Morphological traits of diaspores that can predict their potential attachment to animal coats may help to model epizoochory in plant populations and communities. The present study tested the role of seed mass, shape and the presence of dispersal structures in the attachment potential of a sample of 14 abundant species in Mediterranean grassland and shrubland, using sheep and cattle coats that were shaken mechanically using a standardized protocol. We also tested recently proposed predictive models for the attachment potential of diaspores. Attachment potential measured on cattle hide was low in comparison with sheep wool for all types of diaspores. Differences between vertically and horizontally positioned coats were significant only for cattle, in which attachment potential was higher in vertically positioned coats. Seed weight was highly significant to predict attachment potential in sheep coats, but yielded significant results only in the case of vertically positioned cattle coats. In both cases, light seeds were best retained. Shape yielded marginally significant results only in the case of horizontally positioned cattle coats. In this case, elongated seeds seemed to be best retained. The presence of appendages in diaspores was significant only for sheep, in which attachment potential was higher for seeds with appendages. Recently proposed predictive models for the attachment potential were found to be highly robust for predicting this parameter in sheep coats, and thus support their generalization for this type of coat.

Epizoochory is widely recognized as an effective long-distance seed dispersal mechanism. Nevertheless, few studies have focused on the investigation of its influencing factors. One of the key aspects of epizoochory is the adhesive interaction between seeds and furs. We describe a new method to quantify experimentally and standardize the adhesivity of seeds to animal fur, as a measure of epizoochorous dispersal potential. The method excludes the impact of animal behaviour and environmental factors, and allows the ranking of species according to their adhesivity score. We measured adhesivity scores for 66 species on the furs of seven mammals. Deep furs with long, rough, undulated hairs implanted at a large angle were most suited for seed adhesion, while seeds adhered less well to shallow furs with short, smooth, straight hairs implanted at small angles. Seeds with specialized adhesive appendages had higher adhesivity scores than seeds with unspecialized appendages and seeds without appendages. However, an interaction effect between certain seed and fur types exists. Although seed morphology is a good predictor for seed adhesivity on fur, less well-adhering seed types often still have relatively high adhesivity scores. Therefore, it is likely that nearly all species are, to some extent, able to disperse epizoochorously.

A domestic dog was used in two experiments to elucidate the role of epizoochorous seed dispersal of forest plants. First, the dog was walked through forest vegetation at different times of year and its coat analysed for seeds retained within it. The seed content of the coat was compared to seed frequencies in the vegetation. Secondly, seeds of 11 plant species were placed in different positions on the dog, and their persistence in the coat analysed with respect to distance subsequently travelled. The experiments demonstrate that seeds with morphological adaptations to seed dispersal and small seeds of tall species can be caught effectively by a dog’s coat. The morphologically adapted seeds can be dispersed over large distances as long as the dog moves steadily along a road. The same is true for species with small and smooth seeds if they are deposited on the back of the dog, but not if they are placed on its side. Comparisons of these results with the distribution of forest species in a fragmented landscape, indicated that tall species with small seeds and species with morphological adaptations for epizoochorous dispersal are good at colonizing new forest habitats.

Rain forests at Los Tuxtlas are islands surrounded by anthropogenic vegetation containing invasive species. Epizoochory is an important dispersal syndrome among invasive species but ecological studies are scant. In order to evaluate the movement of epizoochore diaspores into the mature forest, we selected two abundant secondary herbs with contrasting diaspore morphology, and five dispersers with different pelage type. We used stuffed mammals as a dummy and passed them through patches of these two plant species, counting the number of adhered diaspores. For dispersal into the forest, diaspores were manually adhered to dispersers, these were pulled along 100-m forest transects, counting the number of diaspores remaining at distance intervals. We observed the highest amount of detachment in the first 20 m of the forest edge. For Desmodium incanum diaspores with sticky hairs, adhesion to dispersers showed no differences among pelage types, therefore dispersal is determined mainly by diaspore detachment. Dispersers with long pelage (Didelphis marsupialis, Nasua narica and Procyon lotor) move diaspores for shorter distances, while dispersers with short pelage (Silvilagus floridanus and Philander opossum) are more likely to move diaspores for longer distances. In Pavonia schiedeana more diaspores (seeds with spines and hooks) were attached and remained attached on dispersers with long and thick pelage. Mean distance travelled for P. schiedeana diaspores varied between 11 and 2700 m. In D. incanum the distance travelled did not exceed 500 m. Based on comparisons between dispersal patterns and on ecological data of these two plant species and their potential dispersers, P. schiedeana has a higher probability of invading the forest and finding suitable habitats for establishment than D. incanum.