A small, computationally flexible network produces the phenotypic diversity of song recognition in crickets.

Jan Clemens,Berthold Hedwig,Stefan Schöneich,R Matthias Hennig,Konstantinos Kostarakos

doi:10.7554/elife.61475

Abstract

How neural networks evolved to generate the diversity of species-specific communication signals is unknown. For receivers of the signals, one hypothesis is that novel recognition phenotypes arise from parameter variation in computationally flexible feature detection networks. We test this hypothesis in crickets, where males generate and females recognize the mating songs with a species-specific pulse pattern, by investigating whether the song recognition network in the cricket brain has the computational flexibility to recognize different temporal features. Using electrophysiological recordings from the network that recognizes crucial properties of the pulse pattern on the short timescale in the cricket Gryllus bimaculatus, we built a computational model that reproduces the neuronal and behavioral tuning of that species. An analysis of the model's parameter space reveals that the network can provide all recognition phenotypes for pulse duration and pause known in crickets and even other insects. Phenotypic diversity in the model is consistent with known preference types in crickets and other insects, and arises from computations that likely evolved to increase energy efficiency and robustness of pattern recognition. The model's parameter to phenotype mapping is degenerate - different network parameters can create similar changes in the phenotype - which likely supports evolutionary plasticity. Our study suggests that computationally flexible networks underlie the diverse pattern recognition phenotypes, and we reveal network properties that constrain and support behavioral diversity.

Highlights

Many behaviors are driven by the recognition and evaluation of sensory stimuli
We tested whether the delay-line and coincidence-detector network of the cricket G. bimaculatus (Figure 1D) can be modified to produce the known diversity of preference functions for pulse duration and pause in cricket calling songs (Figure 1B)
This network was previously inferred from the anatomical overlap together with the dynamics and the timing of responses of individually recorded neurons to a diverse set of pulse patterns (Kostarakos and Hedwig, 2012; Schöneich et al, 2015)

Summary

Introduction

Many behaviors are driven by the recognition and evaluation of sensory stimuli. For instance, hunting requires the detection and tracking of prey; communication requires the recognition of the sounds, pheromones or visual displays that serve as signals. Neural evolution is well understood when behaviors are driven by signals recognized through the response specificity of primary afferent neurons, where the change in a single amino acid can change the tuning of a specific behavior (Prieto-Godino et al., 2017; Ramdya and Benton, 2010). Many behaviors are driven by complex temporal and spatial signal patterns, whose recognition is based on the processing and comparison of neural activity across time and space, where changes in many parameters define the tuning of the system. For these behaviors, unravelling the underlying neural computation is challenging since it requires a mapping from circuit parameters to recognition phenotype

Objectives

Methods

Results

Conclusion