Modelagem matemática do desenvolvimento foliar em mandioca a campo

Abstract

Cassava (Manihot esculenta Crantz) is an important crop in several regions of Brazil for having good agronomic performance in soils with low fertility and low water availability. Mathematical models adapted and tested in different environmental conditions are tools to describe the interactions between plant and environment. The objective of this study was to simulate the leaf number (LN) with two mathematical models, one that considers the linear effect and one that considers the nonlinear effect of air temperature on leaf appearance on the stems of cassava. Considering two versions in each model, one version uses the daily mean temperature (Tmean) and the other uses the daily minimum and maximum temperature (Tmm).The linear model is the model that uses the concept of phyllochron and nonlinear model is the Wang; Engel (WE) model. Model coefficients were estimated from a data set of accumulated number leaves, on the main stem and on the sympodial first and second order branching (HP, RS1 and RS2) of the variety of cassava "Fepagro RS 13", collected in experiments conducted on the field area, Crop Science Department, Federal University of Santa Maria, Santa Maria, RS, Brazil, in the growing seasons 2005/2006 and 2007/2008.The evaluation of the models was performed with independent data collected in the same location in the growing seasons 2006/2007, 2008/2009, 2009/2010 and 2010/2011, with different rainfall availability. The WE model provided the best estimate of leaf number in cassava, with a root mean square error (RMSE) of 3.2 and 3.5 leaves, compared with the phyllochron model with 3.7 and 4.4 leaves for LN general utilizing temperature versions Tmean and Tmm, respectively. Among the versions of the models, for simplicity the versions that uses Tmean can be used, but in warmer regions the version that uses the daily minimum and maximum temperature (Tmm) is suggested.