Skip to main content

Mockingbird flush distances when approached by humans of differing threat levels


Levey, Douglas (2023), Mockingbird flush distances when approached by humans of differing threat levels, Dryad, Dataset,


These data show flush distances of incubating female mockingbirds on the University of Florida campus (Gainesville, FL) when approached by four different humans, each posing a different level of threat to the bird.  Column A (Nest Location) describes the location of each nest. Multiple rows for the same nest indicate approaches by different humans and/or on different days. 

Column B (Visitor Type) indicates the classification (experimental treatment) of the human. The threat level of that human is provided in Column J (Threat Ranking, which ranges from 2 [high] to 5 [low]).  An “Intruder” approached the nest on 4 sequential days and placed one hand on the rim of the nest for 15 seconds before retreating.  Mockingbirds quickly learned to recognize them and thus Intruders have the highest threat ranking (2).  An “Associate” approached the nest alongside the Intruder but stopped 3m from the nest and waited there until the Intruder returned from the nest. The Associate and Intruder stood facing each other for 15 sec before departing from the nest. Associates had a medium threat ranking (3).  A “Bystander” approached alone on four consecutive days, stopped 3m from the nest and stood quietly without looking at the female for 10 min before departing. Females did not flush during these trial sessions and hence flush data and behavioral data are not provided. However, the Bystander approached the nest and placed one hand on the lip of the nest for 15 seconds on the last day of trials, forcing the female to flush; results from those trials (one per female) are shown. Bystanders had a low threat ranking (4) because they typically didn’t threaten the nest.  Finally, a Control individual who the female had never seen approached the nest on the last day of trials and placed one hand on the lip of the nest for 15 seconds.  Control individuals had the lowest threat ranking (5).

Column C indicates the human’s distance from the nest when the female flushed upon approach.  All humans started to approach the nest from ~30m and walked towards it at a speed of ~1m/sec.

Columns D through I provide behavioral data of the female and her mate (who sometimes wasn’t present; see column M). Female and male mockingbirds cannot be distinguished. Hence data in columns D through I are summed for whatever bird or birds were responding.  Column D shows the number of times birds hovered over the head of the human. Hovers were brief, typically <3 sec, and often ended in a “swoop” (Column E) or an “attack” (Column G). An attack was defined as a downward flight to within 1m of the human. A swoop was a downward flight toward the human but not to within 1m.  Column F (“indirect flights”) shows the number of flights over or around and within 5m from the human, typically from one perch to another.  Column H shows the time of the first alarm call in minutes and seconds from the time the human started to approach the nest.  Column I shows the total number of alarm calls from the start to finish of the trial.

Column K provides an estimate of human activity near the nest (the number of humans who walked within 5m of the nest in a 10 min period).

Column L indicates an irregularity with a trial (0 = protocol tightly followed; 1 = protocol not tightly followed or a problem outside of investigators’ control).

Column M provides miscellaneous notes. Data on presence of the male was inconsistently taken.


Our study took place in Gainesville, Florida on the ~800-ha campus of the University of Florida. Northern mockingbirds commonly nest in isolated shrubs or small trees in open areas frequented by thousands of students daily. They are highly acclimated to the presence of humans, typically showing no response unless approached within 2–3 m. Pairs are socially monogamous and territorial. Territory size is 1–2 ha on campus. Only the female incubates the eggs, which allowed us to tightly control a given bird’s exposure to different humans. Because much of the population was individually color banded and the study occurred during a single breeding season on widely separated territories, we are confident that each female was included only once in our sample.

When the nest of an incubating female was discovered, we first established three 30m paths along which an approaching human could be seen from the nest. In some cases, a person not participating in the trials removed a small number of leaves or twigs near the nest to assure equal visibility of approaching humans along all paths. The purpose of multiple approach paths was to minimize the possibility that the bird might respond to the direction of approach rather than (or in addition to) the identity of the human. On the day that trials at a given nest were initiated, one approach path was randomly selected for use. On the second day, one of the remaining two paths was randomly selected and on the third day, the last path was used. On subsequent days and trials, path use was randomized in the same manner.


Trials on a given nest occurred over five consecutive days. On the first three days (training period) the nest was visited by High and Medium Threat individuals together and at least 3hr before or after by the Low Threat individual. The goal of these trials was to establish the three treatments – i.e., to familiarize the female with each type of human and their threat level. Prior work using a similar protocol established that mockingbirds significantly increase their response to threatening humans after 2 visits to a nest. On the last two days (test period), we measured the female’s response to approaches by a single individual of the four threat levels, including the No Threat individual who had not previously approached the nest. The order and paths of these approaches were randomized. On the last two days, two trials occurred per day, separated by >6 h. In these four trials, all humans behaved identically, except for the approach path used. They approached the nest from 30 m at a rate of 1 m/s, paused ~1 m from the nest for 15 s, touched the rim of the nest for 15 s, and walked away at the same rate and along the same path as they had arrived. None paused 3 m from the nest.


Our experimental design did not employ distinctive clothing or masks with exaggerated features that the birds could associate with treatments. All participants were instructed to wear their normal clothing, regardless of the trial; the only restriction was no hats or sunglasses. Except for the first author, all participants in trials were university students, presumably indistinguishable at the start of trials from the other ~50,000 students on campus at the time. (The study occurred before the COVID pandemic.) Our goal was to mimic as closely as possible the natural variety of humans near each bird’s nest.


Because all trials were conducted during incubation, our experimental design also controlled for the well-known association between nest defense behavior and stage of the breeding cycle. If any eggs hatched or disappeared, we discontinued trials on the impacted nest.


Data have been quality checked but not otherwise processed.


Our experimental protocol was reviewed and approved by the University of Florida Institutional Animal Care and Use Committee (protocol #D884-007).


University of Florida, Award: Katharine Ordway Fund for Ecosystem Conservation

National Science Foundation