Abstract Goats play a significant role in farming communities in semi-arid tropical areas with limited cropping capacity; however, production is limited by endoparasites, especially gastrointestinal nematodes (GINs). Control of GINs is typically mediated by anthelmintic drugs, but can be costly or ineffective where anthelmintic-resistant GINs are present. To manage anthelmintic resistance and improve herd health at lower costs, farmers can implement Targeted Selective Treatment (TST) strategies to treat animals based on performance or health traits. The study aimed to quantify the impacts of plant-based interventions on goat health, nutrition, and parasite infections when applied in a targeted selective feeding regime under an arid environment. Here we trialled a farmer-led TST scheme using a worm diagnostic tool for sheep and goats based on checking five points on the animal body; nose (purulent discharge), eye (colour of the conjunctivae), jaw (subcutaneous pitting oedema), back (body scoring condition, and tail (mild to severe diarrhoea). This method, termed the Five Point Check or FAMACHA, was used to periodically measure goat health, with anthelmintic interventions provided only to individuals in poor condition. In addition to TST with anthelmintic, a plant-TST was trialled on 50% of the farms where goats in borderline or poor condition were supplemented with local bioactive plants (Viscum rotundifolium L. or Terminalia sericea). Plant-TST treatment significantly reduced worm burden (P < 0.001) from a mean of 485 EPG pre-treatment to 269 EPG post-treatment, with 75% of goats having a FEC of ≤ 400 EPG. Further, goats under plant-TST had significantly improved health outcomes (p < 0.001 FAMACHA scores) compared to TST-anthelmintic. Goats under plant-TST were 46.6% less likely to require any anthelmintic treatment. Plant- and anthelmintic-TST had a similar FEC reduction (55.5% and 52.5%, respectively). Plant-TST offers a low-resource means to sustainably manage GINs in in goats in semi-arid conditions.

Refugia for dung-breeding invertebrates can be created by treating only the livestock at the greatest clinical need within a herd or flock with parasiticides, while also safeguarding against risks to production and animal welfare. Although an attractive concept, practical application remains challenging; nevertheless, pragmatic targeted approaches are likely to have positive environmental impacts.

Rapid identification and assessment of animal health are critical for livestock productivity, especially for small ruminants like goats, which are highly susceptible to blood-feeding gastrointestinal nematodes, such as Haemonchus contortus. This study aimed at establishing proof of concept for using bioelectrical impedance analysis (BIA) as a non-invasive diagnostic tool to classify animals at different levels of Haemonchosis. A cohort of 94 intact Spanish bucks (58 healthy; 36 Unhealthy; naturally infected with H. contortus) was selected to evaluate the efficacy of BIA through the measurement of resistance (Rs) and electrical reactance (Xc). Data were collected from live goats using the CQR 3.0 device over multiple time points. The study employed several machines learning models, including Support Vector Machines (SVM), Backpropagation Neural Networks (BPNN), k-Nearest Neighbors (K-NN), XGBoost, and Keras deep learning models to classify goats based on their bioelectrical properties. Among the classification models, SVM demonstrated the highest accuracy (95%) and F1-score (96%), while K-NN showed the lowest accuracy (90%). For regression tasks, BPNN outperformed other models, with a nearly perfect R2 value of 99.9% and a minimal Mean Squared Error (MSE) of 1.25e-04, followed by SVR with an R2 of 96.9%. The BIA data revealed significant differences in Rs and Xc between lightly and more heavily Unhealthy goats, with the latter exhibiting elevated resistance values, likely due to dehydration and tissue changes resulting from Haemonchosis. These findings highlight the potential of BIA combined with machine learning to develop a scalable, rapid, and non-invasive diagnostic tool for monitoring small ruminant health, particularly in detecting parasitic infections like H. contortus. This approach could improve herd management, reduce productivity losses, and enhance animal welfare.

Because it affects economic productivity, food security around the world, and the well-being of animals, livestock health monitoring is an important part of sustainable agriculture. Blood tests and FAMACHA® scoring, which are traditional ways of detecting anemia, have been extensively employed in the management of parasite diseases, especially Haemonchus contortus in small ruminants. While effective, these methods present limitations such as subjectivity, inter-observer variability, and labor-intensive procedures, particularly in large-scale and resource-limited farming systems. Big Data analytics approaches such as Artificial Intelligence (AI) and Machine Learning (ML) methodologies, particularly Natural Language Processing (NLP), are establishing themselves as important instruments for automating, standardizing, and enhancing anemia diagnosis via multi-sensor data integration. This work performed a systematic literature review (SLR) to evaluate the efficacy of AI-driven methodologies, including NLP, deep learning, and classification models (CNNs, SVMs, BPNNs), in improving anemia detection. A structured search across databases (Web of Science, PubMed, Scopus, Google Scholar) identified key advancements in AI-powered FAMACHA® scoring, RF wave-based real-time health monitoring, and BIA applications in parasite detection. Analysis of 1,928 research nodes and 2,897 citation links revealed increasing interest in AI-driven livestock diagnostics, with NLP techniques emerging as a key tool for extracting insights from unstructured veterinary data and scientific literature. Machine learning models have also transformed FAMACHA® scoring by removing human subjectivity. Convolutional neural networks (CNNs) trained on eye mucosa images achieved 92.1% classification accuracy, surpassing traditional FAMACHA® assessments. AI-assisted scoring eliminates observer bias, enhances disease prediction, and enables automated decision-support systems for anemia detection. Similarly, RF-based ultra-wideband radar and RFID sensors allow remote, real-time health monitoring, offering new avenues for precision livestock management. Comparative keyword analysis highlighted 120 mentions of RF waves, 88 mentions of FAMACHA®, and 15 mentions of BIA, confirming that RF-based anemia detection has the most significant research investment. However, NLP remains an underutilized tool in livestock health analytics despite its potential to convert unstructured veterinary data into actionable insights. While promising, BIA, RF-based sensing, and NLP-driven AI models face adoption challenges. Environmental variables, including temperature, humidity, and breed-specific differences, influence BIA and RF signal precision, requiring regular calibration. Moreover, the economic viability and accessibility of AI-driven monitoring systems continue to be issues in commercial cattle management. Future research ought to concentrate on the integration of NLP with multi-sensor AI models, adaptive deep learning algorithms, and mobile veterinary applications to improve scalability, cost, and accessibility in animal health monitoring. Integrating AI-driven NLP with FAMACHA®, RF, and BIA can transform animal health monitoring into a precision-based, automated, and scalable diagnostic solution. These innovations will enhance sustainability, animal welfare, and economic productivity in response to increasing global food demand.

Accurate classification of FAMACHA images is essential for effective anemia detection in small ruminants. However, environmental conditions such as lighting variations (daylight vs. shade) can influence image quality and diagnostic accuracy. This study aimed to evaluate the performance of Support Vector Machine (SVM) and Backpropagation Neural Network (BPNN) models in classifying FAMACHA images taken in daylight (No Shade = 1) and shaded areas (Shade = 0). A dataset of 1000 images (500 per group) was analyzed using machine learning models, with five different data augmentation techniques applied to enhance robustness and using 10 10-fold nested cross-validation techniques. Model performance was evaluated based on accuracy, precision, recall, F1-score, and Cohen’s Kappa statistic to measure classification agreement. The SVM model achieved an overall accuracy of 98%, with a precision and recall of 0.98 for both shaded and non-shaded images. The BPNN model obtained an accuracy of 97%, with precision ranging from 0.94 to 1.00 and recall values from 0.94 to 1.00. Cohen’s Kappa values were 0.959 for SVM and 0.939 for BPNN, indicating strong agreement between predicted and actual classifications. Both models demonstrated high reliability, though SVM performed slightly better in handling lighting variations. The SVM model excelled in balancing classification across both conditions, while BPNN exhibited minor inconsistencies in differentiating shaded images due to slight recall fluctuations. The findings indicate that SVM is the superior model for FAMACHA image classification, particularly in outdoor settings where lighting conditions vary. The high Cohen’s Kappa scores suggest strong inter-model agreement, reinforcing the effectiveness of machine learning in automating FAMACHA classification. Incorporating data augmentation enhanced model generalizability, reducing the risk of misclassification due to lighting inconsistencies. Further research will explore the integration of ensemble learning techniques, combining the strengths of SVM and BPNN to enhance classification robustness. Expanding the dataset to include diverse environmental conditions and using deep learning approaches such as convolutional neural networks (CNNs) may improve performance further. The development of mobile-based AI applications for real-time, automated FAMACHA assessment would significantly benefit precision livestock farming. By leveraging AI-powered image classification, livestock health monitoring can become more efficient, scalable, and accessible, reducing reliance on manual assessments and improving parasite management strategies in small ruminant farming.

Human toxocariasis is a worldwide parasitic disease caused by zoonotic roundworms of the genus Toxocara, which can cause blindness and epilepsy. The aim of this study was to estimate the risk of food-borne transmission of Toxocara spp. to humans in the UK by developing mathematical models created in a Bayesian framework. Parameter estimation was based on published experimental studies and field data from southern England, with qPCR Cq values used as a measure of eggs in spinach portions and ELISA optical density data as an indirect measure of larvae in meat portions. The average human risk of Toxocara spp. infection, per portion consumed, was estimated as 0.016% (95% CI: 0.000-0.100%) for unwashed leafy vegetables and 0.172% (95% CI: 0.000-0.400%) for undercooked meat. The average proportion of meat portions estimated positive for Toxocara spp. larvae was 0.841% (95% CI: 0.300-1.400%), compared to 0.036% (95% CI: 0.000-0.200%) of spinach portions containing larvated Toxocara spp. eggs. Overall, the models estimated a low risk of infection with Toxocara spp. by consuming these foods. However, given the potentially severe human health consequences of toxocariasis, intervention strategies to reduce environmental contamination with Toxocara spp. eggs and correct food preparation are advised.

Parasitic infections in small ruminants, especially gastrointestinal nematodes, pose major risks to animal health, economic growth and productivity. This study examines Fecal Egg Count (FEC), Packed Cell Volume (PCV), and FAMACHA scores at various time intervals to evaluate the incidence of anemia and its correlation with environmental factors like temperature, humidity, and precipitation. Conventional statistical techniques, predictive modeling, and machine learning approaches explored trends, correlations, and forecasting potential. The traditional analysis included descriptive statistics, where the mean FEC was 950 epg (±1120), indicating high variability in parasite loads, while mean PCV was 26.8% (±5.4%), with values as low as 14% in anemic goats. There was a high negative connection (-0.72) between PCV and FEC, and a positive correlation (0.67) between FAMACHA scores and FEC, proving that FEC is reliable source for detecting anemia. Seasonal tendencies were revealed by time series analysis, with warm and humid months exhibiting the highest FEC levels (over 2000 epg). ANOVA results (p < 0.001) showed significant differences in PCV and FEC across FAMACHA score categories, with goats scoring 4 or 5 having an average PCV of 19.6%, significantly lower than those scoring 1 (average PCV: 30.8%). Multiple regression models were developed to predict anemia risk. Linear regression models predicted PCV with an R² of 0.58, considering FEC and environmental factors. Logistic regression classified anemia severity with an 80.2% accuracy, distinguishing between low-risk (FAMACHA 1 & 2) and high-risk (FAMACHA 4 & 5) categories. Advanced machine learning models were implemented to classify FAMACHA scores based on physiological and environmental predictors. Random Forest models achieved 85.4% accuracy, outperforming other classifiers. SHAP analysis revealed that humidity (feature importance: 27%), temperature (22%), and FEC (18%) were the top predictors of anemia risk. A 52-week forecast for PCV, FEC, and FAMACHA scores using ARIMA and Prophet models predicted an increase in anemia risk during weeks 24–38, with expected PCV dropping by 3.5% on average in high-risk months. Finally, clustering analysis (K-means, Hierarchical Clustering) grouped goats into low-risk, moderate-risk, and high-risk clusters, while Kaplan-Meier survival analysis showed that goats with initial PCV below 22% had a 75% probability of developing severe anemia within 6 weeks. This study shows how climate can affect the likelihood of anemia and how decision support systems powered by machine learning can help with sustainable management of small ruminants. This work provides a data-driven paradigm for proactive health monitoring and parasite management in resource-limited situations using prediction models and clustering approaches.

Abstract In recent times, there has been a growing focus on the well-being and healthcare of small ruminants, particularly in relation to the issue of anemia due to infection with blood-feeding gastrointestinal nematodes, such as Haemonchus contortus. The objective of this study was to assess the hematocrit levels in blood samples obtained from small ruminants, specifically goats. Additionally, the study is an attempt to design and create a quick sensor for identifying anemia, which could be conveniently used on farms. A total of 75 mature, intact male Spanish goats were subjected to hematocrit analysis to ascertain their spectrum of hematocrit values and association with anemic conditions. Simultaneously, a unique sensor with user-friendly features was developed to promptly provide farmers with feedback on the anemic status of animals using AI-based machine learning algorithms, enabling timely intervention. The sensor utilized a semi-invasive approach with minimal blood sample requirement. 30 µL of blood was dropped on glycerol soaked Whatman filter paper No. 1, and images were taken at 90 sec, 150 sec, and 270 sec to capture the blood pattern on filter paper soaked in glycerol using a cell phone. These images were correlated with actual hematocrit value obtained from conventional hematocrit analysis. A RMSprop with adjusted learning rate convolution neural network (RMSprop-CNN) based images classification model was developed for the classification of different patterns at different levels of hematocrit. The model was trained on a total of 1,000 images with a training and testing split of 80:20 ratio. For the optimization process, the Adam optimizer with a learning rate of 0.001 is employed. The model is compiled with a categorical cross-entropy loss function, aiming to improve its accuracy metric over training iterations. The initial studies demonstrated a detection accuracy level of 70.31 % at 10 epochs in identifying different hematocrit levels (Level 1; healthy, to Level 5; severely anemic) but improved significantly up to 94.89 % by 100 epochs, resulting in a notable reduction in the time and level of knowledge previously necessary for conducting such evaluations. The present study not only provides valuable insights into the hematological characteristics of small ruminants, but also establishes a foundation for the development of a straightforward and efficient method for detecting anemia. Consequently, this advancement has the potential to enhance the overall care and well-being of animals within agricultural environments.

The 2020 mass mortality of 350 African elephants (Loxodonta africana) in Botswana sparked global concern. These deaths have been linked to cyanobacterial toxins (cyanotoxins) in local watering holes (pans), but evidence remains inconclusive. Our study presents the first detailed spatial analysis that explores the relationship between the ecohydrology of 3,389 regional pans with the locations of deceased elephants. Our findings reveal a significant difference in the distribution of fresh versus older elephant carcasses (p < 0.001), suggesting that the die-off event deviates from typical regional patterns of elephant deaths. We identified twenty pans near the sites of fresh carcasses that experienced more phytoplankton (microalgae or cyanobacteria) bloom events in 2020 (n = 123) compared to the previous 3 years combined (n = 23), exhibiting the highest average phytoplankton biomass of the period 2015-2023 (Normalised Difference Chlorophyll Index > 0.2; p < 0.001). These findings suggest a high risk and likelihood of cyanotoxins as the poisoning source. Our spatial analysis indicates elephants walked an average of 16.5 km (± 6.2 km) and died within 88 hours (± 33 hours) from initial exposure, offering metrics that were previously unknown for elephants. This study presents important new evidence that implicates cyanobacterial toxicity in the 2020 mass die-off and provides a general framework for investigation of future mortality events of large mammals. We underscore the need to integrate spatial analysis and ecohydrological assessments to better monitor and mitigate animal mortality events and inform conservation strategies.

Background Regadenoson is used to induce hyperemia in cardiac imaging, facilitating diagnosis of ischemia and assessment of coronary flow reserve (CFR). While the regadenoson package insert recommends administration of radionuclide tracer 10-20 seconds after injection, peak hyperemia has been observed at approximately 100 seconds after injection in healthy volunteers undergoing cardiovascular magnetic resonance imaging (CMR). It is unclear when peak hyperemia occurs in a patient population. Objectives The goal of this study was to determine time to peak hyperemia after regadenoson injection in healthy volunteers and patients, and whether the recommended image timing in the package insert underestimates CFR. Methods Healthy volunteers (n=15) and patients (n=25) underwent stress CMR, including phase-contrast imaging of the coronary sinus at rest and multiple timepoints after 0.4 mg regadenoson injection. Coronary sinus flow (ml/min) was divided by resting values to yield CFR. Smoothed, time-resolved curves for CFR were generated with pointwise 95% confidence intervals. Results CFR between 60 and 120 seconds was significantly higher than CFR at 30 seconds after regadenoson injection (p < 0.05) as shown by non-overlapping 95% confidence intervals for both healthy volunteers (30 s, [2.8, 3.4]; 60 s, [3.8, 4.4]; 90 s, [4.1, 4.7]; 120 s, [3.6, 4.3]) and patients (30 s, [2.1, 2.5]; 60 s, [2.6, 3.1]; 90 s, [2.7, 3.2]; 120 s, [2.5, 3.1]). Conclusion Imaging at 90 seconds following regadenoson injection is the optimal approach to capture peak hyperemia. Imaging at 30 seconds, which is more aligned with the package insert recommendation, would yield an underestimate of CFR and confound assessment of microvascular dysfunction.

Ferumoxytol is becoming more widely used as an off-label blood-pool contrast agent for MR angiography (MRA) and four-dimensional (4D) flow imaging in pediatric cardiovascular disease. Brand and generic versions of ferumoxytol are available with no information on relative efficacy as a contrast agent and safety profiles. This study evaluates patient safety and image quality of comparable dosages of generic ferumoxytol (GF) versus brand ferumoxytol (BF) with the following hypotheses: (1) Reducing the contrast dosage from 3 to 2 mg/kg will not affect imaging quality and diagnostic accuracy of MRA and four-dimensional 4D flow. (2) GF and BF have similar image quality. (3) GF and BF have similar patient safety profiles. In an IRB-approved retrospective study, changes in vitals/clinical status between baseline, during infusion, and 30 min post-infusion were analyzed in 3 groups: group 1 (3 mg/kg BF, 216 patients, age: 19.29 ± 11.71 years ranging from 2 months to 62 years), group 2 (2 mg/kg BF, 47 patients, age: 15.35 ± 8.56 years ranging from 10 days to 41 years), and group 3 (2 mg/kg GF, 127 patients, age: 17.16 ± 12.18 years ranging from 6 days to 58 years). Both pediatric and adult patients with congenital heart disease (CHD) indications were included within the study. Adverse reactions were classified as mild, moderate, or severe. Quantitative analysis of MR image quality was performed with signal-to-noise ratio (SNR) on MRA and velocity-to-noise ratio (VNR) on 4D flow. Qualitative grading of imaging features was performed by 2 experienced observers. Two-way analysis of variance (ANOVA) and chi-square tests were used for comparison with a P value of ≤ 0.05 used for significance. No statistical difference was found in clinical status and vital signs (P>0.05). No severe reactions were reported. 7.9

Tropospheric helium variations are tightly linked to CO2 due to the co-emission of He and CO2 from natural-gas burning. Recently, Birner et al. (2022a) showed that the global consumption of natural gas has measurably increased the He content of the atmosphere. Like CO2, He is also predicted to exhibit complex spatial and temporal variability on shorter timescales, but measurements of these short-term variations are lacking. Here, we present the development of an improved gas delivery and purification system for the semi-continuous mass spectrometric measurement of the atmospheric He-to-nitrogen ratio (He/N2). The method replaces the chemical getter used previously by Birner et al. (2021, 2022a) to preconcentrate He in an air stream with a cryogenic trap which can be more simply regenerated by heating and which improves the precision of the measurement to 22 per meg (i.e., 0.022 ‰) in 10 min (1σ). Using this “cryo-enrichment” method, we measured the He/N2 ratios in ambient air at La Jolla (California, USA) over 5 weeks in 2022. During this period, He/N2 was strongly correlated with atmospheric CO2 concentrations, as expected from anthropogenic emissions, with a diurnal cycle of 450–500 per meg (max–min) caused by the sea–land breeze pattern of local winds, which modulates the influence of local pollution sources.

Gastrointestinal nematode (GIN) infections are ubiquitous and often cause morbidity and reduced performance in livestock. Emerging anthelmintic resistance and increasing change in climate patterns require evaluation of alternatives to traditional treatment and management practices. Mathematical models of parasite transmission between hosts and the environment have contributed towards the design of appropriate control strategies in ruminants, but have yet to account for relationships between climate, infection pressure, immunity, resources, and growth. Here, we develop a new epidemiological model of GIN transmission in a herd of grazing cattle, including host tolerance (body weight and feed intake), parasite burden and acquisition of immunity, together with weather-dependent development of parasite free-living stages, and the influence of grass availability on parasite transmission. Dynamic host, parasite and environmental factors drive a variable rate of transmission. Using literature sources, the model was parametrised for Ostertagia ostertagi, the prevailing pathogenic GIN in grazing cattle populations in temperate climates. Model outputs were validated on published empirical studies from first season grazing cattle in northern Europe. These results show satisfactory qualitative and quantitative performance of the model; they also indicate the model may approximate the dynamics of grazing systems under co-infection by O. ostertagi and Cooperia oncophora, a second GIN species common in cattle. In addition, model behaviour was explored under illustrative anthelmintic treatment strategies, considering impacts on parasitological and performance variables. The model has potential for extension to explore altered infection dynamics as a result of management and climate change, and to optimise treatment strategies accordingly. As the first known mechanistic model to combine parasitic and free-living stages of GIN with host feed-intake and growth, it is well suited to predict complex system responses under non-stationary conditions. We discuss the implications, limitations and extensions of the model, and its potential to assist in the development of sustainable parasite control strategies.

Goat ownership is prevalent across rural Malawi and provides a vital source of income, nutrition, and food security. However, goat performance is poor, and this presents a risk to individuals and communities who depend on them. Whilst mitigating this through supplementation of persevered forages may be possible, this is a challenge due to limitations of resources and the fact that goats typically free-roam during the dry season. Nutrition is fundamental to health and productivity of any livestock enterprise, it is required to be able to deal with stresses, such as disease, and to enable growth and production. The aim of this study was to characterise the nutritional profile of naturally available forages in Malawi, including the seasonal variation in nutrition. Samples of herbaceous forages and browse were collected over a 17-month period, across four villages (30 farms/smallholders) in Central Malawi. Forages underwent nutritional analysis for crude protein, fibre fractions, and ash/organic matter and NDVI obtained from satellite imagery was used as a measure of forage availability. Forage nutrition and availability were most adequate in the wet season, with higher concentrations of crude protein and a greater availability of herbaceous plants. There were significant differences in low-digestibility fibre fractions between locations, likely due to local factors such as soil and hydrology. The fall in crude protein concentrations from the wet season to the dry season represent a seasonal nutrition-gap which may result in risks to goat health and productivity.### Competing Interest StatementThe authors have declared no competing interest.

Apathy of professionals towards rural communities has severe repercussions with respect to healthcare for both humans and animals. It is observed among the small ruminant (goats and sheep) production community (farmers, research scientists, veterinary doctors, and managers) in the southeastern United States that sericea lespedeza (SL; Lespedeza cuneata), a drought-tolerant fodder, has nutraceutical (nutritional + health) value as a forage for goats and sheep. One of the climate changes (CC) consequences due to global warming is the development of erratic droughts in a temporal manner worldwide. A positive impact of CC is the availability of more land for SL cultivation and increased farmers’ interest towards small ruminant rearing in lieu of cattle. The goal of this study is to develop a geospatial engineering and technology supported SL production suitability model to determine potential areas for cultivation to support profitable small ruminant production. Although SL fodder is a low-maintenance and less climate, terrain, and soil quality sensitive crop, its successful cultivation has specific requirements with respect to weather suitability, such as higher minimum temperature, and soil characteristics, such as non-clay soil with lower bulk density, and open land cover. As the preliminary objective, an automated geospatial model was developed in ArcGIS Pro ModelBuilder platform to determine SL production spatial suitability. As the second objective of this study, this modeling process was followed to determine the suitable SL production locations in the southeastem United States, whose minimum/maximum temperature spatial and annual variation is in an upward trend. A webGIS site was developed as an ArcGIS Online dashboard format, so that farmers interested in switching to small ruminant production in lieu of cattle farming can obtain SL production suitability decision support for their land. This fodder production suitability study in CC scenario would support pasture managers of other countries in the world with similar environmental (weather, climate, soil, and land use) characteristics, as mentioned.

Tropical lands play an important role in the global carbon cycle yet their contribution remains uncertain owing to sparse observations. Satellite observations of atmospheric carbon dioxide (CO) have greatly increased spatial coverage over tropical regions, providing the potential for improved estimates of terrestrial fluxes. Despite this advancement, the spread among satellite-based and in-situ atmospheric CO flux inversions over northern tropical Africa (NTA), spanning 0-24◦N, remains large. Satellite-based estimates of an annual source of 0.8-1.45 PgC yr challenge our understanding of tropical and global carbon cycling. Here, we compare posterior mole fractions from the suite of inversions participating in the Orbiting Carbon Observatory 2 (OCO-2) Version 10 Model Intercomparison Project (v10 MIP) with independent in-situ airborne observations made over the tropical Atlantic Ocean by the NASA Atmospheric Tomography (ATom) mission during four seasons. We develop emergent constraints on tropical African CO fluxes using flux-concentration relationships defined by the model suite. We find an annual flux of 0.14 ± 0.39 PgC yr (mean and standard deviation) for NTA, 2016-2018. The satellite-based flux bias suggests a potential positive concentration bias in OCO-2 B10 and earlier version retrievals over land in NTA during the dry season. Nevertheless, the OCO-2 observations provide improved flux estimates relative to the in situ observing network at other times of year, indicating stronger uptake in NTA during the wet season than the in-situ inversion estimates.

Anemia, often caused by internal parasites like Haemonchus contortus, presents significant health and productivity challenges for small ruminants. The primary goal of this study was to accurately distinguish between healthy and anemic goats using an image classification system focused on eye conjunctiva images. In the initial phase, 1,200 eye conjunctiva images from 75 goats were collected at Fort Valley State University farms over a two-week period using smartphone cameras. These images were randomly divided into training (70%) and testing (30%) datasets, with each group containing three subfolders corresponding to FAMACHA scores of 1, 2, and 3. The validation folder included unique images not found in the other folders. A Convolutional Neural Network (CNN) algorithm was utilized for image analysis, incorporating data augmentation techniques such as Resize, RandomHorizontalFlip, RandomVerticalFlip, and RandomRotation. The CNN model was built on the Google Colaboratory platform using CUDA 11.2 and the PyTorch machine learning framework, incorporating three ConvNet layers. The model training used the Adam Optimizer with a slower learning rate of 0.001 and a weight decay of 0.0001 to prevent exploding gradient issues, alongside ReLU and the cross-entropy loss function over 1000 epochs. Results demonstrate that the Convolutional Neural Network (CNN) model was highly effective in classifying eye conjunctiva images of goats to detect anemia based on FAMACHA scores. The overall precision of 93.9% indicates that the model was accurate in identifying true positive cases. The recall accuracy of 92.1% suggests that the model was successful in capturing most of the true anemic cases from the entire dataset, minimizing the number of false negatives. When examining the precision for each FAMACHA score, the CNN model displayed excellent performance. With a precision of 100% for FAMACHA score 1, the model perfectly identified healthy goats without any false positives. For score 2, the model achieved a precision of 95%, indicating a high level of accuracy in detecting goats with mild anemia. Lastly, for FAMACHA score 3, the precision of the model was 92.9%, demonstrating its effectiveness in identifying goats with more severe anemia. These results show that smartphone-derived images can be a powerful tool in creating an image classification model for monitoring animal health, particularly in detecting anemia in small ruminants. Utilizing smartphone cameras makes the process more accessible, cost-effective, and user-friendly for farmers and veterinary professionals. Despite the impressive performance of the CNN model, the research suggests that there is still room for improvement. By increasing the size of the training dataset, Refining the model development process, such as adjusting the architecture, hyperparameters, or data augmentation techniques, could also contribute to enhanced performance. These improvements would further increase the accuracy and reliability of the model in identifying anemic goats, ultimately leading to better animal health management.