Minimal Model of Prey Localization through the Lateral-Line System


Jan-Moritz P. Franosch, Marion C. Sobotka, Andreas Elepfandt, 
and J. Leo van Hemmen

Physik Department der TU München & Institut für Biologie, Humboldt Universität


The clawed frog Xenopus is a fully aquatic predator catching
prey at night by detecting water movements originating from its
prey. To this end Xenopus uses its lateral-line system, a
mechanoreceptive system found in aquatic amphibians and fish and
analyzing water movements along the animal's body. It comprises,
dependent on the species, hundreds to several thousand small
lateral-line organs dispersed over the trunk. Xenopus laevis
laevis has about 180 of them, positioned along several lines at
both sides of the body, around the eyes, and at a few other locations
of head and neck. A lateral-line organ of Xenopus contains 4-8
small cupulae, gelatinous flags protruding into the water and
deflected by local water flow. The deflection stimulates sensory hair
cells at the base of the cupulae and in this way generates spikes in
the lateral-line nerves, phase-locked to the water
velocity. 

Xenopus' eyes are not adapted to seeing in water and the
animal's lateral-line system has become the central sensory organ for
spatial orientation. When an insect drops onto the water surface it
generates a wave passing along Xenopus and stimulating the
lateral-line organs. As its natural behavior, the frog will then turn
towards the wave's origin, its prey. Not only does the lateral-line
system allow Xenopus to determine the direction of a single
impinging wave but also to resolve the directions of two
simultaneous waves of different frequency that overlap at the
animal, and even to discern wave sources as "food" and
"non-food". Because each lateral-line organ responds to waves from
any direction, localization requires a comparison of
inputs from several lateral-line organs. Since, however, each
lateral-line organ only encodes the local superposition pattern of the
waves at the body surface, the pattern-segmentation ability of
Xenopus requires both a comparison of the encoded superposition
patterns stemming from several organs and a decomposition of the
patterns into their original components.

We present a general method, a `minimal model' based on a
minimum-variance estimator, to explain prey detection through the
frog's many lateral-line organs. The model is robust in the sense
that, even if several lateral-line organs are defunct, it still
matches experimental data. We show how waveform reconstruction
allows Xenopus to determine both direction and character of the
prey and even to distinguish two simultaneous wave sources. A simple
neuronal algorithm with realistic firing rates, number of synapses,
and time constants of postsynaptic potentials suffices to perform
localization and pattern segmentation through waveform reconstruction.

As the model is universal, we expect its results to be applicable to
many aquatic amphibians, fish, and reptiles with a large number of
local stimulus detectors, such as crocodilians, which have recently
been found to take advantage of about 2000 dome pressure receptors at
the surface of their face.