yonder Wal et al. • AVIFAUNA IN AGROFORESTRY IN TABASCO, MEXICO 
559 
may be associated with distance to forests as 
indicated by the authors. Najera and Simonelli 
12010 ) experimentally showed that higher struc¬ 
tural diversity in plantations of African oil palm 
l Elaeh guineensis), obtained by integrating a 
herbaceous and a shrub stratum, increased species 
richness and abundance of the associated avifau¬ 
na However, there were no forest specialists 
among the observed species. Reitsma et al. (2001) 
found a weak correlation between number of tree 
species and trees in the canopy of cacao groves in 
Costa Rica, and the abundance of forest specialists 
and generalists. These parameters together ex¬ 
plained only 24 c t of the variation. Indications ol a 
positive correlation between system structural 
complexity and bird species richness have been 
found, but results are too few, and not sufficiently 
robust to be conclusive. 
Bird communities in agroforestry systems were 
as abundant and species rich as in forests in 
TaJamanca, Costa Rica. However, fewer forest 
specialist species and more open area specialist 
species occurred in agroforestry systems (Harvey 
and Gonzalez-Villalobos 2007). No differences 
among agroforestry systems were found in 
abundance, species richness, and habitat preler- 
ence. Cacao groves in the same region had tower 
forest specialists than forests (Reitsma et al. 
2001). Rubber (Hevea brasitknsis) agroforests 
in Sumatra, Indonesia had fewer forest specialist 
species and more forest generalist species com¬ 
pared to native forests (Bcukemu ct al. 2007). 
file avifauna in a landscape of pasture lands near 
forest fragments in Tabasco, Mexico consisted 
mainly of forest generalists and open area 
specialists, whereas forest specialists were scarce 
•Arriaga-Weiss el al, 2008). The number of 
specialist bird species of the forest understory 
s pecies was small in comparison w ith forests in 
coffee crops under shade trees in Mexico ( Tejeda- 
Cruz and Sutherland 2004). cacao interplanted in 
'orest in Brazil (Faria et al. 2006), and •mate’ tea 
' Hexparaguayensis) under tree cover in Paraguay 
(Cockle et al. 2005). This is also one of the 
findings of Bhagwat ct al. (2008) in their review 
o! research on birds near forest reserves. 
Studies that compare avifauna between agro¬ 
forestry classes are scarce. 1 hesc studies are par¬ 
ticularly of interest where the original vegetation 
cover has been replaced almost completely, as is 
the case in the tropical lowlands o! Tabasco, 
Mexico(Arriaga-Weiss ct al. 2008). Ditterences in 
avian diversity and abundance among agroforestry 
classes would suggest which classes and which 
structural assets should be stimulated most through 
policies to obtain maximum conservation benefits 
of associated biodiversity. 
We compared avifauna among live classes of 
agroforestry systems distinguished by Torquebiau 
(2000). which varied in structural complexity. 
Agroforests resemble forests as they are composed 
of plants of different species and different growth 
forms without a fixed arrangement (e.g.. home 
gardens). Agroforests have the highest structural 
complexity (Torquebiau 2000. Kumar and Nair 
2004). Crops under tree cover are defined by a 
tree upperstory that shades a crop (e.g.. cacao and 
coffee). This class has lower structural complexity 
than agroforests, as the height of trees is 
homogeneous in the upper and lower stratum, 
and trees are generally at fixed distances. Trees 
and crops appear side by side in agrofarestry in a 
linear arrangement. Structural complexity is low, 
as high and low components are separated and 
regularly spaced. Trees and crops in sequential 
agrofarestry do not simultaneously occupy a 
field; cropped and recently abandoned fields are 
structurally simple, relatively open areas. Tree 
species in animal agrofarestry are used to teed 
ruminants. Tree cover is sparse in open fields and 
structural complexity is low. Our objectives were 
to investigate if: (I) species abundance and 
richness of bird communities vary among agro- 
lorcstry classes in the three climatic seasons in 
Tabasco, and (2) (he abundance and species 
richness of foraging guilds and habitat preferences 
vary among agroforestry classes. 
METHODS 
Study Area. —The State of Tabasco in south¬ 
eastern Mexico borders the Gull of Mexico 
between the states of Veracruz and Campeche. It 
is bordered on the south by Chiapas and to the east 
by Guatemala. Its territory consists mainly ot 
tropical lowlands, deposited in the Holocene by 
the Grijalva and Usumacinta rivers (West et al. 
1985). Extensive wetlands occur, roughly paral¬ 
leling the east-west coast line. Alluvial plains are 
succeeded by hills of weathered Pleistocene 
fluvial deposits to the south, and Tertiary 
mountains of sedimentary and volcanic rocks. 
Rainfall is 1.500 mm per year near the coast and 
up to 5.000 mm in the mountains. Three seasons 
are distinguished. The wet season from June to 
November is followed by a season of northern 
winds from December to March, which brings 
