Deconstructing "Boldness": Latent behavioral structure across novelty, predation, and exploration in an endangered species translocation program

PAPER TITLE: Deconstructing “Boldness”: Latent behavioral structure across novelty, predation, and exploration in an endangered species translocation program ABSTRACT: Risk-related behaviours including neophobia, exploration, and predator avoidance are often assumed to covary along a general “boldness” axis. Although some translocation studies have linked one or more of these behaviours with fitness, it is possible that the underlying behavioural structure varies under different selective environments. Conservation breeding programs provide a unique opportunity to evaluate the structure of behavioural correlations in a non-domestic species under artificial conditions. We tested whether risk-related behaviours represent a unified boldness syndrome in a captive population of the endangered Attwater’s prairie chicken (Tympanuchus cupido attwateri). We quantified the repeatable behaviours of 257 juveniles and compared 11 models of behavioural covariation structure using multivariate mixed models and confirmatory factor analysis. There was no support for a single risk-taking axis, but instead we found three dimensions representing responses to novelty and environmental stimuli (Engagement), predator avoidance (Threat Response), and novel maze exploration (Maze Progression). Our results indicate that these risk-related behaviours are not organized along a single axis in captive APCs, challenging assumptions that risk-taking represents a universal behavioural syndrome. To anticipate post-release success accurately, conservation managers should understand the exact behavioural architecture of the captive population rather than assume that the classical bold-shy structure occurs in all contexts. PUBLICATION INFORMATION: Authors: Laney H. Nute & Richard Buchholz Journal Name: Proceedings B doi:10.1098/rspb GENERAL INFORMATION 1. Title of Datasets: behaviour_apcdata.csv 2. Author Information A. Principal Investigator Contact Information Name: Laney H. Nute Institution: Department of Biology, University of Mississippi Email: [e-mail] 3. Dates of data collection: April 2023 - July 2025 4. Geographic location of data collection: NASA Johnson Space Center, Houston, Texas 5. Information about funding sources that supported the collection of the data: Funding was provided by the Wild Animal Initiative [grant number C-2023-00017]; University of Mississippi Center for Biodiversity and Conservation Research Graduate Summer Research Grant [grant numbers 2022 and 2025]; Sigma Xi Grants in Aid of Research [grant number G20230315-5281]; University of Mississippi Biology Department; NASA Mississippi Space Grant Consortium student grant [grant number 80NSSC20M0101]; and the University of Mississippi Graduate School Summer Research Assistantship 6. License information: CC-BY 4.0 7. DOI for data and code repository: 10.5281/zenodo.20738061 Data & File Overview 1. File List: Data behaviour_apcdata.csv: csv file containing raw behavioural data for each tested bird Code manuscript_syndrome_code2026_v2.Rmd: R markdown file containing statistical code used to analyse the behavioural data (including general notes). R Objects mcmc_AIC.rds: R object for the multivariate analysis on the information theory derived predictors. prior_sensitivity_models.rds: R object for the multivariate models using the alternative priors. censored_13trait_mcmc.rds: R object for the final multivariate model used for analysis of the 13 repeatable behaviours. mcmc_apriori.rds: R object for the multivariate model using no predictors aside from year and trial. 2. Relationship between files, if important: manuscript_syndrome_code_2026_v2.Rmd is the R code used to generate the provided R objects and uses the data collected in the behaviour_apcdata.csv file. 3. Additional related data collected that was not included in the current data package: N/A 4. Are there multiple versions of the dataset? No METHODOLOGICAL INFORMATION 1. Description of methods used for collection/generation of data: Birds were tested using a three day testing battery consisting of a Novel Environment & Object assay, a Predator Flyover assay, and a Novel Maze assay at two and four weeks of age. All 10-minute trials were videod and later transcribed using a mix of manual, semi-manual (MTrackJ), and machine learning methods (DeepLabCuts and SIMBA). Trials were run in the mornings, with some exceptions, and temperature, windspeed, weather, and humidity were recorded. Testing order of subjects was randomized when possible. Assays took place in topless, modified brooder cages. Brooder sides were covered in fabric (2023) or cardboard (2024 and 2025) using pink for trial one and blue for trial two. One pink (trial one) or blue (trial two) novel object was placed in the front half of the enclosure and another in the back half (ping pong balls in 2023 and rubber ducks in 2024 and 2025). The floor was gridded and covered with non-slip shelf liners. Each brooder was covered with mesh and each chick was placed in an acclimation chamber at the start of the trial for the novel environment assays. Access to the chamber was also allowed for the Predator Flyover, but not the Novel Maze. 2. Methods for processing the data: We used a mixture of fully manual transcription, manual movement tracking software (MTrackJ version 1.5.1) and machine learning open-source programs (DeepLabCuts version 3.0.0rc7). We used the open-source SIMBA program (version 4.2.8) to identify bouts of escape attempts but were unable to identify bouts of novel object interactions, so we manually recorded those behaviours. We achieved acceptable precision, recall, and F1 scores for all assays and visually checked outputs. 3. Instrument- or software-specific information (including package versions) needed to interpret the data: R version 4.4.3 (2025-02-28) Platform: macOS Sequoia 15.7.9 Running under: x86_64, darwin20 attached packages: package * version date (UTC) lib source ape * 5.8 2024-04-11 [1] CRAN (R 4.4.0) car * 3.1-3 2024-09-27 [1] CRAN (R 4.4.1) carData * 3.0-5 2022-01-06 [1] CRAN (R 4.4.0) coda * 0.19-4.1 2024-01-31 [1] CRAN (R 4.4.0) dplyr * 1.1.4 2023-11-17 [1] CRAN (R 4.4.0) forcats * 1.0.0 2023-01-29 [1] CRAN (R 4.4.0) ggplot2 * 4.0.1 2025-11-14 [1] CRAN (R 4.4.1) lavaan * 0.6-19 2024-09-26 [1] CRAN (R 4.4.1) lme4 * 1.1-37 2025-03-26 [1] CRAN (R 4.4.1) lmerTest * 3.1-3 2020-10-23 [1] CRAN (R 4.4.0) lubridate * 1.9.4 2024-12-08 [1] CRAN (R 4.4.1) Matrix * 1.7-4 2025-08-28 [1] CRAN (R 4.4.1) MCMCglmm * 2.36 2024-05-06 [1] CRAN (R 4.4.0) MuMIn * 1.48.11 2025-04-01 [1] CRAN (R 4.4.1) nlme * 3.1-168 2025-03-31 [1] CRAN (R 4.4.1) performance * 0.13.0 2025-01-15 [1] CRAN (R 4.4.1) purrr * 1.2.0 2025-11-04 [1] CRAN (R 4.4.1) readr * 2.1.5 2024-01-10 [1] CRAN (R 4.4.0) rptR * 0.9.22 2019-03-06 [1] CRAN (R 4.4.0) stringr * 1.6.0 2025-11-04 [1] CRAN (R 4.4.1) summarytools * 1.1.1 2025-02-26 [1] CRAN (R 4.4.1) tibble * 3.3.0 2025-06-08 [1] CRAN (R 4.4.1) tidyr * 1.3.1 2024-01-24 [1] CRAN (R 4.4.0) tidyverse * 2.0.0 2023-02-22 [1] CRAN (R 4.4.0) DATA-SPECIFIC INFORMATION FOR: behaviour_apcdata.csv 1. Number of variables: 78 2. Number of cases/rows: 1438 3. Variable List: Column 1. key = a key for isolating trial one and trial two per individual that includes the individual id, trial type (NE = Novel Environment, NP = Predator Flyover, and NM = Novel Maze), and trial number (APC001-2024_NE1 - APC132-2025_NP1). Column 2. id = individual identity by number per year (APC009-2023 - APC132-2025). Column 3. year = year of trial (2023, 2024, 2025). Column 4. group_id = rearing group of individual (23_NASA4 - 25_NASA109). Column 5. date = date of trial by m/dd/yy format. Column 6. assay = trial type (NE = Novel Environment, NP = Predator Flyover, and NM = Novel Maze). Column 7. trial = trial number (1, 2). Column 8. age = age of individual at trial date (12 - 39). Column 9. start = start time of trial in hh:mm:ss. Column 10. temp = temperature at trial time in Fahrenheit Column 11. wind_speed = wind speed in miles per hour at trial time. Column 12. wind_dir = wind direction at trial time (N, S, E, W, NE, NW, SE, SW, NNE, NNW, SSE, SSW, ENE, ESE, SWW). Column 13. weather = the categorical weather description at time of trial (Isolated Thunderstorm, Sunny, Partially Sunny, Cloudy, Windy, Raining). Column 14. humidity = humidy at time of trial (0.55 - 0.9). Column 15. brooder_id = the testing enclosure/brooder that the chick was tested in (A,B,C) and year (1,2,3). Column 16. catch_time = the time in seconds it took to catch bird after entering home enclosure. Column 17. handling_time = the time in seconds from chick capture until trial began. Column 18. video_length = length of video recording in seconds. Column 19. video_quality = categorical description of video clarity upon post-trial review during behavioural transcription (High, Medium, Low). Column 20. out_of_frame = categorical identifier on if the individual dissapeared from camera view during trial (Y/N). Column 21. lighting = categorical description of the visibility due to lighting upon post-trial review during behavioural transcription (High, Medium, Low). Column 22. light_placement = categorical description of the placement of the light source during the trial (top_right, above, top_left). Column 23. glare = categorical identifier on whether there was the presence of a glare on the recording during post-trial behavioural transcription (Y/N). Column 24. short_video = categorical identifier on if the video was shorter than 600 seconds, or ten minutes (Y/N). Column 25. escaped_enclosure = categorical identifier on whether the individual successfully jumped out of the enclosure during the trial (Y/N). Column 26. enclosure_out_view = categorical identifier on whether all walls of the testing arena were in camera view during post-trial behavioural transciption (Y/N). Column 27. camera_shakes = categorical identifier on whether the video image was shakey due to weather and windspeed during post-trial behavioural transcription (Y/N). Column 28. problem_camview = categorical identifier on whether there was a during post-trial behavioural transcription with the view of the camera (Y/N). Column 29. problem_setup = categorical identifier on whether there was an issue with arena setup visible during post-trial behavioural transcription such as an object out of place or a fallen panel (Y/N). Column 30. emerge_lat = how long it took the chick to emerge from the acclimation chamber in seconds during the Novel Environment and Object trials. Column 31. obj_lat = latency in seconds to approach the novel objects in the Novel Environment and Object trials. Column 32. adj_ttime_obj = the time in seconds spent interacting with the novel objects, represented as a percent of total trial time, in the Novel Environment and Object trials. Column 33. esc_lat = latency in seconds to attempt jumping out of the testing arena. Column 34. adj_ttime_esc = the time in seconds spent attempting to jump out of the testing arena represented as a percent of total trial time. Column 35. distance_mov = the total distance moved during the trial represented as mm. Column 36. velocity = the speed of travel represented as the total distance moved divided by the total trial time. Column 37. mov_1min = the total distance moved within the first minute of the trial. Column 38. vel_1min = the speed of travel within the first minute of the trial represented as the distance moved divided by 60 seconds. Column 39. mov_10min = the total distance moved within the last minute of the trial. Column 40. vel_10min = the speed of travel within the last minute of the trial represented as the distance moved divided by 60 seconds. Column 41. adj_ttime_shelter = the time in seconds spent in the acclimation chamber represented as a percent of total trial time during the Novel Environment and Object trials and the Predator Flyover trials. Column 42. adj_ttime_food = the time in seconds spent in the food tray represented as a percent of total trial time during the Predator Flyover trials. Column 43. bef_adj_esc = the time in seconds spent attempting to jump out of the testing arena before flying over the predator silhoette represented as a percent of trial time before flyover. Column 44. bef_latesc = latency in seconds to attempt jumping out of the testing arena before flying over the predator silhoette. Column 45. bef_adj_shelter = the time in seconds spent in the acclimation chamber before flying over the predator silhoette represented as a percent of trial time before flyover. Column 46. bef_adj_food = the time in seconds spent in the food tray before flying over the predator silhoette represented as a percent of trial time before flyover. Column 47. bef_dist = the total distance moved before flying over the predator silhoette represented as mm. Column 48. bef_vel = the speed of travel before flying over the predator silhoette represented as the total distance moved divided by the total trial time before flyover. Column 49. dur_adj_esc = the time in seconds spent attempting to jump out of the testing arena during flying over the predator silhoette represented as a percent of trial time during flyover. Column 50. adj_lat_esc = latency in seconds to attempt jumping out of the testing arena during flying over the predator silhoette. Column 51. dur_adj_shelter = the time in seconds spent in the acclimation chamber during flying over the predator silhoette represented as a percent of trial time during flyover. Column 52. dur_adj_food = the time in seconds spent in the food tray during flying over the predator silhoette represented as a percent of trial time during flyover. Column 53. dur_dist = the total distance moved during flying over the predator silhoette represented as mm. Column 54. dur_vel = the speed of travel during flying over the predator silhoette represented as the total distance moved divided by the total trial time during flyover. Column 55. aft_adj_esc = the time in seconds spent attempting to jump out of the testing arena after flying over the predator silhoette represented as a percent of trial time after flyover. Column 56. aft_latesc_adj = latency in seconds to attempt jumping out of the testing arena after flying over the predator silhoette. Column 57. aft_adj_shelter = the time in seconds spent in the acclimation chamber after flying over the predator silhoette represented as a percent of trial time after flyover. Column 58. aft_adj_food = the time in seconds spent in the food tray after flying over the predator silhoette represented as a percent of trial time after flyover. Column 59. aft_dist = the total distance moved after flying over the predator silhoette represented as mm. Column 60. aft_vel = the speed of travel after flying over the predator silhoette represented as the total distance moved divided by the total trial time after flyover. Column 61. diff_adj_esc = the difference in the time spent attempting to jump out of the testing arena before the predator flyover and the time spent attempting to jump out of the testing arena after the predator flyover. Column 62. diff_latesc = the difference in the latency in seconds to attempt jumping out of the testing arena before the predator flyover and the latency in seconds to attempt jumping out of the testing arena after the predator flyover. Column 63. diff_adj_shelter = the difference in the time in seconds spent in the acclimation chamber before the predator flyover and the time in seconds spent in the acclimation chamber after the predator flyover. Column 64. diff_adj_food = the difference in the time in seconds spent in the food tray before the predator flyover and the time in seconds spent in the food tray after the predator flyover. Column 65. diff_dist = the difference in the distance in mm moved before the predator flyover and the distance in mm moved after the predator flyover. Column 66. diff_vel = the difference in the speed of travel represented as the total distance moved divided by the total time before the predator flyover and the the speed of travel after the predator flyover. Column 67. np_latesc = the minimum latency to attempt escaping the testing arena before, during, and after flying over the predator silhoette. Column 68. lat_sec2 = latency in seconds to enter section two of the maze. Column 69. lat_sec3 = latency in seconds to enter section three of the maze. Column 70. lat_sec4 = latency in seconds to enter section four of the maze. Column 71. lat_sec5 = latency in seconds to enter section five, or to complete, the maze. Column 72. Emerged = TRUE/FALSE entry on whether the bird entered the novel environment during the trial. Column 73. Object = TRUE/FALSE entry on whether the bird interacted with the novel object items during the trial. Column 74. Escaped = TRUE/FALSE entry on whether the bird attempted to jump out of the enclosure during the trial. Column 75. S2 = TRUE/FALSE entry on whether the bird entered section two during the maze trial. Column 76. S3 = TRUE/FALSE entry on whether the bird entered section three during the maze trial. Column 77. S4 = TRUE/FALSE entry on whether the bird entered section four during the maze trial. Column 78. S5 = TRUE/FALSE entry on whether the bird entered section five (finishing the maze) during the maze trial. 4. Missing data codes: NA