Dispersal ability is of great importance for plants, which commonly occupy spatially and temporally limited substrate patches. Mixed reproductive strategies with abundant diaspore production are favoured in a heterogeneous landscape to ensure successful colonisation at different distances. In bryophytes, long-distance dispersal has been thought to take place primarily by spores, while asexual propagules are important in local dispersal and in the maintenance of colonies. In the present study, we investigated the dispersal potential of two equally sized propagules, sexually formed spores and asexually produced gemmae in the dioecious, epixylic hepatic, Anastrophyllum hellerianum, which inhabits spatially and temporally limited substrate patches. We trapped propagules at different distances (0-10 m) and directions from the source colonies in two experiments: one in a natural habitat within a forest and another involving an artificial set-up in an open habitat. Spore dispersal showed only slight distance dependence both in the open and the forest habitats, presumably as a consequence of wind affecting the dispersal pattern. Gemma dispersal was more strongly distance-dependent in the open habitat than in the forest sites. Considerably more gemmae were deposited during rainy than dry periods, possibly because of the effect of rain drops on gemma release. However, weather conditions had no effect on the dispersal patterns of spores or gemmae. In A. hellerianum, the combination of occasional spore production and practically continuous, massive gemma production facilitates dispersal both on local scale and over long distances. Unlike previously assumed, not only spores but also the asexual propagules may contribute to longdistance dispersal, thus allowing considerable gene flow at the landscape level.

No abstract

No abstract

Fine-scale spatial genetic structure (SGS) of the liverwort, Barbilophozia attenuata, occupying an area characterized by a network of ant trails, was investigated using microsatellite markers. This is the first study investigating SGS in a liverwort. Significant genetic differentiation was detected among colonies along and outside ant trails, and the SGS pattern varied depending on the spatial scale. At short distances, kinship coefficients were significantly positive up to about eight meters, after which they approached zero and turned negative, while at distances greater than 25 m the values were about zero. Thus, nearby individuals are more closely related than expected, at mid-distances less related, and at great distances genotypes show a random distribution. We suggest that the reproductive mode strongly affects SGS in B. attenuata. Asexual propagation by relatively large gemmae allows more effective establishment than sexual reproduction by small-sized spores, and causes an aggregation of similar genotypes, although the inbreeding effect cannot be ruled out. In environments with small-scale disturbances, e.g., ant trails, gemmae are favoured over spores at establishment. Also, the diaspore bank of the forest floor can be activated by disturbances, which may affect SGS. At mid-distances, the isolation by distance effect, presumably related to comparatively ineffective gemma dispersal, is visible, while at greater distances, the role of spores as effective means of dispersal is evident. The Sp statistic values, which quantify the strength of SGS, indicate that outsider colonies possess less SGS than do plant colonies along ant trails, which relates to the more frequent spore production of outsider colonies. Moreover, dispersal from fallen logs or stumps may be more effective than dispersal from ground-level colonies along ant trails.

No abstract