Borges and Silva • A NEW AREA OF ENDEMISM FOR AMAZONIAN BIRDS 
19 
endemics), and the central-western portion of the 
basin (this study). 
DISCUSSION 
We describe a previously undesignated area of 
endemism in central-western Amazonia that is 
supported by the congruent distribution of six 
avian taxa. Haffer and Fitzpatrick (19X5) noted 
that a small number of strongly differentiated 
bird populations were restricted to this region 
but did not recognize it formally as an AOE due 
the reduced number of avian taxa (J. Haffer, 
pers. comm.). Conceptually and operationally, 
however, an AOE could be recognized with just 
two species (Platinick 1991, Harold and Mooi 
1994). 
Amorin and Pires (1999) also identified a small 
AOE bounded by the Rio Solimoes and Rio Negro 
without offering information about which species 
supported this biogeographic component (Amorim 
and Pires 1999: fig. 27). We suggest naming this 
new area of endemism as Jau in recognition of the 
importance of the National Park to biodiversity 
conservation in the Amazon. This area appears 
nominated as Rio Negro Area ol Endemism in 
previous publications (Borges 2007, Ribas et al. 
2011). However, we believe that Jau Area of 
Endemism is more appropriate to the geographical 
setting of the area (Fig. 2). 
The precise geographical limits of this area are 
only partially identifiable. The southern and 
eastern limits probably coincide with middle 
portions of ihe Solimoes and Negro, large rivers 
known to isolate bird populations (Sick 1967, 
Haffer 1992). The northern boundary is more 
complicated, but from the middle Rio Negro 
northward the landscapes dominated by terra 
Jirme forests are extensively replaced by forests 
and fields growing in sandy soils called 'campi- 
naranas' and 'campinas' (Anderson 1981. IBGE 
1997), This more open vegetation occupies 
thousands of square kilometers between the 
middle and the upper Rio Negro. It is possible 
this discontinuity in vegetation serves as a barrier 
(or filler) to dispersal for some bird species, 
especially those that occur more frequently in 
terra firme forest on clay soils. The western 
boundary of Jau AOE also is difficult to identify 
due to the scarcity of bird collections from the 
region between Maraa to the Brazilian-Colombian 
border. We tentatively suggest the courses of the 
Japura or l^a rivers as the western boundary of Jau 
AOE, although additional fieldwork will be 
necessary (Fig. 2). We also note this region of 
faunal turnover could be not coincident with any 
river course as happens in the southern portion of 
the Amazon Basin (Haffer 1992). 
One fundamental concern in recognizing and 
delimiting the Jau AOE is the current taxonomy 
of the birds considered endemics. The use of 
subspecies in biogeographic analyses is problematic 
because an unknown number of bird taxa described 
in the ornithological literature are not discrete 
evolutionary units (Cracraft 19X5, Haffer 1987). 
Some authors have been successful in using 
subspecies for biogeographic analysis (Bates et al. 
1998, Borges 2007). Analyzing polytypical species 
as a single entity tends to over-estimate the 
geographical distributions of taxa and potentially 
ignores relevant units for conservation and bioge¬ 
ography (Bates and Demos 2001). We considered 
only taxa with accentuated morphological differen¬ 
tiation. However, the taxonomic status of Amazo¬ 
nian bird species, including those discussed here, 
needs to be continuously evaluated and their 
phylogenetic relationships clearly established. 
These taxonomic and biogeographic studies will 
require additional ornithological collections in the 
northern portion of the upper Rio Solimoes and in 
the middle to upper Rio Negro, principally along the 
right margin ol' the latter, as these regions are likely 
contact zones. 
The phylogenetic relationships of the species 
and subspecies endemic to the Jau AOE are not 
adequately known. However, three taxa (Pter- 
oglossas azara azara. Picummis lafresnayi pusil- 
Iti.s, and Mynnoboriis myotherinus artlesiacus) 
appear to have their closest relatives in Ihe lmcri 
and Napo AOEs. This is supported by Borges 
(2007), who applied quantitative methods (parsi¬ 
mony analysis of endemicity and cluster analyses) 
to analyze the biogeographic relationships be¬ 
tween avifaunas from different Amazonian AOEs. 
The avifauna of the JNP shares more species and 
subspecies with localities in the Imeri and Napo 
AOEs than with localities in the Inambari and 
Guiana AOEs (Borges 2007). This suggests the 
barrier effect of the Negro and Solimoes rivers is 
stronger than the barrier effect in the western 
border ol the Jau AOE. A working hypothesis to 
the evolutionary history' of the six taxa restricted 
to the Jaii AOE is that they wea* derived through 
parapatric or peri panic speciation from ancient 
species whose ranges were once centered in the 
Imeri and \apo AOEs. Alternatively , tectonic 
events that affect the lower course of the Rio 
