Ethiopian Crop Type 2020 (EthCT2020) dataset: Crop type data for environmental and agricultural remote sensing applications in complex Ethiopian smallholder wheat-based farming systems (Meher season 2020/21).

Crop type observation is crucial for various environmental and agricultural remote sensing applications including land use and land cover mapping, crop growth monitoring, crop modelling, yield forecasting, disease surveillance, and climate modelling. Quality-controlled georeferenced crop type information is essential for calibrating and validating machine learning algorithms. However, publicly available field data is scarce, particularly in the highly dynamic smallholder farming systems of sub-Saharan Africa. For the 2020/21 main cropping season (Meher), the Ethiopian Crop Type 2020 (EthCT2020) dataset compiled from multiple sources provides 2,793 harmonized, quality-controlled, and georeferenced in-situ samples on annual crop types (7 crop groups; 22 crop classes) at smallholder field level across the complex and highly fragmented agricultural landscape of Ethiopia. The focus was on rainfed, wheat-based farming systems. A nationwide ground data collection campaign (GDCC; Source 1) was designed using a stratification approach based on wheat crop calendar information, and 1,263 in-situ data samples were collected in selected sampling regions. This in-situ data pool was enriched with 1,530 wheat samples extracted from a) the Wheat Rust Toolbox (WRTB; Source 2; 734 samples), a database for wheat disease surveillance data [1] and b) an inhouse farm household survey database (FHSD; Source 3; 796 samples). Obtained field data was labelled according to the Joint Experiment for Crop Assessment and Monitoring (JECAM) guidelines for cropland and crop type definition and field data collection [2] and the FAO Indicative Crop Classification [3]. The EthCT2020 dataset underwent extensive processing including data harmonization, mixed pixel assessment through visual interpretation using 5 m Planet satellite image composites, and quality-control using Sentinel-2 NDVI homogeneity analysis. The EthCT2020 dataset is unique in terms of crop diversity, pixel purity, and spatial accuracy while targeting a countrywide distribution. It is representative of Ethiopia's complex and highly fragmented agricultural landscape and can be useful for developing new machine learning algorithms for land use land cover mapping, crop type mapping, agricultural monitoring, and yield forecasting in smallholder cropping systems. The dataset can also serve as a baseline input parameter for crop models, climate models, and crop disease and pest forecasting systems.

Big data, small explanatory and predictive power: Lessons from random forest modeling of on-farm yield variability and implications for data-driven agronomy.

Context: Collection and analysis of large volumes of on-farm production data are widely seen as key to understanding yield variability among farmers and improving resource-use efficiency. Objective: The aim of this study was to assess the performance of statistical and machine learning methods to explain and predict crop yield across thousands of farmers' fields in contrasting farming systems worldwide. Methods: A large database of 10,940 field-year combinations from three countries in different stages of agricultural intensification was analyzed. Random effects models were used to partition crop yield variability and random forest models were used to explain and predict crop yield within a cross-validation scheme with data re-sampling over space and time. Results: Yield variability in relative terms was smallest for wheat and barley in the Netherlands and for wheat in Ethiopia, intermediate for rice in the Philippines, and greatest for maize in Ethiopia. Random forest models comprising a total of 87 variables explained a maximum of 65 % of cereal yield variability in the Netherlands and less than 45 % of cereal yield variability in Ethiopia and in the Philippines. Crop management related variables were important to explain and predict cereal yields in Ethiopia, while predictive (i.e., known before the growing season) climatic variables and explanatory (i.e., known during or after the growing season) climatic variables were most important to explain and predict cereal yield variability in the Philippines and in the Netherlands, respectively. Finally, model cross-validation for regions or years not seen during model training reduced the R2 considerably for most crop x country combinations, while for wheat in the Netherlands this was model dependent. Conclusion: Big data from farmers' fields is useful to explain on-farm yield variability to some extent, but not to predict it across time and space. Significance: The results call for moderate expectations towards big data and machine learning in agronomic studies, particularly for smallholder farms in the tropics where model performance was poorest independently of the variables considered and the cross-validation scheme used.

Impact of land use/cover change and slope gradient on soil organic carbon stock in Anjeni watershed, Northwest Ethiopia.

Today's agri-food systems face the triple challenge of addressing food security, adapting to climate change, and reducing the climate footprint by reducing the emission of greenhouse gases (GHG). In agri-food systems, changes in land use and land cover (LULC) could affect soil physicochemical properties, particularly soil organic carbon (SOC) stock. However, the impact varies depending on the physical, social, and economic conditions of a given region or watershed. Given this, a study was conducted to quantify the impact of LULC and slope gradient on SOC stock and C sequestration rate in the Anjeni watershed, which is a highly populated and intensively cultivated area in Northwest Ethiopia. Seventy-two soil samples were collected from 0-15 and 15-30 cm soil depths representing four land use types and three slope gradients. Soil samples were selected systematically to match the historical records (30 years) for SOC stock comparison. Four land use types were quantified using Landsat imagery analysis. As expected, plantation forest had a significantly (p < 0.05) higher SOC (1.94 Mg ha-1) than cultivated land (1.38 Mg ha-1), and gentle slopes (1-15%) had the highest SOC (1.77 Mg ha-1) than steeper slopes (> 30%). However, higher SOC stock (72.03 Mg ha-1) and SOC sequestration rate (3.00 Mg ha-1 year-1) were recorded when cultivated land was converted to grassland, while lower SOC stock (8.87 Mg ha-1) and sequestration rate (0.77 Mg ha-1 year-1) were recorded when land use changed from cultivation to a plantation forest. The results indicated that LULC changes and slope gradient had a major impact on SOC stock and C sequestration rate over 30 years in a highly populated watershed. It is concluded that in intensively used watersheds, a carefully planned land use that involves the conversion of cultivated land to grassland could lead to an increase in soil C sequestration and contributes to reducing the carbon footprint of agri-food systems.

High-resolution CMIP6 climate projections for Ethiopia using the gridded statistical downscaling method.

High-resolution climate model projections for a range of emission scenarios are needed for designing regional and local adaptation strategies and planning in the context of climate change. To this end, the future climate simulations of global circulation models (GCMs) are the main sources of critical information. However, these simulations are not only coarse in resolution but also associated with biases and high uncertainty. To make the simulations useful for impact modeling at regional and local level, we utilized the bias correction constructed analogues with quantile mapping reordering (BCCAQ) statistical downscaling technique to produce a 10 km spatial resolution climate change projections database based on 16 CMIP6 GCMs under three emission scenarios (SSP2-4.5, SSP3-7.0, and SSP5-8.5). The downscaling strategy was evaluated using a perfect sibling approach and detailed results are presented by taking two contrasting (the worst and best performing models) GCMs as a showcase. The evaluation results demonstrate that the downscaling approach substantially reduced model biases and generated higher resolution daily data compared to the original GCM outputs.

Pathways to wheat self-sufficiency in Africa.

A growing urban population and dietary changes increased wheat import bills in Africa to 9% per year. Though wheat production in the continent has been increasing over the past decades, to varying degrees depending on regions, this has not been commensurate with the rapidly increasing demand for wheat. Analyses of wheat yield gaps show that there is ample opportunity to increase wheat production in Africa through improved genetics and agronomic practices. Doing so would reduce import dependency and increase wheat self-sufficiency at national level in many African countries. In view of the uncertainties revealed by the global COVID-19 pandemic, extreme weather events, and world security issues, national policies in Africa should re-consider the value of self-sufficiency in production of staple food crops, specifically wheat. This is particularly so for areas where water-limited wheat yield gaps can be narrowed through intensification on existing cropland and judicious expansion of rainfed and irrigated wheat areas. Increasing the production of other sources of calories (and proteins) should also be considered to reduce dependency on wheat imports.

Genetic gains in early maturing maize hybrids developed by the International Maize and Wheat Improvement Center in Southern Africa during 2000-2018.

Genetic gain estimation in a breeding program provides an opportunity to monitor breeding efficiency and genetic progress over a specific period. The present study was conducted to (i) assess the genetic gains in grain yield of the early maturing maize hybrids developed by the International Maize and Wheat Improvement Center (CIMMYT) Southern African breeding program during the period 2000-2018 and (ii) identify key agronomic traits contributing to the yield gains under various management conditions. Seventy-two early maturing hybrids developed by CIMMYT and three commercial checks were assessed under stress and non-stress conditions across 68 environments in seven eastern and southern African countries through the regional on-station trials. Genetic gain was estimated as the slope of the regression of grain yield and other traits against the year of first testing of the hybrid in the regional trial. The results showed highly significant (p< 0.01) annual grain yield gains of 118, 63, 46, and 61 kg ha-1 year-1 under optimum, low N, managed drought, and random stress conditions, respectively. The gains in grain yield realized in this study under both stress and non-stress conditions were associated with improvements in certain agronomic traits and resistance to major maize diseases. The findings of this study clearly demonstrate the significant progress made in developing productive and multiple stress-tolerant maize hybrids together with other desirable agronomic attributes in CIMMYT's hybrid breeding program.

Farmers' perspectives on drivers of rice yield in the Fogera Plain of Ethiopia.

In Ethiopia, rice productivity varies over locations. However, there is limited understanding about rice yield drivers for design appropriate policies and strategies to enhance rice productivity. This study focuses on assessing the patterns of rice yield and its drivers. Data were collected from 220 households and field measurements were made accordingly. Descriptive statistics, the Kruskal-Wallis test, and biplot were to assess yield groups, drivers ranking, and driver yield group association, respectively. Four yield groups were identified 2.1 (Y1) t ha-1, 3.0 (Y2) t ha-1, 4.1(Y3) t ha-1, and 5.2 (Y4) t ha-1. Water stress, low soil fertility, lack of draft animals, shortage of credit, pests, weak extension, and weeds were yield affecting drivers in Y1. Similarly, labor shortage, increase input price, credit, and weed was yield-limiting drivers in Y2 while flooding, poor marketing, and the lack of storage were the drivers in Y3. Poor seed system, post-harvest losses, lack of farm tools, price fluctuation, lack of storage, and poor marketing were drivers in Y4. This study showed that the major drivers that significantly affect yield varied among the yield groups. Perceived drivers of the different yield groups have a better understanding and prospect for strategic target policy and intervention support to minimize yield losses thereby increasing productivity.

Climate change impact on wheat and maize growth in Ethiopia: A multi-model uncertainty analysis.

Ethiopia's economy is dominated by agriculture which is mainly rain-fed and subsistence. Climate change is expected to have an adverse impact particularly on crop production. Previous studies have shown large discrepancies in the magnitude and sometimes in the direction of the impact on crop production. We assessed the impact of climate change on growth and yield of maize and wheat in Ethiopia using a multi-crop model ensemble. The multi-model ensemble (n = 48) was set up using the agroecosystem modelling framework Expert-N. The framework is modular which facilitates combining different submodels for plant growth and soil processes. The multi-model ensemble was driven by climate change projections representing the mid of the century (2021-2050) from ten contrasting climate models downscaled to finer resolution. The contributions of different sources of uncertainty in crop yield prediction were quantified. The sensitivity of crop yield to elevated CO2, increased temperature, changes in precipitations and N fertilizer were also assessed. Our results indicate that grain yields were very sensitive to changes in [CO2], temperature and N fertilizer amounts where the responses were higher for wheat than maize. The response to change in precipitation was weak, which we attribute to the high water holding capacity of the soils due to high organic carbon contents at the study sites. This may provide the sufficient buffering capacity for extended time periods with low amounts of precipitation. Under the changing climate, wheat productivity will be a major challenge with a 36 to 40% reduction in grain yield by 2050 while the impact on maize was modest. A major part of the uncertainty in the projected impact could be attributed to differences in the crop growth models. A considerable fraction of the uncertainty could also be traced back to different soil water dynamics modeling approaches in the model ensemble, which is often ignored. Uncertainties varied among the studied crop species and cultivars as well. The study highlights significant impacts of climate change on wheat yield in Ethiopia whereby differences in crop growth models causes the large part of the uncertainties.

Global high-resolution gridded dataset of N2O Emission and mitigation potential from maize and wheat fields.

This data article provides a high-resolution raw data on Nitrous Oxide (N2O) emission and its mitigation potential from global maize and wheat fields. The analytical results, discussion and conclusion thereof is presented in the related manuscript "Model Comparison and Quantification of Nitrous Oxide Emission and Mitigation Potential from Maize and Wheat Fields at a Global Scale" [1]. This raw dataset has a spatial resolution of 0.0833° × 0.0833°, and comprises pixel level baseline emissions estimated using four empirical N2O emission models (CCAFS-MOT, IPCC Tier-I, IPCC Tier-II and Tropical-N2O) and the model results were validated using experimental data extracted from the literature. Spatially explicit soil, climate and crop management data were obtained from various sources detailed in "Experimental Design, Materials and Methods" section below. N2O mitigation potential were then quantified under four scenarios of excess nitrogen reduction (i.e. 25%, 50%, 75% and 100% reduction of excess nitrogen). We believe that the dataset is a valuable source of information to assess N2O emissions and mitigation measures from maize and wheat fields and to make informed decision. Countries can use this dataset to determine emissions reduction targets in their nationally determined contributions (NDCs) from agricultural sector.

Model comparison and quantification of nitrous oxide emission and mitigation potential from maize and wheat fields at a global scale.

Maize and wheat are major cereals that contribute two-thirds of the food energy intake globally. The two crops consume about 35% of the nitrogen (N) fertilizer used in agriculture and thereby contribute to fertilizer-induced nitrous oxide (N2O) emissions. Thus, estimation of spatially disaggregated N2O emissions from maize and wheat fields on a global scale could be useful for identifying emission and mitigation hotspots. It could also be needed for prioritizing mitigation options consistent with location-specific production and environmental goals. N2O emission from four models (CCAFS-MOT, IPCC Tier-I, IPCC Tier-II and Tropical N2O) using a standard gridded dataset from global maize and wheat fields were compared and their performance evaluated using measured N2O emission data points (777 globally distributed datapoints). The models were used to quantify spatially disaggregated N2O emission and mitigation potential from maize and wheat fields globally and the values were compared. Although the models differed in their performance of capturing the level of measured N2O emissions, they produced similar spatial patterns of annual N2O emissions from maize and wheat fields. Irrespective of the models, predicted N2O emissions per hectare were higher in some countries in East and South Asia, North America, and Western Europe, driven mainly by higher N application rates. The study indicated a substantial N2O abatement potential if application of excess N in the maize and wheat systems is reduced without compromising the yield of the crops through technological and crop management innovations. N2O mitigation potential is higher in those countries and regions where N application rates and current N2O emissions are already high. The estimated mitigation potentials are useful for hotspot countries to target fertilizer and crop management as one of the mitigation options in their Nationally Determined Contributions (NDCs) to the United Nations Framework Convention on Climate Change (UNFCCC).

Supplementary irrigation for managing the impact of terminal dry spells on the productivity of rainfed rice (Oryza sativa L.) in Fogera Plain, Ethiopia.

A terminal dry spell is one of the main limiting factors for rice productions. Therefore, this study was conducted to assess the effect of supplemental irrigation for managing the impact of terminal dry spells on the productivity of different rice varieties grown under rainfed conditions in the Fogera Plain. The experiment was designed in a split-plot design with water regimes as main plot factors and rice varieties as a subplot factor with three replications. The water regimes were: dry planted rainfed rice (farmers practice) (FP), transplanted but not irrigated (IWOI), transplanted and irrigated to saturation (SAT), transplanted and ponding to 1 cm water (PD1), and transplanted and ponding to 3 cm water (PD2). The rice varieties were: X-Jigna (V1), Edget (V2), Hiber (V3), Fogera-1 (V4), and Nerica-4 (V5). The combined effect of PD2 with V1 had the highest grain yield (t/ha) (4.35 t/ha) while FP with V3 had the lowest grain yield (2.12 t/ha). The highest (205%) relative grain yield was obtained when V1 was grown under PD2 followed by V4 under PD2 (199%) and V5 under PD2 (192%) compared to FP with V3. Irrigation water productivity (WPIR) varied between water regimes x varieties from as low as 1.84kg grain mm-1ha-1 for V3 in FP to as high as the yield of 3.07kg grain mm-1ha-1 for V1 in PD2. The highest and lowest net benefits were recorded for V1 grown under PD2 (65, 550 ETB) and for V3 grown under TWOI (33, 500 ETB ha-1), respectively. Hence, the combined application of 3 cm ponding depth (PD2) with X-Jigna (V1) and 1cm ponding depth with Fogera-1 (V4) rice varieties could be suggested as effective terminal stress management to increase the yield and profitability of rainfed rice in the Fogera Plain and similar agro-ecologies.

Misidentification by farmers of the crop varieties they grow: Lessons from DNA fingerprinting of wheat in Ethiopia.

Accurate identification of crop varieties grown by farmers is crucial, among others, for crop management, food security and varietal development and dissemination purposes. One may expect varietal identification to be more challenging in the context of developing countries where literacy and education are limited and informal seed systems and seed recycling are common. This paper evaluates the extent to which smallholder farmers misidentify their wheat varieties in Ethiopia and explores the associated factors and their implications. The study uses data from a nationally representative wheat growing sample household survey and DNA fingerprinting of seed samples from 3,884 wheat plots in major wheat growing zones of Ethiopia. 28-34% of the farmers correctly identified their wheat varieties. Correct identification was positively associated with farmer education and seed purchases from trusted sources (cooperatives or known farmers) and negatively associated with seed recycling. Farmers' varietal identification thereby is problematic and leads to erroneous results in adoption and impact assessments. DNA fingerprinting can enhance varietal identification but remains mute in the identification of contextual and explanatory factors. Thus, combining household survey and DNA fingerprinting approaches is needed for reliable varietal adoption and impact assessments, and generate useful knowledge to inform policy recommendations related to varietal replacement and seed systems development.

Can sub-Saharan Africa feed itself?

Although global food demand is expected to increase 60% by 2050 compared with 2005/2007, the rise will be much greater in sub-Saharan Africa (SSA). Indeed, SSA is the region at greatest food security risk because by 2050 its population will increase 2.5-fold and demand for cereals approximately triple, whereas current levels of cereal consumption already depend on substantial imports. At issue is whether SSA can meet this vast increase in cereal demand without greater reliance on cereal imports or major expansion of agricultural area and associated biodiversity loss and greenhouse gas emissions. Recent studies indicate that the global increase in food demand by 2050 can be met through closing the gap between current farm yield and yield potential on existing cropland. Here, however, we estimate it will not be feasible to meet future SSA cereal demand on existing production area by yield gap closure alone. Our agronomically robust yield gap analysis for 10 countries in SSA using location-specific data and a spatial upscaling approach reveals that, in addition to yield gap closure, other more complex and uncertain components of intensification are also needed, i.e., increasing cropping intensity (the number of crops grown per 12 mo on the same field) and sustainable expansion of irrigated production area. If intensification is not successful and massive cropland land expansion is to be avoided, SSA will depend much more on imports of cereals than it does today.

Identifying potential recommendation domains for conservation agriculture in Ethiopia, Kenya, and Malawi.

Conservation agriculture (CA) is being promoted as an option for reducing soil degradation, conserving water, enhancing crop productivity, and maintaining yield stability. However, CA is a knowledge- and technology-intensive practice, and may not be feasible or may not perform better than conventional agriculture under all conditions and farming systems. Using high resolution (≈1 km(2)) biophysical and socioeconomic geospatial data, this study identified potential recommendation domains (RDs) for CA in Ethiopia, Kenya, and Malawi. The biophysical variables used were soil texture, surface slope, and rainfall while the socioeconomic variables were market access and human and livestock population densities. Based on feasibility and comparative performance of CA over conventional agriculture, the biophysical and socioeconomic factors were first used to classify cultivated areas into three biophysical and three socioeconomic potential domains, respectively. Combinations of biophysical and socioeconomic domains were then used to develop potential RDs for CA based on adoption potential within the cultivated areas. About 39, 12, and 5% of the cultivated areas showed high biophysical and socioeconomic potential while 50, 39, and 21% of the cultivated areas showed high biophysical and medium socioeconomic potential for CA in Malawi, Kenya, and Ethiopia, respectively. The results indicate considerable acreages of land with high CA adoption potential in the mixed crop-livestock systems of the studied countries. However, there are large differences among countries depending on biophysical and socio-economic conditions. The information generated in this study could be used for targeting CA and prioritizing CA-related agricultural research and investment priorities in the three countries.