Application of Data Envelopment Analysis to Assess Performance Efficiency of Eight Faba Bean Varieties. (1st July 2017)

• Record Type: Journal Article
• Authors: Etemadi, Fatemeh; Hashemi, Masoud; Shureshjani, Roohollah Abbasi; Autio, Wesley R.
• Is Part Of: Agronomy Journal. Volume 109:Number 4(2017)
• Page Start: 1225
• Page End: 1231
• Publication Date: 2017-07-01
• Subjects: Agronomy -- Periodicals; 630
• DOI: 10.2134/agronj2016.10.0617
• Languages: English
• ISSNs: 0002-1962

Abstract

Abstract : Core Ideas: Seed size along with yield should be considered for evaluation of faba bean varieties. Data envelopment analysis can be implemented for ranking the efficiency of faba bean varieties. Varieties ranked high by data envelopment analysis models can promote faba bean production in the Northeast. Faba bean ( Vicia faba L.) seeds are generally large which limits its adoption as cover crop and/or dual purpose cover crop/cash crop due to the high seed cost. The purpose of this study was to apply data envelopment analysis (DEA) by using seed size as input and fresh pod, fresh seed, and L‐Dopa yield as output to evaluate efficiency of eight faba bean varieties. Eight faba bean varieties were evaluated in a 2‐yr study. Four common methods of DEA were used for ranking faba bean varieties. Aquadulce and Delle Cascine out‐yielded other varieties in both years. Averaged over 2 yr Aquadulce and Delle Cascine produced 16.15 and 16.27 Mg ha −1 fresh pod, respectively. However, Aquadulce had 21% lower seed size than Delle Cascine which significantly reduces the cost of production. L‐Dopa yield ranged from 4 kg ha −1 in Sweet Lorane to 46 kg ha −1 in Aquadulce. Although no significant difference was found in fresh pod yield and fresh seed yield of Aquadulce and Delle Cascine, Aquadulce ranked first in both years while Delle Cascine ranked fourth in 2015 and third in 2016 due to its larger seed size and lower L‐Dopa. Bell bean and Sweet Lorane had the smallest seed size yet their efficiency ranked last due to their low fresh pod yield, fresh seed yield, and L‐Dopa yield. Results revealed that DEA could successfully use multiple traits in a single mathematical model without the need for the specification of tradeoffs among multiple measurements.