Vol. 24 No. 4 (2026): Tạp chí Khoa học Nông nghiệp Việt Nam

The research aimed to select sugarcane accession with superior characteristics in terms of morphology, yield, and quality. Experiments were conducted on 5 Kim Tan purple sugarcane accessions (MM1, MM2, MM3, MM4, MM5) collected from sugarcane monoculture sites in Thanh Hoa province. The experiments were arranged sequentially without repetition under the same planting and care conditions. The results showed that the time from planting to emergence of the 5 purple sugarcane accessions did not differ much (29-33 days), but the MM1 variety had many advantages in morphology, growth and quality compared to the remaining accessions such as: the plant had large, round internodes, even purple color, beautiful shine...; The plant grew well with a germination rate of 81%, plant height (213.1cm), stem diameter (4.8cm), number of internodes/plant (19.6 internodes), leaf area  (62.1 dm2/plant) and individual yield of 2.33 kg/plant. In addition to growth, development and yield for purple sugarcane (fresh sugarcane), accession MM1 had higher quality indicators than the other 4 accessions such as Brix level of 17.1%, lowest fiber content in sugarcane (11.46%) and reducing sugar content of 4.2%.

This study aimed to determine the scientific name of the song mai plant based on selected morphological characteristics and DNA barcoding. The song mai samples were evaluated morphologically in combination with DNA barcode sequence analysis of the psbA-trnH region. The morphological assessment showed that the collected song mai samples had a horizontally spreading growth habit, obovate leaves with acute leaf bases, round fruits with yellow flesh, and broadly elliptical seeds with high surface roughness. In addition, analysis of the psbA-trnH sequences indicated that the song mai samples shared up to 99.68% sequence similarity with Prunus salicina sequences in the NCBI GenBank database. Phylogenetic analysis further demonstrated that the song mai samples clustered with Prunus salicina with a high bootstrap value (73%). Therefore, both morphological characteristics and DNA barcoding of the psbA-trnH region support the conclusion that the scientific name of the song mai plant is Prunus salicina, belonging to the family Rosaceae.

This study aimed to detect the presence of Gyrovirus galga 1 (GyVg1) in domestic dogs from several districts in Hanoi, Vietnam. A total of 63 fecal samples were collected from dogs raised he former Dong Anh, Long Bien, and Gia Lam districts between March 2024 and May 2025. Viral DNA was detected using polymerase chain reaction (PCR). The overall positive detection rate of GyVg1 was 12.7%, with rates of 14.29%, 12.12%, and 11.11% in Dong Anh, Long Bien, and Gia Lam districts, respectively. Partial sequences of the overlapping VP2 and VP3 genes from three representative strains were analyzed, revealing high nucleotide similarity rates ranging from 99.40% to 99.85%. Phylogenetic analysis demonstrated that these strains belong to Gyrovirus galga 1 and they clustered closely with strains previously reported in dogs and cats from China in 2022. Additionally, several amino acid substitutions were identified in the VP2 and VP3 proteins of the Vietnamese strains. These findings provide the first evidence of GyVg1 circulation in domestic dogs in Hanoi and lay the groundwork for future studies on the virus’s epidemiology, genetic diversity, and potential cross-species transmission.

This study was conducted to evaluate the contamination levels and densities of selected food hygiene indicator microorganisms in the meat of the freshwater apple snail in the Mekong Delta and to compare variations among locations and seasons. Freshwater apple snails were collected from 88 sampling points across 4 batches from 5/2022 to 4/2023 in Vinh Long, Hau Giang, Soc Trang and Tra Vinh during both the rainy and dry seasons. Snail meat samples were analyzed for total aerobic microorganisms, Coliforms, and Escherichia coli using standard microbiological testing methods, and the results were compared with the TCVN 7046:2019 standard. This study evaluated the contamination levels of total aerobic microorganisms, Coliforms, Escherichia coli and Salmonella of fresh apple snail meat. Results showed that all snail meat samples were contaminated with microorganisms. Among them, samples from Soc Trang had the highest total  aerobic microorganism count (8,5 × 106 CFU/g), but the lowest Coliforms and E. coli counts (1,3 × 104 CFU/g và 1,0 × 103 CFU/g, respectively). The contamination levels and microorganism counts in snail meat did not differ significantly between the rainy and dry seasons. The contamination levels of total aerobic microorganisms, Coliforms, and E. coli in snail meat exceeded the TCVN 7046:2019 standard by 94.2 times, 55.2 times and 6.8 times (Vinh Long); 38.0 times, 36.0 times and 5.2 times (rainy season); and 31.8 times, 67.8 times and 17.8 times (dry season), respectively. These findings provide a scientific basis for consumers to adopt appropriate preliminary processing and cooking measures to ensure food hygiene and safety when consuming apple  snail meat.

The study was conducted to determine the prevalence and antibiotic resistance profile of colistin-resistant E. coli isolated from chickens sold at several markets in Gia Lam commune, Ha Noi. A total of 50 rectal swab samples were randomly collected from chickens at 10 markets for the isolation of colistin-resistant E. coli using culture on selective media supplemented with colistin (2 mg/l). The antibiotic resistance profiles of the isolates were investigated using the disk diffusion method and multiplex-PCR was performed to detect colistin-resistant genes (mcr) genes. The results showed that 56% (28/50) of the swab samples were positive for colistin-resistant E. coli. The isolates exhibited high resistance rates to ampicillin and tetracycline (96.43%), trimethoprim/sulfamethoxazole (85.71%), and streptomycin (85.71%). None of isolates were resistant to cefepime, ceftazidime, and meropenem. The results of PCR indicate that all isolates carried the mcr-1 gene. The findings of this study reflect the current status of antibiotic resistance in E. coli, providing a scientific basis for proposing measures to control antibiotic use in poultry production, thereby reducing the risk of transmission to the food chain.

This study aimed to assess the current situation and propose solutions for applying the Vietbando Land Information System (VBDLIS) in implementing several tasks of the campaign to enrich and clean up the land database in Hung Yen Province. The methods used include collecting secondary documents and data; surveying 107 students using 24 evaluation criteria based on a 5-point Likert scale; and SWOT analysis. The results show that the Vietbando Land Information System (VBDLIS) played a central role in the land database cleansing process. Specifically, 9 out of 12 criteria were rated as very good, particularly those related to data synchronization, connectivity, sharing, and database updating functions. However, the remaining 3 out of 12 criteria-related to the review, classification, and consolidation of land data grouping results were assessed only at a good level. The SWOT analysis reveals that VBDLIS had significant strengths in automation, data standardization, open architecture, and cross-sectoral integration capacity; however, it also faces challenges related to operational costs, information technology skill requirements, data security risks, and a shortage of dedicated personnel. To address these challenges, the study proposes several solutions aimed at improving the quality, consistency, and operational effectiveness of land database management and utilization.

This study was conducted with the aim of screening and evaluating the antifulgal activity against Alternaria alternata of potential native actinomycete genetic resources. The antagonistic activity of actinomycetes strains against A. alternata was evaluated through methods such as co-cultivation, agar diffusion analysis, and molecular techniques to identify genes involved in the synthesis of bioactive compounds in the selected actinomycete strains. Experimental results showed that, using co-cultivation on PDA medium, nine strains with antagonistic activity against A. alternata were screened from 56 actinimycete strains. Notably, Streptomyces sp. VNUA24 showed the strongest antifungal effect, with an inhibition percentage of 77.62 ± 2.63% after 12 days of  co-cultivation. The 7-day culture filtrate of the strain VNUA24 grown in Gause I medium did not display any antagonistic activity against A. alternata. Interestingly, evaluation of the volatile organic compounds produced by Streptomyces sp. VNUA24 revealed inhibitory effects, resulting in an antagonistic activity of 23.13% after 7 days of incubation. The PCR analysis detected the presence of 5 out of 8 biosynthetic genes associated with the production of bioactive metabolites belonging to the PKS-I, PKS-II, and NRPS gene clusters. Overall, the findings indicate that Streptomyces sp. VNUA24 is a promising candidate for the development of biocontrol products in managing crop diseases caused by A. alternata. 

Artificial Intelligence (AI) has been making great strides and is increasingly becoming a useful tool in supporting the solution of complex problems in all areas of life. This study presented the application of convolutional neural networks (CNN) to recognize users' emotional images from videos of children’s activities. The pre-processing and facial recognition process used Multi-Task Cascade Convolutional Network (MTCNN) to detect and recognize faces. MTCNN helps to detect distinctive faces, then provides data to CNN for emotion classification. The dataset used in this study consisted of videos recording the activities of children aged 3 to 5 years old at a preschool. These videos contained emotional images of children in a classroom, they were used to train the learning model to detect how emotional the children express from the videos. As the results, the CNN network achieved 92% accuracy on the training and 95% on the test. The CNN network demonstrated the better learning than conventional neural networks for the same tasks.

Non-performing loans (NPLs) serve as a key indicator of a bank’s financial strength, risk management, and operational efficiency. Effective management of NPLs is essential for maintaining long-term sustainability, ensuring regulatory compliance, and preserving customer trust. This study examined the factors influencing NPLs, measured by the NPLs ratio (NPLR), using data from 30 commercial banks in Vietnam over the period of 2017–2023. Descriptive statistics were employed to assess the overall performance of the banks, while the fixed effects model (FEM) was used to identify the key determinants of the NPLs. The findings revealed that factors such as credit growth, customer deposits, and bank assets saw substantial increases, although growth was significantly impacted by the Covid-19 pandemic. It was observed that a higher return on equity (ROE) and operating costs were associated with lower NPLs, while rising NPLs from the previous year, a larger bank size, and higher loan-to-asset ratios contributed to higher NPLs. These results provide important insights for banks on how to manage NPLs effectively. It is recommended that banks conduct thorough credit assessments, better manage operation costs, maintain strong profitability, and carefully manage existing loan portfolios to ensure the NPL ratio is within the permissible threshold.

In the context of Vietnam promoting artificial intelligence (AI) as a strategic pillar of digital transformation, the question of workforce readiness for the ethical development and application of AI has become increasingly urgent. This study aims to assess Vietnam’s readiness for AI human resource development in alignment with UNESCO’s Recommendation on the Ethics of Artificial Intelligence. The paper applies five pillars: (1) ethical governance;  (2) human resources and skills; (3) data and infrastructure; (4) fairness and inclusion; and (5) monitoring and enforcement and uses content analysis of national strategies, current policies, and data attracted from published reports and journal articiles. The findings reveal that Vietnam’s level of readiness remains limited, uneven across the pillars. Based on the findings, this study proposes a set of policy recommendations aimed at strengthening institutional foundations, integrating ethics training into AI research and development, enhancing data capacity, and fostering open collaboration models. 

Cattle farming increasingly plays a vital role in enhancing the value of the livestock sector and the choice of farming system contributes significantly to Vietnam’s sustainable agricultural development. This study aimed to clarify the determinants influencing farmers’ decisions on cattle farming systems in northern communes of Ninh Binh province. The analysis was based on primary survey data from 94 cattle farming households and applied propensity score analysis method (empirical application of binary Logit regression function). The t-test was also employed to examine the differences in production outcomes between the two farming systems. The results indicate that households with larger forage cultivation areas tended to engage in beef cattle farming, whereas those with more family labor and agricultural land exhibited the opposite trend. Furthermore, the findings reveal that revenue, cost, and profit of beef cattle farms were higher than those of the breeding cattle farms by VND 368,5; 310 and 58,5 million, respectively. Therefore, to promote the cattle sector toward higher value and sustainability, it is necessary to strengthen programs that support land conversion to forage production, provide guidance on efficient land use, and improve access to preferential credit for investment in housing and equipment.

This study investigated the determinants of employee loyalty in garment enterprises in Thai Nguyen province, Vietnam. Survey data from 282 sewing-line workers in two large garment firms were analyzed using structural equation modeling (SEM). The model showed very good fit (CMIN/df = 1.345; CFI = 0.986; TLI = 0.984;  RMSEA = 0.035). The results indicated that compensation and benefits, working conditions, supervisor support, and career development and training all positively affected employee loyalty (P <0.001). Working conditions (β = 0.274) and career development/training (β = 0.271) were the strongest predictors, followed by compensation/benefits  (β = 0.224) and supervisor support (β = 0.199). The model explained 50.5% of the variance in employee loyalty  (R² = 0.505). The findings suggest that improving working conditions and strengthening training/career pathways should be prioritized to enhance workforce retention in garment enterprises.

Rural and urban areas are always in a close reciprocal relationship. Therefore, new rural construction in the current period must be associated with urbanization - inevitable trend. This study aimed to overview some impacts of urbanization on new rural construction in general, and model new rural construction specifically. Based on secondary sources, the research results built a framework of impacts of urbanization on new rural construction construction, in which the most direct impacts were found on the criteria of production organization and rural economic development, environment, culture, income, and the “smart village” criteria. Several policy recommendations were proposed to take advantage of positive impacts and manage to minimize negative impacts from urbanization on model new rural construction, including planning, preserving rural architecture and culture, environmental management, green and sustainable rural development.

Tourism carrying capacity is a critical tool for sustainable tourism management, particularly important for community-based ecotourism (CBET) in Vietnam. Using a systematic review approach following the PRISMA protocol, from an initial 127 documents, the study selected 36 high-quality publications (2013-2023) from Scopus, Web of Science, and domestic sources to propose an indicator set comprising 23 components across 4 groups: environmental (5 indicators), social (6 indicators), economic (6 indicators), and management (6 indicators), of which 12 are core indicators and 11 are supplementary indicators. The weights of the 4 criteria groups were determined through the Analytic Hierarchy Process (AHP) with 7 experts (CR = 4.78%): environmental carrying capacity 28.9%, management carrying capacity 26.0%, social carrying capacity 23.9%, and economic carrying capacity 20.5%. The Tourism Carrying Capacity Index (TCCI) was calculated using the formula TCCI = 0.289 × ENV + 0.239 × SOC + 0.205 × ECO + 0.260 × GOV, with classification thresholds: 0-0.2 (Extremely overloaded), 0.2-0.4 (Overloaded), 0.4-0.6 (Warning), 0.6-0.8 (Balanced), 0.8-1 (Safe). This is the first multidimensional assessment framework specifically designed for CBET in Vietnam, integrating the central role of local communities and the distinctive characteristics of small-scale ecological conservation.