Login processing...

Trial ends in Request Full Access Tell Your Colleague About Jove


A Laboratory Method to Measure Contagious Yawning in Rats

Published: June 14, 2019 doi: 10.3791/59289


The method described here aims to obtain yawn contagion curves in pairs of familiar or unfamiliar male rats. Cages with holes separated by either clear or opaque partitions with (or without) holes are used to detect whether visual, olfactory, or both types of sensory cues can stimulate yawn contagion.


Communication is an essential aspect of animal social life. Animals may influence one another and come together in schools, flocks, and herds. Communication is also the way sexes interact during courtship and how rivals settle disputes without fighting. However, there are some behavioral patterns for which it is difficult to test the existence of a communicatory function, because several types of sensory modalities are likely involved. For example, contagious yawning is a communicatory act in mammals that potentially occurs through sight, hearing, smell, or a combination of these senses depending on whether the animals are familiar to one another. Therefore, to test hypotheses about the possible communicatory role of such behaviors, a suitable method is necessary to identify the participating sensory modalities.

The method proposed here aims to obtain yawn contagion curves for familiar and unfamiliar rats and evaluate the relative participation of visual and olfactory sensory modalities. The method uses inexpensive materials, and with some minor changes, it can also be used with other rodent species such as mice. Overall, the method involves the substitution of clear dividers (with or without holes) with opaque dividers (with or without holes) that either allow or prevent communication between rats placed in adjacent cages with holes in adjoining sides. Accordingly, four conditions can be tested: olfactory communication, visual communication, both visual and olfactory communication, and neither visual nor olfactory communication. As social interaction occurs between the rats, these test conditions simulate what may occur in a natural environment. In this respect, the method proposed here is more effective than traditional methods that rely on video presentations whose biological validity can raise concerns. Nonetheless, it does not discriminate between the potential role of hearing and roles of smell and vision in yawn contagion.


Traditionally, communicatory behavior has been studied from two perspectives. From one perspective, ethologists observe and record the behavior of animals in natural settings and attempt to recognize its adaptive value1. The particular sense or senses involved have not been the primary interest of these studies. From another perspective, physiologists are more interested in unraveling the mechanisms by which animals communicate1; hence, laboratory studies have provided methods to address the role that sensory modalities play in communication2,3. These two perspectives are indeed complementary, because knowledge of both adaptive value and immediate mechanisms is necessary to gain a comprehensive understanding of communicatory behaviors in the social life of animals.

Yawning behavior is a conspicuous component of the behavioral repertoire in several species of vertebrates4, ranging from fish to primates5. It can be described as a slow opening of the mouth and maintenance of its open position, followed by a more rapid closure of the mouth5. The duration of the whole sequence depends on the species; for example, primates yawn for longer durations than non-primate species6. In many species, with humans being the exception, males tend to yawn more frequently than females7. This feature might underpin the possible communicatory function of yawning, although regular patterns of yawning and its daily frequency may also suggest a physiological function. In rats, spontaneous yawning follows a circadian rhythm, with peaks of high frequency occurring in the morning and afternoon8,9.

One interesting feature of yawning behavior is that it can be a contagious act (when the releasing stimulus of a behavior happens to be another animal behaving in the same way10) in several species of vertebrates11,12,13,14,15,16, including birds17 and rodents18. Furthermore, recent evidence has indicated that contagious yawning may reflect a communicatory role, because the yawning of one rat can affect the physiological state of another when exposed to olfactory cues19. However, whether or not yawning has a communicatory role is still under debate20,21, and analyzing contagious yawning is an essential first step to solve this issue.

On the other hand, contagious yawning has been linked to an animal's ability to empathize with the perspectives of other animals; hence, closely related individuals are more likely to show contagion4. This hypothesis has been frequently tested in laboratory conditions in which animals are presented with yawn stimuli on video12,13; hence, contagion can only occur through visual cues. Other investigations have assessed yawn contagion in more natural conditions using groups of animals14,15. A major problem of this is that socially interacting animals often respond to cues and exchange signals that are conveyed through combinations of sensory modalities. Disentangling the actual senses involved in a given behavior from their combined effects is not always an easy task. Typically, researchers pharmacologically or surgically hinder an animal's use of a given sense, then infer the role of that sense in the relevant behavior2,3,18,22. Fortunately, there are other methods in which only physical barriers are used to either allow or impede communication between animals23,24,25, thereby achieving greater biological validity.

The method proposed here has been specifically designed to study contagious yawning in familiar and unfamiliar rats in a social setting. According to the empathetic hypothesis, the former group should be more susceptible to contagious yawning. The method does not require the animals to be surgically or pharmacologically deprived of any senses. Instead, it works by placing the rats in adjacent cages with holes and physically obstructing their communication using either clear or opaque dividers with or without holes. Thus, four test conditions can be examined: (1) olfactory communication (OC, perforated opaque divider), (2) visual communication (VC, nonperforated clear divider), (3) visual and olfactory communication (VOC, perforated clear divider), and (4) neither visual nor olfactory communication (NVOC, nonperforated opaque divider). Therefore, researchers can compare the relative contributions of olfactory, visual, and to some extent, auditory cues in yawn contagion. This approach is not new, as similar methods have been used to isolate the senses involved in certain communicatory behaviors in animals such as lizards23 and mice26. In fact, Gallup and colleagues27 have used a similar method to demonstrate the role of visual cues in contagious yawning in budgerigars. The main features of these methods are simulation of a social context and the minimal stress inflicted on the animals. Furthermore, the use of interacting animals increases the biological validity of the conclusions.

There are several ways to measure contagious yawning25,28. Dr. Stephen E. G. Lea (personal communication, 2015) helped us numerically adapt a method previously employed by primatologists13,14 for an earlier analysis of the data used here18. Presented in this protocol is an enhanced version of this method with a wider range of applications. It consists of weighting the total number of a rat's yawns, within and outside of a given time window, by the proportion of observation time corresponding to the yawns within and outside the time window.

For example, if it is assumed that rats A and B are observed for 12 min, their yawning is recorded to the nearest minute, and a 3 min time window is set to measure contagious yawning. Next, the following sequences of yawns for each of those rats are considered: rat A (0,0,0,1,0,0,2,0,0,0,2,1) and rat B (0,1,1,0,1,1,0,0,0,0,0,3). It should be noted that each number (0-3) corresponds to the number of yawns scored at each min. For rat A, during minutes 1, 10, and 11 (numbers in bold type), rat B does not yawn within the preceding 3 min (the chosen time window) or within that minute. In those minutes, rat A yawns a total of 2 times. Therefore, the yawn rate of rat A without any yawn stimulus (non-post-yawn yawn rate) is 2/3 (i.e., 0.67 yawns/min). In the remaining 9 min, rat B yawns at least one time in either the same minute or the 3 previous minutes. Rat A yawns a total of four times in those 9 min. Therefore, the yawn rate of rat A in response to a yawn stimulus (post-yawn yawn rate) is 4/9 (i.e., 0.44 yawns/min). The application of the same procedure to rat B yields a non-post-yawn yawn rate of 2/3 (i.e., 0.66) and post-yawn yawn rate of 5/9 (0.55).

On the other hand, if yawning is recorded to the nearest decimal of a minute, yawn contagion will result in an adjusted post-yawn time. For example, if the following yawn times are recorded over a 12 min observation period for rats A and B: rat A (2.3, 5.1, 5.8, 10.4, 10.8, 11.1) and rat B (1.2, 2.4, 4.5, 5.1, 11.2, 11.6, 11.8). For rat A, the time periods over which rat B does not yawn within the past 3 min range from 0 to 1.2 min and from 8.1 to 11.2 min (i.e., 3.1 min), which yields a total of 4.3 min of non-post-yawn time. The number of times that rat A yawns during those times is three (numbers in bold type), so the non-post-yawn yawn rate is 3/4.3 (i.e., 0.69), while the post-yawn yawn rate is 3/7.7 (i.e., 0.38; the denominator from 12-4.3 min). Similarly, for rat B, the time periods over which rat A does not yawn within the past 3 min range from 0 to 2.3 min and from 8.8 to 10.4 min, which yields a total of 3.9 min. The number of times rat B yawns within those periods is one, so the non-post-yawn yawn rate is 1/3.9 (i.e., 0.25). Accordingly, the post-yawn yawn rate is 6/8.1 (i.e., 0.74).

While a near-contemporaneous match in behavior is an ideal criterion to demonstrate the presence of a contagion, aspects such as the constraints on what an individual attends to, time of reaction to a stimulus, distribution of the behavior over time (e.g., yawning may occur in episodes), and time to acclimatize to the experimental setting all give rise to species differences, making it difficult to use a unique time window. This may be the reason why researchers have used time windows that vary from seconds5 to several minutes11, which creates problems when comparing results28. Because of this, it is proposed to repeat the procedure described above for a range of time windows to obtain yawn contagion curves and compare the yawn contagion curves between species.

Equivalent yawn contagion curves can be compared by randomly distributing the number of yawns observed for each rat over the observation period. Thus, the proposed method to measure yawn contagion offers two types of controls: the (1) yawn rate occurring outside of the time window (non-post-yawn time) and (2) artificial yawn contagion curve obtained from the random distribution of the number of yawns. Therefore, this approach to analyze yawn contagion is a step forward from other procedures, such as those comparing the percentage or frequency of yawning within a single time window to that occurring outside this window25, without taking into account the actual times frames. The method is complemented by an R-based program29 to conveniently and objectively calculate the probability of contagious yawning for one or more time windows.

To illustrate the usefulness of this method and advantages of the R-based program, a data set from a previously published study18 is used. The experimental condition consisted of 144 male rats allocated to either a familiar or unfamiliar condition. The rats in each experimental condition were subdivided into four subgroups of nine pairs and exposed to any of the four test situations described above. The yawning behaviors of the rats in each experimental condition and test situation were then recorded over a period of 60 min.

Subscription Required. Please recommend JoVE to your librarian.


The experimental protocols and animal husbandry were conducted in accordance with institutional guidelines.

1. Materials

  1. Find a complete list of the materials used to implement the method in the Table of Materials. Use Figure 1 and seek expert advice to construct an inverted T-shaped table, observation cages, and cage dividers. Follow the safety indications for the use of sharp tools and potentially dangerous materials.
  2. Make an inverted T-shaped table by gluing two rail-like wooden bars (45 cm length, separated 0.6 cm from each other) onto the top and in the middle of a piece of wood (100 cm x 45 cm x 1.5 cm). Then, place a second piece of wood (50 cm x 45 cm x 0.6 cm) vertically between the rail-like bars. Ensure that this second piece of wood prevents the pair of rats on one side from seeing the other pair on the opposite side (Figure 1).
  3. Use glass and acrylic to create eight observation cages. Ensure that each cage (19 cm wide, 19 cm long, 10 cm high, 3 mm thick) has 24 equidistant holes (5 mm diameter) arranged in three rows in the middle of one of its sides. Make this side of acrylic (3 mm thick) and shorten the height of the opposite side by 0.7 cm to allow the rat to breathe.
    1. Make sure each observation cage has a sliding lid made of acrylic to prevent the rat from rearing and distracting itself during the 60 min observation period.
  4. Make four dividers out of acrylic (19 cm wide, 30 cm high, 3 mm thick). Drill 24 holes (5 mm diameter) that match those in the observation cages, each in one clear divider and one opaque divider.
  5. Prepare data sheets in advance to record the occurrence of yawns. Include, in the heading of each data sheet, the name of the observer, experimental condition (i.e., familiar or unfamiliar rats), relevant test situation (OC, VC, VOC, NVOC), the date, and initial and final times of the observation. Divide the rest of the data sheet into two columns, each with the number of the rat written in the upper part, to record the yawning behavior of the rats.

2. Procedure

  1. House 144 male rats in groups of four in plastic cages (home cages) from weaning until they reach 2.5 months in age. Do not use female rats, because they tend to show greater variation in behavior due to hormonal cycles, which may be confused with the effects of the test situation. Make sure that the rats in each cage are not siblings.
    NOTE: It might be difficult to obtain all the rats together. In that case, create blocks of at least 8 familiar rats and 8 unfamiliar rats so that every test situation is represented once in each experimental condition30.
  2. To identify the rats, mark them with symbolic numbers on the tails using a commercial marker. For example, represent the numbers 1 to 4 by combining dots and lines (e.g. ,one dot for number 1 and one line for number 4). Identify familiar and unfamiliar rats using different colors.
    NOTE: Handle the rats following safety directions provided by animal facility staff, and comply with recommendations concerning the correct use of laboratory animals at the institution where experiments will be performed.
  3. Randomly choose which home cages will house the group of familiar rats and which will house the unfamiliar group. For example, suppose there are 16 available rats living in four home cages: number the home cages 1 to 4, then use R (download it at http://cran.r-project.org/) as follows [after the prompt (>)]:
    > sample(4,4)
    [1] 4 3 2 1
    1. Group the familiar (or unfamiliar) rats in home cages 4 and 3, and group the unfamiliar (or familiar) rats in home cages 2 and 1. Make sure the animal facility staff maintains the identity of each cage and handles the rats in the same manner as all other rats in the animal facility.
    2. Use the R program again and randomly choose rats to form each pair in each experimental condition. Ensure that the rats in each pair of familiar rats originate from the same home cage, and ensure that the opposite is true for the pairs in the unfamiliar group. Randomly choose pairs of rats for each test situation in each experimental condition.
  4. Perform two test sessions per day with two of the eight possible test situations (four for each familiar and unfamiliar rats) per test session (i.e., 60 min observation period). Conduct the two test sessions consecutively within 3 h.
    NOTE: Make sure to conduct all test sessions (four pairs of rats per day) either in the morning or afternoon to avoid any confounding factors. Run a complete replicate (eight test situations) of the experiment over two consecutive days.
    1. Make sure to use the test situations in a random sequence for each replicate of the experiment. For example, use numbers 1 to 8 to identify each test situation, then use R as follows:
      [1] ​8 7 4 6 5 1 2 3
    2. Allocate test situations 8 and 7 to test session 1, test situations 4 and 6 to test session 2, and so on. Then, randomly choose the side of the inverted T-shaped table where each pair of rats (test situation) will be placed.
  5. Repeat the same procedure (steps 2.3 to 2.4.2) for a second replicate (i.e., block) of the experiment.
    NOTE: If the aim of the study is to test the effects of a social condition (i.e., two interacting rats) on yawning frequency, use one rat placed in an observation cage next to an empty observation cage as a control for each of the four test situations. Perform this experiment 2x for each test situation for a control group of eight rats.
  6. Set up the test session by transferring the first four rats from the animal facility to the observation room, where they will remain for 15 min to acclimate to the novel environment. Transport the rats in individual cages and keep them separated from each other during transportation and in the observation room.
    NOTE: Follow the safety directions for transporting animals provided by the animal facility staff and use the indicated clothes to work with animals in the laboratory. The rats will not have access to food and water while they are in the observation room.
  7. After the acclimatization period has passed, place the inverted T-shaped table on a larger rectangular table. Make sure there is a ceiling lamp that sufficiently lights the room when making observations.
  8. Put filter paper on the bottom of each observation cage and place the cages in pairs on each side of the inverted T-shaped table. Position the corresponding divider in between each pair of cages.
    NOTE: The filter paper makes it easier to clean the cages and provides a rough surface on which the rats can move without slipping.
  9. Strategically place two digital camcorders such that each records the yawning behavior of each pair of rats. Make sure the camcorders are safely fixed to tripods and correctly orientated to the observation cages. Connect the camcorders to a desktop computer to simultaneously monitor the behavior of the rats.
    NOTE: Store the digital information on flash drives to permanently archive the experimental sessions. Use flash drives with a high storage capacity.
  10. Following the acclimatization period, place the rats in the observation cages according to the allocation previously determined. Set the automatic focus of the camcorder and start the video recording simultaneously with a stopwatch. Stop the video recording at the end of the 60 min observation period.
    NOTE: While the experimenters may be outside of the room during the test session and observing remotely, it is recommended that one person is in the observation room (as far away as possible from the experimental setting) while monitoring the behavior of the rats to ensure that the test session proceeds without interruption.
  11. When the observation is over, return the rats to their home cages in the animal facility. Ask the animal facility staff to ensure that the rats have access to food and water again. Thoroughly clean the observation cages using a nontoxic detergent and prepare the experimental set-up to perform the second and final test sessions of the day.
    NOTE: Clean the wood to remove any scents left by the rats when placed in the observation cages, which could affect the behavior of the next pair of rats tested.
  12. Train one to two volunteers partially blind to the assigned treatments to identify and record yawning using an all-occurrence sampling method. Make sure the observers use the definition of yawning described in the introduction section.
  13. Play back and project each video on a computer screen using any standard playback system. Ask the observer to view each video and observe and record yawning behavior using the previously prepared data sheets, then allow him/her to review the videos at slower speeds to enhance the ability to observe and measure yawning.
    1. Ask the observers to score yawning using a notation system similar to the following: use vertical lines with the minute written as a superscript to represent the occurrence of a yawn (e.g., |3 ||6 |9, indicating one yawn at minute 3, two yawns at minute 6, and one yawn at minute 9). Score yawning as a sequence (e.g.,1.2, 2.2, 3.2, 5.0, 5.8) to record it with greater precision (minutes rounded to one decimal place).
    2. Alternatively, directly input data from the videos into a computer using standard data collection programs. Make sure the observer is acquainted with such programs.
      NOTE: If necessary, allow the observer to view each video in more than one session. Ensure that one observer views all the videos corresponding to one block of experiments. If two different observers viewed these videos, there may be a risk that differences in their abilities to observe and record yawning will confound the effects of the test situations. It is important to score the intra-observer reliability between different observers on 10%-20% of the videos to ensure that yawning is annotated in the same way.

3. Data processing

  1. Transcribe the temporal sequence of yawns in each rat from the data sheets to a spreadsheet. Ensure there is one column for each rat with a short heading (e.g., fr1.l and fr1.r to indicate the first pair of familiar rats on the left and right sides, respectively). Ensure that all columns are the same length by filling the empty cells with either 0 or NA (see below).
    NOTE: Use the general guide provided as a supplementary document to fully understand the following steps to process the data, measure contagious yawning, and obtain contagious yawning curves.
    1. If yawning was recorded to the nearest minute, type in the number of yawns at each relevant minute and use 0 (no yawn) to fill in the cells when no yawn occurred in a given minute. Save the worksheet as a text file (.txt).
    2. If yawning was recorded to the nearest decimal of a minute, type in the sequence top down and use "NA" to fill in the empty spaces of columns (rats) in which the numbers of yawns (rows) is lower than that of the column with the maximum number (rows) of yawns recorded for a rat. Save the worksheet as a text file (.txt).
      NOTE: As R has the ability to work with empty cells, the worksheet can be saved with columns of different lengths. Use the help options provided by R to handle missing data.
  2. Initiate R. Import the data from the file in which they were previously located.
  3. Save the program codes with the extension ".R" (provided as supplementary material), then download the specific program code (see general guide) depending on whether yawns were recorded as integers or fractional numbers. Run the program for each pair of rats and desired number of time frames.
    1. Run the program again, now using a random distribution of the number of yawns from each rat (see general guide). Follow the same steps (steps 3.3 to 3.3.1) for all experimental conditions (i.e., familiar and unfamiliar rats) and/or all the test situations. Proceed as indicated in the general guide and export the results to Excel.
      NOTE: Instead of importing a previously saved program, as suggested above using a text editor, copy and directly paste the program into R's workspace.
  4. To create yawn contagion curves, use the data previously saved in an spreadsheet. Subtract the non-contagion rates from the contagion rates for each pair of rats, time window, and test situation. Separate the analyses of the observed data and artificial (randomly distributed) data.
    1. Next, calculate confidence intervals (CI) for each time window and test situation using a bootstrap procedure (see the general guide). Separate the analyses for familiar and unfamiliar rats. Then, combine the artificial data from the four test situations for each experimental condition and calculate the CI for each time window using a bootstrap procedure.
    2. Next, create plots for each test situation and experimental condition (see Figure 2 and Figure 3). Finally, perform a multiple regression analysis to compare the intensity of yawn contagion between test situations (see below).

Subscription Required. Please recommend JoVE to your librarian.

Representative Results

The rats were selected from a previously produced sub-line of Sprague-Dawley rats that were selected for frequent yawning (approximately 22 yawns per hour31). However, the nine pairs of unfamiliar and nine pairs of familiar male rats (between 2.5 and 3 months of age) used per test situation yawned approximately 12 times per hour, on average18. Therefore, the test situations to measure yawn contagion partially inhibited yawning behavior.

Yawn contagion was measured over a range of time windows that varied from 1 to 10 min. Figure 2 shows the mean difference in the yawn rates between unfamiliar male rats in contagion and non-contagion conditions for each time window and test situation. These yawn contagion curves indicate that only OC rats (in cages with a perforated, opaque divider) showed yawn contagion, which is evident from time window 4 and onwards, because the CI of the averages does not overlap with the CI of the randomly allocated numbers of yawns over the 60 min period. The band indicated, as expected, that a random allocation of the number of yawns in each rat in the observation period produced yawn rates that oscillated at approximately zero, without showing any apparent pattern.

The multiple linear regression model fitted to the data indicated that the curves from the four test situations were significantly different (F1,3 = 11.5, p < 0.0001). Specifically, yawn contagion was stronger in OC rats than: VC rats (rats in cages with a nonperforated, clear divider; t = -3.8, p < 0.001), VOC rats (rats in cages with a perforated, clear divider; t = -5.74, p < 0.0001), and NVOC rats (rats in cages with a nonperforated, opaque divider; t = -2.64, p < 0.01). In all cases, the degrees of freedom were 695, since the analysis took into account not only the four test situations, but also the 10 time windows. As these conditions generated autocorrelated measurements, an autocorrelation term was added (ARMA, autoregressive moving average) to the statistical model. The overall trends over the 10 time windows also differed statistically from a 0 slope (F1,1 = 11.99, p < 0.0001). Figure 3 shows the same analysis as that in Figure 2, applied to familiar male rats. In this case, none of the four test situations stimulated yawn contagion because their CI overlapped with the randomly generated CI band. There were no differences among the four test situations (F1,3 = p = 0.14); although, the overall increase in yawn contagion over the 10 time windows differed from a 0 slope (F1,1 = 9, p < 0.01).

The role of olfaction in the social life of mammals may be the reason why only the OC rats showed yawn contagion and yawned more frequently than the VC rats18. Visual cues may not have facilitated yawning, because albino rats do not see as well as non-albino rats. However, the VC rats yawned twice as much as the group of rats individually placed in an observation cage next to an empty cage18. This finding definitively supports the social nature of yawning and suggests that yawn contagion in rats is dependent on olfactory cues. However, the latter only supports the degree of familiarity between pairs of rats as being involved in yawn contagion, and auditory cues are the likely channel by which yawn contagion occurs18.

Overall, these are unexpected results according to the prediction that familiar rather than unfamiliar rats will show contagious yawning. Although the results presented do not use a positive control (we have no data in which contagious yawning is observed in closely related individuals), it is believed that this method is unbiased for contagious yawning in unfamiliar rats. Several elements support such a claim. First, familiar and unfamiliar rats did not differ in the average frequencies of yawning18; thus, the differences in yawn contagion detected here did not seem to depend on how frequently they yawned. Second, the fact that the same numbers of yawns per rat were randomly allocated to the observation period and that this did not produce yawn contagion in any group renders the finding of contagious yawning in unfamiliar rats more robust. Finally, a cross-correlation analysis (to detect contemporaneous matching of behavior) was previously applied18 to the same dataset used here, and the result coincided with what was found. Therefore, this method is sensitive enough to detect and measure changes in contagious yawning.

Alternatively, it may be questioned whether what was found here is contagious yawning. Such a concern is not exclusive to this study. There are two perspectives from which contagious yawning and its causality have aroused concerns. First, empathy does not seem to be the only factor underlying contagious yawning18, as species not expected to show empathy and/or mental state attribution skills such as sheep and rats can still show contagious yawning. Second, there is growing concern that what is called contagious yawning might not be as such. For many decades (Thorpe, 1956; cited previously10), contagious, stereotyped behavior patterns and contagious yawning in particular have been regarded as an expression of social facilitation. More recently, Kapitány and Nielsen25 suggested through the use of simulated data that what is a perceptual pattern-recognition error may be incorrectly called contagious yawning. Therefore, the problem may not lie in the way contagious yawning is being measured but rather in how it is defined.

In summary, the method proposed here is useful to detect yawn contagion. In addition, it is useful to discard some sensory cues, thereby reducing the number of sensory cues to only one or two. Further experiments to test specific predictions are necessary to distinguish the effects of one sense from the others.

Figure 1
Figure 1: Schematic of the observation set-up to measure contagious yawning. Four individual cages were arranged in pairs on each side of the inverted T-shaped table. In each pair, the cages were facing each other with acrylic dividers in between. The specific test situation determines whether the divider in between the cages has holes and whether it is clear or opaque. The camcorders were strategically placed to record any instance of yawning behavior. Please click here to view a larger version of this figure.

Figure 2
Figure 2: Yawn contagion curves for nine pairs of unfamiliar male rats exposed to (A) olfactory cues, (B) visual cues, (C) olfactory and visual cues, and (D) neither visual nor olfactory cues. Each circle represents the average difference in yawn rate with 95% confidence intervals (CI) from a rat that yawned in the same minute or in the minute (time window) preceding the yawn of the other rat and when it did not yawn. Yawn rate values above the dashed line indicate yawn contagion, whereas values below the dashed line indicate non-yawn contagion. The solid line joining the points at each time window is used to indicate the average difference in yawn rates at each time window when the number of yawns of each rat was randomly allocated to the 60 min observation period. The dark gray band indicates a 95% CI obtained by combining the random dataset from the four test situations at each time window (a continuous shade is used to facilitate the interpretation of the figure). One pair of rats exposed to olfactory cues was removed from the analysis because neither of the rats yawned. Please click here to view a larger version of this figure.

Figure 3
Figure 3: Yawn contagion curves for nine pairs of familiar male rats exposed to (A) olfactory cues, (B) visual cues, (C) olfactory and visual cues, and (D) neither visual nor olfactory cues. Each circle represents the average difference in yawn rate with 95% confidence intervals (CI) from a rat that yawned in the same minute or in the minute (time window) preceding the yawn of the other rat and when it did not yawn. Yawn rate values above the dashed line indicate yawn contagion, whereas values below the dashed line indicate non-yawn contagion. The solid line indicates the average difference in yawn rates at each time window when the number of yawns of each rat was randomly allocated to the 60 min observation period. The dark gray band indicates a 95% CI obtained by combining the random dataset from the four test situations at each time window. One pair of rats exposed to visual cues was removed from the analysis because one of the rats did not yawn. Please click here to view a larger version of this figure.

Subscription Required. Please recommend JoVE to your librarian.


There are critical steps in the method that should be taken into account to obtain successful results. Familiar rats must share home cages for at least 1.5 months after weaning and before running the experiments. However, unfamiliar rats must live in separate home cages. In both cases, the pairs of rats must come from different litters but be as similar in age as possible. Regarding the observation cages, their holes should match those in the dividers, because this is the only way to guarantee olfactory contact between the rats. The dividers, on the other hand, should be clear enough to ensure visual contact or opaque enough to ensure no visual contact. The wooden partition between one pair of rats and the other must be sufficient to prevent the rats on one side from seeing those on the other side. Another crucial aspect is the appropriate design of the experiment. Whenever there is a suspicion that a process may be biased, a random procedure should be implemented30.

The method should not present serious problems to users. Making the holes in the glass was the main technical problem faced, and because of that, acrylic was used instead. Nonetheless, glass may be used for making the entire observation cages, provided professional advice is obtained. It should be ensured that the edges of the holes are filed to avoid glass splinters that may injure the rats. However, modifying the main method is not recommended (e.g., making the holes larger). Additionally, use of a combined group of males and females may make it difficult to detect yawn contagion.

The dividers used here may be insufficient to prevent the rats from using auditory cues, because rats are able to produce and perceive sounds at frequencies at which the materials of the observation cages and dividers likely did not block. However, this situation itself makes it possible to infer that auditory cues caused yawn contagion18 and that olfactory cues only facilitated the recognition of the partner's degree of familiarity. Therefore, the method proposed here still provides reasonable evidence to identify the senses involved in contagious yawning and their intensities.

Earlier methods have been designed to study contagious yawning in laboratory conditions mainly by presenting videos to the experimental individuals12,13, but these were questionable approaches in terms of biological validity. The method presented here solves this concern by using socially interacting animals in conditions more similar to what occurs in the real world. Furthermore, it is possible to simultaneously explore the participation of several sensory modalities in a single experimental set-up. It is recognized that this method does not absolutely discriminate between the effects of auditory cues and other sensory cues. However, a well-designed further experiment may allow researchers to infer the most likely sensory modality involved18. One possible noninvasive solution is the use of white noise to mask sounds and eliminate auditory cues. Similarly, researchers may expose naïve rats to bedding from OC rats to determine the role of olfactory cues, which is a proven procedure in social facilitation studies32.

The use of this method can be extended to study yawn contagion in other species. For example, this set-up can be used after simple modifications with animals such as mice and hamsters to compare yawn contagion curves. Comparisons among different species may reveal unexpected patterns. The basic experimental plan can work with larger animals such as guinea pigs, cats, and rabbits. Likewise, the method can be used to study other potentially contagious behaviors such as grooming and scratching. The R-based program can reduce the time spent calculating yawn contagion for several time windows and can be used to measure yawn contagion in other vertebrate species, provided the user has previously collected relevant data.

In summary, the main advantages of this method are the leading to acquisition of yawn contagion curves and aiding in discrimination between the relative roles of sensory modalities involved. The acquisition of yawn contagion curves is, as far as we know, a novel approach that may be useful to measure the strength of a contagion and observe how this intensity varies among species. Accordingly, the method can also be used with some modifications in other animal species such as sheep16, wolves15, dogs33, snakes34, and fish5. In all these species except snakes, yawning has been previously documented. In fact, a method similar to the one presented here has been used successfully in budgerigars27. This method may also be used to study other types of behavior that are contagious. For example, behaviors such as emotional reactivity, grooming, and scratching in rodents may be contagious. In fact, the method proposed here showed contagious emotional reactivity in familiar rats18.

Subscription Required. Please recommend JoVE to your librarian.


The authors declare no conflicts of interest.


A. M. was partially funded by the Vicerrectoría de Docencia, Benemérita Universidad Autónoma de Puebla. We are especially indebted to the staff of the animal facility “Claude Bernard” for the use of the rats for the shoot. We thank anonymous referees for their comments on early versions of this manuscript. The presentation is less strident and more balanced because of their thoughtful comments


Name Company Catalog Number Comments
Acrylic dividers Handcrafted Not available Two dividers, one clear and one opaque, will have 24 holes each. The other two dividers, one clear and one opaque, will have no holes. See the main text for details of construction.
An R-based program Benemérita Universidad Autónoma de Puebla Not available This is the program used to assess yawn contagion in rats. See the main text for information about the way the program is used.
Data sheets The user can elaborate them Not available These forms will be used for the observer to record the frequency of yawning behaviour by viewing the video recordings. Alternatively, a notebook can be use provided you follow the suggestions given in the main text.
Desktop computer Any maker Not available Make sure the computer has a video card capable of conveniently processing the video recordings of yawning behaviour. 
Digital camcorders Any maker Not available They will be used to video record yawning behaviour of each pair of rats; there will be 2 pairs of rats per experimental session. 
Flash drive Any maker Not available Each experimental session will last 60 min, and so you will require sufficient memory to store the video recording.
Glass cages Handcrafted Not available Each cage (19 X 19 X 10 cm height) will have 24 holes (0.5 cm diameter) forming three rows in the middle of one of its sides. See the main text for more details about their construction. It is recommended to fabricate one extra cage in case one of them is accidentally broken. 
Markers Sharpie or any other maker Not available Permanent markers to number the rats. See the main text to see one way of using painting symbols on the rat's tail. 
Pencils Any maker Not available They are used by the observer to record the frequency of yawning. It is important that the observer has previously been trained to recognize yawning behaviour and operate the video player system. 
R software R Development Core Team Not available Download R at: http://cran.r-project.org/  
Rail-like wooden bars Handcrafted Not available They will be fixed in the middle of the rectangular wooden sheet  forming a track, where a second wooden sheet is placed. See the main text for additional instructions for construction.
Rectangular table Any maker Not available This is the table (approximately 2 x 1 m) where the inverted T-shaped table will be placed for performing the observation of yawning behaviour.
Sprague-Dawley male rats Any local supplier of laboratory animals Not available Nine pairs of male rats per test situation are necessary for each group, familiar and unfamiliar rats, because with this sample size the interindividual variation that might exist in yawning frequency will not severely affect the conclusions drawn from the statistical analysis performed to the data.    
Spreadsheet software Microsoft Not available Excel will be the software used to store the yawning recordings initially recorded on the data sheets. Revise the main text for instructions about the recommended way of doing the transcription.
Square filter papers Any maker Not available They are used for covering the cage's bottom.
Tripods Any maker Not available They will be used for fixing the camcorders in front of each pair of observation cages.
Wooden Inverted T-shaped table Handcrafted Not available Read the instructions in the main text to see the way of constructing it. If preferred, a different material to wood can be used. Make sure any material is as resistant as possible to the transmission of ultrasounds, which the rats might use for communication.  



  1. Cézilly, F. A. History of Behavioural Ecology. Behavioural Ecology. Danchin, E., Giraldeau, L. A., Cézilly, F. , Oxford University Press. 3-27 (2008).
  2. Medina, L. M., Garcia, C. M., Urbina, A. F., Manjarrez, J., Moyaho, A. Female vibration discourages male courtship behaviour in the Amarillo fish (Girardinichthys multiradiatus). Behavioural Processes. 100, 163-168 (2013).
  3. Mekdara, P. J., Schwalbe, M. A., Coughlin, L. L., Tytell, E. D. The effects of lateral line ablation and regeneration in schooling giant danios. Journal of Experimental Biology. 221 (8), 175166 (2018).
  4. Walusinski, O. Contagious Yawning. Encyclopedia of Animal Cognition. Vonk, J., Shackelford, T. K. , Springer Nature. Switzerland. 1-5 (2018).
  5. Baenninger, R. Some comparative aspects of yawning in Bettasplendens, Homosapiens, Pantheraleo, and Papio sphinx. Journal of Comparative Psychology. 101 (4), 349-354 (1987).
  6. Gallup, A. C., Church, A. M., Pelegrino, A. J. Yawn duration predicts brain weight and cortical neuron number in mammals. Biology Letters. 12, 20160545 (2016).
  7. Baenninger, R. On yawning and its functions. Psychonomic Bulletin and Review. 4 (2), 198-207 (1997).
  8. Anías, J., Holmgren, B., Urbá-Holmgren, R., Eguíbar, J. R. Circadian variation of yawning behavior. Acta Neurobiologiae Experimentalis. 44 (4), 179-186 (1984).
  9. Holmgren, B., et al. Food anticipatory yawning rhythm in the rat. Acta Neurobiologiae Experimentalis. 51 (3-4), 97-105 (1991).
  10. Byrne, R. W. The evolution of intelligence. Behaviour and Evolution. Slater, P. J. B., Halliday, T. R. , Cambridge University Press. Cambridge. 223-265 (1994).
  11. Provine, R. R. Yawning as a stereotyped action pattern and releasing stimulus. Ethology. 72 (2), 109-122 (1986).
  12. Provine, R. R. Faces as releasers of contagious yawning: An approach to face detection using normal human subjects. Bulletin of the Psychonomic Society. 27 (3), 211-214 (1989).
  13. Anderson, J. R., Myowa-Yamakoshi, M., Matsuzawa, T. Contagious yawning in chimpanzees. Proceedings of the Royal Society B: Biological Sciences. 271 (Suppl 6), S468-S470 (2004).
  14. Palagi, E., Leone, A., Mancini, G., Ferrari, P. F. Contagious yawning in gelada baboons as a possible expression of empathy. Proceedings of the National Academy of Sciences. 106 (46), 19262-19267 (2009).
  15. Romero, T., Ito, M., Saito, A., Hasegawa, T. Social modulation of contagious yawning in wolves. PLoS One. 9 (8), e105963 (2014).
  16. Yonezawa, T., Sato, K., Uchida, M., Matsuki, N., Yamazaki, A. Presence of contagious yawning in sheep. Animal Science Journal. 88 (1), 195-200 (2017).
  17. Miller, M. L., Gallup, A. C., Vogel, A. R., Vicario, S. M., Clark, A. B. Evidence for contagious behaviors in budgerigars (Melopsittacus undulatus): an observational study of yawning and stretching. Behavioral Process. 89 (3), 264-270 (2012).
  18. Moyaho, A., Rivas-Zamudio, X., Ugarte, A., Eguíbar, J. R., Valencia, J. Smell facilitates auditory contagious yawning in stranger rats. Animal Cognition. 18 (1), 279-290 (2015).
  19. Moyaho, A., Flores Urbina, A., Monjaraz Guzmán, E., Walusinski, O. Yawning: a cue and a signal. Heliyon. 3 (1), e00437 (2017).
  20. Gallup, A. C. Why do we yawn? Primitive versus derived features. Neuroscience&BiobehavioralReviews. 35, 765-769 (2011).
  21. Gallup, A. C., Clark, A. B. Commentary: Yawning, acute stressors, and arousal reduction in Nazca booby adults and nestlings. Frontiers in Psychology. 6, 1654 (2015).
  22. Fentress, J. C. Development of grooming in mice with amputated forelimbs. Science. 179 (4074), 704-705 (1973).
  23. Nicoletto, P. F. The roles of vision and the chemical senses in predatory behavior of the skink, Scincella lateralis. Journal of Herpetology. 19 (4), 487-491 (1985).
  24. Fehér, O., Wang, H., Saar, S., Mitra, P. P., Tchernichovski, O. De novo establishment of wild-type song culture in the zebra finch. Nature. 459 (7246), 564-568 (2009).
  25. Kapitány, R., Nielsen, M. Are yawns really contagious? A critique and quantification of yawn contagion. Adaptive Human Behavior Physiology. 3 (2), 134-155 (2017).
  26. Langford, D. J., et al. Social modulation of pain as evidence for empathy in mice. Science. 312 (5782), 1967-1970 (2006).
  27. Gallup, A. C., Swartwood, L., Militello, J., Sackett, S. Experimental evidence of contagious yawning in budgerigars (Melopsittacus undulatus). Animal Cognition. 18, 1051-1058 (2015).
  28. Campbell, M. W., de Waal, F. B. Methodological Problems in the Study of Contagious Yawning. The mystery of yawning in physiology and disease. Walusinski, O. , Karger Publishers. Basel. 120-127 (2010).
  29. R Development Core Team. R: A language and environment for statistical computing. , R Foundation for Statistical Computing. Vienna, Austria. http://www.R-project.org (2009).
  30. Moyaho, A., Beristain-Castillo, E. Experimental Design: Basic Concepts. Encyclopedia of Animal Behavior. Choe, J. C. 3, Second Edition, Academic Press. 471-479 (2019).
  31. Urbá-Holmgren, R., et al. Genotypic dependency of spontaneous yawning frequency in the rat. Behavioral Brain Research. 40 (1), 29-35 (1990).
  32. Smith, M. L., Hostetler, C. M., Heinricher, M. M., Ryabinin, A. E. Social transfer of pain in mice. Science Advances. 2, e1600855 (2016).
  33. Joly-Mascheroni, R. M., Senju, A., Shepherd, A. J. Dogs catch human yawns. Biology Letters. 4 (5), 446-448 (2008).
  34. Barthalmus, G. T., Zielinski, W. J. Xenopus skin mucus induces oral dyskinesias that promote escape from snakes. Pharmacology Biochemistry and Behavior. 30 (4), 957-959 (1988).


Contagious Yawning Laboratory Method Rats Non-invasive Procedure Application Software Calculation Implementation Observation Case Visual Demonstration Spatial Arrangement Test Session Individual Cages Acclimating T-shaped Table Rectangular Table Ceiling Lamp Filter Paper Observation Cage Pairs Divider Digital Camcorders
A Laboratory Method to Measure Contagious Yawning in Rats
Play Video

Cite this Article

Moyaho, A., Díaz-Loyo, A. P.,More

Moyaho, A., Díaz-Loyo, A. P., Juárez-Mora, O. E., Beristain-Castillo, E. A Laboratory Method to Measure Contagious Yawning in Rats. J. Vis. Exp. (148), e59289, doi:10.3791/59289 (2019).

Copy Citation Download Citation Reprints and Permissions
View Video

Get cutting-edge science videos from JoVE sent straight to your inbox every month.

Waiting X
Simple Hit Counter