CBL - Campus del Baix Llobregat

Projecte llegit

Títol: Determinación de un modelo de predicción del área foliar de tres variedades de uva del Penedés.

Estudiant que ha llegit aquest projecte:

Tutor/Cotutor: GIL MOYA, EMILIO

Departament: DEAB

Títol: Determinación de un modelo de predicción del área foliar de tres variedades de uva del Penedés.

Data inici oferta: 07-09-2022      Data finalització oferta: 07-04-2023

Estudis d'assignació del projecte:
    GR ENG AGRÍCOLA
    GR ENG CIEN AGRONOM

Lloc de realització:
EEABB

Segon tutor (UPC): SANCHEZ SANCHEZ, ELENA

Paraules clau:
Àrea foliar, pes sec, peso fresc, model, vinya

Descripció del contingut i pla d'activitats:
El proyecto pretende realizar un modelo de predicción del área foliar de tres
variedades

de uva (Chardonnay, Macabeo y Cabernet Sauvignon) a través de la relación
fisiológica

de este parámetro con el peso fresco y el peso seco de la hoja. Para ello se buscarán
tres

parcelas distintas de cada una de las variedades de uva, localizadas en el área vitícola
del

Penedés. Se realizará un muestreo semanal durante los meses de mayo a julio, y se

determinará tanto el área foliar como el peso fresco y seco en los laboratorios del
DEAB.

Con este proyecto de investigación se busca facilitar el proceso de determinación de
la

deposición de productos fitosanitarios en hojas de vid. El trabajo será de campo, de

laboratorio y de análisis de datos.

Overview (resum en anglès): The Catalan wine sector is in constant evolution due to different reasons, for example, climate
change or the entry into force of new legislation. Therefore, it is very important to carry out
research projects that advance towards a more environmentally resilient viticulture. The
determination of the area of plant collectors is a parameter used in many studies and is not always
easy to obtain. Hence the need to obtain a model that predicts the leaf area of Vitis vinifera with
the varieties of the area where the studies are carried out, the Penedès.
The main objective of the project is to develop a model to predict leaf area through the physiological
relationship of this parameter with the dry weight and fresh weight of the leaf. Three grape
varieties, Chardonnay, Macabeu and Cabernet Sauvignon, have been used for this purpose, as they
are the most representative of the area.
During a total of 8 weeks, 50 leaves were collected from 9 different plots in the Penedès; three of
Chardonnay, three of Cabernet Sauvignon and three of Macabeu. At the end of the sampling stage,
a total of 3,600 leaves were collected and used to create the database. Once collected, they were
processed by fresh weighing, surface planimetry and drying in an oven for 48 hours to finally obtain
the dry weight.
Two simple regression models were developed to predict leaf area (S), one as a function of fresh
weight (FW) and the other in relation to dry weight (DW). The explanatory variables (FP and PS)
show statistically significant differences between their means. The two validated models show a
good correlation between the explanatory variables and the response variable (S) and have a
relatively low prediction error. However, both the variables and the residuals of the models do not
present a normal distribution, nor homoscedasticity.
The model was validated using historical data provided by the Unitat de Mecanització Agrària
(UMA) and data collected as of the 2022-2023 season. Subsequently, it was verified that the
predicted value of surface area adjusts to the observed one, despite not complying with the
normality assumption, it can be concluded that not always, in statistics, the rule must be applied
without examining the specific case available. The study has validated a model that is functional
and simplifies the task of future research related to these three varieties.

© CBLTIC Campus del Baix Llobregat - UPC