Elsevier

Agricultural and Forest Meteorology

Volume 246, 15 November 2017, Pages 64-77
Agricultural and Forest Meteorology

Climate change determined drought stress profiles in rainfed common bean production systems in Brazil

https://doi.org/10.1016/j.agrformet.2017.06.005Get rights and content

Highlights

  • We investigate future breeding directions for rainfed common bean in Goiás state for the 2030 s (2020–2045);.

  • Four drought stress profiles exist: drought free, reproductive, terminal, and joint reproductive-terminal stress.

  • Climate change is projected to substantially increase probability occurrence of reproductive and/or terminal drought.

  • Research on which traits (e.g. early maturity and deep rooting) are needed for drought adaptation is warranted.

Abstract

Reductions in agricultural productivity with consequences for food security associated to climate change are expected in the absence of adaptation. For common beans, across South America, a decrease in climatic suitability has been projected, with heat and drought stresses being the key drivers for such suitability reductions. Breeding programs will play an important role in the adaptation of common beans to the changing climates. However, breeding targets may vary as climate changes during the 21st century. Here, we assess historical and future (2030) probabilities of occurrence, intensity and impact of seasonal variations of drought stress, which is the most important stress for common beans in the Goiás state. We focus on two rainfed (wet and dry) target population environments (TPEs), which encompass ca. 62% of the bean cropped area in the state for 2016, and address potential breeding implications of future projected changes. The analysis revealed two environment groups for both TPEs (highly favorable environment and favorable environment), and four drought stress profiles within these environmental groups (drought stress free, reproductive stress, terminal stress, and joint reproductive-terminal stress) across all climate and management (cultivars and sowing dates) scenarios. Results suggest that, with respect to the historical (1980–2005) period, climate change will make drought more frequent, but less severe, across the region. For the dry TPE, the probability of occurrence of drought stress situations (reproductive and/or terminal) changes from 29.6% (baseline) to ca. 70% (2030, RCP [Representative Concentrations Pathway] 8.5), whereas for the wet TPE, it increases from 16% (baseline) to ca. 43% (2030, RCP 8.5). Results are consistent across RCPs, although benefits from stringent (RCP 2.6) mitigation are evident. We conclude that drought tailoring under climate change is needed for the Embrapa dry bean breeding program.

Introduction

Beans are crucial for food security in much of Latin America and East and Southern Africa, as they are rich in protein and essential vitamins and minerals (Beebe, 2012, Broughton et al., 2003). Common bean (Phaseolus vulgaris L.) is the most widely grown and consumed grain legume in the world and plays an especially significant role in the human diet (Khoury et al., 2014). It is cultivated in a wide range of production systems, representing different climates, soils, cultivars and levels of technology.

In Brazil, currently the largest single bean producer in the world (∼2.5 million tons in 2013, FAOSTAT, 2015, IBGE, 2015), common bean is a staple crop and constitutes a primary source of protein in the diet of the Brazilian population (per capita consumption estimated at 17.8 kg year−1) (Embrapa Arroz e Feijão, 2015). Some 93% (2.8 million ha) of common bean Brazilian production area is under rainfed conditions (IBGE, 2015). In the Goiás State, one of the main bean-producing states in Brazil, crop production is concentrated in the same geographic area, but spread across three growing seasons, namely, wet (sowing from 1 Nov to 31 Dec), dry (sowing from 1 Jan to 28 Feb) and winter (sowing from 1 May to 30 Jun). Due to the mix of agro-environmental conditions and varietal performance in these seasons, they are also referred to as Target Population of Environments (TPEs, Heinemann et al., 2016). Rainfed bean production − the focus of this paper, occurs mainly in the wet and dry TPEs. Observed yields at farmer level in these TPEs are, on average, 1700 (wet) and 1500 kg ha−1 (dry); these are low compared to the winter TPE (2700 kg ha−1) (IBGE, 2015).

Changes in climate have already been reducing global agricultural production by 1–5% per decade over the last 30 years and will continue to pose challenges to agriculture in the coming decades (Challinor et al., 2014, Porter et al., 2014). Reductions in agricultural productivity with consequences for food security associated to climate change are expected in the absence of adaptation in many parts of South America (Magrin and Marengo, 2014). Bean production systems are no exception to these impacts; modelling studies have projected a systematic decrease in the climatic suitability for common beans cultivation across most of South America (including the Goiás state), with heat and drought stresses being the key drivers for such suitability reductions (Beebe et al., 2011, Ramirez-Cabral et al., 2016).

Cultivar adaptation has been shown to be the most effective adaptation measure for reducing vulnerability to climate change (Challinor et al., 2014), while also promoting sustainable development by for example helping to sustainably close yield gaps (Cassman, 1999, van Ittersum et al., 2016). Hence, targeted breeding programs that develop novel climate-adapted varieties in an anticipated manner can substantially contribute to cropping system adaptation, including improving the tolerance to both biotic and abiotic stresses (McClean et al., 2011, Challinor et al., 2016, Rippke et al., 2016). Nevertheless, for breeding programs to successfully establish priorities under climate change, breeders need to identify and characterize (in frequency and intensity) which stresses are most important and how during the 21st century these stresses change across space and time.

Here, we identify and characterize drought stress profiles for the rainfed common bean TPEs (wet, dry) in the Goiás state, and assess their changes under climate change scenarios by 2030 (as derived from the CMIP5 ensemble). Our ultimate aim is to identify potential future breeding directions for rainfed common beans in Goiás state during the 21st century. Our analysis is the first assessment of breeding priorities for common bean, and contributes important new evidence to a growing body of literature on breeding priorities under climate change.

Section snippets

Current and future weather data

Observed historical weather data for 26 weather stations available within the study region (Fig. 1, Supplementary Table S1) for the period 1980–2005 were gathered from a previous study (Heinemann et al., 2016). The dataset consists of daily observations from the Brazilian Meteorological Institute (INMET, http://www.inmet.gov.br) for which gaps and errors had been thoroughly checked and corrected as described in Heinemann et al. (2016) and D’Afonseca et al. (2012, D’Afonseca et al., 2013a,

Projected changes in seasonal temperature, precipitation and crop yield

Projected changes in precipitation and temperature are shown in Fig. 2, Fig. 3 for the wet and dry TPEs (respectively), for all RCPs, for the period 2020–2045, relative to 1980–2005. The ensemble mean temperature increases for the wet (dry) TPE ranged from 1.10 to 1.66 °C (1.03–1.59 °C) across all RCPs, with RCP 8.5 showing the largest warming. The largest temperature increases are projected to occur in the southwest region of the study area (RCP 8.5), the largest grain production region for both

Climate change impacts across environment groups

This study uses environmental classification based on simulated yields for historical and near future (2020–2045) climatic data set on two distinct rainfed common beans target population environments (wet and dry TPEs) in the Goiás State (Brazil), for determining the frequency of environment groups in the near future.

Historically, the main common bean production region for both TPEs is in the south of the state (Supplementary Fig. S6–S7) mainly due to cooler temperatures (wet TPE) and a longer

Acknowledgments

JRV is supported by the CGIAR Research Program on Climate Change, Agriculture and Food Security (CCAFS), which is carried out with support from CGIAR Fund Donors and through bilateral funding agreements. For details please visit https://ccafs.cgiar.org/donors. The views expressed in this document cannot be taken to reflect the official opinions of these organizations. We acknowledge the World Climate Research Program’s Working Group on Coupled Modelling, which is responsible for CMIP, and we

References (60)

  • S. Beebe

    Common bean breeding in the tropics

  • M.M. Benedetti et al.

    Coverage and potential use of a soil profile database in Brazil

    Rev. Bras. Cienc. Solo

    (2008)
  • K.J. Boote et al.

    Modeling photosynthesis of row crop canopies

    HortScience

    (1994)
  • K.J. Boote et al.

    Simulation of crop growth: CROPGRO model

  • K.J. Boote et al.

    Testing effects of climate change in crop models

  • M.H.C. Borges

    Avaliação agronômica, estabilidade e adaptabilidade de genótipos de feijoeiro comum. MS Dissertation

    (2007)
  • W.J. Broughton et al.

    Beans (Phaseolus spp.) −model food legumes

    Plant Soil

    (2003)
  • CGIAR

    Developing Beans That can Beat the Heat

    (2015)
  • K.G. Cassman

    Ecological intensification of cereal production systems: yield potential, soil quality, and precision agriculture

    Proc. Natl. Acad. Sci. U.S.A.

    (1999)
  • A.J. Challinor et al.

    A meta-analysis of crop yield under climate change and adaptation

    Nat. Clim. Change

    (2014)
  • A.J. Challinor et al.

    Current warming will reduce yields unless maize breeding and seed systems adapt immediately

    Nat. Clim. Change

    (2016)
  • M. Charrad et al.

    NbClust: an R package for determining the relevant number of clusters in a data set

    J. Stat. Software

    (2014)
  • K. Chenu et al.

    Environment characterization as an aid to wheat improvement: interpreting genotype-environment interactions by modelling water-deficit patterns in North-Eastern

    Aust. J. Exp. Bot.

    (2011)
  • L.C. Costa et al.

    Potential forcing of CO2, technology and climate changes in maize (Zea mays) and bean (Phaseolus vulgaris) yield in southeast Brazil

    Environ. Res.

    (2009)
  • D.S. D’Afonseca et al.

    Preenchimento dedados climáticos diários faltantes para os municípios do Estado de Rondônia na base de dados de clima da Embrapa Arroz e Feijão

    (2012)
  • D.S. D’Afonseca et al.

    Preenchimento dedados climáticos diários faltantes para os municípios do Estado de Mato Grosso na base de dados de clima da Embrapa Arroz e Feijão

    (2013)
  • D.S. D’Afonseca et al.

    Metodologia para o preenchimento de dados climáticos diários faltantes: uma aplicação para os municípios de Goiás a partir da base de dados de clima da Embrapa Arroz e Feijão

    (2013)
  • M.J. DelPeloso et al.

    Estabilidade e adaptabilidade de genótipos de feijoeiro comum na região Centro-Sul do Brasil

    (2005)
  • A. Eitzinger et al.

    Assessing high-impact spots of climate change: spatial yield simulations with Decision Support System for Agrotechnology Transfer (DSSAT) model

    Mitig Adapt Strateg Glob Change

    (2017)
  • Embrapa Arroz e Feijão

    Dados de conjuntura da produção de feijão (Phaseolus vulgaris L.) no Brasil 1985–2013

    (2015)
  • Cited by (27)

    • Enviromic prediction is useful to define the limits of climate adaptation: A case study of common bean in Brazil

      2022, Field Crops Research
      Citation Excerpt :

      The major limiting factors in common bean production are drought stress, low soil fertility (Beebe et al., 2011; Müller et al., 2014), and nitrogen deficiency due to poor nitrogen fixation (Rao, 2001). In the future, a general trend toward temperature homogenization will act to homogenize common bean yields and increase drought frequency (Heinemann et al., 2017). In Brazil, the common bean production extends from the South to the North and spreads across three distinct growing seasons: wet, dry, and winter.

    • In silico assessment of sowing dates and nitrogen management in common bean crops

      2022, European Journal of Agronomy
      Citation Excerpt :

      Soil N balance processes are those described by Godwin and Jones (1991) and Godwin and Singh (1998). The CSM-CROPGRO model has been widely evaluated and applied in Brazil in common bean studies for irrigation management optimization (Heinemann et al., 2000, 2002; Teixeira et al., 2017; Justino et al., 2019), environment characterization (Heinemann et al., 2016), climatic change (Heinemann et al., 2017; Antolin et al., 2021), yield prediction (Oliveira et al., 2012) and for soybean to assess the relationship of soil organic carbon content and nitrogen (Silva et al., 2020). For crop model performance analysis, the statistical indicators RMSE (root-mean-square error) and MAE (mean absolute error) were calculated to determine the fitting degree between simulated and observed values.

    • Impact assessment of common bean availability in Brazil under climate change scenarios

      2021, Agricultural Systems
      Citation Excerpt :

      Godfray et al. (2010) and Wheeler and von Braun (2013) emphasized the importance of quantifying risks associated with common bean yields in a world expected to have more than 9 billion inhabitants by 2050. In Brazil, Heinemann et al. (2017) performed an analysis of water stress due to low relative air humidity and predicted increased water stress for common beans in the future. The complexity of common bean cropping systems in Brazil makes their in silico representation a complex task and demands a large amount of basic data.

    • Modelling drivers of Brazilian agricultural change in a telecoupled world

      2021, Environmental Modelling and Software
      Citation Excerpt :

      These changes have come in tandem with significant economic changes in the Brazilian farming system, meaning that many farmers are faced by tough economic decisions to ensure the future viability of their businesses (Silva et al., 2020). Future uncertainty is further exacerbated by the spectre of climate change which may bring increased frequency of drought during the second crop and other conditions unfavourable to consistent production from year-to-year (Heinemann et al., 2017; Hampf et al., 2020). To create a tool for examining the range of possible outcomes given such a range of drivers and uncertainties, we set out to develop a spatially-explicit land use/cover change model capable of representing both production and associated land use of multiple agricultural commodities that could be subsequently linked to a System Dynamics model of global trade (see Millington et al., 2017).

    • Common bean

      2020, Crop Physiology Case Histories for Major Crops
    View all citing articles on Scopus
    View full text