Full text
| |
| Author(s): |
Henrique Rocha de Medeiros
Total Authors: 1
|
| Document type: | Doctoral Thesis |
| Press: | Piracicaba. , gráficos, ilustrações, tabelas. |
| Institution: | Universidade de São Paulo (USP). Escola Superior de Agricultura Luiz de Queiroz (ESALA/BC) |
| Defense date: | 2003-10-17 |
| Examining board members: |
Carlos Guilherme Silveira Pedreira;
Pedro Franklin Barbosa;
Valéria Pacheco Batista Euclides;
Nilson Augusto Villa Nova;
Luiz Gustavo Nussio
|
| Advisor: | Carlos Guilherme Silveira Pedreira |
| Field of knowledge: | Agronomical Sciences - Animal Husbandry |
| Indexed in: | Banco de Dados Bibliográficos da USP-DEDALUS; Biblioteca Digital de Teses e Dissertações - USP |
| Location: | Universidade de São Paulo. Biblioteca Central da Escola Superior de Agricultura Luiz de Queiroz; t633.26; M488a |
| Abstract | |
In agricultural systems, environment, soil, animals and plants are indissociable components of the production process. Furthermore, the production time-frame is relatively long. These characteristics enhance risk and uncertainty associated with decision alternatives made during the process. One tool which can aid in decision-making in these processes is the use of mathematical modeling. Models are simplified representations of reality and allow for the estimation of responses of the system (e.g., production) as the process is altered or for the description and understanding of processes within the system. The objectives of the present study were to (i) test the Stockpol ® model under the conditions and as a predicting tool for ruminant production systems in Brazil, (ii) validate a herbage accumulation model based on climatic variables, and (iii) evaluate models that estimate forage intake on pasture-based animal production systems. For these purposes, the available literature was surveyed in order to identify the datasets that would fulfill the necessary requirements in terms of amount of information needed for the validation and evaluation exercises. The information were then catalogued in three separate datasets: pasture, animal, and economic indicators. This work allowed for the identification of research areas where information is short (such as the seasonal distribution of annual forage production as well as forage intake by grazing animals), besides the need for standardization of research methods and procedures so that research data generated in different environments and by different research groups in Brazil can be compared on the same basis. Despite the many limitations, if the existing parameters are reviewed or adjusted and/or new input variables are incorporated, the Stockpol ® model may become both a research and a decision-making tool for pasture-based animal production systems in Brazil. The Photothermal Units (PU) model proved suitable as it gave adequate predictions of forage yield of Cynodon spp. grasses considering only daylength and temperature as input variables. In this model, predictions seem to be most accurate when the base-temperature is between 13 and 15 ºC. In the future, more input variables, such as the hidric balance (to fine-tune the water budget) may be incorporated into the model to enhance its prediction ability. The model evaluated to predict forage intake by grazing ruminants will likely need reparametrization and/or use green leaf allowance (green leaf dry mass divided by total liveweight at any one point in time) to be useful in predicting intake on tropical pastures. The model used to predict performance gave poor estimates of weight gain, probably due to the overestimation of the energy cost of grazing and/or the underestimation of the nutritive value of the diet selected. It is expected that, once these issues have been resolved, the performance model can operate in conjunction with both the PU and the intake models so that the major components of the system can be simulated under Brazilian conditions (AU) | |