Performing Department
Nutritional Sciences
Non Technical Summary
There are increasing efforts to understand links between ecosystem health and human health, and growing evidence indicates potential relationships between agricultural biodiversity and diet and nutritional diversity, at least a small scales. However, existing research has primarily focused in single countries and small-scale analysis. Previous HATCH work developed a novel method and dataset linking the crop data from 180 countries to the available nutrients provided by those crops, and examined the impact of random crop removal to nutrient availability in a country over time. This work expands previous efforts with aims for three key advances: 1) assessing how specific crops contribute to nutrient availability and stability in different countries; 2) understanding the impact of climate change on crop and nutrient availability in the future; and 3) linking crop diversity with actual diet diversity data across scales. This work will significantly advance our understanding of links between agricultural and nutritional diversity and provide important insights into potential future food security impacts from climate change and other potential market and social disruptions. Such understanding can inform decision-making and food and agricultural policy to safeguard nutrition and biodiversity outcomes.
Animal Health Component
0%
Research Effort Categories
Basic
20%
Applied
70%
Developmental
10%
Goals / Objectives
This project aims to expand upon our novel method of using network science to link crop diversity to nutritional diversity. We seek to further our existing analysis by undertaking three primary objectives and corresponding research:Objective 1: Assess the importance of individual crops to nutrient availability and stability.Objective 2: Link climate change sensitivity of individual crops to nutrient network to assess potential impact of climate change on nutrient availability and stabilityObjective 3: Develop case studies in specific countries to examine the relationship of crop diversity and nutritional stability to the diversity of diets using nutritional intake data
Project Methods
Objective 1: Assess the importance of individual crops to nutrient availability and stabilityObjective 1a: Assess the importance of the top ten most globally produced crops on nutritional stability across countriesThe top ten individual crops produced in the greatest quantity in 2018 were sugar cane, maize, rice, wheat, potatoes, soybeans, cassava, sugar beet, palm fruit and tomatoes (we term "globally important crops") (FAOSTAT n.d.). These crops also correlate highly with the most traded global commodities. We will use these top ten globally important crops as the first scenario for Objective 1, to explore the impact of removal of these crops (in order of production quantity, and randomly) from an individual countryObjective 1b: Assess the importance of the top ten country specific produced crops on nutritional stability across countriesWe will choose a random selection of low, middle, and high-income countries (10 each, as determined by the World Bank classification) to assess the top ten most produced crops in a given country in the most recent year of reliable data (likely 2018). Then, using these top ten crops for each country we will remove each from the crop nutrient network in decreasing order of production amount and randomly, to assess the impact on nutritional stability within a country.Objective 2: Link climate change sensitivity of individual crops to nutrient networks to assess potential impact of climate change on nutrient availability and stabilityA global index of crop climate sensitivity does not currently exist, but there are existing modelling efforts to assess the potential sensitivity and impact of climate change on certain, usually globally produced crops. We will utilize data from globally gridded crop models (GGCM) for available crops. For example, seminal work in this area examined four globally important crops- maize, wheat, rice, and soy (e.g. (Rosenzweig et al 2014). However, more recent datasets have utilized GGCMs in up to 15 crops, including some more traditional crops, such as groundnut (Müller et al 2019). Prior to beginning this work, we will thoroughly review the literature on recent advancements in GGCMs or climate sensitivity indices for crops, to ensure the most recent and robust data are utilized. GGCMs integrate multiple climate prediction models with global crop models to predict yield impacts into the future under regional climate change impacts. These models thus allow for regional variability in climate change intensity and can better predict potential impacts at smaller scales, including countries. Examining the impact of elevated CO2 on crop yields under the RCP8.5 (highest emission CO2 scenario), in 2070-2099 as compared to a 1980-2010 baseline suggests consistent trends in yield shifts globally, with maize and wheat yields increasing in high-latitudes and decreasing in low-latitudes, and the tropics generally not experiencing yield gains (Rosenzweig et al 2014). We will integrate the best available climate impacts data for as many crops as possible (at least 15 based on available current data) to inform the order of removal of crops from our crop-nutrient matrices across countries. We will remove crops first that present the most likely level of crop failure in 2030 and 2050 to assess the relative impacts on available nutrients and nutrient stability in the future across our country dataset.Objective 3: Develop case studies in specific countries to examine the relationship of crop diversity to the diversity of diets using nutritional intake dataWe will focus on case studies of several low, middle and high-income countries to determine the relationship between dietary intake and crop diversity/nutritional stability in a given country. Using existing datasets on dietary intake from different countries, we will link this diet intake data with our existing database of crop diversity and nutritional stability outcomes. We will follow the UNFAO protocol for calculating diet diversity from a number of different datasets spanning our countries of interest. Importantly, all of the nutritional datasets we propose to utilize all capture individual dietary intake data in the previous 24 hours, enabling a standardization of diet data across countries. After developing dietary diversity scores, we will run hierarchical random effects models (similar to (Niles and Brown 2017, Niles et al n.d.) across our case study countries, using countries as a random effect, to predict diet diversity scores as a function of regional crop diversity and a number of relevant demographic household controls (e.g. age, location, race/ethnicity, education). These models, which PI Niles regularly uses in assessing data across countries, will estimate diet diversity as a function of in country crop diversity, nutritional stability, and demographic household controls. Below we detail the specific datasets and case study approaches for each region.Objective 3a: Examine crop diversity and diet diversity/intake in the United StatesWe propose to work in the United States, representative of a high-income country, and utilize the National Health and Nutrition Examination Survey (NHANES) for dietary diversity data. NHANES has been conducted in some format since 1971, and continuously since 1999. This provides a robust dataset to explore over time (at least the last 20 years) to examine whether changes in crop diversity in the United States are related to diet intake. We will utilize the 24 hour dietary recall data from NHANES (Centers for Disease Control n.d.) and calculate diet diversity scores using the UNFAO protocol. We will then link this diet and relevant respondent demographic information and to crop diversity data from the United States over time from at least 1990-2018, including at potentially at the state level.Objective 3b: Examine crop diversity and diet diversity/intake across multiple low and middle income countriesWe then propose to utilize comparable data from the Demographic Health Surveys from the US Agency for International Development (USAID), which are implemented in more than 90 low and middle income countries around the world (Figure 2). The DHS dataset also contains data from multiple rounds of surveys over time in different countries, enabling the opportunity to look at aggregate diet diversity scores from a country over time, in relationship to shifts in crop diversity over time. PI Niles' previous HATCH grant utilized data from the DHS surveys to link diet diversity data (Figure 2) to climate outcomes in 19 countries globally. Here, we will utilize the 24 hour dietary recall data to generate diet diversity scores using the same UNFAO protocol as in the United States. While we have not yet chosen representative low and middle income countries, we will prioritize case studies that have geographic diversity and include countries from multiple regions. Similar to in the United States, we will link the diet diversity scores to relevant respondent demographic information and crop diversity data from these countries over time.