presented by ne parameter set (Tardieu 2003). Analyzing genotypic variation by
means of models is a first step, and has been done with the carbon submodel on a
contrasting population of peach genotypes (Quilot et al. 2004). In addition,
incorporating the virtual peach fruit model in a crop model will be of value in
simulating the combined effects of changes in climate, pest events changing leaf
area or photosynthesis, genotype, and technical operations on profiles of quality
traits, and will thereby help to improve breeding and crop management processes
(Lescourret and Génard 2005).
2.5.3.1 Modeling Crop Quality: A Case of Study in Tomato
Fruit quality at harvest is a complex trait, including size, overall flavor (taste and
texture), and visual attractiveness (color, shape), which depend on both genotype
and environment. The improvement of fresh product quality is slowed down by its
complexity. It is expected that the development of process-based models and their
integrations in ecophysiological models should facilitate quality management,
provided that integration properly accounts for interactions among biological
processes (Bertin et al. 2006).
2.5.4 Applications of Neural Networks in Agriculture
and Biosystems
Agricultural systems, such as environment–plant system, are quite complex and
uncertain and they can be considered as ill-defined systems. They are characterized
by nonlinearities, time-varying properties, climatic interactions, and many other
factors. It is, therefore, difficult to quantify complex relationships between the input
and the output of a system based on analytical methods. Recently, intelligent
system control based on artificial intelligence has been one of the most prosperous
technologies in the complex system science (Hashimoto 1997). ANN is also a
promising tool for predicting crop yield and offers insight into the casual relationships through the use of sensitivity analyses, but the complex parameterization
and optimum model structure require special attention (Park et al. 2005). The
machine learning tools used to model the effects of environmental conditions on
apparent photosynthesis provided a very powerful modeling alternative to ordinary
curve fitting methods. Their major advantages are the flexibility to select between
accuracy and generality and their robustness against outliers and mixtures of differentia responses (Dalaka et al. 2000). Resources capture and plant responses to
energetic stimuli (solar radiation, temperature, CO 2 concentration, and humidity)
present complex ecophysiological mechanisms, which conduct to dynamic modifications of the canopy development and architecture (Vazquez-Cruz et al. 2010).
In agronomic research, ANNs have been presented as alternative methodology
to modeling and simulating crop biophysical properties. ANN models are specially
2 Mathematical Modeling of Biosystems
71
means of models is a first step, and has been done with the carbon submodel on a
contrasting population of peach genotypes (Quilot et al. 2004). In addition,
incorporating the virtual peach fruit model in a crop model will be of value in
simulating the combined effects of changes in climate, pest events changing leaf
area or photosynthesis, genotype, and technical operations on profiles of quality
traits, and will thereby help to improve breeding and crop management processes
(Lescourret and Génard 2005).
2.5.3.1 Modeling Crop Quality: A Case of Study in Tomato
Fruit quality at harvest is a complex trait, including size, overall flavor (taste and
texture), and visual attractiveness (color, shape), which depend on both genotype
and environment. The improvement of fresh product quality is slowed down by its
complexity. It is expected that the development of process-based models and their
integrations in ecophysiological models should facilitate quality management,
provided that integration properly accounts for interactions among biological
processes (Bertin et al. 2006).
2.5.4 Applications of Neural Networks in Agriculture
and Biosystems
Agricultural systems, such as environment–plant system, are quite complex and
uncertain and they can be considered as ill-defined systems. They are characterized
by nonlinearities, time-varying properties, climatic interactions, and many other
factors. It is, therefore, difficult to quantify complex relationships between the input
and the output of a system based on analytical methods. Recently, intelligent
system control based on artificial intelligence has been one of the most prosperous
technologies in the complex system science (Hashimoto 1997). ANN is also a
promising tool for predicting crop yield and offers insight into the casual relationships through the use of sensitivity analyses, but the complex parameterization
and optimum model structure require special attention (Park et al. 2005). The
machine learning tools used to model the effects of environmental conditions on
apparent photosynthesis provided a very powerful modeling alternative to ordinary
curve fitting methods. Their major advantages are the flexibility to select between
accuracy and generality and their robustness against outliers and mixtures of differentia responses (Dalaka et al. 2000). Resources capture and plant responses to
energetic stimuli (solar radiation, temperature, CO 2 concentration, and humidity)
present complex ecophysiological mechanisms, which conduct to dynamic modifications of the canopy development and architecture (Vazquez-Cruz et al. 2010).
In agronomic research, ANNs have been presented as alternative methodology
to modeling and simulating crop biophysical properties. ANN models are specially
2 Mathematical Modeling of Biosystems
71