Túngara frog call-timing decisions arise as internal rhythms interact with fluctuating chorus noise
Data files
Apr 23, 2024 version files 8.50 MB
-
Overlap-Received_Repeatability_and_SMA_Regression.zip
-
READ_ME.docx
-
README.md
-
Single_Caller_Transition_Analysis_GLMM.zip
-
Synchronous_Caller_Transition_Analysis_GLMM.zip
-
Transition_Probability_Repeatability_Analysis.zip
Abstract
For chorusing males, optimally timing their calls relative to nearby rivals’ calls and fluctuations in background chorus noise is crucial for reproductive success. A caller’s acoustic environment will vary by chorus density and the properties of his chorus-mates’ calls and will fluctuate unpredictably due to chorusing dynamics emerging among his chorus-mates. Thus, callers must continuously monitor moment-to-moment fluctuations in the acoustic scene they perceive at the chorus for advantageous times to call. In live experimental choruses, we investigated the factors influencing túngara frog call-timing responses to chorus-mates’ calls on an interaction-by-interaction basis, revealing that intrinsic and extrinsic factors influenced call-timing decisions. Callers were more likely to overlap calls from smaller chorus-mates and chorus-mates at intermediate distances, as well as calls containing lower frequencies and exhibiting lower final amplitude minima. Consequently, variation among males in call properties led to variation in levels of call-interference received when calling in the same social environment. Additionally, callers were more likely to overlap chorus-mates’ calls after experiencing extended periods of inhibition and were less likely to overlap synchronized chorus-mates’ calls relative to single calls. In chorusing species, female choice is influenced by inter-caller dynamics, selecting for male call-timing strategies which, in turn, constitute the selective environment further refining these same strategies. Thus, understanding the specific factors driving call-timing decisions is essential for understanding how sexual selection operates in chorusing taxa.
README: Túngara frog call-timing decisions arise as internal rhythms interact with fluctuating chorus noise
Corresponding author: luke.larter@utexas.edu
Folder ‘Transition Probability Repeatability Analysis’
Contains R markdown file and CSV files to run analysis looking at how repeatable inter-caller transition patterns were over time (corresponding to analysis detailed in the first paragraph of the section titled ‘How Stable are Chorusing Dynamics Over Time?’ in the manuscript).
- ‘210927_001_SHORT_6_males_from_660_to_870_secs_transition_prob_matrix_first70.csv’: a CSV containing a matrix of pair-wise inter-caller transition probabilities among all chorus-mates in chorus ID ‘210927_001’, for the first 70s of the entire 3.5-minute recording. ‘…_second7*0.csv’ and ‘…_third70.csv’* correspond to the second and third of the successive 70s, respectively. There are similar CSV files for each different chorus ID.
- ‘inter-caller_transition_probability_repeatability_analysis.rmd’: Code using the above CSV files to analyse the repeatability of transition probabilities over time. More details are in the code comments.
Folder ‘Overlap-Received Repeatability and SMA Regression’
Contains R markdown file and CSV files to run analysis looking at how repeatable individual variation in the degree of call overlap received from chorus-mates was over time (corresponding to analysis detailed in the second paragraph of section titled ‘How Stable are Chorusing Dynamics Over Time?’ in the manuscript).
- ‘210927_001_SHORT_6_males_from_660_to_870_secs_simple_overlap_proportion.csv’: a CSV containing a data frame with the proportion of each chorus-mate calls that were overlapped by the call of at least one other chorus-mate for each of the successive 70s segments making up the overall 3.5-minute segment in chorus ID ‘210927_001’. Similar CSV files are present for all chorus IDs. ‘segment’ column is which 70s the value corresponds to (s1 = first 70s, s2 = second, s3 = third); ‘chorus_ID’ is chorus identifier; ‘caller_ID’ is the track ID of the chorus-mate; ‘num_calls’ is the number of calls that male produced in that segment, ‘num_overlapped’ is the number of these that were overlapped by at least one of his chorus-mates, ‘not_overlapped’ is the number that was not overlapped at all, and ‘overlap_props’ is the proportion of his calls that received overlap from at least one of his chorus-mates during that segment.
- ‘overlap_received_repeatablitity_analysis.rmd’: Code using the above CSV files to analyse the repeatability of overlap received over time. More details in code comments
Folder ‘Single Caller Transition Analysis GLMM’
Contains R markdown file and CSV files to run analysis looking at how variation in call properties influences how males related to the calls of chorus-mates (corresponding to analysis detailed in the section titled ‘Which Properties of Preceding Calls Influence the Probability of Call Overlap?’ in the manuscript).
- ‘single_caller_transition_data.csv’: Dataframe with rows representing each following call made by all males in all choruses in this study, displaying whether they overlapped the call preceding them or not, and detailing the acoustic properties of that preceding call and the time that has elapsed since the following caller’s last call. ‘ID_focal_following_caller’ is the track ID of the following caller; ‘chorus_ID’ is chorus_identifier; ‘did_focal_follower_overlap_preceding_call_YN’ is the binary response denoting whether the following caller overlapped the preceding caller or not (1 = yes, 0 = no)’; ‘IOI_duration_preceding_focal_call’ is the time that has elapsed (inter-onset interval) since the last call by the following caller; ‘focal_svl’ is the snout-vent-length (body length) of the focal following caller (mm); ‘focal_mass’ is mass of the focal following caller (g); ‘focal_toe_clip_number’ is the permanent ID of the focal following caller (one male appeared in two different choruses, thus had two different ‘ID_focal_following_caller’ track IDs); ‘preceding_caller_svl’ is snout-vent length of preceding caller (mm); ‘preceding_caller_mass’ is mass of preceding caller (g); ‘preceding_caller_focal_toe_clip_number’ is the permanent ID of the preceding caller; ‘max_freq’ is the maximum frequency of the whine of the preceding caller (Hz); ‘min_freq’ is the minimum frequency of the whine of the preceding caller (Hz); ‘after_knee_amplitude_skewness’ is the amplitude skewness of the whine of the preceding caller; ‘min_last_150’ is the minimum final amplitude of the whine of the preceding caller relative to the maximum whine amplitude; ‘distance’ is the inter-caller distance between preceding and following callers (m).
- ‘single_caller_transition_analysis.rmd’: Code using the above CSV to analyse the factors influencing overlap probability during inter-caller transitions from single preceding callers. More details are in the code comments.
Folder ‘Synchronous Caller Transition Analysis GLMM’
- ‘synchronous_caller_transition_data.csv’: Data with rows representing each following call made by all males in all choruses in these studies, displaying whether they overlapped the call preceding them or not, and detailing the number of synchronous callers contributing to that preceding call and the time that has elapsed since the following caller’s last call. ‘ID_focal_following_caller’ is the track ID of the following caller; ‘chorus_ID’ is chorus_identifier; ‘did_focal_follower_overlap_preceding_call_YN’ is the binary response denoting whether the following caller overlapped the preceding caller or not (1 = yes, 0 = no)’; ‘IDs_synchronous_callers_preceding_focal_call’ is the list of the track IDs of the synchronous callers contributing to the preceding call (necessary for the multiple membership term in the GLMM); ‘Number_synchronous_callers_preceding_focal_call’ is the number of synchronous callers contributing to the preceding call; ‘IOI_duration_preceding_focal_call’ is the time that has elapsed (inter-onset interval) since the last call by the following caller; ‘focal_toe_clip_number’ is the permanent ID of the focal following caller (one male appeared in two different choruses, thus had two different ‘ID_focal_following_caller’ track IDs).
- ‘synchronous_caller_transition_analysis.rmd’: ‘single_caller_transition_analysis.rmd’: Code using the above CSV to analyse the factors influencing overlap probability during inter-caller transitions from single or synchronous preceding callers. More details are in the code comments.
Data/Code
Version Information and Packages Used
- RStudio v2021.09.0/351 running R v4.2.2
- tidyverse_v2.0.0
- lme4_v1.1.31
- DHARMa_v0.4.6
- rptr_v0.9.22
- smatr_v3.4.8
- MuMIn_v1.47.5
- sjPlot_v2.8.12
- ggplot2_v3.4.1
- lmerTest_v3.1.3
- car_v3.1.2
- splines_v4.2.2
- lmerMultiMember_v0.11.4
- ggridges_v0.5.4
- sna_v2.7.1
- lsa_v0.73.3
- afex v1.3-0
Methods
In this study, we investigated call-timing interactions among male túngara frogs by recording calling patterns of interacting males within 6 experimental choruses of 6 callers each. We analysed underlying chorusing dynamics and investigated how inter-male variation in call properties influenced how chorus-mates related to one another’s calls.