We determined the seed dispersa] rate of two species of Spondias (S. mombin and S. venulosa) by woolly monkeys at Tinigua National Park, Colombia. Focal sampling was used to quantify the dispersa] pattems during one year. Woolly monkeys disperse seeds of S. mombin more frequently than seeds of S. venulosa (11.2 vs. 0.4 seeds per hour of fruit consumption). In order to find out if differences in shape were correlated with the differences in dispersa] rates of the two species of seeds, we compared severa] morphological features of a sample of 587 seeds and 65 whole fruits, belonging to a total of 23 individualtrees. As seed. width was differerit in the two species; on average, seeds of S. venulosa were wider thanS. mombin seeds. Additionally, a sample of seeds collected frorn the rnonkey's feces indlcated that, on average, seeds dispersed by the rnonkeys are ·srnaller than those available in the forest. The rnorphological differences between the fruits and seeds of the two species of Spondias. are rtpt Iikely to be dlie to .allornetric relationships. The interspecific variation was srnaller than that found between trees of the same species, suggesting a high heritable cornponent in the rnorphological cháracteristics Then, it is possible that woolly rnonkeys in this cornrnunity put a significant selective pressure on the fruit rnorphology of Spondias species. Given their dispersa! patterns, probably they are affecting also the population densities of the plants. Further work on the effect of woolly rnonkeys and other dispersa! agents on the reproductive success of the dispersed is necessary to conf"mn our interpretations.

The activity of dung beetles (Scarabaeidae) associated with woolly monkey (Lagothrix lagothricha) dung, assecondary seed dispersal agents, was studied in Tinigua National Park, Meta, Colombia. From 62 dung sampleswe estimated a mean range of 30-33 small seeds, 3-5 medium seeds and 0-1 large seeds by fragment. The meanduration from the deposition of each dung fragment to dung beetle arrival was 3,2 hours. The means of horizontal17Universítas Scientiarum Vol. 7, N° 1: 17-29and vertical seed postdispersal distance was variable depending on dung beetle type: For the ball roller: 60,6 cm(±1 ,7 cm) and 0,8 cm (±o,2 cm) respectively; for fragment roller: 10,2 cm(± 0,2 cm) and 5,5 cm (±o,6cm) and forthe burrower's vertical remotion was O, 1 cm without horizontal remotion. The longest horizontal removal distan cewas presented by a ball roller.beetle Cantlwn luteicol/is (272 cm). Of 74 individual dung beetles observed, 37were catalogued as piece rollers, 27 as ball rollers and 10 as strict burrowers. The mean arrival times to dungpieces was smaller for rollers than for burrowers. We observed selection of dung material, as well as cooperationand competition between species and within species by dung beetles. The remotion distance of seeds resultedinversely proportional with its size. These observations suggest that not only the initial dispersa! by woolly monkeysand the posterior distribution of dung in the moment of deposition play an important role linked to the success orfailure of the seeds, but the relocation by the dung beetles also contributed to the general dispersa! process.

An intensive study ofseed dispersa! by woolly monkeys (Lagothrit /agotriclza) was made at Tinigua National Park, Colombia. The feeding time in different fruit species recorded o ver focal animals. The number of seeds of each species found in depositions and the spatiallocation of feces was also noted. Results show that woolly monkeys play a majar role in seed dispersa!, because they defecate intact seeds of most of the fruit species consumed. Further indications. that woolly monkeys act as efficient dispersa! agents are their prolonged retention time through the digestive tract, their extensive use of the home range, the large number of depositions with one or two seeds per species and the fact that dispersed seeds are capable to germinate.

The duration of feeding bouts on fruiting trees by animal dispersers may have an effect on seed dispersal effectiveness. In this study we quantified seed removal rates and visit duration for different animal vectors in three trees of Henriettellafissanthera (Melastomataceae). The main purpose was to describe seed removal rates by different dispersers, ánd test the associatiop. of three characteristics of the dipersors that ha ve been proposed to affect bout duration times: crypsis, diet and gut capacity. We observed three primate species and 24 bird species ingesting fruits during a period of 85 h. H. fissanthera fruits correspond to the birdE dispersal syndrome, and they removed the majority of seeds, although primates removed a significant proportion offruits (40%). The largest tree was visited more often than the medium and small trees. Our results indicated that neither crypsis or diet were associated with the average duration of feeding bouts. Probably these results can be explained by the fact that the frugivories are not very exposed to aerial predators when feeding on H. fissanthera fruits, and because particular fruit preferences by frugivories seem to be very important in determining visit times. We found a positive association between average visit time and body weight (as a surrogate variable for gut capacity), but this correlation was not significant, suggesting that other factors may be more important deterrnining the duration of feeding bouts in this plant species ( e.g. instantaneous fruit abundance, tree distribution and interspecific aggressions ).