Module 3 Worksheet - Analysis of bird song

docx

School

University Of Arizona *

*We aren’t endorsed by this school

Course

487L

Subject

Electrical Engineering

Date

Feb 20, 2024

Type

docx

Pages

Uploaded by CountAtomGiraffe8590

ECOL 487L / Module 3 – Analysis of Bird Song Module 3: Analysis of Bird Song Note : This lab will require you to use a laptop computer or equivalent, with Microsoft Excel and Raven lite 2.0 ( which can be used for free ; http://ravensoundsoftware.com/raven- pricing/#raven-lite-license ) installed. The lab exercise can be completed entirely on your computer. *** This module can be written up as a lab report. If you do it as a lab report, you don’t need to submit a worksheet. If you do not do it as a lab report, then submit the worksheet as usual. *** Introduction This module is intended to give you experience with the analysis of acoustic signals in animals. We will analyze the songs of the white-crowned sparrow, a bird that is common along the coast of the Western United States and whose song development and ecology has been studied extensively. Our exercise revolves around a very interesting study that reported that features of white-crowned sparrow song changed with the onset of COVID-19 and the associated lockdown (the publication based on the study is available in this module on D2L). The lockdown reduced levels of anthropogenic noise (for example, car traffic) and the birds responded to that by changing their songs. In this module, we will analyze songs of the white-crowned sparrow before and during lockdown. We cannot hope to achieve the level of resolution that authors did, but you will learn how to analyze sound signals while also learning about the details of this study. A. Trade-offs in producing a song Today, we will analyze the spectrograms of white-crowned sparrow songs. A white-crowned sparrow is a model species that has been studied to understand the song evolution in birds. Only male white crowned sparrows sing songs and their songs act as a signal to attract females or defend their territories from other males. The song is composed of multiple different types of syllables, and specifically, the ‘trill’ at the end of a song has been thought to convey information about the condition of the signaler to the receiver (Figure 1). Figure 1. A spectrogram of white crowned sparrows’ song, magnifying the trill that indicates signaler condition. For example, two acoustic variables are important to assess the quality of signalers: frequency bandwidth (the distance between the highest and lowest frequencies of a trill; FBW), and

ECOL 487L / Module 3 – Analysis of Bird Song trill rate (the number of repeated notes per second). Songs with a wide FBW or a high trill rate considered to be of high quality. However, because of the mechanical constraints, a wide FBW and a high trill rate cannot be achieved at the same time. Birds can produce a FBW that ranges from narrow to wide at a lower trill rate. However, they can only produce narrow FBW, at a higher trill late, but not high FBW. As a result, the distribution of songs will take on a triangular shape as shown in Figure 2, when FBW is plotted as a function of trill rate. The performance limit line represents when the quality of the song is maximized; any deviation below this line is considered to be a decrease in quality. B. Adapting to urban environment Animals produce communicative signals in diverse activities including breeding, parental care, and predation avoidance. The structure of the signal, for instance, the wavelength of light or sound used, affects the signal’s effectiveness. The environment through which the signal travels influences how the receiver perceives the signal and may degrade the information contained in the signal (rev. Endler 2000). The acoustic adaptation hypothesis states that animals adapt signals to transmit efficiently in their local environment (Morton 1975). Anthropogenic noise can be found everywhere in the urban environment, which may hinder the transmission and reception of signals between animals. For example, the highest intensity of vehicle traffic noise is at the lower frequency range (less than 2kHz), which overlaps and potentially masks the low frequency components of bird songs. Consistent with the acoustic adaptation hypothesis, many studies have shown that urban birds sing at a higher pitch than rural birds to compensate for anthropogenic noise pollution. However, increasing song's minimum frequency, and as a result, decreasing the frequency bandwidth, may come at the cost of reducing the overall quality of the song. Even in urban populations, female birds still prefer the low-pitched song of male birds as their mate, indicating that there might be a trade-off between being heard and, once heard, being chosen as a mate. Similarly, male white crowned sparrows respond more to lower-pitch songs of their conspecifics in playback experiments, again implying a trade-off between song salience and transmission efficiency. If that is the case, what will happen to urban sparrows if noise pollution is reduced? Will they be able to begin singing in a low-pitched voice to improve their song quality? In 2020, COVID-19 provided researchers with an unexpected opportunity to explore this question. On March 17 th , San Francisco issued public health order requiring residents to stay at home with only a few exceptions. As a result, noise pollution of the city decreased dramatically, potentially affecting the transmission of bird songs. The songs of white crowned sparrows in 2016 (before lockdown) and 2020 (during lockdown) were compared, and researchers discovered that male sparrows did respond to the change in background noise. Male sparrows sing in a lower frequency during the lockdown, similar to songs from 1970, when the urban environment was much quieter (Derryberry et al. 2020). Today, we will replicate this experiment by comparing sound files of white crowned sparrows' songs found in the Macaulay Library ( https://www.macaulaylibrary.org ) that were recorded in the San Francisco area in 2019 and 2020. Figure . A triangular distribution of songs

ECOL 487L / Module 3 – Analysis of Bird Song References Derryberry, E. P., Phillips, J. N., Derryberry, G. E., Blum, M. J., & Luther, D. 2020. Singing in a silent spring: Birds respond to a half-century soundscape reversion during the COVID-19 shutdown. Science 370: 575-579. Endler, J. A. 2000. Evolutionary implications of the interaction between animal signals and the environment. In: Espmark Y, Amundsen T, Rosenqvist G, Eds, Animal Signals: Signalling and Signal Design in Animal Communication , p. 11-46. Tapir Academic Press, Trondheim, Norway. Morton, E. S. 1975. Ecological sources of selection on avian sound. American Naturalist 109: 17-34. Podos, J. 1997. A performance constraint on the evolution of trilled vocalizations in a songbird family (Passeriformes: Emberizidae). Evolution 51: 537-551. C. Analysis of Song traits: Using a t-test Prior to lab, please read through the description of the Student’s t-test and paired t-test at the following URLs. Focus your attention on the material from beginning up through ‘How to do the test’. Student’s t-test: http://www.biostathandbook.com/twosamplettest.html Paired t-test: http://www.biostathandbook.com/pairedttest.html Based on your reading, answer the following questions: Question 1: What do we use the Student’s t-test for? What kind of data do they require? This is used when there is two samples, a measurable one and a nominal one. The nominal one has only two values. This test sees if the means of the measurement variable are different in the two groups or not. You will need normally distributed observations within each group or you will have higher false positives and inaccurate results. Question 2: Devise two hypotheses about the song frequency in white crowned sparrow before and during the lockdown, one being the alternative hypothesis and one being the null hypothesis. ALT: The white crowned sparrow will adapt to environmental conditions and it’s song will adjust. NULL: The white crowned sparrow will sing the same song regardless of environmental conditions D. Data Collection and Statistical Analyses

Your preview ends here

Eager to read complete document? Join bartleby learn and gain access to the full version

Access to all documents
Unlimited textbook solutions
24/7 expert homework help

ECOL 487L / Module 3 – Analysis of Bird Song 1. Open Raven Lite and open a sound file of white crowned sparrows’ songs (you can download these files from D2L). 2. When you open a sound file, you will see two graphs: a waveform graph (top) and a spectrogram graph (bottom). Uncheck the box for ‘waveform 1’ on the left to remove the waveform graph. 3. Adjust color schemes and the size of the spectrogram graph by clicking ‘change color map’, ‘zoom in X’ and ‘zoom in Y’ buttons on the toolbar. 4. Click ‘create selection mode’ and draw a square selection around the simple trill of the song. 5. Copy cells from the ‘selection table’ and paste it into an Excel spreadsheet. 6. Subtract low frequency from high frequency to get the frequency bandwidth. Divide the number of trill notes in the selection box by the difference between start and end times to get the trill rate. 7: Run a t-test on the low frequency of songs for ‘before covid’ and ‘during covid’ by entering the following formula into an empty cell: =T.TEST(array 1, array 2, tails, type) in which array 1, array 2, tails , and type are dependent on your data and the type of analysis you are using. Array 1 and array 2 : the two columns that include your dependent variables (i.e. measured data) Tails : ‘1’ to determine how the data differs in one direction, lower or higher; ‘2’ to determine how the data differs in both directions, lower and higher (more common in biological studies) Type : ‘1’ for a paired t-test when the measured data of the two treatment groups are dependent of each other (i.e. measured in the same individual) Example : a coyote’s level of activity during the morning versus during the afternoon; an individual beetle’s running speed in the presence and absence of a predator ‘2’ for an unpaired t-test when the measured data of the two treatment groups are independent of each other (i.e. measured in different individuals) and have equal variance Example : the effect of neurotransmitter treatment on rate of learning when the neurotransmitter is administered to one group but not a control group; the effect of diet on butterfly egg maturation in which individuals are either in a fed group or starved group Think about 1) how many tails and 2) which type of t-test is most appropriate for this particular analysis? You may additionally choose to do other analyses, such as whether the birds differ in frequency bandwidth of their songs for ‘before covid’ versus ‘after covid’, or whether trill rates of songs differ. Your instructors can help you with such analyses.

ECOL 487L / Module 3 – Analysis of Bird Song E. Interpretation of Statistical Results Question 3: Plot each data point in the graph of frequency band width and trill rate. Thinking back to the graphs shown in the presentation, is your graph consistent with a tradeoff between frequency bandwidth and trill rate? 26A pri l201 9 2May2 019 6May2 019 25Ma y2019 2June2019 28J une2019 18Ma rch20 20 19Ma rch20 20 29A pri l202 0 2May2 020 9May2 020 10Ma y2020 0 1000 2000 3000 4000 5000 6000 Freq Bandwidth 26A pri l201 9 2May2 019 6May2 019 25Ma y2019 2June2019 28J une2019 18Ma rch20 20 19Ma rch20 20 29A pri l202 0 2May2 020 9May2 020 10Ma y2020 -35 -30 -25 -20 -15 -10 -5 0 5 Trill Rate My graph does look opposite of each other. Question 4: What is the p-value of your t-test on the low frequency of songs (as well as other variables you chose to analyze)?

ECOL 487L / Module 3 – Analysis of Bird Song I spent 4 hours trying to make absolute sure my data was accurate and I’m still not sure I did it correctly but here are the settings I used: T-Score: 0.920896809 DF: 12 Significance Level: .05 Two Tailed Hypothesis The p -value is .375295. The result is not significant at p < .05. Question 5: Interpret the p-value. Can you reject one of your hypotheses? Explain why with respect to your prediction and actual results. With that being said, my values are not what I expected to see. My results are showing that the birds did not change their song due to environmental conditions because of the COVID lockdown. I will reject the alternative hypothesis after my experiment findings. Question 6: Our data categorized songs recorded on March 18 th and 19 th of 2020 as ‘during COVID’ songs. Considering the lockdown began on March 17 th , do you think it is appropriate to analyze those songs in the same category as songs recorded in April or May? Explain your reasoning. Adaptation takes time and birds will not have had that time to change their songs within such a short period of time. Question 7: Why are we comparing songs from the same month range in different years (Why not songs from different month ranges in the same years? e.g., January in 2020 vs April in 2020)? What can we avoid by doing this? So we can see on an annual basis of the progression of adaptation. We avoid not having a varying testable range for our data.

Your preview ends here

Eager to read complete document? Join bartleby learn and gain access to the full version