xander Wal el al. • AVIFAUNA IN AGROFORESTRY IN TABASC O. MEXICO 
567 
where 25% of resident species were forest 
specialists (Arriaga-Weiss 2008). Seven percent 
of species and individuals in the matrix of pasture 
lands near forest fragments were forest specialists. 
Our study confmns die shift from forest specialist 
species in forests and fragmented landscapes to 
generalist species in agroforestry systems as 
documented by several authors (Reitsma et al. 
2001, Hughes et ai. 2002, Cardenas et al. 2003. 
Beukema el al. 2007, Harvey and Gonzalez- 
Villalobos 2007). 
We did not find a relationship between 
structural complexity of agroforestry systems 
and bird species richness. Agroforests had similar 
species richness and abundance as crops under 
tree cover, and forest specialist species were levs 
in both classes. Noise, presence of humans, and 
fowl and other domesticated animals, as well as 
the intense management of the agroforest floor 
resulted in the absence of forest understory 
specialists. Understory specialists were also 
absent in cacao in thinned forest in Brazil (Faria 
et al. 2006) and mate tea plantations under natural 
shade in Paraguay (Cockle et al. 2005). farm 
management has been shown to strongly influence 
avian diversity (Lang el al. 2003, Najera and 
Simonetti 2010). Species richness was not lower 
in animal agroforestry than in agroforests and 
crops under tree cover. 
Our results confirm those of Naidoo el al. 
(2004), which indicated that agroforestry pro¬ 
grams would not increase tree densities to levels 
that would support conservation of forest special¬ 
ists. This is also true for agroforestry systems with 
high tree cover, such as agroforests and crops 
under tree cover. Agroforestry systems would 
only benefit conservation of forest specialists il 
they are close to existing forests (Naidoo 2004). 
Keitsma et al. (2001) found the number of forest 
•'pecialists in cacao groves was negatively corre¬ 
lated with distance to forests, and concluded 
cacao groves cannot substitute lor lorest in the 
conservation of avifauna, although they provide 
habitat for many species that depend on lorests. 
Our results generalize this finding to all agrotor- 
es| ry classes in Tabasco. 
The distribution of species over foraging guilds 
indicates the wide ecological functionality ot all 
ugroforcstry classes (Fig. 2) and the value ol these 
systems for the conservation ot common and (lew ) 
care species. The distribution of birds over guilds in 
lorest fragments was. however, quite dilterent Irom 
those in forest fragments (Arriaga-Weiss et al. 
2008). Arboreal frugivores and arboreal insecti- 
vores/frugivores were more abundant in loiest 
fragments than in agroforestry systems (this study), 
whereas granivorcs were absent in lorest tiag- 
ments. Agroforestry system classes had several 
differences in abundance and species richness of 
foraging guilds. Omnivores were most abundant in 
animal agroforestry and agrotorests. Bark-gleaning 
insectivores were most abundant in crops under 
tree cover. Most raptor species were in crops under 
tree cover, and most leaf-gleaning insectivore 
species were in linear agroforestry and crops under 
tree cover. Most species in animal agroforestry 
were sweeping and sallying insectivores. leal- 
gleaning insectivores. arboreal insectivores, and 
tree frugivores. This differs Iron) Cardenas et al. 
(2003), who found mainly carnivores, granivores, 
and omnivores in animal agroforestry. These 
differences may be due to the abundance of insects 
in the wet climate of Tabasco, distinct species 
composition of trees in both study areas, and 
differences in tolerance of foraging guilds to 
fragmented landscapes (Tscharotke et al. 2008). 
The ecological functionality of the bird com¬ 
munity in agroforestry systems, indicated by the 
distribution "of birds over all foraging guilds in 
all systems, as well as diversity, can be further 
strengthened through management (Lang et al. 
2003). Policies regarding bird conservation in 
agroforestry systems should enhance properties 
that attract bird taxa of interest (Bcrges et al. 
2010). Thus, management of the forest tloor in 
agroforestry systems should be further studied, as 
well as the effect ol' including shrubs in live 
fences (Bernier-Leduc et al. 2009), and epiphytes 
(Beukema cl al. 2007. Gordon el al. 2007) in trees. 
Trees of complementary phenology could be 
integrated into agroforestry systems to assure 
availability of food for frugivores throughout the 
year. Experiments should be conducted on how to 
increase (he contribution from different agrofor¬ 
estry systems to conservation ot avitauna. 
ACKNOWLEDGMENTS 
We thunk farmers for letting ns conduct bird censuses on 
their lands. This research was financed by the National 
Council of Science and Technology and the Tabasco State 
Government through grant TAB-2006-C08-43S67 for the 
project -Eeophysiology and productivity ot agrolorestry 
systems in Tabasco*. Elvia Jimenez gave valuable com¬ 
ments on earlier versions. Gilberto Villanueva Lope/ 
participated in field work and Juan Carlos Jimenez provided 
logistic support. We thank R. K. Colwell and K. A. Babb 
Stanley for constructive comments on an earlier dralt ol the 
