Introduction

Longstanding experience of worldwide anti-locust campaigns during the XX century has shown the futility of wide use of chemical treatments during the peak of the outbreak. Unprecedented measures of chemical control in Kazakhstan during the outbreak of 1997-2003 serve another proof of this fact, because at the beginning of the emergence of insects clouds and swarms phytosanitary service usually is not ready either materially or morally to conduct anti-locust control treatments. As a rule, such measures take place after two or three years following the outbreak, when the locusts have occupied a vast territory. Insecticides usually provide only a temporary reduction in the insects abundance and severity just in the centers of their application, though in general, practically cannot affect dramatically the population dynamics. In contrast, chemical treatments destabilize the ecological situation due to destruction of natural enemies and natural epizootics that extend the periods of mass reproduction for several years (Kurishbaev & Azhbenov, 2013; Yskak, et al., 2012).

The existing approaches in solving the problem of the locust invasion include massive chemical treatments over the large areas in the midst of outbreaks and large-scale migrations of harmful depredators, or compromise actions in response to the risks of problematic foci, i.e. when swarms visitation has already taken place. Such methods cannot be satisfactory. Important strategic disadvantage of massive chemical treatments was their conduct in the “fire fighting” mode, because the initial stages of locusts accumulation in primary foci, especially in remote or inaccessible areas, remain unnoticed (Azhbenov, Baibussenov et al., 2015).

Implementation of the fitosnitarnoy security system at the present stage of country’s agriculture development is inconceivable without the use of forecasts. This is due to the fact that the contemporary chemical control based on the application of the current range of insecticides against locust pests is used in most cases untimely. Such an approach in solving the problem in Kazakhstan and adjucent areas at the turn of the millennium has caused many environmental, economic and social problems.

Kazakhstan areal is inhabited by more than 270 species and subspecies of locust insects. Among them just 15-20 species cause periodically heavy damage to farmland (Nurmuratov, 2000). Fauna of locust pests is presented mainly by Calliptamus italicus L. – one of the most harmful species, Dociostaurus maroccanus Thunb., and Locusta migratoria migratoria (Azhbenov, 2010).

In the territory of Kazakhstan, along with the gregarious locust species, non-gregarious locust species also play the important role. The most common types include: Dociostaurus kraussi Ingen, Dociostaurus brevicollis Ev., Aeropus sibiricus L., Arcyptera microptera F.d.W., Chorthippus albomarginatus Deg. and Stauroderus scalaris F.W., Stenobothrus fischeri Ev. (Akmollaeva, 2005; Baibussenov et al., 2013). Among these spices, the most frequent types are Stenobothrus fischeri Ev., Chorthippus albomarginatus Deg., Aeropus sibiricus L., Dociostaurus brevicollis Ev. They are found at all of the above habitats, waste lands, pastures and hayfields. Other species are spread to a lesser extent (Baibussenov et al., 2013). Whereas gregarious locust types are characterized by distant migrations and invasion from one territory to another, the non-gregarious species are permanent inhabitants of the steppe and cultural habitats. Depending on the cyclicality and weather conditions they are also characterized by massive abundance outbreaks (Uvarov, 1966).

Methodology

Long-term forecasts are developed for a year or a season. Such a forecast is primarily needed for preventative protection of plants, as well as for supporting protective measures. Distribution, development and harmfulness of each species are influenced by many factors that affect reproduction, survival, and behavior of its population. The simulation of the population dynamics of harmful spices highlights important environmental factors, whose quantitative aspect may serve to judge the likely change in the spread and development of populations in country’s specific areas (Dubrovin, et al., 2011; Iliukhin, & Ryabinin, 2010). The used research methods are common in agricultural entomology, insect ecology and plant protection (Suleimenov et al., 2009; Sulejmenov, Yusupov, et al., 2009; Sagitov, et al., 2001). To produce long-term forecasts, longstanding information about the pest population dynamics is needed. Data on the longstanding abundance dynamics was determined through analysis of the longstanding information on the distribution and colonization of harmful non-gregarious locusts on farmlands. Since we analyzed the data taken for over 16 years (Baibussenov et al., 2014), the longstanding official data of the State Agency “Republican Methodical Center of Phytosanitary Diagnosis and Prognosis of the Ministry of Agriculture of the Republic of Kazakhstan” and Republican State Enterprise “Kazakh Research Institute of Plant Protection and Quarantine” served the source of information. To perform this kind of research, in plant protection methodology, namely phytosanitary monitoring and prognosis, there is a special technique to process and analyze longstanding information from plant protection services (Dubrovin, et al., 2011; Polyakov et al., 1984).

Results and Discussion

• Long-term prognosis predictors of population dynamics phase of harmful non-gregarious locusts.

Currently, the issues on the prediction of non-gregarious locusts in some parts of Kazakhstan are studied insufficiently. Today, the northern regions of the country are important for growing grain agricultural products. In turn, the dominant species of harmful non-gregarious locust at their mass reproduction can cause considerable damage to agricultural land in the region. In this regard there is a need for making reasonable long-term forecasts on the relative phase changes of the population dynamics of concerned phytophaga.

The practical value of this work was that the forecasting objects were harmful non-gregarious locusts, which define the activities of above mentioned organizations. Though, to select the more economically important species, we conducted a study of the number of harmful non-gregarious locusts according to their dominance in the steppe cenosis.

Thus, in the research area we conducted surway on the species composition and their quantitative indicators. That is, we determined density of the insects per 1 m² and species dominance indices. In accordance with the methodology of Dubrovin et al., 2011, the dominant forms were considered the species exceeding 16% of all collected species, while subdominant – those varying from 4 to 16%. According to the results presented in Figure 1, the following species were classified as dominant: Dociostaurus brevicollis Ev., Arcyptera microptera (FdW), Podisma pedestris (L.), while subdominant species included Stenobothrus fischeri Ev., Aeropus sibiricus (L.)., Dociostaurus kraussi (Ingen.), Oedaleus decorus (Germ.) and Chorthippus albomarginatus Deg.

Figure 1: Dominance index of various kinds of harmful non-gregarious locusts in Northern Kazakhstan; an average of 2013-2014 Click here to view full figure

In the pests abundance dynamics we revealed regular phase changes in the populations’ qualitative status, predetermining intrapopulational, intraspecific and interspecific relationships. Variability of population dynamics phases is determined primarily by influences of forage and weather conditions on their formation and station distribution (Azhbenov VK, 2001; Azhbenov VK, 2006).

Full cycle of pests reproduction dynamics can be divided into five phases depending on population variability. Most often, said cycle of harmful species population dynamics is incomplete. Thus, the depression phase may occur immediately after the expansion phase, if existence conditions are sharply deteriorating (Polyakov et al., 1984). To produce long-term forecasts, it is extremely important to hold information based on system analysis on the change of population dynamics over a few years.

We have carried out a systematic analysis of the population dynamics of non-gregarious locusts in Northern Kazakhstan for the period from 1998 to 2013. The analysis allowed building and developing a system of numerical indicators that are important for making forecasts. The greatest prognostic value have year-to-year variations of indicators, such as relative and absolute colonization (C_R, C_A), net reproduction (NR), the expansion and reproduction energy (E_EX, E_RP), and progradation rate (PR). The diagnostic features of gradation phases of non-gregarious locusts have been defined depending on the analyzed years (Baibussenov et al., 2014).

Based on the results of these analyzes, pivotal schemes were built (Tables 1, 2 and 3) of diagnostic predictors of non-gregarious locusts populations status and their quantitative aspects at different phases of their dynamics to produce long-term forecasts.

Table 1: Diagnostic predictors of phase status of non-gregarious locusts populations at different dynamic phases (in Northern Kazakhstan)

Table 2: Predictive model of quantitative aspect of non-gregarious locusts populations at different dynamics phases (in Northern Kazakhstan)

It should be noted that gradation data were built based on the system analysis over these 16 years. Thus, the population phase status of harmful non-gregarious locusts depends on the change of several numerical indicators (relative colonization, absolute colonization, etc.). Due to the essential difference between them, the additional quantitative characteristics were introduced (imago density, density of egg-pots in the soil, etc.) to build a long-term forecasting of the expected abundance of phytophaga.

The guidelines for other pests were taken as a basis when building current forecast (Azhbenov, 2014).

The results of analyses of pest population indicators, listed in Table 1, were approximated, because constructing a gradation system of predictors requires systemic sequence of indicators and figures. Table 2 represents a forecast model of quantitative aspect of non-gregarious locusts populations at different dynamic phases. Also, these figures are needed as references to make early forecasts when determining the pest population dynamics phase.

The regression equation coefficients when predicting population dynamics phase of harmful non-gregarious locusts.

The population dynamics of pests is even more important for subsequent prediction of changes in their future abundance. Table 3 shows the relationship of annual variability of absolute colonization (C_A) of harmful non-gregarious locusts on natural farmland depending on the external environmental factors, such as the weighted average annual hydrothermal index and the treated area in thousand hectares

Table 3: Annual variation of the absolute colonization (C_A) of harmful non-gregarious locusts on natural farmland depending on the external environmental factors.

Thus, it is possible to construct a multiple linear regression using given correlation between the annual variability of absolute colonization, C_A (variable Y), the weighted average of the annual hydrothermal index (variable X₁) and the area treated by insecticides and measured in thousand hectares (variable X₂). Here it should be noted that along with these figures, Table 3 presents data on surveyed and colonized areas of natural farmland by harmful non-gregarious locusts. The data on treated area (anthropogenic pressure) come from these additional indicators and have a direct relationship with each other. Also, it should be noted that these initial data are taken on average throughout Northern Kazakhstan, as calculating the regression coefficients requires average data. Besides, such calculation can be made with regard to one specific area or region. But due to the fact that here we took into account the peculiarities of all four areas of Northern Kazakhstan, the calculated coefficients can be used for any of these areas when making forecast. Longstanding information on chemical treatments against phytophaga was taken from anthropogenic pressure since data metering was conducted only on natural farmland, where no interventions of human factors into the ecology of insects have been reported except of chemical treatments. Of course, this kind of anthropogenic pressure is relatively volatile, though it can be used for the calculation of the regression coefficients as one can determine the average long-term variability of this factor when building forecast.

The Excel application for PC offers a special “regression” tool, which makes it possible to determine the regression equation coefficients based on the matrix construction and assuming that the relationship is linear. At that, it is possible to make long-term forecast, as well as to verify the accuracy and the prediction error (Dubrovin, et al., 2011). As a result of conducted regression analysis we obtained the following regression equation:

                                                                     Y = 4.34 – 2.18x₁ + 0.005x₂ (1)

From this regression equation we can conclude that when changing the value of X₁ by one unit, value of Y decreases by 2.18 units, whereas the value of X₂ increases by 0.005 units. Thus, we can predict the future changes in the annual variability of absolute colonization C_A (variable Y) depending on the weighted average of annual hydrothermal index (variable X₁) and the treated area, expressed in thousand hectares (Variable X₂).

Finely, we substitute the data taken for current year into the equation variables:

Y = 4.34 – 2.18 х 1.1 + 0.005 х 564.2;

Y = 4.34 – 2.398 + 2.821;

Y = 4.6

Carrying out calculations by the equation we obtain that in 2015 expected absolute colonization (C_A) of harmful non-gregarious locusts on natural farmland would be approximately 4.6 species per unit of area surveyed. Also, we can make a forecast for several years ahead. In this case, we construct a system of equation chains. Comparing these data with gradation scheme given in Table 1, we can expect either a rise in the abundance or mass reproduction of pests. Although, the refinement of the forecast may show that the population dynamics phase may be different depending on the forecast reliability and accuracy.

When refining the forecast for the coming year, it is possible to determine forecast fidelity (W%) and error (E%).

Suppose calculated forecast gave Y_forcast = 4.6 spec./surveyed area, whereas in the following year the actual value was Y_actual = 3.9 spec./surveyed area.

Y_actual. х 100% 3.9 х 100%

W = ——————— = —————— = 84.7%

 Y_forcast. 4.6

E = 100% – Y_forcast. =100% – 84.7% = 15.3%

 Thus, the forecast reliability (W%) is 84.7%, and the forecast error (E%) is 15.3%.

 Argumentation and planning of protective measures against harmful non-gregarious locusts.

The use of plant protection agents can be recommended only in situations where there is a pest risk and the threat of economically significant losses of crops. At that, there should be no risk of environmental contamination by pesticides. In this regard, the important point is to develop the economic threshold and the feasibility of the use of pesticides, i.e. phytosanitary standards.

The sequence of actions towards the modeling of crop losses and pest risk, as well as development of criteria for the necessity of protective measures to ensure the prevention of post-harvest losses and environmental pollution by insecticides, is provided as follows:

– evaluating and forecasting crop losses from pest species;

– establishing economic threshold of harmfulness;

– defining the criteria for the feasibility of the use of pesticides.

The concept of economic threshold of harmfulness (ETH) and related critical density and the economic damage is important both at the national and international levels. This was noted in publication by Tang, 2010 and Azhbenov, 2007.

When using the economic threshold of harmfulness in plant protection we should highlight two approaches. The first approach involves the simplification of ETH concept and provides its definition depending on the level of permissible losses of 3% or 5%. However, this approach has essential disadvantages: strictly established thresholds when deviating from the average situation.

Noted shortcomings can be avoided if using ETH within the second approach (Azhbenov, 2014). The basis of the general ETH model consists of prediction of crop losses and determination of the saved products, which recover the costs for plant protection. The ETH models are described by equations (2):

ETH = C /P * Y * h (2)

where ETH – is the economic threshold of harmfulness, spec./m², C – is the cost for plant protection, $/ha; P – is the price of products $/ha; Y – is the crop yields  and h – is the coefficient of harmfulness or loss of plant productivity (%) by unit of specimen per m².

Rational planning of protective treatment volumes is an important element in the organization of preventive plant protection. There are two types of planning protective measures: the current planning, intended for a year or a season, and long-term planning, specifying the need for plant protection for the five-year period and more. Both types of planning are based on the respective forecasts of dissemination and evolution of harmful organisms including annual, seasonal and longstanding forecasts (Suleimenov et al., 2009; Azhbenov, 2013).

Planning is based on the data obtained during summer surveys of adult locusts, as well as autumn surveys of their egg-pots. Similarly to the logical model of quantitative aspects of non-gregarious locust populations at different dynamics phases, we can represented the tabular data describing quantitative aspects of non-gregarious locusts depending on the areas of treatment against the insects in Northern Kazakhstan.

Planning of protective measures against locusts is carried out based on certain norms (Tables 4 and 5). If it is planned to apply several treatment methods against locusts, the volumes should be calculated for each method taking into account their percentage ratio and determining the area to be treated by adding the volumes of each method.

Table 4: Agricultural land area to be treated against harmful non-gregarious locusts depending on the population dynamics phase (in Northern Kazakhstan)

Different methods of treatments are used against locusts. They include entire methods, where the whole territory populated by the pest is treated; barrier methods, where treatment is applied just along certain strips of land towards direction of locusts movement leaving the untreated space between the strips; local methods, where only foci of larvae clusters (swarms) are treated; and marginal methods, where treatment is carried out around agricultural crops (Azhbenov, 2013).

Table 5:Planning of area to be treated against locusts using barrier and local methods

Table 5 presents required norms when planning areas to be treated against locusts using entire, barrier and local methods:

The barrier treatment coefficient is calculated as follows: treated area = colonized area: (2),(3),(4);

The local treatment coefficient is calculated as follows: treated area = colonized area (x) (0.4),(0.3),(0.2).The proportions 1 : 1, 1 : 2 and 1 : 3 denote the ratio between the barrier width and the space between barriers.

The maximum possible width of the space between barriers can be 300 m at an average width of the barrier of 100 m.

Irrational use of pesticides in agriculture leads to their accumulation in the soil and food products. However, there is no doubt that the rise in farming culture, improving pesticide application technology, limiting their use, and strict dosage when applying to the soil can significantly reduce their negative impact. Treatment of crops with pesticides should be done during the recommended time period. Especially strictly it is necessary to comply with the terms of final treatments before harvest, specified in the “List of chemical and biological pesticides against pests, plant diseases and weedage”.

Conclusion

The rational and appropriate use of plant protection agents in agriculture economy is very important issue. Conducting preventive measures on pest monitoring and forecasting during their low abundance is vital to avoid large-scale treatments at pest mass reproduction. Chemical treatments with pesticides provide only a temporary decline in the number of locusts, though they cannot dramatically affect the overall biological cycle of reproduction. The results of conducted studies and analysis allowed us to predict the beginning of certain phases in harmful non-gregarious locusts population dynamics that increases the accuracy of the general forecast. With longstanding data analysis of external factors (integrated weather conditions and the area of chemical treatments), affecting the population dynamics of phytophaga, we calculated coefficients of the regression equation. This may serve one of the key factors in long-term forecasting of the population dynamics phase state of harmful non-gregarious locusts in Northern Kazakhstan regions. Also, the developed forecast predictors make it possible to substantiate the area of agricultural land to be treated at different phase states of studied phytophaga.

Based on the results obtained, in the long term perspective, topical issue would be to study the effect of pesticide load on the environment while irrational use of plant protection agents, as well as to reveal the specific impact of total anthropogenic pressure on the overall progress in the harmful non-gregarious locusts population dynamics.

References

1. Akmollaeva, A. (2005). Non-gregarious locusts of Northern Kazakhstan (fauna, ecology, and protective measures). PhD thesis. Almaty.
2. Azhbenov, V. (2000). Mass reproduction and migration of locusts in Kazakhstan. Steppe Newsletter, 6, 16-20.
3. Azhbenov, V. (2001). Mass reproduction of locusts in Kazakhstan and the farmland protecting problem. Science Bulletin of S. Seyfulin Akmola Agrarian University. Astana.
4. Azhbenov, V. (2006). Locusts in Kazakhstan. Plant Protection and Quarantine, 1, 8-11.
5. Azhbenov, V. (2007). Plants protection in the Republic of Kazakhstan: contemporary status and perspectives. Science Bulletin of S. Seifullin Kazakh Agro Technical University, 1(1), 5-14.
6. Azhbenov, V. (2013). Guide to perform monitoring of the Italian locust using GPS-based technology. Astana.
7. Azhbenov, V. (2014). Innovation of pest monitoring and forecasting of especially dangerous pests (Apamea anceps Schiff., Mayetiola destructor Say.) of cereal crops in Northern Kazakhstan on the basis of geo-information and GLONAS/GPS-based technologies to ensure grain production and food security. Research report. Astana.
8. Azhbenov, V., Baibussenov, K., Sarbaev, A., & Harizanova V. (2015). Preventive approach of phytosanitary control of locust pests in Kazakhstan and adjacent areas. Proceedings of IICBE International Conference on Agricultural, Ecological and Medical Sciences (AEMS-2015), Penang, Malaysia.
9. Baibussenov, K., Sarbaev, A., Azhbenov, V., & Harizanova, V. (2014). Environmental features of population dynamics of non-gregarious locusts in Northern Kazakhstan. Life Science Journal, 11(10), 277-281.
10. Baibussenov, K., Azhbenov, V., & Sarbaev A. (2013). Biological and ecological features and the expansion of harmful non-gregarious locusts in Northern Kazakhstan. Proceedings of the National Academy of Sciences of the Republic of Kazakhstan. Series of Agrarian Sciences, 6, 30.
11. Dubrovin, V., Tenyaeva, O., & Kritskaya, V. (2011). Methods of pest monitoring to protect plants against pests. Saratov.
12. Iliukhin G., & Riabinina, G. (2010). Agronomist handbook on plant protection and agroecology. Astana: Foliant.
13. Kurishbaev, A., & Azhbenov, V. (2013). A preventive approach in dealing with locust invasion in Kazakhstan and adjucent areas. Bulletin of the S. Seifullin Kazakh Agro Technical University, 1(76), 42-52.
14. Nurmuratov, T., Azhbenov, V., Kambulin, V., Childebaev, M., Komissarova, I., & Zhumagalieva, G. (2000). Locusts pests of agricultural plants in Kazakhstan and recommendations to limit their abundance. Almaty: Asia Publishing.
15. Polyakov, I., Persov, I., & Smirnov V. (1984). Projected growth of pests and crops diseases. Leningrad: Kolos.
16. Sagitov, A., Azhbenov, V., Evdokimov, C., Yskak, S., & Temreshev, I. (2001). Survey of territories, populated with harmful locusts. Proceedings of the International round table on problems of locust control in Central Asia, Almaty.
17. Suleimenov, S., Yusupova, G., & Suleimenova, Z. (2009). Complex system of measures to protect crops against non-gregarious locusts. Astana.
18. Suleimenov, S., Abdrakhmanov, M., Suleimenova, Z., Kambulin, V., et al. (2009). Guidelines for recordkeeping and identification of harmful and highly dangerous pests of agricultural land. Astana.
19. Tanskyi, V. (2010). Self-regulatory agrobiocenosis. St. Petersburg: VIZR.
20. Uvarov, B. (1966). Grasshoppers and locusts: a handbook of general acridology, Cambridge: Univ. Press.
21. Yskak, S., Agibaev, A., Taranov B., Kalmakbaev, T., & Kambulin, V. (2012). Distribution of gregarious locusts in Kazakhstan and proper protective measures. Proceedings of the National Academy of Sciences of the Republic of Kazakhstan. Series of Agrarian Sciences, 5, 11-20.