Data from: Individual discrimination within, but not between, two vocalization types of the black-capped chickadee
Data files
Apr 28, 2025 version files 121.29 KB
-
Clean_Data.zip
103.26 KB
-
DIS_trial_blocks.csv
334 B
-
Novel_DRs.csv
657 B
-
Probe_tests.csv
2.17 KB
-
README.md
13.90 KB
-
Transfer_tests.csv
966 B
Abstract
Many songbird species use individual vocal recognition in their social behaviours. Individual vocal recognition is often assessed using individual discrimination tasks, commonly using an operant conditioning Go/No-go paradigm. Several black-capped chickadee (Poecile atricapillus) vocalizations contain individually distinct features that may be used for individual discrimination. However, not all such vocalizations have been tested for individual recognition with live birds. Additionally, cross-vocalization generalization of learned individual discrimination has not been tested. Such generalizability would be advantageous for chickadees, as chickadees often communicate outside of visual contact and use vocal communication to guide their social interactions. Here we test whether black-capped chickadees can discern the individual identity of callers in black-capped chickadee chick-a-dee calls. We also aim to answer whether chickadees can generalize learned individual discrimination using chick-a-dee calls to fee-bee songs, and vice versa. Chickadees were trained to discriminate several chick-a-dee calls and several fee-bee songs from one male and one female black-capped chickadee, from calls and songs from different males and females in an operant conditioning Go/No-go paradigm. We then tested for generalization across vocalization types by presenting birds with recordings from the same four individuals, this time of the opposing vocalization type. Chickadees were able to discriminate between individuals using either chick-a-dee calls or fee-bee songs, but were unable to generalize this learning to the opposing vocalization type.
https://doi.org/10.5061/dryad.c866t1ghk
Description of the data and file structure
We conducted an operant conditioning (Go/No-go paradigm) study to investigate individual discrimination by black-capped chickadees using either chick-a-dee calls or fee-bee songs. Individual discrimination by black-capped chickadees has previously been demonstrated in fee-bee songs, but not in chick-a-dee calls. We also tested chickadees' ability to generalize learned individual discrimination from one vocalization type to the other (ie. calls to songs, songs to calls). We initially trained chickadees by rewarding them with food for responding to the calls (True 1) or songs (True 2) of two recorded individuals, and punishing them with 30s of darkness for responding to calls or songs recorded from two other individuals. To test learned individual discrimination we presented birds with novel stimuli of the same vocalization type used during initial training recorded from the same four individuals. In Probe 1 these novel stimuli were neither rewarded nor punished, and birds' proportion of Go responses were recorded for each stimulus. In Transfer 1 we rewarded or punished novel stimuli as done during initial training to compared discrimination accuracy of novel stimuli. We then tested for generalization across vocalization types by presenting birds with novel stimuli recorded from the same four individuals again, this time of the vocalization type not used during initial training. We again tested for this first without training (Probe 2) and then with training (Transfer 2). To serve as controls, we trained other chickadees to respond to a pseudo-random assortment of calls (Pseudo 1) or songs (Pseudo 2) recorded from the same four individuals used for True groups. Because half of the stimuli recorded from each individual were rewarded, and the other half punished, birds could not rely on individual discrimination to solve the task and instead employed rote memorization. Five of the eight Pseudo group birds reached pre-designated fail criteria prior to completing the study. Comparisons among True and Pseudo groups are therefore only conducted for some tests.
Files and variables
File: DIS_trial_blocks.csv
Description: This data compiles the number of trial blocks required for birds to learn the individual discrimination task during initial discrimination training (DIS).
Variables
- BirdID: This variable denotes the bird's specific ID number.
- Bird: This variable denotes the bird's study number.
- Group: This variable denotes the bird's group membership as 1 (True 1), 2 (True 2), 3 (Pseudo 1), and 4 (Pseudo 2).
- First: This variable denotes the number of trial blocks required for the bird to reach a discrimination ratio (DR) of at least 0.80. The DR is calculated as the proportion of correct Go responses (entering the feeder during rewarded trials) out of all Go responses (entering the feeder during both rewarded and unrewarded trials).
- Pass: This variable denotes the number of trial blocks required for the bird to meet stage pass criteria (6 trial blocks with a DR of 0.80 or higher, with the last two consecutive.
File: Clean_Data.zip
Description: This zip folder contains clean data for each bird. Birds' discrimination ratios (DR) for each trial block of each experimental stage are contained within the following files: True 1 blocks, True 2 blocks, Pseudo 1 blocks, and Pseudo 2 blocks. Each file contains the data for only one experimental group (that of the file name). Each bird's data begins with a header row, including the group name (eg. Group 1), the bird's ID number (eg. S-17), their sex (M/F), and unit code (D1). Unit codes are used for internal tracking and do not pertain to the study. Column A denotes the study stage shorthand (DIS = discrimination training, D85 = discrimination training with reduced reinforcement, TRS 1 = the first transfer test, T85 = the first transfer test with reduced reinforcement, and TRS 2 = the second transfer test). Beneath TRS rows DRs are also separated by stimulus type, with previously trained stimuli in the row immediately following the collapsed data (termed DIS stim/DIS and TRS1 stim), followed by novel stimuli (termed TRS stim/TRS 2 stim). Columns B onwards track that bird's DR for consecutive trial blocks within that stage, with the first trial block recorded in column B and so on. At the end of each bird's data, the number of trial blocks required to reach DR = 0.80 (80%) in each stage, as well as the number of trial blocks required to meet all passing criteria, are tabulated.
The 24 files named with a probe number followed by a bird's ID, sex, and experimental group (eg. Probe 1 B-92F T2; where T1 denotes group True 1, and T2 denotes True 2) contain that bird's responses to each stimuli within trial blocks for the named probe test. Columns A - C, rows 1-2, denote bin number (trial block) and size (number of stimuli). Each new trial block begins with these two header rows. Subsequent rows, columns A - H, denote the trial information. A = stimulus track number. B = track duration in ms. C = the stimulus reinforcement instruction (where 1000 instructs the program to reward responses to that stimulus, 0 instructs the program to punish responses to that stimulus, and -1 instructs the program to neither reward nor punish the stimulus (ie. probe stimuli)). D = the number of times that stimulus was presented within the trial block, the bird's response to the stimulus. E = the number of times within that trial block the bird left the perch following stimulus presentation, but did not respond by entering the feeder (No-go). F = the number of times within that trial block the bird did respond to the stimulus by entering the feeder (Go). G = reinforcement duration, should the bird be rewarded during the trial. H = the number of trials the bird left the perch prior to listening to the full stimulus (interrupted trials, termed zaps). In columns K - P, the bird's responses across the first six trial blocks of the probe are summarized. K = the stimulus track number. L = the total number of No-go responses. M = the total number of Go responses. N = the total number of interrupted trials. O = the total number of trials were the bird failed to leave the perch following stimulus presentation (null response termed Perch Sits). P = the total number of active responses (Go + No-go responses). Columns S - Z duplicate columns A - H.
The 12 files named with a probe summary followed by a bird's ID, sex, and experimental group (eg. Probe summaries B-92F T2; where T1 denotes group True 1, and T2 denotes True 2) collect the summary of each bird's responses from Probe 1 and Probe 2 (columns K - P of the above described files) in columns A - F. In columns H - M the bird's proportion of Go responses is calculated for each stimulus. Column H delineates between responses during the first probe test (rows 4 - 63) and during the second probe test (rows 67 - 170), as well as marks probe stimuli, and for the second probe test, previous probe stimuli (P1 and P2 respectively). I = the stimulus track number. J = the total number of responses (Go + No-go). L = the number of times that stimulus played within a trial block. L = the proportion of trials in which the bird responded to the stimulus (Go or No-go) out of all presentations of that stimulus. M = the proportion of responses that were Go responses. Column O = the reinforcement instruction for that stimulus, where 1000 instructs the program to reward responses to that stimulus, 0 instructs the program to punish responses to that stimulus, and -1 instructs the program to neither reward nor punish the stimulus (ie. probe stimuli).
The four files names with a probe number followed by experimental group (eg. Probe 1 True 1) collect all birds belonging to that experimental group's proportion of Go responses to all stimuli within the named probe. Columns B - I, rows 5 - 28, collect responses for probe stimuli only, while columns K - X, rows 5 - 50, record responses to trained stimuli only. For columns B - I, rows 7 - 9 refer to stimuli belonging to the reinforced category (also noted in column B), while rows 10 - 12 refer to stimuli belonging to the nonreinforced (punished) category. Row 5, columns D - I, denotes the bird's ID number and sex. Column C = the stimulus track number. Rows 7 - 22, columns D - I denote the bird's proportion of Go responses across 6 trial blocks (calculated in the previously described summary files). Rows 24 - 28, columns C - J calculate each bird's average proportion of Go responses across all stimuli belonging to the reinforced category (row 25) and all those belonging to the nonreinforced category (row 26). Row 5, columns L - Q also name each bird's ID number for the corresponding column. Column K = the stimulus track number. Rows 7 - 50, columns L - Q record the proportion of Go responses to that stimulus across 6 trial blocks calculated in probe summary files. Rows 23 - 25, columns R - W record each bird's average proportion of Go response to reinforced stimuli (row 24) and nonreinforced stimuli (row 25). The experimental group's average response to all stimuli categories (probe-reinforced = P+, probe-nonreinforced = P-, reinforced = S+, nonreinforced = S-) are recorded in both in rows 41-45, columns G - H, and in rows 50-54, columns G - I.
File: Novel_DRs.csv
Description: This data compiles True group birds' discrimination ratios (DR) for novel stimuli during the first six trial blocks during each stage. DRs are calculated as the proportion of correct Go responses (entering the feeder during rewarded trials) out of all Go responses (entering the feeder during both rewarded and unrewarded trials). DRs for trained stimuli are not included.
Variables
- BirdID: This variable denotes the bird's specific ID number.
- Bird: This variable denotes the bird's study number.
- Group: This variable denotes the bird's group membership as 1 (True 1) or 2 (True 2).
- Stage: This variable denotes the stage the bird's average DR comes from, marked as 1 (initial discrimination training), 2 (within-vocalization type transfer test), or 3 (novel-vocalization type transfer test).
- DR: This variable denotes the bird's average discrimination ratio for novel stimuli within the first 6 trial blocks of each stage. DRs are calculated as the proportion of correct Go responses (entering the feeder during rewarded trials) out of all Go responses (entering the feeder during both rewarded and unrewarded trials).
File: Transfer_tests.csv
Description: This data compiles True group birds' discrimination ratios (DR) for both previously trained and novel stimuli during the within-vocalization type transfer test and the novel-vocalization type transfer test. DRs are calculated as the proportion of correct Go responses (entering the feeder during rewarded trials) out of all Go responses (entering the feeder during both rewarded and unrewarded trials).
Variables
- BirdID: This variable denotes the bird's specific ID number.
- Bird: This variable denotes the bird's study number.
- Group: This variable denotes the bird's group membership as 1 (True 1) or 2 (True 2).
- Stage: This variable denotes the stage the bird's DR comes from, marked as 1 (within-vocalization type transfer test), or 2 (novel-vocalization type transfer test).
- Novelty1N: This variable denotes the novelty of the stimuli the bird's DR comes from, marked as 1 (novel stimuli), or 2 (previously trained stimuli).
- DR: This variable denotes the bird's average discrimination ratio for either novel or previously trained stimuli within the first 6 trial blocks of each transfer test. DRs are calculated as the proportion of correct Go responses (entering the feeder during rewarded trials) out of all Go responses (entering the feeder during both rewarded and unrewarded trials).
File: Probe_tests.csv
Description: This data compiles True group birds' proportion of Go responses for trained and rewarded (sp), trained and punished (sm), novel "rewarded" (pp), and novel "punished" (pm) stimuli during each probe test (6 trial blocks each). Novel stimuli were neither punished nor rewarded in order to test for generalization without training. A Go response is recorded when the bird enters the automated feeder immediately following stimulus presentation.
Variables
- BirdID: This variable denotes the bird's specific ID number.
- Bird: This variable denotes the bird's study number.
- Group: This variable denotes the bird's group membership as 1 (True 1) or 2 (True 2).
- Probe: This variable denotes the probe test the proportion of Go responses comes from with 1 representing the within-vocalization type probe test, and 2 representing the novel vocalization type probe test.
- Types: This variable denotes the stimulus category shorthand for trained and rewarded (sp), trained and punished (sm), novel "rewarded" (pp), and novel "punished" (pm) stimuli.
- Type: This variable denotes the stimulus category for trained and rewarded (1), trained and punished (2), novel "rewarded" (3), and novel "punished" (4) stimuli.
- GO: This variable denotes the bird's average proportion of Go responses for all stimuli within each stimulus category during the respective probe test (six trial blocks).
Code/software
This dataset was processed using R v4.3.1 (R Core Team, 2023). The R code for all associated tests has been uploaded to Zenodo.
The following packages have been loaded (and are included at the top of the R script): tidyverse, ggpubr, rstatix, ggplot2, afex, and emmeans.