Palms are important components in tropical forests, and their fruits and seeds serve as
keystone resources in fruit scarcity periods for most vertebrates and invertebrates, such as
beetles, which are among its main seed predators. Bruchid beetles are able to prey up to 100%
of palm seeds that remain under the parent plants, thus affecting the reproductive success of
several species. Forest fragmentation may affect patterns of palm seeds dispersal and
predation due to changes in abundance or loss of their dispersers. Increased predation by
beetles in forest fragments, where their dispersers are absent or have reduced populations, is
caused mainly by the fact that vertebrates do not remove fruits away from parent plants.
Therefore, it is expected that fragmentation also affects bruchine oviposition patterns due to
high accumulation of seeds under parent palms. Thus, this study evaluated the effects of forest
fragmentation on seed predation of Astrocaryum aculeatissimum by the bruchine Caryoborus
serripes, by comparing two small Atlantic Forest fragments (19 and 57 ha) and two large ones
(2400 and 3500 ha) located in the municipalities of Silva Jardim and Casimiro de Abreu, Rio
de Janeiro. We sampled 53 adult individuals of A. aculeatissimum, 33 in large fragments and
20 in small. To evaluate if bruchine seed predation rates were affected by the size of the
fragments and if size of the endocarp differed between large and small fragments the t test
was used. The Mann-Whitney test was used to verify if the number of bruchine cocoons per
endocarp differed between fragments. The influence of endocarp size and scolytine infestation
rates in number of cocoons was checked through linear regressions. The proportion of seeds
preyed upon by bruchine was higher in small fragments (small: mean 0.47 ± 0.21, large: 0.26
± 0.26; t = 3.12, p < 0.01). Considering the diameter of the collected endocarps, there was no
significant difference between the fragments (small = 32.4 ± 3.4 mm, large = 32.7 ± 5.7 mm; t
= 0.53, p > 0.05). The diameter have a positive and significant relationship with the number of
cocoons at small (r² = 0.11, F= 48.53, p < 0,0001) and large (r²= 0.07, F = 8.59, p < 0.01)
fragments. The number of cocoon per endocarp was lower in small fragments (small: 3.8 ±
2.3, large: 5.6 ± 2.9; U’= 32113, p < 0.0001). The number of scolytine entrance holes
(Coccotrypes sp.) per seed did not affected the number of C. serripes cocoons in the larger
fragments (F = 0.86, p > 0.05). However, at the small ones, the infestation rate by scolytines
have a negative and significant influence in the number of C. serripes cocoons (F = 12.87, p <
0.001). The found results indicate that forest fragmentation affected the seed predation rates
of A. aculeatissimum by C. serripes. Part of this result could be due to alterations in the
oviposition patterns of this beetle as at the small fragments were found less cocoons by
endocarp. The higher predation rates at small fragments, by its turn, suggest that in these
places the recruitment of new individuals of this palm could be negatively affected.
