232 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol 124. No. 2. June 2012 
only one adequate road available. We marked 50 
stops separated by 800 m from each other using a 
Global Positioning System (GPS). Transects were 
selected to cover the largest possible environmen¬ 
tal gradient within the cell. 
Sun-ey Teams .—A team of 21 professional and 
amateur ornithologists from Venezuela and 
abroad was assembled and trained in methods to 
be used during I week in early March 2010. The 
group was divided into seven survey teams with a 
team leader (main observer) and two assistants. 
Each team was assigned several study transects, 
typically within a single ecorcgion. Team leaders 
were professional ornithologists or bird-watchers 
with thorough knowledge and experience in 
identifying the birds in their region. Most surveys 
occurred between 12 and 28 March (-2 weeks) 
but, due to unexpected problems, two transects 
were surveyed between 6 and 10 April. The short 
survey period was designed to minimize the effect 
of time of year or season, as well as to keep costs 
of training and administration low. 
Bird Suneys. —Bird surveys were conducted 
along the marked transects, following an adapta¬ 
tion of the methods of the North American BBS 
(Bystrak 1981, Robbins 2000, Rodriguez ct ul. 
2007). Our survey protocol consisted of a main 
survey (Survey I or SI) based on unlimited radius 
point counts, and a complementary survey 
(Survey 2 or S2) based on cumulative species 
lists, applied on consecutive days (Rodriguez el 
al. 2011). The team leader was the primary 
observer, as the assistants in most cases did not 
have sufficient experience to identify all species 
m the region and were occupied recording the 
information, organizing the sheets, and perform¬ 
ing logistical or road safety tasks. However, in 
cases when the assistants detected and positively 
identified a species not reported by the team 
leader during the count time, it was recorded by 
making a note of the observer who detected it 
Records from the team leader and assistants were 
pooled for the results and analyses presented in 
this paper. Surveys started at first light which, 
depending on the habitat and weather was 
between 0 and 20 min before sunrise. We 
expected surveys to last no more than 5 hrs but 
m some locations, they were longer mostly due to 
poor conditions of the roads or obstacles on the 
way Counts started upon arrival at each stop to 
try to avoid delays. All 50 stops in SI were 
indivvf /'L U , 3 ’ min C ° Um al each st op. All 
ldividual birds seen or heard were recorded and 
notes were made for birds detected >100 m. large 
birds flying high overhead (mostly raptors or 
parrots), and the number of individuals detected 
by sight or sound. Birds not safely identified to 
species were recorded as unidentified and not 
included in the results and data analyses. S2 
covered a selection of 10 points and cumulative 
species lists were made for three 3-min penods 
without counting individuals. We traveled tran¬ 
sects for S2, whenever possible, in the opposite 
direction as for SI to sample individual stops 
along the transect at different times of the 
morning, as bird activity for most species usually 
decreases dramatically after the first 3-4 hrs after 
sunrise. S2 was intended to be complementary to 
the data of SI; thus, the selection of points to lx 
sampled for S2 was made after SI was conducted. 
Point selection was somewhat subjectively direct 
cd to cover the diversity of habitats within 
transects, giving priority to stops with forested 
habitats or water bodies or where it was easier or 
safer to stop the vehicle. All data collected during 
a day were entered into a data base on a laptop 
computer during the afternoon. The team leader 
also kept an extended list of all bird species seen 
or heard, either during the surveys or not 
(hereafter called ‘checklists'). No special effort 
was made to seek other species than those 
recorded in the surveys. 
Sampling Effort .—Total fieldwork time was 
-243 person-days (3 days/location X 27 locations 
X 3 persons). Bird sampling occupied —729 person- 
hrs (mean — 4.5 hrs/day x 2 days/location X 27 
locations X 3 persons). The total effective sampling 
effort, il we exclude driving time and consider that 
mostly only one person made the observations 
(—4% of all observations were not detected by the 
main observer), was 108 person-hrs (27 locations X 
12.5 hrs of SI + 1.5 hrs of S2] X 1 person). 
Potential Species Lists .—The total number of 
species for the country w'as taken from Hilt) 
(2003). 1 he list of potential (expected) species for 
each ecorcgion was estimated by compiling 
several local bibliographic sources (Lentino and 
Goodwin 1991, Goodwin and Lentino 1992. 
Lentino el al. 1996. Sharpe 1997, Rodriguez 
2000. Hilly 2003, Ascanio and Garcia 2005. Sanz 
el al. 2010) and from the lead author's unpub¬ 
lished field observations (GAR). Common and 
scientific names of all species mentioned follow 
Gill and Donsker (2011). 
We classified all potential species into two 
categories: ‘common’ and ‘uncommon or rare'. 
