Collar and Bird • DISCRIMINATION OF ANDEAN IBIS 
461 
TABLE 1. Measurements (mm), Cohen’s d effect sizes and phenotypic difference scores of Theristicus melanopis and 
T. branickii specimens at AMNH. MNHN, NHM, LISNM, and ZMB. Initials “bpw" = belly-patch width, measured as the 
distance from the lower edge of the gray breast-band to the upper edge of the black under-belly (to the nearest 5 mm). 
telaiwpis 
branickii 
Cohen’s d 
Score 
Mean * SD 
Range 
» 
Mean t SD 
Range 
» 
Bill 
139.6 ±7.316 
126-154 
29 
118.4 ± 8.011 
105-130 
29 
2.169 
2 
Tarsus 
79.2 ± 3.573 
72-86 
30 
69.0 ± 3.914 
60-75 
32 
2.706 
2 
Wing 
375.2 ± 14.917 
351-416 
30 
391.0 ± 10.855 
369—409 
32 
-1.209 
1 
Tail 
185.2 ± 9.498 
155-200 
29 
215.4 ± 9.922 
200-238 
31 
-3.112 
2 
Bpw 
71.8 ± 20.947 
25-105 
30 
140.4 ± 19.500 
110-180 
24 
-3.389 
2 
buff (as on the neck in both species) (medium 
difference: 2); and 
(4) the belly-patch in branickii is more extensive 
than in melanopis (difficult to measure in 
specimens but no overlap recorded: Table 1) 
because of its somewhat more restricted 
black underbelly (medium difference 2). 
Other characters mentioned in the literature for 
branickii , rufous-tinged neck-base, paler foreneck 
and breast (both perhaps seasonal differences), 
rather dark gray (or at any rate darker) wing- 
coverts (there is a trend, but some melanopis are 
as gray), and grayer upperparts, do not find clear 
endorsement in the material we have consulted. 
The differences between the two forms are 
considerable. Summing the two greatest morpho¬ 
metric differences (tail length, 2. and bill length, 
2: Table I) and the three strongest plumage 
differences, we have a combined score of 11 
between the two taxa. These differences are, 
following Tobias et al. (2010), more than 
sufficient to recognize the two taxa as separate 
species. T. branickii can actually make a stronger 
claim for specific status, than T. melanopis can 
from T. caudatus, as it more distinct in terms of 
number of differing characters, given that separa¬ 
tion of the latter two species is based (so far as wc 
are aware) on presence or absence of a wattle 
(score 3) and the former’s all-black belly (3) and 
whiter wings (1). 
DISCUSSION 
Theristicus branickii is clearly a species rather 
than a subspecies, based on multiple plumage and 
morphometric characters. However. Tobias et al. 
(2010) also considered the option of scoring real 
differences in behavior and or ecology between 
taxa. We note that branickii is a bird of upland 
puna (3,700-4.500 m; Schulenberg et al. 2007), 
whereas melanopis ranges from sea-level to 
3,000 m (Matheu and del Hoyo 1992). The 
completeness of this elevational disjunction is 
not known, but we speculate the morphometric 
disjunctions in the two taxa (notably the shorter 
bill and legs) are related to differences in foraging 
substrate that arc in turn related to elevation. We 
also note that, from maps and information in 
Matheu and del Hoyo (1992) and Delaney and 
Scott (2006), branickii is resident while melanopis 
undertakes considerable seasonal migrations. 
Puna habitat occupied by Theristicus branickii 
occurs in Ecuador (Antisana, Cotopaxi), Peru 
(Juntn, Huancavelica, Ayacucho, Cuzco, Puno), 
northwest Bolivia (La Paz), and extreme northern 
Chile (Arica) (Slcinbacher 1979). The species 
range-map for Theristicus melanopis in Ridgely et 
al. (2003) suggests an extent of occurrence 
>500,000 km* for branickii. and the combined 
area of the three Central Andean puna ecoregions 
is >480,000 km 2 (World Wildlife Fund 2010), far 
above the threshold (< 20,000 km 2 ; IUCN 2001) 
for possible listing of the species as threatened on 
the IUCN Red List. 
The species has been described as “uncommon 
and very local throughout [its] range” (Matheu 
and del Hoyo 1992: 499), and as “apparently 
declining in some areas” (Stotz et al. 1996:141). 
The evidence in Bolivia and Chile strongly 
suggests numbers cannot have been high. Bolivia 
seems to be listed on the basis of a single bird 
taken at ‘Lagonillas’ in La Paz (or now Cocha¬ 
bamba: Paynter 1992) in July 1901 (Chubb 1919; 
specimen BMNH 1902.3,13.1651). Chile is listed 
on the basis of a specimen taken in June 1853 in 
the Cordillera de Arica, where local people once 
knew it (Goodall et al. 1951) but no longer do so. 
resulting in the view that the form is accidental 
in the country (Martinez Pina and Gonzdlez 
Cifuentes 2004). In Ecuador it has been assumed 
that numbers were “apparently always very 
