render lVeil el al • AVIFAUNA IN AGROFORESTRY IN TABASCO, MEXICO 
563 
TABLE 4. Similarity of bird communities in five agroforestry classes in Tabasco, Mexico in 2008-2009. AF - 
agroforest; AA = animal agroforestry; LA = linear agroforestry; SA = sequential agroforestry; CT = crops under tree 
cover, S = sample; nso = number of species observed; sso = shared species observed; C-J = ( hao-Jaccard estimated 
similarity index. 
s s 
• 1 12 
Rainy season 
Northern winds 
Dry season 
All seasons 
nso 
#1 
nso 
#2 
SSO 
C-J 
nso 
a | 
nso 
1 2 
SSO 
C-J 
nso 
ft i 
nso 
ft 2 
SSO 
C-J 
nso 
# 1 
nso 
#2 
SSO 
C-J 
AF .AA 
27 
40 
18 
0.90 
31 
41 
21 
0.71 
40 
46 
23 
0.72 
53 
64 
38 
0.76 
AF AL 
27 
27 
10 
0.34 
31 
26 
15 
0.51 
40 
40 
22 
1.00 
53 
59 
35 
0.79 
AF AS 
27 
28 
17 
0.54 
31 
23 
14 
0.59 
40 
24 
16 
0.55 
53 
43 
33 
0.75 
AF CT 
27 
25 
14 
0.79 
31 
40 
22 
0.79 
40 
46 
28 
0.84 
53 
60 
39 
0.85 
AA AL 
40 
27 
20 
0.67 
41 
26 
17 
0.47 
46 
40 
25 
0.68 
64 
59 
44 
0.89 
AA AS 
40 
28 
21 
0.79 
41 
23 
20 
0.76 
46 
24 
19 
0.63 
64 
43 
36 
0.87 
AA a 
40 
25 
16 
0.62 
41 
40 
26 
0.64 
46 
46 
28 
0.71 
64 
60 
43 
0.81 
LA SA 
27 
28 
13 
0.50 
26 
23 
13 
0.48 
40 
24 
19 
0.65 
59 
43 
36 
0.98 
LA CT 
27 
25 
12 
0.44 
26 
40 
19 
0.62 
40 
46 
30 
1.00 
59 
60 
40 
0.85 
SA CT 
28 
25 
13 
0.51 
23 
40 
18 
0.72 
24 
46 
20 
0.71 
43 
60 
34 
0.81 
animal agroforestry (Mann-Whitney U- test, P = 
0.002) and crops under tree cover (Mann-Whitney 
6-test. P = 0.017). 
Only one of the 102 species recorded was a 
forest specialist. The Common Black-Hawk 
{Ruteogallus anthmcimis) (I individual) was 
observed in crops under tree cover. Seven 
individuals of four species with habitat preference 
for water bodies were observed. MosI species (75) 
were forest generalists; 22 species were specialist 
°f open areas. The richness of open area 
specialists varied among agroforestry classes 
'ANOVA, F = 5.060. P = 0.003) and was larger 
1(1 :in 'mal agroforestry than in agroforests, linear 
ngroforestry. and crops under tree cover (Tukey’s 
HSO-tests, P < 0.01 in all cases). The abundance 
"• specialists of open areas also varied among 
agroforestry classes (ANOVA. F = 5.685, P = 
^*1 '• and was larger in animal agroforestry than 
ln linear agroforestry and crops under tree cover 
( Tukey's HSD-tests, P = 0.001 and 0,039). and in 
^quential agroforestry than in linear agroforestry 
l|p ~ 0.048). It was also larger in agroforests than 
m linear agroforestry i P = 0.034). No significant 
diflerences were observed in the rainy season 
'Kruskal-Wallis tests. P - 0.163 for richness and 
p = 0.483 for abundance) when comparing 
r 'diness and abundance of forest generalists in 
£ ach season and in the season of northern winds 
'Kruskal-Wallis tests, P = 0.165 for richness and 
^ = 0.167 for abundance). Richness did not vary 
‘imong classes (Kruskal-Wullis test, P — 0.06) in 
dry season, but abundance varied (Kruskal- 
'Vallis test, P = 0.014). Specialists of open areas 
were more abundant in agroforests, animal 
agroforestry, and sequential agroforestry than in 
linear agroforestry and crops under tree cover 
(Mann-Whitney U- tests, P < 0.05). 
The richness and abundance of forest general¬ 
ists did not vary among agroforestry classes in the 
rainy season (Kruskal-Wallis test. P = 0.098 and 
P = 0.737). Richness did not vary among 
agroforestry classes in the season of northern 
winds (ANOVA, F = 1,054, P = 0.395). The F 
statistic indicated difference in abundance among 
classes, but Tukey’s HSD-test did not indicate 
significant differences between particular pairs of 
agrolorestry classes. Average abundance in this 
season was the same in agroforests and crops 
under tree cover (29.3). It was lowest in sequential 
agroforestry (15.8), and had an intermediate value 
in animal agroforestry (21.0). The pattern was 
similar in the dry season; there were no significant 
differences in richness and abundance between 
pairs of agroforestry classes. The highest average 
abundance was observed in agroforesLs, and the 
lowest in sequential agroforestry. Richness and 
abundance of forest generalists also did not vary 
among the agroforestry classes (ANOVA, F = 
2.651. P = 0.051) when data were summed over 
all seasons. Forest generalists were equally 
abundant in animal agroforestry and the other 
classes. Abundance of forest generalists in all 
agroforestry system classes, including animal 
agroforestry, was higher than abundance of open 
area specialists (Fig. 1). Differences between the 
number of forest generalists and open area 
specialists were largest in agroforests, linear 
