Earlier evidence suggests that besides humans, some species of mammals and birds demonstrate visual self-recognition, assessed by the controversial ‘mark’ test.  Whereas there are high levels of inter-individual differences amongst a single species, some species such as macaques and pigeons which do not spontaneously demonstrate mirror self-recognition (MSR) can be trained to do so.  We were surprised to discover that despite being widely used as a model system for avian research, the performance of zebra finches (Taenopygia guttata) on the mark test had not been studied earlier. Additionally, we studied the behavior of another species of passerine songbirds (Indian house crows; Corvus splendens) on the mark test and other self-exploratory behaviors.  Although, a small number of adult male zebra finches appeared to display heightened responses towards the mark while observing their reflections, we could not rule out the possibility that these were specific to the mirror and not a part of general grooming.  In contrast, none of the house crows demonstrated mark-directed behavior or increased self-exploratory behaviors when facing mirrors.  Our study suggests that self-directed behaviors need to be tested more rigorously in adult male zebra finches while facing their reflections and these findings need to be replicated in a larger population, given the high degree of variability in mirror-directed behaviors.  We have also shown that house crows demonstrate early stages of self-awareness but do not pass the mark test.

A total of 6 adult male zebra finches (Taenopygia guttata) and 5 adult house crows (Corvus splendens) were used for the mirror self-recognition experiments.  All experimental birds were housed in aviaries in the Animal Facility at the National Brain Research Centre, Manesar and experimental protocols were approved by the Institutional Animal Ethics committee in accordance with guidelines laid down by the Committee for the Purpose of Control and Supervision of Experiments on Animals (CPCSEA), India and compliant with international standards on animal welfare.  All birds were identified by plastic leg bands and had not been used for any other experiments prior to testing for MSR.  Whereas zebra finches had been bred in our aviaries, the house crows were wild-caught and housed at the Animal Facility, NBRC for ~2 months prior to the MSR experiments.  The aviary for zebra finches was maintained at a temperature of 25ºC-29ºC throughout the year and a 12L:12D light cycle.  House crows were housed individually in cages in a separate aviary with natural lighting, maintained at 25ºC-32ºC.  Most of the experiments to test MSR were performed between May and September, 2014 when the daylight hours are the longest in India and the light-dark cycle was between 13-16L:11-8D, to avoid behavioral changes elicited by alterations in the circadian rhythm. 

MSR task for Zebra Finches

Six adult male zebra finches which had no prior exposure to reflective surfaces were housed in large cages, with unlimited access to food and water prior to the experiments and between trials.  Birds were placed singly in a cage with a perch (Fig. 1A) housed in a closed chamber which was visually and acoustically isolated from other birds. A mirror (20 cm by 15.5 cm) was introduced in the cage and their responses to their reflections were recorded for 30min, using a video camera (Sony HDR-CX240EB).  The method that we used was based on those followed by Prior et al. and Soler et al. (Prior, Schwarz, and Gunturkun 2008; Soler et al. 2020) wherein birds were first exposed to a mirror (mirror exposure phase) followed by a test phase consisting of interleaved trials of sham mirror/covered-mirror, covered-mirror mark and the mark test ( Reiss and Morrison 2017). During the mirror exposure phase, birds were exposed to and habituated to the mirror for the first 4 days of the week. For this phase, birds were placed in the testing chamber with the mirror for a total of 8 sessions (two 30-minute sessions per day) separated by a gap of 4-5 hours every day for 4 days.  Thus, experimental zebra finches were exposed to the mirror for a total of 4 hours for habituation. If birds showed no signs of stress or aggression towards the mirror, they were subjected to the mark test. The mirror exposure phase was followed by a test phase, during which birds were marked with paint or sham-marked (see below).  As was done for the mirror exposure phase, 2 trials were performed daily, separated by a gap of 4-5 hours, and the experiment was repeated 5 days a week.  The test phase comprised of five conditions each of which were tested 4 times, three with a mirror (Mirr) and two with a non-reflective board (Brd) placed in the cage as a control.  Normal (grey) adult male zebra finches have alternating black and white striped feathers on their throats, with a tuft of black feathers at the midpoint. They were very sensitive to tactile stimulation, and removed small stickers placed on the feathers under their beak within seconds. Therefore, red or black odorless, non-toxic water-based tempera colors (washable paint, style: 54 – 1205, Crayola, Funskool) were used to paint spots of color on the throat or head for the mark test (Buniyaadi, Taufique, and Kumar 2019; Prior, Schwarz, and Güntürkün 2008) (Fig. 1C-E) which could not be observed by the bird unless in a reflection as given below:

1. Red spot on the throat facing a mirror: RedThrMirr
2. Red spot on the forehead facing a mirror: RedHdMirr
3. Black spot on the throat facing a mirror: BlThrMirr
4. Red spot on the throat facing a board: RedThrBrd
5. Feathers on the throat were stroked with a brush (sham): CtrBrd

For each condition, while one experimenter gently held the birds’ heads and beaks in such a way that their eyes were covered, the other painted the mark on their throat or forehead (for the RedThrMirr, RedHdMirr, BlThrMirr and RedThrBrd conditions) or stroked feathers on the throat [for the CtrBrd condition; Prior, Schwarz, and Gunturkun 2008; Kraft et al. 2017)]. Whereas the red spots were used to study whether zebra finches could recognize themselves in mirrors, we assumed that the black spot (painted on black feathers on the throat) served as a negative control for color.  For two zebra finches which had white patches on their necks, white spots were painted on their necks as controls for color.  In the last condition, birds were sham-marked by stroking feathers on the throat with a Q tip (also used to apply the paint) to study whether touch rather than visual stimuli evoked preening on the throat.  After painting different colored spots for each condition, zebra finches were placed singly in the test cage in a sound-proof isolation chamber facing the mirror or the non-reflective board and the video-recording was started. Trials of all experiments were interleaved randomly and counterbalanced so that the zebra finches did not get habituated to any of the experimental conditions. 

MSR task for House Crows

House crows (n = 5; 3 males and 2 females) were housed in cages 30” x 21” x 34” in dimension with ad libitum access to food [bread, canned dogfood (Pedigree), boiled egg, multivitamins] and water.  Prior to the experiments on self-recognition, they were placed singly in a cage for two days for habituation, during which time they were also provided food and water (Fig. 1B).  Following habituation, food, water and perches were removed from the cage during behavioral recordings performed during mirror exposure and test phases, as described above for zebra finches. A screen was placed between the house crow being tested and other crows in the aviary.  Whereas experimental birds were visually isolated from other house crows, they were able to hear them vocalize during experiments. House crows have black plumage on their wings, head and just under the beak and grey feathers on their bodies and the back of the head.  Despite the fact stickers have been reported to provide tactile stimulation in jackdaws (Soler, Pérez-Contreras, and Peralta-Sánchez 2014), we found that it took longer to paint spots of color on house crows and remove them after the experiments, leading to greater degrees of stress amongst these birds.  Therefore, to study whether they recognize themselves in a mirror image, we placed yellow stickers (diameter, 9 mm; weight, 12-13 mg) on their foreheads or under the beak, which would be easily visible against their black plumage.  For controls, similarly-sized black stickers were placed on their foreheads, which blended with the color of their feathers (Fig. 1F-H).  Care was taken to ensure that the house crows were unable to see where the mark was placed on their bodies, in the same manner as described for zebra finches by two experimenters.  Additionally, stickers were placed on their wings and bellies to test whether they were capable of removing the ‘marks’ if viewed directly, that is, by perceiving stickers as foreign objects.  The mirror exposure period was followed by the test phase comprising of a set of three different conditions (4 trials, 30 min each) during which crows were placed in front of a mirror or board (control), as follows:

1) Yellow sticker on the forehead with mirror: YelHdMirr

2) Yellow sticker on the throat with mirror: YelThrMirr

3) Black sticker on the forehead with mirror: BlHdMirr 

4) Yellow sticker on the head with a non-reflective board: YelHdBrd

5) Feathers on the forehead were touched as if pasting a sticker facing a non-reflective board (sham): CtrBrd. 

For all conditions, behavior was recorded with a Micromax handset (Canvas A110) 8-megapixel camera at 30 fps.  Trials of all experiments were randomized and counterbalanced so that crows did not get habituated to any of the experimental conditions. 

Behavioral Analysis

The first 30 seconds of the video recordings were discarded to avoid measuring behaviors which may have been affected by the experimenters exiting the area near the experimental chamber or cages. A number of behaviors were scored, including (1) Social responses, which included the (i) time spent in front of the mirror (in seconds) and (ii) the number of songs and calls; (2) contingency testing, that is, behaviors performed in front of the mirror which enables the individual to perceive a visual – kinesthetic match between the behavior and its reflection (Morrison and Reiss 2018). This includes the (i) number of head turns in front of the mirror, including those to one side and back to the midline in front (Suppl. Figure 1A) and from side to side.  Head turns are used to explore the surroundings since birds’ eyes are placed laterally (as described in the schematic; (3) mark-directed behavior including the number of times birds (i) preened their neck, (ii) scratched their neck or head using their claws and (iii) preened their wings while facing the mirror (directed preening).  We also analyzed the (4) preening behavior including the number of times birds (i) preened their neck, (ii) scratched their neck or head using their claws and (iii) preened their wings while facing away from the mirror (undirected preening).  Additionally, we analyzed (5) aggressive/ exploratory responses by counting the number of times birds pecked on the mirror or the frame and (6) search responses wherein the (i) number of times birds looked behind the mirror and the (ii) number of times birds turned clockwise and anti-clockwise were counted.

For the statistical analysis of preening behavior, we combined preening using the beak and scratching/ preening using claws.  Furthermore, for comparison across different conditions in the test phase, in case of CtrBrd and RedThrBrd conditions, directed and undirected preening were collated since birds could not observe their reflections and randomly preened themselves in different parts of the cage.  These were then compared with directed preening behavior in the mirror trials.

Statistics:

The One-Way Repeated Measures Analysis of Variance (RM-ANOVA) test was used to compare different conditions and test for statistical significance, followed by the Tukey’s post-hoc test using SigmaPlot 14.0 (Systat Software, San Jose, CA), which gives adjusted P values for pairwise multiple comparisons.  Data obtained for preening (the most important self-directed behavior for demonstrating whether birds passed the mark test) was tested using the Fisher’s exact test or by multiple comparisons using the Fisher’s exact test with the Bonferroni correction using the R software [version 3.6.0; R library; RVAideMemoire (version 0.9-73.  To test changes in the progression of behaviors such as time spent, vocalization, search responses, pecking and head turns in the mirror exposure phase, data was fitted in a linear regression model to calculate the slope parameter, followed by a t-test for the trend analysis. A Welch's t-test was performed to compare data of contingency testing from mirror versus board conditions for Y45 and G123, the two zebra finches in which preening increased near the marked area. The entire dataset that we obtained for these experiments was analyzed by one of the observers (ANM) who had not performed the experiments.  Another observer analyzed 30 trials and the correspondence between their ratings (r value or inter-rater reliability using the Pearson product moment correlation test; SigmaPlot 14.0) was highly correlated [r range 0.979 - 0.996, n range 30 to 140 , P range 10^-123 to 10^-24; (Clary and Kelly 2016)].  Wherever there were doubts or disagreements regarding the data, the observers discussed the results to reach a consensus for the final analysis. 

There are no missing values.