Data supporting Perceptually salient differences in a species recognition cue do not promote auditory streaming in eastern gray treefrogs (Hyla versicolor)

This dataset corresponds to behavioral choice tests performed on the females of Hyla versicolor to investigate if this species uses perceptually salient differences in species-recognition cue (pulse rise time) to segregate the signal of a potential mate from other overlapping sounds in the environment. Females in this species chose a potential mate based on the properties of the advertisement call. Advertisement calls having slow pulse rise times (time elapsed from the beginning of a pulse to the pulses's maximum amplitude) and slow pulse repetition rate (around 20 pulses/s) are behaviorally attractive to the females. In contrast, calls having fast rise times and fast pulse rates (~40-50 pulses/s) are unattractive. Here, we exploited the subjects' inherent attractiveness for slow rise times and slow pulse rates to design interleaved pulsatile sequences ABAB (repeating at an unattractive pulse rate of 40 pulses/s) having behaviorally attractive slow (pulses 'A'), and behaviorally unattractive fast (pulses 'B') rise-times. We hypothesized that if the rise-times differences between pulses 'A' and 'B' are perceptually salient then the subjects should segregate ABAB into two sequences (A-A- and B-B-, each at an attractive rate of 20 pulses/s each). We first tested (using a two-alternative choice test; Test C3) if the differences between the two rise times were perceptually salient (subjects got a choice between A-A- and B-B-). We then gave the subjects a segregation task (using a four-alternative choice test; Test T1) wherein they got a choice between four alternatives, three of which were designed to be unattractive (AAAA, BBBB and AABB). The fourth alternative ABAB was attractive only if the subjects could segregate A-A- and B-B-, so as to perceive the "attractive" slow pulse rise time at an "attractive" rate of 20 pulses/s. We did an additional test (using a four-alternative choice test; Test C1) to confirm that subjects prefer slow pulse rise times, slow pulse rates and regular pulse-timing patterns within the calls (a four-alternative choice between AAAA, BBBB, AA-- and A-A-). For each test, we recorded if the subject responded by making a choice ('yes' or 'no'), if it responded, which alternative did it chose, and how long did it take to make the choice (choice latency). We also recorded a subject's id, the temperature at which the behavioral test was performed, and the Sound Pressure Level (SPL) of the broadcast stimuli.

本数据集源自针对灰树蛙（Hyla versicolor）雌性个体开展的行为选择实验，旨在探究该物种是否借助物种识别线索（脉冲上升时间，pulse rise time）的感知显著性差异，从环境中重叠的背景噪声中分离出潜在配偶的鸣唱信号。该物种的雌性个体通过求偶鸣唱的声学特征选择潜在配偶：脉冲上升时间缓慢（即脉冲起始时刻至其峰值振幅的时长）且脉冲重复率较低（约20次/秒）的求偶鸣唱，对雌性个体具备行为吸引力。与之相反，脉冲上升速度快、脉冲频率偏高（约40~50次/秒）的鸣唱则无吸引力。
本研究利用受试个体对缓慢上升时间与低脉冲频率的固有偏好，设计了以无吸引力的40次/秒速率重复的交错脉冲序列ABAB：其中脉冲'A'为具备行为吸引力的缓慢上升时间脉冲，脉冲'B'为无吸引力的快速上升时间脉冲。本研究提出如下假设：若脉冲'A'与'B'的上升时间差异具备感知显著性，则受试个体应能将ABAB序列拆分为两个子序列（A-A-与B-B-，二者均以20次/秒的吸引力速率重复）。
首先，我们采用双选项选择实验（实验C3），检验两种上升时间的差异是否具备感知显著性，受试个体可在A-A-与B-B-之间做出选择。随后，我们为受试个体设置了分离任务，采用四选项选择实验（实验T1）：受试个体需在四个选项中做出选择，其中三个选项为无吸引力设计（AAAA、BBBB与AABB）。第四个选项ABAB仅当受试个体能够将序列拆分为A-A-与B-B-时才具备吸引力，此时个体可感知到以20次/秒的吸引力速率呈现的"吸引力型"缓慢脉冲上升时间。
此外，我们开展了补充实验，采用四选项选择实验（实验C1），以验证受试个体偏好鸣唱中缓慢的脉冲上升时间、低脉冲频率与规则的脉冲时序模式，即从AAAA、BBBB、AA--与A-A-四个选项中进行选择。
针对每一项实验，我们记录了受试个体是否做出选择响应（'是'或'否'）、若做出响应则其选择的选项，以及做出选择所需的时长（选择潜伏期）。此外，我们还记录了受试个体的编号、行为实验开展时的环境温度，以及播放刺激的声压级（Sound Pressure Level, SPL）。