Determinants of vulnerability of bean growing households to climate variability in Colombia

ABSTRACT

Climate variability largely affects agriculture in the developing world where rainfed agriculture is highly prevalent, and farmers rely on favourable climatic conditions to grow their crops. In Colombia, interannual climate variability can increase human vulnerabilities. Evidence on the vulnerability of farming households to climate variability at the local scale is, however, scarce. Here, we assessed the climate vulnerability and its determinants for a representative sample of 567 bean growing households in Santander, Colombia. We first applied Multiple Correspondence Analysis to calculate a vulnerability index and its components (exposure, sensitivity and adaptive capacity). The vulnerability index is in turn used to classify households into three vulnerability groups, namely, high, medium, and low. We then estimated a Generalized Ordered Probit Model to assess the probability of falling into each vulnerability category according to the household and farm management characteristics. We find that vulnerability is highly variable in the study region, with up to 65% of households classified as highly vulnerable. Geography, access to agronomic training, crop diversification, the percentage of household members making productive decisions and the gender of the household head are the most important factors determining the probability of being more or less vulnerable.

KEYWORDS:

• Vulnerability index
• smallholders farmers
• Generalized Ordered Probit
• climate variability
• drought
• bush bean

Acknowledgments

We acknowledge support from the Climate Change, Agriculture and Food Security (CCAFS), under the project Agroclimas (http://bit.ly/2i3V0Nh). CCAFS is carried out with support from CGIAR Trust Fund Donors and through bilateral funding agreements. For details please visit https://ccafs.cgiar.org/donors. The views expressed in this paper cannot be taken to reflect the official opinions of these organizations. We also acknowledge support from the Colombian Ministry of Agriculture (MADR), who enable the co-production of climate services in the country. We also gratefully acknowledge the Instituto de Hidrologia, Meteorologia y Estudios Ambientales (IDEAM) for providing access to meteorological station data. We thank the Federación Nacional de Cereales y Leguminosas (FENALCE) for support during the course of this work. We also acknowledge support from the Climate Services for Resilient Development (CSRD, http://www.cs4rd.org/) programme (USAID Award#: AID-BFS-G-11-00002-10 towards the CGIAR Fund –MTO 069018). CSRD brings together public and private organizations and agencies committed to realizing the potential to enhance climate resilience and climate-smart policies and practices throughout the world, particularly in developing countries.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Lisset Perez (MSc) is a researcher in socio-economics in the Agroecosystems and Sustainable Landscapes Research Area at the International Center for Tropical Agriculture (CIAT). Lisset holds a Masters in Economics, and focuses on using econometric methods to understand farmer vulnerability and adoption of agricultural technologies.

David A. Rios (BSc) is a researcher in socio-economics in the Decision and Policy Analysis research area at CIAT. David's expertise in socio-economics contributes to the implementation of CIAT's climate services initiatives in Latin America.

Diana C. Giraldo (MSc) is an agricultural meteorologist at the International Center for Tropical Agriculture (CIAT). Diana has been leading activities and projects on climate services in Latin America in the last 5 years. She is now doing a PhD on participatory climate services in agriculture at the University of Reading (UK).

Dr Jennifer Twyman is a social scientist on gender in agriculture at the International Center for Tropical Agriculture (CIAT). At CIAT, Jennifer leads a team on gender inclusion, which implements projects globally to understand gender roles and address gender equity.

Dr Genowefa Blundo-Canto is a researcher on Impact Assessment at the Centre de coopération Internationale en Recherche Agronomique pour le Development (CIRAD) in Montpellier (France). Genowefa holds a PhD in Environment and Development Economics from the Universitá degli Studi Roma Tre in Italy.

Dr Steven D. Prager is a scientist on integrated modelling at the International Center for Tropical Agriculture (CIAT), where he co-leads the Agricultural and Climate Modelling team. Steve holds a PhD in Geography from Simon Fraser University. Previously to working at CIAT he was a Professor of Geography at the University of Wyoming.

Dr Julian Ramirez-Villegas is a climate impacts scientist with experience in both academic and applied research in the areas of climate change and climate variability impacts and adaptation, climate information services, and crop-climate modelling. Julian co-leads a team of researchers in the area of Agricultural and Climate Modelling.

Notes

1 We assume that adaptive capacity is not over-represented because all variables are combined into a single index that is balanced by the variability of the dataset. Furthermore, literature shows that some of the adaptive capacity variables used can also be used as measures of sensitivity (Supplementary Text S1). Future studies could use more comprehensive surveys to ensure inclusion of a greater number of sensitivity factors.

2 See Supplementary Material Table S3, Table S4, Table S5 and Table S6 for the weight of each dimension in the VI and the weight of each variable in each dimension.

3 The difference between the individuals that are in the margin is not statistically tested since it is not possible to apply a discontinuous regression technique given that the threshold is established somewhat subjectively. On the other hand, a test of difference of means, when considering the tails of the distribution, does not contribute information on the difference of the individuals in the thresholds of each one of the three categories of vulnerability.

Additional information

Funding

This work was supported by CGIAR Research Program on Climate Change, Agriculture and Food Security (CCAFS): [grant number P42]. JRV and SDP were partially supported by the Climate Services for Resilient Development (CSRD)-United States Agency for International Development (USAID) Award#: AID-BFS-G-11-00002-10 towards the CGIAR Fund (MTO 069018).