Anthro-19.qxd 12/13/03 7:19 PM Page 477
CHAPTER NINETEEN
The Tarsier Fovea:
Functionless Vestige or
Nocturnal Adaptation?
Callum F. Ross
THE FOVEA AND HAPLORHINE/ANTHROPOID ORIGINS
The retinal fovea is a specialization of the visual system that is widespread among
vertebrates, but among mammals is restricted to haplorhine primates. All anthropoids have foveae, although that of Aotus the nocturnal owl monkey is morphologically distinct and appears to be degenerate (Webb and Kaas, 1976). The
foveae of tarsiers are reported to be variably (Castenholtz, 1965 Woollard, 1925),
or universally present (Hendrickson, personal communication). Foveae are occasionally present in galagos (one Otolemur crassicaudatus and one Galago senegalensis out of 18 animals [De Bruyn et al., 1980]) (Stone and Johnston, 1981),
although they are less well developed. Reports of foveae in Lemuridae (Lemur
catta and Hapalemur griseus, Pariente, 1975, 1979) are based on ophthalmoscopic investigations and remain to be confirmed using histological techniques.
The retinal fovea of tarsiers and anthropoids is of interest for the study of
anthropoid origins because it may be a synapomorphy of a tarsier–anthropoid
clade (Haplorhini) exclusive of other primates (Cartmill, 1980; Kay et al.,
1997; Martin, 1990; Ross, 1996, 2001; Ross et al., 1998). As such, the fovea
Callum F. Ross ● Department of Anatomical Sciences, Health Sciences Center, Stony Brook
University, Stony Brook, NY 11794–8081
477
Anthro-19.qxd 12/13/03 7:19 PM Page 478
478
Callum F. Ross
may tell us about the ecological context, particularly the diel periodicity
(i.e., nocturnal or diurnal), in which haplorhines arose and diversified.
Although most extant and all basal anthropoids are diurnal (Heesy and Ross,
2002; Kay and Cartmill, 1977; Kay and Kirk, 2000), both extant tarsiers and
the likely sister taxon of the tarsier–anthropoid clade (omomyiforms) are
exclusively nocturnal.1 Consequently, the last common ancestor of tarsiers and
anthropoids is most parsimoniously reconstructed as nocturnal (Heesy and
Ross, 2001). In contrast, some workers hypothesize that the haplorhine fovea
evolved in a diurnal lineage (Le Gros Clark, 1959). Cartmill proposed that the
fovea evolved as an adaptation for improved visual acuity and that it persisted
in the lineage leading to extant Tarsius when that lineage reverted to nocturnality (Cartmill 1980; Kay et al., 1997; Martin 1990; Ross, 1996, 2000). In
support of this argument, a fovea is most often found in diurnal animals
(Cartmill, 1980), tarsiers lack a tapetum, a structure typically present in nocturnal mammals, suggesting that the tarsier lineage lost its tapetum in a diurnal ancestor (Cartmill, 1980; Martin, 1973, 1979), and a fovea may be absent
in some tarsier individuals (Castenholtz, 1965; Woollard, 1925).
This lack of consilience (Lee and Doughty, 1997) between the activity pattern optimized for the tarsier–anthropoid node (nocturnality) and that
expected to be associated with the fovea also present at that node (diurnality),
can be resolved in a number of ways. The functional and comparative arguments mustered by Martin and Cartmill might be incorrect. The haplorhine
fovea may have arisen in a nocturnal environment (Figure 1B and D), be
retained in extant populations of Tarsius because it confers an adaptive advantange under scotopic, or low light, conditions, and have been co-opted (or
exapted) for the adaptive benefits it confers under the diurnal conditions under
which most anthropoids live. Alternately, Martin and Cartmill’s argument
might be correct, the fovea may have arisen in a diurnal ancestor (Figure 1A
and C), the fovea of extant tarsiers might have been co-opted (or exapted) for
adaptive benefits it confers under nocturnal conditions, and the fovea of
anthropoids might have been retained in extant anthropoids because of its
adaptive value under diurnal conditions.
How to choose between these possible scenarios? If one could discriminate
traits that are maintained in lineages by natural selection from those retained
1
Here I follow Ross et al. (1998) in excluding the diurnal form Rooneyia from Omomyiformes,
instead placing it as Primates incerta sedis.
Anthro-19.qxd 12/13/03 7:19 PM Page 479
479
An
t
hr
op
oi
de
a
O
m
om
yi
fo
rm
es
Ta
rs
iu
s
An
t
hr
op
oi
de
a
O
m
om
yi
fo
rm
es
Ta
rs
iu
s
Tarsier Fovea
Nocturnal
Diurnal
Fovea
Diurnal
Fovea
Diurnal
ea
id
po
ro
th
Nocturnal
Fovea
C
An
O
m
om
yi
fo
rm
es
Ta
rs
iu
s
An
th
ro
po
id
ea
B
O
m
om
yi
fo
rm
es
Ta
rs
iu
s
A
D
Fovea
Diurnal
Figure 1. Cladograms of alternate hypotheses of haplorhine (anthropoid–tarsier–
omomyid) relationships and the evolution of activity pattern and retinal fovea assuming that the primitive condition for Primates is nocturnal and afoveate. (A) Phylogeny
advocated by Cartmill (1980), Martin (1990), Kay et al. (1997), Ross et al. (1998)
with most parsimonious optimization of characters. (B) Same phylogeny as in A, but
with character state evolution hypothesized by Cartmill (1980) and Ross (1996,
2000). (C) Alternate phylogeny similar to that advanced by some workers (e.g., Szalay
and Delson, 1979) with Tarsius deriving from some omomyiform lineage. Most parsimonious pattern of character state evolution is illustrated. (D) Same phylogeny as in
C but with diurnality, loss of tapetum and origin of fovea in stem lineage leading to
the last common ancestor of tarsiers and anthropoids. This model assumes that
Cartmill’s evolutionary scenario is correct but his phylogeny is incorrect.
by “phylogenetic inertia,” this question might be resolved. For example, if the
anthropoid fovea evinces traits characteristic of “phylogenetic inertia” and the
tarsier fovea appears to be adaptive, the “nocturnal origin” of both foveae and
crown Haplorhini would be supported. Conversely, if the anthropoid fovea is
adaptive and the tarsier fovea is “inertial,” then “diurnal origins” of both
foveae and crown Haplorhini are indicated. Traits indicative of phylogenetic
inertia might include variable presence in the extant populations, lack of
Anthro-19.qxd 12/13/03 7:19 PM Page 480
Callum F. Ross
480
A
visual axis
optic axis
α
B
GCL
IPL
INL
OPL
ONL
RCL
RE
Ch
Figure 2. Diagram of anatomy of tarsier eye and retina. (A) Eye of Tarsius, redrawn
from Castenholtz (1984). Angle ␣ was calculated using the diameter of the lens and the
distance from the center of the lens to the retina (Castenholtz, 1984). (B) Photograph of
histological section of retina around and at the fovea of Tarsius spectrum. Abbreviations:
Ch, choroid; RE, retinal epithelium; RCL, “rods and cones” layer; ONL, outer nuclear
layer; OPL, outer plexiform layer; INL, inner nuclear layer; IPL, inner plexiform
layer; GCL, ganglion cell layer; ILM, internal limiting layer. The adjectives used to
describe their relative positions, inner and outer, refer to the direction of the vitreous and
the choroid respectively. The outermost layer, adjacent to the choroid, is the nonneural
retinal epithelium.4 Moving towards the vitreous of the eye, the “rods and cones” layer
contains the photosensitive outer segments of the photoreceptor cells, the rods and cones.
The cell bodies of the photoreceptors lie in the outer nuclear layer (ONL) separated from
the rods-and-cones layer by the external limiting membrane, the outer boundary of the
neuroglial cells. The centrally (vitread) directed processes of the photoreceptors synapse
on the axons and dendrites of the bipolar and horizontal cells in the outer plexiform layer
4
In humans this layer is called the retinal pigment epithelium (Williams and Warwick, 1980) but it is
not pigmented in all vertebrates, so the more general term is preferred here.
Anthro-19.qxd 12/13/03 7:19 PM Page 481
Tarsier Fovea
481
plausible function, and functional “degeneracy,” all of which have been argued
to characterize the fovea of Tarsius. This chapter uses biomechanical data to
determine whether the tarsier fovea has a plausible function.
Another way to choose between these scenarios is to use comparative data
to determine whether foveae can arise in nocturnal lineages, or whether they
always arise in diurnal lineages, as has been assumed (Cartmill, 1980; Ross,
1996, 2001). This chapter presents comparative data on the distribution of
foveae among vertebrates and evaluates, using data on phylogenetic relatedness, whether foveae can arise in scotopically adapted lineages.
ANATOMY OF THE FOVEA
The vertebrate eye forms an image by refracting incoming light rays to bring
to a focus on the retina all rays from a common source (Figure 2). The light
rays reflected by the object of fixation are focused onto the retina at the visual
axis, and it is at this point on the retina where a fovea is found in many vertebrates. The fovea of Tarsius spectrum is illustrated in Figure 2 (Hendrickson,
personal communication). The fovea is usually characterized by a relatively
high density of photoreceptors and ganglion cells, a pit in the inner layers of
the retina, and the absence of intraretinal blood vessels.
The retina contains photoreceptor cells (“rods and cones”) that transduce
light energy into neural signals, first order neurons (bipolar cells), and
interneurons (horizontal and amacrine cells) that combine and process the signals from the photoreceptors, and second order neurons (ganglion cells) that
relay these signals to the brain. Compared with other parts of the retina, the
fovea exhibits increases in the density of photoreceptors and ganglion cells,
Figure 2. Continued
(OPL). The cell bodies for the bipolar cells and horizontal cells lie in the inner nuclear
layer (INL) along with the somata of the amacrine cells. The bipolar and amacrine cells
synapse on the ganglion cells in the inner plexiform layer (IPL), and the ganglion cell
somata lie in the ganglion cell layer (GCL). The axons of the ganglion cells form the next
layer, the ganglion cell axon or nerve fiber layer. The innermost layer, next to the vitreous,
is the internal limiting layer or membrane, composed of terminal processes of retinal glial
cells, a basement membrane, and collagen fibrils derived from the vitreous.
Anthro-19.qxd 12/13/03 7:19 PM Page 482
482
Callum F. Ross
Photoreceptors
Photoreceptors
Figure 3. Diagrams illustrating function of the fovea according to Walls (1937).
Top: Concentration of cells in the inner layers of the retina at the area centralis causes
a doming of the retina. This would result in a refraction of incoming light rays such as
to reduce image size. Bottom: The fovea reverses the direction of the light refraction,
enlarging the image.
and in the ratio of ganglion cells to photoreceptors. This is true whether the
foveal photoreceptors are predominantly cones, as in diurnal anthropoids such
as humans and macaques (Curcio et al., 1990; Packer et al., 1989; Perry and
Cowey, 1985; Wikler et al., 1990), or rods as in nocturnal Aotus (Ogden,
1974, 1975) or Tarsius (Hendrickson et al., 2000). In addition to being more
numerous, the photoreceptors have a smaller diameter and are elongated, as
has long been known for primates (Curcio et al., 1987; Hendrickson, 1992;
Hendrickson and Yuodelis, 1984; Polyak, 1941; Rohen and Castenholtz,
1967). Also, in tarsiers and many anthropoids there are fewer shortwavelength sensitive cones (S-cones) in the fovea than in the parafoveal retina
(Calkins, 2001; Hendrickson et al., 2000; Martin and Grünert, 1999).
Anthro-19.qxd 12/13/03 7:19 PM Page 483
Tarsier Fovea
483
Increases in photoreceptor density, ganglion cell density, and in the ganglion
cell:photoreceptor ratio at the fovea suggest that the fovea functions to
enhance visual acuity above and beyond that of the surrounding retina. This is
true whether the photoreceptors involved are cones or rods, although rods do
not afford as great an increase in acuity if they show high summation (i.e.,
many rods connected to few ganglion cells). The decrease in S-cones in the
foveae of some primates also increases acuity by ameliorating the chromatic
abberation that is particularly problematic at short wavelengths (Calkins,
2001) and in large eyes (Martin and Grünert, 1999).
The cells of the vertebrate retina are arranged in layers (Figure 2) (Rodieck,
1988), and the word fovea (Latin for “pit”) refers to the depression in the
retina caused by centrifugal displacement of the inner layers of the retina: The
nerve fiber, ganglion cell, inner plexiform, and inner nuclear layers. In Tarsius,
the inner nuclear layer is still present in the fovea, but is greatly thinned
(Figure 2). Walls (1937) observed that increase in cell numbers in the inner
retinal layers of the area centralis make the retina thickest there, doming the
retinal surface into the vitreous, an effect that has been documented in Lemur,
Indri, and Propithecus (Rohen and Castenholtz, 1967). Because the retina has
a higher index of refraction than the vitreous (Steenstrup and Munk, 1980;
Valentin, 1897; Walls, 1940), this doming of the retinal surface causes incoming light rays to be refracted (“bent”) and to converge onto a smaller area,
spreading the image over a smaller number of photoreceptors, decreasing
visual acuity (Figure 3). Walls (1937) argued that the fovea counters this
decrease in image size by spreading the image over a larger area (Figure 3).
Walls identified two shapes to the foveal pits of vertebrates: The concavesided (concaviclivate) form seen in humans and other primates, including
Tarsius, and the convex-sided (convexiclivate) form found in most birds and
lepidosaurs (Figure 4). When Walls calculated the increase in image size
associated with the two different fovea shapes he found that the amount of
refraction associated with the concaviclivate form was insufficient to produce
a significant increase in visual acuity, while the image produced by the convexiclivate form was significantly enlarged. Walls argued from this that the
concaviclivate fovea was actually a degenerate form of the convexiclivate fovea
and that all forms with concaviclivate foveae were descended from animals
with convexiclivate foveae (Walls, 1940, 1942). Pumphrey (1948) took issue
with Walls’ interpretation of the function of the convexiclivate fovea. He
demonstrated that the highly curved pit at the bottom of a convexiclivate fovea
Anthro-19.qxd 12/13/03 7:19 PM Page 484
484
Callum F. Ross
Concaviclivate
Photoreceptors
Convexiclivate
Photoreceptors
Figure 4. Diagrams of convexiclivate (convex sided) and concaviclivate (concave
sided) foveae.
severely distorts an image lying across it, arguing against Walls’ assumption that
the convexiclivate fovea is a superior version of the concaviclivate fovea.
Pumphrey argued that the convexiclivate fovea transforms a “radially symmetrical image into an asymmetrical one except when there is exact coincidence
between the axes of symmetry of the fovea and the object,” facilitating “maintenance of accurate fixation.” This effect results from the rapid changes in slope
moving away from the center of the convexiclivate fovea. Because the degree
of refraction of a light ray is in part a function of the slope of the interface
between the vitreous and retina, rapid changes in fovea slope result in rapid
changes in the “gain,” or degree of displacement of a light ray where it strikes
the rods and cones layer. Consequently, an image lying across a convexiclivate
fovea will have different parts displaced to different degrees, depending on the
slope of the fovea. Thus, only an image lying symmetrically across the fovea will
be evenly distorted. As soon as an image moves away from this symmetrical
position it will be distorted unevenly, thereby facilitating its fixation.
For these reasons, the convexiclivate fovea might also function as a movement
or focus detector. Rapid changes in gain as a light ray moves across the convexiclivate fovea amplify movement of the cone of rays coming from the lens. This
not only provides “sensitive appreciation of angular movements of a fixated
object” (Pumphrey, 1948; p. 307) but also increases the apparent movement of
the plane of focus, making the eye more sensitive to changes in focus. The convexiclivate fovea is an ideal focus detector because it only subtends a small part
of the visual field, so it can perform its function without distorting the image
over a broad extent of the visual field (Harkness and Bennet-Clark, 1978).
Anthro-19.qxd 12/13/03 7:19 PM Page 485
Tarsier Fovea
485
Notably, these mechanisms can work well in both scotopic (low light level)
and photopic (diurnal) conditions (Harkness and Bennet-Clark, 1978; Locket,
1992), providing plausible explanations for the presence of convexiclivate
foveae among animals living in dim light. Steenstrup and Munk (1980) have
also suggested that the convexiclivate foveae of the deep-sea teleost family
Notosudidae function to break camouflage of mesopelagic fishes using point
sources of light to hide against the background illumination.
The third attribute common to most foveae is the absence of intraretinal circulation. Most foveate vertebrates (lepidosaurs, avians, many euteleost fishes)
lack retinal circulation, their retinae being nourished from the choroid or
directly from the vitreous humor. However, some fishes have blood vessels
lying between the photoreceptor layer and the incoming light rays, either
within the retina or between the retina and the vitreous, and mammals with
thick retinae have two layers of blood vessels within the inner layers of the
retina (Chase, 1982). Anthropoids divert this retinal circulation away from the
fovea (Provis, 2001; Provis et al., 1998; Wolin and Massopust, 1970), tarsiers
have been variably argued to have avascular (Hendrickson et al., 2000) or vascular (Rohen, 1966; Wolin and Massopust, 1970) foveae. Erythrocytes are
thought to diffract light because of the iron attached to hemoglobin molecules, so it has been suggested that the lack of blood vessels in the fovea
improves visual acuity (Weale, 1966). In support of this argument, the retinal
blood vessels are not only deflected around the fovea in anthropoids, but also
around the area centralis of other (afoveate) mammals (Rohen, 1966; Walls,
1942; Weale, 1966; Wolin and Massopust, 1970).
FUNCTIONAL ANALYSIS
Plausible hypotheses for the functions of convexiclivate foveae have been proposed and tested computationally (e.g., Locket, 1992), but the function(s) of
concaviclivate foveae are less well understood. The anatomical bases of the
hypothesized functions for convexiclivate foveae are the steepness of the walls
of the fovea, the sharp reversal in the direction of curvature at the bottom
of the fovea, and the refractive index of the vitreal/retinal surface. The refractive index of the vitreal/retinal surface is fairly constant across vertebrates
(Steenstrup and Munk, 1980) but the slope of the foveal clivus and the shape
of the bottom of the fovea vary widely. As Pumphrey (1948) noted, more gently sloping (concaviclivate) foveae without a sharp reversal at the bottom cannot be assumed to function like convexiclivate foveae. The functions
Anthro-19.qxd 12/13/03 7:19 PM Page 486
486
Callum F. Ross
attributable to convexiclivate foveae may not be applicable to the foveae of
haplorhines and other vertebrates. What then is the function of the tarsier
fovea and of concaviclivate foveae in general? Specifically, does the fovea of
Tarsius refract the incoming light rays enough for it to function as an image
enlarger, focus indicator, or movement detector?
Functional Analysis Methods
If the tarsier fovea functions either to increase visual acuity by increasing the
size of the retinal image, or enhances movement by amplifying it, then the
retina must refract incoming light rays enough to produce gain equal to or
greater than the highest spatial frequency unambiguously available to the
visual system; that is, the Nyquist limit. The Nyquist limit () can be related
to the spacing between ganglion or photoreceptor cells by the formula
⫽ 兹3a
where a ⫽ intercell spacing in a hexagonal array. The Nyquist limit calculated
from the receptive field center diameter of a ganglion cell in cats and macaques
closely matches the limits of spatial resolution measured behaviorally (Wässle
and Boycott, 1991), so it is reasonable to estimate the smallest spatial frequency discernable by the tarsier fovea using the formula for .
The intercell spacing of tarsier ganglion cells is unknown, but Tetreault et al.
(this volume) report a peak ganglion cell density for Tarsius syrichta of
13,300/mm2 near the fovea, suggesting a ganglion cell spacing of 0.75 m.
Corrections of foveal ganglion cell densities due to ganglion cell displacement
in humans reveal higher effective ganglion cell densities in the foveola than on
the foveal rim (Sjörstrand et al., 1999), suggesting that there might be even
higher effective ganglion cell densities in the center of the tarsier fovea. Given
that cone photoreceptors are separated by 2–3 m, a ganglion cell spacing of
0.75 m and higher in Tarsius suggests that more than one ganglion cell is connected to each foveal photoreceptor. If this is the case, then Tarsius resembles
anthropoids in having cones in its fovea connected to more than one ganglion
cell. Hendrickson et al. (2000, personal communication) report cone densities
as high as 50–85,000 mm2 at 50–100 m from the fovea, suggesting that the
“tarsier may have a steep cone gradient centered on the fovea, similar to diurnal
primates” (Hendrickson et al., 2000, p. 727). At present, therefore, the most
conservative estimate of the Nyquist limit of Tarsius is derived using the
Anthro-19.qxd 12/13/03 7:19 PM Page 487
Tarsier Fovea
487
photoreceptor spacing of primate cone cells. Curcio et al. (1990, p. 519) report
a range of values for cone center–center spacing in humans ranging from
1.9–4.8 m. These values yield Nyquist limits of 3.3–8.3 m.
To calculate the increase in image size afforded by the fovea of Tarsius, the
following parameters are required: (a) The degree to which incoming light
rays are refracted as they pass over the retinal surface; (b) the distance from the
vitreal retinal surface to the outer segments of the photoreceptors; and (c) the
angle of incidence of the incoming light rays. Given these three parameters, we
can calculate a parameter, called here gain, defined as the distance on the photoreceptor outer segments (at the inner limiting membrane) that a single light
ray is deflected by the inner retinal surface.
(a) Refraction. Light is refracted as it passes across the interface between
media of different optical densities. Passing from a less to a more dense
medium, as in the case of the vitreous–retina interface, it is bent towards the
perpendicular to the interface. Snell’s law describes the relationship between
the angle of incidence (⍜1), the angle of refraction (⍜2), and the indices of
refraction of the two media (n1 and n2) as follows:
sin ⍜2 ⫽ sin⍜1 * n1/n2
(1)
The refractive index of the retina has not been measured in Tarsius, but
Valentin (1879, in Walls, 1940, pp. 831–832) reported a range of values for
mammals from 1.3407 in the green monkey, to 1.3460 in the dog, with a
mammalian average of 1.34385 and the average of the green monkey and a
baboon of 1.34265. The index of refraction of the vitreous in tarsiers is also
unknown, but in humans it is around 1.336 (Duke-Elder, 1970). Assuming
that the retina of Tarsius has an index of refraction of 1.343 and the vitreous
an index of 1.336, the ratio of the refractive indices of vitreous : retina is 0.995
(Pumphrey, 1948; Valentin, 1879). Therefore ⍜2 can be estimated as
⍜2 ⫽ arcsin(.995 sin ⍜1)
(2)
(b) Retinal depth. The contours of the vitreal surface of the retinal fovea
were digitized from one published photograph of the fovea of Tarsius bancanus (Rohen and Castenholtz, 1967, Figure 8b) and from an image of the
fovea of Tarsius spectrum kindly supplied by Dr Anita Hendrickson, University
of Washington, Seattle, WA (Figure 2B). The digitized data consisted of
“x values” and “y values” for points on the vitread surface of the fovea. In Tarsius
bancanus the distance from the retinal surface to the “rods and cones” layer at
Anthro-19.qxd 12/13/03 7:19 PM Page 488
488
Callum F. Ross
the bottom of the fovea (100 m) was calculated by subtracting the length of
the photoreceptor outer segments (48 m) from the thickness of the retina
at that point (148 m) (Rohen and Castenholtz, 1967). This “y value” was
then applied to the lowest point in the digitized image and the remainder of
the “y values” scaled accordingly. These values represent Y in Figure 5. For the
other foveae included in this study, the scales of the images were calculated
directly from the scale bars in the published figures.
(c) Angle of incidence. An incident light ray that is perpendicular to the
photoreceptor layer makes an angle of incidence (⍜1) (relative to the perpendicular to the vitreal surface) of
⍜1 ⫽ arctan f(x)
(3)
where f(x) ⫽ the slope of the line connecting the digitized points (Figure 5).
However, light arrives at the fovea in a cone. The maximum angles of the
edges of a cone of light arriving at the retina can be estimated from measures
of lens diameter and the distance of the center of the lens from the retina.
Using data in Castenholtz (1984), ␣ ⫽ 24.5⬚ is the highest angle at which rays
Figure 5. Diagram illustrating method for calculating gain. See text for details.
Anthro-19.qxd 12/13/03 7:19 PM Page 489
Tarsier Fovea
489
of light can strike the fovea. If the gain for the perpendicular light rays is
gain ⫽ tan(⍜1 ⫺ ⍜2) f(x)
(4)
or, substituting Eq. (2) and (3) into Eq. (4), as
gain ⫽ f(x) tan{[arctan f(x)] ⫺ [arcsin (.995 sin(arctan f(x)]}
(5)
(See Figure 5), then the gain for the light rays at the periphery of the light
cone can be calculated using the same formula, but adding 25⬚ to the angle of
incidence (⍜1).
These calculations were performed on the digitized images of the foveae of
Tarsius bancanus (Rohen and Castenholtz, 1967, Figure 8b) and Tarsius
spectrum (Hendrickson, personal communication). For comparison, calculations
were also performed for the foveae of some other vertebrates, including some
primates with concaviclivate foveae and some birds and fish with convexiclivate
foveae (Table 1). For vertebrates other than Tarsius, these calculations were
made for light rays perpendicular to the retina only, for lack of precise estimates
of the angle of incidence of rays coming through the periphery of the lens.
Results of Functional Analysis
The gains calculated for the fovea of Tarsius spectrum are plotted in Figure 6
and maximum values of gain for all foveae are given in Table 1. The maximum
Table 1.
Gain produced by fovea according to equation 5
Taxon
Tarsius spectrum
Tarsius bancanus
Cercopithecus nictitans
Homo sapiens
Strix aluco
Asio otus
Asio flammeus
Tyto alba
Alepocephalus bairdi
Falco berigora (nasal fovea)
Falco berigora (temporal fovea)
Halcyon sancta (nasal fovea)
Halcyon sancta (temporal fovea)
a
Gain (m)a
0.6
1.52
0.53
0.30
1.12
0.64
1.35
0.22
3.99
2.63
1.55
12.9
3.12
(1.33 at edge of light cone)b
(2.32 at edge of light cone)b
Highest absolute value of gain (rightward or leftward deflection of a light ray) for a
ray of light perpendicular to the photoreceptor layer.
b
Highest absolute value of gain (rightward or leftward deflection of a light ray) given
by a ray of light arriving at the retinal surface at an angle of 25⬚ to the perpendicular.
Anthro-19.qxd 12/13/03 7:19 PM Page 490
490
Callum F. Ross
Figure 6. Results of gain calculations for Tarsius spectrum. Top: Image of fovea
supplied by Dr A. Hendrickson. Middle: digitized outline of fovea. Light arrives at the
retina in a cone with a maximum included angle of 50⬚ (see text). Bottom: Plot of gain
calculations for the three light rays defining the center and edges of the light cone. The
solid line represents gain for incoming light rays at the center of the cone, the dotted
and dashed lines the gains for the right and left edges of the cone. Gain is plotted at
all positions along the fovea. Note that maximum gain is 1.33 m.
gain produced in the tarsier fovea is 1.5 m for the perpendicular rays, roughly
1.5 times the width of one photoreceptor. The maximum values obtained for
the edges of the cone of light rays are 2.3 m in Tarsius bancanus when the
edge of the cone passes across the sloping side of the fovea. If the minimum
detectable gain required for a detectable optical effect is 3.3–8.3 m, then
Anthro-19.qxd 12/13/03 7:19 PM Page 491
Tarsier Fovea
491
tarsiers are not sensitive to the gain derived from foveal refraction. Notably, nor
does this number exceed the limits of acuity demonstrated for diurnal
macaques, suggesting that the gain produced by the tarsier fovea would be
insignificant to tarsiers, even if they had the visual acuity of diurnal anthropoids.
Low gains are also seen in Cercopithecus, Homo, and the owls Tyto, Asio, and
Strix, forms with shallow foveae with gently sloping walls. Without knowing
maximal pupillary diameter and the distance from the lens center to the retina,
precise values for maximal gain cannot be calculated. However, note that much
higher gains are seen in the convexiclivate foveae of the kingfisher Halcyon
sancta, the falcon, Falco berigora, and the deep-sea teleost, Alepocephalus bairdi.
COMPARATIVE ANALYSIS
Comparative Methods
The evolutionary hypothesis being tested is that foveae are most likely to arise
in diurnal lineages and are retained in nocturnal lineages either by stabilizing
selection or by some kind of unspecified constraint (Cartmill, 1980; Ross,
1996, 2000; Walls, 1942). To evaluate this hypothesis, data on the occurrence
of foveae in vertebrates were gathered from the literature and are summarized
in Tables 2–4. The activity patterns and/or habitats of foveate vertebrates and
their close relatives are also supplied. Hypotheses about the sequence of evolutionary events can only be tested by mapping the distributions of character
states onto phylogenies. The most recent phylogenies for the relevant groups
were obtained from the literature. In one case (lepidosaurs) it was possible to
apply Maddison’s (1990) concentrated changes test to determine whether
foveae are more likely to evolve in those parts of the trees that represent diurnal lineages. Because the number of changes involved in each tree is relatively
small, Schluter et al.’s (1997) maximum likelihood approach was not required.
The concentrated changes test requires full resolution of tree branching patterns, so polytomies were arbitrarily resolved.
Results of Comparative Analysis
Fishes: Eleven genera of bathypelagic fish (fish living below a depth of
1000 m) exhibit a fovea in an all-rod retina (Table 2). Numerous fishes living
near the surface in photopic environments also exhibit foveae. Fishes therefore
provide an excellent group in which to determine the probability of a fovea
evolving in a “nocturnal” or scotopic environment.
Occurrences of foveae and tapeta in Teleostei
Species
Retinal specializations
Recorded depth and diet
References
Elopocephala
Elopomorpha
Anguilliformes
Serrivomeridae
Serrivomer beani
No fovea or area.
Synaphobranchus
kaupi
No fovea or area.
Epipelagic/
abyssopelagic (150–3000 m)
Benthypelagic (400–4800 m);
mobile
piscivore
Collin and Partridge, 1996;
Nicol, 1989
Collin and Partridge, 1996;
Whitehead et al., 1986
Nansenia
groenlandica
Bathylagus benedicti,
B. pacificus
Temporal area, multiple
banks of rods
Concaviclivate temporal
fovea, 6 banks of receptors
in fovea, 3 elsewhere; pure
rod retina
No fovea, pure rod retina
No fovea, pure rod retina
Mesopelagic
Munk, 1966b
Epipelagic and
mesopelagic; gelatinous
predators
(zooplanktonivores)
Epipelagic
Epipelagic
Vilter, 1954a, 1954c and
Munk, 1966b, in
Locket, 1977; Nicol,
1989; Gartner et al., 1997
Munk, 1966b
Munk, 1966b;
Weitzman, 1997
Munk, 1966b;
Weitzman, 1997
Munk, 1966b;
Weitzman, 1997
Munk, 1966b;
Weitzman, 1997
Synaphobranchidae
Clupeocephala
Euteleostei
Protacanthopterygii
Argentiniformes
Argentinoidei
Microstomatidae
Microstomatinae
Bathylagini
Opisthoproctidae
B. stilbius
Bathylychnops exilis
Rhynchohyalus
natalensis
Opisthoproctus
grimaldii
Winteria telescopa
No fovea, pure rod retina
No fovea, pure rod retina
No fovea, pure rod retina
Epipelagic and
mesopelagic (down to 550 m)
Mesopelagic (200–600 m)
Mesopelagic and
Bathypelagic (500–1250 m)
Callum F. Ross
Classification
Anthro-19.qxd 12/13/03 7:19 PM Page 492
492
Table 2.
Convexiclivate temporal
fovea; pure rod retina
Mesopelagic (450–1500 m)
Platytroctes apus
Convexiclivate temporal
fovea; pure rod retina
Deep convexiclivate fovea;
pure rod retina
Deep convexiclivate fovea;
pure rod retina.
Convexiclivate temporal
fovea; pure rod retina
Convexiclivate temporal
fovea; pure rod retina
No fovea; pure rod retina
Bathypelagic (seldom ⬍900 m)
Platytroctegen mirus
Bathylaconidae
Bathylaco nigricans
Alepocephalidae
Bathytroctes
microlepis
Bajacalifornia drakei
Rouleina attrita
Leptoderma macrops
Alepocephalus
rostratus
Alepocephalus bairdi
Conocara murrayi
Xenodermichthys
copei
Convexiclivate fovea; pure
rod retina
Convexiclivate temporal
fovea; pure rod retina
Convexiclivate fovea; pure
rod retina
Convexiclivate temporal
fovea; pure rod retina
Convexiclivate temporal
fovea, pure rod retina
Bathypelagic
Meso- and bathypelagic
Engybenthic/
bathypelagic (1100–2700 m)
—
Engybenthic (1400–2100 m);
mucous
predator
—
Engybenthic (300–3600 m);
Macroplanktonivore,
hover-and-wait
Macroplanktonivore,
hover-and-wait
Engybenthic (1200–2600 m),
Diet unknown.
Mesopelagic (100–1000 m)
Marshall, 1966, in
Locket, 1977, p. 173; Collin
and Partridge, 1996
Collin and Partridge, 1996
Munk, 1966b, in
Locket, 1977: 173
Munk, 1968; Collin and
Partridge, 1996
Collin and Partridge, 1996
Locket, 1985
Collin and Partridge, 1996;
Gartner et al., 1997
Locket, 1977, p. 173,
Figure 61
Collin and Partridge, 1996;
Gartner et al., 1997
Tarsier Fovea
Searsia koefoedi
Anthro-19.qxd 12/13/03 7:19 PM Page 493
Alepocephaloidei
Platytroctidae
Locket, 1992; Gartner
et al., 1997
Collin and Partridge, 1996;
Gartner et al., 1997
Marshall, 1966, in Locket,
1977, p. 173; Collin
and Partridge, 1996
493
Classification
Neoteleostei
Stenopterygii
Stomiiformes
Sternoptychidae
(hatchetfishes)
Stomiidae
Ipnopidae
Notosudidae
Retinal specializations
Recorded depth and diet
References
Argyropelecus aculeatus,
A. sladeni,
A. olfersi
Sternoptyx
No fovea
Mesopelagic
zooplanktonivores
No fovea
Malacosteus niger
(incl. indicus?)
No fovea, tapetum
Mesopelagic;
zooplanktonivores
Mesopelagic;
zooplanktonivore
Munk, 1966b; Collin
and Partridge, 1996;
Gartner et al., 1997
Gartner et al., 1997
Chlorophthalmus
albatrossis
No fovea. Guanine
tapetum
Down to 1440 m
Bathypterois
dubius
(tripodfishes)
No fovea. Yellow eyeshine
Benthic (750–950 m); sit and-wait ambush
micronekton feeder
(copepods)
Ipnops murrayi
Ahliesaurus berryi,
Scopelosaurus hoedti
No fovea
All(?) cone retina;
temporal convexiclivate
fovea in cone area
Scopelosaurus
lepidus
temporal convexiclivate
fovea in cone area; rest of
retina rods; retinal
tapetum around groups
of rods
Meso- and bathypelagic,
hovering and darting
predation
Meso- or benthopelagic
(745–650 m);
zooplanktonivore
Munk, 1977; Gartner
et al., 1997
Tamura, 1957 and
Somiya and Tamura 1971
in Munk, 1977
Collin and Partridge,
1996
Munk, 1959
Nicol, 1989;
Marshall, 1966, in
Locket, 1977, p. 173;
Munk, 1975
Munk, 1977
Callum F. Ross
Aulopiformes
Chlorophthalmoidei
Chlorophthalmidae
Species
Anthro-19.qxd 12/13/03 7:19 PM Page 494
Continued
494
Table 2.
Paralepididae
Notolepis rissoi
Evermannellidae
Evermanella atrata
No fovea, Tapetum
lucidum
Evermanella indica
No fovea
Scopelarchus
güntheri, S. sagax,
Benthalbella infans
No fovea, S. güntheri
tapetum lucidum
Scopelarchus
michaelsarsi
No fovea
Mesopelagic (256–500 m);
active piscivorous
micronectivores
Myctophum
punctatum,
Lampanyctus
macdonaldi
No fovea
Mesopelagic, migratory?;
zooplanktonivores
Lepadogaster candollei
Fovea
Diretmus argenteus
No fovea. Retinal
tapetum
Scopelarchidae
Ctenosquamata
Myctophiformes
Myctophidae
Acanthomorpha
Gobiesociformes
Gobiesocidae
Beryciformes
Diretmidae
Bathypelagic (950–2500 m);
active piscivorous
micronectivores
Mesopelagic; active
piscivorous
micronectivores
?Mesopelagic; active
piscivorous
micronectivores
Mesopelagic; active
piscivorous
micronectivores
Mesopelagic (190–1855 m);
active piscivorous
micronectivores
Munk, 1965;
Frederiksen, 1976 in
Munk, 1977, p. 22;
Gartner et al., 1997
Locket, 1977, p. 167;
Gartner et al., 1997
Brauer, 1908 in Munk,
1977; Gartner et al.,
1997
Munk, 1966b; Gartner
et al., 1997
Locket, 1971; Locket,
1977, p. 158; Munk,
1977; Gartner et al.,
1997
Collin and Partridge,
1996; Collin et al., 1998;
Gartner et al., 1997
Collin and Partridge,
1996; Gartner et al.,
1997
Vrabec, 1969 in Munk, 1975
Mesopelagic (⬍1000 m)
Munk, 1966a
495
Gasterosteiformes
Syngnathoidei
Cone dominated retina. No
fovea, Tapetum lucidum
around each cone; no
reflectors with rods
No fovea, tapetal cell
sheaths
Anthro-19.qxd 12/13/03 7:19 PM Page 495
Omosudis lowei
Tarsier Fovea
Alepisauroidei
Alepisauridae
Continued
Classification
Syngnathidae
Perciformes
Percoidei
Serranidae
Retinal specializations
Recorded depth and diet
References
Entelurus aequoreus,
Nerophis ophidion,
Siphostoma fuscum,
S. acus, S.
tenuiorostris
Fovea
Convexiclivate central
fovea
Fovea
Shallow water
Syngnathus typhle
Temporal, asymmetrical,
slit-like, convexiclivate
fovea, (Two foveae
according to Rauther, 1925)
Shallow water
Munk, 1975; Slonaker,
1897 in Munk, 1975;
Krause, 1889;
Verrier, 1928;
Kahmann, 1936 in
Munk, 1975
Krause, 1886a in
Locket, 1977, p. 173
Kahman, 1934, 1936;
Kolmer, 1936;
Rochon-Duvigneaud,
1943
Locket, 1977, p. 173
Hippocampus
hippocampus,
H. ramulosus
Agonus cataphractus
Fovea
Serranus cabrilla,
S. hepatus,
S. scriba,
Paralabrax
clathratus, P.
maculofasciatus,
P. nebulifer
Fovea
Kahmann, 1936 in
Munk, 1975
Predators of mobile prey
Verrier, 1928,
Kahmann, 1936
Schwassmann, 1968 in
Munk, 1975
Callum F. Ross
Scorpaeniformes
Cottoidei
Agonidae
Species
Anthro-19.qxd 12/13/03 7:19 PM Page 496
496
Table 2.
Scaridae
Trachindoei
Trachinidae
Clinidae
Chaenopsidae
Pholidodidae
Fovea
Verrier, 1933;
Kahmann, , 1934, 1936
in Munk, 1975
Fovea
Ali et al., 1973, in
Munk, 1975
Trachinus draco, T.
vipera
—
Blennius basiliscus,
B. gattorugine, B.
ocellaris, B. pavo, B.
sanguinolentus, B.
tentacularis
Dialammus fuscus,
Mnierpes
macrocephalus
Fovea
Kahmann, 1934, 1936 in
Munk, 1975
—
Rochon-Duvigneaud,
1943
Clinus dorsalis,
Malacoctenus hubbsi,
Paraclinus sini
Acanthemblemaria
crockery, Chaenopsis
alepidota
Pholis gunnellus
—
Fovea
Near shore, benthic
Benthonic, sit-and-wait
ambush predator of
invertebrates, small fish
Fovea
Amphibious fish
specialized for amphibious
vision
Fovea
Munk, 1969; Graham
and Rosenblatt, 1970;
Graham, 1970, 1971 in
Munk, 1975
Munk, 1971
Fovea
Munk, 1971
Fovea
Verrier, 1933;
Kahmann, 1934, 1936 in
Munk, 1975
Tarsier Fovea
Blennioidei
Blenniidae
Coris julis, Julis
lunaris,
Pseudocheilinus
hexataenia,
Thalassoma pavo
Cryptotomus roseus
Anthro-19.qxd 12/13/03 7:19 PM Page 497
Labroidei
Labridae
497
Continued
Classification
Ostraciontidae
Tetraodontoidei
Tetraodontidae
Retinal specializations
Balistapus aculeatus,
Balistes carolinensis
Ostracion cornutus,
O. cubicus
Fovea
Tetrodon fluviatilis
Fovea
Fovea
Recorded depth and diet
References
Kahmann, 1934, 1936 in
Munk, 1975
Kahmann, 1936 in
Munk, 1975
Kahmann, 1934, 1936 in
Munk, 1975
Notes: Definitions of habitat and dietary guilds primarily from Gartner et al. (1997). Information on Blennius from Munk (1971).
Benthic, fishes in physical contact with the bottom and not very mobile; Demersal, fishes spending most of time near bottom (⬍5 m) and moving actively over
it; Benthopelagic, pelagic forms that spend only part of life cycle near the bottom; Pelagic, midwater fishes; Epipelagic, species residing above 200 m;
Mesopelagic, species residing 200 m to 1000 m; Bathypelagic, species residing primarily below 1000 m.
Callum F. Ross
Tetraodontiformes
Balistoidei
Balistidae
Species
Anthro-19.qxd 12/13/03 7:19 PM Page 498
498
Table 2.
Anthro-19.qxd 12/13/03 7:19 PM Page 499
Tarsier Fovea
499
The environments and diets of foveate fishes and their relatives are given in
Table 2. Water attenuates light transmission in various ways and to various
degrees depending on the optical properties of the water. However, under the
best conditions the scotopic threshold lies at a depth of about 1000 m, below
which down-welling light is not visible (Nicol, 1989), and in most circumstances, the mesopelagic environment may be considered scotopic as well. The
phylogenetic relationships of fishes used here derive from Johnson and
Patterson (1993, 1996; Baldwin and Johnson, 1996) and Stiassny et al. (1996).
Protacanthopterygii. Foveae are not reported in elasmobranchs (sharks,
skates, and rays), non-tetrapod sarcopterygians, “chondrosteans”
(Polypteriformes and Acipenseriformes), Ginglymodi (Lepisosteus),
Haleocomorphi (Amia), or in basal teleosts Osteoglossomorpha,
Elopomorpha, and Clupeomorpha. Among euteleosts, foveae first appear
among the Protacanthopterygii, a monophyletic sister taxon to all other
euteleosts (i.e., Eurypterygii, Esociformes, and other stomiiforms) (Johnson
and Patterson, 1996). Protacanthopterygii consist of three groups, the
Salmoniformes, Argentinoidei, and Alepocephaloidei, with the latter two
united in the order Argentiniformes (Figure 7).
The phylogenetic relationships of basal euteleosts are presented in Figure 7
along with data on the presence/absence of foveae and the depths at which
the fish are found (see also Table 2). Visual system data are available for three
taxa that were not included in the most recent phylogenetic analysis (Johnson
and Patterson, 1996). Conocara and Xenodermichthys are here placed at the
base of the Alepocephalidae and Platytroctegen is placed in with the other
platytroctids, Platytroctes and Searsia, in an unresolved trichotomy.
Among basal euteleosts, foveae appear to have evolved twice among argetiniforms: Once among alepocephaloids and once among argentinoids. All alepocephaloids that have been examined possess an all-rod retina and all possess
a temporally positioned convexiclivate fovea except Rouleina, which exhibits a
temporal area with a relatively high concentration of ganglion cells (Collins
et al., 1996). Regardless of the phylogenetic positions of Xenodermichthys,
Conocara, and Platytroctegen (and the retinal anatomy of the six unsampled
genera) a temporally positioned convexiclivate fovea is a synapomorphy of the
Alepocephaloidei. All but one genus of Alepocephaloidei is found in either the
mesopelagic or the bathypelagic zones. Three genera are both bathy- and
mesopelagic (Searsia, Bathylaco, Leptochilichthys), one genus (Bajacalifornia)
ranges from the epipelagic through bathypelagic zones, and Xenodermichthys
Anthro-19.qxd 12/13/03 7:19 PM Page 500
Alepocephalus
Bathyprion
Leptochilichthys
Leptoderma
Photosylus
Rouleina
Rinoctes
Al
Talismania
Narcetes
Bathytroctes
Conocara
Xenodermichthys
F
Bajacalifornia
ge
nt
in
id
ae
M
Bathylagus stilbius
ic
ro
st
Nansenia
om
at
Bathylychnops exilis
id
ae
Winteria
O
pi
Opisthoproctus
st
ho
pr
Rhynchohyalus
oc
tid
Platytroctes
ae
P
Searsia
la
ty
tro
Platytroctegen
ct
id
ae
Bathylaco
Bathylagus pacificus
Argentina
Glossanodon
Salmoniformes
Esociformes
Other Stomiiformes
Eurypterygii
Ostariophysi
Clupeomorpha
Elopomorpha
Ar
ep
oc
ep
ha
lid
ae
Callum F. Ross
500
A
F
Argentinoidei
Binary environment
unordered
Alepocephaloidei
Meso or
Bathypelagic
Epipelagic
F
Polymorphic
Equivocal
A
Fovea
= Afoveate
F
= Foveate
Argentiniformes
Protacanthopterygii
Euteleostei
A
Figure 7. Phylogeny of euteleost fishes with environment mapped on the branches
and most parsimonious hypotheses for gains and losses of foveae indicated with boxes.
Note that foveae evolve twice more among Eurypterygii. See Table 4 for data.
is only recovered from between 100–1000 m. Thus, the stem lineages
of Alepocephaloidei and Alepocephalidae were in possession of a convexiclivate fovea and either mesopelagic (200–1000 m depth) or bathypelagic
(⬎1000 m).
Within the Argentinoidei, two species of Bathylagus, B. pacificus, and
B. benedicti (B. euryops?) have been confirmed as possessing a temporally positioned concaviclivate fovea in an all-rod retina (Locket, 1977; Munk, 1966b;
Vilter, 1954a,b). The retinae of the other seven bathylagine genera are unstudied. The only microstomatine microstomatid genus sampled, Nansenia groenlandica, is afoveate (Munk, 1966b). Five out of a total of six opisthoproctid
genera have been sampled (Opsithoproctus, Dolichopteryx, Rhynchohyalus,
Winteria, Bathylychnops), and they all lack foveae (Ali and Hanyu, 1963; Collin
et al., 1997; Munk, 1966b). Among Argentinidae, Argentina silus lacks a
fovea (Ali and Hanyu, 1963), and Glossanodon is unsampled. Thus, among the
Anthro-19.qxd 12/13/03 7:19 PM Page 501
Tarsier Fovea
501
Argentinoidei, a fovea has only been confirmed in the bathylagines.
Argentinoids tend to reside slightly higher in the water column than alepocephaloids, being predominantly epipelagic and/or mesopelagic, although
Bathylagus and opisthoproctids are sometimes recovered below 1000 m
(Weitzman, 1997).
Thus, it appears that the fovea of the Alepocephaloidei evolved at least in a
mesopelagic environment, and possibly a bathypelagic one. The bathylagine
fovea evolved in either epipelagic, mesopelagic, or bathypelagic conditions.
Bathylagus is sometimes recovered from bathypelagic depths. Where known,
the argentiniforms are macroplanktonivores or zooplanktonivores (Table 2,
data from Gartner et al., 1997), with many of their prey being bioluminescent
(Herring, 1987). Alepocephalus feeds on gelatinous prey and incidental pelagic
prey using a hover-and-wait strategy. The alepocephalids (slickheads) also
exhibit marked expansion of their optic tectum (Wagner, 2001), suggesting
that they may be classified as visual predators. The single exception is
Rouleina, which may obtain organic detritus and tiny organisms from the
water column by suspending mucous from the ventral and dorsal midlines,
then ingesting the mucous with captured food particles (Gartner et al., 1997).
It is notable that Rouleina is the only alepocephaloid genus that is not a visual
predator and the only one lacking a retinal fovea.
The sister taxon of Argentiniformes, Salmoniformes, lives in a photopic
environment but lacks any evidence of a fovea, although only 2 genera and
9 species have been sampled (Ali, 1959; Ahlbert, 1976; Beaudet et al., 1997;
Schmitt and Kunz, 1989).
Euryptergians. Foveae appear again within the basal eurypterygians, the
Aulopiformes. The optic specializations of 10 genera from the two largest families of Aulopiformes, the Chlorophthalmoidei and Alepisauroidei, have been
studied, as have four genera of the closely related mesopelagic order,
Stomiiformes. Among these taxa, only two genera, Ahliesaurus and
Scopelosaurus, from the chlorophthalmoid family Notosudidae, have been
shown to possess foveae. The third genus in this family has not yet been studied. The Chlorophthalmoidei consist of predominantly meso-bathypelagic
predators with significant numbers of cones in the retina. The notosudids possess a temporally positioned convexiclivate fovea characterized by paired cones,
in comparison with the grouped photoreceptors seen elsewhere in the retina.
Foveae have not been reported among other Aulopiformes, or among
Myctophiformes, suggesting that the foveae of acanthomorphs represent at
Anthro-19.qxd 12/13/03 7:19 PM Page 502
502
Callum F. Ross
least a fourth case of independent evolution. However, only one myctophiform
has been sampled and more data on the retina and the visual system in general
are sorely needed for basal Neoteleostei (Wulliman, 1997).
Acanthomorpha. Among acanthomorph fishes, foveae have been documented in numerous taxa from various habitats, summarized in Table 2, but
an exhaustive survey of retinal anatomy is lacking. For example, although one
species of the Gobiesociformes has been reported to have a fovea (Table 2), it
is not clear whether other Paracanthopterygii have been examined. No foveae
have been reported for Beryciformes, but foveae are widespread among percomorph fishes. These foveae vary in shape, although some are documented
to be convexiclivate, even among epipelagic fishes. At present it is impossible
to determine whether a fovea is primitive for Acanthomorpha or not. Notably,
nocturnal predatory acanthomorphs, such as Apogonidae (cardinal fishes) and
Haemulidae (grunts) are not among those documented as possesing foveae,
however, nor is it clear how well they have been sampled.
Summary of Fish Data. A fovea has evolved at least four times among
fishes: A convexiclivate fovea evolved in the mesopelagic or bathypelagic stem
lineage of Alepocephaloidei; a concaviclivate fovea was acquired in the
mesopelagic or epipelagic stem lineage of Bathylagus; a convexiclivate fovea
was gained once in the cone dominated retina of the epi-, meso-, and bathypelagic notosudid Chlorophthalmoidei; and foveae of either shape appear in
several lineages of the Acanthomorpha. The convexiclivate foveae of alepocephaloids almost certainly evolved and were maintained in a lineage of scotopic visual predators. The concaviclivate foveae of bathylagine argentinioids
may have evolved in a scotopic lineage, as all bathylagines have eyes specialized
for scotopic environments, or in the epipelagic and photopic environment
characteristic of bathylagines.
Lepidosauria: The distribution of foveae (Peterson, 1992), wellcorroborated phylogenies (Estes and Pregill, 1988) and ecological studies
(Vitt and Pianka, 1994) are available for lepidosaurs (Figure 8, Table 3).
Lepidosaurs provide an interesting parallel with haplorhine primates. The distribution of activity period among extant squamates suggests that the squamate clade (lizards and snakes) was primitively diurnal, although their sister
taxon is the relict nocturnal predator, Sphenodon (Figure 8). Isolated on tiny
islands off the coast of New Zealand, Sphenodon is the only surviving remnant
of a broader distribution of sphenodontians in the Mesozoic.
Anthro-19.qxd 12/13/03 7:19 PM Page 503
Table 3.
Occurrences of foveae and tapeta in Lepidosauria
Species
Retinal specializations
Habitat
References
Rhynchocephalia
Squamata
Iguania
Iguanidae
Iguanines
Sphenodon punctatus
Concaviclivate fovea
Nocturnal, ambush
Walls, 1942; Vilter, 1951
Iguana iguana
Fovea
Diurnal
Dipsosaurus dorsalis
Fovea
Diurnal, herbivorous
Phrynosoma cornutum,
Central convexiclivate
fovea
Diurnal
Phrynosoma orbiculaire
Fovea
Strictly diurnal
Anoloids
Anolis spp.
Temporal and nasal foveae,
nasal more convexiclivate
Diurnal, sit-and-wait
predators
Basliliscines
Basiliscus plumifrons
Fovea
Diurnal
Meneghini and Hamasaki,
1967, in Peterson, 1992
Peterson, 1981, in
Peterson, 1992
Slonaker, 1897, in Munk,
1970; Detwiler and Laurens
in Peterson, 1992
Rochon-Duvigneaud,
1943
Various authors,
Peterson, 1992; Fite and
Lister, 1981; Makaretz
and Levine, 1980
Kahmann, 1923 in
Rochon-Duvigneaud,
1943
Chamaeleo
Agama tournevillii
Central convexiclivate
fovea
Convexiclivate fovea
Amphibolurus barbatus
Fovea
Physignathus leusieur
Fovea
Diurnal, sit-and-wait
predators
Diurnal, sit-and-wait
predators
Diurnal, sit-and-wait
predators
Diurnal, sit-and-wait
predators
Uromastyx acanthinurus
Horizontal band-shaped
area with fovea
Sceloporines
Acrodonta
Chamaeleonidae
Agamidae
Cajal, 1893, Johnson,
1927, in Peterson, 1992
Verrier, 1933
O’Day, 1939 in Peterson,
1992
Kahmann, 1923 in
Rochon-Duvigneaud,
1943
Kahmann, 1936 in Munk,
1970.
503
Diurnal
Tarsier Fovea
Classification
Continued
Classification
Retinal specializations
Habitat
References
Scleroglossa
Pygopodidae
Delma, Lialis, Aprasia
No fovea
Gekkonidae
Gekkonines
—
Phelsuma spp.
—
Fovea
Nocturnal, some
diurnal activity
—
Diurnal
Aristelliger cochranae
Fovea
Other gekkonines
No fovea
Sphaerodactylus
Fovea
Nocturnal/
crepuscular
Nocturnal/
crepuscular
Diurnal
Underwood, 1970; Walls,
1942
—
Tansley, 1961, 1964 in
Peterson, 1992; Pough
et al., 1998
Underwood, 1970
Gonatodes
Fovea
Diurnal
Ahaetulla (previously
Dryophis)
Dryophiops
Temporal concaviclivate
fovea
Temporal concaviclivate
fovea
Temporal concaviclivate
fovea
Temporal
Diurnal, arboreal
predator
Diurnal, arboreal
predator
Diurnal, arboreal
predator
Diurnal, arboreal
predator
Sphaerodactylines
Serpentes
Colubridae
Thelotornis
Oxybelis
Underwood, 1970
Underwood, 1970, in
Peterson, 1992; Pough
et al., 1998
Underwood, 1970, in
Peterson, 1992; Pough
et al., 1998
Rochon-Duvigneaud,
1943; Walls, 1942
Walls, 1942
Walls, 1942
Fite, personal communication
Callum F. Ross
Species
Anthro-19.qxd 12/13/03 7:19 PM Page 504
504
Table 3.
Cordylidae
Teiidae
Anguimorpha
Varanidae
Helodermatidae
Horizontal band-shaped
area with fovea
Diurnal, herbage,
fruit, blossoms, snails,
insects
Lygosoma (Leiolopisma)
entrecasteauxii
Cordylus (Zonurus)
giganteus
Fovea
Diurnal, active
predator
Diurnal, sit-and wait
predator
Fovea
Kahmann, 1936 in Munk,
1970; Kahmann, 1923 in
Rochon-Duvigneaud,
1943
O’Day, 1939 in Peterson,
1992
Kahmann, 1923 in
Rochon-Duvigneaud,
1943
Fovea lacking in most
Mostly nocturnal
Walls, 1942
Eremias argus, Podarcis
muralis
Fovea
Most lacertids diurnal
Tupinambis sp.
Horizontal band-shaped
area with fovea.
Diurnal
Detwiler, 1923 in
Peterson, 1992; Vilter,
1949
Franz, 1934 in Munk, 1970.
Varanus griseus, V. niloticus
Horizontal band-shaped
area with fovea.
Diurnal, V. niloticus
semiaquatic
Heloderma
No fovea
Nocturnal and diurnal
Tarsier Fovea
Lacertoidea
Xantusidae
Lacertiformes
Lacertidae
Trachydosaurus rugosus
Anthro-19.qxd 12/13/03 7:19 PM Page 505
Scincomorpha
Scincoidea
Scincidae
Kahmann, 1933, 1936 in
Munk, 1970. Kahmann,
1923 in RochonDuvigneaud, 1943
Notes: Squamates: Habitat and behavioral data come from Pough et al., 1998; Worell, 1966; Habitat and behavioral data on vine snakes are summarized by
Henderson and Binder, 1980.
505
Anthro-19.qxd 12/13/03 7:19 PM Page 506
Callum F. Ross
506
F
co
id
ea
tif
or
m
es
F
Helodermatidae
Lanthanotus
Varanus niloticus
Varanus griseus
Xenosauridae
Anguidae
Zonurus
Leiolopisma entrecasteauxii
Trachydosaurus
Xantusidae
Uromastyx acanthinurus
Sc
in
La
ce
r
Tupinambis sp.
Podarca
Eremias
Amphisbaenia
Other serpentes
Se
rp
en
te
s
A
Other colubridae
Vine snakes
Gonatodes
Sphaerodactylus
Other gekkonines
Arostelliger praesignis
Aristelliger cochranae
Phelsuma spp.
Basiliscus plumifrons
Anolis spp.
Phrynosoma orbiculaire
Phrynosoma cornutum
Dipsosaurus dorsalis
Iguana iguana
Physignathus leusieur
Amphibolurus barbatus
Agama tournevillii
Chamaeleo
Sphenodon punctatus
Ig
ua
ni
a
G
ek
ko
ni
da
e
Among lepidosaurs, activity pattern is closely correlated with the presence
or absence of a fovea (Table 3). The presence of a fovea in Sphenodon and the
Iguania places a fovea in the squamate stem lineage and the last common
ancestor of lepidosaurs (Figure 8). However, the activity pattern of the lepidosaurian stem lineage cannot be resolved without data on the fossil outgroups as osteological correlates of activity pattern among lepidsauromorphs
have yet to be identified (Underwood, 1970). Among squamates the change
to nocturnality in gekkonines, Xantusidae, Lanthanotus, and helodermatids
was accompanied by loss of the fovea. Most nocturnal gekkos lack foveae, but
those gekkos that secondarily evolved diurnality also regained a fovea (Röll,
2001). Snakes also lacked a fovea primitively, probably because of a phase as
burrowing animals (Figure 8). However, within Serpentes, several taxa
A
A
A
Activity
Nocturnal
Diurnal
Polymorphic
Scleroglossa
Equivocal
Fovea
A
= Afoveate
F
= Foveate
F
Squamata
Lepidosauria
Figure 8. Phylogeny of Lepidosauria with environment mapped on the branches and
most parsimonious hypotheses for gains and losses of foveae indicated with boxes. See
Table 3 for data.
Anthro-19.qxd 12/13/03 7:19 PM Page 507
Tarsier Fovea
507
referred to as “vine snakes” have evolved a temporal fovea as an adaptation to
arboreal predation (Walls, 1942; Fite, personal communication). Other adaptations to this lifestyle include a key-hole shaped pupil, an aphakic gap in some
taxa (a similarity to bathylagine fishes), and a “siting” groove down the side of
the snout which lines up with the fovea and aphakic gap (Henderson and
Binder, 1980). By all accounts “vine snakes” is a polyphyletic assemblage
(Henderson and Binder, 1980), suggesting that they evolved their foveae and
other visual specializations independently.
One of the few comparative studies of ecological correlates of fovea shape
is on anoline lizards (Fite and Lister, 1981). Anolines are the only non-avian
vertebrates with two foveae in each eye, a convexiclivate central fovea and a
more concaviclivate temporal fovea, as found in many predatory birds, particularly raptors. Fite and Lister report an inverse relationship between the steepness of the fovea centralis and prey size. The close correlation between activity
pattern and fovea presence/absence and between ecology and foveal shape
argue for the strong influence of selection on the fovea in lepidosaurs.
The correlated changes test was applied to determine the probability that the
four gains and one loss of a fovea in diurnal lineages and three losses and one gain
in nocturnal lineages could have occurred at random. This probability was found
to be 0.00015. These results suggest a strong link between fovea presence and a
diurnal activity pattern in lepidosaurs. However, in the absence of information on
activity patterns of stem lepidosaurs, it is not possible to determine whether nocturnality or diurnality characterized their foveate last common ancestor.
Archosauria: When dealing with Archosauria the quality of the data varies
and the limitations of a literature review must be acknowledged. Wood’s
(1917) data derive from ophthalmoscopic investigations only, which cannot
definitively discern the presence or absence of foveae. Good histological cross
sections for a variety of birds are provided by Rochon-Duvigneaud (1943),
Oehme (1961), and Kajikawa (1923). Histological sections of bird eyes are
also provided by workers on more restricted samples, such as Moroney and
Pettigrew (1987), Fite and Rosenfield-Wessels (1975), and Reymond (1985,
1987). Sometimes the data provided are contradictory and/or the classification of the animals is in doubt. For example, Wood found no fovea for Anas
boscas, whereas Rochon-Duvignead found a deep convexiclivate fovea for
“A. boschas.” In such cases the histological data were preferred.
Within Archosauria foveae have not been definitively reported outside of
Aves. Munk (1970) reports Chievitz (1889, 1891) as attributing a fovea to
Anthro-19.qxd 12/13/03 7:19 PM Page 508
508
Callum F. Ross
Crocodylus and Alligator, however, this interpretation is not accepted by
Peterson (1992). Further histological examination of crocodyliforms is
required before the primitive condition for archosaurs can be established.
Most birds possess foveae, and although the evidence for Palaeognathae
requires confirmation with histology, Wood (1917) reports one in most of the
taxa he examined. Thus, a fovea is likely to have been present in the last common ancestor of Aves. In fact, foveae were only reported to be absent in the
Cape Penguin, Spheniscus demersus, several ducks and herons, some galliforms,
some coots, two doves, a cockatoo, and the caprimulgiforms. Moreover,
except in the case of caprimulgiforms and some galliforms where the
reports are based on histological examinations (Pettigrew and Konishi,
1984; Pettigrew, personal communication), the absence of a fovea is based
only on Wood’s ophthalmoscopic investigations. Taking Wood’s data at
face value, these absences can be explained with reference to only six
losses of foveae among birds: Once each in the ancestors of Sphenisciformes,
Caprimulgiformes, and Galliformes, and once each within Columbiformes,
Psitacciformes, and Gruiformes.
Many predatory birds actually possess two foveae in each eye: A concaviclivate fovea located temporally and subserving the binocular visual field, and a
convexiclivate fovea centrally located and oriented into the monocular visual
field. The functional hypotheses advanced by Pumphrey and that of Harkness
and Bennet-Clark predict this distribution, given that binocular focus clues are
not available in the monocular visual field. One might imagine that the convexiclivate fovea can be used to scan the world monocularly for movement
indicating prey, whereas the concaviclivate fovea is used to improve acuity and
depth perception in the binocular field, something vital for a bird attacking
terrestrial prey.
Foveae occur in some nocturnal birds, but not others. Strigiformes (owls)
and their probable sister-group, the Caprimulgiformes (goatsuckers, nightjars,
frogmouths, potoos, and oilbirds), are active in the evening and at night.
(Some owls are active during the day, such as the Snowy Owl, but data on its
retinal morphology are incomplete [Wood, 1917].) A fovea has been
confirmed in all owls that have been examined histologically (Table 4). Those
nocturnal caprimulgiforms that have been examined lack a fovea (Pettigrew,
personal communication). Strigiforms were classified with Falconiformes
(diurnal raptors) for many years, a classification revived by Cracraft (1981),
however, DNA hybridization studies place owls closest to the nocturnal
Anthro-19.qxd 12/13/03 7:19 PM Page 509
Table 4.
Occurrences of fovea and other retinal specializations in Archosauria
Classification
Species
Retinal specializations
Crocodylus intermedius
Horizontal band-shaped
area with trough-like fovea
(?)
Horizontal band-shaped
area with trough-like fovea
(?)
Diel period
References
Crocodylia
Alligator mississipiensis
Struthioniformes
Struthionidae
Rheiformes
Rheidae
Tinamiformes
Tinamidae
Apterygiformes
Apterygidae
Podicipediformes
Podicipedidae
Sphenisciformes
Spheniscidae
Chievitz, 1889, 1891 in
Munk, 1970; Underwood, 1970
Casuarius occipitalis
Fovea, Possibly a small
slit-like fovea
“Somewhat
nocturnal”
Wood, 1917; Kajikawa 1923;
Feduccia, 1996
Struthio camelus
Fovea
Diurnal
Wood, 1917
Rhea americana
Fovea
Diurnal
Wood, 1917
Rhyncotus rufescens,
Calodroma elegans
Fovea?
Apteryx mantelli
Fovea?
Nocturnal
Wood, 1917
Podicipes cristatus
Macula and fovea
Diurnal
Wood, 1917
Spheniscus demersus
No fovea
Diurnal
Wood, 1917; RochonDuvigneaud, 1943
Tarsier Fovea
Aves
Casuariiformes
Casuariidae
Chievitz, 1889, 1891 in
Munk, 1970; Underwood, 1970
Wood, 1917
509
Continued
Classification
Procellariiformes
Procellaridae
Pelecaniformes
Phalacrocoracidae
Retinal specializations
Diel period
References
Puffinus griseus
Linear fovea
Diurnal
Wood, 1917
Phalacrocorax carbo
Phalacrocorax
penicillatus
Pelicanus comspicillatus
Sula bassana
Deep wide fovea centralis
Fovea centralis
Diurnal, diver
Rochon-Duvigneaud, 1943
Wood, 1917
Anser cinereus
Small, not very deep,
elongate fovea
Shallow, round, small
fovea centralis
No fovea?
Fovea
Deep fovea centralis
Diurnal
Diurnal
Kajikawa, 1923
Kajikawa, 1923
Diurnal
Diurnal
Diurnal
Anas boscas Dendrocygna
autumnalis, Aix
galericulata
No fovea
Diurnal
Wood, 1917
Wood, 1917
Rochon-Duvigneaud, 1943;
Wood, 1917
Wood, 1917
Phoenicopterus roseus
Slit-like fovea
Diurnal
Wood, 1917
Mycteria americana
Fovea
Diurnal
Wood, 1917
Anser domesticus
Branta canadensis
Anser caerulescens
Anas boschas
Phoenicopteriformes
Phoenicopteridae
Ciconiiformes
Ciconiidae
Fovea?
Fovea?
Wood, 1917
Wood, 1917
Callum F. Ross
Pelecanidae
Sulidae
Anseriformes
Anatidae
Species
Anthro-19.qxd 12/13/03 7:19 PM Page 510
510
Table 4.
Threskiornithidae
Falconiformes
Accipitridae
Ardea purpurea
Ardea cineraea
Ardea occidentalis
Ixobrychus minutus
Nycticorax nycticorax
Botaurus lentiginosus
Cochlearius cochlearius
Plegadis falcinellus
Platalea leucorodia
Deep fovea centralis
Fovea centralis
No fovea
Fovea centralis
No fovea
Two foveae
No fovea
Fovea
Fovea
Diurnal
Diurnal
Diurnal
Diurnal
Nocturnal
Diurnal
Nocturnal
Accipiter nisus
Accipiter nisus
Fovea
Convexiclivate deep
central fovea and
shallower temporal fovea
Steep convexiclivate fovea
centralis, more gentle
convexiclivate fovea
temporalis
Convexiclivate deep
central fovea and
shallower temporal fovea
Convexiclivate deep
central fovea and
shallower temporal fovea
Two foveae
Steep convexiclivate fovea
centralis, more gentle
convexiclivate fovea
temporalis
Fovea centralis, no
temporal fovea in several
well-preserved eyes
Deep fovea centralis, free
of rods
Diurnal
Diurnal
Kajikawa, 1923
Rochon-Duvigneaud, 1943
Diurnal
Fite and RosenfieldWessels, 1975
Diurnal
Reymond, 1987
Diurnal
Rochon-Duvigneaud, 1943
Diurnal
Diurnal
Wood, 1917
Fite and RosenfieldWessels, 1975
Diurnal
Kajikawa, 1923
Diurnal
Kajikawa, 1923
Aquila audax
Buteo buteo
Buteo latissimus
Buteo jamaicensis
Circus aeruginosus
Circus pratensis
Kajikawa, 1923
Kajikawa, 1923
Wood, 1917
Wood,
Wood,
Wood,
Wood,
Wood,
1917
1917
1917
1917
1917
Tarsier Fovea
Accipiter gentilis
Anthro-19.qxd 12/13/03 7:19 PM Page 511
Ardeidae
511
Continued
Classification
Retinal specializations
Diel period
References
Gypaetus barbatus
Haliaeetus leucocephalus
Bifoveate
Fovea centralis (and
nasalis?)
Two foveae
Bifoveate
Bifoveate
Convexiclivate deep
central fovea and
shallower temporal fovea
Convexiclivate fovea
centralis and temporalis
Convexiclivate temporal
fovea
Deep central fovea and
shallower temporal fovea
Convexiclivate deep
central fovea and
shallower temporal fovea
Convexiclivate deep
central fovea and
shallower temporal fovea
Convexiclivate deep
central fovea and
shallower temporal fovea
Distinct fovea centralis
Diurnal
Diurnal
Wood, 1917
Wood, 1917
Diurnal
Nocturnal
Diurnal
Diurnal
Wood, 1917
Pettigrew, personal communication
Pettigrew, 1983
Rochon-Duvigneaud, 1943
Diurnal
Diurnal
Fite and RosenfieldWessels, 1975
Oehme, 1961
Diurnal
Wood, 1917
Diurnal
Wood, 1917; RochonDuvigneaud, 1943
Hunts in bright
daylight, dawn dusk
and at night
Diurnal
Reymond, 1987
Diurnal
Kajikawa, 1923
Diurnal
Rochon-Duvigneaud, 1943
Haliaeetus leucogaster
Elanus scriptus
Elanus notatus
Milvus migrans
Falconidae
Falco sparverius
Falco tinnunculus
Falco sparverius
Tinnunculus alaudarius
Falco tinnunulus?
Falco berigora
Falco subbuteo
Astur plumbarius
Maybe Asturina?
Astur palumbarius
Convexiclivate deep
central fovea and
shallower temporal fovea
Rochon-Duvigneaud, 1943
Callum F. Ross
Species
Anthro-19.qxd 12/13/03 7:19 PM Page 512
512
Table 4.
Diurnal
Wood, 1917
Meleagris galloparvo
Francolinus lathami
Coturnix histrionica
Gallus domesticus
Guttera pucherani
Crax globosa
One convexiclivate fovea
No macula or fovea
No fovea
No macula or fovea
No macula or fovea
No macula or fovea
No macula or fovea?
Diurnal
Domesticated
Rochon-Duvigneaud, 1943
Wood, 1917
Wood, 1917
Wood, 1917
Wood, 1917
Wood, 1917
Wood, 1917
Rallus aquaticus
Fulica americana
Fulica cristata
Aramides ipecaha
Rhynochetus jubatus
Cariama cristata
Otis tarda, Tetrax tetrax
Fovea centralis
No fovea?
No fovea
Fovea and macula
Nasal fovea
Two foveae?
Fovea centralis
Diurnal
Diurnal
Diurnal
Diurnal
Diurnal
Diurnal
Diurnal
Kajikawa, 1923
Wood, 1917
Wood, 1917
Wood, 1917
Wood, 1917
Wood, 1917
Wood, 1917
Vanellus
Plover
Squatarola squatarola
Oedicnemus scolopax
Cepphus columba
Fratercula arctica
Larus argentatus
Indistinct fovea centralis
Concaviclivate fovea
Fovea
Fovea centralis
Fovea
Diurnal
Diurnal
Diurnal
Nocturnal
Diurnal
Kajikawa, 1923
Rochon-Duvigneaud, 1943
Wood, 1917
Wood, 1917
Wood, 1917
Deep, convexiclivate fovea
centralis, concave temporal
fovea
Distinct, streak-like central
fovea (no second fovea in
6 eyes)
Diurnal
Wood, 1917; Kajikawa, 1923
Diurnal
Kajikawa, 1923
Galliformes
Phasianidae
Numididae
Cracidae
Gruiformes
Rallidae
Rhynochetidae
Cariamidae
Otididae
Charadriiformes
Charadriidae
Alcidae
Laridae
Larus canus
Domesticated
Domesticated
Tarsier Fovea
Fovea centralis and
temporal area
Sagittarius serpentarius
Anthro-19.qxd 12/13/03 7:19 PM Page 513
Sagittariidae
513
Continued
Classification
Scolopacidae
Retinal specializations
Diel period
References
Larus maritimus
Larus ridibundus
Seagull
Deep, elongated fovea
Rather deep fovea centralis
Fovea centralis as in
“Sternes”
Fovea centralis
Convexiclivate fovea
centralis
Large central fovea
Fovea
Steep convexiclivate fovea
centralis, more gentle
convexiclivate fovea
temporalis
Fovea in frontal part of
retina
Convexiclivate deep
central fovea and
shallower temporal fovea
Two foveae
Diurnal
Diurnal
Diurnal
Kajikawa, 1923
Kajikawa, 1923
Rochon-Duvigneaud, 1943
Diurnal
Kajikawa, 1923
Kajikawa, 1923
Gallinago
Tringa totanus
Tringa melanoleuca
Numenius hudsonicus
Sterna albifrons
Diurnal
Kajikawa, 1923
Diurnal
Rochon-Duvigneaud, 1943
Leucosarcia picata
Columba palumbus
Goura victoria
Fovea centralis, shallow,
imperfect
No fovea or macula
Faint, doubtful fovea
No fovea or macula
Diurnal
Diurnal
Diurnal
Diurnal
Kajikawa, 1923
Wood, 1917
Wood, 1917
Wood, 1917
Cacatua roseicapilla
Cactua galerita
Chrysotis amazona
Strigops habroptilus
Fovea centralis
No fovea
Fovea
Temporal area, no fovea
Diurnal
Diurnal
Diurnal
Nocturnal
Kajikawa, 1923
Wood, 1917
Wood, 1917
Wood, 1917
Sterna macrura
Sterna minuta
Sterna hirundo
Columbiformes
Columbidae
Psittaciformes
Psittacidae
Diurnal
Wood, 1917
Wood, 1917
Fite and RosenfieldWessels, 1975
Columba domestica
Callum F. Ross
Sternidae
Species
Anthro-19.qxd 12/13/03 7:19 PM Page 514
514
Table 4.
Diurnal
Walls, 1942
Diurnal
Rochon-Duvigneaud, 1943
No fovea reported
No fovea, tapetum in
pigment epithelium
No fovea?
No fovea
Crepuscular
Crepuscular
Walls, 1942
Walls, 1942; Pettigrew, personal
communication; Martin, 1985
Wood, 1917
Pettigrew and Konishi, 1984;
Walls, 1942; Feduccia, 1996
Nocturnal
Tyto alba
Gentle convexiclivate
fovea temporalis
Fovea temporalis
Fovea temporalis
Fovea temporalis
Gentle convexiclivate
fovea temporalis
Concaviclivate temporal
fovea
Convexiclivate temporal
fovea
Concaviclivate temporal
fovea
Shallow temporal fovea
Calypta anna
Two foveae
Coccyzus americanus
Cuculus canorus
One fovea
One macula, fovea?
Apus apus
Swift
Well-developed fovea
temporalis, weak fovea
centralis
One shallow fovea
Anthro-19.qxd 12/13/03 7:19 PM Page 515
Apodiformes
Apodidae
Caprimulgiformes
Caprimulgidae
Podargus
Goat suckers
Steatornithidae
Caprimulgus europaeus
Steatornis
Bubo virginianus
Bubo bubo
Athene cunicularia
Athene noctua
Glaucidium radiatum
Asio otus
Asio flammeus
Strix aluco
Tytonidae
Trochiliformes
Trochilidae
Cuculiformes
Cuculidae
Nocturnal
Nocturnal
Nocturnal
Nocturnal
Nocturnal
Nocturnal
Nocturnal
Fite and RosenfieldWessels, 1975
Rochon-Duvigneaud, 1943
Wood, 1917
Rochon-Duvigneaud, 1943
Fite and RosenfieldWessels, 1975
Oehme, 1961
Tarsier Fovea
Strigiformes
Strigidae
Crepuscular
Nocturnal, cave
living, echolocation
Rochon-Duvigneaud, 1943;
Oehme, 1961
Wood, 1917; Oehme, 1961
Oehme, 1961
Diurnal
Wood, 1917
Wood, 1917
Wood, 1917
515
Classification
Species
Coraciiformes
Alcedinidae
Alcedo ispida
Halcyon sancta
Forest kingfisher
Halcyon macleayii
Azure kingfisher
Ceyx azureus
Laughing kookaburra
Dacelo gigas
Blue-winged
kookaburra
Dacelo leachii
Upupidae
Common Hoopoe
Meropidae—Bee-eaters
Rainbow bee-eater
Piciformes
Picidae—Woodpeckers
Green woodpecker
Diel period
References
Convexiclivate deep
central fovea and
shallower temporal
Two foveae
Convexiclivate deep
central fovea and
shallower temporal
Convexiclivate deep
central fovea and
shallower temporal
Convexiclivate deep
central fovea and
shallower temporal
Convexiclivate deep
central fovea and
shallower temporal
Convexiclivate deep
central fovea and
shallower temporal
Diurnal and partly
aquatic
Rochon-Duvigneaud, 1943
fovea
Diurnal and partly
aquatic
Wood, 1917
Moroney and Pettigrew, 1987
fovea
Diurnal and partly
aquatic
Moroney and Pettigrew, 1987
Diurnal and partly
aquatic
Moroney and Pettigrew, 1987
Diurnal and partly
aquatic
Wood, 1917; Moroney and
Pettigrew, 1987
Diurnal and partly
aquatic
Moroney and Pettigrew, 1987
fovea
fovea
fovea
fovea
Upupa epops
Shallow fovea centralis
Diurnal
Wood, 1917
Merops ornatus
Convexiclivate deep
central fovea and
shallower temporal fovea
Diurnal, aerial feeder
Moroney and Pettigrew, 1987
Picus minor
Picus viridis
Clearly visible fovea
Convexiclivate deep
central fovea and
shallower temporal fovea
Diurnal
Diurnal
Kajikawa, 1923
Rochon-Duvigneaud, 1943
Callum F. Ross
European kingfisher
Sacred kingfisher
Retinal specializations
Anthro-19.qxd 12/13/03 7:19 PM Page 516
Continued
516
Table 4.
Ramphastidae
Sulphur-breasted toucan
Passeriformes
Corvidae—Crows
Jay
Jay
Diurnal
Diurnal
Diurnal
Kajikawa, 1923
Wood, 1917
Wood, 1917
Convexiclivate deep
central fovea and
shallower temporal fovea
Diurnal
Wood, 1917; RochonDuvigneaud, 1943
Ramphastos sulfuratus
Temporal fovea
Diurnal
Wood, 1917
Garrulus glandarius
Garrulus glandarius
Fovea centralis
Convexiclivate deep
central fovea and
shallower temporal fovea
Fovea centralis
Convexiclivate deep
central fovea and
shallower temporal fovea
Convexiclivate deep
central fovea and
shallower temporal fovea
Fovea centralis within area
Diurnal
Diurnal
Kajikawa, 1923
Rochon-Duvigneaud, 1943
Diurnal
Diurnal
Kajikawa, 1923
Rochon-Duvigneaud, 1943
Diurnal
Rochon-Duvigneaud, 1943
Diurnal
Kajikawa, 1923
Convexiclivate deep
central fovea and
shallower temporal fovea
Corvus corax
Central fovea
Corvus americanus
Central fovea
Cyanocitta cristata (and Gentle convexiclivate
C. stelleri)
fovea centralis
Pica pica
Convexiclivate deep
central fovea and
shallower temporal fovea
Diurnal
Rochon-Duvigneaud, 1943
Diurnal
Diurnal
Diurnal
Wood, 1917
Wood, 1917
Wood, 1917; Fite and
Rosenfield-Wessels, 1975
Rochon-Duvigneaud, 1943
Corvus frugilugis
Corvus frugilegus
Corvus cornix
Jackdaw
Jackdaw
Raven
American crow
Blue jay
Black-billed magpie
Coleus monedula
⫽ Corvus monedula
Coloeus monedula
⫽ Corvus monedula
Diurnal
517
Fovea centralis
Fovea centralis
Fovea centralis
Tarsier Fovea
Rook
Rook
Jynx torquilla
Colaptes mexicanus
Melanerpes
erythrocephalus
Dendrocopus major
Anthro-19.qxd 12/13/03 7:19 PM Page 517
Northern wryneck
Red-shafted flicker
Red-headed
woodpecker
Great spotted
woodpecker
Continued
Classification
Retinal specializations
Diel period
References
Loxia curvirostra
Diurnal
Kajikawa, 1923
Diurnal
Kajikawa, 1923
Fringilla montifringilla
Fringilla coelebs
Round, deep fovea
centralis
Round, deep fovea
centralis
Fovea centralis
Fovea centralis
Diurnal
Diurnal
Diurnal
Kajikawa, 1923
Kajikawa, 1923
Wood, 1917
Pitangus derbianus
Fovea
Hirundo rustica
Two foveae: centralis
containing only rods,
convexiclivate “fovea
externa”
Bluebird
Wheatear
Nectariniidae—Sunbirds
Muscicapa
Sialia sialis
Oenanthe oenanthe
Nectarina chalypea
Fovea centralis
Bifoveate
Fovea centralis
Convexiclivate fovea
centralis
Paridae—True Tits
Crested tit
Parus cristatus
Convexiclivate deep
central fovea and
shallower temporal fovea
Lanius ludovicianus
Bifoveate
Diurnal
Wood, 1917
Parotia lawii
Nasal fovea
Diurnal
Diurnal
Wood, 1917
Passer domesticus
Deep fovea centralis
Fringillidae—Finches
Common (Red)
Crossbill
Carduelis
Brambling
Chaffinch
Tyrannidae
Derby tyrant
Hirundinidae—Swallows
Barn swallow
Diurnal
Muscicapidae—Flycatchers
Laniidae
California Shrike
Paradisaeidae
Law bird of paradise
Passeridae—Sparrows
Domestic sparrow
Wood, 1917; Kajikawa, 1923
Diurnal
Diurnal
Diurnal
Diurnal
Kajikawa, 1923
Wood, 1917
Wood, 1917
Fite and RosenfieldWessels, 1975
Diurnal
Rochon-Duvigneaud, 1943
Kajikawa, 1923
Callum F. Ross
Species
Anthro-19.qxd 12/13/03 7:19 PM Page 518
518
Table 4.
Mimidae
Western mockingbird
Sturnidae—Starlings
European starling
Turdidae—Thrushes
Blackbird
Anthro-19.qxd 12/13/03 7:19 PM Page 519
Sittidae—Nuthatches
Eurasian nuthatch
Diurnal
Sitta europaea
Convexiclivate deep
central fovea and
shallower temporal fovea
Rochon-Duvigneaud, 1943
Mimus polyglottos and
M. orpheus
Bifoveate
Diurnal
Wood, 1917
Sturnus vulgaris
Convexiclivate fovea
Diurnal
Oehme, 1961
Turdus musicus
Turdus merula
Fovea
Convexiclivate deep
central fovea and
shallower temporal fovea
Diurnal
Diurnal
Kajikawa, 1923
Wood, 1917; RochonDuvigneaud, 1943
Tarsier Fovea
Kajikawa (1923) reviews the literature up to 1923. Primary references may be found there. Data from Kajikawa are obtained from a translation by Dr. B. Demes.
Kajikawa describes the position of the area and the shape of the fovea; adjectives of mittel (average, medium); seicht (shallow); schwach (weak); tief (deep). Data
in his table are not compatible with data in his text.
519
Anthro-19.qxd 12/13/03 7:19 PM Page 520
520
Callum F. Ross
Caprimulgiformes. Irrespective of whether caprimulgiforms and owls are sister
taxa or not, they are certainly descended from a diurnal ancestor which, like
most birds, was probably diurnal and foveate.
The only other nocturnal birds exhibiting foveae are the Letter-winged kite
(Elanus scriptus), the Stone plover (“Oedicnemus scolopax”), and possibly the
kiwi (Apteryx) although histology is needed to confirm the latter two. The most
notable example is that of Elanus scriptus, which is bifoveate and nocturnal
(Pettigrew personal communication), with a diurnal bifoveate sister taxon,
Elanus notatus. Several nocturnal birds have lost foveae: The American blackcrowned night-heron (Nycticorax nycticorax), the Boat-billed night heron,
Cochlearius cochlearius, and the New Zealand kakapo, Strigops habroptilus.
The concentrated changes test cannot be applied to the avian data because
the phylogeny is too poorly known. Although further histological work on
palaeognath retinae is needed, it would seem that a fovea was present in the
last common avian ancestor. Whether this ancestor was diurnal or nocturnal
cannot be established at present. Palaeognaths display an array of adaptions
including nocturnality and crepuscularity and the activity patterns of stem
avians have yet to be documented. The fovea has been lost in five lineages
(Galliformes, Sphenisciformes, Columbiformes, and Caprimulgiformes, as well
as within Psittaciformes), two of which are nocturnal and three of which are
not. Possibly three lineages have retained foveae as they evolved from diurnal
to nocturnal environments, their ancestors probably being diurnal, although
for owls and kiwis this need not have been the case. In sum, it is not known
whether the avian fovea originally evolved in a diurnal or a nocturnal lineage,
but it is clear that a nocturnal environment is neither sufficient nor necessary
to eliminate a fovea once established.
DISCUSSION
Functional Analysis
As noted above, hypotheses of foveal function invoke advantages of three
features of foveae: High densities of photoreceptors and ganglion cells combined with a low ratio of ganglion cells to photoreceptors; a depression in the
retina caused by centrifugal displacement of the inner nuclear, inner plexiform,
ganglion cell, and nerve fiber layers (Figure 2B); and a lack of retinal circulation at the fovea.
Anthro-19.qxd 12/13/03 7:19 PM Page 521
Tarsier Fovea
521
Function of Increased Photoreceptor and Ganglion Cell Densities: All
foveate vertebrates that have been examined show increases in the number of
photoreceptors and ganglion cells at the fovea (Collin, 1999; Collin and
Partridge, 1996; Fite and Lister, 1981; Fite and Rosenfield-Wessels, 1975;
Munk, 1966). These concentrations, as well as the low ratio of photoreceptors
to ganglion cells, make the fovea an area of enhanced visual acuity. Both
photopic and scotopically adapted animals exhibit concentrations of photoreceptors and ganglion cells that serve as adaptations for enhanced acuity. For
example, some deep-sea teleosts pack more rods into a smaller area of the
retina by arranging them in multiple banks (Table 2). Bathylagus benedicti has
six banks of rods at the fovea in comparison with three banks elsewhere in the
retina, and the alepocephalid Bajacalifornia drakei has anywhere from 21 to
28 banks of rods at the fovea (Locket, 1985). Clearly these animals are
attempting to maintain acuity while increasing retinal sensitivity. Tarsius
evinces the highest peak ganglion cell density of any nocturnal primate (ca.
13,300 mm2) (Tetreault et al., this volume), and very high cone densities
around the fovea as well (50,000 mm2 and 85,000 mm2 in two individuals)
(Hendrickson et al., 2000; Hendrickson, personal communication). This suggests that the tarsier fovea is also an area of enhanced visual acuity.
In the context of these adaptations for high acuity (high cone and ganglion
cell densities relative to other nocturnal primates), the absence of a tapetum in
Tarsius is expected. Tapeta increase sensitivity by reflecting photons back
through the retina, but they also scatter light rays, degrading the retinal image
and decreasing visual acuity. Consequently, if foveae are adaptations for high
visual acuity, they are expected not to be associated with tapeta. For this reason caprimulgiforms are not only the only avians with a tapetum, but also one
of very few lacking foveae. Indeed, across all vertebrates, a tapetum is only associated with a retinal fovea in Scopelosaurus lepidus (Tables 2, 3, 4). The lack of
a tapetum in a highly visually dependent nocturnal animal like Tarsius is therefore expected. The hypothesis that the tapetum’s absence is purely a primitive
hold-over from a diurnal ancestor (Cartmill, 1980; Martin, 1979; Ross, 1996)
need not be true.
Function of the Retinal Pit: The calculations made here suggest that the
pit-shaped profile of the tarsier fovea does not produce visually detectable
image enlargement or movement enhancement. The validity of this conclusion
depends on the validity of several assumptions regarding the parameters used
in calculating the pit’s optical effect. There are three reasons to believe this
Anthro-19.qxd 12/13/03 7:19 PM Page 522
522
Callum F. Ross
conclusion is robust. First, estimates of the parameters always erred in favor of
the hypothesis that tarsiers have greater acuity than they probably do. Second,
the tarsier fovea is morphologically similar to that of macaques, which have
3–4 ganglion cells per cone at the fovea, and a maximum behavioral visual
acuity of 3.3 m (Wässle et al., 1990). The maximum gain produced by the
tarsier fovea was 2.3 m, below values significant even to macaques, let alone
a nocturnal animal with significant summation of photoreceptors to ganglion
cells (Rohen and Castenholtz, 1967). The tarsier fovea—indeed, any shallow,
concaviclivate fovea—cannot function to enhance movement or improve acuity in the absence of a detectable optical effect. Third, the human fovea is similar to that of Tarsius in the steepness of its clivus, and Walls (1940) found the
gain produced by the human fovea to be functionally insignificant. Similar
comments have been made about the shallow bathylagine fovea (Locket,
1985) and are probably valid for many owls as well (Table 1).
What then is the function of the concaviclivate fovea? Details on the optical
properties of different layers of retinal tissues are scant and these studies have not
separated the effects of cell bodies from those of capillaries. Ohzu and Enoch
(1972) estimated the ability of the foveal retina to transfer an image by determining the modulation transfer function of isolated retinal tissues. The modulation transfer function (MTF) describes the changes (modulation) in quality of an
image transferred by the retina at different spatial frequencies (Ohzu et al.,
1972). Ohzu and Enoch (1972) calculated the MTF for a sinusoidally varying
grating transferred by the foveal and parafoveal regions in freshly obtained retinal
tissues of three humans. They found that the foveal region transferred the image
better than the parafoveal region, evidence that Hughes (1977) cites in support
of the hypothesis that the inner layers of the retina degrade image quality. Ohzu
and Enoch also note that the MTF of the human fovea is better than that of the
albino rat nonfoveal retina reported in Ohzu et al. (1972). However, measures of
MTF are sensitive to the time elapsed since the retina was removed from the eye:
The longer the retina is outside the eye, the worse its MTF. Moreover, the MTF
of the fovea deteriorates slower than the surrounding tissues (Ohzu and Enoch,
1972), suggesting that the MTF of the human fovea might be better than the
parafoveal region because it has deteriorated less postmortem.
The best data available on the optical properties of different regions of the
human retina come from the laboratory of Jean-Marie Gorrand. Rather than
passing an image through the retina, a technique limited to the ex vivo
environment, Gorrand measured image quality by reflecting the image off the
retina in vivo. He found the degree of light scattering in the retina to be least
Anthro-19.qxd 12/13/03 7:19 PM Page 523
Tarsier Fovea
523
at the fovea and to increase away from the fovea and the MTF to be highest
at the fovea and to decrease away from the fovea (Gorrand, 1979). Subsequent
work on two individuals confirmed that the MTF is higher on the fovea than
off it, but that the mean irradiance (image brightness) is not consistently
higher or lower off the fovea than on it (Gorrand, 1989).
Being based only on humans and only on a few individuals, these data cannot be indiscriminately applied to all vertebrates. However, they support the
hypothesis that the concaviclivate fovea improves image quality by increasing
the MTF and hence the quality of the image received by the photoreceptors.
If these data are applicable to all vertebrates, then this is a plausible function
for the gently sloping foveae of some primates (tarsiers and anthropoids), some
lepidosaurs (including Sphenodon), and some birds, including owls.
Function of the Absence of Blood Vessels: Weale (1966) argued that the
inner layers of the retina are not a significant impediment to acuity because
they are highly transparent and are juxtaposed to the photoreceptors. He
instead argued that the real impediment to high acuity vision came from the
capillaries and the blood cells within them refracting and absorbing photons.
Behavioral observations on humans and anatomical observations on Macaca
and Saimiri suggest that the retinal vasculature might indeed be a significant
impediment to high acuity vision in primates. Snodderly and colleagues
(1978) used early eye-tracking technology to precisely map the position on the
retina of images presented in different parts of the visual field. They found that
subjects found it more difficult to see images presented to the parts of the
visual field that projected onto the capillaries within the retina. Moreover,
Adams and Horton (2002) recently noted that the blood vessels also cast a
shadow on the projection of the visual field into primary visual cortex. This
suggests that the blood vessels have the capability of significantly decreasing
image quality or brightness. This result takes on special significance when we
consider that on average 40% of the retinal surface area outside the fovea is
obscured by blood vessels in Macaca fascicularis (Snodderly et al., 1992) and
Saimiri (Snodderly and Weinhaus, 1990). This means that 40% of the photons
reaching the photoreceptors have passed through and been refracted or
absorbed by capillaries and the blood cells in them!
Comparative data support the hypothesis that blood vessels within the
retina or on its vitread surface are an impediment to high acuity vision. All vertebrates except eels2 have blood vessels in the choroid (choriocapilliaris) that
2
Eels have an intraretinal blood supply, like mammals, but possess neither an area nor a fovea.
Anthro-19.qxd 12/13/03 7:19 PM Page 524
524
Callum F. Ross
supply the “rods and cones” layer from behind, but the largest distance over
which oxygen can diffuse from the choriocapillaris is about 143m (data from
humans, Dollery et al., 1969). Thus, vertebrates with thicker retinae, or with
retinal thickenings such as seen in areae centrales around foveae, need to augment retinal nutrition from the vitread side or within the retina. There are various ways of doing this.
In mammals retinal vascularization occurs between the layers of the retina,
not on the surface, and is associated with retinal thicknesses over 143 m
(Chase, 1982). The retinae of some rodents, some bats, and some marsupials
are avascular and must receive their nutrients from the choroid (Buttery et al.,
1990). Primates have intraretinal vasculature (see Michaelson, 1954, Wolin and
Massopust, 1970) except for the foveal region of haplorhines, which is avascular. Tarsius bancanus is reported to have capillaries in its parafoveal region
(Rohen, 1966; Figure 2; Wolin and Massopust, 1970), although Hendrickson
et al. (2000) report the Tarsius spectrum fovea to be avascular. The image of the
fovea of T. spectrum supplied by Hendrickson reveals blood vessels lying in the
outer edge of the inner nuclear layer (Figure 2B). Large caliber vessels are not
seen within 350 m of the center of the foveal pit, and small vessels can only
be seen approximately 100 m from the fovea center. Although these latter vessels are located within the retina underlying the sloping edges of the fovea, in
a position similar to those reported by Wolin and Massopust (1970), the very
center of the fovea (foveola) is devoid of blood vessels.
Some workers have argued that the lack of retinal vasculature in the mammalian fovea accounts for the absence of ganglion cells and interneurons in
the fovea (Rodieck, 1988; Weale, 1966), implying that the pit itself does not
have a function, but is merely an effect of the absence of blood vessels. Provis
et al. (1998) refined this argument by suggesting that in anthropoids the
foveal ganglion cell and inner nuclear layers migrate centrifugally toward the
encroaching retinal vasculature because of the lengthening of the photoreceptor outer segments during ontogeny. Foveal photoreceptors are more densely
packed than in the peripheral retina because they have smaller diameters and
longer outer segments. As a result, the whole photoreceptor layer is thicker in
the fovea than elsewhere in the retina. Photoreceptors are the principal consumers of oxygen from the choroid capillary bed (Provis et al., 1998, p. 575),
so their increased numbers and increased length might be starving the inner
layers of the retina of oxygen. Certainly mammals with thick retinae tend to
have vascular retinae, whereas those with thin retinae are avascular (Chase,
Anthro-19.qxd 12/13/03 7:19 PM Page 525
Tarsier Fovea
525
1982), suggesting constraints on retinal thickness. Moreover, absolute fovea
area is conserved across a wide range of body and eye sizes in anthropoids,
even though this results in the fovea subserving a progressively smaller angle
of the visual field with increasing eye size (Franco et al., 2000). This is also
suggestive of a physiological constraint on fovea size (Franco et al., 2000).
In birds the retina is thicker than that of mammals (300m), but there is
no intraretinal circulation and no vessels traverse the retina at the vitread surface. The retina instead receives its nutrients from (in addition to the choroid)
the vitreous, which is supplied by the pecten, a large vascular appendage projecting into the vitreous from the optic disk. To diffuse nutrients from the
pecten, birds have evolved an unusual cyclotorsional component to saccadic
eye movements (Pettigrew et al., 1990). The absence of blood vessels in front
of the photoreceptor layer in birds seems to have encouraged the evolution of
foveae. The same may also be true of lepidosaurs, among which foveae are
widely distributed, and which resemble birds in lacking retinal vessels and
manifesting a (nonhomologous) richly vascularized cone (conus papillaris)
extending vitread from the optic disk. Among lepidosaurs, only snakes have
blood vessels on the vitread surface of the retina. The relationship of these
blood vessels to the fovea in “vine snakes” is unknown.
Thus, the comparative data suggest that high acuity vision, foveae, and
blood vessels do not mix, but they also suggest that foveae can develop, as in
birds and lizards, even in the absence of intraretinal blood vessels. This calls
into question the hypothesis that the foveal pit develops merely as an effect of
the absence of blood vessels (Provis et al., 1998). It also suggests that birds,
lizards, and therefore possibly primates, derive some benefit from centrifugal
displacement of the inner retinal layers, whether or not this is associated with
exclusion of blood vessels from the fovea. Finally, it is noteworthy that among
teleosts, the relationship of the vitreal circulation to the fovea has only been
reported in one taxon, the sandlance, Limnichthyes fascaitus, in which the convexiclivate fovea is richly vascularized (Collin and Collin, 1988a, 1988b).
Comparative Analysis
The comparative analysis reported here evaluated the hypothesis that foveae
typically arise among diurnal forms. Given the likely functional differences
between convexiclivate and concaviclivate foveae, it is necessary to distinguish
between them.
Anthro-19.qxd 12/13/03 7:19 PM Page 526
526
Callum F. Ross
Foveae have arisen in four lineages of teleost fishes. The mesopelagic or
bathypelagic stem lineage of Alepocephaloidei evolved a convexiclivate fovea;
the mesopelagic or epipelagic stem lineage of Bathylagus evolved a concaviclivate fovea; and the epi-, meso-, and bathypelagic notosudid Chlorophthalmoidei
acquired a convexiclivate fovea as an adaptation to scotopic vision (Steenstrup
and Munk, 1980). Foveae are widespread among the Acanthomorpha but a
lack of precise morphological data on fovea shape and poor sampling in most
taxa renders definitive conclusions impossible. Wulliman (1997) has postulated that there was a rearrangement of the visual brainstem at the origin of
acanthomorphs concomitant with their invasion of reef habitats in the Late
Cretaceous. Possible links between these brainstem arrangements and retinal
anatomy remain to be explored. In sum, among fishes convexiclivate foveae
have evolved in low light levels, with a convexiclivate fovea being strongly
associated with a scotopic environment. The concaviclivate foveae of bathylagines may have arisen in either a scotopic or photopic environment.
Among lepidosaurs there is a strong link between the presence of a fovea and
a diurnal activity pattern. Several nocturnal lineages lose foveae; gekkos reevolve
one on returning to diurnality (Röll, 2001). In the absence of information on
activity patterns of stem lepidosaurs, it is not possible to determine whether nocturnality or diurnality characterized their foveate last common ancestor.
Among archosaurs, a nocturnal environment is neither sufficient nor necessary to eliminate a fovea once it is established. A fovea was probably present in
the last common avian ancestor although it is currently impossible to determine whether this ancestor was diurnal or nocturnal. Palaeognaths display an
array of adaptions including nocturnality and crepuscularity, and histological
work on palaeognath retinae is needed.
Among primates, crown group haplorhines (tarsiers and anthropoids) have
foveae, but tarsiers are dedicated nocturnal animals and anthropoids are predominantly diurnal. The probable outgroup to the tarsier–anthropoid clade,
the Omomyidae, was wholly nocturnal,3 as the stem lineage of primates probably was. The most parsimonious reconstructions of character state evolution
among tarsiers, anthropoids, and omomyids have the fovea arising in a nocturnal lineage (Figure 1).
Overall, the comparative data suggest that convexiclivate foveae can arise
in either diurnal or nocturnal lineages. However, it is noteworthy that there is
3
Rooneyia is not an omomyid (Kay et al., 1997; Ross et al., 1998).
Anthro-19.qxd 12/13/03 7:19 PM Page 527
Tarsier Fovea
527
no instance in which a concaviclivate fovea of the kind possessed by
haplorhines definitively evolved in a scotopic setting. Those scotopically
adapted vertebrates that have gently sloping foveae (Tarsius, Sphenodon,
Strigiformes) are all sister-groups to diverse radiations of diurnal visually
adapted animals, providing some support for Cartmill’s (1980) hypothesis.
The Function of the Tarsier Fovea
What then is the function of the tarsier fovea? The high concentration of cones
and ganglion cells, the high ratio of ganglion cells to photoreceptors, the exclusion of blood vessels from the foveola, and the lack of a tapetum are arguably
adaptations for increased visual acuity related to sit-and-wait ambush predation.
Tarsius and owls share many features cited as adaptations to nocturnal sit-andwait ambush predation, including large eyes protruding from the skull; a long
olfactory tract displaced dorsally by the enlarged orbits; a well-developed auditory
apparatus used for locating prey; loss of optic mobility with concomitant increase
in cervical mobility; and enlarged semicircular canals (Niemitz, 1985). To these
might be added the presence of postorbital ossifications to deflect jaw adductors
around the orbital contents (Ross, 1996, 2000), and concaviclivate retinal foveae.
Although the comparative data suggest that the concaviclivate foveae of
owls, tarsiers and tuataras are useful in nocturnal settings, recent observations
suggest that tarsiers might be active in light levels supporting cone vision.
Gursky reports that in contrast with most nocturnal mammals, spectral tarsiers
increase their activity levels on moonlit nights (1999). Hendrickson et al.
(2000, p. 729) report:
“Recent behavioral observations of T. spectrum in the large home cages of
their colony at Bogor find that they become active several hours before sundown, are relatively quiet in the middle of the night, and then active again just
before and after dawn (D. Sajuthi, personal communication). Light intensities
in the tropics during these hours are high enough to support cone vision.”
These observations suggest that although tarsiers are nocturnal in the
sense of being active at night, they may be adapted to function in light levels
in which their cone rich foveae can provide relatively high acuity vision.
Likely adaptations for these conditions include cone densities as high as
50–85,000 mm2, a fovea, no tapetum, and relatively high degrees of orbital
convergence. Given that these features also perform the same functions in
extant anthropoids, it is reasonable to hypothesize that they were also present
Anthro-19.qxd 12/13/03 7:19 PM Page 528
528
Callum F. Ross
in the last common ancestor of tarsiers and anthropoids. Whether or not that
animal was nocturnal or diurnal, it was adapted for activity in high light levels.
CONCLUSIONS
The increase in photoreceptors, ganglion cells, and the ganglion cell:photoreceptor ratio in the tarsier fovea, and the lack of a tapetum behind it, argue for
a role in improving visual acuity. The fovea does not provide a demonstrable
increase in image size, nor does it provide significant gain for movement detection. Blood vessels are found in the periphery of the tarsier fovea, but not in
the foveola, supporting the hypothesis that the fovea is related to exclusion of
blood vessels from the foveal region. It is therefore plausible that the fovea
(and lack of a tapetum) function to improve image quality by displacing inner
layers of the retina and blood vessels away from the path of incoming light.
Although this effect is of uncertain significance for scotopically adapted animals like tarsiers with extensive summation of photoreceptors to ganglion
cells, the comparative data do indicate that shallow foveae with gently sloping
sides are found in scotopically adapted visual predators, particularly of a sitand-wait ambush variety: That is, owls, tuataras, bathylagine argentinioids, and
tarsiers. These animals might profit from the thinning of the foveal retina,
because it removes either cells, blood vessels, or both from the inner retinal
layers, thereby improving image quality. More and better data are needed on
the optical properties of the retinae, and on the distribution and size of blood
vessels in the foveal regions of these animals. In addition, Woollard (1925) and
Castenholtz (1965) both report that not all tarsiers have a fovea. More data
are needed on the prevalence of foveae among living tarsiers, sphenodons,
slickheads, and owls.
The comparative evidence does not corroborate Walls’ hypothesis that
foveae only arise with a convexiclivate shape and are only concaviclivate as
degenerate “primitive retentions.” Rather, concaviclivate foveae can appear de
novo. However, the comparative data do support the hypothesis that concaviclivate foveae are most likely to evolve in diurnal lineages: There are no definitive instances of concaviclivate foveae arising in a nocturnal lineage, as,
tuataras, owls, and tarsiers have diurnal, foveate close relatives, and bathylgines
migrate up into the epipelagic, photopic environment.
The haplorhine fovea initially functioned to improve the quality of the
image falling on the retina. Image quality in these stem haplorhines would
have been improved by convergence of the optic on the visual axis associated
Anthro-19.qxd 12/13/03 7:19 PM Page 529
Tarsier Fovea
529
with the increased orbital convergence characteristic of primates in general.
Whether the improved image quality facilitated the shift to diurnality in stem
haplorhines or evolved in response to it cannot be determined from the available data. Regardless, it provided the substrate on which the remarkable visual
acuity of extant anthropoids would subsequently be developed.
ACKNOWLEDGMENTS
Brigitte Demes provided helpful discussion, assistance with translation, and
read and commented on the manuscript. Brian Sheehy helped with computations. Pat O’Connor helped with the literature on bird classification. Shaun
Collin, Jack Pettigrew, and Katherine Fite provided unpublished data and
informative discussion. John Allman, Brian Boycott, and Chris Kirk read the
manuscript and gave stimulating comments and discussion that improved the
paper. Luci Betti did the illustrations. Susan Larson assisted with digitizing
foveae. Anita Hendrickson kindly supplied an image of the fovea of Tarsius
spectrum and unpublished data on cone densities in the tarsier fovea. Austin
Roorda provided helpful information on retinal imaging.
REFERENCES
Adams, D. L., and Horton, J. C., 2002, Shadows cast by retinal blood vessels mapped
in primary visual cortex, Science 298:572–576.
Ahlbert, I.-B., 1976, Organization of the cone cells in the retinae of salmon (Salmo
salar) and trout (Salmo trutta trutta) in relation to their feeding habits, Acta Zool.
(Stockholm) 57:13–35.
Ali, M. A., 1959, The ocular structure, retinomotor and photobehavioral responses of
juvenile pacific salmon, Can. J. Zool. 37:965–996.
Ali, M. A., Anctil, M., and Raymond, N., 1973, La rétine de quelques poisssons marins
du littoral brésilien, Revta Biol. Lisb. 9:101–114.
Ali, M. A., and Hanyu, I., 1963, A comparative study of retinal structure in some fishes
from moderately deep waters of the western north Atlantic, Can. J. Zool.
41:225–241.
Baldwin, C. C., and Johnson, G. D., 1996, Interrelationships of Aulopiformes, in:
Interrelationships of Fishes, M. L. J., Stiassny, L. R., Parenti, and G. D., Johnson,
eds., Academic Press Inc, San Diego, pp. 355–404.
Beaudet, L., Novales Flamarique, I., and Hawrhyshyn, C. W., 1997, Cone photoreceptor topography in retina of sexually mature pacific salmonid fishes, J. Comp.
Neurol. 383:49–59.
Brauer, A., 1908, Die Tiefsee-Fische, 2. Anatomischer Teil. Fischer, Jena.
Anthro-19.qxd 12/13/03 7:19 PM Page 530
530
Callum F. Ross
Buttery, R. G., Haight, J. R., and Bell, K., 1990, Vascular and avascular retinae in
mammals. A fundascopic and fluorescin angiographic study, Brain Behav. Evol.
35:156–175.
Calkins, D. J., 2001, Seeing with S cones, Prog. Retin. Eye Res. 20:255–287.
Cartmill, M., 1980, Morphology, Function and Evolution of the Anthropoid
Postorbital Septum, in: Evolutionary Biology of the New World Monkeys and
Continental Drift, R. L., Ciochon, and A. B., Chiarelli, eds., Plenum Press, New
York, pp. 243–274.
Castenholtz, E., 1965, Über die Struktur der Netzhautmitte bei Primaten,
Z. Zellforsch. 65:646–661.
Castenholtz, A., 1984, The Eye of Tarsius, in: Biology of Tarsiers, C., Niemitz, ed.,
Gustav-Fisher Verlag, Stuttgart, pp. 303–318.
Chase, J., 1982, The evolution of retinal vascularization in mammals. A comparison of
vascular and avascular retinae, Ophthalmology 89:1518–1525.
Chievitz, J. H., 1889, Untersuchungen über die Area centralis retinae, Arch Anat.
(Physiol.) 1889:332–366.
Chievitz, J. H., 1891, Über das Vorkommen der Area centralis retinae in den vier
höheren Wirbelthierklassen, Arch. Anat. Entw. 1891:11–334.
Collin, S. P., 1999, Behavioural Ecology and Retinal Cell Topography, in: Adaptive
Mechanisms in the Ecology of Vision, S. N., Archer, M. B. A., Djamgoz, E. R., Loew,
J. C., Partridge, and S., Vallerga, eds., Kluwer Academic Publishers, Dordrecht.
Collin, S. P., and Collin, H. B., 1988a, The morphology of the retina and lens of the
sandlance, Limnichthyes fasciatus (Creeiidae), Exp. Biol. 47:209–218.
Collin, S. P., and Collin, H. B., 1988b, Topographic analysis of the retinal ganglion
cell layer and optic nerve in the sandlance Limnichthyes fascaitus (Creeiidae,
Perciformes), J. Comp. Neurol. 278:226–241.
Collin, S. P., Hoskins, R. V., and Partridge, J. C., 1997, Tubular eyes of deep-sea
fishes: A comparative study of retinal topography, Brain Behav. Evol. 50:335–357.
Collin, S. P., Hoskins, R. V., and Partridge, J. C., 1998, Seven retinal specializations
in the tubular eye of the deep-sea Pearl-Eye, Scopelarchus michaelsarsi: A case study
in visual optimization, Brain Behav. Evol. 51:291–314.
Collin, S. P. and Partridge, J. C. 1996, Retinal specializations in the eyes of deep-sea
teleosts, J. Fish Biol. 49 (Suppl. A): 157–174.
Cracraft, J., 1981, Toward a phylogenetic classification of the recent birds of the world
(class Aves), Auk 98:681–714.
Curcio, C. A., Sloan, K. R. J., Packer, O., Hendrickson, A. E., and Kalina, R. E., 1987,
Distribution of cones in human and monkey retina: Individual variability and radial
asymmetry, Science 236:579–582.
Curcio, C. A., Sloan, K. R. J., Kalina, R. E., and Hendrickson, A. E., 1990, Human
photoreceptor topography, J. Comp. Neurol. 292:497–523.
Anthro-19.qxd 12/13/03 7:19 PM Page 531
Tarsier Fovea
531
DeBruyn, E. J., Wise, V. L., and Casagrande, V. A., 1980, The size and topographic
arrangement of retinal ganglion cells in the galago, Vision Res. 20:315–327.
Dollery, C. T., Bulpitt, C. J., and Kohner, E. M., 1969, Oxygen supply to the retina
from the retinal and choroidal circulations at normal and increased arterial oxygen
tensions, Invest. Ophthalmol. 8:588–594.
Estes, R., and Pregill, C. L., 1988, Phylogenetic Relationships of the Lizard Families,
Stanford University Press, Stanford.
Feduccia, A., 1996, The Origin and Evolution of Birds, Yale University Press, New Haven.
Fite, K., and Lister, B. C., 1981, Bifoveate vision in anolis lizards, Brain Behav. Evol.
19:144–154.
Fite, K., and Rosenfield-Wessels, S., 1975, A comparative study of deep avian foveas,
Brain Behav. Evol. 12:97–115.
Franco, E. C. S., Finlay, B. L., Silveira, L. C. L., Yamada, E. S., and Crowley, J. C.,
2000, Conservation of absolute foveal area in New World monkeys, Brain Behav.
Evol. 56:276–286.
Franz, V., 1934, Vergleichende Anatomie des Wirbeltierauges, in: handbuch der
vergleichenden Anatomie der Wirbeltiere 2, L., Bolk, E., Göppert, E., Kallius, and
W., Lubosch, eds., Berlin & Wien, Berlin, pp. 989–1292.
Frederiksen, R. D., 1976, Retinal tapetum containing discrete reflectors and photoreceptors in the bathypelagic teleost Omosudis lowei, Vidensk. Meddr dansk naturh.
Foren. 139:109–146.
Gartner, J. V., Crabtree, R. E., and Sulak, K. J., 1997, Feeding at Depth, in: Deep-Sea
Fishes, R. J., Randall, and A. P., Farrell, eds., Academic Press, San Diego,
pp. 115–193.
Gorrand, J.-M., 1979, Diffusion of the human retina and quality of the optics of the
eye on the fovea and the peripheral retina, Vision Res. 19:907–912.
Gorrand, J.-M., 1989, Reflection characteristics of the human fovea assessed by
reflecto-modulometry, Ophthal. Physiol. Opt. 9:53–60.
Graham, J. B., and Rosenblatt, R. H., 1970, Aerial vision: unique adaptation in an
intertidal fish, Science 168:586–588.
Harkness, L., and Bennet-Clark, H. C., 1978, The deep fovea as a focus indicator,
Nature 272:814–816.
Heesy, C. P., and Ross, C. F., 2001, Evolution of activity patterns and chromatic vision
in primates: Morphometrics, genetics and cladistics, J. Hum. Evol. 40:111–149.
Heesy, C. P., and Ross, C. F., 2002, Mosaic Evolution of Activity Pattern, Diet, and
Color Vision in Haplorhine Primates, in: Anthropoid Origins: New Visions, C. F.,
Ross, and R. F., Kay, eds. Kluwer Academic/Plenum Publishers, New York,
this volume.
Henderson, R. W., and Binder, M. H., 1980, The ecology and behavior of vine snakes
(Ahaetulla, Oxybelis, Thelotornis, Uromacer): A review, Contrib. Biol. Geol. 37:1–38.
Anthro-19.qxd 12/13/03 7:19 PM Page 532
532
Callum F. Ross
Hendrickson, A. E., 1992, A morphological comparison of foveal development in man
and monkey, Eye 6:136–144.
Hendrickson, A. E., Djajadi, H. R., Nakamura, L., Possin, D. E., and Sajuthi, D.,
2000, Nocturnal tarsier retina has both short and long/medium-wavelength cones
in an unusual topography, J. Comp. Neurol. 424:718–730.
Hendrickson, A. E., and Yuodelis, C., 1984, The morphological development of the
human fovea, Ophthalmology 91:603–612.
Herring, P. J., 1987, Systematic distribution of bioluminescence in living organisms.
J. Biolumin. Chemolumin. 1:147–163.
Hughes, A., 1977, The Topography of Vision in Mammals of Contrasting Life Style:
Comparative Optics and Retinal Organization, in: The Visual System in Vertebrates,
F., Crescitell, ed., Springer-Verlag, Berlin, pp. 613–765.
Johnson, G. D., and Patterson, C., 1993, Percomorph phylogeny: A survey and a new
proposal, Bull. Mar. Sci. 52:554–626.
Johnson, G. D., and Patterson, C., 1996, Relationships of Lower Teleostean Fishes,
in: Interrelationships of Fishes, M. L. J., Stiassny, L. R., Parenti, and G. D., Johnson,
eds., Academic Press Inc, San Diego, pp. 251–332.
Kahmann, H., 1936, Über das foveale Sehen der Wirbeltiere, Albrecht v. Graefes Arch.
Ophthal. 135:265–276.
Kajikawa, J., 1923, Beiträge zur Anatomie und Physiologie des Vogelauges, Albrecht
Von Graefes Arch. Ophthalmol. 112:260–346.
Kay, R. F., and Cartmill, M., 1977, Cranial morphology and adaptations of
Palaechthon nacimienti and other paromomyidae (Plesiadapoidea, ?Primates), with
description of a new genus and species, J. Hum. Evol. 6:19–53.
Kay, R. F., and Kirk, E. C., 2000, Osteological evidence for the evolution of activity
pattern and visual acuity, Am. J. Phys. Anthropol. 113:235–262.
Kay, R. F., Ross, C. F., and Williams, B. A., 1997, Anthropoid origins, Science
275:797–804.
Krause, W., 1889, Die retina II.Die Retina der Fische, Int. Mschr. Anat. Physiol.
6:206–223 & 250–269.
Le Gros Clark, W. E., 1959, The Antecedents of Man, Harper, New York.
Lee M. S. Y., Doughty P., 1997, The relationship between evolutionary theory and
phylogenetic analysis, Biol. Rev. 72:471–495.
Locket, N. A., 1971, Retinal structure in Platytroctes apus, a deep-sea fish with a pure
rod fovea, J. Mar. Biol. Ass. U.K. 51:79–91.
Locket, N. A., 1975, Some Problems of Deep-Sea Fish Eyes, in: Vision in Fishes.
New Approaches in Research, M. A., Ali, ed., Plenum Press, New York, pp. 645–655.
Locket, N. A., 1977, Adaptations to the Deep-Sea Environment, vol. 7/5. in: The
Visual System in Vertebrates. Handbook of Sensory Physiology, F., Crescitelli, ed.,
Springer-Verlag, Berlin, pp. 67–192.
Anthro-19.qxd 12/13/03 7:19 PM Page 533
Tarsier Fovea
533
Locket, N. A., 1985, The multiple bank fovea of Bajacalifornia drakei, an
alepocephalid deep-sea teleost, Proc. R. Soc. Lond., B, 224:7–22.
Locket, N. A., 1992, Problems of deep foveas, Aust. NZ. J. Ophthalmol. 20:281–295.
Maddison, W. P., 1990, A method for testing the correlated evolution of two binary
characters: Are gains and losses concentrated on certain branches of a phylogenetic
tree? Evolution 44:539–557.
Makaretz, M., and Levine, R. L., 1980, A light mircoscopic study of the bifoveate
retina in the lizard Anolis carolensis: General observations and convergence ratios,
Vision Res. 20:679–686.
Marshall, N. B., 1966, Family Scopelosauridae, Sears Found. Mar. Res., Mem.
1,5:194–204.
Martin, G. R., 1985, Eye, in: Form and Function in Birds, A. S., King, and
J., McLelland, eds., Academic Press, New York, pp. 311–373.
Martin, P. R., and Grünert, U., 1999, Analysis of the short wavelength-sensitive
(“blue”) cone mosaic in the primate retina: Comparison of New World and Old
World monkeys, J. Comp. Neurol. 406.
Martin, R. D., 1973, Comparative anatomy and primate systematics, Symp. Zool. Soc.
Lond. 33:301–337.
Martin, R. D., 1979, Phylogenetic Aspects of Prosimian Behavior, in: The Study of
Prosimian Behavior, G. A., Doyle, and R. D., Martin, eds., Academic, New York,
pp. 45–78.
Martin, R. D., 1990, Primate Origins and Evolution. A Phylogenetic Reconstruction,
Princeton University Press, Princeton.
Meneghini, K. A., and Hamasaki, D. I., 1967, The electroretinogram of the iguana
and Tokay gecko, Vision Res. 7:243–251.
Michaelson, I. C., 1954, Retinal Circulation in Man and Animals, Charles C.
Thomas, Springfield.
Moroney, M. K., and Pettigrew, J. D., 1987, Some observations on the visual optics
of kingfishers (Aves, Coraciformes, Alcedinidae), J. Comp. Physiol. 160:137–149.
Munk, O., 1959, The eye of Ipnops murrayi Günther, 1878, Galathea Rep. 3:79.
Munk, O., 1965, Omosudis lowei Günther, 1887, A bathypelagic deep-sea fish with an
almost pure-cone retina, Vidensk. Medd. Fra Dansk naturh. Foren. 128:341–355.
Munk, O., 1966a, On the retina of Diretma argenteus Johnson, 1863 (Diretmidae,
Pisces), Vidensk. Medd. Dan. Haturist. Foren. KBH. 129:73–80.
Munk, O., 1966b, Ocular anatomy of some deep-sea teleosts, Dana-Report No. 70:
1–71.
Munk, O., 1968, On the eye and the so-called preorbital light organ of the isospondylous
deep-sea fish, Bathylaco nigricans Goode and Bean, 1896, Galathea Rep. 9:211–218.
Munk, O., 1969, The eye of the “four-eyed” fish Dialommus fuscus (Pisces,
Blenioidea, Clinidae). Vidensk. Meddr. dansk naturh. Foren. 132:7–24.
Anthro-19.qxd 12/13/03 7:19 PM Page 534
534
Callum F. Ross
Munk, O., 1970, On the occurrence and significance of horizontal band-shaped retinal
areae in teleosts, Vidensk. Meddr. dansk naturh. Foren. 133:85–120.
Munk, O., 1971, On the occurrence of two lens muscles within each eye of some
teleosts, Vidensk. Meddr. dansk naturh. Foren. 134:7–19.
Munk, O., 1975, On the eyes of two foveate notosudid teleosts Scopelosaurus hoedti
and Ahliesaurus berryi, Vidensk. Meddr. dansk naturh. Foren. 138:87–125.
Munk, O., 1977, The visual cells and retinal tapetum of the foveate deep-sea fish
Scopelosaurus lepidus (Teleostei), Zoomorph. 87:21–49.
Nicol, J. A. C., 1989, The Eyes of Fishes, Clarendon Press, Oxford.
Niemitz, C., 1985, Can a primate be an owl?—Convergences in the same ecological
niche, Fort. Ddr Zool. 30:667–670.
O’Day, K. J., 1939, The visual cells of Australian reptiles and mammals, Trans.
Ophthalmol. Soc. Aust. 1:12–20.
Oehme, H., 1961, Vergleichend-histologische Untersuchungen an der Retina von
Eulen, Zool. Jb. Anat. 79:439–478.
Ogden, T. E., 1974, The morphology of retinal neurones of the owl monkey Aotes,
J. Comp. Neurol. 153:399–428.
Ogden, T. E., 1975, The receptor mosaic of Aotes trivirgatus: Distribution of rods and
cones, J. Comp. Neurol. 163:193–202.
Ohzu, H., and Enoch, J. M., 1972, Optical modulation by the isolated human fovea,
Vision Res. 12:245–251.
Ohzu, H., Enoch, J. M., and O’Hair, J. C., 1972, Optical modulation by the isolated
retina and retinal receptors, Vision Res. 12:231–244.
Packer, O., Hendrickson, A. E., and Curcio, C. A., 1989, Photoreceptor topography
of the adult pig-tailed macaque, J. Comp. Neurol. 288:165–183.
Pariente, G., 1979, The Role of Vision in Prosimian Behavior, in: The Study of
Prosimian Behavior, G. A., Doyle, and R. D., Martin, eds., Academic Press Inc.,
New York, pp. 411–459.
Perry, V. H., and Cowey, A., 1985, The ganglion cell and cone distributions in the monkey’s retina: Implications for central magnification factors, Vision Res. 25:1795–1810.
Peterson, E., 1981, Regional specialization in retinal ganglion cell projection to optic
tectum of Dipsosaurus dorsalis (Iguanidae), J. Comp. Neurol. 196:225–252.
Peterson, E., 1992, Retinal Structure, in: Biology of the Reptilia, Volume 17, Neurology
C. Sensorimotor Integration, C. Gans, and P. S. Ulinski, eds., University of Chicago
Press, Chicago, pp. 1–135.
Pettigrew, J. D., 1983, A note on the eyes of the Letter-Winged Kite Elanus scriptus,
The Emu 82 Supplement: 305–308.
Pettigrew, J. D., and Konishi, M., 1984, Some observations on the visual system of the
oilbird, Steatornis caripensis, National Geographic Society Research Reports.
16:439–450.
Anthro-19.qxd 12/13/03 7:19 PM Page 535
Tarsier Fovea
535
Pettigrew, J. D., Wallman, J., and Wildsoet, C. F., 1990, Saccadic oscillations facilitate
ocular perfusion from the avian pectin, Nature 343:362–363.
Polyak, S., 1941, The Retina, University of Chicago Press, Chicago.
Provis, J. M., 2001, Development of the primate retinal vasculature, Prog. Retin. Eye
Res. 20:799–821.
Provis, J. M., Diaz, C. M., and Dreher, B., 1998, Ontogeny of the primate fovea:
A central issue in retinal development, Prog. Neurobiol. 54:549–581.
Pumphrey, R. J., 1948, The theory of the fovea, J. Exp Biol. 25:299–312.
Reymond, L., 1985, Spatial visual acuity of the eagle Aquila audax: A behavioural,
optical and anatomical investigation, Vision Res. 27:1859–1874.
Reymond, L., 1987, Spatial visual acuity of the falcon Falco berigora: A behavioural,
optical and anatomical investigation, Vision Res. 25:1477–1491.
Rochon-Duvigneaud, A., 1943, Les Yeux et La Vision des Vertébrés, Masson, Paris.
Rodieck, R. W., 1988, The Primate Retina, in: Comparative Primate Biology, Volume
4: Neurosciences. H. D. Stecklis, and J. Erwin, eds., Alan R. Liss Inc., New York,
pp. 203–278.
Rohen, J. W., 1966, Zur Histologie des Tarsiusauges, Albrecht v. Graefes Arch. klin.
exp. Ophthal. 169:299–317.
Rohen, J. W., and Castenholtz, A., 1967, Über die Zentralisation der Retina bei
Primaten, Folia primat. 5:92–147.
Röll, B., 2001, Gecko vision—retinal organization, foveae and implications for binocular vision, Vision Res. 41:2043–2056.
Ross, C. F., 1996, An adaptive explanation for the origin of the Anthropoidea
(Primates), Am. J. Primatol. 40:205–230.
Ross, C. F., Lockwood, C., Fleagle, J. G., and Jungers, W. L., 2001, Adaptation and
Behavior in the Primate Fossil Record, in: Reconstructing Behavior in the Primate
Fossil Record, J. M. Plavcan, R. F. Kay, W. L. Jungers, and C. P. van Schaik, eds.,
Kluwer Academic/Plenum Publishers, New York, pp. 1–41.
Ross, C. F., Williams, B. A., and Kay, R. F., 1998, Phylogenetic analysis of anthropoid
relationships, J. Hum. Evol. 35:221–306.
Schluter, D., Price, T., Mooers, A. O., and Ludwig, D., 1997, Likelihood of ancestor
states in adaptive radiation, Evolution 51:1699–1711.
Schmitt, E., and Kunz, Y. W., 1989, Retinal morphogenesis in the rainbow trout,
Salmo gairdneri, Brain Behav. Evol. 34:48–64.
Schwassman, H. O., 1968, Visual projection upon the optic tectum in foveate marine
teleosts, Vision Res. 8:1337–1348.
Slonaker, J. R., 1897, A comparative study of the area of acute vision in vertebrates,
J. Morph. 13:445–492.
Snodderly, D. M., Leung, W. P., Timberlake, G. T., and Smith, D. P. B., 1978,
Mapping Retinal Features in a Freely Moving Eye with Precise Control of Retinal
Anthro-19.qxd 12/13/03 7:19 PM Page 536
536
Callum F. Ross
Simulus Position, in: Frontiers in Visual Science, S. J., Cool, and E. L., Smith, eds.,
Springer, New York, pp. 79–92.
Snodderly, D. M., Weinhaus, R. S., and Choi, J. C., 1992, Neural-vasculature relationships in central retina of macaque monkeys (Macaca fascicularis), J. Neurosci.
12:1169–1193.
Steenstrup, S., and Munk, O., 1980, Optical function of the convexiclivate fovea with
particular regard to notosudid deep-sea teleosts, Optica acta 27:949–964.
Stiassny, M. L. J., Parenti, L. R., and Johnson, G. D., 1996, Interrelationships of Fishes,
Academic Press Inc, San Diego.
Stone, J., and Johnston, E., 1981, The topography of primate retina: A study of
the human, bushbaby and New- and Old-World monkeys, J. Comp. Neurol.
196:205–223.
Szalay, F. S., and Delson, E., 1979, Evolutionary History of the Primates, Academic
Press, New York.
Tamura, T., 1957, A study of visual perception in fish, especially on resolving power
and accommodation, Bull. Jap. Soc. Scient. Fish. 22:536–557.
Tansley, K., 1961, The retina of a diurnal gecko, Pflugers Arch. Ges. Physiol.
272:262–269.
Tansley, K., 1964, The gecko retina, Vision Res. 4:33–37.
Tetreault, N., Hakeem, A., and Allman, J., this volume, The Distribution and Size of
Retinal Ganglion Cells in Cheirogaleus medius and Tarsius syrichta: Implications for
the Evolution of Sensory Systems in Primates, in: Anthropoid Origins: New Visions,
C. F. Ross and R. F. Kay, eds., New York: Kluwer Academic/Plenum Publishers.
Underwood, G., 1970, The Eye, in: Biology of the Reptilia. Volume 2, Morphology B,
C. Gans, and T. S. Parsons, eds., Academic Press, London.
Valentin, G., 1897, Ein Beitrag zur Kenntniss der Brechungsverhältnisse der
Thiergewebe, Pflügers Arch. ges. Physiol. 19:78–105.
Verrier, M.-L., 1928, Sur la Présence et la structure d’une fovea rétinienne chez un
percidé: Serranus cabrilla L., C. R. hebd. Séanc. Acad. Sci. (Paris) 186:457–459.
Verrier, M.-L., 1933, Recherches sur la vision des reptiles, Bull. Biol. France Belg.
67:350–370.
Vilter, V., 1951, Recherches sur les structures foveales dans la retine du Sphenodon
punctatus, Comptes Rendu des Séances de la Société Biologie. 145:26–29.
Vilter, V., 1954a, Différenciation fovéale dans l’appareil visuel d’un poisson abyssal, le
Bathylagus benedicti, Comptes Rendu des Séances de la Société Biologie. 148:59–63.
Vilter, V., 1954b, Relations neuronales dans la fovea à bâtonnets du Bathylagus benedicti, Comptes Rendu des Séances de la Société Biologie. 148:466–469.
Vitt, L. J., and Pianka, E. R., 1994, Lizard Ecology : Historical and Experimental
Perspectives, Princeton University Press, Princeton.
Anthro-19.qxd 12/13/03 7:19 PM Page 537
Tarsier Fovea
537
Vrabec, F., 1969, Sur la présenced’une fovea dans la rétine du Lepadogaster, Vie Milieu,
Ser. A 20:245–250.
Wagner, H.-J., 2001, Brain areas in abyssal demersal fishes, Brain Behav. Evol.
57:301–316.
Walls, G. L., 1937, Significance of the foveal depression, Arch. Ophthal. 18:912–919.
Walls, G. L., 1940, Postscript on image expansion at the fovea, Arch. Ophthal. 22:
831–832.
Walls, G. L., 1942, The Vertebrate Eye and Its Adaptive Radiation, Hafner, New York.
Wässle, H., and Boycott, B. B., 1991, Functional architecture of the mammalian
retina, Physiol. Rev. 71:447–480.
Wässle, H., Grünert, U., Röhrenbeck, J., and Boycott, B. B., 1990, Retinal ganglion
cell density and cortical magnification factor in the primate, Vision Res.
30:1897–1911.
Weale, R. A., 1966, Why does the human retina possess a fovea? Nature 212:255–256.
Webb, S. V., and Kaas, J. H., 1976, The sizes and distribution of ganglion cells in the
retina of the owl monkey, Aotus trivirgatus, Vision Res. 16:1247–1254.
Weitzman, S. H., 1997, Systematics of Deep-Sea Fishes, in: Deep-Sea Fishes, D. J.
Randall, and A. P. Farrell, eds., Academic Press, San Diego, pp. 43–77.
Whitehead, P. J. P., Bauchot, M.-L., Hureau, J.-C., Nielsen, J., and Tortonese, E.,
1986, Fishes of the North-eastern Atlantic and the Mediterranean. Vol. I-III Unesco,
United Kingdom.
Wikler, K. C., Williams, R. W., and Rakic, P., 1990, Photoreceptor mosaic: Number
and distribution of rods and cones in the rhesus monkey retina, J. Comp. Neurol.
297:499–508.
Williams, P. L., and Warwick, R., 1980, Gray’s Anatomy, 36th Edition, Churchill
Livingstone, New York.
Wolff, E., 1940, The Anatomy of the Eye and Its Orbit, Lewis & Co., London.
Wolin, L. R. and Massopust, L. C., 1970, Morphology of the Primate Retina, in:
Advances in Primatology. Volume 1. The Primate Brain, C. R. Noback, and
W. Montagna, eds., Appleton-Century-Crofts, New York, pp. 1–27.
Wood, C. A., 1917, The Fundus Oculi of Birds Especially as Viewed by the
Ophthalmoscope, Chicago, The Lakeside Press.
Woollard, H. H., 1925, The anatomy of Tarsius spectrum, Proc. R. Soc. Lond.
70:1071–1084.
Wulliman, M. F., 1997, Major patterns of visual brain organization in teletosts and
their relation to prehistoric events and the paleontological record, Paleobiology
23:101–114.
Anthro-19.qxd 12/13/03 7:19 PM Page 538
Callum F Ross