Habitat fragmentation and destruction due to anthropogenic land use are the major causes of the increasing extinction risk of many species and have a detrimental impact on animal populations in numerous ways. The long-term survival and stability of spatially structured populations in fragmented landscapes largely depends on the colonisation of habitat patches and the exchange of individuals and genes between patches. The degree of inter-patch dispersal, in turn, depends on the dispersal ability of a species (i.e. the combination of physiological and morphological factors that facilitate dispersal) and the landscape structure (i.e. the nature of the landscape matrix or the spatial configuration of habitat patches). As fragmentation of landscapes is increasing and the number of species is continuously declining, a thorough understanding of the causes and consequences of dispersal is essential for managing natural populations and developing effective conservation strategies.
In the context of animal dispersal, movement behaviour is intensively investigated with capture-mark-recapture studies. For the analysis of such experiments, the influence of marking technique, handling and translocation of marked animals on movement pattern is of crucial importance since it may mask the effects of the main research question. Chapter 2 of this thesis presents a capture-mark-recapture study investigating the effect of translocation on the movement behaviour of the blue-winged grasshopper Oedipoda caerulescens. Transferring individuals of this grasshopper species to suitable but unfamilliar sites has a significant influence on their movement behaviour. Translocated individuals moved longer distances, showed smaller daily turning angles, and thus their movements were more directed than those of resident individuals. The effect of translocation was most pronounced on the first day of the experiment, but may persist for longer. On average, daily moved distances of translocated individuals were about 50 % longer than that of resident individuals because they have been transferred to an unfamiliar habitat patch. Depending on experiment duration, this leads to considerable differences in net displacement between translocated and resident individuals. In summary, the results presented in chapter 2 clearly point out that translocation effects should not be disregarded in future studies on arthropod movement, respectively dispersal. Studies not controlling for possible translocation effects may result in false predictions of dispersal behaviour, habitat detection capability or habitat preferences.
Beside direct field observations via capture-mark-recapture methods, genetic markers can be used to investigate animal dispersal. Chapter 3 presents data on the genetic structure of populations of Metrioptera bicolor, a wing-dimorphic bush cricket, in a spatially structured landscape with patches of suitable habitat distributed within a diverse matrix of different habitat types. Using microsatellite markers, the effects of geographic distance and different matrix types on the genetic differentiation among 24 local populations was assessed. The results of this study clearly indicate that for M. bicolor the isolation of local populations severely depends on the type of surrounding matrix. The presence of forest and a river running through the study area was positively correlated with the extent of genetic differentiation between populations. This indicates that both matrix types severely impede gene flow and the exchange of individuals between local populations of this bush cricket. In addition, for a subsample of populations which were separated only by arable land or settlements, a significant positive correlation between pairwise genetic and geographic distances exists. For the complete data set, this correlation could not be found. This is most probably due to the adverse effect of forest and river on gene flow which dominates the effect of geographic distance in the limited set of patches investigated in this study. The analyses in chapter 3 clearly emphasize the differential resistance of different habitat types on dispersal and the importance of a more detailed view on matrix ‘quality’ in metapopulation studies. Studies that focus on the specific dispersal resistance of different matrix types may provide much more detailed information on the dispersal capacity of species than a mere analysis of isolation by distance. Such information is needed to improve landscape oriented models for species conservation.
In addition to direct effects on realised dispersal (see chapter 3), landscape structure on its own is known to act as an evolutionary selection agent because it determines the costs and benefits of dispersal. Both morphological and behavioural traits of individuals and the degree to which a certain genotype responds to environmental variation have heritable components, and are therefore expected to be able to respond to selection pressures. Chapter 4 analyses the influence of patch size, patch connectivity (isolation of populations) and sand dynamics (stability of habitat) on thorax- and wing length as proxies for dispersal ability of O. caerulescens in coastal grey dunes. This study revealed clear and sex-specific effects of landscape dynamics and patch configuration on dispersal-related morphology. Males of this grasshopper species were smaller and had shorter wings if patches were larger and less connected. In addition, both sexes were larger in habitat patches with high sand dynamics compared to those in patches with lower dynamics. The investments in wing length were only larger in connected populations when sand dynamics were low, indicating that both landscape and patch-related environmental factors are of importance. These results are congruent with theoretical predictions on the evolution of dispersal in metapopulations. They add to the evidence that dispersal-related morphology varies and is selected upon in recently structured populations even at small spatial scales.
Dispersal involves different individual fitness costs like increased predation risk, energy expenditure, costs of developing dispersal-related traits, failure to find new suitable habitat as well as reproductive costs. Therefore, the decision to disperse should not be random but depend on the developmental stage or the physiological condition of an individual just as on actual environmental conditions (context-dependent dispersal, e.g. sex- and wing morph-biased dispersal). Biased dispersal is often investigated by comparing the morphology, physiology and behaviour of females and males or sedentary and dispersive individuals. Studies of biased dispersal in terms of capture-mark-recapture experiments, investigating real dispersal and not routine movements, and genetic proofs of biased dispersal are still rare for certain taxa, especially for orthopterans. However, information on biased dispersal is of great importance as for example, undetected biased dispersal may lead to false conclusions from genetic data. In chapter 5 of this thesis, a combined approach of morphological and genetic analyses was used to investigate biased dispersal of M. bicolor. The presented results not only show that macropterous individuals are predestined for dispersal due to their morphology, the genetic data also indicate that macropters are more dispersive than micropters. Furthermore, even within the group of macropterous individuals, males are supposed to be more dispersive than females. To get an idea of the flight ability of M. bicolor, the morphological data were compared with that of Locusta migratoria and Schistocerca gregaria, which are proved to be very good flyers. Based on the morphological data presented here, one can assume a good flight ability for macropters of M. bicolor, although flying individuals of this species are seldom observed in natural populations.