Author:

Abstract:

Ecological speciation occurs when adaptation to different environments causes the evolution of reproductive isolation. This process is thought to be very important in the evolution of biological diversity. Therefore, insight in the underlying mechanisms is crucial to understand the generation of biodiversity on earth. In ecological speciation species differentiate in phenotypic traits in response to divergent selection associated with the exploitation of different ecological environments. Ecological speciation may give insight into one of the most intriguing open questions in evolutionary ecology: if and how microevolution connects to macroevolution.I performed a series of experiments related to ecological speciation driven by predation in the damselfly genus Enallagma where most species are adapted to the ancestral fish lakes and where some independent speciation events in association with habitat shifts toward fishless lakes with dragonflies as top predators occurred. I studied the selective environments and the genetic variance-covariance matrices of key species related to two independent habitat shifts.In a first part of the thesis, I documented divergent survival selection on important ecological traits (related to escape swimming and general activity) that evolved in relation to different selective regimes imposed by different predators linked to the habitat shift. Dragonfly larvae imposed selection on increased swimming propensity, swimming speed, higher activity levels of arginine kinase and more circular lamellae. Fish imposed selection for a lowered foraging activity and for a decreased swimming propensity. These results confirm the parallel evolution of a higher escape swimming speed associated with the habitat shift from fish lakes to dragonfly lakes. Combining these results with the estimated broad-sense heritability indicated that the evolutionary increase in swimming speed associated with the habitat shift occurred very fast. This would suggest that for predator escape performance, differences at species level that create new species by ecological speciation may have occurred as rapidly as contemporary microevolutionary responses to similar shifts within species.In a second part of the thesis, I reconstructed the genetic variance-covariance matrices for swimming speed and six associated variables of the assumed ancestor and one derived species for each of two independent habitat shifts. Despite the expectation that species experiencing similar multivariate selection regimes will converge in genetic architecture, the matrices did not evolve in parallel during the two independent habitat shifts. Another key finding was that while genetic constraints apparently played a role in shaping the direction of phenotypic evolution during the speciation events in each habitat shift, the exact nature of these genetic constraints differed between both shifts. This indicates that the observed parallelism in the evolution of phenotypes in the independent shifts toward dragonfly lakes cannot be explained by the presence of identical genetic constraints embedded in the genetic architecture of both ancestors and that adaptation in response to similar selection regimes in the derived dragonfly lake habitat was probably the major force driving the parallel phenotypic evolution.