Data from: Evasive maneuvers adopted by mosquitoes reveal plasticity in the echolocation of Myotis riparius (Vespertilionidae)
Data files
Dec 05, 2024 version files 20.93 KB
-
R_script_manuscript_ID_ETH_24_0104.R
12.81 KB
-
README.md
6.43 KB
-
t_Test_approach_phase.txt
560 B
-
T_test_Search_calls.txt
535 B
-
T_test_Terminal_phase.txt
592 B
Abstract
Evasive maneuvers are essential behaviors to escape predator attacks. Several studies describe the advantages of evasive maneuvers when adopted by eared moths, under the risk of attack by bats. However, little is known about this behavior when expressed by dipterans in natural environments. This study describes, for the first time, two distinct behaviors adopted by mosquitoes in flight during attacks by insectivorous bats in riverine environments. Antipredatory behaviors differ greatly among themselves and, consequently, in their success in avoiding predation. In both cases, bats seek to adjust the parameters of their calls to maximize their chances of capturing mosquitoes. The results indicate that mosquitoes are capable of detecting the approach of bats and reacting appropriately, however success in escaping is associated with the type of maneuver and of the microhabitat where the interaction occurs.
README: Breviglieri C. P. B. 2024. Evasive maneuvers adopted by mosquitoes reveal plasticity in the echolocation of Myotis riparius (Vespertilionidae)
https://doi.org/10.5061/dryad.73n5tb35d
Data referring to the average values obtained for each of the attacks recorded by the recorder.
Description of the data and file structure
The files provided include three data tables, with the acoustic parameters obtained during the search phase (T_test_Search_calls), approach (t_Test_approach_phase) and terminal phase (T_test_Terminal_phase) and the script created (R script_manuscript ID ETH_24_0104).
The file "T_test_Search_calls" refers to the acoustic parameters emitted by the bats during the search phase (foraging).
The file "t_Test_approach_phase" refers to the acoustic parameters emitted by the bats during the approach phase, that is, when they detected the prey.
The file "T_test_Terminal_phase" refers to the acoustic parameters emitted by the bats during the terminal phase or as it is also called "Buzz", that is, when the bat attacks the prey.
For all cases, the abbreviation of the parameter names is:
Mosquitoes: Species of mosquitoes studied (Genus)
Fmax: Maximum frequency
Fmin: Minimum frequency
PF: Frequency peak
Durat: Duration of calls
Bandw. Bandwidth
Pulses: Interval between calls
Number: Number of pulses analyzed
The file "R script_manuscript ID ETH_24_0104" presents all the codes used for the manuscript analyses (Script).
######################################################################
To run the analyses, follow the instructions below:
First, load the packages:
library(ggplot2)
library(ggpubr)
###############################################################
To open the data sheet, run the command:
obj=read.table(file.choose(), header=TRUE)
#############################################################################
choose the first sheet, referring to the call parameters search:
T_test_Search_calls
##########################################################
Check if the table was loaded correctly:
obj
Example of table:
> obj
mosquitoes Fmax Fmin PF Durat Bandw Pulse
1 culex 56.76 42.69 46.50 5.08 7.00 80.00
2 culex 58.03 43.72 46.50 5.08 6.50 79.80
3 culex 57.52 42.44 46.50 4.40 7.50 79.60
4 culex 57.52 43.20 46.00 5.08 6.00 80.00
5 culex 55.52 42.69 46.00 4.77 7.50 80.00
6 culex 57.53 43.30 46.30 5.10 7.50 79.90
7 aedes 56.73 43.63 46.56 5.09 7.00 80.02
8 aedes 57.55 43.66 46.48 5.12 6.50 79.80
9 aedes 58.01 43.55 46.55 4.62 7.43 79.60
10 aedes 57.52 43.30 46.39 5.03 6.80 80.00
11 aedes 55.52 43.22 46.50 4.97 7.50 80.00
12 aedes 57.51 43.41 46.44 5.04 7.50 79.90
###############################################
Remember that the abbreviations refer to the following acoustic parameters:
Mosquitoes: Species of mosquitoes studied (Genus)
Fmax: Maximum frequency
Fmin: Minimum frequency
PF: Frequency peak
Durat: Duration of calls
Bandw. Bandwidth
Pulses: Interval between calls
Number: Number of pulses analyzed
#########################################################
Then, start the analysis for each of the acoustic parameters. ## Fmax = maximum frequency
t.test(Fmax ~ mosquitoes, paired = TRUE, data = obj)
Example of the first result:
Result:
Paired t-test
data: Fmax by mosquitoes
t = -0.053176, df = 5, p-value = 0.9597
alternative hypothesis: true mean difference is not equal to 0
95 percent confidence interval:
-0.3289420 0.3156086
sample estimates:
mean difference
-0.006666667
In this case, there were no differences, so the boxplot was not developed.
######################################################
Now, here is an example of the analysis regarding the maximum frequency (kHz) emitted during the approach phase. Differences were detected in this and, therefore, the boxplot was created.
To open the data sheet, run the command:
obj=read.table(file.choose(), header=TRUE)
############################################################
choose the sheet, referring to the search call parameters:
t_Test_approach_phase
############################################################
Check if the table was loaded correctly:
obj
Example of table:
> obj
mosquitoes Fmax Fmin PF Durat Bandw Pulse Number
1 culex 69.27 39.22 48.50 1.65 7.71 9.99 10
2 culex 69.21 38.71 48.50 1.70 7.75 10.02 9
3 culex 69.27 39.04 49.00 1.65 7.74 9.26 10
4 culex 68.93 38.72 48.82 1.63 7.71 9.38 9
5 culex 69.28 38.72 48.89 1.61 7.71 9.89 9
6 culex 69.27 38.23 48.90 1.63 7.75 10.05 10
7 aedes 75.80 39.22 51.50 1.51 5.91 16.05 4
8 aedes 75.21 40.21 51.50 1.82 6.00 15.26 4
9 aedes 76.17 39.24 51.00 1.51 5.93 16.00 4
10 aedes 74.71 39.72 50.82 1.50 6.11 16.04 4
11 aedes 75.28 40.19 51.89 1.51 6.09 16.05 4
12 aedes 75.17 40.23 50.90 1.50 5.94 15.05 4
#########################################################
Then, start the analysis for each of the acoustic parameters. > t.test(Fmax ~ mosquitoes, paired = TRUE, data = obj)
Example of the first result:
Paired t-test
date: Fmax by mosquitoes
t = 34.868, df = 5, p-value = 3.651e-07 (difference detected)
alternative hypothesis: true mean difference is not equal to 0
95 percent confidence interval:
5.729021 6.640979
sample estimates:
mean difference
6,185
# Boxplot
Para criar o boxplot referente a este parametro seguir o comando abaixo:
ggplot(data = obj, aes(x= mosquitoes, y= Fmax, color = mosquitoes)) +
labs(x = "Species of mosquitoes", y = "Fmax - maximum frequency (kHz)") +
scale_y_continuous(limits = c(67.0, 78.0)) +
geom_boxplot(fill=c("coral", "coral"), color="black",
outlier.shape = NA, alpha = 0.5) +
geom_jitter(shape = 16, position = position_jitter(0.1), cex = 20, alpha = 1) +
scale_color_manual(values = c("black", "black")) +
theme_bw(base_size = 20) +
theme(axis.text.y = element_text(size = 20)) +
theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank()) +
theme(legend.position = "none")
##############################################################
Therefore, to redo the analyses and boxplots, follow the available script. Don't forget to open the correct spreadsheets for each of the echolocation phases (search phase -> T_test_Search_calls, approach -> t_Test_approach_phase and terminal phase ->T_test_Terminal_phase)
Methods
To test the first hypothesis, was used a chi-square test to assess differences in the observed and expected number of successful bat attacks, considering the manifestation of each antipredator behaviors (different prey species) as the null hypothesis. To test the second hypothesis, I used Student's t test for paired samples was used to assess whether there were differences in the following acoustic parameters; (i) Fmax - maximum frequency (kHz), (ii) Fmin - minimum frequency (kHz), (iii) PF - peak frequency (kHz), (iv) Bandwidth (kHz), (v) duration of the pulses (ms) and (vi) interval between pulses (ms). These comparisons was being individually for the search, approach and feeding buzz phases emitted by bats during attacks. The paired t-test is the most appropriate analysis because the ultrasonic recordings recorded during the bats attack on each of the two mosquito species were collected at the same sites with standardized sampling effort, making the samples non-independent.
The time parameters of the oscillogram were measured while the frequency parameters were obtained from the power spectrum (Ratcliffe & Jakobsen, 2018). Acoustic data were analyzed using Raven 1.5 software (Charif et al., 2008). Calls were processed with automatic element separation using -20 dB maximum amplitude. The pulses were plotted simultaneously on the spectrogram, showing the relationship between frequency, time and intensity. Spectrograms were constructed from Fast Fourier Transforms size at 512, contrast at 68 and brightness at 63. All spectral parameters are given as mean ± 1 SD.
To test the first hypothesis, was used a chi-square test to assess differences in the observed and expected number of successful bat attacks, considering the manifestation of each antipredator behaviors (different prey species) as the null hypothesis. To test the second hypothesis, I used Student's t test for paired samples was used to assess whether there were differences in the following acoustic parameters; (i) Fmax - maximum frequency (kHz), (ii) Fmin - minimum frequency (kHz), (iii) PF - peak frequency (kHz), (iv) Bandwidth (kHz), (v) duration of the pulses (ms) and (vi) interval between pulses (ms). These comparisons was being individually for the search, approach and feeding buzz phases emitted by bats during attacks. The paired t-test is the most appropriate analysis because the ultrasonic recordings recorded during the bats attack on each of the two mosquito species were collected at the same sites with standardized sampling effort, making the samples non-independent.
The time parameters of the oscillogram were measured while the frequency parameters were obtained from the power spectrum (Ratcliffe & Jakobsen, 2018). Acoustic data were analyzed using Raven 1.5 software (Charif et al., 2008). Calls were processed with automatic element separation using -20 dB maximum amplitude. The pulses were plotted simultaneously on the spectrogram, showing the relationship between frequency, time and intensity. Spectrograms were constructed from Fast Fourier Transforms size at 512, contrast at 68 and brightness at 63. All spectral parameters are given as mean ± 1 SD.