Year: 2023
Type: dissertation
Abstract: One of the most outstanding values of livestock is the nutritional merit of its end products. Foods of animal origin are naturally energy-dense and nutrient-rich, essential for the human body functioning. As a result, there are significant changes to how animal-source foods are produced, processed, and distributed globally. The worldwide trade of livestock products has increased the risk of expanding a local animal disease into a global public health issue, such as the recent avian influenza outbreaks. Besides, all evidence suggests that SARS-CoV-2 has a zoonotic source, and the transmission of the COVID-19 virus to humans more likely happened through animal species handled by humans. The continuing global population growth and dietary changes are the primary forces for an increased demand for meat, egg, milk, and fish products. In addition, poultry meat has a fast-growing consumption rate because it is relatively more affordable and acceptable by religions than other meat sources. Intensification of livestock also follows this greater demand. Self-sufficient small-scale family farms, producing food on their land with little to no external inputs, are replaced by verticalised integration companies that control animal breeding, fattening, feed production and meat processing. For instance, broiler house complexes allocate hundreds of thousands of birds into high-density units. Similarly, hatcheries have been increasing both the capacity of incubators and the density of hatching eggs. The consequence is that traditional stockmanship is no longer suitable. More attention is given to administrative duties and sounds operational logistics for cost-effective production. Therefore, the regular assessment of animals' health and welfare is compromised, creating a negative public awareness of intensive animal farming. In general, this thesis discusses the application of precision livestock farming technology in assisting the monitoring of animal health and welfare, supporting management decisions in contributing to more sustainable livestock units. In livestock, the development and growth of animals follow the same principles as any other living organisms (e.g. humans, plants, and cells). An individual is a unique biological body made of inherited genes with specific traits that are affected by disturbance variables (e.g. micro-environmental conditions) in continuous changes. So far, many relational cause-effect studies have neglected the premise that every living organism (e.g. chicken embryo) will respond individually differently from each other and in a different manner each time a specific stimulus is applied (e.g. embryo development). Therefore, measuring many different incubation physical variables (e.g. air temperature, air humidity, and air velocity) cannot represent the intrinsic variabilities between individual eggs and the time-variant embryo response in a single egg. For example, during the incubation of eggs, the chicken embryo evolves through different stages from a poikilothermic to a homeothermic system. Hence, embryonic thermal responses to changes in the incubator temperature differ daily. Therefore, the individual state of an animal (e.g. thermoregulatory capacity) in intensive livestock systems remains the main obstacle to developing an automated system supporting farm management decisions. For effective and accurate monitoring and management system, objective, measurable bioresponses (i.e. bio-signal of an animal feature) must provide valid and reliable feedback over time. Despite the increased use of information and communications technology in livestock, the majority of livestock systems continue to rely on the measurement of physical environmental variables (e.g., air temperature in broiler houses) without any animal evidence-based measures (i.e. bioresponses) to monitor and control the process (e.g. cooling/heating). In addition, one of the recurrent claims is that existing sensors, information, and computer technology are still expensive for most farmers. The focus of this Ph.D. is on the development of an individual model-based monitoring system for incubated eggs. We ask to what extent the interaction between the growing embryo and the incubation micro-environment can indicate a relevant change during incubation. Also, what set of variables measured directly on the egg non-invasively can be used to represent meaningful information from the chicken embryo's development? We have applied the principles of Precision Livestock Farming (PLF) by first expanding our knowledge of the system of interest (i.e. incubation of eggs) by using low-cost sensor technology (e.g. temperature sensors and cameras) to obtain reliable process-related information. As shown in Chapter 2, registering the eggshell temperature every minute gives precise detection (99% correlation coefficient) of the hatching time of 40 individually monitored eggs. Chapter 3 describes the underlying evaporative cooling effect on eggshells, evidenced by the dynamics observed in the eggshell temperature data during hatching. We then defined proper feature variables measured continuously to monitor the interaction between the egg and the incubator environment to detect changes from the normal status of embryonic development. For example, Chapter 4 uses eggshell temperature as a feature variable to detect embryonic mortality. An offline decision tree induction model classified with 93% accuracy the status of 105 sampled eggs into infertile/early death, middle death, late death, and hatch state. Then, we applied a data-based mechanistic modelling approach with low complexity functions (e.g. system identification techniques) to identify model parameters and give a meaningful biological explanation related to embryonic development during incubation (Chapter 5). This modelling was then used to set individual thresholds for detecting relevant process milestones and predicting future state. In Chapter 6, we classified the status of individuals into the following five milestones of a chicken embryo during development: 1) the transition from endothermic to exothermic status; 2) the plateau phase of metabolic heat production during growing; 3) the starting of lung respiration by the embryo with the internal piping action; 4) The completion of the embryonic stage with the external piping from which chick breaks through the eggshell; 5) the climax of the process through the hatching of the chick emerging from the shell. These results indicate that simple measure variables like eggshell temperature can disclose interesting process information during incubation. It, therefore, offers some basis for developing a more objective monitoring system to support hatchery managers in improving their hatchability and chick quality standards. In this Ph.D. study, using appropriate sensor technology, which does not have to be complex and expensive, in conjunction with real-time modelling, could lead to a more cost-effective automated monitoring and supporting system for management decisions in incubated eggs. In addition, advanced precision livestock monitoring of embryo development throughout incubation based on eggshell temperature measurements can provide an early warning to inform evidence-based (e.g. embryo status, mortality and hatch) strategies for adjusting the incubator environmental conditions and predicting the hatching time of individuals.
Source: Lirias
Author: Bites Romanini, Carlos Eduardo
Language: English
Subfield: Animal Science and Zoology
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7005409308