It is generally assumed that morphological structures like awns, bristles, or hooks increase retention potential of plant diaspores to the coats of animals, but this basic assumption has rarely been tested. In this paper we introduce a tab-experiment in which detachment rates of diaspores from mechanically shaken animal coats are measured using a standardized protocol and compared with detachment rates measured under natural conditions. The tab-experiments were used to assess 'retention potentials' of diaspores for more than 100 plant species. 'Retention potential' is defined as the proportion of diaspores still attached to the animal coat after a certain time period. We assessed the effect of several diaspore traits on retention potential and found that a strong negative effect of diaspore mass overrides the expected positive effect of structures Like awns, bristles or hooks. However, a positive effect of these structures was proven, when analysing the dataset separately for diaspores of different mass classes. Additionally, we found a negative effect of flat appendages. Coat type also affects retention potentials, with higher values in curly sheep wool compared to straight cattle hair. Species with diaspore mass <2 mg had a fair chance to be dispersed in curly wool as well as in straight hair over Long distances, once they get attached to the animal-coat. However, some species with diaspore mass > 2 mg might also be dispersed over long distances, at least by sheep. The frequency distribution of retention potentials clearly shows that a binary classification of species dispersed in animal hair vs. species not dispersed is artificial since differences are gradual. The relative importance of diaspore mass for determining retention potential was higher in straight hair, whereas diaspore morphology had a relatively higher impact in curly wool. Hence, different sets of plant species are dispersed by different animals. (C) 2005 Geseltschaft for Okotogie. Published by Elsevier GmbH. All rights reserved.