Black-capped Chickadee — Dee Count Alarm Calls
The *chick-a-dee-dee-dee* call of *Poecile atricapillus* encodes predator danger level in the number of terminal "dee" notes — with a counter-intuitive inversion: smaller predators trigger more dee notes, not fewer. This is one of the clearest examples of referential encoding of a quantitative property in a non-primate species.
What the Signal Is
The full call has four acoustic unit types (A, B, C, D):
- A, B, C — variants of the "chick-a" portion; not always present
- D — the repeated "dee" note; always present in alarm contexts
The signal that matters for predator encoding is the *number of D notes* and the *interval between the "chick" and "dee" portions*. Normal contact calls carry 2–3 dees. Alarm calls triggered by dangerous raptors carry 4 or more, with a shorter chick-dee interval.
In one documented case, a pygmy owl — the primary avian predator of chickadees — triggered a call with 23 dee notes.
What It Encodes
The dee count is inversely related to predator body size but positively correlated with predator *danger* to the caller. This is not paradoxical: small raptors (pygmy owl, merlin) are more maneuverable in dense forest and are effective chickadee hunters. Large hawks (red-tailed hawk) are formidable but cannot pursue small birds through canopy. The signal encodes effective threat, not raw size.
An analysis of over 5,000 alarm calls confirmed the pattern. The encoding is continuous, not categorical — a gradient, not a lookup table.
Secondary encoding: the C note may signal food location; D notes distinguish predator from conspecific. The full call is compositional at a basic level.
Cross-Species Reception
This signal crosses Umwelten boundaries:
- Red-breasted nuthatches in mixed-species winter flocks respond to chickadee alarm calls and calibrate their response to dee count — stronger flight responses to high-dee calls
- Woodpeckers, kinglets, and vireos in the same flocks also respond
This is a rare documented case of cross-species *graded* signal comprehension — another species not only responds to the alarm but extracts the encoded intensity information.
Honest or Deceptive
Honest. The dee count tracks real predator threat parameters. No documented cases of false inflation for competitive purposes, though the gargle vocalization (used in male aggression) shows that chickadees can deploy vocalizations strategically.
Acquisition
The basic dee/alarm structure appears to be innate — it is produced from birth and does not require social models for the alarm call type. However, the gargle call (used in dominance) is socially learned; sub-gargle precursors develop between days 30–35, shaped by hearing parents and siblings. The two calls represent both ends of the innate/learned spectrum within one species.
Key Researcher / Discovery Story
Studied since the mid-1970s. The systematic dee-count/danger correlation was demonstrated through analysis of 5,000+ alarm calls, studying responses to different raptor species. The counter-intuitive small-predator-more-dees result was the critical finding — it showed the signal tracks ecological danger rather than physical size, implying some form of predator categorization by the caller.
Open Questions
- Is the dee count deliberately modulated by the caller, or a graded autonomic response to arousal? The distinction matters for questions of intentionality.
- Do chickadees across populations show dialectal variation in baseline dee counts, as in the black-capped-chickadee-gargle-dialect (when it exists)?
- What is the upper bound on dee count, and does it saturate at extreme threat levels?
- How does the receiver parse the dee count — is there a counting mechanism, or are they responding to a temporal density measure?