Coexisting plant species often share the same animal mutualists and enemies. The selection preference of an animal on a focal individual is not only directly affected by the focal individual's intrinsic traits, but also indirectly affected by conspecific or interspecific neighbouring individuals. Many studies often show inconsistent neighbour effects by comparison of the results between the presence and absence of neighbours, without considering the contrasting functional traits between the focal individual and the neighbours. Rodent-seed interaction that is ubiquitous in many ecosystems was selected to test the neighbour effect by using artificial seeds with different tannin contents, a vital seed trait that directly influences rodent foraging preference. I compared the seed predation and dispersal preference of rodents by monitoring the fate of 1920 artificial seeds with different absolute tannin contents (from 0 to 25%) among three categories of neighbouring seeds with tannin contents of 0%, 10% and 25%, respectively. Obvious neighbour effects were observed, although the magnitude and direction of the effects differed among different processes of scatter-hoarding behaviour. Furthermore, the neighbour effect did not show a consistent pattern among seeds with different tannin contents. Our results strongly suggested that the neighbour effects of coexisting species should be considered when evaluating the foraging preferences of animal mutualists/predators upon a specific plant species. The contrasting functional traits between focal plants and neighbouring individuals should be considered when studying neighbour effects. Significance statement By using the artificial seed study system, I found that the contrast of tannin content could explain the inconsistent patterns of neighbour effects (i.e. the fate of a specific seed is influenced by its neighbouring seeds) among different seeds. In natural forest systems, plant seeds have a diverse array of traits, all of which markedly affect rodent foraging behaviour. Furthermore, a large spatio-temporal variation in seed production among species may result in a distinct variation in the relative densities of different seeds, which in turn may lead to even more complex neighbour effects at the whole forest community level. Our results strongly suggested that the neighbour effects of coexisting species of seeds should be considered when evaluating seed dispersal and predation of a specific species. The contrasting seed traits between focal plants and neighbouring individuals should be considered when studying neighbour effects.