Dr Erica Chisholm, who is based in AFBI’s Information Systems Branch was awarded a Master’s in Data Analytics passing with distinction. Availing of the ‘Assistance to Study Scheme’, Erica completed the course at Queen’s University, Belfast over two years on a part-time basis. Dr Suzanne Beck who is based in AFBI Fisheries & Aquatic Ecosystems Branch was awarded a PhD with the title “Developing a cost-effective monitoring strategy for coastal cetaceans”. Finally Dr Mark Hawe who now works for Devenish Nutrition and who was supervised by Dr Elizabeth Magowan, AFBI Director of Sustainable & Agri Food Sciences Division achieved a PhD with the title “Identifying, understanding and harnessing the beneficial impact of nutritional interactions to maximise the performance and carcass quality of low birth weight pigs”.
Erica chose to undertake her course to gain a greater insight into how latest analytics technologies and approaches could help AFBI realise the full value of our vast array of disparate data.
The topics covered behavioural analytics, data mining, machine learning and neural networks with a special feature of the course involving “analytathons” using real-life industry data. The last semester involved a practical industry-sponsored project, whereby Erica applied theory and her analytical skills to data produced by Livestock Production Sciences Branch scientist, Gillian Scoley. The aim was to determine the viability of using on-farm precision technologies, together with behavioural data, to predict calf health status. The outcome of this ‘proof of concept’ study was favourable, identifying the important prediction features to be the mean IceRobotic Motion Index, drinking speed, daily weight and birth weight, with evidence to suggest it is potentially easier to predict calf illness after the first 3 weeks of life. As for graduation, there will be no strawberries on the lawn this year with a virtual event taking place on 16 December.
Going forward, Erica is currently contributing to the formulation of AFBI’s Data Science Strategy in addition to leading a DAERA-funded E&I project to explore opportunities to further apply AI/ML to extract insights from data across a range of areas in AFBI in support of the key challenge areas of One Health, Carbon Net Zero, Healthy Water and Precision Livestock Farming.
She would like to extend a special thanks to Dr S Morrison for the support and continued opportunities to further develop the application of AI/ML within AFBI, in addition to recognising AFBI’s support via the Assistance to Study Scheme.
Dr Suzanne Beck achieved her PhD with the title “Developing a cost-effective monitoring strategy for coastal cetaceans”. The overarching aim of this PhD was to further our knowledge on the distribution of cetaceans around Northern Ireland and the surrounding region. These efforts were driven by a need to conserve and protect cetacean species in light of increasing pressures on the marine environment. Four discrete studies were devised to assess historical patterns in cetacean distribution and investigate areas of conservation importance through Passive Acoustic Monitoring (PAM).
The harbour porpoise Phocoena phocoena was found to be the most commonly detected cetacean in Northern Irish waters exhibiting a persistent, but non-uniform distribution. The temporal distribution of this species, monitored in the western Irish Sea and off the north coast of County Donegal, varied across seasons consistently, peaking in winter, while fine scale variations across time of day and tide were found to be site-specific. Further fine-scale spatio-temporal trends (< 5 km) were identified within the Skerries and Causeway Special Area of Conservation (SAC). One deployment location in particular was suggestive of a foraging hotspot, potentially suitable for long-term monitoring through an SAC framework.
It is anticipated that findings on the fine-scale temporal dynamics of harbour porpoise could facilitate local coastal management. While patterns in seasonality, suggestive of an interesting overwintering distribution, pose further questions on the location of these individuals during the summer months. This information places summer derived density estimates into a different perspective and provides another layer of information in which assess the regional population status of this species.
Throughout the project, a critical appraisal of the methods was carried out in the context of suitability for long-term monitoring. Specifically, the presence of fine-scale heterogeneity in distribution (< 5 km) and the association of reduced detection rates with increasing ambient noise were considered. To conclude the thesis, these findings were incorporated into a long-term cetacean monitoring strategy for Northern Ireland.
Finally Dr Samuel Hawes PhD focused on “Identifying, understanding and harnessing the beneficial impact of nutritional interactions to maximise the performance and carcass quality of low birth weight pigs”. Early studies in this PhD project quantified the performance of low birth weight pigs on commercial farms in Northern Ireland and involved a number of high-performance units within the industry. The findings identified low birth weight pigs as a chronic and increasing problem on local farms, with the lactation and nursery periods being critical windows for intervention. Subsequent work involved the development of an improved lactation sow feeding regime. This halved mortality in low birth weight pigs and enabled them to match the suckling behaviour and weaning weight of average birth weight litter mates reared under commercial conditions. However post-weaning growth and feeding behaviour of all piglets was unaffected by how the sow was fed during lactation.
Finally, a targeted post-weaning feeding regime was designed and implemented to enhance the performance of low birthweight pigs during the growing and finishing periods. This fed animals on the basis of bodyweight and improved the growth of low and average birth weight pigs to slaughter age, with no negative impact on body composition. Such a regime was shown to be economically viable in the commercial situation.