This Readme.txt file was generated on 20170210 by Daniel Zielinski ------------------- GENERAL INFORMATION ------------------- 1. Title of Dataset: Acoustically mediated avoidance of silver, bighead, and common carp 2. Author Information Principal Investigator Contact Information: Name: Peter Sorensen Institution: University of Minnesota Address: 135E Skok Hall, 2003 Upper Buford Circle, St. Paul, MN 55108 Email: soren003@umn.edu Associate or Co-investigator Contact Information: Name: Daniel Zielinski Institution: Great Lakes Fishery Commission Address: 11188 Ray Rd., Millersburg, MI 49759 Email: 3. Date of data collection: 20131129 to 20140206 4. Geographic location of data collection: Engineering and Fisheries Laboratory, University of Minnesota, 1955 Fitch Ave., Falcon Heights, MN 55108 5. Information about funding sources that supported the collection of the data: This study was funded by the Minnesota Environmental and Natural Resources Trust Fund as recommended by the Legislative-Citizen Commission on Minnesota Resources (LCCMR). -------------------- DATA & FILE OVERVIEW -------------------- 1. File/Folder List Part1: Sound Data Short description: Folder contains sound pressure and acoustic particle motion measurements within the experimental enclosure. A. Filename: SoundSummary.txt SoundSummary.txt contains the coordinates of all measurement points and the accompanying peak sound pressure level (@ 150 Hz, 750 Hz, and 2000 Hz), and particle acceleration. Table headers descriptions are "PT" - point number, "Inst. Staff X (in)" and "Inst. Staff Y (in)" - coordinates of the center of the instrument staff, "Hydrophone X (cm)" and "Hydrophone Y (cm)" - coordinates of the hydrophone probe [offset from center of instrumentation staff], "Accel. X (cm)" and "Accel. Y (cm)" - coordinate of the accelerometer probe, "SPL @ XXX Hz (dB ref. 1 mPa)" - Sound pressure level at XXX Hz frequency in decibels referenced to 1x10^-6 Pascals, "Par. Accel. (cms/s2) X-dir" and "Par. Accel. (cm/s2) Y-dir" - Acoustic particle acceleration in the lateral (X) and longitudinal (Y) directions, "Part. Accel. Magnitude (dB ref 1 cm/s2)" - magnitude of the particle acceleration in the XY plane in decibels relative to 1 cm/s^2. B. Filename: Sound Pressure.zip The Sound Pressure file contains the raw sound pressure measurements at each measurement point. Each file contains a time stamp and sound pressure (in Pa). The numeral after "x" in each files name corresponds to the measurement point in SoundSummary. C. Filename: Particle Acceleration.zip The Particle Acceleration File contains the raw acoustic particle acceleration measurements at each measurement point. Measurement location is indicated in a similar manner as sound pressure files. Each file contains a time stamp and particle acceleration in the X-, Y-, and Z-directions (in cm/s^2). Part 2: Fish Position Data Short description: Folder contains positional data of fish for all tests. Note, orientation of the speakers changes between common carp and silver/bighead carp tests. A. Filename: Experimental Set-up and Coordinate System_Common Carp.pdf Experimental Set-up and Coordinate System_Common Carp.pdf contains an image of the experimental set-up and location of speakers for tests with common carp. B. Filename: Experimental Set-up and Coordinate System_Silver and Bighead Carp.pdf Experimental Set-up and Coordinate System_Silver and Bighead Carp.pdf contains an image of the experimental set-up and location of speakers for tests with silver and bighead carp. C. Filename: Date_Size_Species_Summary.txt Date_Sizes_Species_Summary.txt contains a summary of the fish size and date of testing for all species. D. Filename: Fish Position Data.zip Fish Position Data.zip contains raw positional data of each fish during tests. Inside the folder, the .txt files contain positions of fish during rest (no sound) and test (sound on) periods. Note that fish are not tracked individually at this stage (i.e. fish are randomly sampled at each time 5 sec time step). The .txt files contain continuously tracked swim paths of individual fish during control or test periods. Fish tracks are designated by date of trial. 2. Relationship between files: Position and orientation of fish relative to the sound field (pressure and particle acceleration) were examined to determine if/how these carp use sound to avoid a sound source. 3. Additional related data collected that was not included in the current data package: Final data analysis of the raw data is detailed in a manuscript "Silver, bighead, and common carp orient to acoustic particle motion when avoiding a complex sound" submitted to PLOSONE. 4. Are there multiple versions of the dataset? no -------------------------- METHODOLOGICAL INFORMATION -------------------------- 1. Description of methods used for collection/generation of data: Experimental methods for this project are provided in the manuscript "Silver, bighead, and common carp orient to acoustic particle motion when avoiding a complex sound" submitted to PLOSONE in Feb. 2017. A. Fish Locations: Experiments were performed in a cylindrical fiberglass tank (3 m diameter, 2 m in depth) into which an internal square opaque plastic enclosure (1.8 m on a side, 50 cm high, 150 microns thick) had been placed to render the testing arena featureless. The center of the arena had a drain pipe which was also shielded with a black plastic box (50-cm high on each side). This tank was supplied with well water to a depth of 30 cm and aerated by airstones positioned outside of the enclosure in each corner. A black plastic tarp covered the entire tank and three infrared floodlights illuminated the inside of the featureless and dark arena (<0.5 lux, 840 nm). Four UW30 speakers (output level 153 dB (ref. 1 µPa) at 1 m, frequency response 0.1 to 10 kHz, Electrovoice, MN, USA) were positioned outside the plastic arena at the center of each side using cables that acoustically separated them from the tank and set the center of the speaker 15-cm above the tank floor. A closed circuit video camera (Interlogix, NC, USA) was mounted 3 m above the tank bottom, and recorded at 30 frames per second through each experiment. Video files were later downloaded from a DVR and a custom Matlab (Mathworks, MA, USA) script was used for frame-by-frame analysis. Tests were conducted between 10:00 and 16:00 h between December 2013 and August 2014. Fish were tested as groups of three individuals of the same species to facilitate natural shoaling behavior and reduce stress. Prior to testing, fish were allowed to acclimate, shoal, and move freely. Acclimation times differed by species and had been determined beforehand by extensive pilot tests as the periods of time required by fish to start to explore tanks and feed when offered food (130 min for common carp, 20 min for silver carp, and 24 h for bighead carp). Water inflow and airstones were turned off 10 min prior to the start of each trial. After the 150-s pre-test period (control), the test sound was played once two individual carp swam within 30-cm of any one of the four speakers, at which time that speaker was turned on for 150-s (treatment). The 30-cm distance was used as a threshold because sound mapping showed it to coincide with both the region of maximum sound pressure and the 1 cms-2 particle acceleration limit for avoidance behaviors. This procedure was repeated for four trials (each with a control and treatment period) until all four speakers had been used once (time between trials varied, and fish could not learn order of testing). After testing, fish were removed and placed into a control tank. Each species was tested 7 times and no fishes were reused. B. Sound Measurements: Both sound pressure levels and particle acceleration were mapped on a Cartesian grid throughout the tank at 5 cm intervals within 30-cm of each of the four speakers and 15 cm above the tank bottom. Sound measurements were made using a PVC probe which contained a C55 hydrophone (usable frequency range of 0.008–100 kHz and a sensitivity of approximately -163.5 dB ref 1V/1x10^-6Pa, Cetacean Research, WA, USA) and PCB model W356A12 triaxial accelerometer (usable frequency of 0.5–5000 Hz and sensitivity of approximately 100 mV/ (m/s2), PCB Piezoelectronics, NY, USA). The sound pressure signal was sampled at 44.1 kHz and fed through a TASCAM US-122mkII (TASCAM, CA, USA) audio interface, digitized, and stored on a windows-based computer. The accelerometer was made neutrally buoyant by embedding it in an extruded polystyrene foam enclosure. The acoustic particle acceleration signal was conditioned using a PCB 482C05 conditioner and fed through a USB-1208FS-Plus data acquisition board (Measurement Computing, MA, USA) sampling each channel at 16 kHz. At each location, a 10 s sample was split into 10 signal ensembles and averaged. Data acquisition hardware was controlled by a custom graphical user interface operating in Matlab, which was also used to analyze and transform the pressure and particle acceleration waveforms into the frequency domain. 2. Methods for processing the data: Data were evaluated in two steps. Step one evaluated fish distribution (and avoidance) while step two determined the tracks that individual carp followed and then evaluated how fish oriented to known sound fields to discern the orientation mechanisms they were using. For the first analysis, the percent time each fish spent within 30-cm of the active speaker (or the soon-to-be active speaker for control periods) was calculated after viewing videos. This was accomplished by recording the x and y coordinates of each fish’s head within each group of three at 5 s intervals (i.e. once every 150 frames). The initial 30-s of each treatment period was not used in the statistical analyses to reduce the role that brief startle responses might have had in the analysis. For each group of fish (and trial), the percent time fish spent within 30-cm of the active speaker was calculated by dividing the total number of times any fish was within 30-cm of the active speaker by the total number of data points. These values were examined for normalcy (Shapiro-Wilk tests) and appropriate paired comparisons performed. Because the data were not normally distributed nonparametric Mann-Whitney U-tests were used to compare differences in the percent time that groups of fish of each species spent within 30-cm of an active speaker between matched control and treatment periods (i.e. Control #1 vs. SPK#1, Control #2 vs. SPK#2, etc.). Significance was determined at P<0.05. The second set of analyses examined the relationship between the orientation of individual fish to different components (sound pressure and particle motion) of the sound field and its source. To accomplish this we calculated both the difference angle between the fish’s bearing relative to the sound source as well as the difference angle to the sensory field (particle acceleration vector). Swimming trajectories were determined for each fish that swam within 30-cm of an active speaker at a 3 Hz sampling frequency. The position of each fish was evaluated 5 s before and after coming within 30-cm of the speaker, so fish were monitored to distances exceeding 30-cm (some up to 125 cm). The entire treatment period was evaluated for each fish found in this space. The x and y coordinates of each fish’s head were used to determine both their distance from the source and orientation relative to measured sound fields. To test whether they orientated differently as they approached and then left the sound field, difference angles were analyzed separately as fish swam towards and away from the speaker. Both the difference angle relative to the speaker, and difference angle relative to the local particle acceleration vector were calculated from the fishes trajectory in the xy-plane. Contributions of particle acceleration in the z-direction were ignored because particle acceleration magnitudes were similar throughout the enclosure in all three-directions and fish movement was laterally restricted. When fish were not located at a specific measurement point, the vector was interpolated linearly. Finally, difference angles were binned at 10-cm increments from the source and circular statistics used to calculate the mean angle, standard deviation, and vector strength. Vector strength was used as a measure of the directional tendency of fish to move in a specific direction relative to either the source or particle acceleration axes (i.e. a value of 0 indicates difference angles were uniformly distributed, while values close to 1 indicates a concentration in one direction). The Rayleigh test was used on each group of binned difference angles to test whether they differed from random (P<0.05). Bearing to the speaker was used to compare swimming trajectories of fish when the sound was off (control) and then while it was on (treatment). This type of comparison could not be made using particle motion as no sound was played during controls. Sound pressure level at the fish’s location was also calculated along each individual swimming track and binned with the mean value and standard deviation calculated to determine if sound pressure might act as a threshold for behavior change. 3. People involved with sample collection, processing, analysis and/or submission: Data collected by: Daniel P. Zielinski, Reid Swanson, Daniel Krause, and Clark Dennis III Data processing by: Daniel Zielinski Data analysis by: Daniel P. Zielinski Data submission by: Daniel Zielinski