Gunnar Klemetsdal

Feed efficiency (FE) is a major determinant of economic and environmental sustainability in Atlantic salmon aquaculture, yet the biological drivers of FE during the saltwater (SW) phase remain poorly understood. This study aimed to develop and validate methods for predicting visceral fat percentage (VFP) and estimate its genetic relationships with growth, fillet lipid deposition, and feed-efficiency traits. A population of 836 fish from 30 families was phenotyped for growth, feed intake, and body composition traits. VFP was predicted with high accuracy using near-infrared spectroscopy on homogenized viscera (cross-validated R2 = 0.91), while fillet fat percentage (FFP) was predicted using Raman spectroscopy (cross-validated R2 = 0.58). Genetic parameters were estimated using animal models based on genomic relationships. VFP was moderately heritable (h2 = 0.28 ± 0.06) and showed a strong positive genetic correlation with feed conversion ratio (FCR; rg = 0.90), indicating that lower visceral lipid was genetically associated with improved FE. When relative visceral weight was partitioned into lipid and non-lipid components, decreased visceral lipid was specifically associated with improved FCR, whereas increased relative non-lipid organ visceral mass was associated with higher specific growth rate. FFP was only moderately genetically correlated with FCR traits (rg = 0.34-0.53). A negative partial genetic correlation between VFP and FFP conditional on body weight (rg = -0.36 ± 0.15) suggested a small allocation tradeoff between lipid compartments. Overall, visceral lipid content is a heritable and informative indicator trait for improving FE in SW Atlantic salmon.

This study aims to evaluate the use of a reaction-norm model, with NDF concentration in the ration as an environmental gradient, and its plasticity as a candidate phenotype for identifying animals that efficiently utilize plant fiber. A dataset from a feeding trial with 40 cows fed different types of silage and concentrate with differing NDF concentrations was used. Single-visit ln-transformed methane production was analyzed using a model accounting for systematic environmental effects; among these a fixed effect of NDF concentration as a fixed effect, along with random effects for individual intercept and slope, both of which were significant. Plasticity, defined as the individual slope of the response in ln-transformed methane production to NDF concentration in the ration, was further examined in relation to underlying data structure and NDF digestibility. Underlying data structure was analyzed with an explanatory factor analysis of mean and variance for variables not included in the reaction-norm model, interpreted as either feed composition or feed intake and activity. The individual slope showed a significant positive relationship to individual NDF digestibility, suggesting that it could serve as an indicator trait to breed cows that efficiently utilize roughage. However, this finding needs to be validated in a larger dataset suitable for genetic analysis.

Kelps are not only ecologically important, being primary producers and habitat forming species, they also hold substantial economic potential. Expansion of the kelp cultivation industry raises the interest for genetic improvement of kelp for cultivation, as well as concerns about genetic introgression from cultivated to wild populations. Thus, increased understanding of population genetics in natural kelp populations is crucial. Genotyping-by-sequencing (GBS) is a powerful tool for studying population genetics. Here, using Saccharina latissima (sugar kelp) as our study species, we characterize the population genetics at a fine geographic scale, while also investigating the influence of marker type (biallelic SNPs versus multi-allelic short read-backed haplotypes) and minor allele count (MAC) thresholds on estimated population genetic metrics. We examined 150 sporophytes from 10 locations within a small area in Mid-Norway. Employing GBS, we detected 20,710 bi-allelic SNPs and 42,264 haplotype alleles at 20,297 high quality GBS loci. We used both marker types as well as two MAC filtering thresholds (3 and 15) in the analyses. Overall, higher genetic diversity, more outbreeding and stronger substructure was estimated using haplotypes compared to SNPs, and with MAC 15 compared to MAC 3. The population displayed high genetic diversity (HE ranging from 0.18–0.37) and significant outbreeding (FIS ≤ − 0.076). Construction of a genomic relationship matrix, however, revealed a few close relatives within sampling locations. The connectivity between sampling locations was high (FST ≤ 0.09), but subtle, yet significant, genetic substructure was detected, even between sampling locations separated by less than 2 km. Isolation-by-distance was significant and explained 15% of the genetic variation, while incorporation of predicted currents in an “isolation-by-oceanography” model explained a larger proportion (~ 27%). The studied population is diverse, significantly outbred and exhibits high connectivity, partly due to local currents. The use of genome-wide markers combined with permutation testing provides high statistical power to detect subtle population substructure and inbreeding or outbreeding. Short haplotypes extracted from GBS data and removal of rare alleles enhances the resolution. Careful consideration of marker type and filtering thresholds is crucial when comparing independent studies, as they profoundly influence numerical estimates of population genetic metrics.

Compared with cow fertility, genetic analyses of bull fertility are limited and based on relatively few animals. The aim of the present study was to estimate genetic parameters for semen characteristics of Norwegian Red bulls at the artificial insemination (AI) center (Geno AI station, Stange, Norway) and to estimate genetic correlations between some of these traits and andrology traits measured at the performance test station. The data from the AI center consisted of records from 137,919 semen collections from 3,145 bulls with information on semen weight, sperm concentration, motility before and after cryopreservation, motility change during cryopreservation, and number of accepted straws made. Data from the performance test station included 12,522 observations from 3,219 bulls on semen volume, concentration, and motility (%) when fresh and after storing for 24 and 48 h. Genetic parameters were estimated using linear animal repeatability models that included fixed effects of year-month of observation, age of bull, interaction between semen collection number, and interval between collections for all traits and type of diluter for postcryopreservation traits. The random effects included test-day, permanent environmental, and additive genetic effects of the bull. Based on records from the AI center, we found that semen weight, sperm concentration, and number of straws were moderately heritable (0.18-0.20), whereas motility had a lower heritability (0.02-0.08). Heritability of motility (%) was higher after cryopreservation than before. Genetic correlations among the semen characteristics ranged from unfavorable (-0.35) to favorable (0.93), with standard errors ranging from 0.02 to 0.22. Among the most precise genetic correlation estimates, number of straws made from a batch correlated favorably with semen weight (0.62 ± 0.06) and sperm concentration (0.44 ± 0.08), whereas sperm concentration was negatively correlated with weight (-0.33 ± 0.09). The genetic correlation between motility (%) before and after cryopreservation was 0.64 ± 0.14, and motility change during cryopreservation had a strong favorable genetic correlation with motility after cryopreservation (-0.93 ± 0.02). The estimated genetic correlation (standard error) between the traits volume, concentration, and motility when fresh measured at the performance test station and their respective corresponding traits at the AI center were 0.83 (0.05), 0.78 (0.09), and 0.49 (0.31). The final product at the AI center (number of accepted straws) correlated genetically favorably with all semen characteristic traits recorded at the performance test station (ranging from 0.51 to 0.67). Our results show that the andrology testing done at the performance test station is a resource to identify the genetically best bulls for AI production.

The aim of this study was to analyse four cohabitation challenge‐test experiments with Mekong striped catfish (Pangasianodon hypophthalmus) against the bacterium Edwardsiella ictaluri. The data were genetically analysed per experiment by three models: 1) a cross‐sectional linear model; 2) a cross‐sectional threshold model; and 3) a linear survival model, at both 50% mortality (for models 1 and 2) and at the end of the test (for all three models). In two of the experiments (3 and 4) that were carried out in two replicated tanks, the predicted family effects (sum of sire, dam and common environmental effects) in each tank were correlated with the family survival in the other replicated tank (cross‐validation). The heritability estimates of resistance to E. ictaluri infection were ≤ 0.012 with the survival model, and up to 0.135 – 0.220 (50% survival) and 0.085 and 0.174 (endpoint survival) for the cross‐sectional linear and threshold models, respectively. The challenge test should aim for an endpoint survival that ceases naturally at 50%. Then, genetic analysis should be carried out for survival at the endpoint (reflecting susceptibility) with a simple cross‐sectional linear model.

Bacillary necrosis is a problematic disease in farming of Mekong striped catfish (Pangasianodon hypophthalmus). The pathogenic bacterium is Edwardsiella ictaluri, causing numerous white spots in swelled liver, kidney and spleen. An alternative to antibiotic treatment and vaccine is to select for improved genetic resistance to the disease that requires to establish a proper challenge test. Here, four challenge tests of Mekong striped catfish against E. ictaluri are reported proposing 3 days of acclimatization of test fish prior to the challenge, with restricted water level in the test, keeping a temperature of 26°C. In the challenge, cohabitant shedders should be released directly into the test tank and make up around ⅓ of the fish, and bacteria should be added directly to water. The last two experiments, with the highest mortality, suggest that any factor involving the dead cohabitants should be removed and that additional experimentation should focus on bacteria (density) and timing for addition of bacteria to water. Genetic analyses revealed that resistance to bacillary necrosis tested in replicated tanks in the same experiment can be considered the same genetic trait.

Background: One objective of this study was to identify putative quantitative trait loci (QTL) that affect indicator phenotypes for growth, nitrogen, and carbon metabolism in muscle, liver, and adipose tissue, and for feed efficiency. Another objective was to perform an RNAseq analysis (184 fish from all families), to identify genes that are associated with carbon and nitrogen metabolism in the liver. The material consisted of a family experiment that was performed in freshwater and included 2281 individuals from 23 full-sib families. During the 12-day feed conversion test, families were randomly allocated to family tanks (50 fish per tank and 2 tanks per family) and fed a fishmeal-based diet labeled with the stable isotopes ¹⁵ N and ¹³ C at inclusion levels of 2 and 1%, respectively. Results: Using a linear mixed-model algorithm, a QTL for pre-smolt growth was identified on chromosome 9 and a QTL for carbon metabolism in the liver was identified on chromosome 12 that was closely related to feed conversion ratio on a tank level. For the indicators of feed efficiency traits that were derived from the stable isotope ratios ( ¹⁵ N and ¹³ C) of muscle tissue and growth, no convincing QTL was detected, which suggests that these traits are polygenic. The transcriptomic analysis showed that high carbon and nitrogen metabolism was associated with individuals that convert protein from the feed more efficiently, primarily due to higher expression of the proteasome, lipid, and carbon metabolic pathways in liver. In addition, we identified seven transcription factors that were associated with carbon and nitrogen metabolism and located in the identified QTL regions. Conclusions: Analyses revealed one QTL associated with pre-smolt growth and one QTL for carbon metabolism in the liver. Both of these traits are associated with feed efficiency. However, more accurate mapping of the putative QTL will require a more diverse family material. In this experiment, fish that have a high carbon and nitrogen metabolism in the liver converted protein from the feed more efficiently, potentially because of a higher expression of the proteasome, lipid, and carbon metabolic pathways in liver. Within the QTL regions, we detected seven transcription factors that were associated with carbon and nitrogen metabolism.

The primary objective of this study was to design a growth profile from 3 mo through puberty to insemination that allows heifers to enter the milking herd at 22 mo of age without impairing milk production over 3 lactations compared with the current rearing practice leading to an age at first calving of 26 mo. Eighty heifers born into the Norwegian University of Life Sciences herd, 40 each from yr 2010 and 2011, were randomly assigned according to birth order either to a high or low intake energy treatment. Each energy group was further subdivided into 2 protein groups, 1 fed according to requirements and 1 fed 10% excess protein, to ensure that metabolizable protein supply would meet the requirements for rapidly growing bone and muscle of today's genetically improved Norwegian Red heifer. Utilizing growth rate and feed composition the energy and protein supply was regulated with roughage quality in a diet containing 1 kg/d of concentrate of 2 qualities. Average daily gain from 3 mo to confirmed pregnancy ranged from 900 to 1,000 g/d among high-energy animals, with high protein-fed animals growing the fastest. Growth rates for low energy animals were <700 g/d. From confirmed pregnancy to first calving, all animals were fed only grass silage to sustain an average daily gain 600 g/d (low energy), excluding fetal and gravid uterus weight, and they reached a postcalving weight of 530 (high energy) to 570 kg (low energy) with body condition score ranging from 3.42 to 3.93 at calving. We have shown that heifers fed a high-energy treatment with the required amount of protein from 3 mo of age to successful insemination combined with an average daily gain of similar to 500 g/d throughout pregnancy will calve at 22 mo without becoming over-conditioned at calving and without impairing performance over 3 lactations. We recommend reducing rearing time by 4 mo, planning for an age at first calving of 22 mo of age. This rearing practice would also improve energy efficiency during the heifer rearing period.

The aim was to carry out a joint genetic analysis of survival and harvest body weight, recorded in a growth performance test in Mekong striped catfish (Pangasianodon hypophthalmus), and susceptibility to bacillary necrosis (caused by Edwardsiella ictaluri), recorded in challenge tests. Data were from two challenge tested year‐classes (~6,000 fish in both) that both had growth test data available for survival and body weight (~13,000 fish each year). Data were analysed with a linear tri‐variate sire‐dam model without the common environmental effect because otherwise genetic correlations were estimated with large standard errors. Susceptibility to bacillary necrosis was found weakly genetically correlated to both growth and survival in the growth test, while growth was found with moderate favourable genetic correlation to growth survival. To defend continued challenge testing of striped catfish in Vietnam, a strong genetic relationship needs to be established between bacillary necrosis and survival under a natural disease outbreak. This requires another field test (in addition to the growth test) with siblings, without antibiotic treatment and the cause of death continuously monitored. Meanwhile, the routine challenge testing with the aim to indirectly improve field survival through selection should continue.