This study examines the causes of emigration from small fragments of suitable habitat in a species that has a distinct metapopulation structure, frequent turnover of local populations, and substantial migration among local populations and currently unoccupied habitat fragments. We conducted a field experiment in which 727 individuals of the Glanville fritillary butterfly (Melitaea cinxia) originating from four regions were marked and released simultaneously in a natural environment. In three of the four source regions, larvae for the experiment were collected from dozens of small local populations, some of which had been established in the previous summer (new populations), whereas the remaining populations were older. In two of the source regions, female butterflies prefer a host plant (Veronica spicata) that is not present in the release area, where there is only Plantago lanceolata, the preferred host plant of females from the other two source regions. We found that migration rate of males was unrelated to any of the factors studied in this experiment. In contrast, two factors influenced the migration rate of females. First, Veronica‐preferring females had higher emigration rate than Plantago‐preferring females from the Plantago‐containing release patches, demonstrating that the individual perception of habitat quality significantly influences the migration rate of females. Second, females from newly‐established populations were more dispersive than females from older populations, supporting the notion that metapopulation processes (recurrent colonizations) select for increased migration. The observed migration rate was not correlated with any body size measurements, and thus the observed differences in migration rate were apparently caused by differences in the behaviour of female butterflies rather than in their flight capacity.

No abstract

No abstract

This study reports a rare example where a native herbivorous insect frequently overexploits local populations of its perennial host. Local dynamics of a flightless weevil (Hadramphus spinipennis, Curculionidae) and its host plant (Aciphylla dieffenbachii, Apiaceae) were assessed for one discrete patch. In this main study site local weevil population structure, dynamics and movement were investigated using a capture-recapture study. Local plant dynamics were studied by mapping plant location, size, sex and the phenological stage of each plant. Regional weevil and plant dynamics were studied for six plant patches using line-transect counts to estimate local weevil numbers and repeated counts of the number of flowering adult plants to assess plant numbers. Dispersal was assessed by regularly searching all plant patches for marked weevils that emigrated from the main study site. Prior to extinction, local weevil abundance, survival and recruitment rates increased continuously. At the same time the feeding damage on the plants increased and the area covered by A. dieffenbachii decreased until no plants were left. An increase in weevil abundance was clearly associated with the extinction of the local host plant population. Weevils stayed in their local host plant patch whilst food was available and dispersed only after local extinction of the plant. Over a 4-year period four local population extinctions were observed. This study was too short to allow explicit conclusions to be drawn about the ratio of extinction to colonisation rates for both the weevil and the host plant populations. However, persistence of this locally unstable system appears possible only in a fragmented habitat where asynchrony in local dynamics is maintained.

No abstract