Rodriguez el id. • VENEZUELAN BIRD SURVEYS 
237 
Abundance .—The total number of individual 
records during the surveys was 12,109 (8,573 
from SI and 3,536 from S2) and the total 
individuals recorded (only from SI) was 15.114 
(Table 4). There was a mean of 1,76 individuals/ 
record and 28.73 individuals/species (from SI). 
The proportion of common species detected was 
27% (37-56% in the different ecoregions) 
(Table 4). Most ‘common’ species were recorded 
>50 times whereas most uncommon or rare 
species were recorded <25 times (Fig. 4). A 
large proportion of the species sampled were 
uncommon or rare (Fig. 4): —95% of species 
were below the 15th prevalence percentile, while 
128 species (24%) were recorded only once (from 
SI). On average, we obtained 318 records/transect 
(min = 95, max = 479) and 6,4 species/3-min 
count (min = 0. max = 18) (from SI). We 
observed a mean of 560 individuals/transect (min 
= 177, max = 1,379) and 11.2 individuals/3-min 
count (min = 0. max = 237) (from SI). The most 
abundant species was Western Cattle Egret 
(Bubulats ibis) with 823 individuals (data from 
SI) and the species with the highest number of 
records was Tropical Mockingbird (Minius gilvus ) 
with 291 records from SI or 360 from SI + S2. 
DISCUSSION 
Our survey recorded 57% of all potential 
species in the 27 transects and 43% of all known 
Venezuelan birds (Table 2) in a little over 2 weeks 
of field work using 108 person-hrs of effective 
bird sampling. However, 6-8 months were 
devoted to planning and logistics prior to the 
survey. We suggest that organizing and executing 
a survey of this type, in a fairly large tropical 
country with an adequate road network but limited 
scientific expertise, can be done in <1 year at 
relatively low costs. The amount of data generated 
and the funds required for implementation 
(—$150,000 US including field equipment, plan¬ 
ning. salaries, and survey costs), is clearly a 
modest effort when compared to long-term 
programs such as the North American UBS and 
the Christmas Bird Counts (Bock and Root 1981, 
Bystrak 1981. Robbins et al. 1986. Robbins 2000). 
We believe our objective of developing and 
testing a rapid and cost-efficient nation-wide bird 
survey for Venezuela was achieved. 
We were also able to obtain reliable richness 
estimates and visualize the composition pattern in 
addition to achieving our cost-efficient objective. 
The highest species richness was in Guayana, 
closely followed by NWCordilleras, which are the 
two ecoregions that include highlands and have 
the widest elevation gradient (Table 4). This is 
not a surprising result, as the wider elevation 
range provides w'idor variety of habitats that host a 
larger diversity. NorthCoast. a lowlands region 
with predominance of dry or desert scrub, had the 
lowest species richness (Table 4), a common 
pattern in arid regions (Stotz et al. 1996. 
Rodriguez-Ferraro and Blake 2008). 
The cluster diagram (Fig. 3) shows that groups 
of locations belonging to the same ecoregion are 
related due to their large number of shared 
species. There are, however, several cases show¬ 
ing a different pattern. The three northwestern 
locations south of the Orinoco River in Bolivar 
Slate (Caicara, Guri, and El Manteco) appear 
closer to locations in the Llanos than the rest of 
Guayanan locations (Fig. I). This is probably due 
to a high number of typical savanna species, the 
predominant habitat in the Llanos and present in 
northern Bolivar (Huber and Oliveira-Miranda 
2010). There has been extensive deforestation in 
this area in the last 25 years. At least 2,000 knr of 
rainforest were lost to logging and burning in 
Bolivar State between 1988 and 2010 (Oliveira- 
Miranda et al. 2010:125). leading to savannization 
in what formerly were forested locations (i.e., El 
Manteco, No. 22; Fig. I). 
The NorthCoast appears to be split into three 
groups, two of which appear closer to the Llanos. 
Species shared between these (Perija, Quebrada 
Arriba. Araya and Faria) and the proper Llanos 
transects are mostly those of farmlands and dry 
scrub, tw'o habitats shared between these two 
ecoregions. whereas the other NorthCoast group 
(Macanao, Paraguana, and Rio Tocuyo) share 
most species typical of desert scrub, the predom¬ 
inant habitat in these three transects and the entire 
ecoregion. The Sur del Lago transect in the 
western Andean foothills, due to its wide altitude 
range, was problematic in terms of ecological 
classification. It included mostly Andean foothills 
(part of the NWCordilleras ecoregion) but also 
pan of the Maracaibo Lake lowlands and overlaps 
with the NorthCoast region with which it shares a 
number of species. A transect (# 57) that appears 
completely isolated (Fig. 1) with a binomial index 
close to 1, is Pinango. despite its ecological and 
geographical relation to the NWCordilleras re¬ 
gion. This location had the lowest number of 
species (only 22 in the surveys) but the highest 
percentage of endemism and of unique species, a 
