Applying Machine Learning Models to Predict Population Dynamics of Harmful Non-Gregarious Locusts Across Kazakhstan Agroclimatic Zones

INTRODUCTION

The digital transformation of the agro-industrial complex (AIC) is currently considered a key factor in the sustainable and innovative development of agriculture. The integration of information and communications technologies into the agricultural production processes contributes to labor productivity and the efficiency of resource use and provides for an export-oriented model by increasing the volume of processed agricultural products (Pakhomova et al., 2025).

Today, locusts hold their position as troublesome polyphagous plant pests worldwide. Kazakhstan, in particular, is home to about 300 species and subspecies of locusts. Of these, about 15-20 species, both gregarious and non-gregarious, are the most harmful and dangerous for crops and lands (Kambulin, 2000; Childebaev et al., 2004b; Kambulin et al., 2010). It causes an increasing need for preventive control and advanced forecasting of the population of these pests to objectively plan the scope of protective measures (Yskak et al., 2002; Azhbenov et al., 2015; Yessenbayeva et al., 2025). Our earlier studies (Baibussenov et al., 2021, 2022a, 2022b) provide data on innovative approaches in the phytosanitary monitoring of the complex of harmful non-gregarious locusts, namely, GIS-based modeling of ecological niches or favorable habitats. This technology allows identifying the places with the highest risk of population by locusts, which need to be prioritized in surveys (Al-Dosari et al., 2024; Csep et al., 2024; Homayoun et al., 2024; Khan et al., 2024; Braun et al., 2025; Flores et al., 2025; Nakamura et al., 2025; Pacheco et al., 2025; Park et al., 2025; Wilson et al., 2025).

There is a satisfactory body of research on the improvement of the phytosanitary monitoring of especially dangerous gregarious locusts, such as Calliptamus itallicus L., Dociostaurus maroccanus Thunb., Locusta migratoria L., and Schistocerca gregaria Fors. (Iranipour et al., 2017; Lachininsky, 2020; Githae & Kuria, 2021; Maeno et al., 2021; Sergeev et al., 2022; Lachininsky et al., 2023; Azhbenov et al., 2024). These studies were used to create large databases on the distribution and migration of these species, as well as models of their ecological niches using remote sensing of the Earth and GIS technologies (Liao et al., 2013; Klein et al., 2022; Wu et al., 2022).

Predicting the mass reproduction and migration directions of gregarious locust species is the goal of several studies (Gómez et al., 2018; Humphreys et al., 2022; Lawton et al., 2022; Sun et al., 2022). Using indices of vegetation, soil moisture, rainfall, and land surface temperature, African researchers employed generalized linear models to examine the intervals between observations of desert locust presence and absence during remission periods in West Africa. According to the study, when the soil was subsequently moistened, the gap grew by three weeks in comparison to vegetation (Sun et al., 2022).

Using a multistage dynamic population model, researchers from Australia (Humphreys et al., 2022; Lawton et al., 2022) uncovered environmental processes explaining large-scale changes in the distribution of Australian desert locusts, accounting for errors in phase determination during field studies. They found that variations in the probability of accumulation are mainly explained by rainfall and vegetation cover, with the response to rainfall in desert habitats being faster and more robust than in agricultural landscapes. This structure is promising for predicting locust outbreaks because it bridges the gap between process-based models, which often lack empirical support, and data-based models, which tend to overlook causation in the migration mechanisms of gregarious locust species.

It is urgently necessary to create contemporary digital methods for phytosanitary monitoring and forecasting that target dangerous non-gregarious locust species as well as gregarious ones. Using automated population forecasting as a foundation for determining the necessary volume of protective measures, a particularly urgent goal is to develop technology to identify population expansion early and establish appropriate preventive plant protection measures.

Today, global science is heavily invested in introducing artificial intelligence, machine learning, and other digital technologies into various fields, including the AIC. These technologies could be very useful for plant protection in research aimed at developing models for the phytosanitary forecasting of plant pest populations. In our studies, this experience was applied to the complex of harmful non-gregarious locusts as the research subject.

The purpose of the research was to identify patterns and underlying predictors in a specific agroclimatic zone of Kazakhstan to forecast the pest population based on machine learning algorithms and GIS technologies.

MATERIALS AND METHODS

Long-term historical data on the distribution and population of non-gregarious locust pests were collected and processed using the methods of statistical analysis adopted in phytosanitary monitoring, forecasting, and signaling of the emergence and development of plant pests (Suleimenov et al., 2009; Dubrovin et al., 2011; Azhbenov, 2013;  Sagitov et al., 2016).

The study utilized the resources of the Center for Technological Competence in the Field of Digitalization of the Agro-Industrial Complex under S. Seifullin Kazakh Agro-Technical Research University, specifically the laboratories of Big Data Analysis, Artificial Intelligence, and GIS Technologies in Agriculture.

1. Study location. Two agroclimatic zones of Kazakhstan were chosen as the location of the study site for subsequent comparison. The research and analysis were conducted within the following agroclimatic zones:

• The highly humid, moderately warm zone spans across the entire territory of the North Kazakhstan region, the northern part of the Kostanay region, the northern and central part of the Akmola region, and the extreme northern part of the Pavlodar region.
• The slightly humid, moderately warm zone, which occupies the territory around Karaganda city and nearby areas in the direction from north to southeast of the Karaganda region.

2. Data collection and aggregation. In order to determine the long-term patterns of the population of harmful non-gregarious locusts, we gathered, examined, and processed long-term data on their distribution and population in the regions of Kazakhstan by agroclimatic zones because the necessary patterns and causes of fluctuations in phytophage population dynamics cannot be established over just two to three years of research. This involved forecasting and signaling the emergence of pests in Kazakhstan, as well as the examination of past data from governmental plant protection services.

The density of non-gregarious locusts in various biotopes and the daily and seasonal dynamics of their number was assessed using a set of accounting methods (Dubrovin et al., 2011; Sagitov et al., 2016):

1 — 100 single (or 50 double) sweeps of the entomological net repeated three to four times. Species abundance was determined using the following scale:

m — the species is abundant (mass), sometimes forming clusters (100 specimens per hour of collection);

f — the species occurs frequently but without forming clusters (21-100 spec./hour);

r — the species occurs rarely (3-20 spec./hour);

un. — only unique individuals of the species occur (1-3 spec./hour).

The dominance index was determined as the ratio of the number of individuals of the species to the total number of all individuals in the collection:

where:

D — domination, %;

K — total number of specimens of all species in all samples, spec.;

k — total number of individuals of the specific species, spec.

According to Pravdin (1978), species that make up over 16% of the total number of collected species are considered dominant forms, and those accounting for 4-16% are categorized as subdominant. In all cases, mass species are always dominant.

3. Building and training the forecasting model. To predict the population of harmful non-gregarious locusts using neural networks and machine learning, we employed a comprehensive approach involving several stages.

First, long-term historical data on the population of locust pests in various agroclimatic zones of Kazakhstan were collected and processed. These data cover information on climate conditions, crops, and pest presence. Thus, the first stage provides the foundation for the forecasting model.

The indicators of changes in the number of individuals of insects over time include the mass appearance of harmful species in specific years. The phases of pest population dynamics include depression, population growth, mass reproduction, peak number, and decline in number (Baibusenov, 2021). To determine the phase of non-gregarious locust pests in the analyzed years, the main indicators of crop pest population—relative (P_rel), average (P_av), and absolute (P_abs) population—were calculated by the following formula:

where

P_rel — relative population, %

A_p — populated area, thousand ha

A_s — surveyed area, thousand ha

where

P_av — average population, pcs./m²

A_p (0-5),..(>30) — populated area (ths. ha) with the given population density (pcs./m²)

N_av (0-5)...030; — average population density, pcs./m²

A_p — populated area, thousand ha

where

P_abs — absolute population, pcs./m²

P_rel — relative population, %

P_av — average population, pcs./m²

The absolute population indicator (P_abs), which best reflects changes in pest abundance, was used as the target variable for model training. The dataset was cleaned, missing values were imputed, categorical variables were encoded, and exploratory data analysis was performed using Python tools (Anand, 2024; Kapoor et al., 2025). Potential climatic and environmental predictors were then screened through correlation analysis and outlier assessment, and the final set of predictors included in the forecasting model is presented in Table 1.

Table 1. Key parameters and predictors used to model the forecast of non-gregarious harmful locust population dynamics

Model development involved splitting the dataset into training and test sets and evaluating four forecasting approaches: multilinear regression, random forest, gradient boosting SVR, and SARIMA (Durgabai et al., 2018; Xiao et al., 2019; de Oliveira Aparecido et al., 2020; Ibrahim et al., 2022; Da Hye et al., 2024; Lima et al., 2024; Lombardi et al., 2024; Verma et al., 2024; Aoyama et al., 2025; Damaševičius et al., 2025; Møller et al., 2025; Papadopoulos et al., 2025; Shafique et al., 2025; Sørensen et al., 2025; Svendsen et al., 2025). The models were implemented in Google Colab using Python-based tools (Shaik et al., 2024; Swarnkar et al., 2024), and their performance was assessed using mean squared error (MSE) and the coefficient of determination (R²). To reduce overfitting, multilinear regression included 3-4 key predictors per zone, random forest was limited to 100 trees with a maximum depth of 3, and gradient boosting used 100 trees, a maximum depth of 3, and a learning rate of 0.1. SARIMA models were fitted to annual population series without an explicit seasonal component; based on AIC, the selected specifications were ARIMA(0,1,0) for zone 14 and ARIMA(0,0,2) for zone 15. Together, these procedures formed the methodological basis for forecasting harmful non-gregarious locust populations in Kazakhstan.

RESULTS AND DISCUSSION

Ecological features of harmful non-gregarious locusts in the studied agroclimatic zones

Phytosanitary services usually monitor gregarious and non-gregarious species separately. Gregarious species of locusts (e.g., in Kazakhstan—Calliptamus itallicus L., Dociostaurus maroccanus Thunb., and Locusta migratoria L.) in years of their mass reproduction usually form large swarms consisting of only one biological species and move in them. However, these swarms may not form uniformly across the entire territory of the country, even in the years of population growth and mass reproduction.

For instance, starting in 2022, an increase in Calliptamus itallicus L. populations was noted in the Aktobe area (Ayteke Bi and Khromtau districts) and districts bordering the Kostanay region with the Aktobe region (Zhetikara and Kamysty districts). The species' population reached its peak in 2024, after which it started to progressively decrease. In the northern and central regions of Kazakhstan, which were selected for our study, Calliptamus itallicus L. was in the solitary or transitional phase and was primarily observed in a mixed population with harmful non-gregarious locust species, whereas in these regions it saw population growth and transition into the gregarious phase. As a result, this species and non-gregarious locusts were taken into account in our observations.

In 2023-2024, route surveys were conducted in the chosen districts within the studied agroclimatic zones to establish the species composition, occurrence, and interspecies domination.

Namely, in the highly humid, moderately warm agroclimatic zone, surveys were conducted in the Fyodorov and Karabalyk districts of the Kostanay region; the Timiryazev, Akkayin, and Ualikhanov districts of the North Kazakhstan region; and the Burabay, Sandyktau, and Akkol districts of the Akmola region. In the slightly humid, moderately warm agroclimatic zone, our study covered the Arshaly district of the Akmola region and the Bukhar-Zhyrau and Karkaraly districts of the Karaganda region. These areas were chosen as the favorable habitats of the target pests. The results of these surveys were used to clarify and establish the set of harmful non-gregarious locust species, their frequency of occurrence, and dominance in the studied agroclimatic zones of Kazakhstan.

In this process, the species composition of harmful locusts in the studied agroclimatic zones of Kazakhstan was clarified. In the studied areas, among the locusts most harmful and economically significant to the AIC, we identified the main 15 harmful species. All these species were found in a community or a complex of species. Further research as part of this focused exclusively on this set of species.

Calliptamus italicus was common in both of the agroclimatic zones investigated. This gregarious species was found everywhere, primarily on fallow and perennial wormwood-cereal plant associations, and it appeared largely in the solitary phase in a mixed population with dangerous non-gregarious locust species. Paracryptera microptera microptera is a common species in the humid, somewhat warm agroclimatic zone that covers the majority of North Kazakhstan. This species was found in tiny quantities in a slightly humid, fairly warm agroclimatic zone. These locusts like needle grass-cereal and needle grass-Volga fescue-cereal plant communities.

Dociostaurus brevicollis and Dociostaurus kraussi occurred frequently in the highly humid, moderately warm agroclimatic zone and moderately often in the slightly humid, moderately warm zone. Dociostaurus brevicollis was found primarily in cereal-wormwood and Volga fescue-wild rye plant associations, while Dociostaurus kraussi appeared mainly in the Volga fescue-wild rye association. Stenobothrus fischeri inhabited the Volga fescue-wormwood association. This species occurred moderately often in the highly humid, moderately warm agroclimatic zone and frequently in the slightly humid, moderately warm zone. Oedipoda miniata was observed in the slightly humid, moderately warm agroclimatic zone, predominantly in needle grass-cereal and cereal-wormwood plant associations, whereas Oedipoda caerulescens was more common in the highly humid, moderately warm agroclimatic zone, mostly in the cereal-wormwood plant association. The remaining species cited are also economically important but occur less frequently in the considered agroclimatic zones. However, in combination with other frequently encountered species, their harmfulness increases.

In addition, we calculated the percentage of dominance of the considered mass harmful locust species within the studied agroclimatic zones (Figures 1 and 2).

According to these data, within the highly humid, moderately warm agroclimatic zone of Kazakhstan, the more dominant species (Figure 1) are Calliptamus italicus, Paracryptera microptera microptera, and Dociostaurus brevicollis (21:12.8:12.8%, respectively). In the slightly humid, moderately warm agroclimatic zone, the most dominant harmful species (Figure 2) are Calliptamus italicus, Oedaleus decorus, and Stenobothrus fischeri (20.4:19.7:19%, respectively).

The remaining species are not dominant in the studied zones, but their combination with the dominant species increases the damage caused by the overall complex of harmful locusts in the area.

These results are crucial for researchers to fully understand the phytosanitary situation with the occurrence and dominance of a set of harmful locusts in the considered agroclimatic zones of Kazakhstan. However, in the practice of plant protection, it is inexpedient to keep separate records for each species, and population and harmfulness are determined for the entire complex of non-gregarious locust pests. The economic threshold of harmfulness is also set based on the entire complex of non-gregarious locust species.

Consequently, all applied work on phytosanitary control and population forecasting was carried out with respect to the complex of harmful species by the studied agroclimatic zones.

Machine learning-based forecasting models for the population of harmful non-gregarious locusts in the studied agroclimatic zones

The issues of forecasting and preventing the outbreaks of mass reproduction of harmful locusts remain pressing in plant protection to this day due to considerable losses of agricultural products.

In any field of science, the forecasting of a phenomenon is founded on the study of repeated events and the main factors affecting the course of events. If a pattern is detected, changes in the dynamics of any future event can be predicted based on the suppression or activation of concomitant factors. Today, these processes can be studied automatically using machine learning or neural networks.

To identify the fundamental predictors of the population of harmful non-gregarious locusts in the studied agroclimatic zones based on machine learning algorithms, we processed and analyzed long-term historical data for the period of 2003-2023, describing the population dynamics of the studied pests depending on the influence of weather-climatic and edaphic (soil) factors.

At the stage of preliminary data processing, the input data were adapted for machine learning models. Categorical variables, such as region and district codes, were converted to a numeric format. The missing values discovered, in particular, in temperatures, solar radiation, and soil humidity, were filled with median values, thus avoiding distortion due to outliers.

The set of standard models tested for the forecasting of pest population included multilinear regression, random forest, gradient boosting SVR, and SARIMA.

In the process of training the forecasting models, together with the analysis of the population of the studied pests, the accompanying weather and climatic environmental factors for the analyzed years were also considered.

Data for the period from 2003 to 2017 were used to train the models, and the 2018-2023 data were left to test the forecast (validation on unseen data). The quality of the models was assessed by mean squared error (MSE) and the coefficient of determination (R²).

The correlation heatmap illustrates the structure of relationships between the variables. As expected, climatic indicators have strong internal links: evaporation correlates positively with air temperatures and "surface temperatures" and negatively with precipitation (forming blocks on the graph with coefficients | r | > 0.7 between, for example, Temp, Skintemp, and Evap). Soil water volume (VSWL) is associated negatively with temperature and evaporation (stronger in the summer months) and positively with precipitation, reflecting moisture balance. Thus, many climatic parameters are multicollinear, which needs to be considered when interpreting the impact of individual factors.

Importantly, the linear correlation of population with each of the climatic parameters was very weak (| r | < 0.1-0.17). In no agroclimatic zone does any factor show a significant pairwise correlation with the "number" indicator. The highest correlation coefficient is observed in the temperature of late summer/early autumn: for example, in the slightly humid, moderately warm agroclimatic zone, r ≈ 0.17 is between "skintemp9" (surface temperature in September) and population. Soil moisture in the spring has a small negative association (r ≈ -0.14 for VSWL4). However, these relationships do not reach statistical significance at the level of individual correlations. These results imply that the pest population is determined by a combination of factors and nonlinear effects undetectable by simple correlation analysis. This proves the need to use multifactor models to catch hidden dependencies.

To identify key climate predictors affecting pest populations, random forest models (at the level of individual districts for all years) were trained, and importance values were calculated for each variable. Figures 3a and 3b show the top 10 most significant features by importance in random forest models in the considered agroclimatic zones.

The heatmap shows that the most influential factor in the slightly humid, moderately warm agroclimatic zone is soil moisture reserve in spring (VSWL4), whose importance reaches ~0.11, noticeably ahead of others. The top ten also includes soil moisture in mid-summer (VSWL7), air temperature in June (Temp6), precipitation in June (Prec6), soil moisture in May (VSWL5), and solar radiation in June (SNSR6). These factors suggest that for the slightly humid, moderately warm agroclimatic zone (more arid), soil moisture at the beginning of the season and conditions in the first half of summer are critical for the development of the pest. Soil water volume in April (VSWL4) leads in importance. Other significant features are soil moisture in July (VSWL7), temperature and precipitation in June (Temp6, Prec6), May moisture (VSWL5), and June solar radiation (SNSR6). These results emphasize the role of the water regime in the arid zone 14 for the development of the studied pests.

The scenario is a little different in the agroclimatic zone that is fairly warm and extremely humid. With a high proportion of ~0.18, precipitation in July (Prec7) is the most significant feature. Soil moisture in August (VSWL8) and July (VSWL7) and air temperature in July (Temp7) and September (Temp9) are next. Therefore, midsummer conditions—including precipitation and soil moisture in July and August as well as temperature during this time—have a substantial impact on population variation in the extremely humid, moderately warm agroclimatic zone. One likely explanation is that excessive humidity and frequent rainfall can either encourage the growth of pests and fodder plants or, on the other hand, repress certain pests or spread illnesses. Because of this, this feature plays a major role in the model (in this case, the years with greater Prec7 saw a bigger number of the examined pests, according to the model's signals of influence).

Next, to estimate future risks of pest populations, the forecast was extrapolated using the chosen gradient boosting model to the period up to 2030. When constructing the long-term forecast, the climatic parameters of recent years were considered. It was assumed that the average weather conditions would remain within the observed ranges, without extreme anomalies. This allows us to obtain an approximate dynamic of the number of pests in each region, subject to the gradual continuation of current trends. The figures below show the forecast curves for the slightly and highly humid, moderately warm agroclimatic zones, together with actual data until 2023 to illustrate the trends.

The gradient boosting model indicates that after the 2019-2020 decline, the population of pests in the highly humid, moderately warm agroclimatic zone will fluctuate around the average level of ~1.3-1.5 c.u., with exceptions in individual years. The forecasted population dynamics for the highly humid, moderately warm agroclimatic zone for 2024-2030 are presented in Figure 4. The forecast shows a moderate increase in population by the middle of the decade: a maximum is expected around 2026, when the average number can reach ~1.7-1.8, approaching the values of 2018. Further on, a temporary decrease is likely: minimums are predicted at the end of the decade (around 2029, the number may decrease to ~1.2). Nevertheless, even in the years of recession, the model does not imply such a sharp drop as was observed in 2020. No significant monotonous trend is expected for zone 15. According to the forecast, the number of pests will remain on a relative plateau with some fluctuations, which indicates the continuing threat of periodic outbreaks. In practice, this means that after the experienced decline in numbers, constant monitoring is vital. The expected new upturn in the mid-2020s signals an increased risk requiring preventive measures in these years, despite possible periods of relief (like the forecasted decline by 2029).

As suggested by the model, in the slightly humid, moderately warm agroclimatic zone, the pest population is expected to continue to recover after the long-lasting low level in the 2000s-2010s. The forecasted population dynamics for the slightly humid, moderately warm agroclimatic zone for 2024-2030 are shown in Figure 5. The forecast indicates a general upward trend in population: after a relatively moderate level in 2024 (~ 1.5 c.u., which is comparable to 2023), further growth is possible in subsequent years. Maximum values are predicted by 2027 and the end of the decade (2030), when the average number may exceed 2.0 c.u. (twice the 2018-2020 level). In some years, small fluctuations are likely: for example, the model suggests a possible temporary decrease in the growth rate around 2028.

Al-Dosari, M., Al-Marri, F., & Al-Sulaiti, A. (2024). A hybrid fuzzy-ML clinical decision support system for breast cancer risk stratification: Design, symbolic/statistical risk fusion, and proof-of-concept validation. Asian Journal of Current Research in Clinical Cancer, 4(2), 106–122. doi:10.51847/Q2DsJjMLMM

Anand, L. (2024). From code to crops: Harnessing bioinformatics and artificial intelligence (AI) in agricultural omics (Doctoral dissertation). University of Kentucky, Lexington, KY, United States.

Aoyama, H., Nakamura, K., & Zhang, M. (2025). Prioritizing preventive oral health: School-based brushing and silver diamine fluoride programs for Cambodian children. Journal of Current Research in Oral Surgery, 5, 161–167. doi:10.51847/egkGNEmAL8

Azhbenov, V. K. (2013). Rukovodstvo dlia vypolneniia monitoringovykh rabot po italianskoi saranche s ispolzovaniem GPS-tekhnologii. Astana.

Azhbenov, V. K., Baibussenov, K. S., Sarbaev, A. T., & Harizanova, V. B. (2015). Preventive approach of phytosanitary control of locust pests in Kazakhstan and adjacent areas. In Proceedings of the International Conference on Agricultural, Ecological and Medical Sciences (AEMS-2015) (pp. 33–37). International Institute of Chemical, Biological & Environmental Engineering.  doi:10.15242/IICBE.C0215122

Azhbenov, V., Shamuratov, D., Niyazbekov, Z., Dinassilov, A., Arystangulov, S., Baibussenov, K., Bashkarayev, N., & Zhumagaliyev, A. (2024). Localized biopesticide application using UAVs for the control of migratory locusts in Kazakhstan. Organic Farming, 10(3), 175–184. doi:10.56578/of100302

Baibusenov, K. S. (2021). Fitosanitarnyi monitoring i prognoz razvitiia i rasprostraneniia nestadnykh saranchovykh vreditelei v Severnom Kazakhstane. Nur-Sultan: S. Seifullin Kazakh Agrotechnical Research University.

Baibusenov, K. S., Azhbenov, V. K., & Bekbaeva, A. M. (2022). Prognostication of distribution of harmful non-gregarious locusts in Northern Kazakhstan using species distribution models. Electronic Scientific Journal “Dnevnik Nauki”, 9(69), 9. doi:10.51691/2541-8327_2022_9_9

Baibussenov, K., Bekbaeva, A., & Azhbenov, V. (2022a). GIS-based potential distribution modeling for harmful non-gregarious locusts in Northern Kazakhstan. In Proceedings of the 24th Paris International Conference on Advances on Chemical, Agricultural, Biological & Environmental Sciences (CABES-22) (pp. 71–78). Eminent Association of Researchers in Engineering & Technology. doi:10.17758/EARES11.EAP0422106

Baibussenov, K., Bekbaeva, A., & Azhbenov, V. (2022b). Simulation of favorable habitats for non-gregarious locust pests in North Kazakhstan based on satellite data. Journal of Ecological Engineering, 23(7), 299–311.  doi:10.12911/22998993/150043

Baibussenov, K., Bekbaeva, A., Azhbenov, V., Sarbaev, A., & Yatsyuk, S. (2021). Investigation of factors influencing reproduction of non-gregarious locust pests in Northern Kazakhstan. OnLine Journal of Biological Sciences, 21(1), 144–153. doi:10.3844/ojbsci.2021.144.153

Braun, D., Eckert, J., Lang, T., Schneider, L., & Wirth, F. (2025). High-resolution HLA-B allele and haplotype frequencies in the Qatari population: Implications for pre-emptive pharmacogenetic screening of drug hypersensitivity. Special Journal of Pharmacognosy, Phytochemistry and Biotechnology, 5, 77–87. doi:10.51847/43hsE3qngg

Castellano-Rioja, C. (2025). Investigating the effect of providing required training to mothers of children with surgery and its effect on mothers’ anxiety. Journal of Integrative Nursing and Palliative Care, 6, 7–11. doi:10.51847/m0J08PS92O

Childebaev, M. K., Pokrovskii, O. M., Zhalmukhamedova, Z. D., & Ivanov, A. I. (2004a). Prognosis of mass reproduction outbreaks of locust using climate forecasting methods. Tethys Entomological Research, 10, 197–208.

Childebaev, M. K., Sagitov, A. O., Akmollaeva, A. S., & Khasenov, S. S. (2004b). Materialy po faune i ekologii saranchovykh Severnogo Kazakhstana. Vestnik selskokhoziaistvennoi nauki, 4, 46–52.

Csep, A. N., Voiţă-Mekereş, F., Tudoran, C., & Manole, F. (2024). Understanding and managing polypharmacy in the aging population. Annals of Pharmaceutical Practice and Pharmacotherapy, 4, 17–23. doi:10.51847/VdKr0egSln

Da Hye, K., Petrenko, O. V., & Desta, H. T. (2024). Public awareness and misbeliefs about oral cancer in Northeast Italy: Influence of gender, education, and sources of information. Journal of Current Research in Oral Surgery, 4, 119–127. doi:10.51847/CwDFViXMTu

Damaševičius, R., Maskeliūnas, R., & Blažauskas, T. (2025). Enhancing virtual medical history taking: Effects of customized guidelines in two serious games for medical education. Annals of Pharmaceutical Education, Safety, Public Health Advocacy Journal, 5, 39–49. doi:10.51847/kNshKQSf5t

de Oliveira Aparecido, L., de Souza Rolim, G., da Silva Cabral De Moraes, J. R., Costa, C. T. S., & de Souza, P. S. (2020). Machine learning algorithms for forecasting the incidence of Coffea arabica pests and diseases. International Journal of Biometeorology, 64, 671–688. doi:10.1007/s00484-019-01856-1

Dubrovin, V. V., Teniaeva, O. L., & Kritskaia, V. P. (2011). Metody fitosanitarnogo monitoringa v zashchite rastenii ot vrednykh organizmov. Saratov State Agrarian University.

Durgabai, R. P. L., Bhargavi, P., & Jyothi, S. (2018). Pest management using machine learning algorithms: A review. International Journal of Computer Science Engineering and Information Technology Research, 8(1), 13–22.

Flores, S., & Ramos, M. (2025). Elucidating the core functional component network and revealing the molecular mechanisms of Longdan Xiegan decoction in uveitis therapy. Pharmaceutical Sciences and Drug Design, 5, 64–84. doi:10.51847/89ndNVfK1D

Frolov, A. N. (2019). Patterns of pest population dynamics and phytosanitary forecast. Plant Protection News, 3(101), 4–33.  doi:10.31993/2308-6459-2019-3(101)-4-33

Githae, E. W., & Kuria, E. K. (2021). Biological control of desert locust (Schistocerca gregaria Forskål). CAB Reviews, 16(013), 1–8. doi:10.1079/PAVSNNR202116013

Gómez, D., Salvador, P., Sanz, J., Casanova, C., Taratiel, D., & Casanova, J. L. (2018). Machine learning approach to locate desert locust breeding areas. Journal of Applied Remote Sensing, 12(3), 036011. doi:10.1117/1.JRS.12.036011

Gurung, N., & Rai, P. (2025). Balancing public health and personal rights: An ethical framework for mandatory outpatient psychiatric treatment. Asian Journal of Ethics in Health and Medicine, 5, 112–120. doi:10.51847/4Ojq3HDF85

Homayoun, S., Salehi, M., ArminKia, A., & Novakovic, V. (2024). The mediating role of individual motivation in the link between digital leadership and organizational agility. Annals of Organizational Culture, Leadership and External Engagement Journal, 5, 1–9. doi:10.51847/WjVc7UJ6Fa

Humphreys, J. M., Srygley, R. B., & Branson, D. H. (2022). Geographic variation in migratory grasshopper recruitment under climate change. Geographies, 2(1), 12–30.  doi:10.3390/geographies2010003

Ibrahim, E. A., Salifu, D., Mwalili, S. M., Dubois, T., Collins, R., & Tonnang, E. H. Z. (2022). An expert system for insect pest population dynamics prediction. Computers and Electronics in Agriculture, 198, 107124.  doi:10.1016/j.compag.2022.107124

Iranipour, Sh., Aalipour, M., Kazemi, M., & Ganbalani, G. N. (2017). Spatial distribution of Italian locust in Iran. Iranian Journal of Plant Protection Science, 48(1), 29–42. doi:10.22059/ijpps.2017.126665.1006643

Islam, F., Hasan, M. R., & Jahan, N. (2024). Healthcare professionals’ perspectives on ethical challenges in enrolling children with cancer in research: Insights from Sweden. Asian Journal of Ethics in Health and Medicine, 4, 181–194. doi:10.51847/5GTyKD7ZWF

Kambulin, V. E. (2000). Locusts in Kazakhstan. Zashchita i Karantin Rastenii, 7, 12–13.

Kambulin, V. E., Yskak, S., & Toleubaev, K. M. (2010). Dynamics of gregarious locust populations in Kazakhstan. Plant Protection and Quarantine, 4, 17–20.

Kapoor, P., Thakur, S., Usmani, H. M., Chauhan, R., Gupta, S., & Chauhan, N. (2025). Crop prediction system using machine learning. AIP Conference Proceedings, 3224, 020030. doi:10.1063/5.0253584

Khan, A., Patel, N., & Verma, R. (2024). Indirect 18F-labeling of estradiol via automated click chemistry: Radiochemical development and in vitro evaluation. Pharmaceutical Sciences and Drug Design, 4, 197–211. doi:10.51847/MiQjUEL35m

Klein, I., van der Woude, S., Schwarzenbacher, F., Muratova, N., Slagter, B., Malakhov, D., Oppelt, N., & Kuenzer, C. (2022). Predicting suitable breeding areas for locust species. International Journal of Applied Earth Observation and Geoinformation, 107, 102672. doi:10.1016/j.jag.2021.102672

Klein, R. S., Romano, L. F., & Bekele, T. M. (2025). Antiviral treatment duration and relapse risk in gastrointestinal cytomegalovirus disease: A retrospective cohort study. Bulletin of Pioneer Research in Medical and Clinical Sciences, 5(1), 182–188. doi:10.51847/OsqOAkAAaX

Lachininsky, A. V. (2020). Practical guidelines on three locust pests in Caucasus and Central Asia. Food and Agriculture Organization. doi:10.4060/cb0879ru

Lachininsky, A. V., Sergeev, M. G., Fedotova, A. A., Childebaev, M. K., Temreshev, I. I., Gapparov, F. A., & Kokanova, E. O. (2023). Moroccan locust Dociostaurus maroccanus: Morphology, distribution, ecology, population management. Food and Agriculture Organization. doi:10.4060/cc7159ru

Latif, M. S., Kazmi, R., Khan, N., Majeed, R., Ikram, S., & Ali-Shahid, M. M. (2022). Pest prediction in rice using IoT and feed-forward neural network. KSII Transactions on Internet and Information Systems, 16(1), 133–152.  doi:10.3837/tiis.2022.01.008

Lawton, D., Scarth, P., Deveson, E., Piou, C., Spessa, A., Waters, C., & Cease, A. J. (2022). Seeing the locust in the swarm: Accounting for spatiotemporal hierarchy improves ecological models. Ecography, 2022(2), e05763. doi:10.1111/ecog.05763

Liao, C., Lv, Y., & Zhang, X. (2013). Locust plague risk assessment in Xinjiang/China using GIS. In Proceedings of the Fifth International Conference on Geo-information Technologies for Natural Disaster Management (pp. 137–142). IEEE.

Lima, J. E., Pacheco, M., & Teixeira, B. (2024). Transition from traditional text to e-text: Medical, dental, and allied health students’ perceptions of e-learning. Interdisciplinary Research in Medical Sciences Special Journal, 4(2), 112–121. doi:10.51847/iMZc7n4TPc

Lombardi, C., Contini, A., & Russo, F. (2024). Undergraduate students’ knowledge, attitudes, and behaviors regarding COVID-19 during emergency remote learning. International Journal of Social Psychology Aspects in Healthcare, 4, 189–197. doi:10.51847/7vTs4Dkjvt

Lopez-Ramos, M., FigueroaValverde, L., Rosas-Nexticapa, M., AlvarezRamirez, M., Mateu-Armand, V., & CauichCarrillo, R. (2024). Interaction of twenty-seven bicyclo derivatives with VEGF receptors as a cancer treatment alternative. Archives of International Journal of Cancer Allied Sciences, 4(2), 18–28. doi:10.51847/m9NoOahmoL

Maeno, K. O., Piou, C., Kearney, M. R., Ould Ely, S., Ould Mohamed, S., Jaavar, M. E. H., & Ould Babah Ebbe, M. A. (2021). A general model of the thermal constraints on the world’s most destructive locust, Schistocerca gregaria. Ecological Applications, 31(4), e02310. doi:10.1002/eap.2310

Møller, A. K., Thomsen, S. L., Svensson, L. H., & Lindberg, E. J. (2025). Comparative assessment of knowledge, attitude, perceived effectiveness, and practice of complementary and alternative medicine among medical and non-medical students in Bangladesh. Journal of Medical Sciences Interdisciplinary Research, 5(1), 93–107. doi:10.51847/5S5IuL2Ln9

Nakamura, H., & Suzuki, K. (2025). Exploring the mediating roles of proactive attitude and employee engagement in the relationship between perceived organizational support, workplace spirituality, and public servant performance in Indonesia. Annals of Organizational Culture, Leadership and External Engagement Journal, 6, 105–118. doi:10.51847/bfcjgYFT14

Novak, S. E., & Svoboda, P. J. (2024). Clinicopathological features, treatment patterns, and survival outcomes in male breast cancer: A multicenter retrospective analysis from the Czech Republic (2007–2017). Archives of International Journal of Cancer Allied Sciences, 4(1), 93–111. doi:10.51847/gDpFxrLDnO

Pacheco, A., Cardoso, J., Faria, M., & Tavares, R. (2025). Population-specific variants in the human kinome: Insights from the IndiGen cohort reveal distinct sequence, structural, and pharmacogenomic implications for drug response in India. Special Journal of Pharmacognosy, Phytochemistry and Biotechnology, 5, 104–125. doi:10.51847/UtxcClcWDU

Pakhomova, J., Tesalovsky, A., Abdullayev, I., Dudynov, S., Maslennikova, N., Shelygov, A., Khatsenko, A., & Shichiyakh, R. (2025). GIS technologies for cadastral monetary valuation of urban land. SABRAO Journal of Breeding and Genetics, 57(4), 1668–1678.  doi:10.54910/sabrao2025.57.4.33

Papadopoulos, G. A., & Markou, E. K. (2025). Is resilience a mediating factor in the relationship between adverse childhood experiences and behavioral and emotional problems? International Journal of Social Psychology Aspects in Healthcare, 5, 168–175. doi:10.51847/b4buvoy4sD

Park, J., Kim, S., & Choi, M. (2025). Stage-dependent atrial fibrillation risk in gastric cancer: A nationwide SEER-based cohort study in South Korea. Asian Journal of Current Research in Clinical Cancer, 5(2), 97–105. doi:10.51847/btmP2WQIL5

Pravdin, F. N. (1978). Ecological geography of insects in Central Asia. Moscow: Nauka.

Sagitov, A. O., Duisembekov, B. A., et al. (2016). Phytosanitary monitoring of harmful and especially dangerous harmful organisms (pests, diseases, weeds) (3rd ed.). Almaty: Kazakh Research Institute of Plant Protection and Quarantine.

Saif, A., Omer, H., & Khan, M. (2024). Clinical staging, management strategies, and outcomes of colorectal cancer among patients in the West Bank: A retrospective evaluation. Bulletin of Pioneer Research in Medical and Clinical Sciences, 4(2), 106–114. doi:10.51847/RX6HDRzhyq

Sergeev, M. G., Childebaev, M. K., Vankova, I. A., Gapparov, F. A., Kambulin, V. E., Kokanova, E. O., Lachininsky, A. V., Pshenitsyna, L. B., Temreshev, I. I., Cherniakhovskii, M. E., et al. (2022). Italian locust Calliptamus italicus (Linnaeus, 1758): Morphology, ecology, distribution, population management. Rome: Food and Agriculture Organization. doi:10.4060/cb7921ru

Shafique, R., Khan, S. H., Ryu, J., & Lee, S. W. (2025). Weather-driven predictive models for jassid and thrips infestation in cotton crop. Sustainability, 17(7), 2803. doi:10.3390/su17072803

Shaik, M. A., Parveen, M., & Qureshi, I. (2024). Leveraging machine learning and drone technology for effective insect pest management in agriculture. In A. D. Dhaygude, S. K. Swarnkar, P. Chugh, & Y. K. Rathore (Eds.), Smart agriculture (pp. 147–162). CRC Press. doi:10.1109/APCIT62007.2024.10673597

Skawsang, S., Nagai, M., Tripathi, N. K., & Soni, P. (2019). Predicting rice pest population occurrence with satellite-derived crop phenology, ground meteorological observation, and machine learning: A case study for the Central Plain of Thailand. Applied Sciences, 9(22), 4846. doi:10.3390/app9224846

Sørensen, R. E., & Holst, M. C. (2025). Economic evaluation of enhancing nurse educational level on in-hospital mortality: Cost per prevented death analysis. Journal of Medical Sciences Interdisciplinary Research, 5(1), 108–120. doi:10.51847/OMW5yrzSDC

Suleimenov, S. I., Abdrakhmanov, M. A., Suleimenova, Z. Sh., Kambulin, V. E., et al. (2009). Methodological recommendations on the accounting and identification of harmful and especially dangerous harmful organisms on agricultural land. Astana.

Sun, R., Huang, W., Dong, Y., Zhao, L., Zhang, B., Ma, H., Geng, Y., Ruan, C., Xing, N., Chen, X., et al. (2022). Dynamic forecast of Desert Locust presence using machine learning with a multivariate time lag sliding window technique. Remote Sensing, 14(3), 747. doi:10.3390/rs14030747

Svendsen, M., Kristensen, O., & Nørgaard, T. (2025). Perceptions of e-learning among Jordanian medical students: A cross-sectional analysis. Interdisciplinary Research in Medical Sciences Special Journal, 5(1), 134–142. doi:10.51847/b9uiL7foKv

Swarnkar, S. K., Rathore, Y. K., & Swarnkar, V. K. (2024). Machine learning models for early detection of pest infestation in crops: A comparative study. In A. D. Dhaygude, S. K. Swarnkar, P. Chugh, & Y. K. Rathore (Eds.), Smart agriculture (pp. 147–162). CRC Press.

Verma, A. K., Malhotra, P. S., & Gupta, R. N. (2024). Evolution of curriculum committees in US pharmacy education: A national survey update from 2011. Annals of Pharmaceutical Education, Safety, Public Health Advocacy Journal, 4, 85–94. doi:10.51847/5xd5wIPmVt

Wilson, E., Brown, N., & Taylor, O. (2025). A population-based approach to evaluating pharmaceutical quality using quality by design. Annals of Pharmaceutical Practice and Pharmacotherapy, 5, 174–185. doi:10.51847/yewG9DO3yi

Wu, R., Guan, J.-Y., Wu, J.-G., Ju, X.-F., An, Q.-H., & Zheng, J.-H. (2022). Predictions based on different climate change scenarios: The habitat of typical locust species is shrinking in Kazakhstan and Xinjiang, China. Insects, 13(10), 942. doi:10.3390/insects13100942

Xiao, Q., Li, W., Kai, Y., Chen, P., Zhang, J., & Wang, B. (2019). Occurrence prediction of pests and diseases in cotton on the basis of weather factors by long short-term memory network. BMC Bioinformatics, 20, 688. doi:10.1186/s12859-019-3262-y

Yessenbayeva, G., Ilyas, A., Kenenbayev, S., Zhanbyrbayev, Y., Kulanbay, K., Dutbayev, Y., & Kultassov, B. (2025). Impact of agritechnologies on the biomass and nutritional parameters of crops: A study of corn, soybean, and pea in Southeast Kazakhstan. OnLine Journal of Biological Sciences, 25(4), 1063–1069. doi:10.3844/ojbsci.2025.1063.1069

Yskak, S., Komissarova, I. A., & Niiazbekov, Zh. B. (2002). On improving the system of crop protection against harmful locusts. In Strategiia nauchnogo obespecheniia: Realnost i perspektivy (pp. 189–190). International Scientific Conference, Almaty.

Zielinska, A., & Kowal, M. (2024). Survival outcomes after cardiac arrest in community-dwelling adults receiving home care versus nursing home residents compared with unsupported individuals. Journal of Integrative Nursing and Palliative Care, 5, 207–218. doi:10.51847/sd6YFareZk