180
Simon B. Laughlin
rate of transmission of information is increasing as the log of SNR. Thus when
parallel channels (in this case synapses) are used to increase the bit rate, the cost
per bit also increases.
The array of photoreceptor-LMC synapses typifies any neural system that has to
improve its information rate by increasing the number of unreliable components
(e.g. transmitter molecules, ion channels, synapses) used to transmit the signal
(Laughlin et al. 1998). As commonly observed in communications engineering,
increasing the rate at which information is transmitted also increases the cost per
bit. Here is another law of diminishing returns that heightens the selective
pressure to reduce the unit cost by reducing performance to the lowest acceptable
level. However, the metabolic costs of capturing and coding information will only
influence the form and function of sense organs and neural codes when these
costs make a significant contribution to an animal's total energy budget. Let us
examine this contribution.
7 The Contribution of Retinal Coding to Metabolic
Rate
Multiplying the rate of A TP hydrolysis per photoreceptor by the total number of
photoreceptors in the blowfly eye gives the total photoreceptor consumption.
Allowing for mitochondrial proton leak (Rolfe and Brown 1997), approximately
five A TP molecules are generated for each oxygen molecule consumed. The A TP
consumed by photoreceptors in daylight converts to 11% of the total oxygen used
by a resting blowfly. Adding the photoreceptor-LMC synapses, 13% of resting
energy consumption is devoted to electrical signalling in the two compound eyes.
This large retinal electricity bill is a good reason to maximize the information
gained from electrical signals by optimizing all components of coding and
processing. The high level of consumption by photoreceptors has also led to
energy saving adaptations. The study of these adaptations takes visual ecology
into the time domain.
8 Dynamics, Cost and a Visual Ecology in Time
Insect compound eyes exhibit differences in response dynamics that economize
photoreceptor energy consumption. In the majority of retinas, invertebrate and
vertebrate, phototransduction accelerates with lightadaptation. In blowfly this
acceleration is carefully adjusted to optimize performance, by converting the
response from slow to brisk (Fig. 2). However, looking across the Diptera, light
adaptation accelerates the responses of some species more than others. The
photoreceptors of tipulids (craneflies) scarcely speed up at all (Laughlin and
Weckstrom 1993). Response dynamics are matched to visual ecology because, as
Simon B. Laughlin
rate of transmission of information is increasing as the log of SNR. Thus when
parallel channels (in this case synapses) are used to increase the bit rate, the cost
per bit also increases.
The array of photoreceptor-LMC synapses typifies any neural system that has to
improve its information rate by increasing the number of unreliable components
(e.g. transmitter molecules, ion channels, synapses) used to transmit the signal
(Laughlin et al. 1998). As commonly observed in communications engineering,
increasing the rate at which information is transmitted also increases the cost per
bit. Here is another law of diminishing returns that heightens the selective
pressure to reduce the unit cost by reducing performance to the lowest acceptable
level. However, the metabolic costs of capturing and coding information will only
influence the form and function of sense organs and neural codes when these
costs make a significant contribution to an animal's total energy budget. Let us
examine this contribution.
7 The Contribution of Retinal Coding to Metabolic
Rate
Multiplying the rate of A TP hydrolysis per photoreceptor by the total number of
photoreceptors in the blowfly eye gives the total photoreceptor consumption.
Allowing for mitochondrial proton leak (Rolfe and Brown 1997), approximately
five A TP molecules are generated for each oxygen molecule consumed. The A TP
consumed by photoreceptors in daylight converts to 11% of the total oxygen used
by a resting blowfly. Adding the photoreceptor-LMC synapses, 13% of resting
energy consumption is devoted to electrical signalling in the two compound eyes.
This large retinal electricity bill is a good reason to maximize the information
gained from electrical signals by optimizing all components of coding and
processing. The high level of consumption by photoreceptors has also led to
energy saving adaptations. The study of these adaptations takes visual ecology
into the time domain.
8 Dynamics, Cost and a Visual Ecology in Time
Insect compound eyes exhibit differences in response dynamics that economize
photoreceptor energy consumption. In the majority of retinas, invertebrate and
vertebrate, phototransduction accelerates with lightadaptation. In blowfly this
acceleration is carefully adjusted to optimize performance, by converting the
response from slow to brisk (Fig. 2). However, looking across the Diptera, light
adaptation accelerates the responses of some species more than others. The
photoreceptors of tipulids (craneflies) scarcely speed up at all (Laughlin and
Weckstrom 1993). Response dynamics are matched to visual ecology because, as