BIOTROPICA 38(6): 778–781 2006 10.1111/j.1744-7429.2006.00209.x The Effect of Dung and Dispersal on Postdispersal Seed Predation of Attalea phalerata (Arecaceae) by Bruchid Beetles 1 Rodrigo S. Rios 2 Herbario Nacional de Bolivia, Instituto de Ecolog´ ıa, Casilla 10077, Universidad Mayor de San Andres, La Paz, Bolivia and Luis F. Pacheco Centro de Postgrado en Ecolog´ ıa y Conservaci ´ on, Instituto de Ecolog´ ıa, Casilla 10077, Universidad Mayor de San Andres, La Paz, Bolivia and Centro de Estudios en Biolog´ ıa Te´ orica y Aplicada–BIOTA, Casilla 9641, La Paz, Bolivia ABSTRACT Low postdispersal mortality of palm seeds in tapir dung is hypothesized to result from the mechanical barrier provided by dung against bruchid infestation and/or from the distance to adult palms at which seeds are dispersed. We tested these hypotheses by distributing endocarps of Attalea phalerata Mart. ex Spreng. in experimental dung piles in Beni, Bolivia. Predation rates were significantly lower for seeds covered by dung than for exposed or partially covered seeds, but did not differ between seeds placed below and 50 m away from palms. Thus, dung, not short-distance dispersal, protects seeds against bruchid beetles, and may ultimately promote survival of palm seeds. RES ´ UMEN La baja mortalidad post dispersi´ on de las semillas de la palmera Attalea phalerata Mart. ex Spreng. encontrada en heces de tapir puede ser el resultado de la barrera mec´ anica que brinda la materia fecal contra la infestaci´ on por br´ uquidos y/o de la distancia a la que son dispersadas las semillas. Pusimos a prueba estas hip´ otesis distribuyendo endocarpos de A. phalerata en pilas fecales experimentales en el Beni, Bolivia. Las tasas de depredaci´ on fueron significativamente menores en semillas completamente cubiertas por materia fecal que en semillas limpias o parcialmente recubiertas, pero no variaron entre semillas debajo y a 50 m de las palmeras madre. Por lo tanto, las heces y no la dispersi´ on de corta distancia protegen a las semillas de br´ uquidos, promoviendo as´ ı la sobrevivencia de semillas. Key words: Attalea phalerata; Bolivia; bruchids; postdispersal seed predation; seed fate; tropical rain forest islands. IN TROPICAL FORESTS, SEED PREDATION IS A PREVALENT PROCESS and one of the most important factors determining seed survival (Howe & Smallwood 1982, Schupp 1988, Crawley 2000). Seed predation levels can approach 100 percent for some species (Wenny 2000, Travaset et al. 2003, Russo & Augspurger 2004). There are, however, instances where levels of seed predation have been found to be very low. Examples include many Neotropical arboreal palms where low postdispersal seed mortality has been found in tapir dung piles. For example, in Attalea maripa only 2 percent of the endocarps in dung piles were infested by bruchid beetles compared to a 77 percent infestation under parental trees (Fragoso 1994, 1997). Similarly, in Attalea phalerata only 6 percent of the endocarps encased in dung piles were infested while 61 percent of the endocarps under parental trees were infested by bruchids (Quiroga-Castro & Rold´ an 2001). At least three hypotheses have been suggested to explain low postdispersal predation of palms seeds in feces. First, feces are a compact mass of organic material that can conceal most seeds in- side; therefore, dung material could serve as a mechanical barrier 1 Received 6 July 2005; revision accepted 23 January 2006. 2 Corresponding author. Current address: Department of Biology, Saint Louis University, 3507 Laclede Avenue St. Louis, MO 63103, U.S.A.; e-mail: riosr@slu.edu against infestation by preventing bruchid penetration to the en- docarps (Fragoso 1994, 1997; Quiroga & Rold´ an 2001; Silvius & Fragoso 2002). Fragoso et al. (2003) found that endocarps of A. maripa were significantly less likely to be infested by bruchids if they were placed in tapir feces; however, because they did not account for seed mortality in endocarps and did not compare predation levels of endocarps inside dung piles to those exposed to the outside, the role dung plays in reducing seed predation remains uncertain. Second, low-predation rates may result from reduced distance- or density- dependent mortality (Janzen 1970, Connell 1971) because feces are typically deposited far away from adult trees (Fragoso 1994, 1997), e.g., in latrines 4–7 km away from conspecific palm aggregations (Fragoso et al. 2003); however, there are no available data on seed fate in fecal samples for shorter dispersal events where the majority of dispersers move seeds (Fragoso 1994, 1997; Silvius & Fragoso 2002). Third, if oviposition by bruchids occurs before fruits are consumed by the dispersal agent, the passage through the digestive tract could kill eggs or developing larvae, resulting in the low seed mortality found in dung piles (Fragoso 1997). This last hypothesis, however, is unlikely to play a major role. Fruits of arboreal palms are significantly less infested by bruchids before consumption by frugivores because removal of the epicarp is crucial for bruchid in- festation (Wright 1983). Rios (1998) found that only 18 percent 778 C  2006 The Author(s) Journal compilation C  2006 by The Association for Tropical Biology and Conservation