Genetic diversity, gene flow, and landscape resistance in a pond-breeding amphibian in agricultural and natural forested landscapes in Norway.

Genetic diversity is a key part of biodiversity, threatened by human activities that lead to loss of gene flow and reduction of effective population sizes. Gene flow is a result of both landscape connectivity and demographic processes determining the number of dispersing individuals in space and time. Thus, the effect of human impact on processes determining the level of genetic diversity must be interpreted in the context of basic ecological conditions affecting survival and recruitment. When the intensity of human impact and habitat suitability correlate, the effect on genetic diversity and gene flow may be challenging to predict. We compared genetic diversity, gene flow and landscape resistance in two contrasting landscapes in Norway for the pond-breeding amphibian Triturus cristatus: a highly human-impacted, agricultural landscape with ecologically productive habitats, and a forested landscape with less productive habitats and lower levels of human impact. Our results show that genetic diversity was higher and gene flow lower within the forested landscape. Microclimatic moisture conditions and vegetation cover were important determinants of landscape resistance to gene flow within both landscapes. There were indications that landscape resistance was increased by minor roads in the forested landscape, which was not the case for the agricultural landscape, suggesting a higher vulnerability to human interference within the landscape matrix for the populations in less productive habitats. Our findings suggest that the effect of human impact on genetic diversity may not be straightforward but modulated by the ecological conditions underlying local demographic processes. Populations within both landscapes seem to be vulnerable to loss of genetic diversity, but due to different mechanisms. This has implications for the choice of relevant management actions, that is, increasing population stability may be more relevant within an agricultural landscape still permeable for dispersal, while conserving dispersal corridors may be more appropriate in the forested landscape, to avoid isolation and increased genetic drift.

Effects of Lake Productivity on Density and Size Structure of Pelagic Fish Estimated by Means of Echosounding in 17 Lakes in Southeast Norway.

Density estimation of pelagic fish was performed by means of single beam echosounding in 17 lakes within a period of 34 years, from 1985 to 2018. Surveys were performed repeatedly (two to fourteen times) in five lakes. The density estimates ranged from 34 to 4720 fish/ha and were significantly correlated with total phosphorus concentration. The high density in relatively phosphorus rich lakes (TP > 10 µg/L) was comprised of small fish (<20 cm) and was partly due to the higher number of pelagic fish species. The number of pelagic species varied from one, Arctic charr, in the most elevated and oligotrophic lakes, and whitefish dominated in less elevated oligotrophic lakes. In lowland lakes characterized as mesotrophic or tending to mesotrophy, smelt, vendace, and two to three cyprinids comprised the pelagic fish stock. These fish species predate zooplankton effectively, and species composition and body size of planktonic cladocerans was affected by fish density. Large species of Daphnia were lacking in lakes with high fish density, and body size of present species, D. galeata, D. cristata, and Bosmina spp. were negatively correlated with pelagic fish density.

Assessing Evolutionary Significant Units (ESU) of the Endangered Freshwater Pearl Mussel (Margaritifera margaritifera) in Southeast Norway on the Basis of Genetic Analysis.

A total of 312 specimens of freshwater pearl mussel (Margaritifera margaritifera) were sampled from 11 populations, located in four different river systems in Southeast Norway, and analyzed for 11 simple sequence repeat (SSR) (microsatellite) markers. All study populations have landlocked brown trout (Salmo trutta) as the only possible host. Several populations had experienced recruitment failure, probably due to low pH (about 6.0) and calcium concentration. STRUCTURE clustering analysis revealed two genetic clusters, of which one cluster occurred mainly in the western river systems, and totally dominated in one population (Fallselva (A-FAL)) that had higher genetic diversity than the others. Cluster 2 completely dominated in the populations of the eastern river systems, and all of them had low genetic diversity. Bottleneck events were indicated in all populations and the inbreeding coefficient FIS was significant in all populations, except for the southernmost population (Sørkedalselva (B-SØR)), which was the only population with genotypes in Hardy-Weinberg equilibrium. FIS were especially high in the populations of the eastern river systems, and maximum shell length was negatively correlated to FIS. If artificially breeding and stocking should become necessary for future preservation, it should be based on single populations; alternatively, the eastern populations should be based on cross-breeding of populations within the cluster to increase their genetic diversity.

Genetic differences between wild and hatchery-bred brown trout (Salmo trutta L.) in single nucleotide polymorphisms linked to selective traits.

To study effects from natural selection acting on brown trout in a natural stream habitat compared with a hatchery environment, 3,781 single nucleotide polymorphism (SNP) markers were analyzed in three closely related groups of brown trout (Salmo trutta L.). Autumn (W/0+, n = 48) and consecutive spring (W/1+, n = 47) samples of brown trout individuals belonging to the same cohort and stream were retrieved using electrofishing. A third group (H/1+, n = 48) comprised hatchery-reared individuals, bred from a mixture of wild parents of the strain of the two former groups and from a neighboring stream. Pairwise analysis of FST outliers and analysis under a hierarchical model by means of ARLEQUIN software detected 421 (10.8%) candidates of selection, before multitest correction. BAYESCAN software detected 10 candidate loci, all of which were included among the ARLEQUIN candidate loci. Body length was significantly different across genotypes at 10 candidate loci in the W/0+, at 34 candidate loci in the W/1+ and at 21 candidate loci in the H/1+ group. The W/1+ sample was tested for genotype-specific body length at all loci, and significant differences were found in 10.6% of all loci, and of these, 14.2% had higher frequency of the largest genotype in the W/1+ sample than in W/0+. The corresponding proportion among the candidate loci of W/1+ was 22.7% with genotype-specific body length, and 88.2% of these had increased frequency of the largest genotype from W/0+ to W/1+, indicating a linkage between these loci and traits affecting growth and survival under this stream's environmental conditions. Bayesian structuring of all loci, and of the noncandidate loci suggested two (K = 2), alternatively four clusters (K = 4). This differed from the candidate SNPs, which suggested only two clusters. In both cases, the hatchery fish dominated one cluster, and body length of W/1+ fish was positively correlated with membership of one cluster both from the K = 2 and the K = 4 structure. Our analysis demonstrates profound genetic differentiation that can be linked to differential selection on a fitness-related trait (individual growth) in brown trout living under natural vs. hatchery conditions. Candidate SNP loci linked to genes affecting individual growth were identified and provide important inputs into future mapping of the genetic basis of brown trout body size selection.