic responses of suprach-
iasmatic area neurons in diurnal degus (Octodon degus) and
nocturnal rats (Rattus norvegicus). Br Res 817:93–1013
Juliusson B, Bergstrom A, Rohlich P, Ehinger B, Veen T van, Szel
A (1994) Complementary cone fields of the rabbit retina. Invest
Ophthal Vis Sci 35:811–818
Kevan PG, Chittka L, Dyer AG (2001) Limits to the salience of
ultraviolet: lessons from colour vision in bees and birds. J Exp
Biol 204:2571–2580
King JA (1956) Social relations of the guinea pig under semi-
natural conditions. Ecology 37:221–228
Kryger Z, Galli-Resta L, Jacobs GH, Reese BE (1998) The
topography of rod and cone photoreceptors in the retina of
the ground squirrel. Vis Neurosci 15:685–691
LaVail MM (1976) Survival of some photoreceptors in albino rats
following long-term exposure to continuous light. Invest
Ophthal Vis Sci 15:64–70
Lee TM, Labyak SE (1997) Free-running rhythms and light- and
dark-pulse phase-response curves for diurnal Octodon degus.
Am J Physiol 273:R278–R286
Lyubarsky AL, Falsini B, Pennesi ME, Valentini P, Pugh EN Jr
(1999) UV- and midwave-sensitive cone-driven retinal responses
of the mouse: a phenotype for coexpression of cone photopig-
ments. J Neurosci 19:442–455
Macdonald D (ed) (2001) The new encyclopedia of mammals.
Oxford University Press, Oxford
Muller B, Peichl L (1989) Topography of cones and rods in the tree
shrew retina. J Comp Neurol 282:581–594
Parry JWL, Bowmaker JK (2002) Visual pigment coexpression in
guinea pig cones: a microspectrophotometric study. Invest
Ophthal Vis Sci 43:1662–1665
Paupoo AAV, Mahroo OAR, Friedburg C, Lamb TD (2000)
Human cone photoreceptor responses measured by the elec-
troretinogram a-wave during and after exposure to intense
illumination. J Physiol (Lond) 529:469–482
Peichl L, Gonzalez-Soriano J (1994) Morphological types of hori-
zontal cell in rodent retinae: a comparison of rat, mouse, gerbil
and guinea pig. Vis Neurosci 11:501–517
Sun H, Macke JP, Nathans J (1997) Mechanisms of spectral tuning
in the mouse green cone pigment. Proc Natl Acad Sci USA
Szel A, Rohlich P (1992) Two cone types of rat retina detected by
anti-visual pigment antibodies. Exp Eye Res 55:47–52
Szel A, Rohlich P, Caffe AR, Veen T van (1996) Distribution of
cone photoreceptors in the mammalian retina. Microsc Res
Tech 35:445–462
Szel A, Lukats A, Fekete T, Szepessy Z, Rohlich P (2000) Photo-
receptor distribution in the retinas of subprimate mammals.
J Opt Soc Am A 17:568–579
Weale RA (1988) Age and the transmittance of the human crys-
talline lens. J Physiol (Lond) 395:577–587
Xu J, Pokorny J, Smith VC (1997) Optical density of the human
lens. J Opt Soc Am A 14:953–960
Yokoyama S, Shi Y (2000) Genetics and evolution of ultraviolet
vision in vertebrates. FEBS Lett 486:167–172
Yokoyama S, Radlwimmer FB, Kawamura S (1998) Regeneration
of ultraviolet pigments of vertebrates. FEBS Lett 423:155–158
Yolton RL, Yolton DP, Renz J, Jacobs GH (1974) Preretinal
absorbance in sciurid eyes. J Mammal 55:14–20
Zhang Y, Brainard GC, Zee PC, Pinto LH, Takahashi JS, Turek
FW (1998) Effects of aging on lens transmittance and retinal
input to the suprachiasmatic nucleus in golden hamsters.
Neurosci Lett 258:167–170
Zigman S, Paxhia T, Waldron W (1985) Biochemical features of the
grey squirrel lens. Invest Ophthal Vis Sci 26:1075–1082
... O. degus can make color discriminations between ultraviolet and visible lights, thanks to the presence of only two types of cones: M cones (507 nm) and UV-sensitive S cones (362 nm) in a 13:1 ratio. Interestingly, thanks to the presence of S cones O. degus lens selectively absorb short wavelength light and shows a progressive increase in optical density as a function of age (Jacobs et al., 2003). In terms of functionality, detection of ultraviolet light is a particularly useful retinal skill in natural environments, as well as for its behavioral significance in communication (Jacobs et al., 2003). ...
... Interestingly, thanks to the presence of S cones O. degus lens selectively absorb short wavelength light and shows a progressive increase in optical density as a function of age (Jacobs et al., 2003). In terms of functionality, detection of ultraviolet light is a particularly useful retinal skill in natural environments, as well as for its behavioral significance in communication (Jacobs et al., 2003). In addition, the ventral part of the degus body, which is usually exposed during alert call, shows significant high UV reflectance, whereas the dorsal part of the body exhibits low UV reflectance, which helps the animal to hide from predators. ...
Integrating the multifactorial processes co-occurring in both physiological and pathological human conditions still remains one of the main challenges in translational investigation. Moreover, the impact of age-associated disorders has increased, which underlines the urgent need to find a feasible model that could help in the development of successful therapies. In this sense, the Octodon degus has been indicated as a 'natural' model in many biomedical areas, especially in ageing. This rodent shows complex social interactions and high sensi-tiveness to early-stressful events, which have been used to investigate neurodevelopmental processes. Interestingly , a high genetic similarity with some key proteins implicated in human diseases, such as apolipoprotein-E, β-amyloid or insulin, has been demonstrated. On the other hand, the fact that this animal is diurnal has provided important contribution in the field of circadian biology. Concerning age-related diseases, this rodent could be a good model of multimorbidity since it naturally develops cognitive decline, neurodegenerative histopathological hallmarks, visual degeneration, type II diabetes, endocrinological and metabolic dysfunctions, neoplasias and kidneys alterations. In this review we have collected and summarized the studies performed on the Octodon degus through the years that support its use as a model for biomedical research, with a special focus on ageing.
... Light was measured using a calibrated spectroradiometer (BLACK-Comet-SR spectrometer; StellarNet Inc., United States). Spectral power densities were converted to effective irradiance for each of the rat photoreceptors (Lucas et al., 2014) by multiplying by the normalized in vivo spectral sensitivity for each photopigment and correcting for the prereceptoral filtering (Jacobs et al., 2003) across the spectrum (Figures 1A-C). The irradiance of filtered light was calculated by applying a transmittance curve for the amber lens ( Figure 1A) on top of rat lens. ...
... The main aim of this work was to verify whether light induced neuronal activity in retinorecipient brain structures is affected by removing short wavelengths (UV-blue) from the polychromatic white light spectrum. The amber lens filters 90% of short wavelengths (cut off at 525 nm), thus mostly reduces activity of S-cones and melanopsin, having maximal sensitivity at approximately 359 and 480 nm, respectively, (Jacobs et al., 2001(Jacobs et al., , 2003Lucas et al., 2014). Intuitively one would expect that such UV/blue light blockade should have profound consequences on the light induced activity. ...
Full-text available
The amount and spectral composition of light changes considerably during the day, with dawn and dusk being the most crucial moments when light is within the mesopic range and short wavelength enriched. It was recently shown that animals use both cues to adjust their internal circadian clock, thereby their behavior and physiology, with the solar cycle. The role of blue light in circadian processes and neuronal responses is well established, however, an unanswered question remains: how do changes in the spectral composition of light (short wavelengths blocking) influence neuronal activity? In this study we addressed this question by performing electrophysiological recordings in image (dorsal lateral geniculate nucleus; dLGN) and non-image (the olivary pretectal nucleus; OPN, the suprachiasmatic nucleus; SCN) visual structures to determine neuronal responses to spectrally varied light stimuli. We found that removing short-wavelength from the polychromatic light (cut off at 525 nm) attenuates the most transient ON and sustained cells in the dLGN and OPN, respectively. Moreover, we compared the ability of different types of sustained OPN neurons (either changing or not their response profile to filtered polychromatic light) to irradiance coding, and show that both groups achieve it with equal efficacy. On the other hand, even very dim monochromatic UV light (360 nm; log 9.95 photons/cm2/s) evokes neuronal responses in the dLGN and SCN. To our knowledge, this is the first electrophysiological experiment supporting previous behavioral findings showing visual and circadian functions disruptions under short wavelength blocking environment. The current results confirm that neuronal activity in response to polychromatic light in retinorecipient structures is affected by removing short wavelengths, however, with type and structure – specific action. Moreover, they show that rats are sensitive to even very dim UV light.
... O. degus can make color discriminations between ultraviolet and visible lights, thanks to the presence of only two types of cones: M cones (507 nm) and UV-sensitive S cones (362 nm) in a 13:1 ratio. Interestingly, thanks to the presence of S cones O. degus lens selectively absorb short wavelength light and shows a progressive increase in optical density as a function of age (Jacobs et al., 2003). In terms of functionality, detection of ultraviolet light is a particularly useful retinal skill in natural environments, as well as for its behavioral significance in communication (Jacobs et al., 2003). ...
... Interestingly, thanks to the presence of S cones O. degus lens selectively absorb short wavelength light and shows a progressive increase in optical density as a function of age (Jacobs et al., 2003). In terms of functionality, detection of ultraviolet light is a particularly useful retinal skill in natural environments, as well as for its behavioral significance in communication (Jacobs et al., 2003). In addition, the ventral part of the degus body, which is usually exposed during alert call, shows significant high UV reflectance, whereas the dorsal part of the body exhibits low UV reflectance, which helps the animal to hide from predators. ...
Integrating the multifactorial processes co-occurring in both physiological and pathological human conditions still remains one of the main challenges in translational investigation. Moreover, the impact of age-associated disorders has increased, which underlines the urgent need to find a feasible model that could help in the development of successful therapies. In this sense, the Octodon degus has been indicated as a ‘natural’ model in many biomedical areas, especially in ageing. This rodent shows complex social interactions and high sensi-tiveness to early-stressful events, which have been used to investigate neurodevelopmental processes. Interest-ingly, a high genetic similarity with some key proteins implicated in human diseases, such as apolipoprotein-E, β-amyloid or insulin, has been demonstrated. On the other hand, the fact that this animal is diurnal has provided important contribution in the field of circadian biology. Concerning age-related diseases, this rodent could be a good model of multimorbidity since it naturally develops cognitive decline, neurodegenerative histopathological hallmarks, visual degeneration, type II diabetes, endocrinological and metabolic dysfunctions, neoplasias and kidneys alterations. In this review we have collected and summarized the studies performed on the Octodon degus through the years that support its use as a model for biomedical research, with a special focus on ageing.
... These arguments, however, are difficult to generalize because each species has presumably taken its own strategy to increase the fitness in its natural habitat, leading to convergent and divergent evolution. On the one hand, UV sensitivity was identified in some mammals that live in a different visual environment than mice, including diurnal small animals such as the degu and gerbil [61,64,65] and even large animals such as the Arctic reindeer [66]. On the other hand, some species showing a similar behavioral pattern as mice do not have the dorsoventral division of the retinal function [12][13][14]. ...
Full-text available
The mouse has dichromatic color vision based on two different types of opsins: short (S)- and middle (M)-wavelength-sensitive opsins with peak sensitivity to ultraviolet (UV; 360 nm) and green light (508 nm), respectively. In the mouse retina, cone photoreceptors that predominantly express the S-opsin are more sensitive to contrasts and denser towards the ventral retina, preferentially sampling the upper part of the visual field. In contrast, the expression of the M-opsin gradually increases towards the dorsal retina that encodes the lower visual field. Such a distinctive retinal organization is assumed to arise from a selective pressure in evolution to efficiently encode the natural scenes. However, natural image statistics of UV light remain largely unexplored. Here we developed a multi-spectral camera to acquire high-quality UV and green images of the same natural scenes, and examined the optimality of the mouse retina to the image statistics. We found that the local contrast and the spatial correlation were both higher in UV than in green for images above the horizon, but lower in UV than in green for those below the horizon. This suggests that the dorsoventral functional division of the mouse retina is not optimal for maximizing the bandwidth of information transmission. Factors besides the coding efficiency, such as visual behavioral requirements, will thus need to be considered to fully explain the characteristic organization of the mouse retina.
... Experiments were performed in pieces of retina perfused continuously with Ringer's AMES medium and obtained from diurnal dichromate rodent Octodon degus (degu). Degus have ~2.9 million cones with middle wavelength-sensitive (M); 221,000 cones with short-ultraviolet wavelength (UV) sensitive (S), and ~6.5 million rods (Chavez et al, 2003;Jacobs et al, 2003;Palacios & Lee, 2013;Peichl, 2005). Data is recollected from 3 different animals and 6 pieces of retina in total. ...
Full-text available
This exploratory project report provides a study of the retina in response to spontaneous and evoked pattern of flashes for different time cycles. These patterns induce changes at individual neural level and network level, while global efficiency, a topological network measure, presents robustness and degenerate states across time. This report also discusses some alternatives to explain changes observed and how it could be related to oscillatory plasticity mechanism.
... If this is the case, they share dichromatic colour vision with other Afrotheria such as manatees (Griebel & Schmid, 1996) and most probably elephants, rock hyraxes and tenrecs (Peichl et al., 2000;Yokoyama et al., 2005;Jacobs, 2009;Emerling & Springer, 2014) as well as with most mammals (Jacobs, 2009). UV vision is used by various vertebrates, for instance by some fishes, amphibians, reptiles, birds and bats, as well as some rodents (Jacobs & Deegan, 1994;Jacobs et al., 2001Jacobs et al., , 2003Jacobs et al., , 2004Yokoyama, 2002;Winter et al., 2003;Siebeck et al., 2010;Ödeen & Håstad, 2013). ...
Little research has been conducted on the senses of sengis (elephant-shrews, Macroscelidea, Afrotheria, Mammalia); behavioural investigations about the animals' vision are completely missing. Other Afrotheria (manatees, elephants, tenrecs, rock hyraxes) are dichromats, having two types of cone photoreceptors in the retina. We tested the hypotheses of dichromatic colour vision in sengis. With choice experiments, we examined the potential of two sengi species to discriminate between trained colours (blue, green, red) and different shades of grey, and to differentiate between trained UV-reflecting (high UV reflectance) and UV-absorbing (low UV reflectance) plates. Both sengi species recognised blue and green and could distinguish them from all shades of grey. The ability to perceive red and ultraviolet could not be proven. Thus, the sengis are most likely dichro-mats based on green and blue sensitive receptor types. We discuss protective mechanisms in the lens as a UV filter explaining maximum sensitivity in the UV suggested for another sengi species after in vitro expression assays.
Full-text available
Phobia against spiders or snakes is common in humans, and similar phobia-like behaviors have been observed in non-human animals. Visual images of snakes elicit phobia in humans, but sensory modalities that cause snake aversion in non-human animals are not well examined. In this study, we examined visually induced snake aversion in two rodent species. Using a three-compartment experimental chamber, reactions to images of snakes were compared between the diurnal precocious rodent Octodon degus and nocturnal laboratory mice. The snakes whose images were presented do not live in the original habitats of degus or mice. Snake aversion was assessed by presenting snake vs. no-image, snake vs. flower, snake vs. degu, and snake vs. mouse images. The time spent in a compartment with the snake image and with the non-snake images were measured. Degus avoided images of snakes in every tests. In contrast, mice did not display snake aversion. Degus are diurnal animals, i.e., visual information is important for their survival. Since mice are nocturnal, visual information is less important for survival. Such behavioral differences in the two species may explain the difference in visually induced aversion to snakes. A principal component analysis of the stimulus images suggests that elementary cues, such as color, do not explain the differences in the species’ aversion to snakes. Finally, snake aversion in degus suggests that aversion is innate, since the animals were born and raised in a laboratory.
Full-text available
The dorsal lateral geniculate nucleus (dLGN) is the main neuronal station en route to higher visual areas. It receives information about environmental light from retinal photoreceptors whose sensitivity peaks are distributed across a visible spectrum. Here, using electrophysiological multichannel recordings in vivo combined with different light stimulations, we investigated short wavelength contribution to the dLGN responses to light and irradiance coding. The results showed that the majority of dLGN cells responded evenly to almost all wavelengths from the 340 – 490 nm spectrum; however, some cells representing extremes of unimodal distribution of Blue-UV index were specialized in the reception of blue or UV light. Moreover, by using alternate yellow and monochromatic light stimuli from blue - UV range, we also assessed the relative spectral contribution to rat dLGN responses to light. Finally, we observed no clear changes in the irradiance coding property of short wavelength-deficient light stimuli, however we noticed a distortion of the coding curves manifested by a significant drop in measure of fit after using short wavelength blocking filter. In conclusion, our data provide the first electrophysiological report on dLGN short wavelength-induced responses under changing light conditions and suggest the presence of colour opponent cells in the rat dLGN.
Full-text available
Knowledge of a species’ visual system has far reaching implications that affect our understanding of a species’ ecology and evolutionary history. As a model taxon, the heteromyid rodent genus Dipodomys has been valuable in elucidating patterns and mechanisms in biomechanics, ecology, adaptive physiology, biogeography, and more. Although studied extensively, the visual system of Dipodomys has not been described beyond anecdotal mention of their large eyes. Here, the transmittance parameters of the cornea and lens of Ord’s kangaroo rat (Dipodomys ordii) were analyzed and photoreceptor proteins (opsins) expressed in the retina were identified with immunohistochemical (IHC) labeling. Retina maps were constructed to illustrate the relative densities of photoreceptor cells expressing short wavelength (SWS1) opsins, middle/long wavelength (MW/LW) opsins, and rhodopsin (RH1). The retina of D. ordii has variable densities of SWS1 opsin with the highest density being ventral to the optic nerve, high density of MW/LW opsin, and uniform distribution and high density of RH1 across the retina. Our results suggest that D. ordii has a UV-sensitive visual system. Composition and densities of MW/LW- and SWS1-expressing cells resemble that of a crepuscular/diurnal species thereby supporting previous authors who have reported such activity patterns. Uniform retinal distribution of RH1 indicates visual acuity at night, also confirming the paradigm of D. ordii as primarily a nocturnal species and suggesting visual acuity at all times of the day in the species. These results demonstrate not only that the species is capable of UV vision and has a retina characteristic of a diurnal mammal, but that many previously unknown photic niche selective advantages likely have shaped the evolution and ecology of this model taxon.
Full-text available
Recent years have seen an upsurge of interest in the study of ultraviolet vision in vertebrates. The presence of retinal photopigments that allow for ultraviolet vision appear to be relatively common among birds and fishes, and there is evidence that such pigments are to be found in some species from each of the following classes: amphibians, reptiles, and mammals. Following a review of the distribution and nature of ultraviolet vision in vertebrates the issue of the utility of this capacity is discussed.
Full-text available
Tetrachromatic colour vision, based on four ‘main’ colours and their combinations, is probably the original colour vision in terrestrial vertebrates. In addition to human visible waveband of light (400–700 nm) and three main colours, it also includes the near ultraviolet part of light spectrum (320–400 nm). The ecological importance of ultraviolet (UV) vision in animals has mainly been studied in the context of intra- and inter-sexual signalling, but recently the importance of UV vision in foraging has received more attention. Foraging animals may use either UV cues (reflectance or absorbance) of food items or UV cues of the environment. So far, all diurnal birds studied (at least 35 species), some rodents (4 species), many reptilians (11 species) and amphibians (2 species) are likely able to see near UV light. This probably allows e.g. diurnal raptors as well as frugivorous, nectarivorous and insectivorous birds to use foraging cues invisible to humans. The possible role of UV cues and existing light conditions should be taken into account when food selection of vertebrate animals is studied, particularly, in experiments with artificial food items.
Full-text available
Forests exhibit much variation in light environments, and this can affect communication among animals, communication between animals and plants, photosynthesis, and plant morphogenesis. Light environments are caused by , and can be predicted from, the geometry of light paths, the weather conditions, and the time of day. The structure of forests leads to four major light habitats when the sun is not blocked by clouds: forest shade, woodland shade, small gaps, and large gaps. These are characterised by yellow-green, blue-gray, reddish, and white ambient light spectra, respectively. When the sun is blocked by clouds, the spectra of these four habitats converge on that of large gaps and open areas, so the single light environment during cloudy weather will be called open/cloudy. An additional light environment(early/late) is associated with low sun angles(near dawn or dusk; it is purplish) Each light environment is well defined and was found in forests of Trinidad, Panama, Costa Rica, Australia, California, and Florida. Scattered literature references suggest that fruits depend on the interaction between ambient light colour and the reflectance colour of the animal or plant parts. As a result an animal or plant may have a different appearance in each environment, i.e. a colour pattern may be relatively cryptic in some light environments while relatively conspicuous in others. This has strong implications for the joint evolution of of visual signals and vision, as well as microhabitat choice. Plant growth and form may also be affected by variation in the colour of forest light.
PURPOSE. To determine the visual pigment content of the rods and cones of the guinea pig (Cavia porcellus) and to quantify the level of coexpression of pigments within individual cones. METHODS. Microspectrophotometry was used to measure the absorbance spectrum of visual pigments in individual rods and cones from three retinal regions: dorsal, ventral, and a subequatorial transition zone. Partial bleaching was used to establish whether two spectrally distinct visual pigments were present within a single cone. RESULTS. Rods possessed a pigment with a wavelength of maximum absorbance close to 500 nm. A population of middle-wave-sensitive cones (lambda(max) cones) contained a pigment with lambda(max) at approximately 530 nm, and a short-wave-sensitive cone population (S cones) contained a pigment with lambda(max) close to 400 nm. The majority of cones in all regions were M cones. Approximately 10% of cones in the transition region were found to coexpress the M and S cone pigments in a ratio of approximately 4:1. Coexpression was not detected in S cones. In C. porcellus, coexpression of cone pigments occurs in a small number of cells but is biased in favor of the M pigment. Given the relatively low level of coexpression, detectable in only approximately 10% of the cones in the transition region, it is unlikely to cause any significant detriment to dichromatic color vision.
Behavioral discrimination tests and electroretinogram (ERG) flicker photometry were used to measure spectral sensitivity and to define the spectral mechanisms of the guinea pig (Cavia porcellus). Results from these 2 approaches converge to indicate that guinea pig retinas contain rods with peak sensitivity of about 494 nm and 2 classes of cone having peak sensitivities of about 429 nm and 529 nm. The presence of 2 classes of cones suggests a retinal basis for a color vision capacity. Behavioral tests of color vision were conducted that verified this prediction: Guinea pigs have dichromatic color vision with a spectral neutral point centered at about 480 nm. The cone pigment complement of the guinea pig is different from that known to characterize other rodents.
Data on activity and social behavior of the Chilean degu, Octodon degus, were gathered by direct observation of animals, some of which had been marked for individual recognition. Data from autopsies and external inspection of trapped animals suggested that most reproduction occurs in September at the latitude of Santiago. Degus are diurnal and show morning and evening activity peaks. Social organization is based on group territories, at least during the period after emergence of the young. Mound building (collecting a pile of sticks, stones, and cow dung) was associated with territorial marking. Females of the same social group may rear their young in a common nest burrow. Octodon degus burrows are sometimes used by Abrocoma bennetti, a similar sized rodent, and on two occasions nest burrows were found to be shared by young and mothers of both species.
We report here the regeneration of the visual pigments of mouse, rat, goldfish and pigeon, which have wavelengths of maximal absorption at 359 nm, 358 nm, 359 nm, and 393 nm, respectively. The construction and functional assays of the ultraviolet or near-ultraviolet pigments from a wide range of vertebrate species will allow us to study the molecular bases of ultraviolet vision for the first time.
Many vertebrates achieve UV vision through a class of photoreceptors in the retina containing UV-sensitive visual pigments with optimal light absorption (λ max ) at 360∼370 nm. UV vision has been used for such basic behaviors as communication, foraging, and mating. Despite its biological importance, the molecular bases and the evolution of UV vision are not well understood. In this study, we first identified the major amino acid residues that determine the spectral sensitivity of UV pigments in vertebrates. We also provided strong supportive evidence that UV pigments, unlike all the other visual pigments, are based on an unprotonated Schiff base chromophore. We then inferred the amino acid sequences of the ancestral pigments of UV/violet pigments in vertebrates. These ancestral pigments were then reconstructed by introducing the necessary mutations into the contemporary pigments and their absorption spectra evaluated by using an in vitro assay. Our results demonstrated that the common ancestor of vertebrates and most other ancestors had UV-sensitive pigments, and most contemporary UV pigments maintained the ancestral function. The ancestral pigments of birds achieved violet sensitivity by four amino acid substitutions, and some extant avian species regained UV vision by a change at another amino acid site. In conclusion, this study not only revealed the molecular mechanisms for spectral tuning in UV pigments, but also elucidated the evolutionary history of UV vision in vertebrates.