Successful implementation of SensOor® technology in research/contract herds in all partner countries: This is the basis for an identical automatically recording technique of health and behaviour traits across country borders. On the basis of SensOor® records from multiple breeds across country borders, we identified genomic regions of interest and potential candidate genes for behaviour and welfare traits.
Most relevant functional traits within specified trait categories (based on e.g. incidences, heritabilities, and economic importance) have been selected and bundled within a trait recording atlas, defining the traits and recording guidelines (organic Eprints).
The ‘trait atlas’ is a basis for harmonized phenotyping across country borders and production systems for further functional traits (conformation, health, behaviour, fertility). We showed that novel physiological traits (e.g., respiration rate, body temperature, surface temperature) have a moderate heritable component.
Detailed characterization of grazing herd environments: installation of data loggers to record temperature and humidity, combined with the recording of meteorological local weather stations.
Additionally, a rising platemeter (E10) was used to measure sword height, pasture classifications (e.g. percentage of herbs), and determination of fodder ingredients (protein content, energy content) was assessed. This is the unique data basis to estimate genetic values along environmental gradients.
A data base system was developed by the agricultural engineering group, including a broad range of information about animal traits (including SensOor® data), pedigrees, genotype data, farm characteristics, climate data, etc., with programming tools for on-farm analyses.
Multiple trait herd cluster models have been developed, in order to study genotype by environment interactions (GxE) for German black and white (DSN), Holstein Friesian (HF), Original Braunvieh (OB), and Brown Swiss (BS) cattle across production systems. For some functional traits (somatic cell score, fat to protein ratio) genetic correlations were lower than 0.80, indicating GxE. Hence, we suggest a borderless clustering genetic evaluation system for dual-purpose breeds, allowing specific breeding values for specific environments.
The GxE-concept was further developed by considering additional environmental and herd descriptors: herd size, calving age, latitude of farm, production level (milk yield), somatic cell count level, genetic DSN percentage, percentage of artificial insemination and natural service sires. Furthermore, we enhanced statistical multiple-trait models by considering genetic relationships and inbreeding, in order to define sustainable breeding strategies.
Genetic values from the multiple trait models are available and can be used for selection decisions within farms (cows and natural service sires), and for optimization of selection of elite animals (additional tool for breeding organizations to select bull sires and bull dams). In this regard, we developed and economically evaluated breeding program designs for dual-purpose cattle in organic production systems.
Breed comparisons for dual-purpose cattle versus dairy cattle breeds from the same country/ same pasture-based production system were conducted for DSN and HF cattle for production (milk-kg, protein-%, fat-%) as well as health indicator traits (SCC, FPR) by applying linear and generalized linear mixed models. Only minor differences were found between DSN and HF for health traits (SCC, FPR), but differences were significantly for physiological traits.
Genotyping across country borders allowed sophisticated studies on genetic diversity in dual-purpose cattle breeds. Based on principal component analyses, different breeds were clearly separated, and genetic similarities were identified and explained in the context of the breed history.
SensOor®-validations: The agricultural engineering group installed alternative cattle behaviour devices in participating research herds: Rumiwatch halters, pH-boluses and pedometers for SensOor®-validation
