Since 1960 crop production has become increasingly exposed to drought shocks, particularly in Zambia and other southern African countries. The majority of droughts are short duration (<1 month) events that occur during the crop growing season and cover small areas (<1000 km²). These shocks, which can occur along with others such as outbreaks of the crop pest fall armyworm, can reduce staple crop yields by over 20%. Such losses have substantial negative impacts on household food security in rural agricultural communities, but their impacts on people's access to food, whether through production or purchase, are modified by interactions with policy measures and trade networks, with varying effects depending on the location, spatial extent, and time period. Several government actions have mitigated drought impacts on food production. A policy encouraging membership in local farmer cooperatives has substantially boosted yields while reducing marketing costs for participating farmers.  A program to electronically distribute input subsidies also improved yields by encouraging the use of fertilizer and improved cultivars, while a process for certifying new maize cultivars speeds the release of seed varieties that are potentially more drought-tolerant.  However, these policies have spatially variable effects that can undermine many rural households' ability to cope with drought, particularly in remote regions. Poorer farmers are often excluded from cooperatives as they cannot afford to join them, while electronic subsidies primarily benefit farmers who live near larger markets that are better connected to transportation networks. In terms of people's ability to purchase food, government-backed grain stockpiles smooth food prices over the course of a single year at a national scale, reducing the cost of food by 7% during the lean season, but do not affect price stability over longer time periods. At the local scale, rural households improve their food security by purchasing food during lean periods when self-produced supplies become depleted, but this option is more available to households living near tarmac roads where food markets are more accessible. Villages further away from road networks compensate for lack of market access by sharing food between households. Patterns of household food sharing can be explained by a gravity model, which is typically used to understand trade between nations, indicating that food exchange networks are similarly structured from local to global scales.

This project developed an extensive, country-scale, longitudinal dataset of household and market characteristics, including high frequency data collected through a unique cell phone survey method. A number of novel methods were also developed to detect and evaluate drought and its biophysical impacts. These included coupling a new high resolution hydrological model with remote sensing, machine learning, and numerical crop models. These methods were used to identify drought events at spatial and temporal scales that are undetectable by conventional monitoring techniques, and to map the impact of drought events on crop yields at field scales and national extents. A gravity model of food flows was used to explain household food sharing, which was combined with a framework for assessing the tradeoffs between efficiency and resilience in food trade networks. Building on these datasets, a high-performance machine learning model for predicting food insecurity with up to 84% accuracy was also developed. This combination of data and methods facilitated new insights into how drought, governance, and trade interact to affect food security across a broad range of scales, from the short term and local to the longer term and national. These insights can be used to improve food security outcomes. 

Yearly summary reports on key findings, such as regional patterns of farmer seed choice, annual changes in food prices, and regional precipitation patterns, were delivered to 40 communities where survey work was conducted. Recipients included community representatives, agricultural extension agents, and district and provincial representatives from the Ministry of Agriculture.

Last Modified: 01/17/2022
Modified by: Lyndon Estes