Full text
| |
| Author(s): |
Robson Mateus Freitas Silveira
Total Authors: 1
|
| Document type: | Doctoral Thesis |
| Press: | Piracicaba. |
| Institution: | Universidade de São Paulo (USP). Escola Superior de Agricultura Luiz de Queiroz (ESALA/BC) |
| Defense date: | 2025-07-18 |
| Examining board members: |
Iran Jose Oliveira da Silva;
Daniella Jorge de Moura;
Irenilza de Alencar Naas
|
| Advisor: | Iran Jose Oliveira da Silva |
| Abstract | |
The aim of this study was to apply intelligent predictive models to the adaptive and productive responses of farm animals, focusing on the sustainability of the production system and animal welfare in face of climate change. This thesis is divided into five chapters. 1) An exploratory review was carried out to provide a longitudinal perspective on the current state of academic research in the emerging field of machine learning applied to sustainable animal production in times of climate change. 1,082 papers published in the last 70 years were selected. It was found that the concept of sustainable animal production emerged from the principles of animal welfare and climate change established in the UN 2030 Agenda and that the main keywords for future studies on this topic include omics sciences, greenhouse gases, energy efficiency and animal welfare; 2) Conceptualizing sustainable animals and developing a methodology to identify these animals. Thermoregulatory, morphological, hematological, serum biochemical, hormonal and productive responses, carcass traits and feed efficiency measures of 62 bulls were collected. A typology of bulls with sustainable resilience and adaptive capacity to the environment, productive sustainability with heavy carcasses and high yield and feed sustainability was identified - the synergism of these traits is the concept of a sustainable animal. Rectal temperature, mean corpuscular volume, leukocytes, precocity, birth weight, carcass yield and residual feed intake are the main phenotypic biomarkers to identify these animals; 3) Adaptive profiles and their phenotypic biomarkers were identified in the animal model. Data from 76 dairy goats were used. The main results were: (i) hematological variables represent important biomarkers to study animal adaptation; (ii) thermoregulatory, hematological and immunological mechanisms vary according to coat thickness; and (iii) age, body condition score and lactation order are factors that influence adaptive responses; 4) Modeling complex patterns of relationships or multiphenotypic differences between the thermal environment and the adaptive and productive responses of 30 lactating cows. Moderate relationships (0.300 ≤ rc 2 ≤ 0.628) were found with low influence (0.141 ≤ rc2 ≤ 0.384) between indicators (thermal environment, thermoregulatory responses, biochemistry, hormonal profile, hematological responses and milk composition). The prediction of adaptive responses according to climatic variables demonstrated that all adaptive responses are important for the study of animal adaptation and 5) Assessing the impact of the thermal environment on the adaptive responses of livestock in both hemispheres and simulating future climate scenarios to project the conditions of the impact of climate change on livestock production in the coming decades. Twelve databases collected in different hemispheres of the globe from different livestock species were used. It was found that (i) the thermal environment of the Northern Hemisphere has and will have a greater impact on the thermoregulatory responses of animals than the thermal environment of the Southern Hemisphere; (ii) laying hens and quails are more susceptible to heat stress when compared to ruminants; (iii) erythrocyte morphology; thyroid hormones; liver and kidney function; and hair density are phenotypic biomarkers of animal adaptation; (iv) beef cattle and goats are animals with phenotypic plasticity; and (v) ruminants reared in the Southern Hemisphere may increase rectal temperature as an evolutionary way to cope with high temperatures in the tropics. It was concluded that machine intelligence can be applied in order to understand the relationships between adaptive and productive responses of farm animals, contributing to sustainability and animal welfare in the face of climate change, as identified sustainable animals and key biomarkers for adaptation and animal production. Global simulations indicated regional differences in the impact of climate change on farm animals, with the greatest impact on animals raised in the Northern Hemisphere. (AU) | |
| FAPESP's process: | 22/14250-8 - Machine intelligence applied to animal bioclimatology: Predictive models between the thermal environment and adaptive and productive responses for farm animals |
| Grantee: | Robson Mateus Freitas Silveira |
| Support Opportunities: | Scholarships in Brazil - Doctorate |