Summary1. Habitat fragmentation in agricultural landscapes has reduced the population sizes of many plant species while increasing their spatial isolation. Restoration or maintenance of the connectivity by gene flow between the fragmented patches may be determinant to sustaining viable populations, especially for insect-pollinated species. Functional biological corridors facilitating pollen flow between remnants in a human-dominated matrix might achieve this. 2. Dye dispersal was investigated for the extremely fragmented insect-pollinated herb Primula vulgaris, using fluorescent dye particles as pollen analogues, in a study site comprising 20 populations, of which 13 pairs were physically connected by a linear landscape elements (LLEs, ditches), and 11 pairs were not connected by an LLE. The dye deposition events were used to fit a model of pollen dispersal at the landscape level. We examined whether existing LLEs in the intensively used agricultural landscape act as functional corridors for pollen dispersal. The effects of LLE length and size and plant density of the recipient population on the dispersal patterns were tested. 3. Dye dispersal showed a leptokurtic decay distribution, with 80% of the dye transfers occurring at less than 85.1 m, and a maximal distance of 1010.8 m. The mean distance travelled by fluorescent dye particles based on the dye dispersal model was d = 87 m. 4. Dye dispersal between populations was found to be significantly higher when populations were connected by an LLE, than when populations were unconnected. For the group of population pairs connected by an LLE, dye deposition significantly decreased with the distance to dye source, but was not related to recipient population size and plant density. 5. Synthesis. Our study is, to our knowledge, the first to demonstrate that existing LLEs in an intensively used farmland may act as functional biological corridors facilitating pollen dispersal through pollinator movements. The maintenance or restoration of a network of populations connected by LLEs, but also by other landscape structures (e.g. population relays in vegetation patches and networks of small elements allowing indirect connections) should be strongly encouraged.

Habitat fragmentation is known to generally reduce the size of plant populations and increase their isolation, leading to genetic erosion and increased betweenpopulation genetic differentiation. In Flanders (northern Belgium) Primula vulgaris is very rare and declining. Populations have incurred strong fragmentation for the last decades and are now restricted to a few highly fragmented areas in an intensively used agricultural landscape. Previous studies showed that small populations of this long-lived perennial herb still maintained high levels of genetic variation and low genetic differentiation. This pattern can either indicate recent gene flow or represent historical variation. Therefore, we used polymorphic microsatellite loci to investigate genetic variation and structure in adult (which may still reflect historical variation) and seedling (recent generation, thus affected by current processes) life stages. The recent generation (seedlings) showed a significant loss of observed heterozygosity (H o ) together with lower expected heterozygosity (H e ), a trend for higher inbreeding levels (F IS ) and higher differentiation (F ST ) between populations compared to the adult generation. This might result from (1) a reduction in effective population size, (2) higher inbreeding levels in the seedlings, (3) a higher survival of heterozygotes over time due to a higher fitness of heterozygotes (heterosis) and/or a lower fitness of homozygotes (inbreeding depression), (4) overlapping generations in the adult life stage, or (5) a lack of establishment of new (inbred) adults from seedlings due to degraded habitat conditions. Combining restoration of both habitat quality and gene flow between populations may be indispensable to ensure a sustainable conservation of fragmented populations.

In flowering plants, pollen dispersal is often the major contributing component to gene flow, hence a key parameter in conservation genetics and population biology. A cost-effective method to assess pollen dispersal consists of monitoring the dispersal of fluorescent dyes used as pollen analogues. However, few comparisons between dye dispersal and realized pollen dispersal have been performed to validate the method. We investigated pollen dispersal in two small populations of the insect-pollinated herb Primula elatior from urban forest fragments using direct (paternity analyses based on microsatellite DNA markers) and indirect (fluorescent dyes) methods. We compared these methods using two approaches, testing for the difference between the distance distributions of observed dispersal events and estimating parameters of a dispersal model, and related these results to dye dispersal patterns in three large populations. Dye and realized (based on paternity inference) pollen dispersal showed exponential decay distributions, with 74.2-94.8% of the depositions occurring at <50 m and a few longer distance dispersal events (up to 151 m). No significant difference in curve shape was found between dye and realized pollen dispersal distributions. The best-fitting parameters characterizing the dye dispersal model were consistent with those obtained for realized pollen dispersal. Hence, the fluorescent dye method may be considered as reliable to infer realized pollen dispersal for forest herbs such as P. elatior. However, our simulations reveal that large sample sizes are needed to detect moderate differences between dye and realized pollen dispersal patterns because the estimation of dispersal parameters suffers low precision.

Despite the importance of tetraploid species, most population genetic studies deal with diploid ones because of difficulties in analysing codominant microsatellite data in tetraploid species. We developed a new software program-atetra-which combines both the rigorous method of enumeration for small data sets and Monte Carlo simulations for large ones. We discuss the added value of atetra by comparing its precision, stability and calculation time for different population sizes with those obtained from previous software programs tetrasat and tetra. The influence of the number of simulations on the calculation stability is also investigated. atetra and tetrasat proved to be more precise when compared with tetra, which, however, remains faster. atetra has the same precision than tetrasat, but is much faster, can handle an infinite number of partial heterozygotes and calculates more genetic variables. The more user-friendly interface of atetra reduces possible mistakes.

Ceratophyllum spp., Callitriche spp., Zannichellia spp. and Potamogeton pectinatus L. are widespread submerged macrophyte species, often occurring at high abundance and forming an integral part of the vegetation of many types of shallow aquatic systems. Several species occur in both freshwater and brackish water habitats. Most have a mixed reproduction system and can reproduce sexually by seeds and propagate asexually by rhizomes, turions, root tubers or axillary tubers. It is hypothesized that sexual propagules are more important than vegetative fragments to ensure long-distancedispersal, which in case of frequent bird or water flow-mediated dispersal should lead to lowered genetic differentiation. At a regional level, we used dominant ISSR markers in a multi-species approach and observed the largest clonal differentiation between brackish water and freshwater populations of the western European lowland (Belgium). Differentiation was pronounced at taxon level (e.g. Zannichellia), as a salinity gradient (P. pectinatus) or as a coastal-to-inland conductivity gradient (Callitriche obtusangula). These differences and trends suggested a very limited dispersal at regional level across both habitats and regions. To test the hypothesis whether vegetative reproduction and dispersal may have an important function in maintenance of the species at local scale, we investigated the microsatellite diversity and clonal distribution within and between populations of P. pectinatus from a single catchment, representing upstream forest ponds and downstream river sites along the Woluwe (Brussels, Belgium). Clonal diversity was low on average, however, with a higher number of multilocus genotypes in upstream forest ponds than in downstream river sites. A few but abundant clones were present along various stretches of the river indicating clonal spread and establishment over larger distances within the river. Clonal dispersal at a local scale was more pronounced in river than in pond habitats, indicating a higher relative importance of water flow than birdmediated dispersal for establishment of P. pectinatus in river sites. Dispersal of seeds and establishment of seedlings were assumed more effective within ponds than in river habitats. Upstream forest ponds can be regarded as source populations and refuges of clonal diversity for recolonization of the more stressful downstream river habitat.