In many taxa, receivers use signals to detect and discriminate among mates. Signal detection and discrimination thus has important fitness consequences for individuals. Noise is defined as any factor that prevents detection or discrimination of signals. The noise produced by groups of signaling animals is a well-known impediment to signal detection and discrimination in animals, but how many signals produce the emergent, masking effects of noise? This dataset was generated to explore how receivers discriminate among signals in noisy, multi-choice environments. Subjects were female Australian field crickets, Teleogryllus oceanicus. We performed a series of phonotaxis (movement toward sound) assays in which we manipulated the number of long chirps in the signal. First, we assessed female preferences for the number of long chirp pulses and found that receivers preferred more long chirp pulses to fewer. Then we gave receivers a choice between a preferred, 7-pulse signal and either 1, 3, 5, or 7 presentations of the non-preferred, 2-pulse signal ("the multi-choice experiment"). We observed the probability that subjects left the release point, the probability that subjects responded to playback, and the probability of choosing the preferred stimulus. We also recorded the subject's latency to leave the release point and the latency to respond to playback. Because the angular separation between speakers decreased with increasing number of playback speakers in the multi-choice experiment, we then conducted an experiment ("the angular separation experiment") to determine whether observed effects were due to the spatial configuration of speakers or due to the emergent noise of multiple playback speakers. 

This dataset was produced using phonotaxis experiments of adult, female Australian field crickets (Teleogryllus oceanicus) aged 10-13 days post eclosion. Data were collected under red light in an anechoic room at a controlled temperature of 26 degrees during the scotoperiod.

The testing arena was a 2.5 m x 2.5 m square constructed from interlocking foam mats. Speakers were arranged on the floor on the perimeter of a circle measuring 1 m in radius (but not demarcated by a physical barrier within the arena).

Subjects were given phonotaxis trials in randomized order and given at least 10 minutes and up to an hour of "downtime" between trials. We tested individuals in blocks that corresponded to a single day of behavioral testing, and for each block, randomized the trial order and which speaker played the preferred signal.

Playback of all stimuli for a given trial began simultaneously and at haphazard times during the signal's period to prevent any systematic relationship between the preferred and non-preferred signals' spatial relationship and relative timing.

Full details are provided in the associated manuscript.

The dataset consists of three spreadsheets, each of which corresponds to one of the experiments described in the accompanying paper as the "preliminary experiment," the "multi-choice experiment," and the "angular separation experiment." See the ReadMe file and the associated paper in Behavioral Ecology for complete details.