The global prevalence of junk and nonorganic foods has made it increasingly difficult to find sustainable organic food options, despite their potential health benefits. This issue is exacerbated by factors such as production scarcity, climate change, and the widespread availability and affordability of junk food. Additionally, there is a lack of research and attention toward orphan crops, which have high nutritional value and significant health benefits, resulting in food and nutrition insecurity. One such underutilized crop is teff, a small grain cereal grown in Eritrea that is highly nutritious, resilient to climate variability, gluten‐free, and rich in dietary fiber. Teff requires low agricultural input, making it a superfood for humans and a valuable animal feed source. The lack of scientific research on teff’s agronomic characteristics, yield potentials, and nutritional benefits hinders its global recognition and distribution. To address this, the Addey Agricultural and Food Cooperative Research and Development team has reviewed existing scientific knowledge on teff. This review covers the botanical description, origin, production, climatic and soil requirements, agronomic practices, and improvement techniques of teff. Also, we explore teff’s nutritional and phytochemical composition, potential health benefits, and use as animal feed. The review highlights teff’s global potential and identifies gaps in current farming practices, recommending modern, technology‐based methods for sustainable production and productivity.

In Eritrea, efforts are being made to tackle the widespread land degradation and promote natural resources and the agricultural sector. However, these efforts lack digital resources assessment, mapping, planning and monitoring. Thus, we developed soil organic carbon (SOC) predictor models for the Western Lowlands of the country, employing 6 machine learning models with different input variables (36, 27, 15, and 08) obtained following these variables selection strategies: (1) all proposed SOC predictor variables; (2) very high multicollinearity (≥0.900 **) reduction; (3) high multicollinearity (≥0.700 **) reduction; (4) the Boruta feature selection algorithm. The results revealed that SOC levels were generally low (mean = 0.43%). Grazing lands, rainfed croplands, and irrigated farmlands all exhibited similarly low SOC values, attributed to unsustainable land management practices that deplete soil nutrients. In contrast, natural forestlands exhibited significantly higher SOC concentrations, highlighting their potential for soil carbon sequestration. Among the tested models, the XGBoost algorithm using 27 covariates achieved the highest predictive performance (RMSE = 0.118, R² = 0.758, RPD = 2.252), whereas the multiple linear regression (MLR) model with 8 variables yielded the lowest performance (RMSE = 0.141, R² = 0.742, RPD = 1.883). Compared to the Boruta-based feature selection, the MLR, PLS, XGBoost, Cubist, and GB models showed performance improvements of 10.41%, 10.06%, 6.72%, 6.50%, and 3.15%, respectively. Rainfall emerged as the most influential predictor of SOC spatial variability in the study area. Other important predictors included temperature, soil taxonomy, SWIR2 and NIR bands from Landsat 8 imagery, as well as sand and clay contents. We conclude that reducing very high multicollinearity is essential for improving model performance across all tested algorithms, while reducing moderate multicollinearity is not consistently necessary. The developed SOC prediction models demonstrate robust predictive capabilities and can serve as effective tools for supporting soil fertility management, land restoration planning, and climate change mitigation strategies in the Western Lowlands of Eritrea.

Population growth and urbanization have resulted in land use/land cover (LULC) changes and affect land surface temperature (LST), which contributes to global warming. The study aimed to detect the LULC changes across Mendefera, Eritrea, from 2002 to 2022, and examine their impacts on LST. Two Landsat-7 ETM and One Landsat OLI_TRI-8 images from 2002, 2012 and 2022 were utilized. A supervised vector machine algorithm was used to classify LULC. Overall accuracy and Kappa coefficient were calculated. Linear regression was performed to show the relationship between Normalized Difference Vegetation Index (NDVI) and LST. The study found remarkable LULC changes in the study area 2002-2022. Built-up and agricultural areas increased by 113.5 and 64.4% respectively, whereas natural vegetation and open areas decreased by 77.6 and 24.8%, respectively. The highest and lowest mean LST were recorded in built-up (35oC) and natural vegetation (29.7oC) areas, respectively. Generally, mean LST decreased by 3.3oC in the study area from 2002 to 2022, and a negative correlation was observed between NDVI and LST. Thus, the study concludes that urbanization causes vegetation degradation, abrupt built-up growth and urban heat islands. The study will help planners and decisions-makers in planning appropriate mechanisms in land use planning and climate change mitigation.

Conversion of natural forests into conventional agricultural lands may lead to significant soil organic carbon losses. Soil organic carbon stock assessment for such land use changes is very crucial for appropriate land use management, soil fertility improvement, ecosystem restoration and climate-change mitigation measures. However, information on the status of soil organic carbon stocks for such land use types is limited in Eritrea and in the Horn of Africa. Thus, the study aimed to assess soil organic carbon stocks for natural forests, continuous cropping, shifting cultivation, and grazing land use types. Fifty-one surface soil samples were collected from these four types of land use around Adi Hakin, Laelay Gash, Eritrea, and analysed. One-way analysis of variance (ANOVA) test results showed that land use changes had highly significant effect on soil organic carbon stock (p < 0.001). The natural forest and continuous cropping land use types recorded the highest (51.69 Mg·ha^-1) and lowest (21.23 Mg·ha^-1) mean soil organic carbon stocks, respectively. Grazing and shifting cultivation had 22.74 and 23.57 Mg·ha^-1 soil organic carbon stocks, respectively. Conversion of natural forest into continuous cropping, grazing, and shifting cultivation in the study area in the long run resulted in losing 58.93, 56.00 and 54.40% of soil organic carbon stocks, and emitting 111.79, 106.25 and 103.20 Mg·CO₂ ha^-1, respectively, to the atmosphere. Thus, the study concludes that conventional agriculture contributes to the atmospheric CO₂ concentration through soil carbon emission. On the contrary, conservation of natural forests is crucial for soil carbon sequestration and atmospheric CO₂ mitigation endeavors.

In Eritrea, in response to widespread land degradation, rehabilitation works are underway. However, the impacts of such works on soil carbon remain unknown. This study assessed the impacts of enclosures by comparing four land uses: permanent enclosure, oxen enclosure, communal grazing, and rainfed farming. Seasonal Sentinel-2 time-series data were extracted using Google Earth Engine and soil organic carbon (SOC) concentration and stock (SOCs) predictive models were developed using soil, topographic, and Sentinel 2 data using machine learning. We compared Random Forest versus Ridge Regression models across four seasons with correlation-based versus Boruta algorithm-based selected features. Kruskal-Wallis H and Dunn post-hoc tests showed that conservation enclosures significantly enhanced soil carbon compared to grazing and rainfed cropping lands. The highest median SOC (1.28%) and SOCs (37.66 Mg ha-1, up to 30 cm soil layer) were found in the permanent enclosure followed by the oxen enclosure, while rainfed farming had the lowest values (0.45% and 14.85 Mg ha-1). These improvements correspond to potential carbon sequestration gains of up to 153.6% when converting degraded land to permanent enclosure. Modelling revealed that SOC was predicted with excellent accuracy (R2 = 0.85, RPD = 2.57) using Ridge Regression and with very good accuracy (R2 = 0.80, RPD = 2.22) with Random Forest with Boruta selected winter season features. For SOCs, the best models of both models achieved similar fair accuracy (R2 = 0.60, RPD = 1.58). Key predictors for SOC were bulk density, Sentinel 2 bands B5 and B8A, while for SOCs B5, Soil Organic Carbon Index (SOCI), B11, and soil pH. Winter (Jan-Mar) was the optimal window season and Boruta-selected features gave the best results for SOC and SOCs prediction. The study demonstrates that enclosure systems are effective for soil carbon restoration in dryland landscapes. This framework highlights the potential for cost effective satellite based monitoring of restoration outcomes. These findings provide evidence based support for scaling up enclosure based restoration in Eritrea and similar dryland regions, contributing to both land degradation neutrality and climate change mitigation goals.

In Eritrea, conventional agriculture is the main stay for most of the population for livelihood where crop productivity is very low, and food insecurity and poverty are common. Though initiatives are being taken to promote the agricultural sector in the country, soil resources are not addressed. Soil studies are very limited, and the country lacks digital soil mapping. Thus, the study aims to develop robust soil organic carbon (SOC) prediction model/s for the western lowland soils where most of the agricultural activities of the country are carried out. We employed MLR, PLS, Cubist, RF, GB and XGB algorisms, and regressed multiple soil, climatic, Landsat 8 (L8) bands and spectral indices against SOC (n=178) through machine learning. The SOC modelling was done in 3 steps with 25, 14, and 06 independent environmental variables to identify the main SOC driver variables. Models performances were evaluated using the RMSE, R2, and RPD metrics. The SOC content in the study area was low with an average of 0.44%, which needs effective soil carbon improvement planning. The accuracies of all the tested models were good enough in all the three steps. The PLS model with 14 input variables gave the highest accuracy (RMSE = 0.1128%, R2 = 0.8268, RPD = 2.4393), and RF model with 06 input variables recorded the lowest (0.1435%, 0.7032, 1.9173). MLR and XGB models improved but GB model worsened with dimensionality reduction. PLS, Cubist, and RF models gave better results with 14 input variables. According to the RPD category, the PLS, XGB, and Cubist models were very good, and RF was good in all the three steps. MLR improved from good to very good but GB deteriorated from very good to good. Rainfall was the most important variable for SOC spatial variability prediction in the study area. Temperature, Green and SWIR2 Landsat 8 bands, NDSI, BR2, Sand, and MSAVI2 also had good capacity to predict the SOC spatial variability. We conclude that all the developed models have good predictive accuracies to be employed in short-mid-long-term planning and monitoring of soil fertility and productivity improvements, ecosystem restoration, and climate-change mitigation action. The study, being the first of its kind in the country, has laid the foundation for digital soil mapping (DSM) and management in the country, and more detailed SOC modelling studies are advised with more soil samples in the study area and other parts of the country for better results.