본 연구는 환경 요인을 바탕으로 절화용 국화 생장 예측을 위한 최적의 모델을 개발하는 것을 목표로 하였다. 이를 위해 13개의 모델(Linear Regression, Lasso Regression, Ridge Regression, ElasticNet Regression, K-Nearest Neighbors (KNN), Support Vector Regression (SVR), Neural Network, Decision Tree, Random Forest, XGBoost, AdaBoost, CatBoost, Stacking)의 성능을 R2, MAE, RMSE를 평가 지표 로 비교하였다. 단일 모델 중에서는 Decision Tree가 가장 우수한 성능을 보였으며, R2값은 0.90에서 0.91 사이였다. 앙 상블 모델 중에서는 CatBoost가 가장 높은 성능을 보였으며 (R2=0.90~0.92) Random Forest와 XGBoost 또한 유사한 성 능을 보였다. 전체적으로 트리 기반 앙상블 모델이 국화 생장 예측에 적합한 모델로 나타났다.

Hypertension is characterized by excessive renin-angiotensin system activity, leading to blood vessel constriction. Several synthetic compounds have been developed to inhibit renin and angiotensin-converting enzyme (ACE). These drugs often have adverse side effects, driving the exploration of plant protein-derived peptides as alternative or supplementary treatments. This study assessed the phenolic compound and amino acid content and the antioxidant and antihypertensive activity of 5 South Korean staple crops. Sorghum had the highest phenolic compound content and exhibited the highest antioxidant activity. Millet grains, particularly finger millet (38.86%), showed higher antihypertensive activity than red beans (14.42%) and sorghum (17.16%). Finger millet was found to contain a large proportion of branched-chain, aromatic, and sulfur-containing amino acids, which are associated with ACE inhibition. In particular, cysteine content was positively correlated with ACE inhibition in the crops tested (r=0.696, p<0.01). This study confirmed that the amino acid composition was more correlated with the antihypertensive activity of grains than the phenolic compound content. Finger millet mainly contained amino acids, which have higher ACE inhibitory activity, resulting in the strongest antihypertensive activity. These findings underscore the antihypertensive potential of select crops as plant-based food ingredients, offering insight into their biological functions.

증산은 적정 관수 관리에 중요한 역할을 하므로 수분 스트레스에 취약한 토마토와 같은 작물의 관개 수요에 대한 지식이 필요하다. 관수량을 결정하는 한 가지 방법은 증산량을 측정하는 것인데, 이는 환경이나 생육 수준의 영향을 받는다. 본 연구는 분단위 데이터를 통해 수학적 모델과 딥러닝 모델을 활용하여 토마토의 증발량을 추정하 고 적합한 모델을 찾는 것을 목표로 한다. 라이시미터 데이터는 1분 간격으로 배지무게 변화를 측정함으로써 증산 량을 직접 측정했다. 피어슨 상관관계는 관찰된 환경 변수가 작물 증산과 유의미한 상관관계가 있음을 보여주었다. 온실온도와 태양복사는 증산량과 양의 상관관계를 보인 반면, 상대습도는 음의 상관관계를 보였다. 다중 선형 회귀 (MLR), 다항 회귀 모델, 인공 신경망(ANN), Long short-term memory(LSTM), Gated Recurrent Unit(GRU) 모델을 구 축하고 정확도를 비교했다. 모든 모델은 테스트 데이터 세트에서 0.770－0.948 범위의 R2 값과 0.495mm/min－ 1.038mm/min의 RMSE로 증산을 잠재적으로 추정하였다. 딥러닝 모델은 수학적 모델보다 성능이 뛰어났다. GRU 는 0.948의 R2 및 0.495mm/min의 RMSE로 테스트 데이터에서 최고의 성능을 보여주었다. LSTM과 ANN은 R2 값이 각각 0.946과 0.944, RMSE가 각각 0.504m/min과 0.511로 그 뒤를 이었다. GRU 모델은 단기 예측에서 우수한 성능 을 보였고 LSTM은 장기 예측에서 우수한 성능을 보였지만 대규모 데이터 셋을 사용한 추가 검증이 필요하다. FAO56 Penman-Monteith(PM) 방정식과 비교하여 PM은 MLR 및 다항식 모델 2차 및 3차보다 RMSE가 0.598mm/min으로 낮지만 분단위 증산의 변동성을 포착하는 데 있어 모든 모델 중에서 가장 성능이 낮다. 따라서 본 연구 결과는 온실 내 토마토 증산을 단기적으로 추정하기 위해 GRU 및 LSTM 모델을 권장한다.

농가에서 곤충자원을 활용한 치유농업프로그램의 개발 요구가 증가함에 따라 기존 곤충 돌보기 중심의 치유프로그램과 더불어 활용할 수 있는 야간 곤충채집 활동을 치유농업프로그램으로 개발하 고 시범 적용하였다. 야간곤충채집 활동은 대학생 9명에게 사전에 참여 동의를 획득하고, 전남대학교 학술림(장성)에서 1박 2일간 진행하였다. 18시부터 24시까지 야간 채집 활동과 익일 오전에는 채집한 곤충을 이용한 곤충 표본 제작 활동으로 프로그램을 구성하였다. 프로그램의 효과 분석을 위해 사전 (활동 전), 중간(채집활동 직후), 사후(표본제작 직후)에 뇌파, 맥파 검사 그리고 설문조사를 실시하였 다. 뇌파 분석 결과, 야간 채집활동 직후 뇌파의 기초율동이 향상(10.67→11.44)하였다(p<0.05). 요인 분석을 통한 스트레스 분석 결과, 내적 스트레스는 3.22(사전)→2.96(사후)로 감소(p<0.1)하였고, 문제 수행능력 관련 스트레스는 2.92(중간)→2.70(사후)로 감소하였다. 이러한 결과는 야간 채집 활동 속에 서 이용되는 근육의 움직임이 뇌 기능 향상과 스트레스 감소에 영향을 줄 것으로 생각된다. 참여자가 적은 연구의 한계를 극복하기 위해 향후 연령대별, 성별 등 다양한 참여자를 대상으로 추가적인 채집 프로그램의 개발 및 적용이 필요하다.

We analyzed comparatively the ground beetle communities of 1993 and 2023, with 30 year interval, in broadleaf and coniferous stands of the Gwangneung forest, a biosphere reserve in South Korea. 2,906 individuals (53 species, 38 genera, and 15 families) were collected from broadleaf stands and 1,782 individuals (36 species, 32 genera, and 14 families) from coniferous stands in 1993. The dominant species was Synuchus cycloderus in both stands. In 2023, 1,682 individuals (110 species, 78 genera, 21 families) were collected from broadleaf stands and 990 individuals (92 species, 67 genera, 22 families) from coniferous stands. In 2023, the dominant species were Eucarabus (Parhomopterus) sternbergi sternbergi in broadleaf stands and Coptolabrus jankowskii jankowskii in coniferous stands. Species richness (R'), species evenness (E'), species diversity (H'), and dominance (DI) were all higher than in broadleaf (R'=15.472, E'=0.336, H'=1.601, and DI=0.371) and coniferous (R'=13.756, E'=0.369, H'=1.324, and DI=0.326) forests in 2023.

Until now, in case of Palaearctic region, only one species, Nesendaeus monochrous in this genus was recorded from China. but the biology and life cycle are not clearly known, yet. This species collected from the bud of Glochidion chodoense in Korea for the first time and closely related to Heterochyromera imerodeus Kojima & Morimoto which was recorded only in Japan,. But Nesendaeus monochrous is easily distinguishable from H. imerodeus by the relatively longer elytra with bisinuately round apices. Heterochyromera imerodeus Kojima & Morimoto can be found from Eurya emarginata and Glochidion obovatum in Japan. So, this two species are closely related each other and need more taxonomic study about the generic synonymy or changing the generic position.

The genus Silpha Linnaeus, 1758 in Silphinae includes 33 species worldwide. This group is flightless and known to feed on invertebrate carcasses from the ground surface. In South Korea, 3 species have been recorded, and S. koreana Cho& Kwon, 1999 is an endemic species, distributed only Mt. Taebaek and Mt. Mindung. Up to date the immature stages of the species has never been studied although the species has significant value as an endemic species. The present study is the very first one ever conducted describing the larval morphological characters of S. koreana along with images of larval stages. Furthermore, we report observations on developmental lengths for each stages of the species.

This study was aimed to isolate bacterial inoculants producing chitinase and evaluate their application effects on corn silage. Four corn silages were collected from four beef cattle farms to serve as the sources of bacterial inoculants. All isolates were tested against Fusarium graminearum head blight fungus MHGNU F132 to confirm their antifungal effects. The enzyme activities (carboxylesterase and chitinase) were also measured to isolate the bacterial inoculant. Based on the activities of anti-head blight fungus, carboxylesterase, and chitinase, L. buchneri L11-1 and L. paracasei L9-3 were subjected to silage production. Corn forage (cv. Gwangpyeongok) was ensiled into a 10 L mini silo (5 kg) in quadruplication for 90 days. A 2 × 2 factorial design consists of F. graminearum contamination at 1.0104 cfu/g (UCT (no contamination) vs. CT (contamination)) and inoculant application at 2.1 × 105 cfu/g (CON (no inoculant) vs. INO (inoculant)) used in this study. After 90 days of ensiling, the contents of CP, NDF, and ADF increased (p<0.05) by F. graminearum contamination, while IVDMD, acetate, and aerobic stability decreased (p<0.05). Meanwhile, aerobic stability decreased (p<0.05) by inoculant application. There were interaction effects (p<0.05) on IVNDFD, NH3-N, LAB, and yeast, which were highest in UCT-INO, UCT-CON, CT-INO, and CT-CON & INO, respectively. In conclusion, this study found that mold contamination could negatively impact silage quality, but isolated inoculants had limited effects on IVNDFD and yeast.

A total of 31 bacterial strains were isolated from the Geum River basin in the Republic of Korea during our investigation of indigenous prokaryotic species. The isolated bacterial strains had high 16S rRNA gene sequence similarity (>98.7%) with those of validly published bacterial species, which have not been reported in Republic of Korea. The 31 bacterial strains were phylogenetically diverse and assigned to 4 phyla, 8 classes, 18 orders, 21 families, and 27 genera. At the genus level, the unreported species were affiliated with Kineococcus, Pedococcus, Rhodoluna, Salinibacterium, Rhodoluna, Arthrobacter, Williamsia, Nakamurella, Nocardioides of the class Actinobacteria, Patulibacter of the class Thermoleophilia, Pontibacter, Hymenobacter of the class Cytophagia, Flavobacterium of the class Flavobacteriia, Geomicrobium of the class Bacilli, Brevundimonas, Gellertiella, Rhizobium, Paracoccus, Taonella, Sphingomonas of the class Alphaproteobacteria, Burkholderia, Polaromonas, Hydrogenophaga, Chitinilyticum, Azospira, Zoogloea of the class Betaproteobacteria, and Pseudomonas of the class Gammaproteobacteria. The unreported bacterial species were further characterized by examining their morphological, cultural, physiological, and biochemical properties. The detailed descriptions of the 31 bacterial strains were provided.

Global concerns have grown regarding emerging infectious diseases (EIDs) caused by previously unknown pathogens. Considering that strengthening surveillance capacity for unknown diseases is one of the core capacities for preparedness and early response to EIDs, identifying areas with poor capacity could be beneficial to prioritize regions for the improvement of surveillance. In this regard, we aimed to develop prediction models to identify high risk areas for low surveillance capacity for unknown diseases in a global scale. Unexplained death events reported between 2015 and 2019 were collected from two internet-based surveillance systems, ProMED-mail and Global Public Health Intelligence Network. From the reports, the number of reported unexplained deaths at the first report and the time gap between death and report were extracted as measures for sensitivity and timeliness of surveillance capacity, respectively. Using geographical locations of the reports and published global scale spatial data, including demographic, socioeconomic, public health and geographical variables, we fitted two boosted regression tree models to predict regions with the low sensitivity and timeliness. The performance of prediction model for the low sensitivity showed moderate validity, but in terms of the model for timeliness, the performance was unreliable. Therefore, we provided predicted risk only for low sensitivity. The mean predicted risks of low sensitivity were, respectively, 45.2%, 37.4%, 12.5%, and 3.0% in low-income, lower middle-income, upper middle-income, and high-income countries. Enhancing surveillance capacity in low-income countries is highly required, given the predicted low level of sensitivity despite the importance of early response.