Search

Abstract Over the last decades, numerous natural habitats have been converted into settlement areas, agricultural land, and tree plantations on a large spatial scale. As a result, natural ecosystems have been destroyed. In consequence, many ecosystems exist today as small and geographically isolated remnants. To what extent the original species diversity can persist in such small habitat patches is questionable and strongly depends on the ecology of the species. A prominent example of severe habitat destruction are the species-rich tropical cloud forests of Taita Hills in southern Kenya, which have been deforested almost completely during past decades. However, there still exist typical forest species in the few remaining forest fragments. In this study, we investigate the population ecology and behaviour of two butterfly species present in the cloud forest remnants of Taita Hills, Protogoniomorpha parhassus and Precis tugela. Over a period of one month, we conducted Mark-Release-Recapture to study population sizes and demographic structures, lifespan, dispersal, and behaviour. We found that both species exhibited medium population sizes and are sedentary. However, some individuals performed dispersal throughout the forest. The behaviour of the two species differs: While P. tugela was mostly observed basking with open wings, P. parhassus was mostly sitting under leaves with closed wings. The life span was rather long for butterflies. Implications for insect conservation This study documents the population ecology and behaviour of these two Afrotropical butterflies and underlines the relevance of the conservation of cloud forest remnants to preserve species, which mainly depends on these habitat remnants.

Genetic diversity is of crucial importance for population fitness and the potential of populations to adapt to environmental change. Population-level genetic diversity is expected to be reduced in specialists having small and isolated populations. We analysed genetic structure and diversity in (pre-)Alpine populations of four Copper butterfly species, differing in the degree of habitat specialisation. Despite substantial variation among genetic markers and molecular indices, genetic diversity and thus evolutionary potential tended to be highest in Lycaena hippothoe and lowest in L. tityrus. Microsatellite analyses revealed that genetic diversity tended to be higher in L. hippothoe and L. virgaureae than in L. helle and L. tityrus. MDH sequencing indicated only few differences among species, while GAPDH sequencing showed higher genetic diversity in L. hippothoe and L. tityrus than in L. virgaureae and L. helle. In L. helle, populations showed strong genetic differentiation but without clear spatial structure, while the other species showed genetic differentiation among clusters according to cardinal directions. These results do not match our a priori predictions of a high genetic diversity in L. tityrus and a particularly low one in the glacial relict L. helle. We thus document that evolutionary potential is not necessarily constrained in relict taxa, and that predicting genetic diversity is challenging. Conservation status does not necessarily reflect genetic diversity at the species level. It thus seems to be more meaningful to focus on specific populations of concern to assess genetic vulnerability to environmental change.

Isolated populations represent one of the main focuses in conservation biology. Long-term isolation often causes losses of genetic diversity and as a consequence might reduce individual fitness. Morphometric characters can be used as suitable markers to analyse ecological stress and individual fitness of local populations. Asymmetry in bilateral symmetry is used as a measure for developmental instability of populations and is often negatively correlated with population size and low genetic diversity. As a study system, we selected the endangered butterfly Parnassius apollo, which occurs in small and isolated remnant populations in Central Europe, but also in fairly large metapopulations in the Alps. We analysed wing morphometrics (shape and size characters) of 812 individuals representing (1) already extinct, (2) highly isolated and (3) still interconnected populations. Seventeen landmarks on veins were used to analyse morphological variances in the wing shape. Our data show significant deviations between landmarks on the left and right wing side within individuals and strong morphological variance among individuals. The highest morphological variability could be found for individuals in the Alps, however, the level of asymmetry was very similar for all populations analysed. The higher morphological variability found in the Alps can be interpreted as a consequence of the higher level of the genetic diversity detectable for this region. Analysis on morphological variance of P. apollo individuals of the Mosel valley using time series ranging from 1895 until today showed no significant rise in asymmetry and no decline of morphological variability over time, although, anthropogenic habitat destruction has caused severe bottlenecks in this population.

Agricultural intensification caused a fragmentation of flower-rich extensively used meadows which resulted in the reduction of the abundance of species of these habitats. The abundance pattern and the dispersal behaviour of species influence the connectivity of local populations. In this context exchange rates can be directly measured by mark release recapture (MRR) studies and indirectly by genetic analysis. Both approaches are used in our study in a comparative way. As a test species to investigate the influence of habitat interconnectivity on a local and regional scale, we selected the butterfly Melanargia galathea (Lepidoptera, Nymphalidae), a widely distributed species, which is common on flower-rich meadows in our study area in western Germany. We marked 3,175 individuals of four neighbouring sites in a mark-release-recapture study and analysed 18 allozyme loci for 644 individuals sampled over 17 sites. Only 3.3% of the total genetic variance was found among samples, thus supporting the detected between-patch movements. Both approaches revealed a high exchange rate among local populations. Moderate between- and high within-patch movements were recorded by MRR analysis, dependent on the geographical distance. The two analytical tools showed high estimated effective population sizes for all populations. In the light of conservation biology, the combination of the MRR and allozyme data support the assumption that high dispersal ability and habitat interconnectivity countervail genetic differentiation and enable the maintenance of a high level of genetic differentiation.

The habitats of many species are fragmented. Therefore, the survival in a metapopulation depends on the stability of the single populations and the amount of movements between patches. We chose the calcareous grassland specialist butterfly species Polyommatus coridon as a model. As study area, we selected a mosaic-like landscape in Rhineland-Palatinate (western Germany) with several well preserved calcareous grassland fragments. We marked a total of 2,211 individuals during July and August 2003. The overall recapture ratio was 7.1%. The estimated mean butterfly densities over the whole flight season ranged from 52 to 487 individuals per hectare. The within-patch movements were relatively low (13.3%) compared with the between-patch movements (3.2%). Therefore, the metapopulation structure appears to be intact in our study area.