In the last decade, small unmanned aerial vehicles (UAVs/drones) have become increasingly popular in the airborne observation of large areas for many purposes, such as the monitoring of agricultural areas, the tracking of wild animals in their natural habitats, and the counting of livestock. Coupled with deep learning, they allow for automatic image processing and recognition. The aim of this work was to detect and count the deer population in northwestern Serbia from such images using deep neural networks, a tedious process that otherwise requires a lot of time and effort. In this paper, we present and compare the performance of several state-of-the-art network architectures, trained on a manually annotated set of images, and use it to predict the presence of objects in the rest of the dataset. We implemented three versions of the You Only Look Once (YOLO) architecture and a Single Shot Multibox Detector (SSD) to detect deer in a dense forest environment and measured their performance based on mean average precision (mAP), precision, recall, and F1 score. Moreover, we also evaluated the models based on their real-time performance. The results showed that the selected models were able to detect deer with a mean average precision of up to 70.45% and a confidence score of up to a 99%. The highest precision was achieved by the fourth version of YOLO with 86%, as well as the highest recall value of 75%. Its compressed version achieved slightly lower results, with 83% mAP in its best case, but it demonstrated four times better real-time performance. The counting function was applied on the best-performing models, providing us with the exact distribution of deer over all images. Yolov4 obtained an error of 8.3% in counting, while Yolov4-tiny mistook 12 deer, which accounted for an error of 7.1%.

This paper presents the standardized precipitation evapotranspiration index (SPEI)-based approach to agricultural drought monitoring (ADM-SPEI approach) combining well-known methods, expert’ critical opinions, and local agro-climatic specificities. The proposed approach has been described in detail in three phases. This allows its application in any region and modification according to different agro-climatic conditions. The application of the ADM-SPEI approach has resulted in obtaining a modified SPEI for different crops (agricultural drought SPEI (AD-SPEI_crop)) in the Vojvodina region. In the first phase of the proposed approach, analytical hierarchy process (AHP) was used to obtain an experts’ group decision regarding the most suitable method for calculating evapotranspiration for a particular analyzed region. In the second phase, SPEI was modified and adjusted to the conditions in Vojvodina, where ET₀ was replaced by ET_c. In the validation phase, the results of the application of AD-SPEI_crop were compared to crop yields and well-known indices and evaluated by the experts’ feedback. The statistically significant correlations were achieved between AD-SPEI_crop and crop yields. The highest correlations were achieved in the months when the analyzed crops were in the developmental stages when they are most sensitive to drought. The AD-SPEI_crop better correlates to the crop yields compared to SPEI. The comparison of AD-SPEI_crop to the standardized precipitation index (SPI), SPEI, and self-calibrated Palmer drought severity index (SC-PDSI) shows that it can successfully detect dry and wet periods. The results have indicated that the proposed approach can be successfully applied, and AD-SPEI_crop has shown a good performance for agricultural drought monitoring.

Conserving clean and safe freshwater is a global challenge, with nitrogen (N) and phosphorus (P) as frequent limiting factors affecting water quality due to eutrophication. This paper provides a critical overview of the spatiotemporal variability in both nutrient concentrations and their total mass ratio (TN:TP) in the canal network of the Hydro system Danube–Tisza–Danube at 21 measuring locations monitored by the Environmental Protection Agency of the Republic of Serbia over a length of almost 1000 km, collected once a month during the last decade. A spatiotemporal variation in nutrient concentrations in the tested surface water samples was confirmed by correlations and cluster analyses. The highest TN concentrations were found in winter and early spring (non-vegetation season), and the highest TP concentrations in the middle of the year (vegetation season). The TN:TP mass ratio as an indicator of the eutrophication pointed out N and P co-limitation (TN:TP 8–24) in 64% of samples, N limitation (TN:TP < 8) was detected in 27% and P limitation (TN:TP > 24) in the remaining 9% of water samples. Such observations indicate slow-flowing, lowland water courses exposed to the effects of non-point and point contamination sources as nutrient runoff from the surrounding farmlands and/or urban and industrial zones, but further investigation is needed for clarification. These results are an important starting point for reducing N and P runoff loads and controlling source pollution to improve water quality and underpin recovery from eutrophication in the studied watershed.

The aim of this study was to investigate the efficacy and phytotoxicity of herbicides in combinations and to determine their economic feasibility in the soybean crop. The trials were placed at two locations, Pobeda and Budisava (Serbia), during 2016 and 2018, organized in a randomized block design with four replicates of all herbicide combinations (metribuzin + S-metolachlor; imazamox + oxasulfuron + typhenesulfuron-methyl; metribuzin + S-metolachlor + imazamox + oxasulfuron + typhenesulfuron-methyl; and bentazon + imazamox + typhenesulfuron-methyl). At the first location, 16 weed species were determined, while in the second location, 14 were determined. The highest reduction in the number of weeds was achieved at the first location, using a combination of herbicides metribuzin + S-metolachlor + imazamox + oxasulfuron + typhenesulfuron-methyl with a total efficacy that ranged from 96.98% to 97.40%. Only on the second location bentazone + imazamox + tifensulfuron-methyl showed passable phytotoxicity on soybean, during both years. Based on the established economic justification, in the combinations of metribuzin + S-metolachlor + imazamox + oxasulfuron + typhenesulfuron-methyl, yield was increased for 2350 kg/ha or 1.91 times more, according to the untreated variant at first location. The economic injury level at the most effective combination of herbicides was 52.70 weeds/m².

Soil organic carbon (SOC) has a high impact on the sustainability of ecosystems, global environmental processes, soil quality and agriculture. Long-term tillage usually leads to SOC depletion. The purpose of this study was to determine the impact of different land uses on water extractable organic carbon (WEOC) fractions and to evaluate the interaction between the WEOC fractions and other soil properties. Using an extraction procedure at 20°C and 80°C, two fractions were obtained: a cold water extractable organic carbon (CWEOC) and a hot water extractable organic carbon (HWEOC). The results suggest that there is a significant impact from different land uses on WEOC. A lower relative contribution of WEOC in SOC and a lower concentration of labile WEOC fractions are contained in arable soil compared to forestlands. Chernozem soil was characterized by a lower relative contribution of WEOC to the SOC and thus higher SOC stability in contrast to Solonetz and Vertisol soils. Both CWEOC and HWEOC are highly associated with SOC in the silt and clay fraction (<53 µm) and were slightly associated with SOC in the macroaggregate classes. The WEOC fractions were highly and positively correlated with the SOC and mean weight diameter.

The objective of this study is to assess the possibility of using unmanned aerial vehicle (UAV) multispectral imagery for rapid monitoring, water stress detection and yield prediction under different sowing periods and irrigation treatments of common bean (Phaseolus vulgaris, L). The study used a two-factorial split-plot design, divided into subplots. There were three sowing periods (plots; I—mid April, II—end of May/beginning of June, III—third decade of June/beginning of July) and three levels of irrigation (subplots; full irrigation (F)—providing 100% of crop evapotranspiration (ETc), deficit irrigation (R)—providing 80% of ETc, and deficit irrigation (S) providing—60% of ETc). Canopy cover (CC), leaf area index (LAI), transpiration (T) and soil moisture (Sm) were monitored in all treatments during the growth period. A multispectral camera was mounted on a drone on seven occasions during two years of research which provided raw multispectral images. The NDVI (Normalized Difference Vegetation Index), MCARI1 (Modified Chlorophyll Absorption in Reflectance Index), NDRE (Normalized Difference Red Edge), GNDVI (Green Normalized Difference Vegetation Index) and Optimized Soil Adjusted Vegetation Index (OSAVI) were computed from the images. The results indicated that NDVI, MCARI1 and GNDVI derived from the UAV are sensitive to water stress in S treatments, while mild water stress among the R treatments could not be detected. The NDVI and MCARI1 of the II-S treatment predicted yields better (r² = 0.65, y = 4.01 tha⁻¹; r² = 0.70, y = 4.28 tha⁻¹) than of III-S (r² = 0.012, y = 3.54 tha⁻¹; r² = 0.020, y = 3.7 tha⁻¹). The use of NDVI and MCARI will be able to predict common bean yields under deficit irrigation conditions. However, remote sensing methods did not reveal pest invasion, so good yield predictions require observations in the field. Generally, a low-flying UAV proved to be useful for monitoring crop status and predicting yield and water stress in different irrigation regimes and sowing period.