Running Head: COMBINATORY ACTIONS IN CORVIDS AND PARROTS
1
Supplemental Materials
Combinatory Actions During Object Play in Parrots (Psittacus erithacus) and Corvids
(Corvus)
by A. M. I. Auersperg et al., 2014, Journal of Comparative Psychology
http://dx.doi.org/10.1037/a0038314
A) Object Manipulation Ethogram
OBJECT-RELATED BEHAVIORS
 Object–object & object–substrate combinations (see Analysis/Results)
o Free object–object combinations

Balancing (ba-cb): The bird is lying on its back and is balancing a
combination of two or more objects with its feet.

Carry (ca-cb): Holding two or more objects in the bill while the bird is
moving along ground, or holding an object in feet while flying.

Combining (cb): Placing one object into physical contact with another.
Can be subcategorized as above and below. Ambiguous object–object
combinations are coded as cb.

Grab with foot (gf-cb): Two or more objects are grabbed with the foot
and simultaneously held up to the bill.

Put on top (pot): Placing one (or more) object on top of another object.

Pick up (pi-cb) Two objects are seized together with the bill and lifted.
Objects grabbed with the foot are coded as ‘gf’
o Object–substrate combinations

Caching (ch): Making an object disappear out of immediate sight.
COMBINATORY ACTIONS IN CORVIDS AND PARROTS
2
Includes putting objects underneath, between, behind and inside any
aviary features except for the tubes, holes and poles on the activity
plates or the aviary wire grid. NOT to be confused with ‘inserting’.

Inserting

Dipping (dp): Inserting or retrieving an object from an open
cavity without releasing it (not to be confused with probing).

Inserting (in): An object is inserted into an opening or a hole in
the activity plates or through the aviary wire cage walls (not to
be confused with caching).

Probing (pr): Inserting an object repeatedly into a cavity
without letting go.

Ring-stacking (rst): Putting a ring on top of another object e.g., placing
a ring over a stick or over a pole (Object inserted into a rings lying on
the floor are however coded as ‘inserting’).
(we originally included ‘recover’ as a object–substrate combination, when
objects were recovered from tubes/holes/caches; these data were not used
here).

Object actions involving only one item were recorded but not included into
Analysis/ Results.
B) Apparatus Dimensions
OBJECT abbreviations:
Big Cube bc
Medium Cube mc
Small Cube sc
COMBINATORY ACTIONS IN CORVIDS AND PARROTS
Big Ring br
Medium Ring mr
Small Ring sr
Big ball bb
Medium ball mb
Small Ball sb
Tiny Ball tb
Stick 1 biggest (grabbed at middle or end ) (s1m, s1e)
Stick 2 (s2m, s2e)
Stick 3 (s3m, s3e)
Stick 4 (s4m, s4e)
Stick 5 (s5m, s5e) thinnest
All objects come in 3 colors: blue (b), red (r), yellow (y)
Playground objects
Big tube horizontal bth
Medium tube horizontal mth
Small tube horizontal sth
Tiny tube horizontal tth
Big tube vertical btv
Medium tube vertical mtv
Small tube vertical stv
3
COMBINATORY ACTIONS IN CORVIDS AND PARROTS
Tiny tube vertical ttv
Big Pole bp
Small Pole sp
Big Hole bh
Medium Hole mh
Small Hole sh
Tiny Hole th
Dimensions wooden plates for tubes, poles, and holes
400x400 mm
Length tubes, sticks
100 mm
Inner diameter tubes
TT 22 mm
ST 28 mm
MT 37 mm
BT 48 mm
Diameter poles
BP 20 mm
SP 12 mm
Inner diameter holes
TH 17 mm
SH 25 mm
4
COMBINATORY ACTIONS IN CORVIDS AND PARROTS
5
MH 33 mm
BH 40 mm
Inner diameter ring
BR 45 (Thickness 10 mm)
MR 32 (Thickness 10 mm)
SR 20 (Thickness 7 mm)
Diameter sticks
S1 20 mm
S2 14 mm
S3 12 mm
S4 8 mm
S5 4 mm
Diameter balls
BB 40 mm
MB 29 mm
Sb 25 mm
Sidelength cubes
BC 39 mm
MC 30 mm
SC 25 mm
C)
Object Color and Shape Preferences
To determine possible shape and color preferences for objects we noted (if clearly
visible) the color/shape of each object subjects interacted with (this was not restricted to
combinatory actions but included all object actions). For each subject, we calculated the
COMBINATORY ACTIONS IN CORVIDS AND PARROTS
6
mean percentage of objects being of one of the three colors or of the four shapes. We
conducted two GLMMs including all six species studied with ‘subjects’ as random factor and
species, sex, age class, color /shape as fixed factors.
We found a significant interaction for species*color (F10=3.695, p<0.0001) and
species*shape (F15=13.701, p<0.0001).
Bonferroni-Holms corrected posthoc tests (Wilcoxon singed rank tests) were
conducted for each species for the three colors and four shapes (threshold values for color:
p<0.0167 and for shape: p<0.0083).
The respective analysis revealed that Goffins interacted significantly more with
yellow objects than with red (Wilcoxon test, Z=2.979; p=0.003) and jackdaws preferred blue
over red objects (Wilcoxon test, Z=2.521, p=0.012). Goffins also chose sticks significantly
more than cubes (Wilcoxon test, Z=2.988, p=0.003). NCCs preferred balls over all other
objects (Wilcoxon test, Zstick=3.823, p<0.0001; Zcube=3.822, p>0.0001; Zring=3.823,
p>0.0001). NCCs also selected sticks (Wilcoxon test, Z=2.689, p=0.007) and rings
(Wilcoxon test, Z=2.999, p=0.003) more frequently than cubes.
COMBINATORY ACTIONS IN CORVIDS AND PARROTS
7
Figure S1. Mean percentage of objects interacted with for particular shapes (A: Percentage of sticks, balls,
cubes and rings on x-axis) and colors (B: Percentage of blue, yellow, red on x-axis). The different colors of the
bar chart indicate each different species.