Deep Learning Enables Instant and Versatile Estimation of Rice Yield Using Ground-Based RGB Images.

Rice (Oryza sativa L.) is one of the most important cereals, which provides 20% of the world's food energy. However, its productivity is poorly assessed especially in the global South. Here, we provide a first study to perform a deep-learning-based approach for instantaneously estimating rice yield using red-green-blue images. During ripening stage and at harvest, over 22,000 digital images were captured vertically downward over the rice canopy from a distance of 0.8 to 0.9 m at 4,820 harvesting plots having the yield of 0.1 to 16.1 t·ha-1 across 6 countries in Africa and Japan. A convolutional neural network applied to these data at harvest predicted 68% variation in yield with a relative root mean square error of 0.22. The developed model successfully detected genotypic difference and impact of agronomic interventions on yield in the independent dataset. The model also demonstrated robustness against the images acquired at different shooting angles up to 30° from right angle, diverse light environments, and shooting date during late ripening stage. Even when the resolution of images was reduced (from 0.2 to 3.2 cm·pixel-1 of ground sampling distance), the model could predict 57% variation in yield, implying that this approach can be scaled by the use of unmanned aerial vehicles. Our work offers low-cost, hands-on, and rapid approach for high-throughput phenotyping and can lead to impact assessment of productivity-enhancing interventions, detection of fields where these are needed to sustainably increase crop production, and yield forecast at several weeks before harvesting.

Effect of soybean husk supplementation on the fecal fermentation metabolites and microbiota of dogs.

In vitro fermentation and in vivo feeding experiments were conducted to characterize the effects of soybean (Glycine max) husk on the fecal fermentation metabolites and microbiota of dogs. An in vitro fermentation study using feces from three Toy Poodle dogs (6.5 ± 3.5 months in age and 2.9 ± 0.4 kg in body weight) revealed that the fecal inoculum was able to ferment soybean husk (supplemented at 0.01 g/mL culture) and increased levels of short chain fatty acids (SCFA) and Bifidobacterium, irrespective of pre-digestion of the husk by pepsin and pancreatin. In a feeding experiment, four Shiba dogs (7-48 months in age and 7.5 ± 1.7 kg in body weight) fed a commercial diet supplemented with 5.6% soybean husk showed an increase in SCFA, such as acetate and butyrate, and lactate, and a decrease in indole and skatole in the feces compared to those fed a 5.6% cellulose diet. Real-time PCR assay showed that soybean husk supplementation stimulated the growth of lactobacilli, Clostridium cluster IV including Faecalibacterium prausnitzii, Clostridium cluster XIVa, Bacteroides-Prevotella-Porphyromonas group but inhibited the growth of Clostridium cluster XI. Both in vitro and in vivo experiments indicated that soybean husk supplementation improves gastrointestinal health through optimization of beneficial organic acid production and increase of beneficial bacteria. Therefore, soybean husk is suggested to be applicable as a functional fiber in the formulation of canine diets.

Factors underlying genotypic differences in the induction of photosynthesis in soybean [Glycine max (L.) Merr].

Crop leaves are subject to continually changing light levels in the field. Photosynthetic efficiency of a crop canopy and productivity will depend significantly on how quickly a leaf can acclimate to a change. One measure of speed of response is the rate of photosynthesis increase toward its steady state on transition from low to high light. This rate was measured for seven genotypes of soybean [Glycine max (L.) Merr.]. After 10 min of illumination, cultivar 'UA4805' (UA) had achieved a leaf photosynthetic rate (Pn ) of 23.2 µmol · m(-2) · s(-1) , close to its steady-state rate, while the slowest cultivar 'Tachinagaha' (Tc) had only reached 13.0 µmol · m(-2) · s(-1) and was still many minutes from obtaining steady state. This difference was further investigated by examining induction at a range of carbon dioxide concentrations. Applying a biochemical model of limitations to photosynthesis to the responses of Pn to intercellular CO2 concentration (Ci ), it was found that the speed of apparent in vivo activation of ribulose-1:5-bisphosphate carboxylase/oxygenase (Rubisco) was responsible for this difference. Sequence analysis of the Rubisco activase gene revealed single nucleotide polymorphisms that could relate to this difference. The results show a potential route for selection of cultivars with increased photosynthetic efficiency in fluctuating light.