190
Eric J. Warrant
to how well resolved a compound eye is. Smaller values of L1p indicate better
resolution (Table 1 ). Very often (but not always), eyes which strive for sensitivity
sacrifice resolution and have wider acceptance angles (Warrant and Mcintyre 1992).
In apposition eyes (Fig. 1A) each ommatidium is isolated from its neighbours by
a sleeve of light-absorbing screening pigment, thus preventing light reaching the
rhabdom from all but its own small corneal lens. It is this small lens aperture (only
15 Jlm wide in some species) which severely limits the amount of light that can
reach each rhabdom. Apposition eyes are therefore typical of bright habitats. This
limitation is slightly offset in higher flies which possess a special variation of the
apposition design known as neural superposition (Kirschfeld 1967). In this design
rhabdomeres in seven neighbouring ommatidia view the same direction in space,
and send their axons to the same cartridge in the first optic neuropil, the lamina. This
improves sensitivity and contrast discrimination since each visual channel receives
light through seven lenses.
In superposition eyes (Fig. 1B), the pigment sleeve is withdrawn and a wide
optically transparent area (the clear zone) is interposed between the lenses and the
retina. This clear zone - and specially modified crystalline cones - allow light from a
narrow region of space to be collected by a large number of ommatidia (comprising
the superposition aperture) for focusing onto a single rhabdom. Unlike the crystalline
cones of apposition eyes, those of some superposition eyes have evolved refractive
index gradients (Exner 1891), those of others reflective surfaces (Vogt 1975, Land
197 6), and in still others a combination of both (Nilsson 1988). The optics of such
cones allows as many as 2000 ommatidia to collect light for a single rhabdom (as in
some large nocturnal moths: Warrant, unpubl. data). Not surprisingly, superposition
eyes are typical of animals active in dim habitats.
2.2 Design Principles, Constraints and Trade-Offs
2.1.1 Acute Zones and Matched Filtering
Despite their incredible adaptability, compound eyes are rarely more than a few
millimetres wide, and this is a serious drawback. Light reaches the rhabdoms through
tiny corneal lenses that are frequently no more than a few tens of micrometres in
diameter (A in Fig. 1A). This is not only a problem for photon collection in dim light.
Even in bright light such small lenses cause diffraction, and thus blurring, of the
incoming light rays, thereby limiting spatial resolution. A solution to both problems
is to make the lenses larger: diffraction is reduced and light capture improves.
Unfortunately, for the same eye radius R, larger lenses mean a wider interommatidial
angle: L11j>= AIR radians. This, in tum, means poorer anatomical resolution. A simple
way out of this dilemma is to scale up the size of the eye (i.e. makeR larger), and
allow larger lenses without sacrificing L11j>: but there are limits to this, too. Animals
with compound eyes are generally rather small and there is a limit to how big an eye
they can carry. Apart from anything else, the metabolic cost of having a larger eye is
quite enormous and sets a serious limit to how big a particular eye can be (see
Eric J. Warrant
to how well resolved a compound eye is. Smaller values of L1p indicate better
resolution (Table 1 ). Very often (but not always), eyes which strive for sensitivity
sacrifice resolution and have wider acceptance angles (Warrant and Mcintyre 1992).
In apposition eyes (Fig. 1A) each ommatidium is isolated from its neighbours by
a sleeve of light-absorbing screening pigment, thus preventing light reaching the
rhabdom from all but its own small corneal lens. It is this small lens aperture (only
15 Jlm wide in some species) which severely limits the amount of light that can
reach each rhabdom. Apposition eyes are therefore typical of bright habitats. This
limitation is slightly offset in higher flies which possess a special variation of the
apposition design known as neural superposition (Kirschfeld 1967). In this design
rhabdomeres in seven neighbouring ommatidia view the same direction in space,
and send their axons to the same cartridge in the first optic neuropil, the lamina. This
improves sensitivity and contrast discrimination since each visual channel receives
light through seven lenses.
In superposition eyes (Fig. 1B), the pigment sleeve is withdrawn and a wide
optically transparent area (the clear zone) is interposed between the lenses and the
retina. This clear zone - and specially modified crystalline cones - allow light from a
narrow region of space to be collected by a large number of ommatidia (comprising
the superposition aperture) for focusing onto a single rhabdom. Unlike the crystalline
cones of apposition eyes, those of some superposition eyes have evolved refractive
index gradients (Exner 1891), those of others reflective surfaces (Vogt 1975, Land
197 6), and in still others a combination of both (Nilsson 1988). The optics of such
cones allows as many as 2000 ommatidia to collect light for a single rhabdom (as in
some large nocturnal moths: Warrant, unpubl. data). Not surprisingly, superposition
eyes are typical of animals active in dim habitats.
2.2 Design Principles, Constraints and Trade-Offs
2.1.1 Acute Zones and Matched Filtering
Despite their incredible adaptability, compound eyes are rarely more than a few
millimetres wide, and this is a serious drawback. Light reaches the rhabdoms through
tiny corneal lenses that are frequently no more than a few tens of micrometres in
diameter (A in Fig. 1A). This is not only a problem for photon collection in dim light.
Even in bright light such small lenses cause diffraction, and thus blurring, of the
incoming light rays, thereby limiting spatial resolution. A solution to both problems
is to make the lenses larger: diffraction is reduced and light capture improves.
Unfortunately, for the same eye radius R, larger lenses mean a wider interommatidial
angle: L11j>= AIR radians. This, in tum, means poorer anatomical resolution. A simple
way out of this dilemma is to scale up the size of the eye (i.e. makeR larger), and
allow larger lenses without sacrificing L11j>: but there are limits to this, too. Animals
with compound eyes are generally rather small and there is a limit to how big an eye
they can carry. Apart from anything else, the metabolic cost of having a larger eye is
quite enormous and sets a serious limit to how big a particular eye can be (see