The effect of ants on the seed dispersal cycle of the typical myrmecochorous Ricinus communis

Author(s)

Martins, Valeria Forni ^[1] ; Guimaraes, Jr., Paulo Roberto ^[1] ; Baptista Haddad, Claudia Regina ^[2] ; Semir, Joao ^[2]

Affiliation

^[1] Univ Estadual Campinas, UNICAMP, Inst Biol, Programa Posgrad Ecol, BR-13083970 Campinas, SP - Brazil

^[2] Univ Estadual Campinas, UNICAMP, Inst Biol, Dept Plant Biol, BR-13083970 Campinas, SP - Brazil

• Document type: Journal article
• Source: PLANT ECOLOGY; v. 205, n. 2, p. 213-222, DEC 2009.
• Web of Science Citations: 8

Abstract

The directed dispersal hypothesis has two components: (1) non-random seed deposition by a predicted vector, which removes greater amounts of seeds to specific sites, and (2) higher seed survival and seedling establishment in these specific sites. Several studies suggest that ants perform both tasks. This study was designed to cover the processes from post-dispersed seeds to established juvenile plants of a typical ant-dispersed species. Our main objective was to determine whether Ricinus communis benefits from directed dispersal by ants to maintain its populations in previously colonized habitats. We examined whether there were differences between ant nest pile mounds and their vicinities in the: (1) densities of seeds with and without elaiosome, seedlings and juveniles; (2) performance of post-dispersed seeds (without elaiosome), which may be affected by seed density, a key feature attracting seed predators; (3) nutrient quantities; (4) number of germinated seeds and juvenile biomass measurements; and (5) ant protection of seedlings from herbivores. There were more seeds without elaiosome, seedlings and juveniles in pile mounds, and seeds with elaiosome were equally distributed. There was no difference in the number of non-removed seeds in pile mounds and in their vicinities, and there was no tendency for this difference to increase or decrease with time or with initial seed density. Apparently, there was no difference in nutrient contents in soils of pile mounds and their vicinities. Likewise, there was no difference in the number of seeds germinated and in the biomass measurements of juveniles in both soils. Ants did not provide differential protection for seedlings in pile mounds against potential herbivores. The dispersal of Ricinus seeds by ants had a marked effect on the distribution pattern of the seeds, seedlings and juveniles of this species. However, there were no additional advantages for the seeds, seedlings and juveniles in pile mounds and, therefore, Ricinus does not benefit from directed dispersal by ants to maintain its populations in the study sites. (AU)