Over the years, the growth in chicken production has brought with it certain problems. Pollutants from poultry farms cause environmental and health problems. Various measures must be taken to properly manage these issues and mitigate their effects. Proper ventilation and litter management within the broiler house are the primary measures. Additionally, the treatment of pollutants in litter using UV rays is an alternative method. In this study, NH3 and CO2 gases originating from litter material obtained from a broiler farm operating in the Bursa region were treated using UV rays on a laboratory scale. Litter samples were exposed to UV-A and UV-C rays at different exposure times and distances, and their treatment efficiencies were determined. The data obtained as a result of the study show that UV rays are effective in reducing livestock-derived pollutant gases. When the treatment efficiencies of the parameters are evaluated independently, the highest CO2 treatment efficiencies in terms of exposure time were observed at the 120th minute (93.56% for UVA and 100% for UVC). For NH3 gas, the highest efficiencies over time were recorded at the 60th minute (90.50% for UV-A) and at the 90th minute (95.33% for UV-C). When the application distances were examined individually, it was determined that the highest purification performance for both gases and lamp types (UVA and UVC) was achieved at 40 cm. In this context, it has been determined that the UV-C lamp shows the highest treatment efficiency for CO2 gas, and the UV-A lamp for NH3 gas, at a distance of 40 cm for both gases.
Field pea (Pisum sativum L.) is an important legume crop whose productivity is strongly influenced by environmental conditions, making the evaluation of genotype performance and stability essential for breeding and sustainable cultivation. The aim of the present study was to evaluate local field pea varieties grown under the climatic conditions of Northeastern and Central Southern Bulgaria with respect to yield-related traits and to assess the effects of genotype, environment, and genotype × environment interaction on trait expression and stability. Field experiments were conducted during 2024 and 2025 at two locations (Ruse and Sadovo) using four local field pea cultivars arranged in a randomized complete block design with three replications. Plant height, number of primary branches, number of productive pods per plant, number of seeds per pod, number of seeds per plant, and seed weight per plant were recorded. The data were analyzed using multifactor analysis of variance (ANOVA), Tukey's HSD test, and stability statistics including Shukla's stability variances, Wricke's ecovalence, and Kang's rank-sum criterion. The results demonstrated significant effects of genotype, year, and their interactions on most of the evaluated traits, indicating a substantial influence of environmental conditions on yield formation. In general, higher values for several productivity-related traits were observed under the conditions of the Sadovo environment. The cultivar Ruse 1 exhibited the highest stability for the traits number of seeds per plant and seed weight per plant and achieved favorable rankings according to Kang's stability criterion, whereas Mir showed superior performance for several yield-related characteristics. The obtained results provide useful information for the evaluation of local field pea germplasm and may support breeding programs and the selection of genotypes adapted to different environmental conditions in Bulgaria.
The increasing pressure on agricultural soils due to intensive land use and excessive fertilization practices threatens soil quality and long-term food security. This study aimed to determine the effects of biochar application on soil carbon dynamics as well as on the chemical and biological properties of a soil cultivated with wheat for three consecutive years. Biochar, produced from woody residues via gasification, was applied at a rate of 5.0 t ha⁻¹ and incorporated into the topsoil (0–20 cm). Wheat was cultivated consecutively for three years following biochar application. Soil samples were collected from control and biochar-amended plots and analysed for mineral associated organic carbon (MAOC), particulate organic carbon (POC), bulk soil organic carbon, chemical properties, nutrient availability and microbial activity. Biochar was found to significantly increase organic carbon in all three fractions, with the greatest increase observed in POC. Biochar application did not significantly affect soil pH but reduced electrical conductivity while increasing soil organic carbon and cation exchange capacity. Exchangeable Ca, Mg and DTPA-extractable Mn contents were significantly higher in biochar-amended soils, whereas plant-available P and DTPA-extractable Cu and Zn decreased. In terms of soil biology, the amendment reduced basal respiration while increasing microbial biomass carbon. This suggests a potential for enhanced carbon stabilisation and altered microbial dynamics. These findings demonstrate that the residual effects of a single biochar application persist for several years, influencing soil carbon pools, nutrient availability and microbial activity. The contrasting responses of different nutrients emphasise the importance of optimising biochar management strategies to support long-term soil fertility and sustainable crop production.
In the face of climate change, drought stress represents a critical challenge to the sustainability of maize (Zea mays L.) cultivation. This study explored the potential of Arbuscular Mycorrhizal Fungi (AMF) to enhance maize tolerance to drought. The experiment was conducted in a glass greenhouse of the Soil Fertilizer and Water Resources Central Research Institute, Ankara, employing a factorial design with five AMF treatments under three water stress levels (70%, 50%, and 30% of field capacity). Maize plants inoculated with AMF demonstrated notable improvements in growth parameters under drought conditions. At 70% field capacity, the AMF treatment with Rhizophagus intraradices (M3) significantly increased plant height (PH) by up to 104.2 cm and fresh weight (FW) by 35.7 g, compared to control values of 95.7 cm and 27.5 g, respectively. Similarly, at 50% field capacity, AMF treatments sustained higher stem diameter (SD) measurements, with a notable instance being Rhizophagus irregularis (M2) at 5.44 cm, against a control value of 4.74 cm. The resilience of AMF-treated plants was further underlined by the improved root dry weight (RDW) in the face of severe drought stress (30% field capacity), where Glomus iranicum (M5) treatments resulted in RDW of 1.20 g compared to 0.73 g in non-AMF-treated plants. These findings substantiate the hypothesis that mycorrhizal symbiosis can significantly mitigate the effects of drought stress on maize, suggesting a viable strategy to enhance crop resilience and productivity in water-limited environments.
This study evaluated the effect of harvest time on the essential oil (EO) yield and chemical composition of Helichrysum italicum (Roth) G. Don subsp. italicum. The research was conducted in Afyonkarahisar during the 2023 growing season. Inflorescences were harvested at seven different times between 06:00 and 18:00 at 2-hour intervals. EOs were extracted via hydrodistillation and analyzed using GC/FID-MS. Results showed that EO yield varied between 0.30% (06:00) and 0.47% (14:00). The highest number of volatile compounds (36) was observed at 08:00. Major constituents exhibited distinct diurnal accumulation patterns. γ-curcumene reached its maximum level at 06:00 (23.60%). Neryl acetate peaked at 14:00 (23.59%), while α-pinene reached its highest concentration at 18:00 (28.97%). Additionally, 3-decen-2-one reached 16.73% at 12:00. Multivariate analyses revealed descriptive intraday shifts between sesquiterpene- and monoterpene-rich profiles. These findings provide a scientific basis for defining optimal harvest windows to meet specific industrial and pharmacological standards.
The study provides the first baseline assessment of the mangrove ecosystem in Barangay Tomongtong, Enrique B. Magalona, Negros Occidental, Philippines, by evaluating vegetation structure, species diversity, regeneration status, aboveground biomass, and carbon stock. A total of 16 true mangrove species belonging to eight families were recorded, indicating moderate site-level species richness. Community structure analysis revealed strong dominance of Avicennia marina across all stations and zonation belts, with high importance values and dense mature stands. Despite excellent tree height conditions, crown cover and regeneration were consistently poor, suggesting limited recruitment and reduced structural complexity. Carbon stocks estimation varied substantially among stations with Station 3 (7 601 167.14 kg), followed by Station 2 (752 392.43 kg) and Station 1 (276 661.19 kg). Similarly, the middle zone stored the greatest amount of carbon (3 675 035.30 kg), followed by the seaward (1 085 827.69 kg) and landward (10 017.67 kg). The increasing dominance of stress-tolerant pioneer species suggests a simplification of the mangrove community, potentially reducing long-term ecological resilience under disturbances and climate stressors. These findings provide critical baseline data to support local government coastal and fisheries management intervention, including assisted natural regeneration, enrichment planting, hydrological restoration of abandoned fishponds, and mitigation of anthropogenic pressures to enhance biodiversity conservation, carbon sequestration, and fisheries sustainability.
Coastal communities in Indonesia face structural challenges such as economic vulnerability, environmental degradation, and limited market access due to less sustainable conventional agribusiness models. This research offers a novel blue economy-based agribusiness model that not only emphasizes product downstreaming but also integrates institutional strengthening, business diversification, adoption of environmentally friendly technologies, and SMART indicators and Key Performance Indicators (KPIs) as measurement instruments for success. The research was conducted using a mixed methods approach, including a quantitative survey of 200 coastal households in Sidoarjo and Banyuwangi, in-depth interviews, and focus group discussions (FGDs) with approximately 30 key informants. The results indicate heterogeneity between locations: Sidoarjo, dominated by intensive cultivation, faces input and environmental constraints, while Banyuwangi, with a more diverse economic structure, shows strong potential for diversification and ecotourism. The resulting model consists of input–process–output–outcome components with measurable KPIs in the short, medium, and long term, as well as an eight-step implementation strategy that is adaptive to the local context. This research contributes to national policies ranging from the 2021–2045 Blue Economy National Action Plan to the SDGs. The limitations of this research lie in the limited scope of the study area and the validation of the model which is still at the conceptual stage, so further research is needed to test implementation in other locations as well as the use of more varied indicators and quantitative data processing.
Dieffenbachia is widely used in indoor landscaping due to its variegated leaf structure and high adaptability to indoor conditions. This study aimed to evaluate the effects of combinations of cytokinins (BAP) and auxins (NAA and 2,4-D) on vegetative growth during the in vitro micropropagation of Dieffenbachia sp. Single-node shoot explants were transferred to MS medium. Basic morphological parameters such as shoot length, number of leaves, number of roots, and root length were evaluated in the study. Statistically significant differences were found among the treatments (p<0.05). The highest shoot growth was observed in the T1 (6.00 cm) treatment, whereas the T1 (5.00 per plantlet) and T3 (5.00 per plantlet) treatments exhibited a higher number of leaves. In terms of rooting performance, the highest number of roots (8.08 per plantlet) occurred in the T2 treatment, while the longest roots occurred in the T1 treatment (8.67 cm). No vegetative growth occurred in the treatments containing a high concentration (0.6 mg/L) 2,4D (T14, T16, T18, T20). This study demonstrates that low and moderate levels of cytokinin-auxin combinations enhance in vitro regeneration efficiency in Dieffenbachia exotica cv. Tropic Snow. in in vitro regeneration and provides an important basis for optimizing protocols the commercial micropropagation of this species.
In recent years, novel, effective, and environmentally sustainable approaches have been developed to promote plant growth and enhance plant protection. Plant extracts, which comprise novel, natural, and multicomponent products, have emerged as promising tools for these purposes. This study was designed to determine the effects of S. triloba extracts and biofertilizer applications on the germination, shoot, and root characteristics of fenugreek, as well as phytotoxicity tests. Ten traits were examined in the study, and except for the relative radicle growth (RRG) values, statistically significant differences were determined among the traits. The highest germination percentage, fresh and dry shoot values, and dry root weight with the relative seed germination (RSG) and germination index (GI) values were obtained from T8 (4 mg ml-1 extract + biofertilizer), while the maximum root length and fresh root weight values were found in T5 (0.5 mg ml-1 extract + biofertilizer). Moreover, the relationships among the examined traits were determined by multivariate analysis as principal coordinate (PCA) and heat map analyses. The results of the PCA showed over 75% variations, and heatmap analysis revealed that most of the treatments were found to be associated with the properties. In conclusion, T8 treatment can be used as a biological fertilizer to improve the germination percentage, fresh shoot weight, dry shoot and root weight, and RSG and GI values of the fenugreek.
This study compared insect diversity in Allium cepa var. aggregatum G. Don, Solanum lycopersicum L., and Capsicum annuum L. agroecosystems Parhaboan District, North Tapanuli, and Juhar District, Karo, North Sumatra, Indonesia, aiming to assess species composition, abundance, and ecological roles. Using pitfall, yellow sticky, and blue ball traps, hand picking, and direct observation, we identified pest and beneficial insects. In A. cepa var. aggregatum fields, Spodoptera litura (Fabricius) and Liriomyza chinensis (Kato) were dominant pests, with L. chinensis populations peaking during specific crop growth stages, highlighting the impact of crop phenology and pesticide use. S. lycopersicum plantations showed moderate diversity, dominated by Bactrocera spp. and Formicidae, with blue ball traps capturing high densities of Bactrocera spp., suggesting strong visual attraction. C. annuum crops had abundant natural enemies, including Paratrechina longicornis (Latreille) and Scelio spp., alongside pests like S. litura and Aphis spp., demonstrating a complex ecological balance. Despite moderate diversity, high evenness of natural enemies indicated a stable predator community. Intercropping C. annuum with A. cepa var. aggregatum or S. lycopersicum revealed varied insect compositions. A critical imbalance between herbivores and natural enemies was observed across all agroecosystems, emphasizing the need for integrated pest management (IPM). This research highlights the importance of insect diversity for sustainable agriculture in North Sumatra, advocating for habitat diversification, biological control, and strategic pesticide use. Future studies should examine seasonal dynamics, crop-specific variations, and long-term agricultural impacts to refine IPM and promote biodiversity conservation.
The increasing demand for raw materials in the paper and fermentation industries has intensified pressure on forest resources, highlighting the urgent need for sustainable alternatives. Hemp (Cannabis sativa L.), a fast-growing, non-woody fiber crop with additional medicinal and nutritional benefits, has emerged as a promising candidate. This two-year field study, conducted using a split-split plot design with three replications, evaluated the effects of four planting densities (3.6, 5.5, 9.5, and 20 plants m-2) and four nitrogen levels (0.150, 300, and 450 kg ha-1) on three hemp genotypes: Futura 75 (Germany), Isfahan, and Yazd, in the Alborz region of Iran. Significant effects were observed for all main factors and their interactions on key morphological and yield-related traits. The highest dry matter yield (22.559 kg ha-1) was obtained from the Isfahan genotype under 300 kg ha-1 nitrogen and a plant density of 20 plants m-2.The Futura 75 genotype recorded the greatest stem diameter (5.31 cm) and number of lateral stems (70 per plant) under low density (3.6 or 9.5 plants m-2) and high nitrogen input (450 kg ha-1). Leaf dry weight peaked at 103.66 g plant-1 in the Isfahan genotype with high nitrogen and low density. Yield components were generally higher in the first growing season, suggesting a notable influence of favorable environmental conditions. These findings emphasize the importance of optimizing planting density and nitrogen application according to genotype and environmental conditions. The Isfahan genotype, combined with high plant density and 300 kg ha-1nitrogen, is particularly recommended for enhancing biomass yield in semi-arid regions.
In this study, an effective regeneration and micropropagation protocol was developed from segregating the F4 industrial hemp genotype Narlısaray. The apical shoot meristem and nodal (from different parts of the plant) explants were cultured at various concentrations of TDZ, BAP or Kin. Shoot regeneration was observed in all treatments, including the control treatment. The maximum number of shoots per explant (4.7) was obtained at 1.0 mg l⁻¹ TDZ, and was statistically different from 0.5 mg l⁻¹ (4.3) and 2.0 mg l⁻¹ TDZ (4.0). The longest shoots (6.5 cm) were recorded on 1.0 mg l-1 TDZ. The position of the nodal explant source on the plant also affected its regeneration ability. The 1st stem node explant was more responsive to in vitro regeneration than the 2nd stem node. All tested concentrations of IBA induced root regeneration, with frequencies ranging from 62.5% to 87.5%. The regenerated plants were successfully acclimatized, reaching a maximum survival rate of 66.7%. This propagation technique can contribute to future hemp rapid breeding programs.
This study evaluated the effects of basal leaf removal at different phenological stages on yield, fruit quality, phytochemical composition, and winter bud physiology of the Karaerik grape cultivar under the Baran training system. Basal leaf removal was applied by removing five basal leaves at full flowering, pea-size berry, and veraison, and compared with an untreated control. The experiment was carried out over two consecutive growing seasons (2023 and 2024) using a randomized block design. The timing of leaf removal significantly influenced yield components and berry characteristics. Control vines produced the highest fresh grape yield (8.98 kg vine⁻¹), whereas vines defoliated at full flowering showed a moderate yield reduction (7.43 kg vine⁻¹). Despite this reduction, full flowering treatment substantially decreased cluster compactness (OIV-204: 4.58) compared with the control (7.58) and reduced the incidence of cracked and decayed berries by approximately 80%. Early defoliation also promoted berry growth, resulting in the highest single berry weight (5.91 g). Technological maturity parameters were positively affected by early leaf removal. The full flowering treatment resulted in the highest total soluble solids (19.07%), maturity index (27.42), and total phenolic content (504.57 mg L⁻¹). Moreover, leaf removal at full flowering and at pea-size berry significantly increased soluble carbohydrate concentrations in winter buds and reduced primary bud cold injury and necrosis. Overall, basal leaf removal at full flowering emerged as an effective vineyard management strategy to improve fruit quality, enhance phytochemical accumulation, and increase winter bud cold tolerance in Karaerik grapes cultivated under the Baran training system.
The present research aims to assess the biochemical characteristics and environmental stress adaptation of the Bacillus mojavensis AOB29 strain, along with its influence on the growth and nutrient uptake of arugula (Eruca vesicaria). The B. mojavensis AOB29 strain used in this study was isolated from soil samples collected from an agricultural field. The morphological, physiological, metabolic, and environmental characteristics of the isolate were evaluated through Gram staining, spore staining, motility assessment, various enzyme activity tests, BIOLOG GEN III analysis, antibiotic susceptibility testing, and tolerance assays under different salt, heavy metal, and pH conditions. In addition, its effect on plant growth was investigated through pot experiments. Pot experiments demonstrated that the application of B. mojavensis AOB29 resulted in significant improvements in plant growth compared to the control group. In particular, the treatment led to a 57.8% increase in dry biomass and a 40.9% increase in fresh biomass relative to the control. Nutrient analyses revealed increases in potassium, magnesium, phosphorus, zinc, and manganese levels, while decreases in calcium, iron, and copper levels were detected. In conclusion, the B. mojavensis AOB29 strain was found to possess the potential to promote plant growth, enhance nutrient uptake, and improve tolerance to environmental stress conditions. These findings indicate that this strain may be considered for use as a biofertilizer or biocontrol agent in agricultural applications.
In this study, phenolic compounds of methanol extracts obtained from stems (fruit pedicels), leaves, and flowers of two different varieties (0900 Ziraat and Regina) of cherry (Prunus avium L.) in Çankırı (Eldivan) region were investigated. Component determination was performed using an LC-MS/MS instrument. The data obtained revealed that different plant parts of both varieties differed in phenolic content. The major components identified were quinic acid, epicatechin, catechin, chlorogenic acid, and rutin. Accordingly, the major components were found to be higher in 0900 Ziraat extracts rather than Regina extracts. The highest values of the major components, catechin (44222 ± 480.74 µg g-1, 41288 ± 176.38 µg g-1) and epicatechin (9577.78 ± 38.49 µg g-1, 9377 ± 101.83 µg g-1) were determined in the stems of 0900 Ziraat and Regina varieties, respectively. The highest concentrations of rutin, chlorogenic acid, and quinic acid were obtained from the leaf extracts of both varieties. The highest phenolic content was observed in stem extracts of both varieties, followed by leaf and flower extracts. This result indicates that stem tissue has a more active metabolic profile for phenolic accumulation. These findings suggest that the stem (fruit pedicel), leaf, and flower tissues of cherry may be relatively rich in phenolic compounds and provide a baseline dataset supporting their potential evaluation as natural sources of antioxidants.
Climate change is listed as the biggest disaster threatening the world in today's conditions. The effects of climate change are felt in all sectors, all living beings, and every point of the universe. However, the greatest harm of these effects is more intense on freshwater resources, which are the source of life for living beings and an indispensable element for ecosystems. Global warming, which causes climate change, reveals the results of the decrease and pollution of freshwater resources. Therefore, the effects of climate change on freshwater resources should be clearly known, and effective water use policies should be put forward to reduce these effects in the climate crisis. Thus, this study, which examines the effects of climate change on freshwater resources through the World assessment on Türkiye scale, aims to bring the ways of effective water use in the climate crisis to the literature in the light of various solution suggestions and action plans.
Tomato mosaic virus (ToMV) is one of the most important viral pathogens affecting tomato production worldwide. This study investigated the occurrence, molecular characterization, and phylogenetic relationships of ToMV isolates infecting tomato plants in the Samarkand region of Uzbekistan. Field surveys conducted in May 2024 identified typical virus-like symptoms in tomato plants, including mosaic patterns, chlorosis, leaf deformation, and fruit malformations. Sixty symptomatic samples were collected and analyzed using reverse transcription–polymerase chain reaction (RT-PCR) with universal tobamovirus primers and ToMV-specific primers targeting the coat protein (CP) gene. RT-PCR screening revealed that 42 samples (70%) were infected with tobamoviruses, of which 29 (48.3% of all tested plants) were specifically positive for ToMV. Two representative isolates, Samarkand-1 and Samarkand-2, were sequenced and deposited in GenBank under accession numbers OR400153 and OR400155. Phylogenetic analysis based on CP gene sequences demonstrated that the two Uzbek isolates belong to distinct evolutionary lineages: Samarkand-2 clustered with East Asian isolates, whereas Samarkand-1 grouped with European and Middle Eastern isolates, suggesting independent introduction events likely associated with international seed trade. Amino acid comparison of the CP revealed two substitutions in the Samarkand-1 isolate, while Samarkand-2 was identical to the reference isolate DSMZ PV-0143 from the USA. The results indicate considerable genetic diversity of ToMV circulating in Uzbekistan and highlight the importance of molecular surveillance, the use of certified virus-free seeds, and strict phytosanitary measures to prevent the further spread of diverse ToMV lineages in Uzbekistan's tomato production systems.
Lima bean (Phaseolus lunatus L.) remains an underutilized legume in Nigeria, partly due to limited agronomic recommendations for optimizing its productivity. The present study investigated the effects of staking methods and plant population density on some agronomic properties of two accessions of Lima bean. The experiment was conducted out at Institute of Agricultural Research and Training research station situated at Ikenne, Ogun state in 2023 and 2024. The treatment combinations consisted of two lima bean accessions (2005-012A and 2006-005A) planted at three population densities: 17777, 13333 and 10000 plant ha⁻¹ respectively and two staking alternatives (trellis staking and vertical staking) with no stakes as control. The experimental design was a split-split plot in a randomized complete block with accession, staking methods and population densities assigned as the main plot, sub-plot and sub-sub plot respectively. Data were analysed using ANOVA with mean separated at α=0.05. Vertical staking consistently enhanced vine length, leaf production, biomass accumulation, pod parameters, and grain yield relative to trellis and no staking treatments. Grain yield ranged from 253.1 (10000 plants ha⁻¹) to 347.7 kg ha⁻¹ (13333 plants ha⁻¹). Accession 2006-005A outperformed 2005-012A in grain yield by 25%. No stakes resulted in delayed flowering and the lowest yield (225.14 kg ha⁻¹) compared to vertical (361.94 kg ha⁻¹) and trellis staking (313.08 kg ha⁻¹). The study recommends vertical staking combined with a plant population density of 13333 plants ha⁻¹ for improved lima bean productivity in the study area.
Trogoderma granarium Everts (Coleoptera: Dermestidae) is recognized as one of the most damaging quarantine insects on stored grains. However, with the increasing international trade of gluten-free products, it is important to increase our current level of knowledge to better understand their potential susceptibility to infestation by this quarantine insects. For this, the study examined the development potential of T. granarium on three internationally traded gluten-free products, quinoa, chia, and linseed, when subjected to whole and flour forms of these products. The adults of T. granarium were allowed to grow on whole and flour forms of each product, and the progeny development were recorded after 65 days. Results on population growth, frass production and commodity weight loss, differed significantly on different products as well as their forms. Progeny development, egg production, and pupal weight of T. granarium as well as frass production and commodity weight loss, were found to be significantly higher in quinoa, especially in flour, compared to other products. Chia seeds, both in whole and flour form, recorded the lowest results on above parameters. Linseed was found to be support highest insect pupal weight as well as frass production. These results demonstrate that both the commodity itself and physical form strongly affect T. granarium population dynamics and damage. The findings point out quinoa flour as a high-risk commodity for infestation and highlight the need to incorporate emerging gluten-free commodities into postharvest risk assessment programs.
The rising demand for reliable plant tissue cultures in agriculture underscores the critical need for early detection of microbial contamination in culture vessels. Conventional detection methods often lack sensitivity in identifying contamination at initial stages, which can result in compromised cultures and substantial economic losses. This research focuses on developing an automated detection system for microbial contamination in plant tissue cultures. The study involved the design and implementation of a macroscopic image acquisition prototype of culture vessel contamination of Oryza sativa. The acquired images underwent preprocessing, including normalization and enhancement techniques, followed by feature engineering for effective data extraction of color and textural features. The extracted features were subsequently analyzed using supervised and unsupervised machine learning models. Supervised machine learning models such as support vector machine, K-nearest neighbors and Discriminant analysis showed high accuracy with Discriminant analysis scoring 99.6%, which is the highest. K-means clustering accurately identified the grouping by distinguishing contaminated from non-contaminated samples, while the Isolation forest indicated the presence of microbial contamination. Overall, the approach presented in this research proves the capability of detecting microbial contamination with macroscopic image processing and machine learning.
