Brevia

Video Cameras on Wild Birds

See allHide authors and affiliations

Science  02 Nov 2007:
Vol. 318, Issue 5851, pp. 765
DOI: 10.1126/science.1146788

Abstract

New Caledonian crows (Corvus moneduloides) are renowned for using tools for extractive foraging, but the ecological context of this unusual behavior is largely unknown. We developed miniaturized, animal-borne video cameras to record the undisturbed behavior and foraging ecology of wild, free-ranging crows. Our video recordings enabled an estimate of the species' natural foraging efficiency and revealed that tool use, and choice of tool materials, are more diverse than previously thought. Video tracking has potential for studying the behavior and ecology of many other bird species that are shy or live in inaccessible habitats.

Building on recent advances in microelectronics and communication technologies, we developed miniaturized, animal-borne video cameras for studying the undisturbed behavior of wild, free-ranging birds. Our cameras can record bird behavior in its full ecological, physiological, and social context and in locations and circumstances where conventional observation techniques fail.

New Caledonian crows (Corvus moneduloides) are renowned for using tools for extractive foraging (1, 2), but the ecological and evolutionary significances of this unusual behavior are largely unknown. These crows are sensitive to human disturbance and inhabit forested, mountainous areas, where visibility is limited and close following of subjects is impractical. Thus, most current knowledge of the species' biology comes either from observing birds at baited sites or from collecting artifacts (1), but these techniques are not suitable for investigating natural foraging behavior and diets.

We designed tail-mounted camera units (Fig. 1A) that do not interfere with movement and ensure safe shedding of the tag with regular molt (3). Units transmit a color video signal with sound to custom-built receivers and incorporate very high frequency (VHF) radio tags for simultaneous positional tracking. We deployed cameras on 18 different crows (12 males) in our dry forest study site (21°33'50″S, 165°19'27″E), capturing 451 min of analyzable video footage from 12 subjects (38 ± 5 min per bird, mean ± SE; maximum of 60 min).

Fig. 1.

(A) Position and viewing angle of a video camera (red). The unit is taped to the upper side of two inner tail feathers. The lens at the distal end is protruding through the feathers, peeking forward through the legs. (B) Part of a day's foraging trajectory (crow is CC1; 20 December 2006; 5:22 to 6:44 p.m.; movie S2). Live footage was captured by two ground-operated receivers (R1 and R2); a VHF radio tag in the camera unit enabled positional tracking of the bird. Lines show main movements, and oval shapes give approximate locations of the subject before (blue) and during (red) the video shoot. Video-recorded behavior can be linked to the bird's trajectory (see four video stills). Shaded areas indicate shrubs and trees. (C) Activity profile of crow CC1 (percentage locomotor activity in 10-s time bins) during the video shoot (n = 242 bins with uninterrupted video reception). Food items in parentheses were not consumed. (D) Tool-assisted ground foraging by a wild crow (EK1; 17 December 2006; 3:56 to 5:06 p.m.; movie S3). Images show a tool (left) during and (right) after a long flight (note distinct curved tip).

Our footage highlighted the importance of ground foraging for these crows (movie S1). Six birds were terrestrial for 13 to 73% (median of 50%) of analyzable footage (the remaining six were almost exclusively observed in trees), and all video-recorded food items, with the exception of some fruit, were collected on the ground. A case study illustrates how the combination of animal-borne cameras with conventional VHF radio telemetry (video tracking) can generate a rich record of behavior along a known trajectory, yielding data on activity patterns, foraging modes, prey encounters, and diet choice (Fig. 1, B and C, and movie S2).

Video data enabled us to estimate the species' natural foraging efficiency. Our pooled prey sample translates into an encounter rate of eight small food items per hour of ground foraging (118 min of video footage showing birds on the ground across seven subjects with 16 items consumed). Prey items collected during long bouts of ground foraging were considerably smaller than the woodboring beetle larvae (fig. S1) that these crows often hunt with stick tools [Supporting Online Material (SOM) text]. These larvae, a hidden and otherwise unexploited food source, might be an important component of crow diet and could have contributed to the evolution of tool use in this species.

Recordings from two adult males (EK1 and CC1) revealed previously unknown aspects of the species' natural tool use. Crow EK1 used at least three different tools for probing loose substrate on the ground during 45 min of scorable footage (Fig. 1D and movie S3), a foraging mode that has not been reported before (1). The bird (i) used one tool for >18 min, traveling >100 m with it; (ii) briefly put aside this tool on several occasions to use its beak, resuming tool use thereafter; and (iii) transported the tool in flight between prolonged bouts of ground-based tool use and after extracting prey with it. Because sticks are plentiful in forest habitats, these observations indicate that these crows may keep particularly good tools for future use. Both crows made tools on the ground from what appeared to be dry, grasslike stems (movies S2 and S3). This tool material was hitherto unknown (1), and no such tool was found at typical larvae fishing sites in our study area (circa 280 tools collected).

Video tracking is a research tool for studying wild birds that are shy or live in inaccessible habitats and offers three main applications: (i) discovery of unknown behaviors, (ii) context-specific quantification of behavior, and (iii) calibration of other telemetry devices (SOM text, fig. S2, and movie S4).

Supporting Online Material

www.sciencemag.org/cgi/content/full/1146788/DC1

Materials and Methods

SOM Text

Figs. S1 and S2

References and Notes

Movies S1 to S4

References

View Abstract

Stay Connected to Science

Navigate This Article