Above-Ground Biomass Estimation in Oats Using UAV Remote Sensing and Machine Learning

Current strategies for phenotyping above-ground biomass in field breeding nurseries demand significant investment in both time and labor. Unmanned aerial vehicles (UAV) can be used to derive vegetation indices (VIs) with high throughput and could provide an efficient way to predict forage yield with high accuracy. The main objective of the study is to investigate the potential of UAV-based multispectral data and machine learning approaches in the estimation of oat biomass. UAV equipped with a multispectral sensor was flown over three experimental oat fields in Volga, South Shore, and Beresford, South Dakota, USA, throughout the pre- and post-heading growth phases of oats in 2019. A variety of vegetation indices (VIs) derived from UAV-based multispectral imagery were employed to build oat biomass estimation models using four machine-learning algorithms: partial least squares (PLS), support vector machine (SVM), Artificial neural network (ANN), and random forest (RF). The results showed that several VIs derived from the UAV collected images were significantly positively correlated with dry biomass for Volga and Beresford (r = 0.2–0.65), however, in South Shore, VIs were either not significantly or weakly correlated with biomass. For Beresford, approximately 70% of the variance was explained by PLS, RF, and SVM validation models using data collected during the post-heading phase. Likewise for Volga, validation models had lower coefficient of determination (R2 = 0.20–0.25) and higher error (RMSE = 700–800 kg/ha) than training models (R2 = 0.50–0.60; RMSE = 500–690 kg/ha). In South Shore, validation models were only able to explain approx. 15–20% of the variation in biomass, which is possibly due to the insignificant correlation values between VIs and biomass. Overall, this study indicates that airborne remote sensing with machine learning has potential for above-ground biomass estimation in oat breeding nurseries. The main limitation was inconsistent accuracy in model prediction across locations. Multiple-year spectral data, along with the inclusion of textural features like crop surface model (CSM) derived height and volumetric indicators, should be considered in future studies while estimating biophysical parameters like biomass.

Development and evaluation of a meat mitochondrial metagenomic (3MG) method for composition determination of meat from fifteen mammalian and avian species

Background Bioassessment and biomonitoring of meat products are aimed at identifying and quantifying adulterants and contaminants, such as meat from unexpected sources and microbes. Several methods for determining the biological composition of mixed samples have been used, including metabarcoding, metagenomics and mitochondrial metagenomics. In this study, we aimed to develop a method based on next-generation DNA sequencing to estimate samples that might contain meat from 15 mammalian and avian species that are commonly related to meat bioassessment and biomonitoring. Results In this project, we found the meat composition from 15 species could not be identified with the metabarcoding approach because of the lack of universal primers or insufficient discrimination power. Consequently, we developed and evaluated a meat mitochondrial metagenomics (3MG) method. The 3MG method has four steps: (1) extraction of sequencing reads from mitochondrial genomes (mitogenomes); (2) assembly of mitogenomes; (3) mapping of mitochondrial reads to the assembled mitogenomes; and (4) biomass estimation based on the number of uniquely mapped reads. The method was implemented in a python script called 3MG. The analysis of simulated datasets showed that the method can determine contaminant composition at a proportion of 2% and the relative error was < 5%. To evaluate the performance of 3MG, we constructed and analysed mixed samples derived from 15 animal species in equal mass. Then, we constructed and analysed mixed samples derived from two animal species (pork and chicken) in different ratios. DNAs were extracted and used in constructing 21 libraries for next-generation sequencing. The analysis of the 15 species mix with the method showed the successful identification of 12 of the 15 (80%) animal species tested. The analysis of the mixed samples of the two species revealed correlation coefficients of 0.98 for pork and 0.98 for chicken between the number of uniquely mapped reads and the mass proportion. Conclusion To the best of our knowledge, this study is the first to demonstrate the potential of the non-targeted 3MG method as a tool for accurately estimating biomass in meat mix samples. The method has potential broad applications in meat product safety.

Estimating target strength of young chum salmon (Oncorhynchus keta)

Although chum salmon (Oncorhynchus keta) is an important fishery resource in Japan, acoustic methods cannot be applied to biomass estimation because the target strength (TS) is unknown. This study clarified the TS for each fork length (FL: 5.5–33.5 cm) of young chum salmon inhabiting the Japanese coastal area to the Bering Sea by measuring free-swimming fish. The size dependences of the TS values were TSmean = 20 log10 FL – 68.0, for both 38 and 120 kHz. This facilitated the estimation of biomass of young salmon using acoustic methods.

Estimating Forest Aboveground Biomass Using Gaofen-1 Images, Sentinel-1 Images, and Machine Learning Algorithms: A Case Study of the Dabie Mountain Region, China

Quantitatively mapping forest aboveground biomass (AGB) is of great significance for the study of terrestrial carbon storage and global carbon cycles, and remote sensing-based data are a valuable source of estimating forest AGB. In this study, we evaluated the potential of machine learning algorithms (MLAs) by integrating Gaofen-1 (GF1) images, Sentinel-1 (S1) images, and topographic data for AGB estimation in the Dabie Mountain region, China. Variables extracted from GF1 and S1 images and digital elevation model data from sample plots were used to explain the field AGB value variations. The prediction capability of stepwise multiple regression and three MLAs, i.e., support vector machine (SVM), random forest (RF), and backpropagation neural network were compared. The results showed that the RF model achieved the highest prediction accuracy (R2 = 0.70, RMSE = 16.26 t/ha), followed by the SVM model (R2 = 0.66, RMSE = 18.03 t/ha) for the testing datasets. Some variables extracted from the GF1 images (e.g., normalized differential vegetation index, band 1-blue, the mean texture feature of band 3-red with windows of 3 × 3), S1 images (e.g., vertical transmit-horizontal receive and vertical transmit-vertical receive backscatter coefficient), and altitude had strong correlations with field AGB values (p < 0.01). Among the explanatory variables in MLAs, variables extracted from GF1 made a greater contribution to estimating forest AGB than those derived from S1 images. These results indicate the potential of the RF model for evaluating forest AGB by combining GF1 and S1, and that it could provide a reference for biomass estimation using multi-source images.

Tree Biomass Estimation in Karst Forest of West Papua, Indonesia

Indonesia is estimated to have 14,5 million hectares of karst areas. The characteristic of karst vegetation is specific, one of which is the dominance of small trees. With all of the potency, their vegetation acts as a significant carbon sequester and store it in biomass. This study aims to estimate and discuss biomass estimation in the karst forest within the Nature Recreational Park of Beriat, a protected area in South Sorong, West Papua. A total of 28 plots were made in the forest using the purposive random sampling method. Tree biomass (DBH ≥10 cm) was estimated using five different allometric equations. The results showed that the biomass was estimated at ca. 264 Mg ha-1 (95 % CI: 135-454 Mg ha-1). While small trees (DBH 10 – 30 cm) only contribute 30 % of the total biomass, about 38 % of the biomass is the contribution of large trees (DBH >50 cm), where Pometia pinnata contributes ca. 39 % of the biomass at plot-level. The use of various allometric equations results in different biomass estimates and biases with deviations ranged from -14.78 % to +17.02 % compared to the reference equation. Therefore, the selection of allometric equations used must be considered carefully to reduce uncertainties in biomass estimation.

Determining the Effects of Nanonutrient Application in Cabbage (Brassica oleracea var. capitate L.) Using Spectrometry and Biomass Estimation with UAV

Geospatial technologies are presented as an alternative for the monitoring and control of crops, as demonstrated through the analysis of spectral responses (SR) of each species. In this study, it was intended to determine the effects of the application of nanonutrients (Zn and Mn) in cabbage (Brassica oleracea var. capitate L.) by analyzing the relationship between the vegetation indices (VI) NDVI, GNDVI, NGRDI, RVI, GVI, CCI RARSa and the content of chlorophyll (CC), from two trials established in the field and in the greenhouse, together with the calculation of dry biomass production in the field through the use of digital models and its further validation. The results indicated that for greenhouse experiments no significant differences were found between the VIs in the implemented treatments, rather for their phenological states. Whereas in the field assays it was evidenced that there were significant differences between the VIs for the treatments, as well as for the phenological states. The SR issued in the field allowed the evaluation of the behavior of the crop due to the application of nanonutrients, which did not occur in the greenhouse, in the same way. The SR also enabled the spectral characterization of the crop in its phenological states in the two trials. All this information was stored in a digital format, which allowed the creation of a spectral library which was published on a web server. The validation of the dry biomass allowed, by statistical analysis, the efficiency of the method used for its estimation to be confirmed.

PEMETAAN BIOMASSA TEGAKAN HUTAN HUJAN TROPIS DI BUKIT MANDIANGIN MENGGUNAKAN CITRA SENTINEL-2 MSI

The preservation of standing biomass is one of the most vital elements for environmental sustainability and the sustainability of the forest itself. One of the actions that can be taken in an effort to maintain the sustainability of forest stand biomass is to map the distribution of biomass, and monitor changes or dynamics of stand biomass from time to time in a sustainable manner. This study aims to build a model based on remote sensing imagery to estimate the total biomass of tropical rainforest stands in Mandiangin Hill, South Kalimantan. The models developed in this study are based on vegetation indices extracted from Sentinel-2 MSI Imagery. A total of ten vegetation indices were tested in this study. For the construction process and validation of stand biomass estimation models, biomass information was measured directly in the field using a number of measuring plots. Stand biomass estimation models were made by correlating stand biomass information from the field with vegetation indices from Sentinel-2 MSI Imagery. The results showed that the most accurate model for estimating the biomass of tropical rainforest stands was 9.5806.exp (0.1454.PSSRa). Where PSSRa is Pigment Specific Simple Ratio. This model has a correlation coefficient (R2) of 0.876, a Mean Absolute Percentage Error (MAPE) of 16.8%, and a Root Mean Square Error (RMSE) of 32.6. The estimation results show that the total biomass of the Bukit Mandiangin tropical rainforest stands is between 11.7 to 998.5 Mg/ha, with an average biomass of 135.8 Mg/ha. Furthermore, the estimation of stand biomass in this study is limited to woody vegetation with a DBH of 10 cm and above. The PSSRa model with various improvements can be used to accurately estimate stand biomass

Root Growth Was Enhanced in China Fir (Cunninghamia lanceolata) after Mechanical Disturbance by Ice Storm

Accurate estimation of forest biomass and its growth potential could be important in assessing the mitigation potential of forest for climate change. However, severe mechanical disturbance such as stem breakage imposed significant changes to tree individuals in biomass structure, which could bring new inaccuracy to biomass estimation. In order to investigate the influence of severe mechanical disturbance on tree biomass accumulation and to construct accurate models for biomass and carbon storage estimation, this paper analyzed the relationship between tree size and biomass for China fir (Cunninghamia lanceolata (Lamb.) Hook) which suffered stem breakage from, and survived, an ice storm. The performance of independent variables diameter (D) and height (H) of China fir, were also compared in biomass estimation. The results showed that D as an independent variable was adequate in biomass estimation for China fir, and tree height was not necessary in this case. Root growth was faster in China fir which had suffered breakage in the main stem by the ice storm, than China fir which were undamaged for at least 7 years after the mechanical disturbance, which, in addition to biomass loss in stem, caused changes in the allocation pattern of the damaged trees. This suggests biomass models constructed before severe mechanical disturbance would be less suitable in application for a subsequent period, and accurate estimations of biomass and forest carbon storage would take more effort.