Open AccessEstimation of common bean (Phaseolus vulgaris) leaf area by a non-destructive method
Author(s) -
Valéria Pohlmann,
Isabel Lago,
Sidinei José Lopes,
Jéssica Taynara da Silva Martins,
Caren Alessandra da Rosa,
Milena Caye,
Diego Portalanza
Publication year - 2021
Publication title -
semina. ciências agrárias
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.268
H-Index - 25
eISSN - 1679-0359
pISSN - 1676-546X
DOI - 10.5433/1679-0359.2021v42n4p2163
Subject(s) - phaseolus , mathematics , cultivar , leaflet (botany) , mean squared error , horticulture , crop , statistics , botany , agronomy , biology
The aim of this study was to develop mathematical models to estimate the leaf area of common bean (Phaseolus vulgaris) in irrigated and non-irrigated water regimes from linear dimensions. An experiment was carried out in a completely randomized design with a 3×2 factorial arrangement (three cultivars: Triunfo, Garapiá and FC 104; two water regimes: irrigated and non-irrigated) with 25 replicates each. A total of 523 trifoliates were collected throughout the crop cycle. The length (L, cm) and width (W, cm) of the central leaflet of the trifoliate were measured and their product (LW) (cm²) calculated. Then, the leaf area of these trifoliates was determined by digital photography methods using ImageJ® software, and using leaf discs. The number of samples required to estimate the leaf area of a trifoliate was determined to define which method is the most accurate to be used as the real leaf area in generating equations to estimate the leaf area in common bean. The relationship between area by digital photographs and the dimensions of the central leaflet of the trifoliate (L, W and LW) was fitted by linear, quadratic and power models. Subsequently, the predictive capacity of the equations was assessed by the root mean square error (cm2 trifoliate-1), mean absolute error (cm2 trifoliate-1), index of agreement and Pearson’s correlation coefficient. Sample size varied between cultivars, water regimes and evaluation methods. It is more appropriate to use the leaf area provided by ImageJ® as real for comparison purposes in generating models to estimate leaf area from linear measurements, in common bean. The general equation LA = 1.092L1.945 can be used in the tested regimes without accuracy losses.