Bocconea 7 - 1997 153 Reproductive biology and gene flow in a fragmented landscape Hans C. M. den Nijs & J. Gerard B. Oosterrneijer Gene flow Abstract den Nijs, H. C. M. & Oostenneijer, J. G. B.: Reproductive biology and gene tlow in a fragmented landscape. - Boeeonea 7: 153-165. 1997. - ISS N 11 20-4060. After a presentation of recent studies on gene tl ow and on pattems of genetie variation in pl ant s, a brief evalu ation is given of the eff ects of fr agmentation of th e landseape on th e genetic variation in pianI species and the consequential effects or in breed in g depression. Data show th at reduced fitness in small and isolated populations oceurs, ow in g to reduced reproduetive success (Iowe r fruiI or seed set) and beeause of inbreeding. It is also suggested th at the stili poorly studied mechani sm of outbreeding depression may also have nega ti ve effects. In partieular, outbreeding speeies that used to be widespread and common, suffer more from fr agmenta ti on and isolation than ende mi e and naturally rare species or predominant selfe rs. More emphasis shoul d be plaeed on th e eonservation of new rare species, beeause their ecotypic variation and manifold, but often undiscovered local adaptations may get lost, and eonsequently also the ir long-term evolutionary perspectives. Sincc Ehrlich & Raven (1969) expressed their doubt rega rding the importance of gene flow for the cohesion of species and for the evolutionary perspectives of taxa, much debate on thi s topic has arisen. The behaviour of foraging animals, thc dispersal of pollen and seeds by wind and animals, a nd th e isolating distances kept by seed growers also led Levin & Kerster (1974) to conclude that ge ne tlow must be limited, a nd could not account for the long di stance genetic coherence of plant species. Levin (1979) stated that many of th e research data point to a limited amount of ge ne tlow, which made him pose the question whether a pl ant species constitutes a coherent unit of evolution or just comprises a group of separate subunits sharing few genetic and eco logical characters (as for example found in Triticum turgidum by Nevo & a l. 1988, Nevo & Beiles 1989). Recent studies report very limited ge ne tlow distances indeed in several spec ie s: onli approximately 1.5 metres for Plantago (Bos & al. 1986, Ton sor 1985), and approximately 14 metres for Lathyrus (Godt & Ham ri ck 1993), Vogel & Westerkamp (1991) also assume